Triple tissue culture fusion

The present invention relates to cultures of fused cellular tissues .

Background

Bagley et al. (Nat Methods, 2017, 14 (7) : 743-751) and WO 2018/197544 Al describe fused dorsal-ventral cerebral organoids.

Ao et al. (Lab Chip, 2021, 21: 688) describe the fusion of human forebrain and human mid-brain organoids by acoustofluidics within a hexagonal acoustofluidic device that allows rotation and transportation of an organoid. Xiang et al. (Cell Stem Cell 24, 487-497, 2019) describe a fusion organoid of developing thalamus or cortex.

Anderson et al. (Cell 183 (7) (2020) : 1913) describes an assembly of 3D cultures derived from cortical, spinal and skeletal muscle spheroids. WO 2019/246436 Al describes a fusion of spinal cord spheroids with skeletal muscle cells. The fused spheroid is then subjected to a second fusion with a cortical spheroid.

Miura et al. (Nature Protocols, 2022, 17: 15) and WO 2021/087145 Al describe the generation of 3D spheroids resembling specific domains of the nervous system and their fusion to an assemble id.

Knoblich et al. (Protocol exchange (2017) DOI: 10.1038/pro- tex .2017.064 ) describes a fusion of dorsal and ventral forebrain-like tissues.

Choi et al. (Biomaterials 34 (1.2) (2013) : 2938-2946) discloses the generation of three-dimensional neurospheres.

Choi et al., (Biomaterials 31 (15) (2010) : 4296-4303) discloses aggregating murine ES cells.

Sun et al., IEEE (2016) : 339-342 describes ECM-coated microwells for characterizing different cell properties.

KR 2013 0013537 A (English abstract) describes a microwell array with an anti-adhesion coating.

There remains a need to provided tissue cultures that recapitulate improved in vivo characteristics, in particular at tissue interactions, of tissues in contact with different tissues.

Summary of the invention

The invention provides an in vitro method of producing a fused tissue culture with at least three different tissues comprising the steps of selecting a spatial shape for attaching the at least three di f ferent tissues to each other, placing the at least three di f ferent tissues into contact in said spatial shape , culturing the at least three di f ferent tissues under conditions that maintains tissue fusion of the at least three di fferent tissues ; and/or comprising the steps of attaching each of the at least three tissues to at least one other of the at least three tissues with applied force pushing tissues together, culturing the at least three di f ferent tissues under conditions that maintains tissue fusion of the at least three di f ferent tissues . The force may be applied over a period of at least 4 sec .

Further provided is a tissue culture obtainable by such a method . Provided is a tissue culture comprising a planar or linear fusion of at least three di f ferent neural or neuronal tissues .

Further provided is an in vitro method of generating a ventral hindbrain tissue culture , comprising inducing neural or neuronal di f ferentiation in pluripotent cells with a neural induction medium comprising at least one TGF-beta inhibitor and a Wnt activator, and allowing the cells during said induction to form a cell aggregate ; culturing the cell aggregate to form a tissue culture .

Further provided is a method of investigating neural formation or neural di f ferentiation, comprising introducing a neural stem cell , neural progenitor cell , neuronal cell , neural cell , glial cell , astrocyte or oligodendrocyte into a tissue culture of the invention and monitoring cell di f ferentiation and/or cell growth .

Further provided is a method of testing or screening a candidate compound for influencing neural cells of a fused tissue culture , comprising contacting cells or a tissue in a method of the invention with the candidate compound or contacting a tissue of the invention with the candidate compound and maintaining said contacted tissue in culture , and observing any change in the neural cell as compared to said tissue without contacting by said candidate compound .

Further provided is a carrier, in particular a mold, for tissue fusion comprising a top surface and a recess in said top surface , wherein said recess is elongated and comprises a deepest depression and at least one slope along the direction of the elongation, the at least one slope descending to the deepest depression, wherein the carrier or mold at least at the deepest depression comprises a low- or non-cell-adhesion surface .

Further provided is a kit comprising a carrier or mold of the invention and neural induction medium .

All embodiments of the invention are described together in the following detailed description and all preferred embodiments relate to all embodiments , aspects , methods , tissue cultures , organoids , materials , like carriers or molds and media, uses and kits alike . E . g . kits and materials or their components can be used in or be suitable for inventive methods . Any component used in the described methods can be part of the kit or describe the materials or their uses . Inventive tissue cultures or organoids are the results of inventive methods or can be used in inventive methods and uses . Preferred and detailed descriptions of the inventive methods read alike on suitability of resulting or used organoids or tissue cultures of the invention . All embodiments can be combined with each other, except where otherwise stated .

Detailed description

The present invention provides an in vi tro method of producing a fused tissue culture with at least three di f ferent tissues comprising selecting a spatial shape for attaching the at least three di f ferent tissues to each other, placing the at least three di f ferent tissues into contact in said spatial shape , culturing the at least three di f ferent tissues under conditions that maintain tissue fusion of the at least three di f ferent tissues . The tissues are cellular tissues . The cells should be alive and functional in order to study their behaviour in the fused tissue culture . The invention allows to select special shapes in which the di f ferent tissues can be fused together, in the shape during fusion . This shape will also determine functionality of the produced fused tissue culture as the connectivity between or orientation of the di f ferent tissue has an ef fect of possible functions of the fusion product , i . e . the fused tissue culture . The inventive spatial placement allows the coterminal fusion of the at least three di f ferent tissues . In previous methods mentioned in the background section only double fusions were possible at a given time . Such fusions were possible one at a time, and such fusions could be repeated to achieve multiple tissues, but no simultaneous fusions were described in a spatially defined manner.

For example, the inventive spatial placement allows contacts or connections between all three of at least three different tissues (or of more tissues) or more limited contacts, such as with at least one, preferably at least two, of the at least three different tissues being in contact with only one of the (other) at least three different tissues. Such a shape is e.g. a chain of different tissues that are fused to each other. A possible spatial shape is a planar or linear shape.

A linear shape of different tissues or tissue parts of different origin in the produced fused tissue culture is to be understood as a shape wherein a line can be drawn through the arrangement of the different tissues or fused tissue culture wherein the line passes through all or some of the at least three different tissues or tissue parts. At least 3, preferably at least 4, at least 5 at least 6 or more of the at least three different tissues or tissue parts may comprise one line that can be drawn through the arrangement of all the different tissues or tissue parts. In some embodiments all of the different tissues to be fused or tissue parts are found on a shape consisting of 1, 2, 3, 4, 5, 6, 7, 8 or more lines, which lines each can be drawn through the arrangement of the different tissues or tissue parts wherein each line passes through at least 2, preferably at least 3, of the at least three different tissues or tissue parts .

A planar shape of different tissues or tissue parts of different origin in the produced fused tissue culture is to be understood as a shape wherein a plane can be drawn through the arrangement of the different tissues or fused tissue culture wherein the plane passes through all or some of the at least three different tissues or tissue parts. At least 3, preferably at least 4, at least 5 at least 6 or more of the at least three different tissues or tissue parts may comprise one plane that can be drawn through the arrangement of all the different tissues or tissue parts. In some embodiments all of the different tissues to be fused or tissue parts are found on a shape consisting of 1, 2, 3, 4, 5, 6, 7, 8 or more planes, which planes each can be drawn through the arrangement of the different tissues or tissue parts wherein each plane passes through at least 3, preferably at least 4, of the at least three different tissues or tissue parts. The shape may be a surface that is bent as the shape on which the tissues rest on the carrier or recess (bottom surface) , or the spatial shape may follow the curvature of a carrier on which the at least three different tissues are placed .

The step of culturing the at least three different tissues under conditions that maintains tissue fusion of the at least three different tissues may comprise standard culturing conditions that allow fusion by growth or attachment through intercellular connections of the different tissues, while maintaining the spatial shape of the at least three different tissues. It may comprise steps to enhance fusion speed, e.g. by using a cellular adhesive, such as extracellular matrix (ECM) , such as Mat- rigel, or extracellular matrix components, such as collagen and/or laminin, and/or at least partially removing surface fluid, e.g. by simply removing the medium from the culture, which increases stickiness of the different tissues and contacting the tissues to each other (in the selected spatial shape) . With gentle handling the addition of a cellular adhesive, like ECM, is not needed. Cells may form their own ECM in time. The shape during attachment is easily maintained by letting the at least three different tissues rest. Preferably there is no medium stirring or agitation either mechanically or acoustically, that would move the at least three different tissues.

The at least three different tissues may be 3, 4, 5, 6, 7, 8 or more different tissues. The number of different tissues (n) minus 1 (n-1) is the minimum number of fusions that are required to connect all of the different tissues. So, for at least three different tissues, at least n-1 fusions between the different tissues happen. Preferably 2, 3, 4, 5, 6, 7 or more fusions between different tissues happen during the steps of "placing the at least three different tissues into contact in said spatial shape, culturing the at least three different tissues under conditions that maintains tissue fusion of the at least three different tissues". Preferably the different tissue fusions happen simultaneously .

The at least three different tissues may be tissue spheroids or organoids or biopsy tissues or 3D cell aggregates. The different tissues may have a size of 0.1 mm to 10 mm, such as 0.2 mm to 8 mm or 0.3 mm to 6 mm, in their longest dimension. "Size" refers to the longest dimension in 3D space ("length") . Preferably the tissues are globular in shape, in particular with the shortest dimension being not less than 20% of the longest dimension, in particular not less than 30% or not less than 40% of the longest dimension. Other possible shapes are squared or cylindrical, among others. Preferably the volume of the at least three different tissues is at least l><10 ⁶ pm ³ , in particular preferred at least 2*10 ⁶ pm ³ , at least 4*10 ⁶ pm ³ , at least 6*10 ⁶ pm ³ . The fused tissue culture may have the same sizes, shapes and volumes or even larger sizes and volumes, such as a volume of at least 8*10 ⁶ pm ³ , at least 10*10 ⁶ pm ³ , at least 15*10 ⁶ pm ³ and/or a length of at least 250 pm, especially preferred at least 350 pm.

The at least three different tissues are usually small aggregates of cells and may have a size of at most 12.5 mm, preferably of at most 10000 pm or at most 8000 pm, e.g. with volumes of at most 200 mm ³ , at most 150 mm ³ , or at most 125 mm ³ . In some embodiments, the fused tissue culture may be larger with a size of at most 100 mm, preferably of at most 70 mm or at most 50 mm, e.g. with volumes of at most 6000 mm ³ , at most 3000 mm ³ , or at most at most 1000 mm ³ .

In preferred embodiments, the at least three different tissues are placed onto a carrier, wherein the carrier has a slope gradually descending to a deepest depression, wherein the at least three different tissues sink to the bottom of the carrier and then are pushed together due to a downslope force caused by an inclination of the slope. A carrier with a depression can be used to push the at least three different tissues together by gravity and thereby maintain the spatial shape. The carrier may have a recess or groove that forms the spatial shape in a depression. The recess may be linear, bent, curved or planar, with even more complex forms depending on the number of different tissues to be fused, such as multi-linear of 2, 3 or more lines that intersect, such as star-shaped.

The slope may be lower inclined closer to the deepest depression than further away from the deepest depression of the carrier. At the deepest depression, the inclination may be zero or a horizontal at the bottom of the depression. There may be one or more deepest depressions in the carrier for a given location for the formation of a fused tissue culture. The slope may gradually descend to a deepest depression. Preferably the gradually descending slope is a curved slope. The slope may also be or comprise a linear or planar slope.

The inventive method may comprise placing the at least three different tissues in a recess of a carrier, wherein said recess is elongated and comprises a deepest depression and at least one slope, the at least one slope descending to the deepest depression. Placement of the at least three different tissues into the recess may force contact of each the at least three different tissues with at least one other of the at least three different tissues through gravity in the recess. Through gravity and/or through removal of the residual media in which the at least three tissues are transferred, the at least three different tissues sink to the bottom of the carrier and then are pushed together due to a downslope force caused by an inclination of the slope. The recess helps to put the at least three different tissues into the spatial shape. The recess may have walls, i.e. further inclinations, in sides on which the tissues should not be able to move freely, i.e. to prevent a deviation from the spatial shape. Such walls are e.g. side walls. In particular preferred embodiments, medium can be removed to increase attachment of the different tissue to their respective proximal tissue in contact, as is described further below.

In preferred embodiments the recess has a length of 1 mm to 40 mm, preferably 2 mm to 30 mm, especially preferred 3 mm to 20 mm, and/or a width of 0.5 mm to 16 mm, preferably 1 mm to 12 mm. Such dimensions are usually suitable to accommodate the most common at least three different tissues that are to be fused, such as organoids.

The recess may have a depression (e.g. depth from a top side or top surface of the carrier) of at least 0.2 mm, preferably at least 0.5 mm or even more preferred at least 1 mm, over a length of at least 0.5 mm, preferably at least 2 mm, and/or a depression of at least 0.2 mm, preferably at least 0.5 mm or even more preferred at least 1 mm, over a width of at least 0.3 mm, preferably at least 0.5 mm, especially preferred at least 0.7 mm. Due to the slope, the depression/depth may vary but in these preferred embodiments, over at least the specified length/width (preferably longer/wider) the given dimension is preferably adhered to.

In preferred embodiments, in a cross-sectional view of the carrier the at least one slope comprises a region with a length of at least 2 mm in which region all tangents to the slope enclose an angle between 1° and 20° with a top surface / top side of the carrier. Thus, in a significant part of the recess/the depression, the slope of the depression is at an angle of from 1° to 20°. The top surface / top side is considered as a horizontal. Thus, the angle can also be regarded as 1° and 20° to the horizon. The angle is preferably 1° to 15°, especially preferred 2 ° to 10°.

The invention further provides an in vitro method of producing a fused tissue culture with at least three different tissues comprising attaching each of the at least three tissues to at least one other of the at least three tissues with applied force pushing tissues together, preferably with the force being applied over a period of at least 4 sec, especially preferred at least 5, 6, 7, 8, 9, 10, 12, 14, 16, 18, 20 or more seconds or any range between these values or more, culturing the at least three different tissues under conditions that maintains tissue fusion of the at least three different tissues. This aspect of the invention relates to an alternative to the spatial arrangement described above. Still, it relates to the same principles as the spatial shape aspect and all embodiments described above for the spatial shape aspect also relate to this aspect. Of course, this alternative defining the force can also be combined with the spatial shape mentioned above, e.g. the at least three tissues can be pushed together with the force in the spatial shape. Preferably, the force between two of the at least three different tissues is 0.01 pN to 1000 pN, preferably 0.1 pN to 400 pN, or up to 200 pN. The force may be due to the downslope force acting on the at least three different tissues, e.g. on the carrier/in the recess as described above. The force is preferably applied over a period of at least 10 sec, especially preferred at least 15 sec.

Preferably the carrier or mold (provided as such, or carrier or mold in the kit or uses thereof) has a low- or non-cell-adhe- sion surface, in particular at the deepest depression or even other regions which may be contacted by cells or tissue cultures, such as the recess. A low- or non-cell-adhesion surface is also referred to as a non-adherent surface. A suitable carrier preferably has a non-adherent surface. A non-adherent surface is a surface on which the cells are placed, and which has little or no adhesion tendency to the cells. Thus, the cells do essentially not attach and/or essentially not adhere to this surface. Without wishing to be bound by theory, use of a non-adherent surface provides a driving force for the cells or tissues to not adhere to the surface, but instead adhere to each other, thus forming or not disturbing a cell aggregate or fused tissue culture in the present invention. The non-adherent surface preferably does not interact negatively with a tissue. It may be biotolerant or bioinert.

Various materials are suitable for the carrier or mold, such as elastomers, thermoplastic polymers, resins, ceramics, hydrogels, silicones, non-corrosive metals or thermoplastics such as polylactide/polylactic acid (PLA) , polycaprolactone (PCL) , polystyrene or poly ( lactic-co-glycolic acid) (PLGA) .

A non-adherent surface may be formed by coating a material with a non-adherent biological or artificial material, or a non- adherent surface may be obtained by suitably shaping a non-adherent material, or by other means known in the art. A carrier or mold on or in which the tissue cultures are maintained or fused will from hereon be called a scaffold. Low-adhesion surfaces are preferably hydrophilic, neutrally charged or nonionic. They may have a hydrogel layer or coating that repels cell attachment. Such low-adhesion carriers and surfaces are known in the art, e.g. described in WO 2019/014635 or WO 2019/014636 (incorporated herein by reference) .

Scaffolds with a non-adherent surface are made of or are coated with, for example, a silicone, siloxane or polysiloxane, preferably polydimethylsiloxan (PDMS) , ethylene oxide, propylene oxide, polyethylene glycol, ( PEG) - ( co ) polymers (for instance PLL-g- (PEG) ) , poly ( ethylene oxide) (PEO) ( co ) polymers , agarose hydrogels, temperature-responsive materials below their Lower Critical Solution Temperatures (LCST) (for example Poly(N-iso- propylacrylamide) ) , hydrophobic materials (for example olefin polymers) , cell-repellent micro- and nanotopographies. Preferred materials of the scaffold (carrier or mold) are PDMS, poly (methyl methacrylate (PMMA) , styrene-ethylene-butylene-styrene (SEBS, Flexdym™) , polysiloxane, polyethylene glycol (PEG) , glass, polytetrafluoroethylene (PTFE) , titanium, platinum. A non-adherent surface may comprise a coating with albumin, e.g. bovine albumin, agar, a polyvinyl alcohol (PVA) , or a surfactant. Any of these materials can be used for the inventive mold. The scaffold (carrier or mold) may comprise or consist of any of these materials. Optionally, in addition a non- adherent coating may be used for the non-adherent surface. PDMS is a preferred material as it has excellent flexibility, anti-adhesive and biocompatible properties for the inventive tissue cultures and organoids .

Thus, forming a tissue culture according to the present invention is preferably achieved in a non-adherent scaffold. A non-adherent scaffold has at least one surface that essentially does not allow for the adherence of cells. Preferably, this is the side on or in which cells to form the tissue culture to be formed or fused are placed. A non-adherent scaffold can be formed from a non-adhering material, or can be formed from another material coated with a non-adherent material. A non-adherent carrier, petri dish, tube, or recessed carrier may for example be used as scaffold, but preferably, apart from the groove, the scaffold has a plate-like shape, such as for instance a more or less hexagonal, pentagonal, square, rectangular, triangular, oval or round shape. The carrier/scaf fold preferably is made by a material which is bioinert or biocompatible, thus having no or low unwanted interaction with the tissues besides providing a carrier to rest on or holding the medium.

In preferred embodiments of the inventive method a contact point between two of the at least three different tissues is higher than a level of medium fluid. Accordingly, the contact between the at least three different tissues is not immersed in the medium fluid. Of course, medium may be present due to surface tension and capillary effects but most of the medium fluid will descend below the contact point due to gravity. With a nonimmersed contact point, the different tissues have an increased stickiness and a higher propensity to attach to each other, thereby increasing adherence and fusion of the tissues. In practical embodiments, medium fluid may be removed from the tissues and/or from the recess. In other embodiments, the contact point between different tissues is immersed in a medium for fusion.

The present invention is preferably used on brain tissues, such as neural tissues or neuronal tissues. E.g., preferably the at least three different tissues comprise 1, 2 or 3 different brain tissues or neural tissues or neuronal tissues. Other non- neural or non-neuronal tissues may also be fused to at least one brain tissue or neural tissue or neuronal tissue. Such non-neu- ral or non-neuronal tissues are further described below and may e.g. stem from biopsies. They may be from tissues that form connections, in particular axon connections, with the brain tissue or neural tissue or neuronal tissue.

"Neural" is often referred to cells of the brain, which is not limited to neurons, whereas "neuronal" refers to neurons and cells involved with neurons. Astrocytes and glia cells are included in the group of neural cells, but not neuronal cells. Neural stem cells e.g. can give rise to neurons, glial cells (e.g. astrocytes, oligodendrocytes) and other neural stem cells. Neuronal on the other hand is exclusively for neurons. Neuronal stem cell only give rise to neurons, but not glial cells. A neuronal tissue comprises neurons and/or neuronal stem cells. The term neural cells includes neuronal cells. The term neural tissue includes neuronal tissue. Neural differentiation may lead to neural and/or neuronal cells. Neuronal cells are one of the more interesting research subjects of the invention, as they are involved with complex interactions between the different tissues, i.e. the tissue parts of the fused tissue culture that stem from the different tissues originally used in the inventive method. Such interactions are e.g. cell migration from one tissue type to another (as e.g. described in Bagley et al., Nat Methods, 2017, 14 (7) : 743-751 and WO 2018/197544 Al) and/or axon formation. Axons may reach from one of the different tissues/ tissue type to one, two or more of the different tissues/ tissue types. "Reaching from" means that the cell bodies are in this tis- sue/tissue type. The generation of differentiated brain tissues known in the art, e.g. as in the background section and in Xiang et al., Seminars in Cell and Developmental Biology 111 (2021) 40-51; WO 2017/160234 Al. The present invention further provides methods for the generation of differentiated brain tissues. Any differentiated brain tissue, such as organoids or spheroids, generated by a prior art method can be combined with different tissues generated by the present invention.

Preferably, the at least three different tissues are human tissues; likewise, preferably the (fused) tissue culture is a human or non-human primate culture, e.g. a human or non-human primate cell aggregate. The at least three different tissues may stem from culturing human or non-human primate cells, such as pluripotent cells, such as induced pluripotent cells (IPS cells) or embryonic stem cells. Human cells, tissues and cultures are especially preferred.

Gene names or gene symbols as used herein refer to the human genes and are described in databases such as GeneCards (www.genecards.org) or the HGNC database (www.genenames.org) . Gene symbols are defined e.g. by the "HUGO Gene Nomenclature Committee" (HGNC) . Other designations, such as long names, can be found at their website.

In particular preferred embodiments at least one of the different tissues contains dopaminergic neuronal tissue or a serotonergic neuronal tissue. A dopaminergic neuronal tissue may be for example a ventral midbrain tissue. A dopaminergic neuronal tissue may comprise TH+ cells. A serotonergic neuronal tissue may be for example a ventral hindbrain tissue. A serotonergic neuronal tissue may comprise TPH2 positive cells (Tryptophan Hydroxylase 2 positive cells) or 5-HT positive cells (serotonin positive cells) . In optionally combinable embodiments, the tissue may comprise cells expressing the striatal neuron markers dopamine receptor type 1 and/or 2 (DRD1+ and/or DRD2+) , and display heterogeneous dopaminergic neuron subtypes expressing for example- but not limited to- GIRK2, CALB1, 0TX2, S0X6, ALDH1A1 or GABA, or combinations thereof.

The dopaminergic neuronal tissue may comprise dopaminergic cells. The dopaminergic cells in some embodiments are A9 dopaminergic neurons and/or A10 dopaminergic neurons. A9 dopaminergic neurons may be in a tissue resembling functionally and/or morphologically the substantia nigra compacta. The resemblance may be with regard to A9 dopaminergic neuron function and/or cell connectivity. A9 dopaminergic neurons are associated with the nigrostriatal pathway and project into the dorsolateral striatum, where they have a crucial function in fine motor control (Arenas, Denham and Villaescusa, Development 142, 1918-36, 2015) . These A9 dopaminergic neurons degenerate in Parkinson's disease (Lees, Shin and Revesz, Lancet 373, 2055-66, 2019) . A10 dopaminergic neurons may be in a tissue resembling functionally and/or morphologically the ventral tegmental area. A10 dopaminergic neurons may project to the striatum tissue, especially a ventral striatum, as well as into the cortex tissue, majorly into the limbic system and prefrontal cortex. A10 dopaminergic neurons may be in a tissue resembling functionally and/or morphologically the ventral tegmental area. The resemblance may be with regard to A10 dopaminergic neuron function and/or cell connectivity. A10 dopaminergic neurons are associated with affective encoding ("dopaminergic reward pathway") (Lio, Shin and Ikemoto, Neuropsychopharmacology 33, 2182-94, 2014) .

The invention further provides an in vitro method of generating a ventral hindbrain tissue culture, comprising inducing neural differentiation in pluripotent cells or multipotent neural stem cells with a neural induction medium comprising at least one TGF-beta inhibitor and a Wnt activator, and allowing the cells during said induction to form a cell aggregate; culturing the cell aggregate to form a tissue culture. The pluripotent or multipotent neural stem cells are preferably human cells, such as induced pluripotent cells (IPS cells) . The ventral hindbrain tissue culture may be a ventral hindbrain organoid. The inventive ventral hindbrain tissue culture may be one of the at least three different tissues used in the fusion. The fused tissue culture may comprise parts that stem from the inventive ventral hindbrain tissue culture. The generated ventral hindbrain tissue culture preferably has serotonergic neurons.

The TGF-beta inhibitor may be a TGF-beta signaling pathway inhibitor as it inhibits TGF-beta function. It may be an inhibitor of the TGF-beta superfamily pathways. Preferably one or - more preferred - two, or even more, TGF-beta inhibitors are used. In some preferred embodiments it is comprised of or comprises one or more inhibitors of the TGF-beta receptor or a TGF- beta pathway. In preferred embodiments the at least one TGF-beta (pathway) inhibitor is at least one SMAD inhibitor, preferably DMH1 (dorsomorphin homolog 1) ) and/or SB431542 (4- [4- (2H-1, 3- Benzodioxol-5-yl ) -5- (pyridin-2-yl ) -lH-imidazol-2-yl] benzamide) . Further preferred TGF-beta inhibitors, that are alternatives or combinable with those mentioned before, are Noggin (a Protein that binds and inactivates BMP proteins, which belong to the TGF-beta superfamily) , A 83-01 (a small molecule) , LDN 193189 (a small molecule) and/or Dorsomorphine . Further TGF-beta inhibitors are known in the art, e.g. as disclosed at www.med- chemexpress . com/Targets/TGF- (beta) %20Receptor. html .

Wnt activators are known in the art and described e.g. in Nusse and Clevers, Cell 169, 2017: 985-999. Nusse and Clevers discuss the Wnt/b-catenin signalling pathway and its manipulation in stem cells. The Wnt activator may be a GSK3 inhibitor. Further WNT activators are disclosed at the website web. Stanford . edu/ group/ nusselab/ cgi-bin/ wnt/ smallmolecules . In preferred embodiments the Wnt activator is a WNT ligand, such as WNT-3a, or CHIR99021 ( 6- [ [ 2- [ [ 4- ( 2 , 4-dichlorophenyl ) -5- ( 5-methyl- IH-im- idazol-2-yl) -2-pyrimidinyl ] amino] ethyl] amino] -3-pyridinecarboni- trile) . Further preferred Wnt activators are WAY-316606, ABC99, IQ1, QS11, SB-216763. The pluripotent cells are preferably as described above, especially embryonic stem cells (ESCs) or induced pluripotent stem cells (iPSCs) , in particular preferred human ESCs or iPSCs.

In preferred embodiments, the pluripotent cells or multipotent cells, that emerge from pluripotent cells, are treated with an activator of sonic hedgehog signaling; preferably a Smooth- ened agonist, e.g. SAG ( 3-Chloro- - [ trans-4- (methylamino ) cyclohexyl ] -N- [ [ 3- ( 4-pyridinyl ) phenyl ] methyl ] benzo [b thiophene- 2 -carboxamide ) , retinolic acid and/or purmorphamine . This treatment is preferably before treatment with the TGF-beta inhibitor and/or Wnt activator. The activator of sonic hedgehog signaling is preferably a Smoothened agonist, especially preferred SAG (N- Methyl-N'- ( 3-pyridinylbenzyl ) -N'- ( 3-chlorobenzo [b] thiophene-2- carbonyl) -1, 4-diaminocyclohexane) or purmorphamine (PMA) , or a combination thereof.

In further preferred embodiments, the cell aggregate is treated with dissolved extracellular matrix, such as Matrigel, or an extracellular matrix component (or Matrigel component) , preferably selected from collagen and/or laminin. Dissolved extracellular matrix means that it is in the liquid state. Even more preferred, the cell aggregate is not embedded in an extracellular matrix in a solid or gel state. The aggregate state of the extracellular matrix component may be controlled by the concentration. In stronger diluted form/at lower concentration, the extracellular matrix component may be in liquid or dissolved state and in lower diluted form/at higher concentration, the extracellular matrix component may form a gel and turn solid. The pluripotent cells are preferably as described above , especially embryonic stem cells (ESCs ) and induced pluripotent stem cells ( iPSCs ) , in particular human ESCs and iPSCs , in particular human IPS cells .

In a further preferred embodiment , the cells or cell aggregate is not treated with a gamma-secretase inhibitor such as DAPT . This allows a more cell growth driven/ less arti ficially- forced di f ferentiation .

Also provided is an in vi tro method of generating a ventral midbrain tissue culture , comprising inducing neural di f ferentiation in pluripotent cells or multipotent neural stem cell with a neural induction medium comprising at least one TGF-beta inhibitor and a Wnt activator and a ROCK inhibitor and allowing the cells during said induction to form a cell aggregate ; culturing the cell aggregate to form a tissue culture . The TGF-beta inhibitor may be any one as described above , preferably Noggin and/or SB431542 , which work particularly well in this method . The Wnt activator may be any one of the above , preferably CHIR99021 . The ROCK inhibitor is preferably Y-27632 , which works well in this method . In a preferred embodiment , culturing the cell aggregate comprises treating the cell aggregate with a FGF ( fibroblast growth factor ) , preferably FGF8 , and/or an activator of sonic hedgehog signaling . The activator of sonic hedgehog signaling may be SAG, especially preferred SAG at a concentration of 100 nM to 1000 nM, preferably 200 nM to 600 nM . In further preferred embodiments , the cell aggregate is treated with dissolved extracellular matrix, such as Matrigel , or an extracellular matrix component ( or Matrigel component ) , preferably selected from collagen and/or laminin . Dissolved extracellular matrix means that it is in the liquid state . Even more preferred, the cell aggregate is not embedded in an extracellular matrix in a solid or gel state . The aggregate state of the extracellular matrix component may be controlled by the concentration . In stronger diluted form/at lower concentration, the extracellular matrix component may be in liquid or dissolved state and in lower diluted form/at higher concentration, the extracellular matrix component may form a gel and turn solid . The pluripotent cells are preferably as described above , especially embryonic stem cells (ESCs ) and induced pluripotent stem cells ( iPSCs ) , in particular human ESCs and iPSCs . In a further preferred embodiment , the cells or cell aggregate is not treated with a gamma-secretase inhibitor such as DAPT . This allows a more cell growth driven/less artificially-forced differentiation.

Further provided is an in vitro method of generating a striatal tissue culture, comprising inducing neural differentiation in pluripotent cells or multipotent neural stem cell with a neural induction medium comprising at least an activator of sonic hedgehog signaling and a WNT pathway inhibitor and ROCK inhibitor and allowing the cells during said induction to form a cell aggregate and culturing the cell aggregate to form a tissue culture. The activator of sonic hedgehog signaling is preferably SAG. The WNT pathway inhibitor is preferably IWP2. The ROCK inhibitor is preferably Y-27632, which works well in this method. In preferred embodiments, the cell aggregate is treated with dissolved extracellular matrix, such as Matrigel, or an extracellular matrix component or Matrigel component, preferably selected from collagen and/or laminin. Even more preferred wherein the cell aggregate is not embedded in an extracellular matrix in a solid or gel state. Reference is made to the preceding paragraph on the state of extracellular matrix component. In a further preferred embodiment, the cells or cell aggregate is not treated with a gamma-secretase inhibitor such as DAPT. This allows a more cell growth driven/less-artif icially-forced differentiation. The pluripotent cells are preferably as described above, especially embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs) , in particular human ESCs and iPSCs.

WNT pathway inhibitors are disclosed e.g. in Nusse and Clevers, Cell 169, 2017: 985-999. Further WNT activators are disclosed at the website web.stan- ford.edu/group/nusselab/cgi-bin/wnt/smallmolecules. The WNT pathway inhibitor may be a Porcupine inhibitor or a tankyrase inhibitor and/or Axin inhibitor. The WNT pathway inhibitor is preferably selected from Wnt-C59, IWR-1, XAV939, IWP-2, IWP-4, DKK1 or a combination thereof. Particular preferred is IWP-2.

Preferably, culturing the cell aggregate comprises treating the cell aggregate with a FGF (fibroblast growth factor) , preferably FGF4, FGF8, FGF17, FGF18. Especially preferred is FGF4, especially in the method of generating a ventral hindbrain tissue culture. The FGF treatment may be following the formation of a cell aggregate. Some cells of the aggregate may have differentiated towards a neural fate (such as neuronal precursor cells or even further differentiated to neuronal cells) . Said differentiation towards a neural fate may have occurred before FGF treatment starts and/or occurs during FGF treatment.

The cell aggregate may be an embryoid body. An embryoid body is a cell aggregate which has the three different germ layers (Ectoderm, Mesoderm, Endoderm) . As it is of interest to increase the ectoderm content and reduce or remove mesoderm and endoderm, the inventive methods (e.g. the usage of neural induction medium from early on, preferably from the beginning) results in structures which are mostly comprised of neuroectoderm and no or very low endoderm/mesoderm.

Extracellular matrix (ECM) may comprise collagens, laminins, entactins, glycosaminoglycans (e.g. hyaluronan) , proteoglycans (e.g. neurocans) , glycoproteins (e.g. tenascin-R) , heparin-sul- fated proteo-glycans , fibronectin, vitronectin or any combination thereof. Extracellular matrix may be from the Engelbreth- Holm-Swarm tumor or any component thereof such as laminin, collagen, preferably type 4 collagen, entactin, and optionally further heparan-sulf ated proteoglycan or any combination thereof. Such an ECM is Matrigel. Matrigel is known in the art (US 4,829,000) and has been used to model 3D heart tissue previously (WO 01/55297 A2 ) or neural tissue (WO 2014/090993) . Contrary to these previous uses, the present invention does not require a 3D matrix formation by ECM. A dissolved ECM or component thereof may be used as treatment agent or coating for or after fusion. Preferably the matrix comprises laminin, collagen and entactin, preferably in concentrations 30%-85% laminin, 3%-50% collagen and optionally entactin, usually 0.5%-10% entactin. Laminin may require the presence of entactin to form a gel if collagen amounts are insufficient for gel forming. For a dissolved ECM, entactin may be absent or below gel forming concentration. Even more preferred, the ECM comprises in parts by weight about 50%- 85% laminin, 5%-40% collagen IV, optionally l%-10% nidogen, optionally l%-10% heparan sulfate proteoglycan and 0%-10% entactin. All %-values given for the matrix components are in wt.-%. Entactin is an optional bridging molecule that interacts with laminin and collagen.

The inventive ventral hindbrain tissue culture, or any of the at least three different tissues, which are preferably organoids, can be obtained from culturing pluripotent stem cells. In principle, the cells may also be totipotent, if ethical reasons allow, e.g. for non-human mammalian cells.

A "totipotent" cell can differentiate into any cell type in the body, including the germ line following exposure to stimuli like that normally occurring in development. Accordingly, a totipotent cell may be defined as a cell being capable of growing, i.e. developing, into an entire organism.

The cells used in the methods according to the present invention are preferably not totipotent, but pluripotent or at least multipotent. The multipotent cell may be a multipotent neural stem cell. For a dopaminergic tissue, the multipotent cell may be a midbrain dopaminergic (mDA) neuron progenitor cell. Such midbrain dopaminergic (mDA) neuron progenitor cell may for example be used in the generation of ventral midbrain tissues, like spheroids or organoids, as starting cell in the inventive method.

A "pluripotent" cell is not able of growing into an entire organism, but is capable of giving rise to cell types originating from all three germ layers, i.e., mesoderm, endoderm, and ectoderm, and may be capable of giving rise to all cell types of an organism. Pluripotency can be a feature of the cell per se, e.g. in embryonic stem cells, or it can be induced artificially. E.g. in a preferred embodiment of the invention, the pluripotent stem cell is derived from a somatic, multipotent, unipotent or progenitor cell, wherein pluripotency is induced. Such a cell is referred to as induced pluripotent stem cell herein. The somatic, multipotent, unipotent or progenitor cell can e.g. be used from a patient, which is turned into a pluripotent cell, that is subject to the inventive methods. Such a cell or the resulting tissue culture can be studied for abnormalities, e.g. during tissue culture development according to the inventive methods. A patient may e.g. suffer from a neurological disorder or cerebral tissue deformity. Characteristics of said disorder or deformity can be reproduced in the inventive tissue cultures and investigated.

A "multipotent" cell is capable of giving rise to at least one cell type from each of two or more different organs or tissues of an organism, wherein the said cell types may originate from the same or from different germ layers, but is not capable of giving rise to all cell types of an organism. An example of multipotent cells are multipotent neural stem cell, such as midbrain dopaminergic (mDA) neuron progenitor cells.

In contrast, a "unipotent" cell is capable of differentiating to cells of only one cell lineage.

A "progenitor cell" is a cell that, like a stem cell, has the ability to differentiate into a specific type of cell, with limited options to differentiate, with usually only one target cell. A progenitor cell is usually a unipotent cell, it may also be a multipotent cell.

The inventive tissues and/or cells are preferably mammalian, e.g. human or nun-human primate.

With decreasing differentiation capabilities, stem cells differentiate in the following order: totipotent, pluripotent, multipotent, unipotent. During development of the inventive tissue cultures, stem cells differentiate from pluripotent (also totipotent cells are possible) into multipotent neural stem cells, further into unipotent stem cells and subsequently into non-stem tissue cells. Tissue cells may e.g. be neural or neuronal cells or neuroepithelial cells, such as glial cells.

In preferred embodiments, the inventive tissue cultures (the fused tissue culture and the preceding at least three different tissues) are in vitro grown. Since tissues are grown from human or mammalian, preferably primate, pluripotent stem cells, this allows growth of human or primate tissue without obtaining human fetal tissue samples.

In a particular preferred embodiment, the cells of the present invention (including all embodiments related thereto) , are human cells.

Preferably, the pluripotent cells during aggregate formation are cultured on a low-cell-adhesion surface, and/or as 3D culture. Low-cell-adhesion surfaces, also referred to as non-adher- ent surfaces, have been described above and the same applies to this embodiment of the invention.

Culturing as 3D culture means that the cells and/or the forming cell aggregates and tissues are not impeded in their growth by binding to a surface. They may remain free floating in a suspended condition so that expansion in all 3D directions is uniformly possible, or, if settled onto the bottom of a carrier surface, do not attach to that surface, again so that expansion in all 3D directions is uniformly possible. Such culturing is in a low-adhesion culture so that the cells, aggregates or tissues do not attach to culture vessel walls. Low-adhesion surfaces are preferably hydrophilic, neutrally charged or non-ionic. They may have a hydrogel layer or coating that repels cell attachment, forcing cells into a suspended state. Such low-adhesion culture vessels are known in the art, e.g. described in WO 2019/014635 or WO 2019/014636. Preferably the bottom of the culturing carrier is round, especially concave. Alternatively, it may be flat or have a V-bottom shape.

Preferably, the neural induction medium comprises glutamine, non-essential amino acids, heparin, transferrin, insulin, vitamins, salts or any combination thereof. Neural induction medium has been described by Eiraku et al. (Cell Stem Cell (2008) 3: 519-532) , US 2011/0091869 Al, WO 2011/055855 Al, Lancaster et al. (Nature (2013) 501: 373-379) , and WO 2014/090993 Al. Neural induction medium may comprise DMEM/F12 and/or N2 supplement (Price and Brewer. Protocols for Neural Cell Culture: Third Edition. (2001) : 255-64) , glutamine, non-essential amino acids, heparin, or any combination thereof. In preferred embodiments, the neural induction medium comprises selenin and/or transferrin, and/or insulin. Neural induction medium preferably lacks growth factors that would differentiate neural tissue to a particular neural fate. Such absent growth factors may be any one of Hedgehog, Wnt, TGF-beta (e.g. Bmp signaling) , Notch, retinoids, or FGF, or any combination thereof. In other embodiments, e.g. for region-specific differentiation, specific, dose-controlled and temporally restricted Hedgehog, Wnt, TGF-beta, Notch, retinoids, or FGF species or small molecules activating corresponding ( sub) pathways may be present. For example, neural induction medium may comprise inorganic salts: CaC12, Fe(NO ₃ )3, KC1, MgCl ₂ , NaCl, NaHCO ₃ , NaH ₂ PO ₄ , ZnSO ₄ ; D-Glucose; HEPES; pyruvate; Amino acids: L-Alanine, L-Arginine, L-Asparagine, L-Cys- teine, L-Glutamine, L-Glutamate, Glycine, L-Histidine, L-Isoleu- cine, L-Leucine, L-Lysine, L-Methionine, L-Phenylalanine, L-Pro- line, L-Serine, L-Threonine, L-Tryptophan, L-Tyrosine, L-Valine; Vitamins: D-Calcium pantothenate, Choline chloride, Folic acid, i-Inositol, Niacinamide, Pyridoxal, Riboflavin, Thiamine, Vitamin B12; or any combination thereof. The medium preferably comprises 1, 2, 3, 4 or more of the group of inorganic salts, 1, 2, 3, 4 or more of the group of Amino acids, and 1, 2, 3, 4 or more of the group of Vitamins. Furthermore, glucose and/or pyruvate are preferred. Insulin, preferably human insulin, may be present to increase glucose uptake. The medium may comprise phenol red. The medium may comprise heparin. Further components of the medium may be selected from transferrin, especially human transferrin, progesterone, putrescine, selenite, or a combination thereof. Preferably the medium comprises (a) Sodium chloride at a concentration of between 70 and 150 mM;

(b) a neuroactive inorganic salt at a concentration of between 0.000001 and 10 mM, selected from the group consisting of Potassium Chloride, Calcium Chloride, Magnesium Sulfate, Magnesium Chloride, Ferric Nitrate, Zinc sulfate, Cupric sulfate, Ferric sulfate, and combinations thereof; (c) Glycine at a concentration of between 0.0001 and 0.05 mM; (d) L-alanine at a concentration of between 0.0001 and 0.05 mM; and (e) L-serine at a concentration of between 0.001 and 0.03 mM. Any component may be present at a concentration that maintains the survival and neural functionality of a neural cell cultured in the medium (EP 2 986 715 Bl) .

The invention further provides a tissue culture comprising a planar or linear fusion of at least three different neural or neuronal tissues. Such a tissue culture may be obtained from the inventive method, described above as the fused tissue culture. The at least three different tissues mentioned herein for the inventive method form part of the (fused) tissue culture and contribute as different tissue parts of the (fused) tissue culture) . They may be different tissue types, i.e. different kinds of tissues. For the tissue culture of the invention these are preferably neural or neuronal tissues, i.e. the "at least three different neural or neuronal tissues" mentioned above. The tissue culture is a cellular tissue culture and is an aggregate (or aggregation) of cells. It may be an organoid, stemming from fused organoids (as at least three different neural or neuronal tissues) . The tissue culture may therefore also be referred to as "fused organoid", "organoid fusion" or "organoid assembloid" herein. The same options, embodiments and preferred elements described for the at least three different tissues as described above also apply to the fused tissue culture and relate to its parts and/or at least three different tissues. The tissue culture may of course be cultured further and achieve larger sizes, as possible for in vitro tissue cultures, in particular organoids. No limit in culturing duration of organoids is known to date, with cultivation times of more than 2 years with still functional tissues.

The inventive tissue culture may be used to study interactions between the different parts (the at least three different neural or neuronal tissues) . Preferably, the tissue culture comprises one or more neural cells with a cell body in one of said at least three different neural tissues and an axon to another one of said at least three different neural tissues (axonal outgrowth) . The axons may interact with target neurons, such as dendrites or cell bodies (axonal innervation) . In preferred embodiments, a tissue comprises neuronal cells that project their axons into at least two of the other neural or neuronal tissues. Axon activity and/or axon formation are particular interesting objects of study of the inventive tissue culture. The inventive tissue may be just formed, in particular by fusion and not yet comprise such axons that reach into different tissues. However, preferably the fused tissue culture may have the cells that are capable to form such axons or are in the process to form such axons or contain cells that give rise to cells that are capable to form axons. Examples include axons of dopaminergic cells or serotonergic cells, which are described further herein. Briefly, such axons may form from specific tissues, such as ventral midbrain tissues (for dopaminergic cells, among others) or ventral hindbrain tissues (for serotonergic cells, among others) .

In preferred embodiments, the at least three different neural or neuronal tissues are selected from central nervous system tissue, preferably of a telencephalic tissue, in particular preferred of cortical, striatal, lateral ganglionic eminence (LGE) , medial ganglionic eminence (MGE) or caudal ganglionic eminence (CGE) tissue; preferably of a diencephalic tissue, especially preferred thalamic, hypothalamic, epithalamic, subthalamic or optic cup tissue; preferably mesencephalic tissue, especially preferred tectum, tegmentum, ventral tegmental area (VTA) , substantia nigra (SN) , such as SN pars compacta (SNpc) , Nucleus ruber tissue; preferably rhombencephalon tissue, especially preferred cerebellum, medulla, pons, raphe nuclei tissue; preferably spinal cord tissue; or peripheral nervous system tissue; or any combination thereof. Any of these tissues may be combined in the inventive tissue culture, among others. These tissues may also be used in the inventive method for fusion. These tissues may be in the form of an organoid for fusion in the inventive method .

The at least three different tissues of the inventive method or the at least three different brain, neural or neuronal tissues of the inventive fused tissue culture may be fused with non-neuronal or non-neural tissues, e.g. muscle tissue, heart tissue, intestinal tissue, bone, vascular tissues, liver tissue, spleen tissue, kidney tissue, skin tissue. Tissues fused in the inventive tissue culture can be biopsies. Biopsies can be neural or neuronal (like a biopsy of the brain) or non-neuronal or non- neural. An example is to fuse brain regions associated with motor control with spinal cord and/or muscle. Preferably the non- neuronal or non-neural tissues to be fused may comprise nerveendings that connect with the neuronal or neural tissues in the tissue culture.

In preferred embodiments, the (fused) tissue culture comprises cortical circuit tissue; neural retina tissue; dorsoven- tral forebrain tissue, preferably lateral ganglionic eminence (LGE) tissue, medial ganglionic eminence (MGE) tissue or caudal ganglionic eminence (CGE) tissue; olfactory tissue, preferably comprising olfactory bulb neurons, amygdala tissue, cortex tissue or hippocampus tissue; limbic system tissue, preferably thalamus, hippocampus, amygdala, hypothalamus; dopaminergic system tissue, preferably ventral tegmental area (VTA) , substantia nigra (SN) , such as SN pars compacta (SNpc) tissue, nucleus ac- cumbens, olfactory tubercle, prefrontal cortex, cortical, caudate nucleus, putamen, hypothalamus, brainstem, spinal cord tissue; serotonergic system tissue, preferably rostral serotonergic group tissue, such as different tissues of the cortex, striatum, amygdala, substantia nigra, pons, hippocampus, entorhinal cortex, locus coeruleus or caudal serotonergic group tissue, such as trigeminal motor nucleus, dorsal nucleus of vagus nerve, in- termediolateral nucleus, medulla, mesencephalon; cerebellar pathway tissue, preferably cortical, thalamic, pons, vestibular nuclei tissue; noradrenergic pathway tissue, preferably cerebellum, spinal cord, hippocampus, basal ganglia, cortex tissue; or any combination thereof. Any of these tissues may be combined in the inventive tissue culture , among others . These tissues may also be used in the inventive method for fusion . These tissues may be in the form of an organoid for fusion in the inventive method . Of course , any of these tissues may be combined in a ( fused) tissue culture with the tissues mentioned in the preceding paragraph .

The inventive tissues , or the parts that stem from these tissues in the fused tissue culture , may comprise neurons of such a tissue . E . g . a striatum tissue comprises striatal neurons , a ventral hindbrain tissue comprises ventral hindbrain neurons , a ventral midbrain tissue comprises ventral midbrain tissue neurons , a cortex tissue comprises cortex neurons , etc . for any of the above tissues or tissue parts .

In particular preferred embodiments , the at least one of the neural tissues is selected from ventral hindbrain, ventral midbrain, striatum, cortex . These tissues are particular preferred originating tissues and receiving tissues of dopaminergic or serotonergic axons . The originating tissues have cell bodies of dopaminergic or serotonergic cells that form the axons which protrude into the receiving tissue . The receiving tissues have the axon ends with the nerve ends of these cells that connect to cells in the receiving tissues . Originating tissues may be ventral hindbrain and/or ventral midbrain . Receiving tissues may be striatum and/or cortex . Notably, reciprocal connections or reciprocal proj ections from receiving tissues into originating tissues can and ideally should also occur, mimicking feedbacksignaling and circuit formation in the brain .

Preferably the tissue culture comprises ventral midbrain tissue in contact with striatum tissue and said striatum tissue being in contact with cortex tissue . "Contact" refers to the tissue body contact , e . g . as controlled by the fusion or attachment during fusion . Axons are not counted as contacts . In fact , axons may reach into all other tissues starting from cells in one of the originating tissues that comprise the cell bodies of the axon forming cells . The ventral midbrain tissue in contact with striatum tissue and said striatum tissue being in contact with cortex tissue can be the product of a linear fusion in the order : ventral midbrain tissue - striatum tissue - cortex tissue . Axons may form from the ventral midbrain tissue ( contains the cell bodies ) to the striatum tissue and/or cortex tissue . Although the ventral midbrain tissue may not be in contact with the cortex tissue , axons may still form between these tissues since axons can be long and reach through or proj ect beside other tissues (here through/beside the striatum tissue ) . In preferred embodiments , this tissue culture may be further combined with a ventral hindbrain tissue . The ventral hindbrain tissue may be in contact with the ventral midbrain tissue . Such a tissue culture may be product of a linear fusion of : ventral hindbrain tissue - ventral midbrain tissue - striatum tissue - cortex tissue . Ventral hindbrain tissue may contain cells that form axons that reach to cells of the other tissues selected from ventral midbrain tissue , striatum tissue , and cortex tissue . Such axons may be serotonergic axons .

The tissue culture may have a si ze of 100 pm to 100 mm, preferably 300 pm to 30 mm, in its longest dimension . The tissue culture may be larger than the di f ferent tissues from which it is formed . The tissue culture may be cultured after fusion and enlarge through growth . "Si ze" refers to the longest dimension in 3D space , as mentioned above . The tissue culture may have an elliptical shape with a longest dimension ( length) and a widest dimension perpendicular to the longest dimension (width) . The width may be 70 pm to 20 mm .

Preferably the tissue culture comprises a neural cell from one of the at least three di f ferent neural or neuronal tissues in at least one of the other at least three di f ferent neural or neuronal tissues . The neural cell from one of the at least three di f ferent neural or neuronal tissues may have migrated to at least one of the other at least three di f ferent neural or neuronal tissues . The provided tissue may have migrating or migrated cells . Such cells may have moved to another tissue type (migration recipient tissue ) that is not of the same tissue type . The migrated cells may constitute not more than 5% of the cells of the migration recipient tissue ( % of cell numbers ) .

Preferably the at least three di f ferent neural or neuronal tissues or the tissue culture , respectively, comprise a neural tissue of a neural or neuronal tissue type and another neural or neuronal tissue of said neural or neuronal tissue type with its cells being modi fied by a genetic modi fication ( genetically engineered cells ) . Tissue types are mentioned as above and are selected e . g . from ventral hindbrain tissue , ventral midbrain tissue, striatum tissue, cortex tissue, olfactory tissue, neural retina tissue, etc. According to this embodiment there may be two or more tissues of these types, one being genetically modified (or: genetically engineered) and another being non-modified or having a different genetic modification. Behaviour, growth and activity (e.g. axon formation and or axon activity, such as neuron signalling) of cells of the genetically modified vs. nonmodified or differently modified cells can be studied with such a tissue culture. The tissues of this type (mutated, non-mu- tated, different mutated) may be in contact with a different tissue, such as a tissue that is a recipient for axons of the tissues of this type (genetically modified, non-modified, differently modified) .

Preferably the tissue culture comprises dopaminergic neurons and/or serotonergic neurons. Such neurons are a particularly interesting object for study. Dopaminergic neurons may originate from ventral midbrain tissue. Serotonergic neurons may originate from ventral hindbrain tissue. Preferably the dopaminergic neurons connect between ventral midbrain tissue and striatum tissue and/or connect between ventral midbrain tissue and cortex tissue .

Preferably the tissue culture comprises TH+ (Tyrosine Hydroxylase positive) cells, preferably wherein said TH+ cells have axons in a striatal and/or cortical tissue of the tissue culture. Such cells may be dopaminergic cells. Preferably the tissue culture comprises TPH2 positive cells or 5-HT positive cells, preferably wherein said TPH2 positive cells or 5-HT positive cells have axons in a striatal and/or cortical tissue and/or ventral midbrain tissue of the tissue culture. Such cells may be serotonergic cells.

The invention further provides a method of investigating neuron formation, comprising introducing one or more neural stem cell, neural progenitor cell, neuronal cell, neural cell, glial cell, astrocyte or oligodendrocyte into a tissue culture of the invention and monitoring cell differentiation and/or cell growth. In some embodiments, a mixture or combination of these cells may be introduced, e.g. by injection, into a tissue culture of the invention. The introduction of the one or more cells or combinations thereof and monitoring developments, such as neuron maturation and/or activity can be used to study modified (e.g. genetically) or diseased cells or tissue cultures of diseased tissue cultures that receive the introduction of a healthy cell or a diseased cell or a treated diseased cell. Such a disease may be Parkinson's disease (Hiller et al., npj Regenerative Medicine (2022) 24) or Amyotrophic Lateral Sclerosis (ALS) . Such studies may be used to investigate a potential cell therapy (e.g. by introducing a healthy cell to a diseased tissue culture) or to increase understanding of diseased cells or diseased tissue cultures. It is also possible to study healthy cells for cell therapy in a normal environment (healthy tissue culture) , to study their properties.

The invention further provides a method of testing or screening a candidate compound for influencing neural cells of a fused tissue culture, comprising contacting cells or a tissue in a method of the invention with the candidate compound or contacting a tissue of the invention with the candidate compound and maintaining said contacted tissue in culture, and observing any change in the neural cell as compared to said tissue without contacting by said candidate compound.

Such a method may comprise treating the cells or fused tissue culture with at least one candidate compound. The effects of this method may be compared to the method without the respective at least one candidate compound as control comparison. Otherwise, the control comparison is performed like-wise as is common for controls in order to evaluate the effect of the at least one candidate compound only. The neural cells can be influenced in various ways e.g. morphologically, genetically, in their expression pattern, or through environmental exposure to e.g. drugs or neural or neuronal circuit activity level (e.g. optogenetic, chemogenetic, electrical stimulation, calcium event duration etc.) or epigenetically . The mentioned observation or any study of the inventive tissue or its cells can detect e.g. changes in differentiation, neural morphological features (such as cell shape, axon shape, dendrite shape, neurite shape, processes shape) , in particular axonal varicosities, changes in neural activity (such as neuronal activity) , transcriptional changes, protein expression changes, epigenetic, genetic changes and/or neuron maturation changes. Any such change can be observed in the inventive method. In particular preferred, the varicosity diameter of axons, e.g. of TH+ axons or of TPH2+ cells or 5-HT+ cells, is observed. Any of the observed parameters and effects can be used for the comparison.

Compounds that are expected to cause an observable change or influence and that may be tested or used as comparative control in the inventive method are drugs that are addictive to a human, dopamine-releasing drugs, dopamine reuptake inhibitors, dopamine transporter ligands or inhibitors, cocaine (as shown in the examples) , amphetamines and compounds which generally interact with dopaminergic neurons or their target neurons.

Further compounds that are expected to cause an observable change or influence and that may be tested or used as comparative control in the inventive method are drugs that interfere with the serotonergic systems, such as compounds which are used to reduce depression or reduce serotonine syndrome, or might cause depression, or might cause serotonine syndrome, or have an effect on serotonergic signaling, or substances which can cause the serotonergic syndrome. Preferably for observing these changes or influences the tissue comprises TPH2 positive cells or 5-HT positive cells and serotonergic axons.

Serotonine syndrome is typically caused by serotonergic candidate compounds. Serotonergic effects and the risk of causing serotonine syndrome may be screened or tested with the inventive method. Example possible serotonergic candidate compounds include selective serotonin reuptake inhibitor (SSRI) , serotonin norepinephrine reuptake inhibitor (SNRI) , monoamine oxidase inhibitor (MAOI) , tricyclic antidepressants (TCAs) , amphetamines, pethidine (meperidine) , tramadol, dextromethorphan, buspirone, L-tryptophan, 5-hydroxytryptophan, St. John's wort, triptans, ecstasy (MDMA) , metoclopramide, or cocaine.

In addition, the invention allows to study withdrawal symptoms on neural circuits. E.g. in the inventive method after contacting the cells or tissue with the candidate compound there can be a step of maintaining said contacted tissue in culture without contacting it with the candidate compound (withdrawal) . Such a maintaining step without candidate compound exposure can be compared with a maintaining step with further candidate compound contact and/or compared with a tissue that has not been contacted by the candidate compound in the screening method as mentioned above. Differences apparent by the comparison, as mentioned above, can be observed. E.g. the contacting with the candidate compound can be a single exposure or a for one or more days, e.g. up to 200 days, preferably 40 to 130 days, and/or the maintaining step without candidate compound exposure can be for 10 to 100 day, preferably 20 to 80 days. The maintaining step without candidate compound exposure can also be longer, in particular longer than 100 days, to determine recovery from the previous contracting step. This may take few or many days, depending on the changes and effects caused by the candidate. Such a recovery time may be observed.

A treatment with a candidate compound can be in tissue culture for a time sufficient to elicit a response. Example treatments are an exposure with the candidate compound for 10 min to 24 h, or 20 min to 12 h, or 30 min to 6 h, or 40 min to 4h, or 50 min to 3h. Exposures may be daily, twice daily, or every 2 ^nd or 3 ^rd or 4th day. Such treatment options can be combined with the above-mentioned contacting durations, e.g. as shown in Fig. 17A.

Preferably a tissue culture with TH+ cells is used to observe dopaminergic effects of the candidate compound. Preferably a tissue culture which comprises serotonergic axons, TPH2 positive cells or 5-HT positive cells, and/or ventral hindbrain is used to observe serotonergic effects of the candidate compound.

The inventive method allows to study a pathology of neural circuits across neural tissue types, i.e. connecting cells between different tissue types. The candidate compound can be used to screen effects on the pathology, in particular, if the pathology is ameliorated. With the inventive method, compounds useful to ameliorate pathological conditions, like Parkinson's disease or depression or serotonine syndrome, can be screened and tested.

The invention further provides a mold or carrier for tissue fusion comprising a top surface and a recess in said top surface, wherein said recess is elongated and comprises a deepest depression and at least one slope along the direction of the elongation, the at least one slope descending to the deepest depression, wherein the mold/carrier at least at the deepest depression comprises a low- or non-cell-adhesion surface. Low- or non-cell-adhesion surfaces are described above and the same applies here. E.g. a preferred surface is of or comprises polydi- methylsiloxan (PDMS, such as provided in the Sylgard 184 Elastomer Kit) . In preferred embodiments, the low- or non-celladhesion surface is of a coating with a low- or non-cell-adhe- sion material. The mold or carrier may have the dimensions and/or the recess as mentioned above. The mold or carrier may have one or more recesses of such dimensions, such as 2, 3, 4, 5, 6, 7, 8, 9, 10 or more recesses.

The invention further provides a kit comprising a mold of the invention and neural induction medium. Both, the mold/car- rier and the neural induction medium are described above and the same applies to these components of the kit. The kit may further comprise instruction for use, e.g. to perform any method of the invention .

The following numbered embodiments are preferred according to the invention. Any of the numbered embodiments can of course be combined with embodiments and preferred options as described above, or with the corresponding elements of the example section .

1. An in vitro method of producing a fused tissue culture with at least three different tissues comprising selecting a spatial shape for attaching the at least three different tissues to each other, placing the at least three different tissues into contact in said spatial shape, culturing the at least three different tissues under conditions that maintains tissue fusion of the at least three different tissues .

2. The method of 1, wherein at least one, preferably at least two, of the at least three different tissues are in contact with only one of the at least three different tissues.

3. The method of 1 or 2, wherein the spatial shape is a planar or linear shape or the spatial shape follows the curvature of a carrier on which the at least three different tissues are placed .

4. The method of any one of 1 to 3, wherein the at least three different tissues are placed onto a carrier, wherein the carrier has a slope gradually descending to a deepest depression, wherein the at least three different tissues sink to the bottom of the carrier and then are pushed together due to a downslope force caused by an inclination of the slope.

5. The method of 4, with the slope being lower inclined closer to the deepest depression than further away from the deepest depression of the carrier

6. The method of 4 or 5, wherein the slope gradually descending to a deepest depression is a curved slope.

7. The method of any one of 1 to 6, comprising placing the at least three different tissues in a recess of a carrier, wherein said recess is elongated and comprises a deepest depression and at least one slope, the at least one slope descending to the deepest depression, wherein the placement of the at least three different tissues into the recess forces contacts of each of the at least three different tissues with at least one other of the at least three different tissues through gravity in the recess.

8. The method of 7, wherein the recess has a length of 1 mm to 40 mm and/or a width of 0.5 mm to 16 mm.

9. The method of 7 or 8, wherein the recess has a depression of at least 0.2 mm over a length of at least 0.5 mm and/or a depression of at least 0.2 mm over a width of at least 0.3 mm.

10. The method of any one of 4 to 9, wherein in a cross-sectional view of the carrier the at least one slope comprises a region with a length of at least 2 mm in which region all tangents to the slope enclose an angle between 1° and 20° with a top surface of the carrier.

11. The method of any one of 4 to 10, wherein the carrier at least at the deepest depression has a low- or non-cell-adhesion surface .

12. The method of any one of 1 to 11, wherein a contact point between two of the at least three different tissues is higher than a level of medium fluid.

13. The method of any one of 1 to 12, wherein the at least three different tissues comprise 1, 2 or 3 different brain or neural tissues or neuronal tissues.

14. The method of any one of 1 to 13, wherein at least one of the different tissues contains dopaminergic neural or neuronal tissue or a serotonergic neural or neuronal tissue.

15. The method of any one of 1 to 14, wherein the at least three different tissues are tissue spheroids or organoids or biopsy tissues or 3D cell aggregates; and/or wherein the at least three different tissues are 0.1 mm to 10 mm in size in their longest dimension . 16. An in vitro method of generating a ventral hindbrain tissue culture, comprising inducing neural differentiation in pluripotent cells or multipotent neural stem cells with a neural induction medium comprising at least one TGF-beta inhibitor and a Wnt activator, and allowing the cells during said induction to form a cell aggregate; culturing the cell aggregate to form a tissue culture.

17. The method of 16, wherein a) the at least one TGF-beta inhibitor is at least one SMAD inhibitor, preferably DMH1 (dorso- morphin homolog 1) ) and/or SB431542 (4- [4- (2H-1, 3-Benzodioxol-5- yl) -5- (pyridin-2-yl ) -lH-imidazol-2-yl] benzamide) ; and/or b) the Wnt activator is a WNT ligand, such as WNT-3a, or CHIR99021 (6- [ [2- [ [4- (2 , 4 -di chlorophenyl ) -5- ( 5-methyl-lH-imid- azol-2-yl) -2-pyrimidinyl ] amino] ethyl] amino] -3-pyridinecarboni- trile) .

18. The method of 16 or 17, wherein culturing the cell aggregate comprises treating the cell aggregate with a FGF (fibroblast growth factor) , preferably FGF4.

19. An in vitro method of generating a ventral midbrain tissue culture, comprising inducing neural differentiation in pluripotent cells or multipotent neural stem cells with a neural induction medium comprising at least one TGF-beta inhibitor, preferably Noggin and/or SB431542, and a Wnt activator, preferably CHIR99021, and a ROCK inhibitor, preferably Y-27632, and allowing the cells during said induction to form a cell aggregate; culturing the cell aggregate to form a tissue culture, preferably wherein culturing the cell aggregate comprises treating the cell aggregate with a FGF (fibroblast growth factor) , preferably FGF8 and/or an activator of sonic hedgehog signaling, preferably SAG, especially preferred SAG at a concentration of 10 nM to 100 pM, preferably 200 nM to 1000 nM; especially preferred wherein the cell aggregate is treated with dissolved extracellular matrix or an extracellular matrix component selected from collagen and/or laminin; even more preferred wherein the cell aggregate is not embedded in a extracellular matrix in a solid or gel state .

20. An in vitro method of generating a striatal tissue culture, comprising inducing neural differentiation in pluripotent cells or multipotent neural stem cells with a neural induction medium comprising at least an activator of sonic hedgehog signaling, preferably SAG, and a WNT pathway inhibitor, preferably IWP2, and ROCK inhibitor, preferably Y-27632, and allowing the cells during said induction to form a cell aggregate; culturing the cell aggregate to form a tissue culture; especially preferred wherein the cell aggregate is treated with dissolved extracellular matrix or an extracellular matrix component selected from collagen and/or laminin; even more preferred wherein the cell aggregate is not embedded in an extracellular matrix in a solid or gel state; and/or further especially preferred wherein the cells or cell aggregate is not treated with a gamma-secretase inhibitor such as DART.

21. The method of any one of 16 to 20, wherein the pluripotent cells or multipotent neural stem cells during aggregate formation are cultured on a low-cell-adhesion surface, and/or as a 3D culture.

22. The method of any one of 16 to 21, wherein the pluripotent cells or multipotent neural stem cells are treated with an activator of sonic hedgehog signaling; preferably SAG (3-Chloro- -

[ trans- 4- (methyl amino ) cyclohexyl ] -N- [ [ 3- ( 4-pyridinyl ) phenyl ] methyl ] benzo [b thiophene-2-carboxamide ) , retinolic acid and/or purmorphamine .

23. The method of any one of 16 to 22, wherein the neural induction medium comprises inorganic salts, glutamine, non-essential amino acids, heparin, vitamins, or any combination thereof.

24. A tissue culture comprising a planar or linear fusion of at least three different neural or neuronal tissues.

25. The tissue culture of 24, comprising neural cells with a cell body in one of said at least three different neural or neuronal tissues and an axon to another one of said at least three different neural or neuronal tissues, preferably wherein a tissue comprises neural cells that project their axons into at least two of the other neural or neuronal tissues.

26. The tissue culture of 24 or 25, wherein the at least three different neural or neuronal tissues are selected from central nervous system tissue, preferably of a telencephalic tissue, in particular preferred of cortical, striatal, lateral ganglionic eminence (LGE) , medial ganglionic eminence (MGE) or caudal ganglionic eminence (CGE) tissue ; preferably of a diencephalic tissue, especially preferred thalamic, hypothalamic, epithalamic, subthalamic or optic cup tissue ; preferably mesencephalic tissue, especially preferred tectum, tegmentum, ventral tegmental area (VTA) , substantia nigra (SN) , such as SN pars compacta (SNpc) , Nucleus ruber tissue; preferably rhombencephalon tissue, especially preferred cerebellum, medulla, pons, raphe nuclei tissue; preferably spinal cord tissue; or peripheral nervous system tissue.

27. The tissue culture of any one of 24 to 26, comprising cortical tissue; dorsoventral forebrain tissue, preferably lateral ganglionic eminence (LGE) tissue, medial ganglionic eminence (MGE) tissue or caudal ganglionic eminence (CGE) tissue; olfactory tissue, preferably comprising olfactory bulb neurons, amygdala tissue, cortex tissue or hippocampus tissue; limbic system tissue, preferably thalamus, hippocampus, amygdala, hypothalamus; dopaminergic system tissue, preferably ventral tegmental area (VTA) , substantia nigra (SN) , such as SN pars compacta (SNpc) tissue, nucleus accumbens, olfactory tubercle, prefrontal cortex, cortical, caudate nucleus, putamen, hypothalamus, brainstem, spinal cord tissue; serotonergic system tissue, preferably rostral serotonergic group tissue, such as cortex, striatum, amygdala, substantia nigra, pons, hippocampus, entorhinal cortex, locus coeruleus or caudal serotonergic group tissue, such as trigeminal motor nucleus, dorsal nucleus of vagus nerve, in- termediolateral nucleus, medulla, mesencephalon; cerebellar pathway tissue, preferably cortical, thalamic, pons, vestibular nuclei tissue; noradrenergic pathway tissue, preferably cerebellum, spinal cord, hippocamus, basal ganglia, cortex tissue .

28. The tissue culture of any one of 24 to 27, wherein the at least one of the neural or neuronal tissues is selected from ventral hindbrain, ventral midbrain, striatum, cortex.

29. The tissue culture of any one of 24 to 28 having a size of 100 pm to 100 mm, preferably 300 pm to 30 mm, in its longest dimension .

30. The tissue culture of any one of 24 to 29, comprising ventral midbrain tissue in contact with striatum tissue and said striatum tissue being in contact with cortex tissue. 31 . The tissue culture of any one of 24 to 30 comprising a neural cell from one of the at least three di f ferent neural or neuronal tissues having migrated to at least one of the other at least three di f ferent neural or neuronal tissues .

32 . The tissue culture of any one of 24 to 31 , wherein the at least three di f ferent neural or neuronal tissues comprise a neural or neuronal tissue of a neural or neuronal tissue type and another neural or neuronal or neuronal tissue of said neural or neuronal tissue type with its cells being modi fied by a genetic mutation .

33 . The tissue culture of any one of 24 to 32 , containing dopaminergic neurons and/or serotonergic neurons .

34 . The tissue culture of 33 , wherein the dopaminergic neurons connect between ventral midbrain tissue and striatum tissue and/or connect between ventral midbrain tissue and cortex tissue .

35 . The tissue culture of any one of 24 to 34 comprising TH+ cells , preferably wherein said TH+ cells have axons in a striatal and/or cortical tissue of the tissue culture .

36 . A method of investigating neuron formation, comprising introducing a neural stem cell , neural progenitor cell , neural cell , glial cell , astrocyte or oligodendrocyte into a tissue culture of any one of 24 to 35 and monitoring cell di f ferentiation and/or cell growth .

37 . A method of testing or screening a candidate compound for influencing neural cells of a fused tissue culture , comprising contacting cells or a tissue in a method of any one of 1 to 23 with the candidate compound or contacting a tissue of a culture of any one of 24 to 35 with the candidate compound and maintaining said contacted tissue in culture , and observing any change in the neural cell as compared to said tissue without contacting by said candidate compound .

38 . A mold for tissue fusion comprising a top surface and a recess in said top surface , wherein said recess is elongated and comprises a deepest depression and at least one slope along the direction of the elongation, the at least one slope descending to the deepest depression, wherein the mold at least at the deepest depression comprises a low- or non-cell-adhesion surface .

39 . The mold of 36 wherein the low- or non-cell-adhesion surface is of a coating with a low- or non-cell-adhesion material.

40. A kit comprising a mold of 38 or 39 and neural induction medium.

Throughout the present disclosure, the articles "a", "an", and "the" are used herein to refer to one or to more than one (i.e., to at least one) of the grammatical object of the article .

As used herein, words of approximation such as, without limitation, "about", "substantial" or "substantially" refer to a condition that when so modified is understood to not necessarily be absolute or perfect but would be considered close enough to those of ordinary skill in the art to warrant designating the condition as being present. The extent to which the description may vary will depend on how great a change can be instituted and still have one of ordinary skill in the art recognize the modified feature as still having the required characteristics and capabilities of the unmodified feature. In general, but subject to the preceding discussion, a numerical value herein that is modified by a word of approximation such as "about" may vary from the stated value by e.g. ±10%.

As used herein, the words "comprising" (and any form of comprising, such as "comprise" and "comprises") , "having" (and any form of having, such as "have" and "has") , "including" (and any form of including, such as "includes" and "include") or "containing" (and any form of containing, such as "contains" and "contain") are inclusive or open-ended and do not exclude additional, unrecited elements or method steps. The "comprising" expressions when used on an element in combination with a numerical range of a certain value of that element means that the element is limited to that range while "comprising" still relates to the optional presence of other elements. E.g. the element with a range may be subject to an implicit proviso excluding the presence of that element in an amount outside of that range. As used herein, the phrase "consisting essentially of" requires the specified integer (s) or steps as well as those that do not materially affect the character or function of the claimed invention. As used herein, the closed term "consisting" is used to indicate the presence of the recited elements only.

The present invention is further illustrated by the following examples, without being limited to these embodiments of the invention .

Figures

Figure 1 : Patterning schematic

Figure 1A: A patterning schematic for striatal , ventral Midbrain and ventral Hindbrain organoids .

EB formation is performed in media which is permissive or promotes di f ferentiation into ectoderm and neuroectoderm . Such media can be neural induction media (NI ) .

Both Wnt inhibition and low sonic hedgehog activation at the right time of the protocol were performed to achieve striatal organoids . For ventral midbrain and ventral hindbrain organoids , dual SMAD inhibition is performed by the addition of two TGF- p inhibitors in the first days of the protocol . Such SMAD inhibitors might be Noggin and SB431542 . Additionally, Wnt activation is performed . Wnt activation can be performed using the small molecule CHIR99021 . After di f ferentiation into neural lineages is started, sonic hedgehog signaling is activated to introduce floor plate , which, together with the right Wnt activation, gives rise to ventral midbrain floor plate progenitors , or at higher posteriori zation, hindbrain serotonergic neurons . Sonic hedgehog signaling might be activated by the small molecule SAG ( smoothened Agonist ) . A region-speci fic FGF such as FGF8 might be applied for increasing the yield of dopaminergic neurons , as has been demonstrated before . A region-speci fic FGF such as FGF4 might be applied for increasing the yield of serotonergic neurons , as has been demonstrated before .

Figure IB : Overview of the position of the regions relevant to the dopaminergic system, namely the Cortex, LGE ( Lateral Ganglionic Eminence , which gives rise to striatal proj ection neurons ) and ventral Midbrain, giving rise to mDA (midbrain dopaminergic neurons ) . Figure IB also provides an overview of the patterning axes and molecules used in the generation of striatal and ventral midbrain organoids .

Figure 2 : Generation of ventral midbrain organoids

Figure 2A-D : qPCR analysis of markers of ventral midbrain of the dose curving experiment to generate ventral midbrain organoids . Two separate dose curving experiments were performed, where first the correct dosage of sonic hedgehog activation was tested ( Figure 2A, floor plate marker F0XA2 ) . Secondly, we performed qPCR to test for appropriate concentrations of posteriori zation using Wnt activation . Dopaminergic neurons are born in the very ventral floor plate of the posterior midbrain, next to the midbrain-hindbrain boundary, thus the Forebrain and Midbrain marker OTX2 should still be expressed, whereas the hindbrain is OTX2 negative . We found a concentration of 0 . 8- 1 pM suf ficient to introduce midbrain ( Figure 2B ) , while the floor plate marker FOXA2 was still expressed ( Figure 2C ) . qPCR for the dopaminergic marker TH ( Tyrosine hydroxylase , the rate-limiting enzyme for dopamine synthesis ) confirmed this concentration of Wnt activation, with a sharp drop of TH expression starting at 1 . 2 pM CHIR ( Figure 2D) .

All qPCR data are relative to the expression of TBP .

Figure 2E-G : Additional stainings by Immunohistochemistry ( IHC ) confirmed the expression of FOXA2 and OTX2 in 20-day old organoids . Already on day 20 , the first expression of TH ( Tyrosine Hydroxylase , the rate-limiting step for dopamine synthesis and a marker for dopaminergic neurons ) can be noticed, and TH positive cells can be broadly found on day 40 .

Figure 3 : Further ventral midbrain data

Panel A: Organoid formation of Cortical , Striatal and ventral Midbrain organoids at day 10 . All protocols result in homogeneously si zed organoids in H9 embryonic stem cells .

Panel B : While striatal and cortical organoids express the forebrain marker FOXG1 , correctly patterned ventral midbrain organoids do not express FOXG1 .

Panel C : The ventral Midbrain protocol has been tested with one iPSC and two H9 embryonic stem cell derived cell lines .

Panel D : FOXA2 staining of day 20 old ventral midbrain patterned organoids of multiple cell lines shows robust generation of floor plate patterned tissue .

Panel E : Staining for OTX2 and TH indicates robust production of TH+ dopaminergic neurons from multiple cell lines .

Figure 4 : Generation of striatal organoids

Panel A: Schematic of dorsoventral patterning gradients important for striatal neurogenesis ( the caudali zing and dorsali z- ing factor Wnt , and the ventrali zing factor SHH) and markers expressed during neurogenesis of striatal neurons from the LGE (lateral ganglionic eminence) to young bourn neurons to mature striatal neurons.

Panel B: Wnt inhibition together with a dose curve of sonic hedgehog activation with SAG demonstrates that low concentrations of SHH activation were sufficient to induce GSX2+ PAX6~ LGE neural stem cells in 28 day old organoids.

Panel C, D: Quantifications of GSX2+ rosettes and PAX6~ rosettes, indicating LGE identity.

Panel E-F: Immunolabeling of CTIP2, an important early striatal marker, in two conditions with promising numbers of GSX2 + rosettes, but low PAX6 expression. Notably, CTIP2 expression in 25 nM SAG treated organoids is much lower than in SAG10 treated organoids (Panel F) .

Panel G-J: Further marker characterization of day 28 striatal organoids with 10 nM SAG. IWP2-SAG10 treated organoids displayed GSX2+ rosettes which produce DLX5+ young bourn LGE-derived neurons (Panel G' ) . Further LGE or young striatal neuron markers such as ASCL1 (Panel H) , CTIP2 (Panel I) and ISLET (Panel J) are readily expressed in SOX2+ neural rosettes.

Panel K: Striatal neurons express DARPP32, a marker for dopa- mine-sensitive neurons. DARPP32 can be broadly found in day 60 old LGE-patterned organoids and form DARPP32+ clusters (white arrows) .

Panel L: We could surprisingly find different expression levels of cells in DARPP32 clusters versus DARPP32+ cells which are scattered through the organoid, which could indicate immature and mature striatal neurons.

Panel M: Staining for FOXP1, another important striatal marker, together with DARPP32.

Panel N: Staining for GAD1 and DARPP32 in 80-day old striatal organoids. Striatal neurons are majorly GABAergic, and DARPP32+ neurons indeed expressed the marker GAD1 (rate-limiting enzyme for GABA synthesis) .

Panel 0: Quantification of DARPP32+ neurons being positive for FOXP1 and CTIP2.

Panel P: Quantification of DARPP32+ neurons being GABAergic (GAD1 + ) .

Figure 5: Further striatum data Panel A: patterning protocol for the generation of striatal organoids

Panel B : The striatal protocol has been tested with multiple cell lines , example EBs of two H9 derived cell lines and 176- 1 . Panel C : GSX2 staining of day 33 striatal patterned organoids of 7 cell lines ( 5 iPSC lines , 2 H9 derived embryonic stem cell lines ) .

Panel D : 6 out of 7 cell lines displayed GSX2 + rosettes on day 33 , although with variation in the total percentage of GSX2 + rosettes .

Panel E : Striatal patterned organoids do not express the cortical marker TBR1 at day 80 , with few exceptions (white arrows ) . This is of particular importance , as many striatal markers ( DARPP32 , CTIP2 ) are also expressed in cortical neurons .

Panel F : Striatal patterned organoids express the striatal marker DARPP32 at day 80 (white arrows ) .

Panel G : The MGE and interneuron marker Nkx2 . 1 is expressed scattered throughout striatal organoids at day 60 , indicating some tissue in striatal organoids might be of MGE origin .

Figure 6 : Linear organoid fusions

Panel A: Generation of fusion molds for fusions of 3 or more organoid regions in a spatially defined manner . Embedding molds were designed using Tinkercad . First , a negative is designed and then sliced using XYZprint® to prepare for 3D printing . The negative was printed using the XYZprinting printer da Vinci Color in full filling mode and high resolution .

Subsequently, the negative was surface treated with a heat gun to melt the surface and create a smooth finish . The positive was then cast using the silicone PDMS ( Polydimethylsiloxane ) .

Prior to usage , the positive mold was washed with 70% Ethanol and coated with an anti-adherence solution to further increase the nonstick characteristics of PDMS .

Panel B : For spatially defined/ linear fusions , organoids were linearly put into the molds . Media is then removed and the organoids were attached for 30sec - Imin . Subsequently, a small volume of Matrigel is added to the organoids . After solidi fication of the Matrigel at 37 ° C, the organoids can be washed out carefully with Improved+A media . For the first 2 days , organoids were cultured without shaking, after 2 days the organoids can be transferred back on an orbital shaker. Panel B' : a plate with triple fusions after transfer. B' ’ : magnified view on linearly fused organoids. B' ’ ’ : a linear fusion with Cortex (unlabeled) : : Striatum (CAG-tdtomato) : : ventral Midbrain (CAG-GFP) . Fluorophore channels were added to the red or green channel respectively. Panel C: Triple fusions on the day of fusion, 12 days after fusion and 88 days after fusion. The three different tissues start to grow together and form a homogeneous tissue.

Panel D: Triple fusions where the ventral midbrain is labelled with GFP (H9 CAG-GFP) . Only 12 days after fusion, first axon bundles emerge and project into striatal and cortical tissue. Panel E: Triple fusions on day 109, the striatal and cortical tissue is highly innervated.

Panel F: GFP labeled projections from the ventral midbrain broadly express TH in striatal and cortical tissues (quantifications in Panel G) .

Panel H: TH+ projections highly innervate TBR1+ cortical regions and DARPP32+ striatal regions.

I: In CtxcAG-GF? : : Striatum: : vMidcAG-tdtomato fusions, the striatum is highly innervated in triple fusions by cortical and ventral midbrain, indicating the formation of complex neuronal circuits.

Figure 7 : Embedding mold design

Panel A: Tilted side view of the PDMS embedding mold.

Panel B: Top view of the PDMS embedding mold.

Panel C: Side view (side A) .

Panel D: Side view (Side B) .

Panel E: Design of individual molds. Focus has been put on a low curvature at the side where organoids are positioned, allowing organoids to be fused in a straight line, but tilted sides, which push organoids together. Examples of triple and quadruple fusions are shown.

Panel F: Side view.

Figure 8 : scRNAseq

A, scRNAseq of day 60 fused organoids (in this case of MISCOs) shows all major populations of the dopaminergic circuit (mDA neurons, striatal neurons, cortical neurons) were present, together with clusters of cortical and LGE progenitors, MGE and CGE derived cells and VM GABAergic and glutamatergic neurons (n=3 pooled fused organoids) . B, Correlation of the cortical excitatory neuron and striatal neuron cluster with the BrainSpan dataset of the developing human brain (PCW20-25) . STR...Striatum, NCx... Neocortex, HIP... Hippocampus , DTH... Dorsal Thalamus, CB... Cerebellum, AMY...Amygdala . C, Voxhunt spatial similarity mapping of the cortical excitatory neuron, Striatal neuron and mDA neuron clusters onto E13.5 Allen Developing Mouse Brain Atlas data with sections colored by scaled expression similarity scores. D, Voxhunt spatial similarity mapping on dopaminergic clusters of E13.5 sagittal mouse brains colored by scaled expression similarity scores.

E, Density plots of floor plate (F0XA2, SHH) and dopaminergic neuron (EN1, TH, DAT, PITX3} as well as a TH, EN1 and F0XA2 joint density plot indicating the cluster of mDA neurons. F, The striatal neuron cluster is GABAergic (VGAT, GAD1) and expresses the striatal markers CTIP2, F0XP1, MEIS2 and ZFHX3. The joint density plot of ZFHX3, GAD1 and F0X1 indicates the cluster of striatal neuron identity. F' -F' ' , The striatal cluster expresses markers of DRD1 medium spiny neurons (F' , TAC1, ISL1, EBF1) as well as DRD2 medium spiny neurons (F' ' , SIX3, SP9, GRIK3) .

G, Density plots for cortical progenitor markers (PAX6, EMX2) , the intermediate progenitor marker EOMES and the cortical neuronal markers NEUR0D6, VGLUT1 and TBR1. H, Forebrain (cortical and GE patterned) tissues expressed the forebrain marker F0XG1. Clusters for LGE progenitors (GSX2) as well as MGE (LHX6') and CGE-derived (NR2F2) cells were observable. Additionally, clusters with the identities of oligodendrocyte (-progenitors) (OLIGl) , dividing progenitors (mKI67) and glial cells (S100B) clusters were present. I, Dot plot of top marker genes expressed in at least 20% of cells in individual clusters ranked by p value, identified by Wilcoxon rank sum test.

Figure 9: Tissue clearing and morphological reconstruction

Panel A: Whole-mount 2ECi tissue clearing and z projection of a triple fusion with immunostaining for TH. Left side: low intensity demonstrates TH+ clusters in the ventral midbrain of the fusion. Right side: High intensity demonstrates innervation of striatal and cortical tissue.

Panel B: Whole-mount 2ECi tissue clearing reveals TH+ axon bundles projecting from TH clusters into striatal and cortical tissues .

Panel C: high magnification recording of a TH+ axon bundle originating from the ventral midbrain side and projecting to the striatum.

Panel D: 2ECi tissue clearing with immuno labeling for TH+ axons in the striatum from day 40 to day 120. Over time, more putative axonal boutons appear on TH+ axons .

Panel E: Quantification of putative axonal boutons per 10pm of axon length. Over time, the amount of putative axonal boutons significantly increases in both striatal and cortical tissues. Panel F: TH+ axons not only innervate neuronal tissues but avoid neural progenitor regions such as the subventricular zone (SVZ) . L=Lumen .

Panel G: Dopaminergic neurons often, but not always, organize in clusters which can display a multitude of different morphologies.

Panel H: TH+ neurons can also be found in the striatum and cortex, however they can be morphologically distinguished from ventral midbrain dopaminergic neurons. Multiple reports of TH positive interneurons in the human cortex exist, thus we grew fusions using a DLX5-GFP cell line, which labels forebrain interneurons. We could confirm that TH+ cells in striatum and cortex were GFP+ and thus TH+ interneurons .

Panel I: various morphologies of dopaminergic neurons could be observed such as multipolar (images a, b) and bipolar (c, d) with variations in cellular morphology and cell body size.

Panel J-L: 2D Immunostaining for TH and DAT (Dopamine transporter) in ventral midbrain (J) , Striatum (K) and Cortex (L) . Double positive cells (J) are highlighted with a white arrow, dopaminergic axons (K and L) which express DAT are highlighted with yellow arrows. DAT is vital for dopamine transport and indicates that dopaminergic neurons matured to a functional level, releasing and re-uptaking dopamine from its synapses.

Figure 10: tissue clearing and morphological reconstruction

Panel A: A triple fusion where the striatum is labeled with CAG- GEP. GFP+ projections in 2ECi cleared organoids partially were positive for GABA (white arrows) , indicating striatal GABAergic long-range connections. Notably, cortical regions often had GFP GABA double positive cells (yellow arrows) and axons originating from these cells, which are absent in ventral midbrain tissues. Panel B: 2D immunohistochemistry characterization of a representative ventral hindbrain organoid with serotonergic neurons. Serotonergic neurons express TPH2 and produce Serotonin (5-HT) . Both markers are broadly expressed in ventral hindbrain organoids .

Panel C: 2ECi cleared quadruple fusion of Ctx : : Str : : ventral midbrain : : ventral hindbrain, with the ventral hindbrain being labeled constitutively with CAG-GFP. Three different intensity levels are displayed, to highlight various features of innervation. Notably, ventral midbrain received substantially more innervation from the ventral hindbrain than striatum and cortex, but all three target regions were densely innervated.

Panel C' : Immunolabeling of TPH displays serotonergic neurons in the ventral hindbrain part of the quadruple fusion.

Panel D: Same organoid as in Panel C, immunostaining for TH. Notably, the ventral midbrain shows the highest signal of TH, with innervation of the ventral hindbrain, striatum and cortex.

Panel E: Live imaging of 150-day old organoids infected with a Synapsin-Archl-GFP AAV. The fusion protein is membrane-bound and allows reconstruction of neuronal morphologies. We could observe various morphologies in the different regions of the fusion. Panel F: High magnification live imaging of a dendritic tree of a cortical neuron with readily observable dendritic spines.

Figure 11: neural activity

Panel A: Organoids were grown with an H9 Syn-GCAMP cell line in either the cortical, striatal or ventral midbrain region. Recordings were performed for 6.5 min continuously with a frame every 65 ms. Different modalities in activity could be observed in all three regions, with the cortex often displaying synchronous network events, whereas the striatum displayed asynchronous activity, but with often longer calcium events than the other two regions. Ventral midbrain recordings displayed events of highly synchronous activity, lasting for seconds to more than a minute- in this example at the end of the recording.

Panel B: Example traces from A indicating different activity modalities in all three regions.

Panel C: A recording in a striatal organoid with a Ctx _WT : : Str _WT : : ventral Midbrain _GC MP fusion. Network events from axons from the ventral midbrain could be observed ( fluorescence change in C' versus C' ' ) . No GCAMP+ cell bodies were visible in the area of recording .

Panel D : Cumulative fluorescence over time from the recording of Panel C . Notably, no detectable synchronous activity occurs in the first minute of the recording, in the second minute synchronous events can be observed, but not in the axons in focal plane of the recording, and in the third minute the axons coming from ventral midbrain contribute to a high- frequency synchronous network event which lasts until the end of the recording .

Panel E : To confirm i f ventral midbrain axons can functionally interact with neurons in the striatum and cortex, organoids were infected with an AAV containing CAG-hChr2-H134R-tdtomato .

We then used a silicon probe setup ( left side of Panel E ) and optogenetic activation of the midbrain ( right two images of Panel E ) to activate neurons in the ventral midbrain and recorded from either the striatum or cortex .

Panel F : Stimulation setup . l Omin of baseline in either the striatal or cortical part was recorded . Then, for 10 minutes , the ventral midbrain side was stimulated every 10 seconds for 500ms .

Panel G : example trace of a striatal recording without (blue ) and with a 90-250 Hz bandpass filter (black) and the RMS ( red) . Stimulation started at 615 sec . This recording indicates that neurons from the ventral midbrain make functional connections with neurons in the striatum and cortex . 3/ 3 recordings in striatum and 3/ 3 recordings in the cortex displayed similar behavior .

Panel H : Analysis of an individual burst (wideband/bandpass 90- 250 Hz , RMS ) . Note that the burst duration corresponds to the duration of ventral midbrain stimulation ( 500 ms ) .

Panel I : Average burst duration of the recording, corresponding to the 500 ms stimulation setup .

Panel J : Optogenetic setup for measuring dopamine release . Organoids were co-infected with an AAV containing Syn-ChrimsonR- tdTomato and an AAV containing Syn-GRAB-DA4 . 4 . Thus , stimulation of ventral midbrain dopaminergic neurons should induce dopamine release , which can be measured with GRAB-DA4 . 4 , a fluorescent dopamine sensor .

Panel K : cumulative fluorescence without stimulation in green and cumulative fluorescence with stimulation subtracted from cumulative fluorescence without stimulation in red. Notably, a change in fluorescence was majorly observed on the membrane of cells and neurites (white arrows) .

Panel L: Fluorescence change over time of neurons recorded in the striatum. Notably, upon stimulation several regions displayed an up to 10% increase in fluorescence.

Figure 12: mDA injections

Panel A: Schematic of the injection setup. Midbrain dopaminergic neuron progenitors (day 16) with endogeneous Cre in the TH locus and infected with a f lox-stop-f lox GFP lentivirus were grafted into the ventral midbrain side of a triple fusion using a microinjector. 40000 cells in roughly 200 nl were injected in organoids between 60 and 150 days old. Injection of one organoid takes approximately 1-2 minutes.

Panel B: A triple fusion with grafted mDA one month after injection. The graft is easily visible in the ventral midbrain.

Panel C: Surface recording of an organoid with 30-day old graft. Notably, the ventral midbrain region is highly innervated, but striatum and cortex also receive significant levels of innervation .

Panel D: dopaminergic bundle formation from grafted neurons, projecting into the striatal tissue.

Panel E: Live-imaging experiment which allows the movement of axonal structures to be visualized, which could be indicative of axonal transport.

Panel F: Live imaging of two dopaminergic axonal growth cones. Notably, one axonal growth cone extends a filopodium or lamel- lipodium in the direction of the other dopaminergic axon and starts interacting (green arrow) . After the left axon continues projecting, a bleb remains at the site of interaction (yellow arrow) . This happens a second time with another axon (red arrow) . The middle axon does not project further during and after interaction and a bleb is formed at its end.

Panel G: 2ECi recording of an injected graft, immunolabeled for FOXA2 and TH. The grafts form fiber bundles and start projecting .

Figure 13: Bulk RNAseq A, PCA of Bulk RNAseq of individual VM, striatal and cortical , patterned organoids on day 60 (n=3-4 organoids per group ) . B , Genes with highest loading on PCI and PC2 . C , Voxhunt spatial similarity mapping of bulk RNAseq data of VM, striatal and cortical organoids to E13 . 5 Allen Developing Mouse Brain Atlas data with sections colored by scaled expression similarity scores . Data show mean ±SD . D , Striatal organoids do not have PAX6 ⁺ regions and are negative for the cortical neuron marker TBR1 , unlike cortical organoids .

Figure 14 : MISCOs display heterogeneous dopaminergic subtypes A DARPP32 ⁺ neurons in 120 day old organoids in the striatum expressed the striatal subtype markers DRD1 and DRD2 ( representative images , n=6- 8 organoids of 2-3 batches ) B-C Dopaminergic neurons in MISCOs expressed the dopaminergic subtype markers GIRK2, CALB1 , ALDH1A1 , 0TX2 , GABA and SOX6 (n=5-7 organoids of 3-4 batches ) . D Quanti fication of fusion ef ficiency of 5 batches and a total of 365 fusions . In average , 96% ( ±2 . 4 % SD) of fusions remained intact after the fusion procedure across multiple batches .

Figure 15 : Organoid fusions form structurally mature neural circuits

A, Striatal tissue displayed stronger innervation than cortical tissue . B , Quanti fication of the peak fluorescence of 2ECi cleared organoid recordings in striatal and cortical tissue (n=5 organoids , p=0 . 0125 ) . C , Day 60 VM _WT -Str _C AG-GFP ^_ Ctx _WT fused organoids allow to study striatal innervation and show reciprocal GABA+ innervation from striatal into VM I and cortical (C' ) tissue (white arrow) , indicating reciprocal connectivity between VM and striatal tissue . Additionally, migrated interneurons into cortical tissues could be observed ( yellow arrows ) . N=8 / 8 60 day old fused organoids . D , Day 60 VM _WT - Str _WT -Ctx _C AG-GFP fused organoids display innervation from cortical tissue into both VM (D ) as well as striatal (D' ) , displaying cortical proj ection neuron axon bundles (white arrows in D' ) in the striatum as well as axonal innervation of VM tissues with mDA neuron clusters (n=6/ 6 organoids ) . E , Striatal DARPP32 ⁺ clusters got readily innervated from cortical and VM tissues in day 40 VM _t dtomato -Str _WT -Ctx _GFP fused organoids ( 6 of 6 organoids ) . Figure 16 : Fused organoids display neuronal activity and functional dopaminergic connectivity

A, Schematic illustrating optogenetic stimulation of VM tissue and extracellular recording for the striatal tissue . Organoids were transduced with the optogenetic construct AAV-RG . AAV-CAG- hChRlHl SiR-tdtomato . 140- 171 day old fused organoids were stimulated with a glass fiber focusing the light on VM tissue . B , Representative active channel raster plots in striatal tissue ± 100 seconds from the initiation of optogenetic stimulation ( orange box ) . 460-nm LED light pulses set at an interval of 10 seconds with a 500-millisecond duration (blue ) . A 20-second interval of the recordings is shown to highlight optogenetic neural population responses ( right ) . C , Normali zed firing rate (Hz ) changes in striatal tissue across 10-minute baseline and 10-mi- nute optogenetic stimulation periods , calculated per organoid per active channels (n=7 independent experiments across two organoid batches , a total of 315 active channels , Wilcoxon signed- rank test , p=0 . 016 ) . D , Percentages of active electrodes which are responsive to optogenetic stimulation for VM and recording in striatal tissue ( 235/ 315 active responding channels from 7 independent experiments across two organoid batches ) . E , Schematic illustrating optogenetic stimulation of VM tissue and extracellular recording from the cortical tissue , following the same experimental setup as in (A) . F , Representative active channel raster plots in cortical tissue ± 100 seconds from the initiation of optogenetic stimulation ( orange box ) . G, Normali zed firing rate (Hz ) changes in cortical tissue across 10-mi- nute baseline and 10-minute optogenetic stimulation periods , calculated per organoid from active channels (n = 8 independent experiments across 3 organoid batches , a total of 391 active channels , Wilcoxon signed-rank test , p=0 . 039 ) . H , Percentage of responsive active electrodes to optogenetic stimulation for VM- cortical ( 292 / 391 active responding channels from eight independent experiments across three organoid batches ) recordings . Data shown as mean ± s . e . m . I , Schematic illustrating optogenetic stimulation of VM tissue and simultaneous fluorescent confocal recording of striatal and cortical tissue . Organoids were transduced with the optogenetic construct AAVl-pAAV-Syn- ChrimsonR-tdTomato and the fluorescent genetic encoded dopamine sensor GRAB-DA2m (AAV2-pAAVss-hSyn-GRAB-DA4 . 4 ) . J-K, Recordings of striatal and cortical tissue in 130 day old fused organoids show an increase in fluorescence of the dopaminergic sensor upon optogenetic stimulation of the VM . L-M, Timelapse recording of individual cortical and striatal regions in fused organoids demonstrates an increase of dopamine release upon stimulation of VM tissue in cortical and striatal tissues (N=3 cortical and 3 striatal recordings ) . Bottom : Average in blue with SD in gray . N-O , Heatmap of activity regulated response genes and dopamine signaling response genes in striatal (N) and cortical ( 0) tissue of fused organoids . The maj ority of activity regulated genes were upregulated in bulkRNAseq of striatal and cortical tissues of separated fused organoids on day 60 in comparison to day 60 individual organoids . PRGs... primary response genes . SRGs... secondary response genes . P , Representative active channel raster plots in forebrain tissue ± 100 seconds from the initiation of optogenetic stimulation in VM tissue . 460-nm LED light pulses set at an interval of 10 seconds with a 500-millisecond duration (blue ) . After a 5min incubation with a cocktail of synaptic blockers ( D-AP5 , CNQX, Gabazine , SCH-23390 , Sulpiride ^-7 - ) , population responses in forebrain were absent during stimulation of VM . Q , Normali zed mean firing rate in fused organoids before and after light application, and repeated with synaptic blocker applications (n= 127 channels of 2 organoids ) . p<0 . 0001 . Samples were analyzed with one-way ANOVA followed by Tukey' s multiple comparisons test .

Figure 17 : Cocaine treatment of fused organoids allows to study perturbations of the dopaminergic system in vitro

A, Schematic for the treatment of fused organoids with cocaine . Organoids were treated with 0 . 7pM cocaine hydrochloride for Ih every 3 days from day 40 until day 130 ( Chronic ) and until day 105 (Withdrawal condition) . Functional , morphological and transcriptional analysis was performed on day 130 . B , Representative images of dopaminergic ( TH ⁺ ) axons in striatal and cortical tissue in control , chronic and withdrawal fused organoids . C , Schematic for the extraction of varicosity density and varicosity diameter in dopaminergic axons . D , Quanti fications of TH ⁺ varicosity density in striatal and cortical tissue . N=6- 8 organoids per condition, 39-49 axons . E , Measurement of varicosity diameter (axon average) in striatal and cortical tissue. N= 6-8 organoids, 36-47 axons per condition and 1055 | 1552 | 1347 (striatal) and 923 | 1063 | 1340 (cortical) boutons measured. F, Schematic of the parameters' frequency and duration from GCAMP traces after extraction with CalmAn. G-H, fused organoids with GCAMP6S expression in either VM, striatal or cortical tissue were recorded on day 130 in control, chronic and withdrawal conditions (10-21 organoids per condition) . G, Analysis of calcium event duration showed a significant decrease in striatal and increase in cortical neuron calcium event duration. 1776 | 2144 | 4254 individual calcium events (striatal) and 1437 | 6931 | 8000 individual calcium events (cortical) . H, Analysis of calcium event frequency of striatal (n=248 | 245 | 680) and cortical ( 175 | 936 | 885) neurons in control, chronic and withdrawal conditions. I-J, Volcano plot comparing bulk RNAseq data of day 130 forebrain organoids chronic versus control (I) and withdrawal versus control (J) (n=6-9 organoids of 2-3 batches) . K, GO term overrepresentation analysis of genes downregulated in forebrain chronic and withdrawal versus forebrain control. Data show mean ±SD. Samples were analyzed with one-way ANOVA followed by unpaired Mann-Whitney test.

Figure 18: Cocaine treatment of Fused organoids

A, Representative 130 day old GCAMP recordings of fused organoids with Syn-GCAMP expression either in the cortical, or striatal, or VM tissue in control (left) , chronic (middle) and withdrawal (right) condition. B, Representative GCAMP-recording of a fused organoids with Syn-GCAMP in striatal tissue before (control) and after TTX application (0.5pM) , visualized as cumulative neuronal activity in a 6.5min recording and displayed as AF/F. C, Calcium event duration of VM neurons. 876 | 2058 | 1068 individual calcium events of 7-17 organoids. D, Analysis of calcium event frequency of VM (n=100 | 278 | 115) neurons in control, chronic and withdrawal conditions. Data show mean ±SD. Samples were analyzed with one-way ANOVA followed by unpaired Mann-Whitney test (C,D) . E-F, Volcano plots of differentially expressed genes in VM tissue in chronic versus control I and withdrawal versus control (F) . G-I, GSEA analysis of three GO terms associated with neural circuit formation (Dendrite Development, Axon Development and Synapse Organization) in forebrain chronic vs control (top) and forebrain withdrawal versus control (bottom) .

Figure 19: Alternative fusion conditions

A, Schematic for the generation of linear fusions in an optimized method using Matrigel (MG) . B, Changes for the generation of linear fusions without the addition of Matrigel between the transfer step and the step of fused organoids on a shaker as illustrated in A. Organoids are transferred into PDMS Anti-Adherence treated molds and media is gently added on top. Organoids are then incubated 1-3 days in an incubator, before the fusions are rinsed out and the PDMS embedding mold is removed. Fusions can then be continued to be cultured the same way as fusions with Matrigel. C, Fusions without Matrigel 1 day after fusion. D, Fusions without Matrigel 6 days after fusion.

Examples :

Example 1 : Material & Methods : Example 1.1: Stem cell culture:

The hESC lines (Hl (WaeOOl-A) , H7 (WA07) , H9 (WA09) , RC17 (Rce021-A) ) and in-House generated iPSC lines (176/1, 178/5, 178/6) were cultured feeder-free on growth factor-reduced Matrigel (Corning, Cat.# 354277) in mTESRl (Stemcell Technologies, Cat.# 85875) . Routine genome integrity tests were performed every 10 passages. Cells were split after 3-5 days with avoiding confluently coated wells, in a ratio of 1:6-1:10 using PBS ^_/_ °- ^{5mM EDTA} (PBS: Bio Trend PBS-1A, EDTA: Sigma-Aldrich, Cat.# E6758) and incubated in an incubator until colonies start to break apart (4-8min after incubation) .

Example 1.2: Media formulations: -Neural induction media

Media basis: DMEM/F12 (Invitrogen, Cat.# 11330-057) , 1% N2 Supplement (ThermoFisher, Cat.# 17502001) , 1% GlutaMAX-I (ThermoFisher, Cat.# 35050-038) , 1% MEM-NEAA (Sigma-Aldrich, M7145) , 1:1000 Heparin solution (Sigma-Aldrich, Cat.# H3149-100KU) , 1% PenStrep (Sigma-Aldrich, Cat.# P4333) -Improved-A media

Media Basis: 50:50 DMEM/F12 : Neurobasal (Gibco, Cat.# 21103049) . 0.5% N2 supplement, 2% B27-A (ThermoFisher, Cat.# 12587010) , 1:4000 Insulin (Sigma-Aldrich, 19278) , 1% GlutaMAX, 0.5% MEM- NEAA, 1% Antibiotic-Antimycotic (ThermoFisher, Cat . #15240062 ) -Improved +A media:

Media Basis: 50:50 DMEM/F12 : Neurobasal (Gibco, Cat.# 21103049) . 0.5% N2 Supplement, 2% B27+A (ThermoFisher, Cat.# 17504044) , 1:4000 Insulin (Sigma-Aldrich, 19278) , 1% GlutaMAX, 0.5% MEM- NEAA, 1% Antibiotic-Antimycotic (ThermoFisher, Cat . #15240062 ) , 1% Vitamin C solution (40mM stock in DMEM/F12) (Vitamin C: Sigma-Aldrich, Cat.# A4544) , 1 g/liter sodium bicarbonate (Sigma-Aldrich, Cat.# S5761) -Brainphys : Media Basis: BrainPhys Neuronal Medium (Stemcell Technologies, Cat. #05790) , 2% B27+A, 1% N2 Supplement, 1 ml CD Lipid Concentrate (Thermo Scientific, Cat . #11905031 ) , 1% Antibiotic-Antimycotic, 1:147 20% Glucose Solution, 20ng/ml BDNF (Stemcell Technologies, Cat . #78005.3) , 20ng/ml GDNF (Stemcell Technologies, Cat.# 78057.3) , 1 mM db-cAMP (Santa Cruz Biotechnology, Cat.# sc-201567C) .

Example 1.3: Viruses:

AAV8 pAAV-Syn-Archonl-KGC-GFP-ER2 (Addgene Cat.# 115892-AAV8)

AAV1 pAAV-Syn-ChrimsonR-tdT (Addgene Cat.# 59171-AAV1)

AAV-RG. AAV-CAG-hChR2-H134R-tdTomato (Addgene Cat.# 28017-AAVrg)

AAV2 (pAAVss_hsyn-GRAB-DA4.4) : 3.61E+13 VG/mL (gift from Boehringer Ingelheim)

Example 1.4: Antibodies :

* For 3D Immunohistochemistry, 5-10x this concentration was used (dependent on antibody quality and tissue size to be stained)

Secondary antibodies were the standard panel of Jackson ImmunoResearch and Invitrogen (Alexa Fluor 488, 568 and 647 polyclonal donkey antibodies raised against species of primary antibodies) with the option of affinity-purification where available.

Example 1.5: Generation of cortical organoids:

For cortical organoid preparation, cells were grown for 2-3 days after splitting in MG coated 6-well plates and dissociated using 600pl Accutase (Sigma-Aldrich, Cat.# A6964) . Cells were washed off with 1.4ml stem cell medium and spun down at 150g for 5min. Cells were resuspended in 1ml stem cell medium +1:100 RI Y27632 (Selleck Chemicals, Cat.# S1049) and counted. 9000 cells in 150ul of media +l:100RI per well were then transferred into a 96 well ultra-low attachment plate (Thermo Scientific, Cat.# 136101 and Corning, Cat. #7007) . After 3 days, lOOpl of media was removed and replaced with 150pl of fresh media. On days 5, 7 and 9, 150pl of media was exchanged with 150pl of neural induction media. On day 11, up to 30 Ebs were transferred into a 10cm plate which was coated with anti-adherence rinsing solution (Stemcell Technologies, Cat.# 07010) . Ebs were transferred into 10ml Improved-A media containing 2% liquid Matrigel (Corning, Cat.# 356235) in the media. Notably, media has to be fridge-cold when Matrigel is added and used the same day. Ideally, until Ebs are added the media does not reach temperatures above 10°C to prevent Matrigel from polymerizing. On day 12, 3pM CHIR99021 (Merck Millipore, Cat.# 361571) was added to the media. On day 13, media was exchanged with fresh media containing 3pM CHIR99021 and not changed for three days. Organoid media was then exchanged every 3 days, with a change to Improved+A media around day 16. Organoids were transferred to orbital shakers (Inforce Celltron HT, 42RPM) on day 20 and cultured in Improved+A media until day 60, where a gradual shift to Brainphys media (25%, 50%, 75%, 100%) is performed. As soon as organoids were transferred to orbital shakers, media volume was increased to 30ml.

Example 1.6: Generation of ventral midbrain organoids :

For ventral midbrain organoid preparation, cells were grown for 2-3 days after splitting in MG coated 6well plates and dissociated using 600pl Accutase. Cells were washed off with 1.4ml stem cell media and spun down at 150g for 5min. Cells were resuspended in 1ml stem cell media +1:100 RI and counted. 9000cells were then transferred into 150ul neural induction media containing 200ng/ml Noggin (R&D Systems, Cat.# 6057) , lOpM SB431542 (Stemgent, Cat.# 04-0010-10) and IpM CHIR99021 and ROCK inhibitor in an ultra-low attachment U-shaped 96well plate (Corning Cat#) . On day 2, lOOpl of this media was removed and replaced with 150pl fresh media of the same formulation.

On Day 4, 150pl of media was exchanged, with the addition of 200ng/ml Noggin, lOpM SB431542, IpM CHIR99021 and 300nM SAG (Merck, Cat.# US1566660) and lOOng/ml FGF8 (R&D Systems, Cat.# 5057-FF) to the newly added media. On day 6, 150pl of media was exchanged, containing 300pM SAG and lOOng/ml FGF8.

On day 8, organoids were transferred into 10cm plates coated with anti-adherence rinsing solution. Media was exchanged with 12ml Improved-A media containing 2% Matrigel (added to COLD media and used within Ih) and 300pM SAG and lOOng/ml FGF8 for the first feed.

Organoid media was then exchanged every 3 days, with media lacking Matrigel, and with a change to Improved+A media around day 16.

Organoids were transferred to orbital shakers on day 20 and cultured in Improved+A media until day 60, where a gradual shift to Brainphys media (25%, 50%, 75%, 100%) is performed. As soon as organoids were transferred to orbital shakers, media volume was increased to 30ml.

Example 1.7: Generation of striatal organoids:

For striatal organoid preparation, cells were grown for 2-3 days after splitting in MG coated 6well plates and dissociated using 600pl Accutase. 9000cells were then transferred into 150pl neural induction media containing lOnM SAG and 2.5pM IWP2 (Sigma- Aldrich, Cat.# 10536) as well as 1:100 ROCK inhibitor. Media was exchanged on day 2 by replacing lOOpl media with 150pl fresh NI media containing lOnM SAG, 2.5pM IWP2 and 1:100 ROCK inhibitor. On day 4, media was exchanged again by replacing 150pl media with 150pl fresh NI media with lOnM SAG and 2.5pM IWP2. On day 6, 150pl of media were exchanged with NI media without factors. On day 8, organoids were transferred into 10cm plates coated with anti-adherence rinsing solution. Media was exchanged with 12ml Imp-A media containing 2% Matrigel (added to COLD media and used within Ih) . Organoid media was then exchanged every 3 days, with media lacking Matrigel, and with a change to Improved+A media around day 16. Organoids were transferred to orbital shakers on day 20 and cultured in Improved+A media until day 60, where a gradual shift to Brainphys media (25%, 50%, 75%, 100%) is performed. As soon as organoids were transferred to orbital shakers, media volume was increased to 30ml.

Example 1.8: Generation of ventral hindbrain organoids :

For ventral hindbrain organoid preparation, cells were grown for 2-3 days after splitting in MG coated 6well plates and dissociated using 600pl Accutase. Cells were washed off with 1.4ml stem cell media and spun down at 150g for 5min. Cells were resuspended in 1ml stem cell media +1:100 RI and counted. 9000cells were then transferred into 150pl neural induction media containing 2pM DMH1 (R&D Systems, Cat.# 4126) , 2pM SB431542 and 1.4pM CHIR99021 and ROCK inhibitor in an ultra-low attachment U-shaped 96well plate. On day 2, lOOpl of this media was removed and replaced with 150pl fresh media of the same formulation.

On Day 4, 150pl of media was exchanged, with the addition of 2pM DMH1, 2pM SB431542, IpM CHIR99021, 300nM SAG and 50ng/ml FGF4 (R&D Systems, Cat.# 7460-F4-025/CF) to the newly added media. On day 6, 150pl of media was exchanged, containing 300pM SAG and 50ng/ml FGF4.

On day 8, organoids were transferred into 10cm plates coated with anti-adherence rinsing solution. Media was exchanged with 12ml Imp-A media containing 2% Matrigel (added to COLD media and used within Ih) and 300pM SAG and 50ng/ml FGF4 for the first feed .

Organoid media was then exchanged every 3 days with media lacking Matrigel, and with a change to Improved+A media around day 16.

Organoids were transferred to orbital shakers on day 20 and cultured in Improved+A media until day 60, where a gradual shift to Brainphys media (25%, 50%, 75%, 100%) is performed. As soon as organoids were transferred to orbital shakers, media volume was increased to 30ml.

Example 1.9: Generation of PDMS fusion embedding molds : Embedding molds have been designed in Tinkercad (www. Tinker- cad, com) and were adjusted in diameter and length based on organoid size on the day of fusion. Files were exported as .stl files and loaded into the slicer software XYZ print 1.4.0. The negative was printed using transparent PLA with 100% infill density and 0.1mm layer height and 215°C nozzle temperature. After printing, the molds were treated with a Heatgun (Bosch Hot Air Blower 1800W) at 550°C to carefully melt the surface of the mold, creating a smooth finish and removing the typical rough surface of 3D printing.

The positive was then casted using polydimethylsiloxane (PDMS) . In brief, 5ml of curing agent and 45ml of Monomer (both Sylgard® 184 Elastomer Kit, VWR) were intensively mixed. The mixture was then spun down to remove air bubbles and used directly.

To reduce the extent of bubbles formed during curing, the molds were first cast at room temperature (24°C) . For this, the 3D printed negative is placed in a 10cm plate with the wells looking up. The PDMS was then carefully poured on the middle of the mold and the plate is filled with approx. 45ml of PDMS. After pressing the negative down to remove air bubbles, the molds were cured overnight at RT . For faster curing, the mold was alternatively transferred into a 55°C incubator, however this can increase the release of gas and thus bubble formation into the PDMS . When curing at RT overnight, the cast was transferred to 55°C for 2-3 hours the next day to finalize curing.

The mold was then cut out of the plate and washed in 70% Ethanol for 30min and then dried.

The negative was re-used as it is not damaged in the casting procedure .

Example 1.10: Generation of mold assisted fused organoids:

Molds were coated in an anti-adherence solution to increase the non-stick behavior of PDMS further. After coating, the molds were washed once in PBS .

For linearly fusing organoids, organoids were transferred group by group into the embedding molds, with up to 32 fusions at once. For this, the first group (exemplary cortical organoids) was transferred into all molds first, transferring as little media as possible with the organoid. Secondly, striatal organoids were transferred to the right of the cortical organoids, and third, the procedure was repeated with ventral midbrain organoids. Optionally, ventral hindbrain organoids were added as well .

After transfer, residual media of the organoids was removed entirely, paying close attention to not disturbing the positioning of the different organoids. This step allows the attachment of organoids linearly (without media, organoids become sticky and attach to each other) . Optionally, after roughly 30sec- 1 minute, a small amount of liquid Matrigel (about 15pl) is added to the fusions. The fusions were transferred into an incubator (37°C, 5% CO2, humidified) for 20min. After this incubation step, the linear fusions can be easily washed out of the embedding molds. The embedding molds can be directly used for fusing organoids again or washed in PBS and 70% Ethanol (cell-culture grade) and then dried. Fused organoids were cultured in 10cm plates without shaking for the first 2 days, before being transferred to an orbital shaker in 30ml of media with reduced shaking speed (Inforce shakers, 42rpm instead of 56rpm) .

Matrigel may act as a glue for the tissues but it is an animal- derived product which might be problematic for clinical applications. Thus, we also provide a procedure which does not use Matrigel. Steps between the transfer into the mold can be replaced as illustrated in Fig. 19 (A, the method with Matrigel for reference; B, replaced steps in the method without Matrigel) . When transferring organoids into PDMS embedding molds and removing media for a minute, tissues become pre-attached. Media can then carefully be positioned on top of the tissue, with special focus on not breaking tissues apart. Subsequently, pre-fused organoids were incubated in an incubator for at least over night to let the tissues attach (incubation of 3-4 days would be possible at this step) . Fusions were then washed out of the plate and the protocol was continued as in the method with Matrigel.

Example 1.11: scRNAseq protocol:

Organoid fusions were separated by a scalpel. Ventral midbrain, striatal and cortical tissue were processed separately. Before dissociation, necrotic material from the core of the organoid was washed off in PBS ^_/_ . Organoids were then transferred into a 1.5ml Eppendorf tube and all PBS was removed. 1ml of Trypsin-Ac- cutase (1:10) (Trypsin: Thermo Fisher, Cat.# 15090046) was added to the organoids and the tubes were transferred to a tube shaker at 37°C 700RPM. Tubes were flipped every 3-4 minutes until the tissue was dissolved, or for a maximum of 40min. After dissociation, 500pl of ice-cold PBS ^_/_ were added on top. Cloudy material on top of the solution, which forms occasionally, can be removed with a pipet. Tubes were then spun down in a precooled (4°C) table centrifuge at 400rpm for 5min. Then, as much liquid as possible was removed without damaging the cell pellet. Pellets were carefully resuspended in 400ul of PBS ^_/_ and filtered through a 40pm cell strainer.

Cells were then counted. Live cell count had to be above 80% and above a concentration of 200. OOOcells/ml, but not more than 2.7mio cells/ml. Samples were diluted down, where needed. Multiseq labeling using MULTI-seq Lipid-Modif led Oligos was performed directly after counting following the manufacturer' s protocol (Sigma-Aldrich, Cat.# LM0001) . In brief, 40ul of anchor solution including barcode was mixed with the sample and incubated on ice for 5min. 40pl of co-anchor were then added and mixed and incubated on ice for 5min. The reaction was then topped up with 1ml of ice-cold PBS ^_/_ containing 2% BSA, and mixed to bind residual anchor and co-anchor.

Cells were spun down at 400-600RPM for 5min at 4°C. The supernatant was removed and cells were resuspended in 200pl PBS ^_/_ with 2% BSA and supplemented with 0.8pg/ml DAPI .

Cells were then FACS sorted for live cell sorting (DAPI-) in a FACS Aria III machine and a 70pm nozzle (approx. Ini per cell) into a 1.5ml Eppendorf tube with 20pl of PBS ^-/- and 2% BSA.

Library preparation was performed using the service of the Vienna BioCenter NGS facility using a lOx Genomics stranded kit and NovaSeq SI Asymmetric 10X Illumina sequencing.

Example 1.12: scRNAseq analysis: scRNAseq analysis was performed using R Studio and the Seurat toolset of the Satija lab (satijalab.org/seurat/) . Cutoffs for cells were defined as mRNA (20%) , feature_RNA (6000) and ribosomal 40%.

Example 1.13: Cryoprocessing of tissue:

Organoid tissue was rinsed once in PBS and fixed in 4% Formaldehyde solution in lx PBS for 4h at RT . The tissue was then incubated for one day in 30% sucrose at 4°C. Tissue is subsequently transferred to a bed of OCT (Scigen, Cat. #4586) and OCT is carefully swirled around the organoid. After 5min incubation, organoids were transferred into cryomolds and transferred onto a metal block on dry ice. Once a small ring of solidified ^frozen) OCT appears, the cryomolds were filled up with OCT and were completely frozen. The resulting cryoblocks were then wrapped in aluminum foil and were transferred to at least -70°C until cryosectioning .

Frozen organoid tissue was sliced into 20pm (regular organoids) or 30pm (for better representation of axons) using a cryostat, and collected on cryoslides.

Sections were dried for at least 3h before processing for immunostaining, or alternatively dried overnight and then stored at -20°C.

Example 1.14: 2D IHC:

For Immunohistochemical fluorescent labeling, slides were thawed and dried for Ih at RT . Residual OCT was then washed off by washing 5min in PBS on an orbital shaker. Permeabiliza- tion/blocking was performed by gently dropping PBS containing 5% BSA and 0.3% TX100 (Sigma-Aldrich, Cat. #93420) on the slides and incubating them at RT in a humidification chamber for 30min. Permeabilization/blocking solution was sterile filtered and then frozen, and fresh aliquots have been used for every staining round .

Primary antibodies were added at desired dilutions in staining solution (5% BSA, 0.1% TX100, in PBS, sterile filtered and frozen until needed) and incubated at 4 °C overnight in a humidification chamber. On the second day, the slides were rinsed 3x with PBS and then washed 3x for lOmin in PBS-T (PBS+ 0.1% TX100) at RT on an orbital shaker. Secondary antibodies were added at a 1:500 dilution and slides were incubated for 2h at RT in a humidification chamber. DAPI (2pg/ml in PBS) was added for 7min, then slides were rinsed 3x in PBS and then washed 2x in PBST. The last washing step was performed using only PBS. Coverslips were mounted using fluorescent mounting medium (DAKO, Cat.# S3023) . Slides were stored at RT for at least 4h before using for microscopy and at 4°C for storing.

Example 1.15: 3D IHC:

2Eci tissue clearing was performed as previously described (Mas- selink & Reumann, et al., Development 146, 2019) . In brief, organoids were fixed with 4% formaldehyde in PBS for 4h at RT . For 3D immunohistochemistry, PBS-TxDB was prepared: lOx PBS, 5% BSA and 2% TX100 were pre-mixed and filled up with distilled water to achieve approx. 70% of aimed volume (e.g. for 11 of PBS-TxDB, 100ml of lOx PBS, 50g BSA and 20ml TX100 were used and filled up to 700ml with distilled water. 20% DMSO was then added slowly under heavy stirring - rapid addition of DMSO may result in irreversible BSA aggregation. The solution was filled up to 100% with distilled water and sterile filtered. PBS-TxDB is stable at RT for at least several weeks, or at 4°C for up to a year.

Organoids were blotted/permeabilized on a rotor or shaking plate for 1-2 days in 10ml PBS-TxDB at RT . Organoids (dependent on the size, between 3 and 6) were then transferred into 2ml Eppendorf tubes containing primary antibody solution in PBS-TxDB (between 500pl to 1ml) and transferred to a rotor or shaking plate for 5 days. Organoids were then washed in 10ml PBS-TxDB for 2 days (lx rinse, 3x PBS-TxDB exchange) and the same steps were repeated for secondaries. After washing off the residual secondaries, organoids were washed in PBS for Ih and then fixed in 4% Formaldehyde solution for 30min-2h. Organoids were then ready to be dehydrated in a 1-Propanol gradient (30%, 50%, 70%, 100%, 100%, 4-8h per step) . As no endogenous fluorescence had to be maintained, pH adjustment of the 1-Propanol was skipped. After dehydration, organoids were transferred into Ethyl Cinnamate and were ready for imaging as soon as completely transparent (l-3h or overnight at RT) . Cleared organoids were stored at room temperature and should not be transferred to 4°C, as Ethyl Cinnamate crystallizes out at lower temperatures.

Example 1.16: Microscopy (2D, 3D, timelapse) :

Cell culture microscopy was performed using a Zeiss Axio Vert. Al widefield microscope with the Axiocam Ere 5s camera (Zeiss GmbH) .

2D and 3D tissue clearing recordings were performed using an Olympus Spinning Disk system based on the Olympus 1X3 Series (1X83) inverted microscope. The system is equipped with a dualcamera Yokogawa W1 spinning disk using 405nm, 488nm, 561nm and 640nm lasers and recorded with the Hamamatsu Orca Flash 4.0 camera. Objectives used were 10x/0.3 (Air) WD 10mm, 10x/0.4 (Air) WD 3.1mm, 20x/0.75 (Air) WD 0.6mm and 40x/0.75 (Air) WD 0.5mm. For live imaging, the attached incubator setup was used (37°C, 5% CO2 ) . For high magnification live imaging with air objectives, the additional 3.2x magnification of the Orca Flash 4.0 camera was used.

For dopamine sensor live imaging, a Visiscope Spinning Disc Confocal (Visitron SystemsGmbH) was used. This system is based on a Nikon Eclipse Ti E inverted microscope, equipped with a Yokogawa W1 spinning disc and an incubator setup (used at 37°C, 5% CO2 ) . Components are controlled by the Visiview software. Lasers used were 405nm 120mW, 488nm 200mW, 561nm 150mW and 640nm 150mW using the Andor Ixon Ultra 888 EMCCD camera (13pm pixel, 1024x1024 pixels) or the PCO Edge 4.2m sCMOS camera (6.5pm pixel, 2024x2024 pixels) .

Example 1.17: Image processing:

Images were processed using the FIJI distribution of the open- source image processing application Image J (version 1.53c) . For 3D reconstruction, the open community software Icy was used (icy.bioimageanalysis.org/) . Example 1.18: GCAMP setup:

Organoids were grown until day 120. GCAMP was recorded using an Olympus Spinning Disk (see Microscopy section) . In brief, 20x magnification and an exposure time of 65ms per frame (continuous recording) were used for up to 6.5min of continuous recording. For trace extraction, recordings were scaled down from 2048x2048 to 512x512pixels . Trace extraction was performed using the open- source software package CalmAn (Flatiron Institute) .

Example 1.19: Extracellular recordings and optogenetic stimulation :

Organoids were transferred in a batch recording chamber (PC-41 LP, Warner Instruments) and mounted with small amounts of 4% agar in Phosphate Buffer (PB) . The recording chamber was then transferred to a pre-warmed dish incubator (DH-351L, Warner Instruments) and perfused with 5ml/min (95% 02 5%CO2 infused) ACSF at 37°C (controlled by TC-344C, Warner Instruments) through a peristaltic pump (PPS2, Multichannel System) . For extracellular spike detection, signals were recorded using silicon probes with two shanks and 16 recording sites each (total of 64 recording sites, Probe ASSY-77 P-2, Cambridge NeuroTech) . Probes were reused for recordings and were washed in 1% Tergazyme (Merk, Cat.# Z273287) in MilliQ water for 30min at 60°C. Probes were then washed with iso-Propanol and rinsed in MilliQ water.

Probes were positioned on the organoids and acquisition was performed at 16bit resolution on a 64-amplifier chip (RHS2000, Itan Technologies) with the open-source tool open-ephys. The signal sampling rate was set to 30kHz per channel and filter to 0.1Hz high-pass and 5kHz.

For optogenetic stimulation, a light source (Sola SMII, Lumen- cor) was connected to a collimator with insertable emission filters (630/69nm and XXX, Thorlabs) . From there, the filtered light was transported with a fiber optic glass (400pM, NA=0.39, Thorlabs) to the organoid. For stimulation, the fiber optic glass was positioned less than 1mm away from the site of stimulation .

Pulses were generated using Pulse Pal v2 (Sanworks) .

Prior to recording, the probes were positioned on the organoid and incubated for 5min. Then, a lOmin baseline recording was performed. Subsequently, the organoids were stimulated for lOmin following a 500ms-10sec interval (500ms stimulation, lOsec pause) .

Electrophysiology analysis was performed using the "minibrain" pipeline (github.com/JoseGuzman/minibrain) by Jose Guzman.

Example 1.20: Viral transduction protocol:

For individual organoids, IxlO ¹³ vg was used. For triple fusions, 3xl0 ¹³ were used. Viral titer might have to be significantly adjusted based on organoid size, or transduction efficiency of the virus .

Organoids were transferred into 24 well plates and were incubated with the virus in 400pl of media for 4h. Subsequently, organoids were transferred into 6cm plates together with the media containing the virus. Organoids of the same group can be transferred together. Media was then added to achieve a volume of 6ml (individual organoids) or 7ml (fusions) .

Organoids were transferred to an orbital shaker overnight. The next day, organoids were transferred back into 10cm plates containing 30ml of media.

Example 1.21: qPCR:

For each condition, 6-8 organoids were collected at indicated time points into 2ml RNAse-free tubes. RNA was extracted using the Rneasy mini kit (Qiagen) . cDNA synthesis was performed using Ipg of total RNA and Superscript II (Invitrogen) enzyme, following protocols provided by the manufacturer. qPCR was performed using Sybr Green master mix (Promega) on a BioRAD 384-well machine (CXF384) . The protocol used is: (1) 95°C 3min, (2) 95°C lOsec, (3) 62°C lOsec, (4) 72°C 40sec, (5) go to 2, repeat 40x, (6) 95°C Imin, (7) 50°C lOsec.

Quantification was performed in excel by calculating ACt relative to TBP. Data are represented as expression level (2~ ^ACt ) relative to TBP.

Example 1.22: Statistical analysis:

For statistical analysis, GraphPad Prism 8.1.1 was used.

Example 1.23: mDA graft injections:

16-day old mDA progenitors were thawed in a water bath until a small sliver of ice is remaining. 500pl of room temperature DMEM/F12 with 20% KOSR (ThermoFisher, Cat.# 10828028) and 1:100 RI was added drop by drop on top of the cells. Cells were then transferred into 10ml of DMEM/F12 with 20% KOSR and 1:100 RI and spun down at 500g for 5min. The supernatant was removed and cells were resuspended in exactly 1ml of media and counted. Cells were then spun down again and resuspended in the required volume for injecting (20.000 cells per lOOnl of media) . Organoids were injected using a Nanoinjector (Nanoject II, Drummond) and pulled glass capillaries (Drummond™ Capillaries for Nanojet II™ injectors) which were pulled using a Micropipette puller (Model P-97, Sutter Instrument) . A total of 207nl was injected in three pulses (69nl each) in the fast injection mode. Prior to injection, organoids were transferred into an empty plate and all media was removed. After injection, organoids were not touched for one minute to allow closure of the injection site. Organoids were then transferred back into 10cm plates with media and incubated without shaking for one day, before being transferred back to an orbital shaker.

Example 2 : Recreation of dopamine-associated brain regions in cerebral organoids

The human dopaminergic system is majorly comprised of dopaminergic neurons in the ventral midbrain. From there, two major dopaminergic cell types project into the striatum and cortex. The first group of dopaminergic neurons are A9 dopaminergic neurons in the substantia nigra compacta. A9 dopaminergic neurons are associated with the nigrostriatal pathway and project into the dorsolateral striatum, where they have a crucial function in fine motor control (Arenas, Denham and Villaescusa, Development 142, 1918-36, 2015) . These A9 dopaminergic neurons preferentially degenerate in Parkinson's disease (Lees, Shin and Revesz, Lancet 373, 2055-66, 2019) . The other, more abundant population of dopaminergic neurons are A10 dopaminergic neurons from the ventral tegmental area. A10 dopaminergic neurons project to the ventral striatum as well as into the cortex, majorly into the limbic system and prefrontal cortex. A10 dopaminergic neurons are associated with affective encoding ("dopaminergic reward pathway") (Lio, Shin and Ikemoto, Neuropsychopharmacology 33, 2182-94, 2014) .

During development, dopaminergic axons first send out axons dorsally and are then attracted rostrally. They then project longitudinally through the midbrain and diencephalon and form the medial forebrain bundle (MFB) . Once dopaminergic axons reach the telencephalon, they particularly innervate the striatum and project further into the developing cortex (Arenas, Denham and Villaescusa, Development 142, 1918-36, 2015) .

While the dopaminergic system is relatively well understood in rodents, dopaminergic neurons project into the arguably evolutionary most different regions in the human brain: striatum and cortex (Florio and Huttner, Development 141, 2182-94, 2014) (Balsters et al., Elife 9, 1-24, 2020) . Additionally, rodents generally do not develop Parkinson's disease, which makes artificial lesion models necessary (Stroker& Greenland, Chapter 5 Table 3, 2018) .

The targeted regions and the connectome of both A9 and A10 dopaminergic neurons are relatively well understood, however, not much is known about the factors which contribute to A9 or A10 differentiation, as well as why A9 dopaminergic neurons are so much more affected in Parkinson's disease.

We thus decided to work on an in vitro system to be able to model human-specific aspects of the dopaminergic system. Such a system would ideally include the generation of a diverse population of dopaminergic neurons from the ventral midbrain, but also its interaction partners (dopamine sensitive neurons) in the striatum and the cortex. Additionally, such a system should be scalable and allow for the formation of dopaminergic circuits to be studied on a functional level.

While many axonal outgrowth models exist (e.g. coated surfaces) , we decided to aim for a 3D environment, which should also provide more appropriate chemotactic signaling and an environment more similar to that in vivo. We also wanted to challenge this system by investigating if ventral midbrain dopaminergic grafts would behave similarly to the in vivo situation and innervate their target regions.

To achieve the goal of a 3D model system for the ventral midbrain, we used brain organoids. Brain organoids depict the early stages of neurogenesis, but also allow us to study the maturation of neurons and their properties, for example by looking at neuronal activity. Additionally, several attempts have been made to study brain region interactions using fused brain organoids, such as in the example of pallial : : subpallial fusions (Bagley et al., Nat Methods 14, 743-751, 2017) (Birey et al., Nature 545, 54-59, 2017) and cortex : : thalamic fusions (Xiang et al., Cell Stem Cell 24, 487-497, 2019) .

As a first step, we tried to recreate the three major regions which are associated with A9 and A10 projections: ventral midbrain, striatum and cortex.

As cortical organoids are already a well-established protocol and many publications exist which use cortical organoids (Giandomenico et al., Nat. Neurosci. 22, 669-679, 2019, Eichmuller et al., Science 375, 6579, 2022) , we developed patterning protocols for the ventral midbrain (Figure 1A, 2, Figure 3) and striatum (Figure 1A, 4, Figure 5) by patterning organoids along the anterior-posterior axis and dorsoventral axis (Figure IB) .

To test the versatility of the method, we further developed a novel protocol for the generation of ventral hindbrain organoids which contain serotonergic neurons. Serotonergic neurons, similar to dopaminergic neurons, broadly innervate forebrain structures and are vital for, among others, mood control, reward and learning. Dysfunctions in the serotonergic pathways are associated with a range of diseases such as depression and Amyotrophic Lateral Sclerosis (ALS) , but also many Parkinson's disease (PD) patients show a decline in serotonergic neurons, which is associated with the non-motor symptoms of PD (anxiety, depression, dementia, sleep disturbances) .

Example 3 : Ventral midbrain organoids

Ventral midbrain dopaminergic neurons are born at the caudal floor plate of the ventral midbrain, close to the hindbrain (Arenas, Denham and Villaescusa, Development 142, 1918-36, 2015) . For the generation of ventral midbrain protocols, we modified existing protocols from 2D and 3D differentiation approaches (Jo et al., Stem Cell 1-11, 2016, Nolbrant et al., Nat Protoc 12, 1962-1979, 2017) . In a dual smad inhibitory context, we first defined the amount of sonic hedgehog activation using the small molecule SAG (Figure 2A) and found that a range from 200-1000pM reliable had a floor plate induction effect, as confirmed by qPCR. Preferably we used 300nM SAG. The addition of FGF8 was additionally performed to control for slight off-patterning and increase mDA yield, as described before (Arenas, Denham and Villaescusa, Development 142, 1918-36, 2015) .

For Wnt activation as posteriorizing factor, we found a concentration of IpM to be sufficient for midbrain induction (Figure 2 B-D, the forebrain and midbrain marker OTX2 is still expressed, the dopaminergic marker TH is still expressed) . Immunolabeling confirmed the presence of these markers and thus confirms that the modified protocol using SAG300nM CHIR1. OpM produces ventral midbrain dopaminergic neurons (Figure 2E-G) . We found that a concentration of 300nM SAG from day 4-11, together with dual SMAD inhibition and Wnt activation, was sufficient to introduce maximal FOXA2 expression levels by day 20. These organoids were positive for both the dopaminergic marker tyrosine hydroxylase (TH) , a key enzyme for dopamine synthesis, and F0XA2 on day 44. These TH+ mDA neurons widely expressed mDA markers EN1 and LMX1A indicating correct differentiation into mDA neurons. We additionally confirmed the robustness of the protocol by using 4 cell lines with two different genetic backgrounds and origin (176-1: iPSC, H9: hESC, Figure 3C-E) .

To summarize, we are able to grow ventral midbrain organoids with the ability to produce dopaminergic neurons.

Example 4 : Striatal organoids

Striatal neurons are born majorly from the lateral ganglionic eminence (LGE) , from where they migrate and form the developing striatum (Fjodorova, Noakes and Li, Neurogenesis, 2 (1) , 2015, Ornorati et al., Nat. Neurosci. 17, 1804-15, 2014) . To pattern into LGE, we performed a dose curve experiment in a Wnt inhibitory context (IWP-2, which anteriorized and inhibits dorsal fate) similar to previous 2D differentiation approaches (Wu et al., Stem Cell Reports 11, 635-650, 2018) . In brief, we activated sonic hedgehog signaling with the small molecule SAG at concentrations between lOnM and lOOnM (Figure 4B) . We performed a SAG dose-response curve in the presence of the Wnt inhibitor IWP2 in an otherwise growth-factor free neural induction media. Notably, already the lowest concentration of SAG resulted in a high level of rosettes expressing the LGE marker GSX2 (Figure 4G) while having a low number of rosettes expressing the more dorsal marker PAX6 (Figure 4D) . While SAG25nM had similar numbers, we noticed that the striatal (and cortical) marker CTIP2 was already significantly lower in SAG25nM treated organoids (Figure 4E, F) .

We additionally performed immuno labeling experiments for a range of striatal progenitor and young born striatal neuron markers, among others SOX2, ASCL1, DLX5 and ISLET1 (Figure 4G-J) . We additionally performed this protocol in a range of cell lines (both iPSCs and hES cells) and could confirm the existence of GSX2+ rosettes in all but one cell line (Figure 5B (exemplary) , Figure 5C) . However, as levels of endogenous SHH signaling might vary in each cell line, different concentrations of sonic hedgehog activation (or even inactivation) might be needed to achieve LGE induction.

Next, we looked at marker expression of typical striatal neurons. Striatal neurons express the marker DARPP32, which labels dopamine-sensitive cells (Figure 4L) and is broadly expressed in the striatum, as well as FOXP1, another striatal marker (Figure 4M, 0) (Arlotta et al., J. Neurosci. 28, 622-632, 2008) . Additionally, we investigated if DARPP32+ cells also start to become GABAergic, and indeed on day 80 the majority of DARPP32+ cells expressed GAD1, a marker for GABAergic neurons (Figure 4N, P) . We could confirm the expression of DARPP32 in multiple cell lines (Figure 5 F) .

We additionally found multiple cells in striatal organoids expressing Nkx2.1, one of the markers of MGE-derived cells, indicating that not just striatal medium spiny neurons, but also interneurons were produced in striatal organoids (Figure 5 G) . We confirmed the capability of this protocol to produce striatal neurons on three hiPSC and three hESC lines and found consistent organoid formation and induction of striatal neurogenesis. To summarize, we are able to grow striatal organoids with the ability to produce striatal neurons.

Example 5 : Ventral Hindbrain Organoids

Serotonergic neurons of the raphe nuclei, similar to dopaminergic neurons, emerge from the ventral part of the neural tube and are part of the raphe nuclei, from where they project to almost the entire central nervous system (Ren et al., Elife 8, 1-36, 2019) . For the generation of a ventral hindbrain protocol, we modified an existing protocol for the generation of 2D striatal neurons (Lu et al., Nat Biotechnol 34, 89-94, 2016) . Using two small molecules for DualSmad inhibition (DMH1 and SB431542) and a higher concentration of CHIR (1.4pM) from day 0 until day 6, we patterned into the hindbrain (Figure 1A) . On day 4 we applied the small molecule SAG at 300nM, together with FGF4, as both activation of SHH and FGF4 signaling have been shown to activate the serotonergic program (Ye et al., Cell, 93, 755-66, 1998) . Using this protocol, we could achieve organoids which were comprised of serotonergic neurons (TPH2 and 5-HT positive, ) (Figure 10B) .

Thus, we were able to generate a protocol which allows us to study features of ventral hindbrain serotonergic neurons in 3D neuronal culture.

Example 6 : Organoid fusion

We next developed a robust method to bring cortical, striatal and ventral midbrain organoids together. For this, we thought of a system which would allow us to position the organoids in their anterior-posterior orientation, or in the sequence dopaminergic neurons start to innervate. We developed PDMS linear fusion molds, which allow the positioning of three, or more, organoids in a linear manner (Figure 6A, B) . To do so, we first created a CAD designed negative which was printed using a 3D printer with PLA filament. We then cast the positive using Polydimethylsiloxane (PDMS) (Figure 6A) .

For linearly fusing organoids, we positioned the cortical, striatal and ventral midbrain (and, in the case of quadruple fusions, the ventral hindbrain) one by one in a row into the molds. We then removed the media and waited for at least 30 seconds to let the tissues stick to each other, before adding a small droplet of Matrigel. The addition of Matrigel is optional and the method can be done without it (Fig. 19) . After polymerization, the linearly fused organoids were gently washed out of the mold with Improved+ A media.

The resulting tissues allowed us to observe that the different organoids nicely grew together over time (Figure 6C) . Notably, over several batches we achieved a fusion efficiency of roughly 98% .

By the introduction of a constitutive GFP (CAG-GFP) expressing cell line into the ventral midbrain, we could observe first axonal outgrowth and axon bundles forming already 12 days after fusion (Figure 6D) , with an increase of innervation over time to densely innervated striatal and cortical regions ( Figure 6E ) . We could additionally confirm by immuno labeling that the maj ority of axons from the ventral midbrain were of dopaminergic nature ( Figure 6F, G) and that these dopaminergic axons innervated both TBR1+ ( cortical ) as well as DARPP32+ ( striatal ) regions .

By growing Cortex ( GFP ) -Striatum (unlabeled) -ventral Midbrain ( tdtomato ) fusions , we could additionally confirm that striatal tissues were not j ust innervated by ventral midbrain, but also by cortical axonal proj ections .

To summari ze , we found a highly ef ficient and scalable method to fuse organoids in a spatially controlled manner, which is of essence i f interactions with more than two regions have to be observed .

Example 7 : ScRNAseq

To investigate tissue identities in triple fusions , we performed single-cell RNA sequencing of day 60 linear fusions . To be able to separate cells from di f ferent regions unbiased from their expression profile , we separated the cortical , striatal and ventral midbrain tissues of fusions using a scalpel and labelled cells coming from the cortex, striatum or ventral midbrain individually using Multi-seq . We sampled roughly 1000- 1800 cells per region . We next performed analysis of marker expression and found clusters of cortical , striatal and ventral midbrain neurons ( Figure 8A) . We additionally found other neural populations such as MGE and CGE derived interneurons , midbrain glutamatergic and GABAergic neurons and small clusters of glial cells ( Fig . 8D-H) . Correlating the striatal and cortical cluster with the Brainspan dataset of the developing human brain at PCW20-25 uncovered a strong similarity index of cortical neurons with the developing neocortex, while striatal neurons were closest to the developing striatum . We additionally performed Voxhunt spatial similarity brain mapping of the dopaminergic, striatal and cortical clusters against E13 . 5 Allen Developing Mouse Brain Atlas data ( Fig, 8C ) . All three clusters correlated strongly with their corresponding tissue region in vi vo, further confirming VM mDA, striatal and cortical excitatory neuron identity .

Cortical neural progenitors readily expressed the markers PAX6 , FOXG1 and EMX2 ( Figure 8G-H) . Interestingly, we found a cluster of EOMES positive cortical intermediate progenitors which was more closely clustering with cortical neurons (FOXG1, EMX1, TBR1, SLC17A7) .

Striatal progenitors were positive for FOXG1 and ASCL1, and, when terminally differentiating, start to express GAD1, GAD2, FOXP1, BCL11B and ZFHX3 (Figure 8F) . The striatum is comprised of more than 85% spiny projection neurons (SPNs) , and while mature medium spiny neuron markers such as dopamine receptor type 1 and 2 (DRD1, DRD2, or DI and D2 ) were not broadly expressed on day 60, we could indeed find markers of DI (TAC1, ISL1 and EBF1) and D2 specific (SP9, SIXS, GIRK3) differentiation (Figure 8F' and F' ' ) .

We next checked ventral midbrain specific markers and found FOXA2 positive cells, which also expressed key markers of dopaminergic differentiation; amongst others EN1, SLC6A3 (DAT) , PITX3, as well as TH (Figure 8E) . Surprisingly, we only found a small number of TH+ neurons, which we speculate could be due to the fact that dopaminergic neurons are highly fragile and hard to dissociate, and improved protocols might be needed to increase the fraction of dopaminergic neurons.

We next investigated the remaining clusters in the data set and found that one cluster expressed genes typical for medium ganglionic eminence (MGE) derived interneurons, such as Nkx2.1, LHX6 and SOX6. We additionally found first cells being positive for astrocyte markers (S100B) or oligodendrocyte markers (OLIG1) , indicating that the switch from neurogenesis to astrogenesis and oligodendrogenesis already occurred in a small population of neural progenitors (Figure 8H) .

We also observed a population of neurons from ventral midbrain organoids being positive for GABAergic markers, hinting at the existence of midbrain GABAergic neurons in ventral midbrain patterned organoids (Figure 8A) .

This dataset confirms that we have all required cells of the dopaminergic system in the linear fusions and indicates that day 60 is a good time point to study neurogenesis of all three regions, as progenitors and neurons are still broadly present.

In summary, these data demonstrate that fused organoids are composed of dopaminergic, striatal and cortical neurons in a spatially organized manner, which can be used to study the innervation of dopaminergic neurons into striatal and cortical tissues in vitro. Example 8 : 2Eci tissue clearing allows morphological reconstruction of dopaminergic neurons and shows maturation features of dopaminergic axons

We next tried to recapitulate morphological aspects of the dopaminergic circuit and used a recently published tissue clearing protocol to perform whole-mount immuno labeling and 3D recordings of dopaminergic neurons and their proj ections .

We could find that dopaminergic neurons usually aggregated in clusters ( Figure 9A, G) , a phenomenon which can be found in vivo but to our knowledge has not been described in vitro . Additionally, we could confirm that both striatal and cortical regions were highly innervated by dopaminergic axons ( Figure 9A' ) and that dopaminergic axons often formed axon bundles ( Figure 9B, C ) , which sometimes were more than a millimeter in length ( Figure 9B ) .

Axonal boutons are the sites where synapses typically occur, which makes them an important metric for structural neuronal maturation . We noticed that between day 40 and 120 the number of axonal boutons on dopaminergic axons drastically increased ( Figure 9D, E ) to more than 3 axonal boutons/ 10pm on average on day 120 . While the striatum had signi ficantly more axonal boutons on day 40 , at later timepoints the density equali zed out between both striatum and cortex .

We could also notice that axons generally avoided regions of neurogenesis ( Figure 9F) , indicating the existence of chemotactic repellents similar to in vi vo .

Dopaminergic neurons have di f ferent morphologies dependent on their location and function in the brain, although still poorly described in the developing human brain . However, we found that dopaminergic neurons clustered together ( Figure 9A, G) and displayed a variety of morphologies ( Figure 91 ) , indicating a heterogeneous population of dopaminergic neurons . Surprisingly, we also found a TH+ population in striatal and cortical organoids . As previous reports exist of TH positive interneurons , we grew fusions containing a DLX561-GFP interneuron reporter cell line in the striatal part (which we speculated would be the only interneuron producing region) . Go-staining for GFP and TH could indeed confirm that TH+ cells in the striatum and cortex were of interneuron origin ( Figure 9H) . As another investigation of dopaminergic maturation, we immunolabeled for both TH and dopamine transporter ( DAT ) in 2D cryosectioned triple fusions . We could find that dopaminergic neurons in the ventral midbrain ( Figure 9J) expressed DAT , but also dopaminergic axons in the striatum ( Figure 9K) and cortex ( Figure 9L ) . DAT is vital for dopamine transport and indicates that dopaminergic neurons matured to a functional level , releasing and re-uptaking dopamine from its synapses .

By fusing three fusions with a constitutive GFP expressing striatal organoid ( CAG-GFP ) , we could see that there were reciprocal connections from the striatum to cortex and ventral midbrain, and by immuno labeling for GABA we could confirm that some of these proj ections were positive for GABA, confirming that there are GABAergic long-range connections in the fusions - one of the features of striatal proj ection neurons ( Fig . 15C ) . Notably, these GABAergic proj ections could be found in both the striatum and cortex, but while in the ventral midbrain usually only GFP+ axons could be observed, many cells in the cortex were GFP and GABA double positive , indicating migrated interneurons . Thus , we speculate that many GFP GABA double positive axons in the cortex are not of striatal proj ection neuron identity, but of interneuron identity .

To summari ze , these experiments show that we can recapitulate morphological features of the dopaminergic system to an unprecedented level and that reciprocal connections form between Ctx : : Str : : vMid fusions .

Example 9 : Addition of ventral hindbrain organoids allows the study of serotonergic pathways

To further test the fusion system, we developed a protocol for ventral hindbrain organoids containing serotonergic neurons ( Figure 10B ) and made linear fusions composed of Cortex : : Striatum : : ventral Midbrain : : ventral Hindbrain fusions using our previously developed fusion molds . We used a cell line with constitutive GFP expression to label the ventral hindbrain ( Figure I OC ) . We could observe high levels of innervation in the midbrain, with medium levels of innervation in striatal , and lower levels of innervation in the cortical regions . We additionally stained for TPH, a serotonergic marker and found serotonergic neurons in the hindbrain ( Figure I OC' ) . Immunolabeling for TH in the same fusions indicated abundant dopaminergic neurons in the ventral midbrain, and TH+ proj ections into the hindbrain, striatum and cortex can be observed ( Figure 10D) .

Fusions where both dopaminergic and serotonergic systems are available and can be studied simultaneously could potentially have huge implications in pharmaceutical research, as both systems have strong associations with maj or depressive disorders and the availability of two monoaminergic systems could allow for a broader investigation of potential drugs .

To summari ze , the methodology of recreating the dopaminergic system can be expanded to other, or more groups of di f ferent ( or the same ) brain regions with connective properties , allowing the study of many brain region interactions- and in this speci fic example allows the study of both serotonergic and dopaminergic pathway-speci fic features in one tissue .

Example 10 : Morphological reconstruction of neurons

Infecting organoids with an AAV expressing a membrane-bound Archl-GFP allowed us to morphologically reconstruct neurons from di f ferent brain regions and allowed us to observe discrete neuronal morphologies in Cortex, Striatum and ventral Midbrain ( Figure 10E ) . Additionally, it allowed us to observe intricate neuronal morphologies , such as dendritic spines on a dendritic tree of a cortical neuron ( Figure 10F) .

Example 11 : Neural activity measurements indicate heterogeneous region-specific activity, and functional dopaminergic long-range connections

To investigate features of neuronal activity in the linear fusions , we first investigated i f spontaneous neuronal activity could be observed . For this , we grew organoid fusions with an H9 cell line expressing the fluorescent calcium indicator GCAMP6s under a synapsin promoter, which restricts GCAMP expression to neurons . Notably, we always only used this cell line in one of the three regions , which restricts GCAMP expression to cells which received either cortical , striatal or ventral midbrain patterning . On day 120 , organoids were recorded for 6 . 5 min and the recordings were processed using an adapted version of the CalmAn calcium imaging analysis pipeline ( Flatiron Institute ) .

We could observe di f ferent modes of activity in Cortex, Striatum and Midbrain . Cortical organoids often, but not always , displayed repetitive synchronous network events ( Figure 11A, B ) , whereas neuronal activity in the striatum was usually more disorgani zed, and the calcium transients lasted longer . Ventral midbrain neurons displayed events of high- frequency synchronous calcium transients ( Figure 11A, B, end of the recording) , which was unique to recordings from the ventral midbrain .

We additionally looked at fusions with GCAMP in the ventral midbrain and investigated i f we could observe calcium transients in other regions in axons proj ecting from the ventral midbrain . As can be seen in Figures 11C and D, calcium transients could well be observed in axons proj ecting from the ventral midbrain into the striatum, confirming that neuronal activity of one region can also spread to other regions .

To confirm functional axonal connections , we combined the opto- genetic tool Channelrhodopsin2 with extracellular recordings using silicon probes . In brief , we infected organoids with an AAV- RG containing a CAG-hChR2-H134R-tdtomato and incubated them for 2 weeks . We then stimulated the ventral midbrain side of a triple fusion using an optogenetic setup and recorded from striatum and cortex using silicon probes ( Figure HE, F) . Notably, we could get direct responses from ventral midbrain stimulations in the striatum ( Figure 11G, H, I ) , indicating the establishment of functional long-range connections from the ventral midbrain . We could additionally observe a similar phenotype in the cortical part of the fusion .

To confirm not only the establishment of functional long-range connections , but also to test whether these connections were of dopaminergic identity, we performed another optogenetic stimulation experiment in composition with a fluorescent genetic dopaminergic sensor ( GRAB-DA4 . 4 ) . We infected organoids with AAV1 containing Syn-ChrimsonR-tdTomato ( as hChR2 is incompatible with GRAB-DA4 . 4 ) and incubated them for two weeks . We then recorded the fluorescence intensity of the ventral midbrain before and during optogenetic stimulation and could observe a change in fluorescence in many of the recorded neurons , indicating dopamine release ( Figure 11K, L ) .

To summari ze , we could show that we have spontaneously active neurons ( GCAMP ) , that they proj ect into other regions , form functional connections with the target neurons ( silicon probes ) and that stimulation of the ventral midbrain organoid results in a release of dopamine in target regions ( dopamine sensors ) . This allows us to ask dopamine circuit-speci fic questions .

Example 12 : Using triple fusions to test GMP grade mDA progenitors for cell therapy

Some of the most promising therapies for the currently uncurable Parkinson' s disease is the transplantation of dopaminergic neurons into patient brains . However, for highly ef ficient cell therapy, several steps still have to be improved and several caveats limit current studies .

Firstly, mostly A9 dopaminergic neurons are af fected ( although other dopaminergic neurons such as ol factory bulb dopaminergic neurons and A10 dopaminergic neurons can also be af fected but are of less vital importance ) . However, current grafts contain more A10 than A9 dopaminergic neurons , and it is currently unknown how to increase the proportions of A9 dopaminergic neurons in grafts . A system which allows fast screening of conditions , but also allows the interaction of neurons with its targets - an important step in dopaminergic maturation - could be ideal for testing such new conditions , without the need to go into slow, expensive and tedious animal experiments .

Secondly, dopaminergic grafts are currently grafted into the striatum, as grafts in the ventral midbrain did not show the potential to re-innervate the ( in the human) long distance into the striatum .

However, recent studies have shown that overexpression of axonal chemoattractants , such as GDNF, could recruit dopaminergic axons into the striatum in mouse experiments (Niamh et al . , Cell Stem Cell , 2022 ) . While this system seems to work in mice , dopaminergic axons have to migrate over much longer distances in the human brain, suggesting more sophisticated recruitment strategies might be necessary to recruit dopaminergic axons over longer distances .

We here propose the usage of linearly fused organoids to study mDA grafts in a 3D environment , which would allow high throughput and better accessibility than animal model systems . To investigate this possibility, we inj ected 40 . 000 GMP grade RC17 embryonic stem cell derived mDA progenitors in 200nl into the ventral midbrain part of a triple fusion ( Figure 12A, B ) . These cells were modified with an endogenous Cre in the TH locus, and double infected with a Lentivirus containing a floxed GFP, thus labeling dopaminergic neurons with GFP. One month after injection, the graft can be readily observed in the ventral midbrain. Surface confocal microscopy revealed high levels of axonal innervation of the midbrain, but also axon bundles projecting to the striatum, and innervation of striatal and cortical regions (Figure 12C) .

We next performed live-imaging of dopaminergic axons and could observe movement along the axon (Figure 12E) , which could be indicative of axonal transport, as well as dopaminergic axon-axon interactions (Figure 12F) .

We performed 3D tissue clearing and could recreate the dopaminergic grafts (GFP+) and F0XA2+ cells from the graft, with some of them already differentiating into TH+ neurons (Figure 12G) .

Example 13: Bulk RNAseq confirms VM, striatal and cortical tissue identity.

To validate the identity of the organoids generated using our patterning protocols, we performed bulk RNAseq at day 60 on VM, striatal and cortical organoids. Cortical organoids were grown according to recently published protocols (Eichmuller et al. Science (1979) 375, (2022) ; Esk et al. Science (1979) 370, 935- 941 (2020) ; Lancaster et al. Nat Biotechnol 35, 659-666 (2017) ) . We found that individual organoids clustered into their respective patterning groups when performing principal component analysis (PGA) (Figure 13 A) . Furthermore, among genes with the strongest contribution to PCI we found candidates of anterior- posterior identity, such as TH and RMST for posterior/VM identity, as well as NEUR0D2, NEUROD6 and BCL11A for anterior (forebrain) identity. Features with highest loading in PC2 mostly indicated forebrain dorsoventral patterning, including the dorsal markers NEUROD6 and NEUR0D2, while GAD1, DLX6-AS1 and LHX6 are classic ventral forebrain derived neuronal markers (Figure 13B) . To further confirm correct patterning, we analyzed the bulk RNAseq data with the spatial similarity analysis pipeline Voxhunt using the E13.5 Allen Developing Mouse Brain Atlas dataset as reference. We found a strong correlation between the patterned organoids with their corresponding tissues in vivo (VM, striatal and cortical identities) (Figure 13C) . Notably, striatal organoids were negative for the cortical marker TBR1 ( Figure 13D) .

Example 14 : Characterization of dopaminergic neuronal subtypes When investigating dopaminergic subtype identities , we found that the maj ority of TH+ neurons expressed the mDA markers GIRK2 and CALB1 . Furthermore , we found dopaminergic neurons being positive for the A10 VTA mDA neuron speci fication marker 0TX2 as well as ~ 10% S0X6 ⁺ mDA neurons- a marker associated with A9 SNc mDA neuron speci fication . We further found populations of dopaminergic neurons that expressed the dopaminergic subtype markers ALDH1A1 and GABA ( Figure 14B, C ) , indicating that neural multitissue organoids depict the heterogeneity of ventral midbrain dopaminergic neurons .

In consecutive batches , we achieved 96% ( ±2 . 4 % SD) fusion ef ficiency three days after fusion, after which we generally did not see tissue separation anymore ( Figure 14D) .

In this example , we have shown multiple dopaminergic subtypes which can be found in the human ventral midbrain, and that striatal neurons can mature and express dopamine receptors type 1 and 2 .

Example 15 : Dopaminergic innervation is stronger in str than in ctx

Striatal tissues were generally displaying denser dopaminergic innervation from vMid tissue , whereas dopaminergic innervation in the cortical tissue was less dense , which resembles innervation densities in vi vo ( Figure 15A) .

To investigate reciprocal connectivity, we generated chimeric fused organoids from striatal or cortical organoids that constitutively expressed GFP under the GAG promoter (Bagley et al . Nat Methods 14 , 743-751 ( 2017 ) ) . GFP expression in striatal tissues highlighted GFP+ GABA+ reciprocal connections from striatal into VM, as well as GFP+ interneurons which migrated into the cortical tissue ( Figure 15C ) . Notably, striatal proj ections into the VM are dis-inhibitory and play a crucial role in decision-making and motor control in vi vo . By constitutive labeling with GFP of the cortical tissue , extensive innervation was found in both striatum and VM tissue , as also found in vi vo . Intriguingly, innervation from the cortical into the striatal organoid mainly showed the formation of a significant amount of axon bundles in the striatum, which were however not detected in VM, and which appear to represent the formation of cortico-striatal fiber tracts (Figure 15D) .

Here, we have shown that dopaminergic innervation is stronger in str than in ctx, which corresponds to the situation in vivo. Reciprocal connectivity forms, which is an important factor for neural circuit formation and functional regulation. Formation of putative corticostriatal fiber tracts can be observed and it is shown that striatum gets innervated by both Ctx as well as vMid.

Example 16: optogenetical stimulation and response

We next investigated whether these long-range projections from VM functionally participate into the fused organoid neural circuitry. To test this, fused organoids were transduced with an adeno-associated virus (AAV) containing the light-sensitive opsin hChR2. We then optogenetically activated only the VM tissue section by spatially restricted optical illumination using a glass fiber and simultaneously recorded extracellular neural activity from either the striatum or cortex tissue using silicon neural probes (Fig. 16A, E) . The optogenetic stimulation of VM increased firing rates in the striatal (Fig. 16B) and cortical tissue (Fig.l6F) in synchrony with the 460nm light flashes. Overall, VM stimulation increased the firing rate higher in the striatum than in the cortical tissue (Fig. 16C, G) . 56.3% of all active channels displayed a firing rate change >10% from baseline in the striatum (Fig. 16 B-D) , as well as 48.4% of the extracellular units in cortical recordings (Fig. 16 F-H) . Together, this data confirms the ability of neural multi-tissue organoids to develop functional neural networks across multiple regions. The optogenetic stimulation of VM increased multi-unit firing rates in both striatal (Fig. 16A-C) and cortical tissue (Fig.l6E-G) with at least 74% of active channels responding to 460 nm light pulses (Fig. 16D, H) . To exclude that these findings were due to antidromic signal propagation through reciprocal axons, we broadly inhibited synaptic transmission with a cocktail of synaptic blockers (D-AP5, CNXQ, Gabazine, SCH-23390 and Sulpiride-/-) . In the presence of this cocktail, population responses to VM stimulation were absent and normalized firing rate per channel decreased significantly, indicating trans- synaptic long-range connectivity ( Fig . 16P, Q) . Together, this data confirms the ability of fused organoids to develop functional long-range connections and the formation of neural networks between fused brain regions .

To directly confirm dopamine release from VM mDA neurons using the fluorescent dopamine sensor GRAB-DA2m ( Sun et al . Nat Methods 17 , 1156-1166 ( 2020 ) ) , fused organoids were transduced with both AAVs for the light-sensitive opsin ChrimsonR (Klapoetke et al . Nat Methods 11 , 338-346 ( 2014 ) ) as well as GRAB-DA2m ( Figure 161 ) . We optogenetically stimulated VM tissue and recorded fluorescent changes from striatal and cortical tissues . VM stimulation increased dopamine levels in most striatal and cortical neurons ( Fig . 16 J-M) . Notably, the increase in dopamine release persisted during the period of stimulation ( Fig . 16L, M) .

To examine i f the fusion of organoids , and the formation of long-range connections , would lead to change in activity-regulated genes compared to unfused organoids , we performed bulk RNAseq of striatal and cortical tissue of fused organoids and compared activity-regulated genes , as well as dopamine signaling response genes in fusions versus individual , non- fused organoids . Remarkably, striatal and cortical tissues of fused organoids showed an upregulation of activity-regulated genes such as FOS and BDNF in comparison to non- fused individual organoids ( Fig . 16 N, 0) .

In summary, these data demonstrate the formation of functional circuits within fused organoids with mDA neurons from the VM tissue that can release dopamine into striatal and cortical tissues .

Here we demonstrated the functional connectivity between vMid and Str as well as vMid and Ctx . Striatum gets more connected than cortex . Further, stimulation of vMid triggers dopamine release showing functional dopaminergic innervation . Neural multitissue organoid fusions have increased expression of activity ( and dopamine signaling) related genes .

Example 17 : Neural multi-tissue organoids can be used as a model for the perturbation of the dopaminergic system and show longterm changes upon chronic stimulation with cocaine .

Addictive substances increase dopaminergic signaling, either by direct or by indirect regulation of the dopaminergic system . Here , we used the neural multi-tissue organoids to test the effects of cocaine on the establishment of the dopaminergic system . Cocaine is a dopamine reuptake inhibitor, which binds to the dopamine transporter ( DAT) and inhibits the reuptake of dopamine after release and thus increases the available levels of dopamine to target cells . While the ef fects of cocaine on the adult brain are relatively well understood, not much is known about the ef fects of cocaine on the development of the CNS - predominantly because of the lack of an ef fective model system . We tested the ef fects of chronic cocaine exposure and its potential ef fect on the establishment of the dopaminergic system . For treatment of fused organoids , we used 0 . 7pM cocaine hydrochloride ( IC50 of cocaine hydrochloride ) , a concentration well within the physiological range in humans , and treatments were performed every 3 days for Ih before exchanging medium to simulate the pharmacokinetics of cocaine concentration in humans . We chronically treated fused organoids with cocaine from day 40 until day 130 and generated a withdrawal condition which was not exposed to cocaine from day 105 onwards ( Fig . 17A) .

Cocaine exposure in vi vo has previously been shown to impact neuronal plasticity, as characteri zed by morphological changes and increased density of dopaminergic varicosities . To investigate whether cocaine treatment could cause such morphological alterations in fused organoids , we quanti fied the density of TH ⁺ varicosity in individual axons of striatal and cortical tissue . The varicosity diameter served as a readout for varicosity si ze ( Fig . 17C ) . Strikingly, we found that the density of varicosities on axons was signi ficantly increased in both striatal and cortical tissues , with the withdrawal condition showing similar changes and only a tendency of recovery ( Fig . 17B, D) . Moreover, the diameter of TH+ varicosities was signi ficantly reduced in both striatal and cortical tissue indicating varicosities of reduced volume , while 25-day long withdrawal also failed to rescue the neuromorphological ef fects of cocaine ( Fig . 17E ) . Thus , cocaine exposure of neural multi-tissue organoids recapitulates morphological phenomena of dopaminergic axons which have previously been described in vi vo .

We next investigated i f these morphological observations could cause a functional phenotype . In vi vo, cocaine exposure has been associated with changes in neural activity . To investigate i f we could recapitulate functional changes , the neuronal activity of treated chimeric fused organoids with region-speci fic and neuron-restricted GCAMP6S expression was recorded on day 130 ( Fig . 18A, B ) . Using the calcium imaging analysis pipeline CalmAn, we extracted a total of 3662 neurons in control , chronic and withdrawal of VM, striatal and cortical tissues of fused organoids . We analyzed the duration and frequency of calcium event duration from each condition ( Fig . 17 F) . Using this extensive dataset , we found that the calcium event duration in the striatum became signi ficantly shorter in the chronic condition ( Fig . 17G) . Conversely, cortical patterned neurons showed a signi ficant increase in calcium event duration . Surprisingly, the withdrawal condition in both cases was not signi ficantly di f ferent from the chronic condition, strengthening our observations that exposure to cocaine causes long-term changes in neuronal activity . Interestingly, VM neurons displayed a decrease in calcium event duration without recovery upon cocaine withdrawal ( Fig . 18C ) . We further found that neurons in VM, striatal and cortical patterned neurons were generally displaying fewer events per minute and did only partially recover in the withdrawal condition ( Fig . 17H, 18 D) .

Finally, we performed bulk RNAseq to investigate potential transcriptional changes which could be correlated with the observed morphological and functional phenotypes . Fused organoids were separated into forebrain ( Fb ) and VM and sequenced separately . Although only little transcriptional change was found in the VM tissue ( Fig . 18E-F) , a broad list of genes was di f ferentially expressed in the chronic and withdrawal conditions in the forebrain ( Fig . 171 , J) . Amongst the top hits of upregulated genes in chronic and withdrawal conditions , we found genes associated with interferon response ( such as I FITM2 and I FITM3 ) . There is a correlation between cocaine and interferon response in neural stem cells and astrocytes . Conversely, we found a downregulation of genes associated with neural circuit formation, which was also apparent in gene ontology ( GO) term analysis ( Fig . 17K) . While chronic treatment showed the strongest transcriptional changes regarding circuit formation, withdrawal still displayed signi ficant changes . To further investigate the transcriptional ef fects of cocaine on neuronal circuits , as well as its long-term ef fects , we performed gene set enrichment analysis ( GSEA) of transcriptional changes upon treatment and identi fied GO terms associated with neural circuit formation ( dendrite development , axon development and synapse organi zation) to be consistently and signi ficantly downregulated ( Figure 18G- I ) . The downregulation of associated genes in all three GO terms was consistent in both chronic exposure as well as withdrawal , further highlighting the long-term ef fects of cocaine treatment on neural circuit formation . Taken together, these data suggest that a chronic exposure to physiological concentrations of cocaine af fects neuronal activity, as well as circuit formation on the morphological and transcriptional level . These ef fects most likely are due to an increased exposure to the neuromodulator dopamine . Strikingly, correlating phenotypes were detected after withdrawal of cocaine , suggesting that the perturbations we found might have a long-lasting impact on the development of the dopaminergic system- and potentially other circuits . While these findings suggest lasting neuronal circuit modi fications , obtaining cell-type speci fic readouts ( e . g . DRD1 /DRD2 medium spiny neurons , A9/A10 mDA neuron contribution) allows to elucidate these perturbations in more detail .

Here we show that multi-tissue organoids can be used to screen drugs . In the example of cocaine as test drug, this had morphological , functional and transcriptional consequences . These consequences were long-term, indicating that neural circuits remain altered . This demonstrates that temporary or permanent ef fects of drugs can be studied .