EP4698629A1 - Derivation of primary epithelial organoids - Google Patents

Abstract

The invention relates to a method of generating tissue-specific primary fetal organoids from an epithelial stem cell population wherein said method comprises: i) isolating a viable population of cells comprising epithelial stem cells from a fetal fluid; ii) seeding the population of cells comprising the epithelial stem cells into an extracellular matrix (ECM) and culturing the seeded cells in a cell growth medium for the formation of clonal organoids from the epithelial cells; iii) extracting a clonal organoid from the ECM and dissociating the clonal organoid into single cells for passaging; iv) culturing the single cells such that they expand into tissue-specific primary fetal organoids; and related methods and products.

Description

Derivation of Primary Epithelial Organoids The present invention relates to a method of generating primary fetal epithelial organoids from the fetal fluids. These organoids are a model for development, disorders or disease state in tissues or organs, and uses for screening for efficacy of pharmacological and genetic manipulations for treatments. Background Modern prenatal screening routinely adopts sophisticated genetic and imaging analyses that are increasingly effective at detecting and characterising congenital anomalies. Despite this, it remains challenging to predict the functional severity of many complex conditions such as CDH, spina bifida (SB), cystic fibrosis (CF), polycystic kidney disease (PKD) and others, after their identification. Consequently patient-specific parental counselling is limited. Stratification of prenatal therapy is very relevant and has been used in the last few years to select patients for fetal intervention that may reverse the natural history of some of these conditions. There is level 1 evidence for improved outcomes in conditions such as CDH and SB, while for other conditions such as in vesico-amniotic shunting for lower urinary tract obstruction (LUTO) is technically possible, but appropriate patient selection remains the main hurdle. The lack of autologous / personalised systems that are capable of accurately recapitulating the complexity and functional characteristics of developing human tissues is the main bottleneck to implementing personalised regenerative medicine strategies that would deliver a significant impact in the lives of babies affected by these conditions. Consequently, in terms of directing prenatal therapy, the field of prenatal functional diagnosis remains underdeveloped. Organoids provide a reliable three-dimensional tissue model, which can recapitulate some of the biological and pathophysiological features of the patient’s tissues in vitro. Autologous organoids can be derived from human embryonic stem cells or during prenatal and postnatal life through reprogramming to iPSCs. These cells are then committed to the specific tissue- type of interest and expanded into stable lines. With respect to human developmental conditions, iPSC-derived organoids have been successfully generated from fetal cells within the amniotic fluid (AF) of both healthy fetuses and those with congenital anomalies. Overall, human iPSC-derived organoids have some advantages in terms of patient-specificity, but their reliance on considerable manipulation reduces fidelity to the patient’s individual condition, and strict quality control bears significant cost and time implications that hamper their applicability to personalised disease modelling in order to target therapy. Primary organoids have previously been derived from numerous human tissues. Being directly derived from the target tissues, these require significantly fewer in vitro manipulations, hence carrying lower safety burdens. More recently, primary organoids have been derived from discarded postnatal biological samples (e.g., urine, menstrual flow, PAP brush, bronchoalveolar lavage). This has clear advantages over the use of biopsies, allowing generation of autologous primary organoids from samples that are anyway acquired in the course of routine clinical care. In the context of prenatal medicine, primary organoids have been successfully derived from several fetal tissues collected post-mortem or through biobanks such as the Human Developmental Biology Resource (HDBR). The use of primary fetal organoids offers many unprecedented benefits to study human development and congenital diseases. Access to fetal tissues, however, has ethical and legal restrictions in many parts of the world that prohibit their use in research. Destructive methods of deriving primary fetal organoids limit their use for autologous disease modelling, prenatal functional diagnostics and personalised therapeutics. What is required is an improved method for deriving tissue-specific primary fetal organoids, that can be implemented prenatally, is compatible with the developmental timeline of the human fetus, and allows for continuation of pregnancy. Statements of Invention According to a first aspect of the present invention, provided herein is a method of generating tissue-specific primary fetal organoids from an epithelial stem cell population wherein said method comprises: i) isolating a viable population of cells comprising epithelial stem cells from a fetal fluid; ii) seeding the population of cells comprising the epithelial stem cells into an extracellular matrix (ECM) and culturing the seeded cells in a cell growth medium permissive to the formation of clonal organoids from the epithelial cells; iii) extracting a clonal organoid from the ECM and dissociating the clonal organoid into single cells for passaging; iv) culturing the single cells such that they expand into tissue-specific primary fetal organoids. Until now, primary fetal organoids have only been derived with destructive methods, limiting their use for autologous disease modelling, prenatal functional diagnostics and personalised therapeutics. This invention presents the derivation of primary human fetal epithelial organoids of multiple tissue identities (e.g. intestinal, renal and pulmonary) from fetal fluids, for example collected during the second and third trimester of gestation. Such fetal fluids (e.g., Amniotic and Tracheal) are already sampled as part of routine prenatal diagnosis and therapeutic intervention, and the process permits the generation of organoids alongside the continuation of pregnancy. The amniotic fluid (AF) surrounds, supports and protects the human fetus during development. The origin and recirculation of this fluid follows complex dynamics, evolving together with the development of the various fetal and extraembryonic organs that contribute to its production. As consequence, the amniotic fluid contains cells shed from a variety of origins29. The amniotic fluid harbours multipotent stem cells that are ascribed to the epithelial, mesenchymal and haematopoietic niches30–32. Amniotic fluid stem cells (AFSCs) have successfully been used in various animal models of disease, where they have demonstrated the capacity to induce regeneration through cell transplantation, conditioned media and activated AFSC-derived extracellular vesicles. Whilst stem cells with mesenchymal and hematopoietic potential have clear therapeutic promise, most of the cells present in the amniotic fluid manifest an epithelial identity. The amniotic fluid is highly heterogeneous in origin and composition and includes secretions and cells shed from various tissues, including the fetal kidney, lung and gastrointestinal tract29,37. Due to the complexity of the epithelial culture systems, a detailed map of the amniotic fluid epithelial population has not yet been compiled. The invention herein validates the presence of amniotic fluid epithelial cells (AFEC), highlighting that these shed from a multiplicity of developing tissues, and demonstrates that this population contains stem/progenitor cells capable of forming tissue-specific primary fetal organoids. Advantageously, with the adoption of fetal surgery procedures such as Fetal Endoluminal Tracheal Occlusion (FETO) to treat CDH, we can now sample and expand viable epithelial cells from tracheal fluid (TF) at various stages in the therapy 39–41. The invention herein provides evidence that diverse epithelial stem cell populations are shed from a number of different tissues into the fetal fluids during development. It is shown that these cells are capable of forming tissue-specific primary fetal organoids. Autologous derivation of primary fetal organoids during a continuing pregnancy broadens the possibility of conducting research at later gestational stages which may be beyond the limits of termination of pregnancy, allowing the development of functional diagnosis, personalised counselling, and providing innovative tools for designing autologous prenatal and perinatal therapies. Step i): isolating a population of cells comprising epithelial stem cells from a fetal fluid; The fetal fluid may be collected during gestation, such as during the second or third trimester of gestation. The fetal fluid may be collected at any time post-conception, for example from at least 2-10 weeks post-conception, or from at least 5 weeks post-conception. In one embodiment, the fetal fluid may be collected from at least 10 weeks post-conception. The fetal fluid may be collected for autologous derivation of primary fetal organoids during a continuing pregnancy. The methods herein may comprise the step of collection of the fetal fluid, or providing or receiving the fetal fluid that has been collected, for example by a third party. The fetal fluid may have been stored, which may comprise cooling at 4°C or cryopreservation. In another embodiment, the fetal fluid is freshly collected. The amount of fetal fluid collected or provided may be sufficient to provide viable epithelial stem cells for culture. In one embodiment, at least 0.5ml fetal fluid is collected or provided. In one embodiment, between about 0.5ml and 2500ml fetal fluid is collected or provided. The step of isolating the population of cells from the fetal fluid may be ex vivo or in vitro. Preferably the fetal fluid is ex vivo when the population of cells is isolated. Preferably the population of cells is a viable population or cells, or at least comprises viable epithelial stem cells. The step of isolating the population of cells from the fetal fluid may comprise the isolation of a viable population of cells or isolation of viable epithelial stem cells. Preferably the population of cells is not derived from induced pluripotent stem cells (iPSCs) or embryonic stem cells (ESCs). Preferably the epithelial stem cell population is not derived from induced pluripotent stem cells (iPSCs) or embryonic stem cells (ESCs). Preferably the method does not involve embryo destruction or harvesting. The fetal fluid may be one or more of amniotic fluid (AF), tracheal fluid (TF), vernix caseosa, fetal urine, meconium, saliva, ascites fluid, pleural fluid, aqueous humor, lacrimal fluid, and nasal mucus. The fetal fluid may be cerebrospinal fluid. In one embodiment, the fetal fluid is amniotic fluid (AF) and/or tracheal fluid (TF). In one embodiment, the fetal fluid may be one or more of vernix caseosa, fetal urine, meconium, saliva, ascites fluid, pleural fluid, aqueous humor, lacrimal fluid, and nasal mucus. The epithelial stem cells may be amniotic fluid epithelial cells (AFEC). In another embodiment, the epithelial stem cells may be from tracheal fluid, for example from tracheal fluid taken during Fetal Endoluminal Tracheal Occlusion (FETO). In another embodiment, the epithelial stem cells may be from one or more of the fetal fluids comprising amniotic fluid, tracheal fluid vernix caseosa, fetal urine, meconium, saliva, ascites fluid, pleural fluid, aqueous humor, lacrimal fluid, and nasal mucus. In one embodiment, the epithelial stem cells may isolated from the population of cells, for example prior to seeding step ii). In particular, only epithelial stem cells may be isolated for the seeding step ii). The epithelial stem cells may be isolated using fluorescence-activated cell sorting (FACS) or magnetic cell sorting (MACS). In one embodiment, the epithelial stem cells are isolated using fluorescence-activated cell sorting (FACS). In an alternative embodiment, the epithelial stem cells are isolated using fluorescence-activated cell sorting (MACS). Prior to cell sorting, the fetal fluid may be filtered through a cell strainer and/or centrifuged to form a pellet comprising the epithelial stem cells. The pellet may be resuspended in a buffer. Preferably, the heterogeneity of the cells may be preserved during FACS by not selecting the epithelial stem cells by Forward Scatter or Side Scatter. Viable epithelial stem cells may be isolated. Viable cells may be identified and isolated, for example away from non-nucleated cells, cell debris, urea crystals and/or other non-cellular particles. Viable cells may be identified and isolated by the incorporation of a viable cell stain, such as Hoechst, which can be used to identify nucleated cells. Additionally or alternatively, an apoptotic cell stain such as propidium iodide (PI) may be used to exclude dead cells and/or cells with a damaged plasma membrane. The epithelial stem cells may be identified and isolated based on the expression of one or more of the epithelial marker genes, such as EPCAM, ECAD (CDH1) and KRT genes. In one embodiment, the epithelial stem cells may be identified and isolated based on the expression of two or more of the epithelial marker genes such as EPCAM, ECAD (CDH1) and KRT genes. Optionally, the epithelial stem cells may further be identified and/or isolated before seeding as organ-specific progenitor cells, such as intestinal-, renal-, or pulmonary-specific progenitor cells, for example by the expression of organ-specific markers. The skilled person in the art will be familiar with organ-specific markers that may be used for identifying and/or isolating organ-specific progenitor cells. For example, co-expression of: SOX2 and ASCL2, MUC2, FABP1, LRIG1 for intestine, co-expression of: PAX8 or LHX1 for kidney, and at least two of: SOX2, SOX9, NKX2-1, FOX2A for lung. Specific epithelial stem cell subtypes may be isolated for seeding, for example for deriving organoids of a specific region of the organs. For example for the kidney: nephron progenitor cells generating proximal tubule and distal tubule organoids or ureteric bud progenitor to derive collecting duct organoids. In one embodiment, the organoid may comprise a neurosphere organoid. A neurosphere is a culture system composed of free-floating clusters of neuroepithelial stem/progenitor cells. The epithelial stem cells may be neuroepithelial stem/progenitor cells. In one embodiment, only one type of organ-specific progenitor cells (e.g. intestinal-, renal-, neural- or pulmonary-specific progenitor cells) may be isolated and seeded for formation into a tissue-specific primary fetal organoid. In one embodiment, the epithelial stem cells may not be cell sorted, for example by FACS/MACS. In particular, the skilled person will recognise that the fetal fluid may provide sufficient purity and levels of viable epithelial stem cells, which may be used to seed the ECM without cell sorting. In an embodiment wherein the fetal fluid is tracheal fluid, the epithelial stem cells of the tracheal fluid may be isolated for example by a cell wash step (e.g. centrifugation and resuspension in a buffer or medium), and seeded into the ECM. The epithelial stem cells may be concentrated, for example by centrifugation and resuspension, prior to seeding. Advantageously, the centrifugation and resuspension step allow the concentration of cells to be controlled or enhanced, and allows for the elimination of undesired cells (i.e. immune, red blood cells), debris, and mucous. This further enhances cell viability and purity. One or more epithelial stem cells may be isolated from the fetal fluid. In one embodiment, at least about 50 epithelial stem cells may be isolated from the fetal fluid for seeding. In another embodiment, at least about 500 epithelial stem cells may be isolated from the fetal fluid for seeding. In another embodiment, at least about 5000 epithelial stem cells may be isolated from the fetal fluid for seeding. In another embodiment, at least about 10,000 epithelial stem cells may be isolated from the fetal fluid for seeding. Step ii): Seeding the population of cells comprising the epithelial stem cells into an ECM and culturing the seeded cells in epithelial medium The ECM may be in the form of droplets, for example with the epithelial stem cells seeded therein. The skilled person will recognise that the volume of the ECM to be used may be decided based on the number of isolated cells for seeding and/or the type of culture container, such as a multi-well Petri dish, used. The seeding may be in 96, 48, 24, 12, or 6 multi-well flat bottom plates, or 96 well V-bottom or U-bottom plates, depending on the experiment to be performed. In one embodiment, the ECM, e.g. in the form of droplets, is about 5-50µl in volume. In another embodiment, the ECM, e.g. in the form of droplets, is about 5-10µl in volume. In another embodiment, the ECM, e.g. in the form of droplets, is about 5-30µl in volume. Advantageously, the dome shape provided by a droplet allows the creation of a 3D microenvironment where the cells can subsequently grow as 3D organoids. Further Advantageously, the volume/size of the ECM droplet may be sufficiently small such that the diffusion of the medium nutrients is not significantly impaired into the centre of the ECM droplet, causing the suffering of the cells. The cells may be plated to a density that does not permit cell-to-cell contact during culture. The epithelial stem cells may be seeded in an amount up to about 6x104 live cells/ECM substrate, such as an ECM droplet. In one embodiment, the epithelial stem cells may be seeded in an amount up to about 6x104 live cells/30µl ECM droplet. In one embodiment, the epithelial stem cells may be seeded in an amount of about 5000 live cells/ECM droplet, such as an 30µl ECM droplet. In another embodiment, the epithelial stem cells may be seeded in an amount of about 50 to about 6x104 live cells/ECM droplet, such as an 30µl ECM droplet. In another embodiment, the epithelial stem cells may be seeded in an amount of about 50 to about 6x104 live cells/ECM droplet, such as an 5µl ECM droplet. In another embodiment, the epithelial stem cells may be seeded in an amount of about 50 to about 1000 live cells or more, per ECM droplet, such as an 5µl ECM droplet. In another embodiment, the epithelial stem cells may be seeded in an amount of about 50 to about 1x104 live cells/ECM droplet, such as an 30µl ECM droplet. In another embodiment, the epithelial stem cells may be seeded in an amount of about 50 to about 6000 live cells/ECM droplet, such as an 30µl ECM droplet. The skilled person will recognise that where the volume of the droplet is adjusted, the number of cells may equally be adjusted to maintain substantially the same concentration of cells per µl of ECM droplet. The extracellular matrix (ECM) may be a hydrogel. The ECM may be a mixed matrix hydrogel. In one embodiment, the mixed matrix hydrogel comprises proteins such as one or more of laminin, entactin, nidogen, collagen and heparan sulfate proteoglycans. In one embodiment, the ECM comprises a natural polymer, such as collagen and/or fibrin. The ECM may comprise collagen, connective tissue glycoproteins and proteoglycans. In another embodiment, the ECM may be a synthetic hydrogel, for example comprising or consisting of synthetic peptides, or peptide/adhesion functionalized polysaccharides. In an alternative embodiment, the ECM may be a Basement Membrane Extract (BME), such as Matrigel™, Geltrex™ and Cultrex® BME, or a generic equivalent thereof. The BME may comprise or consist of basement membrane purified from Engelbreth-Holm-Swarm (EHS) tumor. The BME may comprise or consist of laminin, collagen IV, entactin, and heparin sulfate proteoglycans. The ECM may comprise or consist of a gelatinous protein mixture derived from cells, such as tumour cells. In one embodiment, the ECM comprises or consists of ECM gel derived from decellurized tissues. In one embodiment, the ECM comprises or consists of solubilized basement membrane matrix secreted by Engelbreth-Holm-Swarm (EHS) mouse sarcoma cells. In one embodiment, the ECM comprises or consists of Matrigel®, Cultrex® BME, Geltrex® Matrix, LunaGel™, or a generic equivalent thereof. In one embodiment, the ECM comprises or consists of Extracellular matrix derived hydrogels of animal origin. The animal origin ECMs that may be used in the present invention are described by Giobbe et al. (Nature Communications volume 10, Article number: 5658 (2019)0, or in patent publication number WO2021089472, which are herein incorporated by reference. The ECM may further comprise growth factors, such as TGF-beta and/or EGF. Culturing The culturing of the seeded cells may be for a period until single clonal organoids are formed, or until formed single clonal organoids reach an average size of at least about 50 µm. The size of the single clonal organoid is understood to be the average of the largest diameter of the single clonal organoids. The culturing of the seeded cells to form single clonal organoids may be for about 1-3 weeks, which may depend on the proliferation potential of the cells. Medium may be changed one or more times during the culture period. Medium may be changed periodically, for example to maintain an adequate level. In one embodiment, medium is changed about every 3-4 days. The cell growth medium at step ii) The cell growth medium may not comprise components, such as cytokines, that provide tissue- specific signals for differentiation. In one embodiment, the cell growth medium is not tissue specific. Preferably the cell growth medium allows the formation of organoids, without providing specialised tissue-specific signals. The cell growth medium may be epithelial medium. The skilled person will understand that epithelial medium may be cell growth medium which is capable of supporting the maintenance or growth of epithelial cells. The epithelial medium may comprise the AFO expansion medium as described herein (e.g. as described in Supplementary Table 2). In another embodiment, the cell growth medium may be dependent on the type (i.e. tissue type) of the organoid being cultured. For example, the cell growth medium may be appropriate for lung epithelial cells, when culturing lung primary fetal organoids. In another embodiment, the cell growth medium may be appropriate for kidney epithelial cells, when culturing kidney primary fetal organoids. In another embodiment, the cell growth medium may be appropriate for intestinal epithelial cells, when culturing intestinal primary fetal organoids. In another embodiment, the cell growth medium may be appropriate for neuronal cells, when culturing neurosphere organoids. For example, cell growth medium may comprise neurobasal medium. The cell growth medium appropriate for neuronal cells may comprise the neurosphere medium described herein. In an embodiment wherein the fetal fluid is tracheal fluid, the epithelial stem cells of the tracheal fluid may be cultured in medium appropriate for lung organoid development, such as the human fetal lung organoid medium described herein (e.g. as described in Supplementary Table 2). The cell growth medium may be supplemented with Rho-kinase inhibitor, such as ROCKi. One example of a ROCKi that may be used is Y-27632. The concentration of ROCKi may be in an amount sufficient to inhibit Rho-kinase activity, such as about 10μM. The concentration of ROCKi may be from about 5μM to 20μM. The cell growth medium may be further supplemented with one or more antibiotics, such as a broad spectrum antibiotic. The antibiotic may be one of the antibiotics as described herein. Step iii): Extracting a single clonal organoid from the ECM and dissociating the single clonal organoid into single cells Single clonal organoids may be extracted from the ECM and resuspended in medium for further culturing. To form clonal organoid lines, single clonal organoids may be picked (e.g. manually picked) from the ECM. The single clonal organoids may be picked at passage 0. The single clonal organoids may be extracted at any suitable time, such as from day 7-21, or more. The single clonal organoids may be extracted at any suitable time, such as from day 7 or more, such as for as long as the cells of the single clonal organoids are viable and no noticeable contact between two organoids can be observed (e.g. the clonal organoids are discreet from each other). Single clonal organoids may be extracted from the ECM by depolymerising (melting) the ECM, for example by using a cell recovery solution that enables the recovery of cells/organoids cultured in an ECM, such as Matrigel™. Each single organoid may be collected, for example in different eppendorf tubes, and washed in medium. The washing may remove any ECM leftover and/or dead/single cells. Single clonal organoids may alternatively be extracted from the ECM by physically/mechanically scraping them from the ECM and/or the extracted single clonal organoids may be suspended in a medium, such as cell culture or basal medium, and centrifuged to separate the single clonal organoids from the ECM and form free single cells. After picking/extracting, the clonal organoid can be subject to disaggregation to single cells, which may be manual and/or enzymatic disaggregation (otherwise termed “dissociation” herein). The enzyme may be trypsin, trypLE, or accutase, or other enzymes capable of aiding the dissociation of cells; or combinations thereof. In one embodiment the clonal organoid is subject to enzymatic disaggregation, together with mechanical disaggregation, for example by pipetting. For example, a clonal organoid may be transferred to an incubation tube, such as a 0.5ml tube, and optionally the tube pre-coated with BSA or other agent capable of reducing the adherence of cells to the surface of the tube. The BSA may comprise 1% BSA. The clonal organoid may be resuspended in an enzyme composition capable of dissociating the cells of the clonal organoid to single cells, optionally with manual manipulation, such as pipetting. Following dissociation, the single cells may be washed, such as by a centrifugation and resuspension in a buffer or medium. Washing may be repeated, such as at least twice, for example to inhibit the enzymatic activity due to treatment for disaggregation. Step iv: Culturing the single cells such that they expand into tissue-specific primary fetal organoids. The single cells may be cultured in ECM. The ECM may be as described herein, for example in accordance with the ECM used in step ii). In one embodiment, the single cells are resuspended in the ECM and plated, such as in a multi-well plate (e.g. a 24 or 48 well plate), with a cell growth medium, which may be a tissue-specific medium. In one embodiment, the seeded ECM is cultured in cell growth medium, which may be supplemented with Rho-kinase inhibitor, such as ROCKi. In one embodiment, the cell growth medium is supplemented with Rho-kinase inhibitor, such as ROCKi, for at least the first 3 days of culture. One example of a ROCKi that may be used is Y-27632. The concentration of ROCKi may be in an amount sufficient to inhibit Rho-kinase activity, such as about 10μM. The concentration of ROCKi may be from about 5μM to 20μM. The single cells may be cultured for about 10 days. The medium may be replaced one or more times, such as periodically. In one embodiment, medium is replaced about every 3-4 days. Following culture, growing primary fetal organoids may be passaged. Passaging of primary fetal organoids The primary fetal organoids may be passaged after clonal expansion, for example to a new incubation plate, such as a 12 or 24 well plate. The primary fetal organoids may be split, such as 1:2 to 1:10. The splitting may be after about 7–14 days of culture. Passaging may comprise washing of the primary fetal organoids. The washing may comprise the removal medium and disruption/dissolution of any ECM. The primary fetal organoids may be centrifuged and resuspended in buffer or medium. The primary fetal organoids may be mechanically and/or enzymatically disaggregated, for example using a pipette and/or an enzyme such as trypsin, trypLE, or accutase, or other enzymes capable of aiding the dissociation of cells; or combinations thereof. The cells of the disaggregated primary fetal organoids may be further washed and resuspended in ECM, such as Matrigel™ (or a collagen comprising equivalent thereof), and plated for further incubation. The ECM may be as described herein, for example in accordance with the ECM used in step ii). The single cells of the primary fetal organoids may be cultured for about 10 days. The medium may be replaced one or more times, such as periodically. In one embodiment, medium is replaced about every 3-4 days. The cell growth medium may be supplemented with Rho-kinase inhibitor, such as ROCKi. The cell growth medium may be supplemented with Rho-kinase inhibitor, such as ROCKi, for at least the first 3 days of culture. One example of a ROCKi that may be used is Y-27632. The concentration of ROCKi may be in an amount sufficient to inhibit Rho-kinase activity, such as about 10μM. The concentration of ROCKi may be from about 5μM to 20μM. Organoid maturation / differentiation The primary fetal organoids may be further matured and/or differentiated, for example into a tissue/organ type, such as lung, kidney or intestine. The primary fetal organoids may be further matured and/or differentiated into neuronal specific cell types. The primary fetal organoids may be further matured and/or differentiated into small intestinal AF organoids (siAFO), kidney AF organoids (kAFO), lung AF organoids (lAFO), skin organoids, placental organoids, corneal organoids, oesophageal organoids, gastric organoids, large intestine organoids, or lacrimal gland organoids. In one embodiment, the primary fetal organoids are further matured and/or differentiated into small intestinal AF organoids (siAFO), kidney AF organoids (kAFO), or lung AF organoids (lAFO). To form small intestinal AF organoids (siAFO) after passaging, organoids may be seeded in ECM, such as Matrigel™ (or a collagen comprising equivalent thereof), and cultured in medium, such as generic organoids culture medium, e.g. AFO expansion medium described herein (see s.Table 2). After about 7 days, human small intestine medium may be used (e.g. as described in supplementary Table 2) for a period until siAFO are formed, such as for about 14 days. Notch inhibitor (such as DAPT at about 10 ^m) may be added to the basal culture medium to stimulate differentiation. The organoid may be exposed to the notch inhibitor for a period of about 48hrs. The basal culture medium may not comprise glycogen synthase kinase 3 inhibitor, such as CHIR99021. To form kidney AF organoids (kAFO) after passaging, organoids may be seeded in ECM, such as Matrigel™ (or a collagen comprising equivalent thereof), and cultured in medium, such as generic cell culture medium. After about 7 days, distal/collecting duct kidney differentiation medium may be used (e.g. as described in supplementary Table 2) for a period until kAFO are formed, such as for about 14 days. To form lung AF organoids (lAFO) after passaging, organoids may be seeded in ECM, such as Matrigel™, and cultured in medium, such as generic cell culture medium. After about 7-10 days, lung proximal differentiation may be used, such as PneumaCult™ ALI Medium (StemCell Technologies, #05001). For distalisation into pulmonary organoids with a distal (alveolar) identity, human distal lung medium (e.g. as described in supplementary Table 2) may be used, such as for about 14 days. Identifying cell and organoid types Proliferating cells may be identified for example by staining and/or studying their cell marker expression, such as proliferative marker Ki67. The lack of apoptotic cell death may be identified for example by staining the cells and/or studying their cell marker expression, such as caspase 3. The epithelial identity of the primary fetal organoids may be confirmed, for example by immunofluorescent staining and/or studying their cell marker expression. Immunofluorescent staining may be used for identifying standard pan-epithelial human cell markers, such as one or more of EpCAM, ECAD, and Pan-Cytokeratin. Additionally or alternatively, the identity of the primary fetal organoids may be confirmed by sequencing, for example by RNA sequencing. The primary fetal organoids may demonstrate upregulation of pathways related to the tissues/organ type. Small intestinal primary fetal organoids may be identified by the formation of crypt, or crypt- like, structures, for example after expansion over several passages, such as over 10 passages. Small intestinal primary fetal organoids may express one or more typical intestinal stem/progenitor cell genes, such as one or more of LGR5, OLMF4, LRIG1, and SMOC2, as well as one or more paneth (LYZ), goblet (MUC2, CLCA1) and endocrine (CHGA) cell markers. Enterocyte cell markers, such as one or more of ALPI, FABP1, VIL1, EZR, KRT20, and ATP1A1, may also be expressed by small intestinal primary fetal organoids. Small intestinal primary fetal organoids may also lack lung and kidney specific markers NKX2-1 and/or PAX8, respectively. Kidney tubule AF-derived organoids (kAFO) may be identified by RNA sequencing with colocalization with the fetal tissue-derived kidney organoids control PCA cluster. Kidney primary fetal organoids may be identified by their morphology, for example after expansion over several passages, such as over 10 passages. Additionally or alternatively, kidney primary fetal organoids may be identified by the presence of renal markers, such as one or more of canonical renal epithelial development/progenitor and nephron progenitor-specific genes, for example PAX2, PAX8, LHX1, and JAG1; expression of one or more distal tubule genes, for example PCBD1, SLC41A3, and POU3F3; and/or expression of one or more proximal tubule markers, for example ABCC1, ABCC3, ABCC4, and CUBN. Additionally or alternatively, kidney primary fetal organoids may be identified by the expression of collecting duct marker GATA3. The kidney primary fetal organoids may manifest a tubuloid-like phenotype and may comprise markers belonging to multiple segments of the renal tubules. Kidney primary fetal organoids may express one or more of the renal epithelium progenitor markers PAX8 and/or LHX1; kidney segment-specific protein markers such as GATA3 and/or ECAD (distal tubule/collecting duct), and Lotus tetragonolobus lectin (LTL, proximal tubule). Kidney primary fetal organoids may comprise polarised tubular microvilli, for example confirmed by immunofluorescence of acetylated tubulin. Kidney primary fetal organoids may be functionally assessed for kidney associated functions. For example, functional potassium channels may be confirmed, such as by demonstrating thallium intake following addition of a voltage-gated potassium (K+) ion channel stimulator, The kidney primary fetal organoids may be further differentiated to promote maturation of the distal/collecting duct lineage. The differentiation may comprise stimulation with vasopressin and arginine aldosterone. The stimulation may be for a period of time sufficient to promote a morphological change and expression of markers of the principal cells of the collecting duct (e.g. AQP2) and of the distal tubules (e.g. SLC12A1 and/or CALB1). The period of time of the stimulation may be about 10 days or more, preferably 14 days or more. Proliferating cells may be identified by expression of Ki67 and/or EDU/BRDU. Lung AF-derived organoids may be identified by expression of stem/progenitor cell markers (such as one or more of NKX2-1, FOXA2, SOX2, SOX9, TP63, GATA6), as well as of both Alveolar Type 1 (such as one or more of HOPX, PDPN, AGER, AQP5) and Alveolar Type 2 cells-related genes (such as one or more of SFTPA1, SFTPA2, SFTPB, SFTPC, SFTPD, ABCA3, LAMP3). Additionally or alternatively, lung AF-derived organoids may be identified by expression of the stem cell markers NKX2-1, SOX2, and the basal cell marker P63. The lAFO may undergo terminal proximal and/or distal lung differentiation. Proximal differentiation may be determined by the appearance of a polarised epithelium with motile cilia on the luminal surface of the organoids. Immunofluorescent staining on the proximally differentiated lAFO may show the presence of Ac- ^TUB positive cilia on the luminal side of the organoids, confirming both differentiation and polarisation of the epithelia. The nuclear expression of the ciliated epithelia marker FOXJ1 may further be determined. In addition, the appearance of the mature basal cell marker keratin 5 (KRT5), and secretory marker mucin (MUC5AC), concomitantly with the maintenance of SOX2, may indicate the active proximal lung differentiation process. Additionally or alternatively, lung AF-derived organoids may be identified by increased expression of airway markers such as FOXJ1, TUBA1A and KRT5. Lung AF-derived organoids may be pushed towards a distal phenotype, which may be identified by increased protein expression of the AT2 cell marker SFTPB and/or observation of lamellar bodies with a normal structure and a core composed of multi-lamellar membranes, which are typical features of distal lung cells. In one embodiment, a lack of mesenchymal features in the primary fetal organoids may be demonstrated by determining the absence of PDGF receptor alpha expression. The formation of polarised epithelium in the primary fetal organoids may be determined by the presence of one or more of basolateral integrin b4 (ITG ^4), apical F-Actin and the tight junction marker ZO-1, for example on the luminal side of the primary fetal organoids. Identity of the organoids may be confirmed by a combination of cell marker expression and phenotypic observations, for example those described herein. The fetal fluid, epithelial stem cells and primary fetal organoids may be mammalian, preferably human. In an alternative embodiment, the fetal fluid, epithelial stem cells and primary fetal organoids may be from a non-human animal, such as simian, rodent, equine, bovine, porcine, canine, fish, bird or feline. The non-human animal may be a sheep or mouse. The primary fetal organoids may be primary autologous fetal organoids. Culturing the cells or primary fetal organoids described herein may be under suitable cell growth conditions, for example at 5% O2, 5% CO2, and 37°C. Other Aspects According to another aspect of the present invention, there is provided a primary fetal organoid formed by the methods described herein. According to another aspect of the present invention, there is provided a disease model, wherein the disease model comprises a primary fetal organoid according to the invention that has been treated with an agent to induce a disease state in the primary fetal organoid. The disease may be a congenital disease, such as Congenital Diaphragmatic Hernia (CDH). Congenital diseases may be studied by deriving the primary fetal organoids from fetal fluid of a fetus having, at risk of, or suspected of having, the congenital disease. Pathological features of the disease may be studied and/or differences to non-congenital disease controls may be determined. The fetal organoids may be derived from the fetal fluid of fetuses with a congenital disorder, such as Congenital Diaphragmatic Hernia (CDH), spina bifida, oesophageal atresia, twin to twin transfusion syndrome, etc. (as listed, but not limited to Supplementary Table 1). In one embodiment the primary fetal organoids may be derived from the fetal fluid of Congenital Diaphragmatic Hernia (CDH) fetuses. The primary fetal organoids may further comprise cells from a healthy donor, which may be engrafted. In one embodiment, an agent is used to induce a disease state in the primary fetal organoid. In one embodiment the agent comprises or consists of a growth factor or cytokine. The agent, such as a cytokine. In another embodiment, disease or disorder may be induced genetically, such as by genetic modification, genetic overexpression, or by siRNA silencing, or by treatment with other proteins or pharmacological agents. The amount of agent and incubation time with the agent may be an amount and time sufficient to cause disorder in the primary fetal organoid. According to another aspect of the present invention, there is provided the use of the disease model to identify agents capable of preventing or treating the disease, wherein the primary fetal organoid is treated with a potential agent before, during or after the primary fetal organoid is treated with a disease-inducing agent. In another embodiment, disease induced by an agent may be prevented by pharmacological or genetic manipulation of the primary fetal organoid. The use may comprise the determination of whether the treatment by the potential agent has any effect in inhibiting or reducing the development of disease, or the reduction in disease after it has developed. According to another aspect of the present invention, there is provided a method of screening for agents capable of preventing or treating a disease, for example using the primary fetal organoid described herein, wherein the primary fetal organoid is genetically manipulated or treated with a potential agent before, during or after the primary fetal organoid is treated to induce disease; and determining if the potential agent or genetic target has any effect in inhibiting or preventing the development of disease in the primary fetal organoid, or the reduction in disease after it has developed in the primary fetal organoid. The agent to be screened or investigated may be dosed at a physiological relevant amount. The agent to be screened or investigated may be dosed at a therapeutically relevant amount. Combinations of agents may be investigated. The determination may be relative to an untreated primary fetal organoid (i.e. not treated with the potential agent) and/or relative to a control or reference value. In one embodiment the agent to be investigated is a small molecule (e.g. less than 900Da), nucleic acid, antibody therapy, cellular therapy, drug compound, metabolite or peptide. In one embodiment the agent to be investigated is a small molecule (e.g. less than 900Da), nucleic acid or peptide. The peptide may comprise or consist of an antibody. In another embodiment the agent to be investigated is a genetic manipulation agent, such as siRNA, shRNA, CRISPR- CAS9, lentiviral or retroviral vectors, for example for over expression. A method is provided for seeding of the organoids with cells from a donor, and tracking the cells to assay one or more of survival, proliferation and isolation of engrafted cells, for example for downstream functional testing. The cells maybe tracked by fluorescent markers or tags. Tracking the cells may comprise the use of a fluorescent cell tracking system. In one embodiment, the primary fetal organoids according to the invention may be used to produce a 3D culture model, for example a model resembling a disease affecting a donor of the epithelial stem cells. According to another aspect of the present invention, there is provided a method for producing a product from the primary fetal organoids according to the invention, the method comprising the incubation of the primary fetal organoids in vitro, and harvesting the product produced from the primary fetal organoids. The product may be produced naturally by the primary fetal organoids or induced. The product may be harvested by separating them from the cells of the primary fetal organoids. The product may comprise cells, tissue, proteins, nucleic acids or other biological molecules. According to another aspect of the present invention, there is provided a method of screening for biomarkers of a disorder in an organ or tissue, the method comprising the monitoring of biomarkers released from the primary fetal organoids or cells engrafted therein, or biomarkers in tissue or cellular extracts of the primary fetal organoids. The disorder may be a disease state. The disorder may be an infection, genetic disorder, immunological response, cancer state, biological pathway irregularity, or biochemical irregularity. A disorder may develop, or be induced in the primary fetal organoid, whereby changes to the biomarker profile may be determined and linked to the disorder. The biomarkers may comprise proteins, glycoproteins, glycans, peptides, nucleic acids, or any cellular product which may indicate a disorder of the primary fetal organoid or engrafted cells therein. The biomarkers may be cell markers, such as surface proteins and/or secretions. In another aspect, there is provided a kit comprising one or more, such as all of, the cell culture mediums described herein, and/or the ECM, which are required to carry out any method disclosed herein. The kit may comprise recombinant growth factors and/or cytokines. The recombinant growth factors and/or cytokines may be provided in combined solutions in the kit, or two, three, four, five, six, seven or more separate solutions in the kit. The kit may further comprise a set of instructions. The instructions will enable the reader to perform any method disclosed herein. The methods herein may employ any step of any other method recited herein. In yet a further aspect, there is provided cells obtained by any method disclosed herein, such as one or more specific cell types that may be harvested from the fetal primary organoids according to the invention. The organoids described herein may be further differentiated into a tissue type, and assembled further into assembloids. Therefore, according to another aspect of the present invention, there is provided an assembloid comprising two or more organoids, wherein at least one organoid is according to the invention herein, or is produced by the method of the invention herein. According to another aspect of the present invention, there is provided a method of generating an assembloid, the method comprising incubating two or more different organoids, wherein at least one of the organoids comprises an organoid of the invention described herein, or is produced by the method of the invention. One or more of the organoids may be further differentiated into a tissue type, such as proximal lung prior to, or after, assembling into an assembloid with the same tissue type or a different tissue type or organoid. In some embodiments, for example for generating a neurosphere organoid, the method of the invention may be used, but without seeding an ECM. Therefore, according to another aspect of the present invention, there is provided a method of generating tissue-specific primary fetal organoids from an epithelial stem cell population wherein said method comprises: i) isolating a viable population of cells comprising epithelial stem cells from a fetal fluid; ii) culturing the cells in a cell growth medium for the formation of organoids from the epithelial cells; and optionally iii) further passaging and culturing the organoids for expansion, or extracting an organoid from the cell culture and dissociating the organoid into single cells for passaging, and culturing the single cells such that they expand into tissue-specific primary fetal organoids. Definitions An “assembloid” is intended to refer to an in vitro model that combines two or more organoids, spheroids, or cultured cell types to recapitulate structural and functional properties of an organ. References to “inhibition” or similar, may comprise a reduction in activity or presence of a molecule or the block of a biological pathway, such as a signaling pathway. The inhibition may be total (i.e.100%) or at least a substantial inhibition. The inhibition may be partial inhibition. Partial inhibition may comprise significant inhibition in order to affect the desired outcome of the inhibition. It is understood that “epithelial stem cell” refers to a type of stem cell that is responsible for generating and maintaining the epithelium. It is understood that “chemically defined medium” is a growth medium suitable for the in vitro cell culture of human or animal cells in which all of the chemical components are known. As used herein, the term “organoid” may be taken to describe a self-renewing 3-dimensional, multi-lineage cellular structure that resembles a specific organ or tissue. Unlike naturally occurring organs and tissue, organoids are grown ex vivo, in miniature and in multiples, typically forming multiple structures that are a micrometers to millimeters in diameter, such as the size of between 50 µm and 5 mm, or between 800 µm and 2 mm. A key distinguishing feature of organoids is that they are grown reproducibly with multiple replicates, enabling experimental studies. The organoids can contain cell types that are transcriptionally representative of specific tissues or organs. The skilled person will recognise that the term “organoid” may alternatively be termed “spheroid” or “microtissue”. As used herein, the term “primary fetal organoid” may be taken to describe an organoid that is directly derived from primary cells of a fetus, without the need for reprogramming. The skilled person will understand that optional features of one embodiment or aspect of the invention may be applicable, where appropriate, to other embodiments or aspects of the invention. Embodiments of the invention will now be described in more detail, by way of example only, with reference to the accompanying drawings. Figure 1 – Single cell analysis of the AF content. (a) Graphical representation of the AF sampling (top). The bottom plot shows the sorting strategy utilised to collect the living cell fraction, negative for Propidium Iodide and positive for Hoechst. The UMAP shows the content of the amniotic fluid of multiple patients obtained across the second trimester of pregnancy (n=11 patients 15-34GA; 40919 single cells are presented). The UMAP shows the content of the amniotic fluid of multiple patients obtained across the second trimester of pregnancy (n=11 patients 15-34GA; 36726 single cells analysed). Highlighted in orange the epithelial cluster, as identified by the SingleR cell labelling package using the human primary cell atlas dataset as reference. The violin plots show the level of expression of the pan-epithelial specific genes EPCAM, CDH1(ECAD), KRT8, KRT10, KRT17 and KRT19 (data presented as normalised counts per million, CPM). (b) The UMAPs show the expression of the same epithelial markers, within the epithelial cluster identified in a. (c) Representative flow cytometry analysis of EPCAM and ECAD (CDH1) expression in live-sorted cells from the AF, grey line represents unstained controls. (d) The violin plots highlight the occurrence, and level of expression of intestinal, renal and pulmonary cell markers within the epithelial cell cluster highlighted above (data presented as normalised CPM). (e) The UMAPs highlight in red, the presence of cells co-expressing tissue specific epithelial stem/progenitor cells markers: SOX2 and ASCL2, MUC2, FABP2, LRIG1 (Intestine), PAX8 or LHX1 (kidney), and two of SOX2, SOX9, NKX2-1, FOX2A (Lung). Figure 2 – Generation of primary epithelial fetal AFO. (a) Phase contrast images showing the formation of the organoids starting from 3D cultured viable AF cells. The growth of full-size organoids of different morphologies can be observed at day 14 (scale bar: 200 ^m). (b) Analysis of the formation efficiency (formed organoids/viable cells plated) and of the size (area) of the organoids at isolation (Passage 0) (n=18 AF samples analysed for the efficiency plot and n=128 organoids for area plot; median and quartiles). Dot plot also shown representing organoid generation efficiency (organoids formed / live cells plated) at various gestational ages, with colour and size representing total number of organoids generated per sample. R2 = 0.03 and standard error shown. (c) Phase contrast images showing the multiplicity of morphologies presented by clonal AFO in expansion, from Passage 1 to 5 (scale bars: 200 ^m). (d) Quantification of the formed organoids per field of view at 7-15 days of culture quantified over 5 passages (ns: non-significant; n =33 organoids from ^8 AF samples, median and quartiles; 2-way ANOVA with multiple comparison). (e) X-ray phase contrast computer tomography (PC-CT) of two of the organoids’ phenotypes observed, parenchymatous and cystic (scale bars: 25 ^m). (f) The immunofluorescent staining shows expression of the proliferative marker Ki67 and lack of apoptotic cells stained with cleaved caspase 3 in the AFO at P3. Counterstain with Hoechst was used to highlight the localisation of the cell nuclei. (Scale bars: 50 ^m). (g) Whole mount immunofluorescent staining showing AFO at P3 expressing the epithelial markers EpCAM, ECAD and pan-cytokeratin; together with the lack of the mesenchymal marker platelet-derived growth factor receptor ɑ (PDGFRɑ); the immunofluorescent staining also shows AFO’s polarisation, highlighted by the presence of the epithelial tight junction protein zonula-occludens 1 (ZO-1) on the luminal surface, and integrin beta 4 (ITGβ4) on the basolateral layers. Actin filaments are also displayed by staining with Phalloidin (F-ACT). (h) Principal component analysis plot showing the unsupervised clustering of the organoids into three main clusters (n=105 organoid lines from n=17 AF samples). These cluster show co-localisation with primary fetal tissue-derived control organoids (n=11) produced from Lung (Yellow), Small Intestine (Cyan) and Kidney (purple) biopsies. (i) Gene ontology analysis of top 10 upregulated pathways in each cluster compared to rest of the dataset, after removal of the fetal tissue-derived controls. Figure 3 – Generation and maturation of small intestine AFO. (a) Phase contrast images depicting the expansion of SIAFO from passage 0 to passage 5 (scale bar: 200 ^m). The immunofluorescent staining for EdU (P5) shows the localisation of proliferating cells at the basis of crypt-like structures (scale bar: 50 ^m). (b) The dot plot shows the gene expression analysis performed on the small intestinal organoids using bulk RNA sequencing showing the presence of all the major small intestine markers in the siAFO (n=2 patients, n=4 siAFO lines in intestinal maturation medium, n=2 siAFO rings, n=4 fetal-derived small intestinal organoids as control). (c) Immunofluorescent staining for the intestinal crypt stem cell marker olfactomedin 4 (OLFM4), villi enterocyte marker cytokeratin 20 (KRT20) and luminal integrin-β4 (ITGβ4). The presence of Paneth cells and enterocytes is also highlighted in the immunofluorescent staining for lysozyme (LYZ), epithelial cadherin (ECAD) and fatty acid binding protein 1 (FABP1) respectively; (scale bar: 50 ^m). (d) Quantification of positive cell type-specific markers siAFO (n=2 independent biological samples; ^4 organoids per sample; mean ± SEM). (e) Functional assessment of Dipeptidyl peptidase IV enzyme in the siAFO; nitroaniline production was monitored within 1 hour of exposure to Gly-Pro p-nitroanilide hydrochloride (n=2 independent biological samples; n=5 clonal organoid lines at passages 6 and 10; n=1 pediatric small intestinal ileal organoid as a positive control; mean ± SEM). (f) Phase contrast image depicts the siAFO after maturation, showing budding morphology (scale bar: 200 ^m). The immunofluorescent stainings confirm the progression of the maturation by the appearance of Chromogranin A (CHGA)-positive enteroendocrine cells and Mucin 2 (MUC2)-positive secretory cells. Counterstaining with F-actin (F-ACT) and Hoechst shows organoids’ lumen and nuclei respectively (scale bar: 50 ^m). Figure 4 – Generation and differentiation of kidney AFO. (a) Phase contrast images depicting the establishment of KAFO starting from passage 0 up to long-term expansion at passage 7 (scale bar: 200 ^m). The immunofluorescent staining on the right highlights the presence of the proliferative marker Ki67 (scale bar: 50 ^m). (b) The dot plot shows the RNAs sequencing analysis showing the presence of a broad range of kidney markers in the kAFO (n=16 patients, n=2 lines in differentiation medium, n=3 independent fetal kidney-derived organoids as a control). (c) Immunofluorescent staining shows presence of developmental kidney nephron progenitor markers PAX8 and LHX1, counterstained with F-actin (F-ACT). The panel also shows positivity for the distal tubule/collecting duct marker GATA3, for the proximal tubule marker LTL, as well as the presence of ECAD and apical Ac- ^TUB-positive (acetylated ^-tubulin) cilia confirmed the renal epithelial identity of the kAFO. (Scale bar: 50 ^m). (d) The bar graph shows quantification for the early developmental renal markers PAX8, LHX1, and GATA3 in kAFO (n=3 independent biological samples, ^4 organoids per sample; mean ± SEM). (e) Potassium ion channel assay was performed on n=2 kAFO independent biological samples, n=1 fetal kidney organoids, n=1 fetal lung organoids as negative control; medium fluorescence intensity was calculated. (f) Phase contrast image and immunofluorescent staining highlighting the morphological changes, as well as the expression of the mature renal markers AQP2, SLC12A1 and CALB1 in kAFO exposed to a renal differentiation medium (scale bars: 200 ^m and 50 ^m). (g) Quantification of CALB1 positive cells in kAFO cultured in expansion medium (EM) compared to differentiation medium (DM) (n=2 biological samples, mean ± SEM). (h) Bar graph showing CALB1 gene expression of differentiated kAFO compared to undifferentiated controls (CTR) based on the RNAseq plot presented in b (n=2 differentiated independent biological samples; mean ± SEM; ****P<0.0001 with unpaired t-test). Figure 5 – Generation and differentiation of lung AFO. (a) Phase contrast images depicting the establishment of lAFO from passage 0 up to long-term expansion at passage 5 (scale bar: 200 ^m). The immunofluorescence on the right highlights the presence of the proliferative marker Ki67 (scale bar: 50 ^m). (b) The dot plot shows the gene expression analysis performed via RNA sequencing on the organoids (n=18 patients, n=4 differentiated lines, n=4 control tissue-derived primary fetal lung organoids). (c) The immunofluorescent stain highlighting the expression of the lung stem/progenitor cell markers NKX2-1 and SOX2. P63-positive basal cells are also present in the lAFO, while proSFTPC secreting cells are absent; images were counterstained with the structural marker F-Actin (F-ACT) and Hoechst, quantified in (d) (n=3 biological samples, ^4 organoids per sample; mean ± SEM) (Scale bar: 50 ^m). (e) Immunofluorescence staining was performed on lAFO after 14 days of exposure to proximal differentiation protocol. The occurrence of ciliation in the proximally differentiated lAFO is demonstrated by the polarised expression of the ciliary protein Ac- ^TUB, co-localised with the basal cell marker P63 and the ciliated cell marker FOXJ1. The panel also shows the expression of the mature basal cell marker KRT5, together with the occurrence of Mucin 5AC goblet cells and maintenance of SOX2 progenitor cells (Scale bar: 50 ^m). (f) Quantification of FOXJ1 and KRT5 positive cells within the lAFO in expansion (EM) versus proximal differentiation media (DM) (n=3 biological samples, ^4 organoids per sample; mean ± SEM; *P=0.0461 with unpaired t-test). (g) Bar graphs showing gene expression of FOXJ1, TUBA1A and KRT5 in proximally differentiated lAFO (DIFF) compared to undifferentiated controls (CTR), based on the RNAseq plot presented in b (n=2 biological sample; mean ± SEM; ***P=0.0003, ****P<0.0001 with unpaired t-test). (h) TEM imaging showing a cross section of proximal lAFO with cilia inside the lumen (red asterisk) (scale bar 2 ^m); In cross-section, the axonemes were normal in structure showing outer (blue arrow) and inner (white arrow) dynein arms (scale bars 100 nm). (i) Immunofluorescent staining displaying distalised lAFO showing the occurrence of surfactant-secreting cells (SFTPB); The two images represent two different SFTPB patterning observed in different organoids (granular – left, and luminally secreted – right). Nuclei were counterstained with Hoechst (Scale bar: 50 ^m). (j) Bar graph showing SFTPB gene expression of distalised lAFO (DIFF) compared to control in expansion (CTR) based on the RNAseq plot presented in b (n=2 biological samples; mean ± SEM; ns= nonsignificant). (k) TEM of distalised lAFO showing lumen (red asterisk) and cells containing lamellar bodies (red arrows); on the right, magnification of lamellar body containing multi-lammellar membranes (scale bars: 1 ^m and 500 nm respectively). Figure 6 – Generation, differentiation and characterisation of lung AFO and TFO from Congenital diaphragmatic hernia pregnancies. (a) Schematic of the amniotic and tracheal fluid sampling from CDH pregnancies. (b and c) Phase contrast images depicting the establishment of CDH AFO and TFO starting from passage 0 and expanding up to passage 3 and 5 respectively (scale bar: 200 ^m). The immunofluorescent staining panel highlights the positivity of the CDH organoids for the proliferative marker Ki67 and the lung stem/progenitor cell markers NKX2-1, SOX2, and the basal cell marker P63 (scale bar: 50 ^m); nuclei were counterstained with Hoechst. (d) Quantification of the organoids formation efficiency and area of CDH AFO vs. CDH TFO at isolation (n=9 CDH AF and n=5 CDH TF independent samples analysed for the efficiency plot and n ^56 organoids for area plot; median and quartiles; **P=0.0014 with unpaired t-test). (e) Quantifications of immunofluorescent stainings are presented in the bar plot (n=2 CDH AF samples, n=3 CDH TF samples, ^4 organoids per sample; mean ± SEM). (f) Immunofluorescent stain showing the expression of SOX9 exclusively in the CDH lTFO. It is also possible to observe the presence of the pro-surfactant protein C (proSFTPC) in both the CDH conditions, and its absence in the CDH lTFO derived after FETO. (g) Dot plot showing the expression of lung-associated markers in organoids derived from CDH patients’ tracheal (TF, 29- 34 GA, 3 patients) and amniotic (AF, 28-34 GA, 4 patients) fluids alongside GA- matched control lung organoids derived from non-CDH AF (27-34 GA, 3 patients). (h) Volcano plots showing differentially expressed genes (DEGs) amongst CDH organoids and GA-matched controls before and after FETO, significant (p-value < 0.01) lung- associated markers are also labelled. (i) Pathway analysis showing the 20 top up- and down regulated GO classes in CDH organoids vs. GA-matched controls (j) The panel depicts proximal lung differentiation of the CDH LAFO. The phase-contrast image shows proximally differentiated CDH LAFO that display presence of cilia, as highlighted by the immunofluorescence for Ac- ^TUB together with the ciliary transcription factor FOXJ1. Nuclei were counterstained with Hoechst (Scale bar for phase contrast: 200 ^m; immunofluorescence: 50 ^m). (k) Quantification of the ciliary beating frequency (Hz) performed using a high-speed camera (n ^22 cilia per organoids, n=1 CDH patient; n=2 non-CDH control patient; median and quartiles, ****P<0.0001 with unpaired t-test). (l) The immunofluorescent staining displays the distally differentiated CDH lAFO showing the occurrence of surfactant secreting cells (SFTPB); nuclei were counterstained with Hoechst (Scale bar: 50 ^m). Figure 7 – AF single cell RNA Sequencing: (a) Phase contrast image of fresh AF sample at collection. FACS gating strategy for purifying viable amniotic fluid cells and a further FACS plot showing heterogeneity of cells post-sorting (b) scRNAseq UMAP analysis of n=11 AF samples; middle UMAP shows second and third trimester labelling; right panel shows labelling by gestational age (GA) weeks. (c) UMAPs depict expression of general epithelial keratins within the AF epithelial cell cluster. (d) Upregulated pathways in the epithelial-labelled cluster of the scRNAseq AF from DEGs calculated when compared to all other clusters. Figure 8 – AFO derivation: (a) Percentage of samples that generated organoids at passage 0 (mean and 95% confidence interval). (b) Phase contrast images showing the recovery of AFO at passage 4 after cryopreservation and their expansion until passage 8. (c) Phase contrast images depicting two additional independent patients AFO lines (scale bar: 200 ^m). (d) First PCA shows all sequenced AFO with 3 unknown samples identified; the bottom PCA shows clustering of AFO based on their gestational age (GA). (e) Euclideanclustered heatmap confirming the tissue-typing labelling of the AFO. Figure 9 – siAFO: (a) Phase contrast images showing the thawing of siAFO at passage 6 and their expansion after cryopreservation until passage 10 (scale bar: 200 ^m). (b) Whole mount immunofluorescent staining showing the absence of NKX2-1 (lung) and PAX8 (kidney) in siAFO (scale bar: 50 ^m). (c) Whole mount immunofluorescent staining for Chromogranin A (CHGA) and Mucin 2 (MUC2) on siAFO in expansion medium (scale bar: 50 ^m). (d) RT–qPCR analysis of siAFO cultured in siAFO maturation medium without (Basal), with the GSK-3β inhibitor CHIR99021 (CHIR) or the Notch inhibitor DAPT (n=4 siAFO clonal lines from n=1 patient; mean } SEM; ***P=0.0005). Figure 10 – kAFO: (a) Phase contrast images showing the thawing of kAFO at passage 7 and their expansion after cryopreservation until passage 10 (scale bar: 200 ^m). (b) Full set of sequenced kAFO. (c) Whole mount immunofluorescent staining showing double positive GATA3 and LTL cells confirm a mixed tubular phenotype; in the lower panel a different phenotype of kAFO was observed with GATA3 positive and LTL negative cells (scale bar: 50 ^m). (d) Distal and collecting duct markers are absent or poorly present in kAFO in expansion medium; differentiation induces expression of CALB1, SLC12A1 and AQP2 (n=2 biological samples); nuclei counterstained with Hoechst (scale bars: 50 ^m). Figure 11 – lAFO: (a) Phase contrast images showing the thawing of lAFO at passage 3 and their expansion after cryopreservation until passage 6 (scale bar: 200 ^m). (b) Different morphological phenotype observed in lAFO lines (scale bars: 200 ^m). (c) Full set of sequenced lAFO. (d) Proximal (FOXJ1, ACATUB, MUC5AC, KRT5) markers are absent or poorly expressed at protein level in lAFO cultured in expansion medium; conversely, they are expressed in proximally differentiated lAFO (n=3 biological samples); (scale bars: 50 ^m). (e) TEM showing normal cilia rootlets with associated mitochondria (green arrow); longitudinal section of the ciliary axonemes confirms a normal central microtubule pair (yellow arrow) and radial spokes (red arrow) (scale bars: 100 nm). (f) Distalised lAFO produce SFTPB compared to control in expansion medium; nuclei counterstained with Hoechst (scale bars; 50 ^m). Figure 12 – CDH organoids: (a) Derivation of amniotic fluid and tracheal fluid organoids from different CDH patients before (left) and after (right) FETO fetal surgery (scale bar: 200 ^m). (b) Organoid generation efficiencies at various gestational age for CDH AF and TF. (c) PCA showing CDH organoids within alongside all the AFO clusters. (d) Heatmap and comparison of lung surfactant related genes between CDH organoids isolated before and after FETO. (e) Comparison between up and downregulated genes in CDH lAFO and lTFO before and after FETO. (f) DEG comparison of CDH lAFO and lTFO before and after FETO. Figure 13 – Further CDH organoid analysis. Volcano plots showing significant (p<0.05, |LFC| > 2) DEGs. (a) Comparison of organoids generated from AF or TF sampled after FETO with those before FETO. (b) Comparison of AFO and TFO from before and after FETO with age-matched controls. (c) Comparison of AFO with TFO, before and after FETO. Figure 14: Schematic diagram summarising the steps of the method of the invention. Figure 15: Isolation of mouse amniotic fluid-derived organoids (mAFO). a) Derivation of mouse amniotic fluid-derived primary organoids (mAFO). b) Different mAFO morphologies. c) Expansion of mAFO over weeks. d) Quantification of organoid formation efficiency (organoids per viable cells). d) Immunofluorescent staining showing epithelial identity of mAFO (scale bar 50um). Figure 16: Derivation and expansion of sheep amniotic fluid-derived primary organoids. (scale bar 200um). Figure 17: Generation of airway assembloids from fetal fluid-derived lung organoids. Phase-contrast images showing that differentiated lung AF organoids cultured in floating condition self-assemble into complex assembloids. Scale bar 200um. Figure 18: Characterization of fetal fluid-derived lung assembloids. Immunofluorescence images showing that assembloids maintain epithelia identity (ZO- 1, ECAD) and contain main airway epithelial cell types such as ciliated (Ac-tub) and basal cells (KRT5). Scale bar 50um. Figure 19: Phase-contrast images showing the derivation of neurosphere organoids from human fetal fluids. Example 1 - Single cell-guided prenatal derivation of primary epithelial organoids from the human amniotic and tracheal fluids. Summary Despite advances in prenatal diagnosis, it is still difficult to predict severity and outcomes of many congenital malformations. New patient-specific prenatal disease modelling may optimise personalised prediction. We and others have described the presence of mesenchymal stem cells in amniotic fluid (AFSC) that can generate induced pluripotent stem cells (iPSCs). The lengthy reprogramming processes, however, limits the ability to define individual phenotypes or plan prenatal treatment. Therefore, it would be advantageous if fetal stem cells could be obtained during pregnancy and expanded without reprogramming. Using single cell analysis, we characterised the cellular identities in amniotic fluid (AF) and identified viable epithelial stem/progenitor cells of fetal intestinal, renal and pulmonary origin. With relevance for prenatal disease modelling, these cells could be cultured to form clonal epithelial organoids manifesting small intestine, kidney and lung identity. To confirm this, we derived lung organoids from AF and tracheal fluid (TF) cells of Congenital Diaphragmatic Hernia (CDH) fetuses and found that they show differences to non-CDH controls and can recapitulate some pathological features of the disease. Amniotic Fluid Organoids (AFO) allow investigation of fetal epithelial tissues at clinically relevant developmental stages and may enable the development of therapeutic tools tailored to the fetus, as well as to predicting the effects of such therapies. RESULTS Single cell mapping of the human amniotic fluid reveals the presence of fetal intestinal, renal and pulmonary epithelial progenitors With the aim of mapping the cellular content of the human AF and investigating the presence of epithelial progenitors from multiple tissues, we collected AF from 11 pregnancies ranging between 15-34 weeks’ gestational age (GA) (Supplementary Table 1). As more than 97% of the cells shed in the AF are dead or undergoing apoptosis, we first isolated the viable cell fraction using fluorescence-activated cell sorting (FACS). To preserve the heterogeneity of the AF cellular content, we did not select the cells by Forward Scatter or Side Scatter. The incorporation of Hoechst was used to identify nucleated cells and exclude debris, urea crystals and other non-cellular particles. In conjunction, we used propidium iodide (PI) to exclude dead cells and cells with a damaged plasma membrane (Figure 1a; Figure 7a). After confirming cell viability through Live/Dead Acridine Orange/Propidium Iodide fluorescent Luna cell counter, we performed 3’ single cell RNA sequencing (scRNAseq) using the 10x Genomics Chromium platform. The data were processed, batch corrected and filtered to produce normalised count matrices, enabling the generation of an unsupervised uniform manifold approximation and projection (UMAP) using Seurat v4 (Figure 1a middle and Supplementary Figure 1b). We applied the SingleR package to perform automated labelling of any AFEC based on human primary cell atlas data42. We then probed our dataset for the expression of the epithelial genes EPCAM, CDH1, KRT8, KRT10, KRT17 and KRT19, which showed high levels of expression (Figure 1a right; Figure 7c). We validated that the cluster expressing these epithelial markers was the same as highlighted by SingleR (Figure 1b). A further validation of these gene expression data was performed by flow cytometry on viable, unselected AF cells, confirming the broad presence of EpCAM (EPCAM) and ECAD (CDH1) in the majority of live cells present in the AF (Figure 1c). We proceeded to interrogate our scRNAseq data for the expression of distinct intestinal, renal, and pulmonary epithelial markers within the AFEC cluster. The violin plots presented in Figure 1d highlight the presence of specific markers of the three tissues within the AF epithelial cluster. Finally, we assessed expression of tissue- specific stem / progenitor cell markers in the dataset. The UMAPs highlight the presence of cells co-expressing: SOX2 and ASCL2, MUC2, FABP1, LRIG1 (Intestine), PAX8 or LHX1 (kidney), and at least two of SOX2, SOX9, NKX2-1, FOX2A (Lung) (Figure 1e). This demonstrated the presence of intestinal, renal and pulmonary stem / progenitor cells in the AF. Based on this evidence, we went on to culture primary fetal organoids from the AF. Generation of primary fetal epithelial human amniotic fluid organoids (AFO) We isolated unselected viable AF cells with the method described above. The cells were seeded in Matrigel droplets and cultured in an ad hoc-defined epithelial medium, developed to allow the formation of organoids, without providing specialised tissue-specific signals. Using this method, we derived >200 organoid lines from 18 AF samples ranging from 16-34 weeks GA (Supplementary Table 1). Organoid formation was observed in 82% of the AF samples (18/22), with a median formation efficiency of 9 organoids formed/105 live cells plated (range 30.9). However, we found no meaningful association between the GA at AF sampling and the AFO formation efficiency (Figure 2b; figure 8a). Starting from day zero, some cells began proliferating and self-organising to form 3D structures, that became visible within the first week. During the first 2 weeks of culture, the formation of organoids became evident, and it was possible to pick single clonal organoids (Figure 2a-c). Once dissociated into single cells and replated, the organoids started reforming and expanding. AFO showed multiple morphological features, could be expanded at least until passage 5 and successfully cryopreserved, providing initial evidence of self-renewal and long-term culture (Figure 2b-d; Figure 8b-c). Two distinct organoid phenotypes were observed through X-ray phase contrast computed tomography (PC-CT) and were determined to be parenchymatous and cystic (Figure 2e). To further confirm the presence of proliferating cells within the AFO, as well as the lack of apoptotic cell death, we performed immunostaining for the proliferative marker Ki67 and the apoptotic marker caspase 3 (Figure 2f). To confirm the epithelial identity of the organoids, we performed immunofluorescent staining for standard pan-epithelial human cell markers (EpCAM, ECAD, and Pan-Cytokeratin). Concomitantly, we have confirmed the lack of mesenchymal features by demonstrating absence of PDGF receptor alpha expression. Interestingly, AFOs form a polarised epithelium as demonstrated by the presence of basolateral integrin b4 (ITG ^4), apical F-Actin and the tight junction marker ZO-1 on the luminal side of the organoids (Figure 2g). To gather initial evidence on the tissue identity of the organoids, we isolated RNA from clonally expanded organoid lines and performed bulk RNA sequencing (n=105 from 17 independent AF samples). The bulk RNA sequencing data was subjected to an unsupervised principal component analysis (PCA; Figure 2h; Figure 8d). Primary fetal organoids were derived from fresh fetal human intestine, stomach, lung and kidney, and included in the dataset as internal controls (n=14 with a minimum of 3 per tissue type). The PCA highlights the formation of three clusters of samples, colocalising with the intestinal, pulmonary and renal controls respectively (confirmed by Euclidean clustering, Figure 8e). Moreover, a Gene Ontology analysis performed on each cluster against the rest of the dataset (after removal of the control organoids) showed upregulation of pathways related to the three tissues in the respective cluster (Figure 2i). Overall, this provides evidence that AF cells can give rise to clonal small intestine, lung and kidney AFO. We then went on investigating the gene and protein expression of tissue specific markers, as well as the organoids’ maturation potential within the three AFO clusters. Generation and maturation of small intestinal Amniotic Fluid Organoids (siAFO) Small intestinal AFO (siAFO) expanded consistently for over 10 passages, showing the formation of crypt-like structures (Figure 9a). Moreover, an EdU incorporation assay indicated the presence of proliferating cells at the base of the intestinal organoids’ crypts (Figure 3a). RNA sequencing of 23 siAFO showed expression of typical intestinal stem/progenitor cell genes (LGR5, OLMF4, LRIG1, SMOC2) as well as Paneth (LYZ), goblet (MUC2, CLCA1) and endocrine (CHGA) cell markers. Enterocyte cell markers (ALPI, FABP1, VIL1, EZR, KRT20, ATP1A1) were also highly expressed by siAFO (Figure 3b). We validated the RNA sequencing data at a protein level via immunostaining for the crypt stem cell marker Olfactomedin 4 (OLFM4) and the intestinal epithelial cytokeratin 20 (KRT20), highlighting the presence of a crypt-villus axis. Interestingly, siAFO express markers of numerous intestinal cell types such as Paneth cells, as shown by immunostaining for Lysozyme (LYZ), and enterocytes stained for Fatty Acid Binding Protein 1 (FABP1) (Figure 3c and d). siAFO also lack lung and kidney specific markers NKX2-1 and PAX8, respectively (Figure 9b). To evaluate the functional capacity of these siAFOs, we assessed digestive enzyme activity. We demonstrated dipeptidyl peptidase IV activity (Figure 3e), a small intestinal brush border enzyme, indicating that siAFOs are capable of peptide hydrolysis. We then performed a maturation assay by placing the siAFO in an intestinal specific medium, in which, upon long-term culture (passage 9) and maturation, siAFO displayed more budding structures, acquiring the typical small intestinal crypt-like organisation. We observed several Chromogranin A-positive enteroendocrine cells, as well as the strong presence of the intestine-specific goblet cell marker Mucin 2 (MUC2) compared to siAFO in the expansion medium (Figure 3f; figure 9c). In addition, RT-qPCR analysis, following gamma-secretase inhibitor (DAPT) treatment, demonstrated downregulation of both notch target genes (HES1, OLFM4) and stem cell, Paneth cell and Wnt target genes (LGR5, LYZ, AXIN2). Enterocyte and goblet cell markers (FABP1, ALPI, MUC2) were upregulated. Surprisingly, with notch inhibition we did not see an increase in the expression of the master transcription regulator ATOH1 nor its downstream target DLL1. (figure 9d) Generation and differentiation of kidney tubule Amniotic Fluid Organoids (kAFO) Similar to what has been presented for siAFO, we expanded, characterised and differentiated kidney tubule AF-derived organoids (kAFO), identified by RNAseq within the kidney PCA cluster. Upon expansion, kAFO manifested a more compact morphology distinguishable from the one observed for the siAFO. kAFO could be cryopreserved and expanded long-term (up to passage 10) while maintaining proliferation ability, as highlighted by the diffuse expression of Ki67 (Figure 4a; Figure 10a). We then probed our RNAseq dataset for the presence of renal markers, which were present in a total of 44 organoids from 16 patients spanning 18-34 weeks GA. kAFO express canonical renal epithelial development/progenitor and nephron progenitor- specific genes (PAX2, PAX8, LHX1, JAG1). We also detected high expression levels of distal tubule genes (PCBD1, SLC41A3, POU3F3) as well as some proximal tubule markers (ABCC1, ABCC3, ABCC4, CUBN). Collecting duct marker GATA3 was also detected, while canonical Loop of Henle marker UMOD was not present. Podocyte markers (WT1, NPHS1, NPHS2), except for PODXL, were not expressed in kAFO. Interestingly, some kAFO lines expressed ureteric bud marker (RET), while early cap mesenchyme cell genes (SIX2, CITED1, GDNF) were not observed (Figure 4b; Figure 10b). Based on this profile, we concluded that kAFO manifest a tubuloid-like phenotype and are rich in markers belonging to multiple segments of the renal tubules. As additional validation, we performed immunofluorescent staining to confirm protein expression for the renal epithelium progenitor markers PAX8 and LHX1. Moreover, kAFO displayed kidney segment-specific protein markers such as GATA3 and ECAD (distal tubule/collecting duct), and Lotus tetragonolobus lectin (LTL, proximal tubule). The presence of polarised tubular microvilli was also confirmed by immunofluorescence of acetylated tubulin (Ac-αTUB; Figure 4c-d). Interestingly, kAFO exhibited a mixed tubular phenotype with some organoids co-expressing GATA3 and LTL or presenting only GATA3 (figure 10c). Functional assessment of kAFO was performed evaluating thallium intake. Following addition of a voltage-gated potassium (K+) ion channel stimulator, kAFO showed increased intracellular thallium fluorescence compared to positive control fetal kidney organoids (FKO) and negative control fetal lung organoids (FLO), indicating presence of functional potassium channels (Figure 4e). To further confirm the renal phenotype of the organoids, we have adapted a previously reported differentiation assay to promote maturation of the distal/collecting duct lineage44. After 14 days of stimulation with vasopressin and arginine aldosterone, kAFO manifested a slight morphological change and expressed markers of the principal cells of the collecting duct (AQP2) and of the distal tubules (SLC12A1 and CALB1) compared to kAFO in expansion medium (Figure 4f; figure 10d). Moreover, differentiated kAFO displayed a higher percentage of CALB1 positive cells (19.3 ± 7.6) which was reflected in an increased CALB1 expression (Figure 4g-h). Generation and differentiation of lung Amniotic Fluid Organoids (lAFO) The lungs are one of the major cellular contributors to the AF due to the continual release of TF, rich in pulmonary cells into the amniotic cavity. Consequently, we hypothesised that lung AF-derived organoids (lAFO) would be readily formed. lAFOs can have a cystic appearance with a thin epithelial cell layer and a lumen or can present with a more parenchymatous morphology. We isolated and clonally expanded 69 lAFO lines from 18 AF samples spanning 16-34 weeks GA, that propagated for over 6 passages maintaining high proliferation ability, as confirmed by the strong expression of Ki67 (Figure 5a; Figure 11a-b). RNAseq-based marker analysis indicated the presence of multiple cellular identities within the 38 sequenced lung organoids, with consistent expression of stem/progenitor cell markers (NKX2-1, FOXA2, SOX2, SOX9, TP63, GATA6), as well as of both Alveolar Type 1 (HOPX, PDPN, AGER, AQP5) and Alveolar Type 2 cells-related genes (SFTPA1, SFTPA2, SFTPB, SFTPC, SFTPD, ABCA3, LAMP3). Mature basal cell markers TROP2 and NGFR were not detected when lAFO were cultured in expansion medium. The distal marker KRT5 was occasionally detected in some lines, as well as the specific ciliated cell transcription factor FOXJ1, which showed sporadic low expression. In addition, early secretory cell marker SCGB3A2 was highly expressed in comparison to the mature marker SCGB1A1, which was not expressed. MUC5AC secretory cells were present in the control fetal lung organoids, while lAFO did not express the marker. Lastly, neuroendocrine cell marker ASCL1 was present but not highly expressed by lAFO while few lines showed low expression of CHGA (Figure 5b; Figure 11c). We then performed protein expression validation via immunostaining, which showed homogenous and strong presence of the stem cell markers NKX2-1, SOX2, and the basal cell marker P63. The prosurfactant protein C (SFTPC) was instead absent (Figure 5c, quantification in 5d). The potential of the lAFO to undergo terminal proximal and distal lung differentiation was also assessed. Remarkably, when pushed towards proximal differentiation, we observed the appearance of a polarised epithelium with motile cilia on the luminal surface of the organoids. Immunofluorescent staining on the proximally differentiated lAFO showed the presence of Ac- ^TUB positive cilia on the luminal side of the organoids, confirming both differentiation and polarisation of the epithelia. The nuclear expression of the ciliated epithelia marker FOXJ1 further corroborated our observation (Figure 5e). In addition, the appearance of the mature basal cell marker keratin 5 (KRT5), and secretory marker mucin (MUC5AC), concomitantly with the maintenance of SOX2, indicated the active proximal lung differentiation process (Figure 5e; Figure 11d). Indeed, the percentage of FOXJ1 positive cells (60.7% ± 13.7) was higher in the differentiated lAFO compared to controls in expansion, while the number of KRT5-positive cells (3.1% ± 2.6) did not significantly increase in number (Figure 5f). Moreover, proximal lAFO led to an increased expression of airway markers such as FOXJ1, TUBA1A and KRT5 (Figure 5g). Respiratory motile cilia were analysed in detail by transmission electron microscopy (TEM; Figure 5h). Cilia displayed normal rootlets and associated mitochondria. The ciliary axonemes also displayed normal internal structures such as radial spokes and a normal central microtubule pair. The axonemes were normal in structure, showing outer and inner dynein arms (Figure 5i; Figure 11e). On the other hand, when pushed towards a distal phenotype, lAFO showed increased protein expression of the AT2 cell marker SFTPB, presenting with different cellular localisation in independent lAFO lines. We observed different distribution of SFTPB in different lAFO lines. Some presenting granules, while others had accumulation of surfactant droplets in the lumen, possibly indicating a higher state of maturation (Figure 5j; Figure 11f). This observation was however not reflected in an associated increase in SFTPB gene expression (Figure 5k). However, ultrastructural analysis revealed that distalised lAFO contain lamellar bodies with a normal structure and a core composed of multi- lamellar membranes, typical features of distal lung cells (Figure 5l). Lung organoids derived from amniotic and tracheal fluid (AF / TF) of fetuses with CDH manifest a substantially different phenotype compared to gestational age-matched controls CDH is a congenital malformation where the diaphragmatic muscle fails to close (OMIM: 142340, 222400, 306950), with a consequent herniation of the fetal abdominal organs into the chest. Consequently, the fetal lungs are subjected to a mechanical compression, limiting their physiological growth and leading to developmental impairments of the respiratory and vascular compartments45,46. To investigate the suitability of our platform for fetal disease modelling, we derived lung organoids from both AF and TF of fetuses diagnosed with severe or moderate CDH undergoing fetal intervention at our institutions (Figure 6a; Figure 12a; Supplementary Table 1). AF and TF were obtained at the time of Fetoscopic Endoluminal Tracheal Occlusion (FETO)3,4. The fluid collected was subjected to the organoid derivation method described above. As the volume of TF samples collected was low (1-3mL) but with high levels of cell viability (60-75%) we opted to omit the sorting/purification part of the protocol to preserve cell numbers. Similar to the non-CDH lAFO presented in Figure 5, we successfully generated CDH lAFO from 9 CDH fetuses (9/14; 64.3%) and lTFO from 5 (5/9; 55.6%) CDH fetuses (Figure 8a). Both CDH lAFO and lTFO expanded for multiple passages (up to P5; figure 6b-d) and expressed NKX2-1, SOX2 and P63 along with Ki67 (Figure 6e). Interestingly, only lTFO expressed SOX9, suggesting a more prominent stem/progenitor identity25 (Figure 6f). Moreover, PCA analysis confirmed that all TF-derived organoids had lung identity and demonstrated clustering of CDH lAFO and lTFO with an associated shift of the latter from non-CDH lAFO (Figure 12c). We then investigated the gene expression profile of CDH organoids (lAFO and lTFO) using RNAseq and found evidence of expression of most lung epithelial stem/progenitor markers at levels similar to control lAFO (GA-matched; Figure 6g). DEG analysis was performed to explore differences between CDH and control GA- matched lAFO. There was a clear reduction in number of DEGs (p-value < 0.01, |LFC| >2) identified between organoids generated from samples before and after the FETO procedure (181 vs.47 total DEGs, Figure 6h). Gene ontology analysis identified a clear upregulation in pathways related to surfactant production and metabolism in CDH organoids before FETO treatment vs. age-matched controls, including upregulation of genes related to phosphatidylcholine metabolism (Figure 6i), as well as a downregulation of pathways related to wound healing, growth and differentiation (p-value < 0.05, |LFC| >1). Interestingly, DEG analysis of post-FETO CDH organoids compared to GA-matched controls revealed fewer differences beside surfactant metabolism (Figure 6i; Figure 12d). We then subjected CDH lAFO to the same proximal and distal differentiation protocols previously mentioned. Similar to controls, proximal CDH lAFO showed the formation of motile cilia and expressed of acetylated ^-tubulin (Ac- ^TUB), FOXJ1 and SOX2 proteins (Figure 6j). By recording the ciliary movement using a high-speed camera, we analysed the ciliary beating frequency (CBF) of lAFO from two non-CDH patients and one with CDH. Although remaining within in a physiological range, the CDH lAFO showed a lower CBF of 10 ^ 0.17 Hz vs 13 ^ 0.16 Hz (Figure 6k). Finally, when subjected to distal differentiation, CDH lAFO expressed surfactant protein B (Figure 6l), similar to that observed in non-CDH (Figure 5j). We generated lAFO and lTFO from CDH pregnancies from samples taken at the time of the two interventions related to FETO: I) insertion of the occlusive endotracheal balloon (n=9, 26-31 weeks’ GA, n=41, 6 patients); II) removal of the balloon (n=5, 32-34 weeks’ GA, n=13, 3 patients). Paired analysis was not possible owing to sample availability and therefore pooled analysis was performed herein. We noted that there were fewer DEGs in the post-FETO CDH organoids compared to control lAFO of the same gestational age (Figure 6 h, i). Immunofluorescent staining for proSFTPC showed no expression in the CDH lTFO obtained after FETO but was positive in all other CDH samples (figure 6f), highlighting some differences between these samples. Individual comparative DEG analyses between controls, lAFO and lTFO before and after FETO are provided in Figure 13. DISCUSSION In this article, we describe a reliable method for derivation of autologous primary fetal organoids from multiple epithelial tissues using AF and TF sampled for clinical diagnostic and therapeutic purposes, while allowing the continuation of pregnancy. To guide our experiments, we generated an scRNAseq atlas of the human AF from 11 patients (15 to 34 GA), a resource not currently available in the literature. While it is known that most cells present in the AF are epithelial in nature, their cellular origin has been long debated, and often ascribed to the skin, kidneys and fetal membranes47. Our AF cell map confirms that human AFECs are heterogeneous in nature, originating from multiple tissues and consequently distinct from the previously reported placental-derived hAECs48,49. In detail, we provide evidence of the presence, within the AF epithelial cluster, of amniotic fluid-derived stem / progenitor cells of intestinal, renal and pulmonary origin. Once cultured in permissive conditions, AFEC demonstrated their capacity to generate primary clonal epithelial organoids. AFO are amenable to long term expansion, through methodologies similar to that described for fetal organoids produced with destructive approaches. When placed in conditions favourable to promote maturation, the renal, intestinal and pulmonary AFO successfully acquired differentiation hallmarks typical of their tissue of origin. The generation of AFO uses widely available samples, requires minimal manipulation, and applies only routine organoid culture techniques. The timeline from fluid sampling to full characterisation and expansion of the organoids is currently below 4 weeks, providing a tool that can be applied in a timeline relevant to prenatal counselling and potential therapy, compared with iPSC-dependent methods16. The advancement of non-invasive prenatal diagnostic techniques, such as the detection of cell free DNA of fetal origin in the maternal blood stream, has reduced the need to perform amniocenteses for primary diagnostic purposes. However, AF samples remain accessible, with approximately 30,000 amniocenteses performed each year in the UK50. This remains necessary in order to perform advanced diagnostics such as array analysis or whole exome sequencing and as confirmatory diagnosis 2. Moreover, invasive procedures such amniodrainage, are routinely used as treatment for polyhydramnios51,52, and laser treatment for Twin-to-Twin Transfusion Syndrome (TTTS)53. Finally, spina bifida repair, and FETO for CDH, provide further access to the fluid during pregnancy3–5. Intestinal organoids have been isolated from both fetal and adult tissues and used for modelling intestinal development, regeneration, and repair54–56. Interestingly, once pushed towards maturation, siAFO acquired features observed in small intestinal organoids derived from primary tissues56. Derivation of small intestinal organoids was rare, with success in only two of our samples (16 and 17 GA), one of which was obtained from a termination of pregnancy. The occurrence of AFEC with intestinal stem cell features is an interesting finding given the widely held presumption that after the breakdown of the anal membrane, the anal sphincters retain fetal intestinal content within the gastrointestinal tract during normal development from 12 weeks GA57. However, colonic mucosal cells are present in second trimester AF58 indicating that cells originating from the gastrointestinal tract are present in AF after the time at which the anal sphincters are thought to retain fetal intestinal content. Thus, there may be the possibility of generating siAFO for therapeutic use in instances where prenatal diagnosis of a congenital disorder potentially resulting in short-bowel syndrome has been made. In contrast, kAFO and lAFO were easily derived in large numbers across all gestational ages studied. This is presumably related to the regular circulation of fluid through the fetal lungs and renal tract. AF volume increases until the 34th week of pregnancy and then remains relatively constant until term. The production rate of the fetal urine in the human fetus at term (800–1200ml/day), is sufficient to completely replenish the entire amniotic volume every 12– 24 hours 59,60. In keeping with the latter, mesenchymal cells of renal origin have been found in third-trimester AF but are generally difficult to isolate61. So far, there has been only one report of isolation of potential podocytes62. When cells isolated from kidney tissue and urine are maintained in culture, they are able to generate tubuloids63. Tubuloid AFO recapitulate some of the characteristics observed previously, such as the expression of both developmental kidney nephron progenitor markers and, upon differentiation, specific functional proteins. Similarly, while fetal breathing movements are associated with inward and outward movements of fluid through the trachea, there is overall a net efflux of lung fluid which it is likely to continuously provide lung progenitors to AF, while limiting the influx of cells from other origins into the fetal64. This flow is interrupted during FETO treatment and could explain some of the differences between lAFO and lTFO59. CDH represents a condition that is clinically relevant for study on account of reliable prenatal diagnosis, along with current clinical practice beginning to offer prenatal therapy for high-risk infant3,4. Current stratification of patients relies on prenatal imaging parameters which are simple to acquire and reliable; although development of sophisticated MRI techniques such as 3-dimensional Lung Volume hold some promise to improve prognostic65. The FETO procedure is proven to be of considerable survival benefit to the most severely affected patients; however, there is the need to identify predictive functional biomarkers, as survival remains poor. Notably, around 60% of severe patients do not survive to discharge from hospital66. Previous work on tracheal fluid samples taken before and after FETO showed a different miRNA signature in patients who responded to the intervention, with higher expression of miR- 20067. This specific pathway could not be robustly analysed in our dataset, since our organoid media contained both A83-01 and Noggin, respectively inhibiting TGFβ and BMP signalling, previously suggested to be an underlying basis for the success of FETO. Importantly, AFO/TFO-derived from fetuses affected by CDH manifest some features of this condition such as alterations in the expression of the surfactant protein genes (SFTPA1, SFTPA2, SFTPB, SFTPC, SFTPD) as presented in Figure 6i. Moreover, increased AT2/AT1 has been suggested as a signature for hypoplastic and CDH lungs in previously published preclinical work68,69. Indeed, our RNAseq analysis demonstrated higher expression of AT2 genes in a greater proportion of cells in the CDH organoids compared to their age-matched controls, with marked upregulation of gene pathways of surfactant metabolism evident on pre-FETO lAFO vs. controls. This suggests that lAFO may be used for disease modelling, drug testing and potentially therapeutically. If successfully implemented, these complex disease models would be amenable for severity prediction, enabling informed clinical decisions to be made before the baby is born. Ultimately, when derived from pregnancies affected by a congenital condition (CDH), AFO and TFO manifested an overall altered phenotype, that was, in part, brought closer to the control transcriptome, in samples taken after treatment with FETO, the current gold standard prenatal approach for this condition (Figure 6h; Figure 13). It must be noted that, AF sampling post-FETO is conducted before the removal of the balloon, so AFO may not reflect the changes undergone by the fetal lung in response to the FETO procedure, while TFO being sampled from the occluded trachea will do so70. While sampling of TF at the time of FETO is a less widely applicable means of accessing fetal cells the CDH TF samples giving rise to organoids (5 out of 9), only originated lung organoids (5 out of 5) confirming the single tissue origin of the cells contained in the TF. With rapid advances in organoid biology, particularly with regards of organoids derived from fetal tissue, we expect that in the near future, it will be possible to generate organoids from other tissues exposed to AF. Moreover, we expect that further work will allow different types of organoids to be obtained for each organ. As an example, it is possible to observe the broad distribution of the lung organoids in the PCA presented in Figure 2, as well as the formation of two renal clusters. The core advantage of the technology presented in this article, is to offer the ability of deriving fetal organoids prenatally, without the need for accessing the fetal tissue. Currently, the only viable alternative is derivation of organoids from iPSC. Although the iPSC route offers flexibility, and virtually unlimited expansion potential, it comes with a series of technological trade-offs prevent widespread clinical use. Some of these issues are intrinsic to reprogrammed cells, such as insertional mutagenesis of the reprogramming transgenes, acquisition of genomic aberrations during expansion, aneuploidy, sub-chromosomal copy number variants point mutations and alteration of epigenetic marks71. Moreover, the use of mutated iPSCs, could influence their successful recapitulation of disease phenotypes72. From our perspective one of the major technological burdens of using reprogramming to produce fetal organoids is the time required. If the aim is to model or treat a condition before birth in order to provide personalised prognostic information or autologous organoid-based therapy, it must be possible to implement this strategy within the 40 weeks of gestation. This timeline is further shortened by the fact that AF / chorionic villi sampling is not normally performed until the end of the first trimester; and furthermore, the majority of prenatal therapies are currently ideally delivered prior to 30 weeks. Several examples of iPSCs-organoid derivation highlight the need of at least 21 weeks to produce the organoids73. Our approach benefits from the already committed progenitors present in the fetal fluids, which require minimal manipulation to lead to the production of a large amount of primary autologous fetal organoids in approximately 4 to 6 weeks. Conclusions In conclusion, we report derivation of epithelial organoids of different tissue identity through a minimally invasive approach, from continuing pregnancies and within a broad GA window. AFO of intestine, kidney and lung origin are expandable and can be functionally differentiated with great potential for functional diagnosis, regenerative medicine, and disease modelling. An example of the latter could be observed in CDH, where AFO can recapitulate some of the original features observed in the affected fetuses, suggesting their use for personalised medicine. REFERENCES 1. Davidson, J. R. et al. Fetal body MRI and its application to fetal and neonatal treatment: an illustrative review. Lancet Child Adolesc Health 5, 447–458 (2021). 2. Lord, J. et al. Prenatal exome sequencing analysis in fetal structural anomalies detected by ultrasonography (PAGE): a cohort study. Lancet 393, 747–757 (2019). 3. Deprest, J. A. et al. Randomized Trial of Fetal Surgery for Moderate Left Diaphragmatic Hernia. N Engl J Med 385, 119–129 (2021). 4. Deprest, J. A. et al. Randomized Trial of Fetal Surgery for Severe Left Diaphragmatic Hernia. N Engl J Med 385, 107–118 (2021). 5. Adzick, N. S. et al. A randomized trial of prenatal versus postnatal repair of myelomeningocele. N Engl J Med 364, 993–1004 (2011). 6. Morris, R. K. et al. Percutaneous vesicoamniotic shunting versus conservative management for fetal lower urinary tract obstruction (PLUTO): a randomised trial. Lancet 382, 1496–1506 (2013). 7. Takasato, M. et al. Directing human embryonic stem cell differentiation towards a renal lineage generates a self-organizing kidney. Nature Cell Biology 201316:116, 118–126 (2013). 8. Takahashi, K. et al. Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell 131, 861–872 (2007). 9. Hofer, M. & Lutolf, M. P. Engineering organoids. Nat Rev Mater 6, 402–420 (2021). 10. Spence, J. R. et al. Directed differentiation of human pluripotent stem cells into intestinal tissue in vitro. Nature 470, 105–110 (2011). 11. Kim, J., Koo, B. K. & Knoblich, J. A. Human organoids: model systems for human biology and medicine. Nat Rev Mol Cell Biol 21, 571–584 (2020). 12. McCracken, K. W. et al. Modelling human development and disease in pluripotent stem-cell-derived gastric organoids. Nature 516, 400–404 (2014). 13. Dye, B. R. et al. In vitro generation of human pluripotent stem cell derived lung organoids. Elife 4, 1–25 (2015). 14. Takasato, M. et al. Kidney organoids from human iPS cells contain multiple lineages and model human nephrogenesis. Nature 526, 564–568 (2015). 15. Michielin, F. et al. The Microfluidic Environment Reveals a Hidden Role of Self- Organizing Extracellular Matrix in Hepatic Commitment and Organoid Formation of hiPSCs. Cell Rep 33, (2020). 16. Kunisaki, S. M. et al. Human induced pluripotent stem cell-derived lung organoids in an ex vivo model of the congenital diaphragmatic hernia fetal lung. Stem Cells Transl Med 10, 98–114 (2021). 17. Sato, T. et al. Long-term expansion of epithelial organoids from human colon, adenoma, adenocarcinoma, and Barrett’s epithelium. Gastroenterology 141, 1762– 1772 (2011). 18. Huch, M. et al. Long-term culture of genome-stable bipotent stem cells from adult human liver. Cell 160, 299–312 (2015). 19. Bartfeld, S. et al. In vitro expansion of human gastric epithelial stem cells and their responses to bacterial infection. Gastroenterology 148, 126-136.e6 (2015). 20. Sachs, N. et al. Long-term expanding human airway organoids for disease modeling. EMBO J 38, (2019). 21. Nunez-Nescolarde, A. B., Nikolic-Paterson, D. J. & Combes, A. N. Human Kidney Organoids and Tubuloids as Models of Complex Kidney Disease. Am J Pathol 192, 738–749 (2022). 22. Cindrova-Davies, T. et al. Menstrual flow as a non-invasive source of endometrial organoids. Commun Biol 4, (2021). 23. Lõhmussaar, K. et al. Patient-derived organoids model cervical tissue dynamics and viral oncogenesis in cervical cancer. Cell Stem Cell 28, 1380-1396.e6 (2021). 24. Giobbe, G. G. et al. SARS-CoV-2 infection and replication in human gastric organoids. Nat Commun 12, (2021). 25. Nikolić, M. Z. et al. Human embryonic lung epithelial tips are multipotent progenitors that can be expanded in vitro as long-term self-renewing organoids. Elife 6, (2017). 26. Fordham, R. P. et al. Transplantation of expanded fetal intestinal progenitors contributes to colon regeneration after injury. Cell Stem Cell 13, 734–744 (2013). 27. Cao, K. X., Booth, A., Ourselin, S., David, A. L. & Ashcroft, R. The legal frameworks that govern fetal surgery in the United Kingdom, European Union, and the United States. Prenat Diagn 38, 475–481 (2018). 28. de Coppi, P. et al. Regenerative medicine: prenatal approaches. Lancet Child Adolesc Health 6, 643–653 (2022). 29. Underwood, M. A., Gilbert, W. M. & Sherman, M. P. Amniotic fluid: Not just fetal urine anymore. Journal of Perinatology 25, 341–348 (2005). 30. de Coppi, P. et al. Isolation of amniotic stem cell lines with potential for therapy. Nat Biotechnol 25, 100–106 (2007). 31. Steigman, S. A. & Fauza, D. O. Isolation of mesenchymal stem cells from amniotic fluid and placenta. Curr Protoc Stem Cell Biol Chapter 1, (2007). 32. Ditadi, A. et al. Human and murine amniotic fluid c-Kit+Lin- cells display hematopoietic activity. Blood 113, 3953–3960 (2009). 33. Loukogeorgakis, S. P. & de Coppi, P. Concise Review: Amniotic Fluid Stem Cells: The Known, the Unknown, and Potential Regenerative Medicine Applications. Stem Cells 35, 1663–1673 (2017). 34. Mellows, B. et al. Protein and Molecular Characterization of a Clinically Compliant Amniotic Fluid Stem Cell-Derived Extracellular Vesicle Fraction Capable of Accelerating Muscle Regeneration Through Enhancement of Angiogenesis. Stem Cells Dev 26, 1316–1333 (2017). 35. Li, B. et al. Activation of Wnt signaling by amniotic fluid stem cell-derived extracellular vesicles attenuates intestinal injury in experimental necrotizing enterocolitis. Cell Death Dis 11, (2020). 36. Antounians, L. et al. Fetal lung underdevelopment is rescued by administration of amniotic fluid stem cell extracellular vesicles in rodents. Sci Transl Med 13, (2021). 37. Beall, M. H., van den Wijngaard, J. P. H. M., van Gemert, M. J. C. & Ross, M. G. Amniotic fluid water dynamics. Placenta 28, 816–823 (2007). 38. Haniffa, M. et al. A roadmap for the Human Developmental Cell Atlas. Nature 597, 196–205 (2021). 39. Amonkar, G. M. et al. Primary culture of tracheal aspirate-derived human airway basal stem cells. STAR Protoc 3, (2022). 40. Lu, J. et al. Rho/SMAD/mTOR triple inhibition enables long-term expansion of human neonatal tracheal aspirate-derived airway basal cell-like cells. Pediatric Research 2020 89:389, 502–509 (2020). 41. Wagner, R. et al. A tracheal aspirate-derived airway basal cell model reveals a proinflammatory epithelial defect in congenital diaphragmatic hernia. doi:10.1101/2022.11.10.515365. 42. Aran, D. et al. Reference-based analysis of lung single-cell sequencing reveals a transitional profibrotic macrophage. Nature Immunology 201920:220, 163–172 (2019). 43. Sachs, N., Tsukamoto, Y., Kujala, P., Peters, P. J. & Clevers, H. Intestinal epithelial organoids fuse to form self-organizing tubes in floating collagen gels. Development 144, 1107–1112 (2017). 44. Uchimura, K., Wu, H., Yoshimura, Y. & Humphreys, B. D. Human Pluripotent Stem Cell-Derived Kidney Organoids with Improved Collecting Duct Maturation and Injury Modeling. Cell Rep 33, (2020). 45. Congenital diaphragmatic hernia. Nature Reviews Disease Primers 20228:18, 1–1 (2022). 46. Zani, A. et al. Congenital diaphragmatic hernia. Nat Rev Dis Primers 8, (2022). 47. Hoehn, H. & Salk, D. Morphological and biochemical heterogeneity of amniotic fluid cells in culture. Methods Cell Biol 26, 11–34 (1982). 48. Miki, T., Lehmann, T., Cai, H., Stolz, D. B. & Strom, S. C. Stem Cell Characteristics of Amniotic Epithelial Cells. Stem Cells 23, 1549–1559 (2005). 49. Parolini, O. et al. Concise Review: Isolation and Characterization of Cells from Human Term Placenta: Outcome of the First International Workshop on Placenta Derived Stem Cells. Stem Cells 26, 300–311 (2008). 50. Amniocentesis and chorionic villus sampling. https://obgyn.onlinelibrary.wiley.com/doi/epdf/10.1111/1471-0528.16821. 51. Dickinson, J. E. et al. Amnioreduction in the management of polyhydramnios complicating singleton pregnancies. Am J Obstet Gynecol 211, 434.e1-434.e7 (2014). 52. Erfani, H. et al. Amnioreduction in cases of polyhydramnios: Indications and outcomes in singleton pregnancies without fetal interventions. European Journal of Obstetrics & Gynecology and Reproductive Biology 241, 126–128 (2019). 53. Senat, M.-V. et al. Endoscopic laser surgery versus serial amnioreduction for severe twin-to-twin transfusion syndrome. N Engl J Med 351, 136–144 (2004). 54. Sato, T. et al. Single Lgr5 stem cells build crypt-villus structures in vitro without a mesenchymal niche. Nature 2009459:7244459, 262–265 (2009). 55. Sprangers, J., Zaalberg, I. C. & Maurice, M. M. Organoid-based modeling of intestinal development, regeneration, and repair. Cell Death & Differentiation 202028:128, 95– 107 (2020). 56. Meran, L. et al. Engineering transplantable jejunal mucosal grafts using patient- derived organoids from children with intestinal failure. Nat Med 26, 1593–1601 (2020). 57. Nievelstein, R. A. J., van der Werff, J. F. A., Verbeek, F. J., Valk, J. & Vermeij-Keers, C. Normal and Abnormal Embryonic Development of the Anorectum in Human Embryos. Teratology 57, 70–78 (1998). 58. Chitayat, D., Marion, R. W., Squillante, L., Kalousek, D. K. & Das, K. M. Detection and enumeration of colonic mucosal cells in amniotic fluid using a colon epithelial- specific monoclonal antibody. Prenat Diagn 10, 725–732 (1990). 59. Glibert, W. M., Moore, T. R. & Brace, R. A. Amniotic fluid volume dynamics. Fetal Matern Med Rev 3, 89–104 (1991). 60. Rabinowitz, R., Peters, M. T., Vyas, S., Campbell, S. & Nicolaides, K. H. Measurement of fetal urine production in normal pregnancy by real-time ultrasonography. Am J Obstet Gynecol 161, 1264–1266 (1989). 61. Rahman, M. S. et al. The presence of human mesenchymal stem cells of renal origin in amniotic fluid increases with gestational time. Stem Cell Res Ther 9, 113 (2018). 62. da Sacco, S. et al. A novel source of cultured podocytes. PLoS One 8, (2013). 63. Gijzen, L. et al. Culture and analysis of kidney tubuloids and perfused tubuloid cells- on-a-chip. Nature Protocols 202116:416, 2023–2050 (2021). 64. Muniraman, H. Fetal lung fluid: Not the same as amniotic fluid. Pediatr Pulmonol 56, 1270 (2021). Davidson, J. et al. Motion corrected fetal body magnetic resonance imaging provides reliable 3D lung volumes in normal and abnormal fetuses. Prenat Diagn 42, 628–635 (2022). Wagner, R. et al. Proteomic Profiling of Hypoplastic Lungs Suggests an Underlying Inflammatory Response in the Pathogenesis of Abnormal Lung Development in Congenital Diaphragmatic Hernia. bioRxiv 2022.03.07.483298 (2022) doi:10.1101/2022.03.07.483298. Pereira-Terra, P. et al. Unique Tracheal Fluid MicroRNA Signature Predicts Response to FETO in Patients With Congenital Diaphragmatic Hernia. Ann Surg 262, 1130– 1140 (2015). Davey, M. Surfactant Levels in Congenital Diaphragmatic Hernia. PLoS Med 4, e243 (2007). Donahoe, P. K., Longoni, M. & High, F. A. Polygenic Causes of Congenital Diaphragmatic Hernia Produce Common Lung Pathologies. Am J Pathol 186, 2532– 2543 (2016). Evrard, V. A. et al. Intrauterine tracheal obstruction, a new treatment for congenital diaphragmatic hernia, decreases amniotic fluid sodium and chloride concentrations in the fetal lamb. Ann Surg 226, 753–758 (1997). Andrews, P. W. et al. Assessing the Safety of Human Pluripotent Stem Cells and Their Derivatives for Clinical Applications. Stem Cell Reports 9, 1–4 (2017). Poetsch, M. S., Strano, A. & Guan, K. Human Induced Pluripotent Stem Cells: From Cell Origin, Genomic Stability, and Epigenetic Memory to Translational Medicine. Stem Cells 40, 546–555 (2022). Khalaj, K. et al. Treatment with Amniotic Fluid Stem Cell Extracellular Vesicles Promotes Fetal Lung Branching and Cell Differentiation at Canalicular and Saccular Stages in Experimental Pulmonary Hypoplasia Secondary to Congenital Diaphragmatic Hernia. Stem Cells Transl Med 11, 1089–1102 (2022). Ginsberg, M., Schachterle, W., Shido, K. & Rafii, S. Direct conversion of human amniotic cells into endothelial cells without transitioning through a pluripotent state. Nature Protocols 201510:1210, 1975–1985 (2015). Ginsberg, M. et al. Efficient direct reprogramming of mature amniotic cells into endothelial cells by ETS factors and TGFβ suppression. Cell 151, 559–575 (2012). 76. Giobbe, G. G. et al. Extracellular matrix hydrogel derived from decellularized tissues enables endodermal organoid culture. Nature Communications 201910:110, 1–14 (2019). 77. Schutgens, F. et al. Tubuloids derived from human adult kidney and urine for personalized disease modeling. Nature Biotechnology 201937:337, 303–313 (2019). 78. Jacob, A. et al. Differentiation of Human Pluripotent Stem Cells into Functional Lung Alveolar Epithelial Cells. Cell Stem Cell 21, 472-488.e10 (2017). 79. Jafree, D. J., Long, D. A., Scambler, P. J. & Moulding, D. Tissue Clearing and Deep Imaging of the Kidney Using Confocal and Two-Photon Microscopy. Methods Mol Biol 2067, 103–126 (2020). 80. Paganin, D., Mayo, S. C., Gureyev, T. E., Miller, P. R. & Wilkins, S. W. Simultaneous phase and amplitude extraction from a single defocused image of a homogeneous object. J Microsc 206, 33–40 (2002). 81. Atwood, R. C., Bodey, A. J., Price, S. W. T., Basham, M. & Drakopoulos, M. A high- throughput system for high-quality tomographic reconstruction of large datasets at Diamond Light Source. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 373, (2015). 82. Limaye, A. Drishti: a volume exploration and presentation tool. https://doi.org/10.1117/12.9356408506, 191–199 (2012). 83. Savvidis, S. et al. Monitoring tissue engineered constructs and protocols with laboratory-based x-ray phase contrast tomography. Acta Biomater 141, 290–299 (2022). 84. Aran, D. et al. Reference-based analysis of lung single-cell sequencing reveals a transitional profibrotic macrophage. Nature Immunology 201920:220, 163–172 (2019). All references are herein incorporated by reference. MATERIALS AND METHODS Amniotic fluid collection and isolation of the viable cell fraction AF samples (amniocenteses and amniodrainages) were collected from the University College London Hospital (UCLH) Fetal Maternal Unit (FMU) and UZ Leuven as part of standard patient’s clinical care (REC 14/LO/0863 IRAS 133888). Written informed consent was obtained prior to the procedure. After collection, fluids were stored at 4°C until processing. AF samples were passed through a 70 μm and 40 μm cell strainer and transferred in 50ml tubes before being centrifuged at 300 g for 10 min at 4°C. Supernatant was discarded, pellet resuspended in 5-10 mL of FACS blocking buffer containing 1% FBS and 0.5 mM EDTA in PBS and transferred to FACS tubes. Cells were incubated with 5 μg/mL Hoechst (Sigma- Aldrich, 33342) for 40 min at 37°C and then counterstained with 2 μg/mL propidium iodide (PI) (Sigma-Aldrich, P4170) for 5 min at RT. Viable cells were sorted using a FACSAria III (BD), unselected for side and forward scatter, but gated for Hoechst+ and PI-. Derivation and culture of human amniotic fluid organoids (AFO) Viable amniotic fluid cells were resuspended in 30 μL of Matrigel (Corning) and plated 6x105 live cells/droplet onto a pre-warmed 24 well plate. Cells were cultured in an ad hoc defined generic medium (Supplementary Table 2) supplemented with Rho-kinase inhibitor (ROCKi; Tocris) and Primocin. To establish clonal organoid lines, single organoids formed were manually picked at passage 0 under the microscope to be clonally expanded. Each organoid was transferred in a 0.5 mL tube pre-coated with 1% BSA (Sigma-Aldrich). Organoids were resuspended in TrypLE (Thermo) and incubated for 5 min at 37°C. After digestion, organoids were disaggregated by pipetting and additional 400 μL of ice-cold Advanced DMEM/F-12 supplemented with Glutamax, P/S and Hepes (ADMEM+++) were added. Organoids were precipitated with a minicentrifuge for 2 min and a second washing passage was repeated. After centrifugation the pellet was resuspended in 20 μL Matrigel (Corning) and plated in a 48 well plate. The plate was incubated for 20 min at 37°C and generic medium was added with ROCKi (Tocris) and Primocin for the first 3 days. Medium was replaced every 3-4 days. After approximately 10 days, grown organoids were passaged as described below. Derivation and culture of human tracheal fluid organoids (TFO) Tracheal fluids (TF) were collected during procedures of Fetoscopic Endoluminal Tracheal Occlusion (FETO) carried out at the University College London Hospital (UCLH) or KU Leuven, kept refrigerated and processed within 24-48 hours. We collected TF samples before the insertion of the balloon and after its removal. Due to the nature of the TF samples, mostly small and containing a majority of living cells, FACS sorting was not performed. TF was transferred into a 15 mL tube on ice, washed with ice-cold ADMEM+++ and centrifuged at 300 g for 5 min at 4°C. Supernatant was discarded and cells were resuspended in 1 mL of ADMEM+++. Cells were counted and plated in Matrigel droplets. Plates were incubated for 20 min at 37°C and human fetal lung organoid medium (Supplementary Table 2) supplemented with ROCKi and Primocin was added. Medium was changed every 3 days. TFO were clonally expanded and passaged as described below. Passaging of organoids Depending on number and size, organoids were passaged to a 24 or 12 well plate after clonal expansion. Afterwards organoids were usually split 1:2 to 1:3 after 10–14 days of culture. The medium was aspirated and ice-cold ADMEM+++ was added to each well. Matrigel droplets were disrupted and collected into a 15 ml tube on ice. Organoids were washed with 10 mL of cold ADMEM+++ and centrifuged at 300 g for 5 min at 4°C. Big and cystic organoids were resuspended in 1 mL of ADMEM+++ and mechanically disaggregated using a P1000 pipette. If small, organoids were disrupted enzymatically. Medium was aspirated and organoids pellet was resuspended in 300 μL of TryplE. After incubation for 5 min at 37 °C, organoids were pipetted with a P200 to break them down into single cells. Cold ADMEM+++ was added up to 10 mL and the sample was centrifuged at 300 g for 5 min at 4°C. Supernatant was discarded and cell pellet was resuspended in Matrigel and plated. The plate was incubated for 20 min at 37°C to allow the Matrigel to solidify, upon which generic culture medium was added with ROCKi. Medium was changed every 3 days. Fetal tissue samples collection and derivation of control primary fetal organoids Control fetal tissue samples were sourced via the Joint MRC/Wellcome Trust Human Developmental Biology Resource under informed ethical consent with Research Tissue Bank ethical approval (Project 200478: UCL REC 18/LO/0822 - IRAS ID 244325; Newcastle 18/NE/0290 - IRAS ID 250012). The derivation of control fetal organoids was conducted as follows: i) Human fetal small intestinal organoids. Fetal small intestines were processed as previously described76. Tissue was washed with PBS, cleared of any mesenteric tissue and fat, then cut longitudinally. The villi were scratched away using a glass coverslip. The remaining tissue was cut into 2–3 mm pieces, washed vigorously, and incubated in 2 mM EDTA in PBS for 30 min for 5 min on an orbital shaker. The supernatant containing the intestinal crypts was centrifuged at 800 g for 5 min at 4 °C. After being washed in ADMEM+++ and centrifuged, the pellet was resuspended in Matrigel and plated in presence of Primocin and ROCKi. Medium recipe is in Supplementary Table 2. ii) Human fetal kidney tubular organoids The process was adapted following a previously published protocol for deriving adult tubuloids77. Briefly, fetal kidneys were harvested, washed in ice-cold HBSS and minced to isolate the cortical tissue. This was washed in 10 mL of basal medium and supernatant was removed when the tissue pieces were sedimented. After being washed several times in ADMEM+++, tubular fragments were isolated by 1 mg/mL collagenase digestion (C9407, Sigma) on an orbital shaker for 30-45 min at 37°C. Fragments were further washed in basal medium with 2% FBS and centrifuged at 300 g for 5 min at 4°C. Pellet was resuspended in Matrigel and cultured in Kidney medium (Supplementary Table 2) supplemented with ROCKi and Primocin. iii) Human fetal lung organoids Fetal lung tissue was processed adapting a previously published protocol20. Briefly, fetal lungs were minced and washed in ADMEM+++. Tissue fragments were digested in ADMEM+++ containing 1 mg/mL of collagenase (C9407, Sigma) on an orbital shaker at 37°C for 30-60 min. The digested tissue was shacked vigorously and strained over a 100 μm filter. Tissue fragments were washed in ice-cold basal medium with 2% FBS and centrifuged at 300 g for 5 min at 4°C. Supernatant was discarded and pellet was resuspended in Matrigel and cultured in Lung medium (Supplementary Table 2) supplemented with ROCKi and Primocin. iiiv) Human fetal stomach organoids Fetal stomach organoids were isolated from specimens following an established dissociation protocol24. Briefly, stomachs were cut open and mucus was removed with a glass coverslip and mucosa was stripped from muscle layer. Mucosa samples were cut into pieces of 3-5 mm and washed in HBSS until the supernatant was clear. Tissue was incubated in chelating buffer supplemented with 2 mM EDTA for 30 min at RT. Tissue fragments were squeezed with a glass slide to isolate the gastric glands which were transferred in ADMEM+++, strained through at 40 μm and centrifuged at 300 g for 5 min at 4°C. Pellet was resuspended in Matrigel and plated. Gastric medium (Supplementary Table 2) was added with ROCKi and Primocin. Organoid cryopreservation and thawing After 7-10 days of culture, organoids were dissociated enzymatically as described above. The final cell pellet was resuspended in 1:1 ADMEM+++ and freezing medium (80% FBS and 20% DMSO). Cryovials were stored at -80°C overnight and then transferred to LN2 for long term storage. For organoids thawing, cryovials were equilibrated on dry ice and then placed at 37°C. Vial content was rapidly transferred to 15 mL falcon tube containing 9 mL of ice-cold ADMEM+++, then centrifuged at 300 g for 5 min at 4°C. Supernatant was discarded and pellet resuspended in cold Matrigel (Corning). After 20 min of incubation at 37°C, medium was supplemented with ROCKi and replaced after 3 days. Evaluation of organoids formation efficiency and area Organoid formation efficiency was determined by counting the number of organoids at passage 0. The total number of formed organoids per well was manually counted approximately 14 days after seeding of the AF cells in Matrigel. The efficiency was determined by calculating the total number of grown organoids divided by the number of viable single cells initially plated. The organoids area was determined by measuring the perimeter of each organoid in different 5x fields acquired at the Zeiss Axio Observer A1 and using ImageJ software and normalised by field size. Organoid maturation / differentiation Small intestinal AF organoids (siAFO): after manual passaging, organoids were seeded in triplicate in Matrigel and cultured in generic medium. After approximately 7 days, human small intestine medium was used (Supplementary Table 2) for 14 days. Basal culture medium was the same but without addition of CHIR99021, DAPT (notch inhibitor) 10 ^m was added to basal culture medium for 48 hours to stimulate differentiation. Kidney AF organoids (kAFO): after either manual or enzymatic passaging, organoids were seeded in triplicate in Matrigel and cultured in generic medium. After approximately 7-10 days, distal/collecting duct kidney differentiation medium (Supplementary Table 2) was used for 14 days. Lung AF organoids (lAFO): after manual passaging, organoids were seeded in triplicate in Matrigel and cultured in generic medium for approximately 10 days. For lung proximal differentiation, PneumaCult™ ALI Medium (StemCell Technologies, #05001) was then used for 14 days. For distalisation, previously reported medium78 was used for 14 days as well (Supplementary Table 2). Whole-mount immunofluorescence Prior to fixation, organoids were removed from Matrigel using Cell Recovery Solution for 45 min on ice. Organoids were harvested into a 15 mL tube pre-coated with 1% BSA in PBS and fixed with 4% PFA for 20 min at RT. Samples were washed 3 times with PBS for 5 min and spun down at 300 g for 5 min at 4°C. Whole-mount immunostaining was performed by blocking and permeabilising the organoids with PBS-Triton X-1000.5% with 1% BSA for 1 hour at RT. Primary antibodies were incubated in blocking/permeabilisation buffer for 24 h at 4 °C in rotation. After being extensively washed with PBS-Triton 0.2%, organoids were incubated with secondary antibodies overnight at 4 °C in rotation. Organoids were further washed and resuspended in PBS for confocal imaging. For tissue clearing of siAFO ring, previously published protocol was adapted79. EdU staining was performed with the Click-iT EdU Alexa Fluor 568 Imaging kit (Life Technologies) following the manufacturer’s protocol and images were acquired using a Leica SP5 confocal microscope. Full list of antibodies is available in Supplementary Table 3. Image acquisition Phase-contrast images were acquired using a Zeiss Axio Observer A1. Immunofluorescence images of whole-mount staining and sections were acquired on a Zeiss LSM 710 confocal microscope using 25x, 40x, and 63x immersion objectives. Image analysis and z-stacks projections were generated using ImageJ (Rasband, W.S., ImageJ, U. S. National Institutes of Health, Bethesda, Maryland, USA, https://imagej.nih.gov/ij/, 1997-2018). X-Ray phase contrast computed tomography The imaging of the organoids was performed using phase contrast computed tomography (PC- CT) at beamline I13-1 (coherence branch) of the Diamond Light Source (Didcot, UK). The x- ray energy was 9.7 keV and the system resolution 1.6^^ . The organoids were imaged embedded in Histogel (Epredia™ HistoGel™). The PC-CT scan entailed the acquisition of 2000 equally spaced projections through a 180º rotation of the specimen. The total scan time was approximately 1h. The “single image” phase retrieval operation80 was applied to the acquired projections, with the estimated phase to attenuation ratio (refer to as δ/β ratio) set at 250. Both phase retrieval and tomographic slice reconstructions were performed using Savu81 whilst the 3D images were generated using Drishit82,83. siAFO dipeptidyl peptidase IV assay Organoids were plated in 48-well plates, 15 ^l BME/well, in triplicate. Organoids were washed in PBS and then incubated at 37 ^C with 200 ^l /well in Gly-Pro p-nitroanilide hydrochloride (Sigma G0513) dissolved in PBS at a concentration of 1.5mM (or PBS alone in control wells). During incubation, samples were agitated on an orbital shaker (60rpm) and supernatants were sampled at 20, 40 and 60mins. Absorbance (415nm) was measured with a plate reader (Biorad) and concentration determined by comparison to a 4-nitroaniline (Sigma 185310) standard curve (0-200µg/ml) and normalised mg-1 organoid lysate protein (Pierce BCA Protein Assay Kit – ThermoScientific). kAFO Potassium ion channel assay kAFOs were expanded at least in triplicate in a 96 well plate. FluxOR™ II Green Potassium Ion Channel Assay was performed according to the manufacturers’ instructions (F20017, Thermo). Briefly, medium was removed and 80 µL of 1X Loading Buffer were added to each well and incubated for 30 min at RT and 30 min at 37 ºC to facilitate dye entry. After removing the Loading Buffer, 80 µL of Assay Buffer were added to each well. The plate was read using a microplate reader every 5 sec for 5 min, with an excitation wavelength of 480 nm and an emission wavelength of 545 nm. After 5 min of recording, voltage-gated channels were stimulated with 20 µL of High Potassium Stimulus Buffer containing 2 mM Thallium Sulphate (Tl2SO4) and 10 mM Potassium Sulphate (K2SO4). The plate was read once again every 5 sec for 5 min. The average of replicates was normalised on the control (Assay Buffer) and the number of cells. Ciliary beat frequency analysis Organoids were seeded into 8 well glass bottom slide and differentiated towards the lung proximal lineage as described above. For ciliary beat frequency (CBF) analysis, motile cilia grown inside organoids were observed using an inverted microscope system (Nikon Ti-U) with a digital high-speed video camera (Prime BSI Express, Teledyne photometrics). Videos were recorded at a rate of approximately 87 frames/second using a 60x objective. 8-10 videos of individual organoids for each patient were acquired, and the total region of interest (ROI; 1024 x 1024) of each video was then divided into 16 smaller ROIs (sROIs; 256 x 256) for analysis. Every sROI containing clearly visible cilia was counted for 5 full sweeps and the number of frames for 5 full sweeps was integrated into the calculation of CBF (Hz) followed by [87 / (number of frames for 5 sweeps)] x 5. Transmission electron micrograph Organoids in matrix were fixed in a mix of 2.5% Glutaraldehyde and 4% Paraformaldehyde in 0.1M Sorensen’s buffer, pH7.3, and washed with 0.1M Sorensen’s buffer. They were postfixed in 1% aqueous solution of osmium tetroxide, washed and dehydrated through the increasing series of ethanol solutions, followed by propylene oxide (Merck). Organoids were embedded in TAAB812 resin (TAAB Laboratory Equipment Ltd) and sectioned to approximately 70nm thick using a Leica UC7 Ultramicrotome (Leica Microsystems). Sections were collected onto copper mesh grids and contrasted for 2min with 4% uranyl acetate solution in methanol (VWR), followed by 2min in lead citrate (Reynold’s solution). Samples were viewed on a JEOL JEM-1400 TEM (JEOL UK) with an accelerating voltage of 120kV. Digital images were collected with a Xarosa digital camera using Radius software (both from EMSIS). Flow cytometry Cold-stored AF was centrifuged at 300 g for 10 min at 4°C. Supernatant was discarded, pellet resuspended in FBB and strained over a 70 μm and then at 40 μm filter. Cells were incubated with the following fluorochrome-conjugated antibodies: APC/Fire™ 750 anti-human CD324 (E-Cadherin) (Biolegend 324122), PE anti-human CD326 (EpCAM) (324205). RNA isolation and RT-qPCR Organoids were collected from Matrigel with Cell Recovery Solution for 45 min at 4°C. Cells were then washed in ice-cold PBS to remove Matrigel leftover. Organoids were centrifuged at 300 g for 5 min at 4 °C and supernatant discarded. Pellet was resuspended and lysed with RLT buffer (Qiagen). Total RNA was isolated with RNeasy Micro or Mini Kit (Qiagen) following manufacturer’s instructions. RNA concentration was quantified using a Nanodrop (Thermo). cDNA was prepared using High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, #4368813). Quantitative real-time PCR detection was performed using PowerUp™ SYBR® Green Master Mix (Applied Biosystems, A25742) and StepOnePlus Real-Time PCR System (Applied Biosystems). Assays for each sample were run in triplicate and were normalised to the housekeeping gene β-actin. Primer sequences are listed in the Supplementary Table 4. Bulk RNA Sequencing RNA extraction was performed as described above and stored at -80°C until processing. NEBNext Low Input RNA library preparation and sequencing was performed by the UCLgenomics facility. Single-end bulk RNA sequencing was performed on an Illumina NextSeq 2000.100 cycles were run to achieve an average of 5million reads per sample. Transcriptome bioinformatic analysis Quality control was conducted on the FASTQ raw sequences using v0.11.9 FastQC (https://github.com/FelixKrueger/TrimGalore) was used to trim low quality reads (quality 20, length 70). v2.7.1a STAR was applied to align the FASTQ sequences to the NCBI human reference genome GRCh38.p13 (https://www.ncbi.nlm.nih.gov/assembly/GCF_000001405.39/). v1.6.3 featureCounts (https://doi.org/10.1093/bioinformatics/btt656) quantified the expression of individual genes to generate the raw count matrix, using the GRCh38.104 gene annotation (ensembl.org/Homo_sapiens/Info/Index). Default parameters used for both alignment and quantification. The generated count martrix was further processed with a custom R script. Genes will less than 10 reads across 3 samples were removed. Gene IDs were included using the added BioMart package (ensembl.org/info/data/biomart/biomart_r_package.html). Counts per million (CPM) normalisation was completed with the edgeR package (bioconductor.org/packages/release/bioc/html/edgeR.html). ComBat_seq batch correction (rdrr.io/bioc/sva/man/ComBat_seq.html) was applied between the 4 batches. ggplot2 was used throughout for graph generation, including principal component analyses (PCA) and dot plots generated from the normalised count matrix. Clustered heatmaps were generated with Pheatmap (cran.r- project.org/web/packages/pheatmap/), to enable hierarchical comparisons. For the comparison of CDH- and healthy-derived AFO, the pheatmap hierarchical clustering was used to separate the samples for differentially expressed genes (DEGs) analysis using DESeq2 (bioconductor.org/packages/release/bioc/html/DESeq2.html). Volcano plot was generated with ggplot2 to highlight statistically significant DEGs. Single-cell RNA Sequencing Library generation was done following the 10x Genomics Chromium Next GEM Single Cell 3ʹ Reagent Kits v3.1 (Dual Index) kit. Libraries were sequenced using NovaSeq 6000. Data processing was completing with CellRanger. Analysis was done with Seurat V4.1.1 within a custom R script was used for major downstream processing. Cells with less than 150, or more than 8000 features were removed to prevent doublets or cells of low quality. Dying cells with more than 30% mitochondrial genes detected were also removed. Normalisation was then carried out using the NormalizeData function, with a LogNormalize method and a scale factor of 10,000. Batch correction was completed through Seurat’s IntegrateData function, after assessing for integration anchors based on the 2000 most variable features. The object was scaled using ScaleData, RunPCA and FindNeighbors determined for 20 PCs. UMAPs and violin plots were generated using ggplot2 and normalised gene expression always shown, with violin plots showing averaged normalised gene expression within the identified epithelial cluster. v1.6.1 SingleR84 was used to label a single cell experiment object using Human Primary Cell Atlas Data (humancellatlas.org), accessed via celldex (github.com/LTLA/celldex). Statistics and reproducibility Statistical analysis was conducted on data from at least three independent experimental or biological replicates wherever possible, as stated in the figure legends. Results are expressed as the mean ± S.E.M or as the median and quartiles (25% and 75% percentiles) or range. Statistical significance was analysed using unpaired Student’s t-tests for comparisons between two different experimental groups. Statistical significance was assessed using one-way ANOVA with Dunnett’s or Tukey’s post hoc multiple-comparisons test for analysis among more than two groups. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 were considered significant. Exact P values are stated in each relevant figure legend where appropriate. Statistical analysis was carried out using R and GraphPad Prism 9 software. Tables Supplementary Table 1: List of amniotic fluid samples. AF = Amniotic Fluid, TF = Tracheal Fluid, CDH = Congenital Diaphragmatic Hernia, TTTS = Twin to Twin Transfusion Syndrome, FETO = Fetoscopic Endoluminal Tracheal Occlusion, M = Male, F = Female. Sampl GA Type Proce Comor Gend Organ lAFO kAF siAF Organ Sc- e code of dure bidities er oids gener O O oids RN fluid gener ated gener gener seque Aseq ated ated ated nced HO695 15+3 AF M N / / / / Yes 15993 16+0 AF TOP HDBR Y N N Y Yes No fetus HO713 16+0 AF Amni N / / / / Yes odrai nage HO777 16+3 AF Amni MCDA Y Y N N Yes No ocent -Twins esis post- laser HO773 17+0 AF Amni TTTS Y Y N Y Yes No ocent (Stage2 esis ) HO721 18+0 AF Amni MCDA M Y N Y N Yes No odrai TwinsT nage TTS HO680 19+0 AF Amni TTTS M N / / / / Yes odrai nage HO689 19+1 AF Amni M N / / / / Yes odrai nage HO719 19+2 AF Amni TTTS N / / / / Yes odrai nage HO728 20+5 AF Fetal Spina F Y N Y N Yes No surge Bifida ry HO754 20+6 AF Amni TTTS F Y Y Y N Yes No odrai nage HO725 21+0 AF Amni TTTS N / / / / Yes odrai nage HO742 22+2 AF Amni TTTS Y N Y N Yes No odrai (stage2 nage ) HO734 22+3 AF Amni TTTS M Y N Y N Yes No odrai nage HO724 24+0 AF Amni Triplets Y N Y N Yes No odrai nage #6 SB 25+0 AF Fetal Spina Y Y Y N Yes No surge Bifida ry HO727 25+4 AF Fetal Spina F Y Y Y N Yes No surge Bifida ry HO744 27+2 AF Amni MCDA M Y Y Y N Yes No odrai -Twins nage TTTS #11 27+6 AF Pre- CDH Y Y N N Yes No CDH FET O HO760 28+1 AF Pre- CDH-L F Y Y Y N Yes No FET O HO730 28+5 AF Pre- CDH-R M Y Y N N Yes No FET O #4 28+6 AF Pre- CDH-L Y N N N No No CDH FET O #2 29+3 AF Pre- CDH-R F Y Y N N Yes No CDH FET O #5 29+4 AF Pre- CDH-R M Y Y Y N Yes No CDH FET O HO768 29+6 AF Pre- CDH-L F Y No No FET O HO758 30+2 AF Amni Polyhy M Y Y Y N Yes Yes odrai dramin nage ous HO741 30+6 AF Amni Late F Y N Y N Yes Yes odrai onset nage Hydrop s #10 31+4 TF Pre- CDH Y Y N N Yes No CDH FET O HO696 32+0 AF Suspect M N / / / / Yes ed syndro me - cardiac/ bilatera l talipes/ polyhy droamn ios HO730 32+6 TF Post- CDH-R M Y Y N N Yes No FET O HO768 33+6 TF Post- CDH-L F Y Y N N Yes No FET O HO768 33+6 AF Post- CDH-L F Y Y N N Yes No FET O #9 33+6 AF Post- CDH-L M Y Y N N Yes No CDH FET O #9 33+6 TF Post- CDH-L M Y Y N N Yes No CDH FET O HO687 34+4 AF Amni M Y Y Y N Yes Yes odrai nage The following tables detail medium that may be used in various steps of the methods described herein. The methods herein may use the components as listed, or alternative components may be provided. The skilled person will recognise that components can be substituted with the same specified or generic components from different sources/manufacturers or brands. The skilled person will appreciate, that the exact content and quantities in such mediums may be varied without affecting the ability of the medium to support a particular cell culture, maturation or differentiation. For example, at least 10% variation in quantities may be tolerated. Furthermore, particular agents, substances, inhibitors, growth factors and cytokines may be substituted with another which may perform a similar function. Supplementary Table 2: media compositions AFO expansion medium Component Final Concentration Advanced DMEM (Thermo 12634) To volume HEPES (Thermo 15630080) 10 mM Glutamax (Thermo 35050061) 2 mM Pen/Strep (Thermo 15140122) 1% B-27 supplement minus Vitamin A (Thermo 12587010) 1X n-acetylcysteine (Sigma A9165) 1.25 mM Wnt-3A (Peprotech 315-20) 100 ng/mL R-spondin 1 (Peprotech 120-38) 500 ng/mL Noggin (Peprotech 120-10C) 100 ng/mL EGF (Thermo PMG8043) 50 ng/mL TGFb inhibitor (A83-01) (Sigma SML0788) 5 μM GSK-3 inhibitor (CHIR 99021) (Tocris 4423) 3 μM ROCK inhibitor Y-27632 (Tocris 1254) (Add to single cells) 10 μM (alternatively 10μg/ml) Human fetal lung organoids and lTFO medium Component Final Concentration Advanced DMEM (Thermo 12634) To volume HEPES (Thermo 15630080) 10 mM Glutamax (Thermo 35050061) 2 mM Pen/Strep (Thermo 15140122) 1% B-27 supplement minus Vitamin A (Thermo 12587010) 1X n-acetylcysteine (Sigma A9165) 1.25 mM N2 (Thermo 17502048) 1X R-spondin 1 (Peprotech 120-38) 100 ng/mL Noggin (Peprotech 120-10C) 100 ng/mL EGF (Thermo PMG8043) 50 ng/mL FGF10 (Peprotech 100-26) 100 ng/mL FGF7 (Peprotech 100-19) 100 ng/mL TGFb inhibitor (SB 431542) (Tocris 1614) 10 μM GSK-3 inhibitor (CHIR 99021) (Tocris 4423) 3 μM ROCK inhibitor Y-27632 (Tocris 1254) (Add to single cells) 10 μM (alternatively 10μg/ml) Human fetal small intestinal organoids medium Component Final Concentration Advanced DMEM (Thermo 12634) To volume HEPES (Thermo 15630080) 10 mM Glutamax (Thermo 35050061) 2 mM Pen/Strep (Thermo 15140122) 1% B-27 supplement minus Vitamin A (Thermo 12587010) 1X n-acetylcysteine (Sigma A9165) 1.25 mM Wnt-3A (Peprotech 315-20) 100 ng/mL R-spondin 1 (Peprotech 120-38) 500 ng/mL Noggin (Peprotech 120-10C) 100 ng/mL EGF (Thermo PMG8043) 50 ng/mL Gastrin (Sigma G9020) 10 nM TGFb inhibitor (A83-01) (Sigma SML0788) 500 nM GSK-3 inhibitor (CHIR 99021) (Tocris 4423) 3 μM P38 inhibitor (SB202190) (Sigma S7067) 10 μM Prostaglandin E2 (Cambridge cay14010) 10 nM ROCK inhibitor Y-27632 (Tocris 1254) (Add to single cells) 10 μM (alternatively 10μg/ml) Human siAFO maturation medium Component Final Concentration Advanced DMEM (Thermo 12634) To volume HEPES (Thermo 15630080) 10 mM Glutamax (Thermo 35050061) 2 mM Pen/Strep (Thermo 15140122) 1% B-27 supplement minus Vitamin A (Thermo 17504044) 1X Nicotinamide (Sigma N0636) 10 mM n-acetylcysteine (Sigma A9165) 1 mM Wnt-3A (Peprotech 315-20) 0.1 nM R-spondin from 293T-HA-Rspol-Fc cells 20% vol/vol Noggin from HEK293 mNoggin-producing cells 10% vol/vol EGF (Thermo PMG8043) 50 ng/mL Gastrin (Sigma G9020) 10 nM TGFb inhibitor (A83-01) (Tocris 2939) 500 nM GSK-3 inhibitor (CHIR 99021) (Tocris 4423) 3 μM P38 inhibitor (SB202190) (Sigma S7067) 10 μM ROCK inhibitor Y-27632 (Tocris 1254) (Add to single cells) 10 μM (alternatively 10μg/ml) Human fetal kidney organoid medium Component Final Concentration Advanced DMEM (Thermo 12634) To volume HEPES (Thermo 15630080) 10 mM Glutamax (Thermo 35050061) 2 mM Pen/Strep (Thermo 15140122) 1% B-27 supplement minus Vitamin A (Thermo 12587010) 1X n-acetylcysteine (Sigma A9165) 1 mM R-spondin 1 (Peprotech 120-38) 100 ng/mL EGF (Thermo PMG8043) 50 ng/mL TGFb inhibitor (A83-01) (Sigma SML0788) 2 μM GSK-3 inhibitor (CHIR 99021) (Tocris 4423) 3 μM FGF10 (Peprotech 100-26) 100 ng/mL GDNF (Peprotech 450-10) 50 ng/mL Heparin (Sigma H3393) 1 μg/mL LDN193189 dihydrochloride (Cambridge Bioscience SM23-5) 200 nM ROCK inhibitor Y-27632 (Tocris 1254) (Add to single cells) 10 μM (alternatively 10μg/ml) Human fetal stomach organoid medium Component Final Concentration Advanced DMEM (Thermo 12634) To volume HEPES (Thermo 15630080) 10 mM Glutamax (Thermo 35050061) 2 mM Pen/Strep (Thermo 15140122) 1% B-27 supplement minus Vitamin A (Thermo 12587010) 1X n-acetylcysteine (Sigma A9165) 1.25 mM Wnt-3A (Peprotech 315-20) 100 ng/mL R-spondin 1 (Peprotech 120-38) 500 ng/mL Noggin (Peprotech 120-10C) 100 ng/mL EGF (Thermo PMG8043) 50 ng/mL Gastrin (Sigma G9020) 10 nM FGF10 (Peprotech 100-26) 200 ng/mL TGFb inhibitor (A83-01) (Sigma SML0788) 5 μM GSK-3 inhibitor (CHIR 99021) (Tocris 4423) 3 μM ROCK inhibitor Y-27632 (Tocris 1254) (Add to single cells) 10 μM (alternatively 10μg/ml) Human distal tubule/collecting duct kidney differentiation medium Component Final Concentration Advanced DMEM (Thermo 12634) To volume HEPES (Thermo 15630080) 10 mM Glutamax (Thermo 35050061) 2 mM Pen/Strep (Thermo 15140122) 1% Aldosterone (Sigma A9477) 10 nM Vasopressin (Sigma V9879) 10 nM Human distal lung medium A) cSFDM (Complete serum free differentiation media) Component Final Concentration IMDM (Thermo 12440053) 75% Ham’s F12 (Cellgro 10-080-CV) 25% B-27 (with RA) supplement (Invitrogen 1750444) 1% N-2 supplement (Invitrogen 17502048) 0.5% BSA (Sigma A7030) 0.05% Primocin (Thermo NC9392943) 100 μg/mL Glutamax (Thermo 35050061) 1X Ascorbic Acid (Sigma A4544) 50 μg/mL MTG (Sigma M6145) 450 μM B) CK + DCI media Component Final Concentration cSFDM Base To volume GSK-3 inhibitor (CHIR 99021) (Tocris 4423) 3 μM FGF7 (Peprotech 100-19) 10 ng/mL Dexamethasone (Sigma D4902) 50 nM 8BrcAMP (Sigma B7880) 0.1 mM IBMX (Sigma I5879) 0.1 mM Supplementary Table 3: list of antibodies Antibody/conjugated molecules Dilution EpCAM (Abcam ab71916) 1:100 E-cadherin (BD 610182) 1:200 PDGF Receptor α (Cell signaling 3174) 1:200 Integrin β-4 (Abcam ab110167) 1:100 Zonula occludens-1 (Invitrogen 40-2200) 1:100 Pan Cytokeratin (Abcam ab7753) 1:100 Ki-67 (Abcam ab15580) 1:100 Cleaved Caspase-3 (Cell Signaling 9661) 1:100 Olfactomedin 4 (Cell Signaling 14369T) 1:50 Cytokeratin 20 (Proteintech 60183-1-lg) 1:100 Lysozyme (Biorad 5790-4110) 1:50 Fatty Acid Binding Protein 1 (R&D AF1565) 1:100 Chromogranin A (Abcam ab15160) 1:1000 Mucin 2 (Santa Cruz sc-15334) 1:200 PAX8 (Abcam ab191870) 1:200 LIM1/LHX1 (Abcam ab229474) 1:100 GATA-3 (R&D AF2605) 1:200 Lotus Tetragonolobus Lectin, Fluorescein (Vector Laboratories FL-1321-2) 1:300 Acetylated α Tubulin (Santacruz sc-23950) 1:200 Calbindin 1 (Abcam ab108404) 1:100 SLC12A1 (Abcam ab171747) 1:100 Aquaporin-2 (Biotechne NB110-74682) 1:300 TTF1 (NKX2-1) (Abcam ab76013) 1:200 p63 (Abcam ab53039) 1:100 SOX2 (Abcam ab97959) 1:200 SOX9 (R&D AF3075) 1:100 Prosurfactant Protein C (Merck AB3786) 1:500 FoxJ1 (R&D AF3619) 1:100 Keratin 5 (BioLegend 905501) 1:100 Mucin 5AC (Invitrogen MA5-12178) 1:100 Surfactant Protein B (Thermo PA5-42000) 1:500 Phalloidin 488 (Thermo A12379) 1:100 Phalloidin 647 (Sigma Aldrich 65906) 1:100 Alexa Fluor Donkey anti-Mouse 488 (Thermo A21202) 1:300 Alexa Fluor Goat anti-Rabbit 488 (Thermo A11008) 1:500 Alexa Fluor Donkey anti-Rabbit 568 (Thermo A10042) 1:300 Alexa Fluor Donkey anti-Mouse 546 (Thermo A10036) 1:300 Alexa Fluor Donkey anti-Goat 633 (Thermo A21082) 1:300 Alexa Fluor Donkey anti-Rabbit 647 (Thermo A31573) 1:300 Alexa Fluor Goat anti-Rabbit 647 (Thermo A21244) 1:300 Hoechst 33342 (Thermo H1399) 1:300

Supplementary Table 4: list of primers Primer Name Forward Reverse human β-ACTIN ATGGTGGGCATGGGTCAGA GCAACGTACATGGCTGGGG human OLFM4 CTTTCCAAAGTGAGGGAATATGTC GATGTCAATTCGGACAGTTAGG human LGR5 CAGTGCAGTGTTCACCTTCC AGTGCCAGAACTGCTATGGT human LYZ CATTGTTCTGGGGCTTGTCC TCATTACACCAGTAGCGGCT human MUC2 CAACAACTCCGAAGCTGTG CAAATGTTTCTCGGTCACC human ALPI TCATCATGAGGGTGTGGCTT TGTAGGCTTTGCTGTCCTGA human FABP1 AAGACAGTGGTTCAGTTGGAAG TGAGTTCGGTCACAGACTTGAT human HES1 AGCACACTTGGGTCTGTGC TGAAGAAAGATAGCTCGCGG human AXIN2 GACAGGAATCATTCGGCCAC CCTTCAGCATCCTCCGGTAT human DLL1 ACTCCTACCGCTTCGTGTGT CAGGGTTGCACACTTTCTCC Human ATOH1 GCAGGAGGAAAACAGCAAAA ACTTGCCTCATCCGAGTCAC Example 2 - Derivation of primary organoids from mouse and sheep fetal fluid Mouse and sheep amniotic and tracheal fluids are collected from pregnant mice or sheep and placed in tubes containing ADMEM+++ at 4 ^C. The fluid is filtered over a 70 ^m cell strainer and centrifuged at 300g for 10min at 4 ^C. Supernatant is removed, and pellet resuspended in 1mL ADMEM+++. After cell count and viability check, the cells are washed by adding further 9ml ADMEM+++. Cells are centrifuged at 300g for 10min at 4 ^C and supernatant aspirated. Cell pellet is resuspended in Matrigel at a density of 100.000 cells/30 ^L Matrigel and seeded into a pre-warmed well plate. After 30 min of incubation at 37 ^C, generic medium supplemented with ROCK inhibitor and Primocin is added. Once grown, mouse or sheep fetal fluid-derived organoids are then cultured and expanded clonally as described for the human fetal fluid-derived organoids. Example 3 – Generation of assembloids from fetal fluids-derived primary organoids Lung fetal fluid-derived organoids are cultured for 7-10 days in expansion medium and subsequently differentiated into proximal lung. After 10-12 days of differentiation, organoids are gently collected with a P1000 pipette and transferred to a BSA-pre coated tube in 1mL of Cell Recovery solution. To allow Matrigel depolymerization organoids are incubated for 1h at 4 ^C. Organoids are then washed in 10mL of ADMEM+++ and centrifuged at 100g for 1 min at 4 ^C. Supernatant is discarded, organoids resuspended in proximal lung medium at a ratio of 400 ^L medium per 15-20 ^L original Matrigel droplet and plated into ultra-low attached 24 well plate (500 ^L per well. Medium is changed every 2-3 days by removing 400 ^L of used medium and by replacing 400 ^L fresh medium. Assembloids are cultured in suspension for 12-15 days. Example 4 - Derivation of primary neurospheres from fetal fluids Human amniotic or cerebrospinal fluid is collected and transferred to 15 or 50mL falcon tube depending on fluid volume and centrifuged for 10min at 4 ^C. Supernatant is removed, and pellet washed in ADMEM+++. After being centrifuged at 300g for 5min at 4 ^C, the supernatant is removed, and pellet resuspended in 1mL of neurosphere medium (Table below). After cell count and viability check, the cell suspension is plated in ultra-low-attachment (or BSA pre- coated) 6 well plates (2.2 ml per well). ROCK inhibitor is added for the first 3-4 days of culture. After 2-4 days of culture, debris are removed and re-plated in another 6 well plate. Medium is added to both the original and second plate. Once neurospheres are grown, they are picked and expanded as bulk cultures or clonally. Medium for fetal fluid-derived neurospheres. The following table details medium that may be used in various steps of the methods described herein. The methods herein may use the components as listed, or alternative components may be provided. The skilled person will recognise that components can be substituted with the same specified or generic components from different sources/manufacturers or brands. The skilled person will appreciate, that the exact content and quantities in such mediums may be varied without affecting the ability of the medium to support a particular cell culture, maturation or differentiation. For example, at least 10% variation in quantities may be tolerated. Furthermore, particular agents, substances, inhibitors, growth factors and cytokines may be substituted with another which may perform a similar function. Component Stock conc. Final conc. in 50 ml Advanced DMEM F-12 (Thermo 12634) - To volume _{24 ml} HEPES (Thermo 15630080) 1 M 10 mM _{250 ul} Glutamax (Thermo 35050061) 100 X 2 mM _{250 ul} Pen/Strep (Thermo 15140122) 100% 1% _{250 ul} Neurobasal medium 24 ml Pen/Strep (Thermo 15140122) 250 ul Primocin (Thermo Fisher NC9392943) 50 mg/mL 100 mg/mL _{100 ul} B-27 supplement minus vitamin A (Thermo 12587010) 50 X 1 X 1 ml N2 supplement 100X 1X 500 ul MEM non essential amino acid solution 100X 1X 500 ul hFGF-10 100 µg/mL 50 ng/ml 25 ul hFGF-2 100 µg/mL 40 ng/ml 20 ul hEGF 500 microg/ml 50 ng/ml 5 ul

Claims

CLAIMS 1. A method of generating tissue-specific primary fetal organoids from an epithelial stem cell population wherein said method comprises: iii) isolating a viable population of cells comprising epithelial stem cells from a fetal fluid; iv) seeding the population of cells comprising the epithelial stem cells into an extracellular matrix (ECM) and culturing the seeded cells in a cell growth medium for the formation of clonal organoids from the epithelial cells; v) extracting a clonal organoid from the ECM and dissociating the clonal organoid into single cells for passaging; vi) culturing the single cells such that they expand into tissue-specific primary fetal organoids.

2. The method according to claim 1, wherein the fetal fluid is collected during gestation.

3. The method according to claim 1 or 2, wherein the fetal fluid is one or more of amniotic fluid (AF), tracheal fluid (TF), vernix caseosa, fetal urine, meconium, saliva, ascites fluid, pleural fluid, aqueous humor, lacrimal fluid, nasal mucus, and cerebrospinal fluid.

4. The method according to any preceding claim, wherein the fetal fluid is amniotic fluid (AF) and/or tracheal fluid (TF).

5. The method according to any preceding claim, wherein viable cells are identified and isolated away from non-nucleated cells, cell debris, urea crystals and/or other non-cellular particles.

6. The method according to any preceding claim, wherein the epithelial stem cells are isolated from the population of cells prior to seeding step ii), optionally wherein the epithelial stem cells are isolated using fluorescence-activated cell sorting (FACS) or magnetic cell sorting (MACS).

7. The method according to claim 5, wherein the heterogeneity of the cells is preserved during FACS by not selecting the epithelial stem cells by Forward Scatter or Side Scatter.

8. The method according to any preceding claim, wherein the epithelial stem cells may be identified and isolated based on the expression of one or more of the epithelial marker genes.

9. The method according to any one of claims 1-4, wherein the viable population of cells are seeded into the ECM without cell sorting.

10. The method according to any preceding claim, wherein the ECM is in the form of a droplet.

11. The method according to any preceding claim, wherein the ECM is about 5-50µl in volume.

12. The method according to any preceding claim, wherein the epithelial stem cells may be seeded in an amount up to about 6x104 live cells/ECM substrate, such as an ECM droplet.

13. The method according to any preceding claim, wherein the epithelial stem cells are seeded in an amount of about 50 to about 6x104 live cells/ECM droplet, such as an 30µl ECM droplet.

14. The method according to any preceding claim, wherein culturing the single cells such that they expand into tissue-specific primary fetal organoids comprises culturing the single cells in ECM in cell growth medium.

15. The method according to claim 14, wherein the cell growth medium is supplemented with Rho-kinase inhibitor.

16. The method according to any preceding claim, wherein the primary fetal organoids are further matured and/or differentiated into a tissue or organ type.

17. The method according to any preceding claim, wherein the primary fetal organoids are further matured and/or differentiated into small intestinal AF organoids (siAFO), kidney AF organoids (kAFO), lung AF organoids (lAFO), skin organoids, placental organoids, corneal organoids, oesophageal organoids, gastric organoids, large intestine organoids, or lacrimal gland organoids.

18. The method according to any preceding claim, wherein the fetal organoids are derived from the fetal fluid of fetuses with a congenital disorder, such as Congenital Diaphragmatic Hernia (CDH) fetuses, spina bifida, oesophageal atresia, or twin to twin transfusion syndrome.

19. The method according to any preceding claim, further comprising the formation of a 3D culture model comprising one or more of the primary fetal organoids.

20. A primary fetal organoid formed by the method according to any preceding claim.

21. A primary assembloid model formed from two or more primary fetal organoids according to claim 20, or from two or more primary fetal organoids derived from the method according to any of claims 1-19.

22. A disease model, wherein the disease model comprises a primary fetal organoid according to claim 20 or a primary assembloid according to claim 21, optionally wherein the primary fetal organoid has a disease characteristic, such as a congenital disorder, and/or the primary fetal organ has been treated with an agent to induce a disease state.

23. Use of the disease model according to claim 22 to identify agents capable of preventing or treating the disease, wherein the primary fetal organoid is treated with a potential therapeutic or prophylactic agent; and optionally determining if the treatment by the potential therapeutic or prophylactic agent has any effect in inhibiting or reducing the development of disease, or the reduction in disease after it has developed.

24. A method of screening for agents capable of preventing or treating a disease, the method using the primary fetal organoid according to claim 20, or a primary assembloid according to claim 21, or the disease model of claim 22, wherein the primary fetal organoid is genetically manipulated or treated with an agent before, during or after the primary fetal organoid is treated to induce disease; and determining if the agent or genetic manipulation has any effect in inhibiting or preventing the development of disease in the primary fetal organoid, or the reduction in disease after it has developed in the primary fetal organoid.

25. A 3D culture model comprising a primary fetal organoid according claim 20 or a primary assembloid according to claim 21.

26. A method for producing a product from a primary fetal organoids according to claim 20 or a primary assembloid according to claim 21, the method comprising the incubation of the primary fetal organoids or primary assembloid in vitro, and harvesting the product produced from the primary fetal organoids or primary assembloid.

27. A method of screening for biomarkers of a disorder in an organ or tissue, the method comprising the monitoring of biomarkers released from the primary fetal organoids according to claim 20 or a primary assembloid according to claim 21, or cells engrafted therein, or biomarkers in tissue or cellular extracts of the primary fetal organoids according to claim 20 or a primary assembloid according to claim 21.

28. A method of producing one or more specific cell types, wherein the one or more specific cell types are harvested from a fetal primary organoid according to claim 20 or a primary assembloid according to claim 21.

29. A method of generating tissue-specific primary fetal organoids from an epithelial stem cell population wherein said method comprises: i) isolating a viable population of cells comprising epithelial stem cells from a fetal fluid; ii) culturing the cells in a cell growth medium for the formation of organoids from the epithelial cells; and optionally iii) further passaging and culturing the organoids for expansion, or extracting an organoid from the cell culture and dissociating the organoid into single cells for passaging, and culturing the single cells such that they expand into tissue-specific primary fetal organoids.