Prenatal diagnosis of congenital anomalies relies on genetic and imaging analyses, but predicting severity remains challenging. Personalized prenatal therapies have shown improved outcomes for conditions like CDH, TTTS, and MMC, but patient selection for interventions like vesico-amniotic shunting for LUTO remains a hurdle. The lack of autologous models of developing human tissues is a major limitation. Organoids, three-dimensional in vitro models, offer a potential solution. While autologous organoids can be derived from human embryonic stem cells or induced pluripotent stem (iPS) cells, the extensive manipulation and lengthy differentiation protocols reduce patient fidelity. Primary organoids, derived from postnatal samples (urine, menstrual flow, etc.), require minimal manipulation but are limited by ethical and legal restrictions surrounding the use of fetal tissue obtained post-mortem. The current methods are destructive, limiting their use for prenatal modeling, diagnostics, and regenerative medicine, and primarily focus on early gestational stages. This research aims to derive primary human fetal epithelial organoids from fetal fluids collected during prenatal diagnostic and therapeutic interventions (second and third trimesters), enabling organoid generation alongside the continuation of pregnancy. Amniocenteses, amniodrainage for polyhydramnios and TTTS, and procedures for MMC repair and FETO for CDH provide access to AF and tracheal fluid (TF). AF is heterogeneous, containing cells from various fetal tissues (gastrointestinal tract, kidney, lung), but a detailed map of its epithelial populations is lacking. This study investigates whether AF harbors lineage-committed progenitors capable of forming tissue-specific primary fetal organoids, expanding the analysis to TF from CDH cases during FETO.

Numerous studies have explored the use of organoids for disease modeling and therapeutic development. iPS cell-derived organoids have been generated from reprogrammed AF-derived fetal cells, but the extensive manipulation and lengthy differentiation protocols limit their fidelity and applicability for prenatal disease modeling and targeted therapy. Primary organoids derived from discarded postnatal biological samples have shown promise, but their application in prenatal medicine has been hampered by ethical considerations surrounding the use of fetal tissues obtained after pregnancy termination. Existing fetal tissue-derived organoid models offer valuable insights but are limited in terms of accessibility, ethical approvals, and the ability to model later gestational stages.

Amniotic fluid (AF) samples were collected from 12 pregnancies (15-34 gestational age weeks) after written informed consent. Viable nucleated cells with intact cell membranes were isolated using fluorescence-activated cell sorting (FACS). 3' single-cell RNA sequencing (scRNA-seq) was performed, and unsupervised UMAP analysis was used to characterize the cellular composition of the AF. The SingleR package and expression of pan-epithelial marker genes were used to identify epithelial clusters. scGSEA was used to identify cells with gastrointestinal, kidney, and lung signatures. Viable AF cells were seeded in Matrigel droplets and cultured in a defined epithelial medium to generate amniotic fluid organoids (AFOs). Clonal lines were established by picking individual AFOs, dissociating them into single cells, and replating. Bulk RNA sequencing and scRNA-seq were used to characterize AFOs. Small intestinal AFOs (SIAFOs), kidney tubule AFOs (KAFOs), and lung AFOs (LAFOs) were characterized based on their morphology, gene expression profiles, and functional assays. Lung organoids were also derived from tracheal fluid (TF) of fetuses with CDH. Organoid maturation assays, immunostaining, functional assays (dipeptidyl peptidase IV and disaccharidase activity for SIAFOs, potassium ion channel assay and epithelial barrier integrity assay for KAFOs, ciliary beat frequency analysis for LAFOs), and imaging techniques (X-ray phase-contrast computed tomography (PC-CT), microCT, and transmission electron microscopy (TEM)) were used. Statistical analyses included t-tests, ANOVA, and correlation analysis.

Single-cell analysis of AF revealed the presence of epithelial cells originating from multiple developing tissues, including gastrointestinal, renal, and pulmonary progenitors. These progenitors formed clonal AFOs that could be expanded for multiple passages and cryopreserved. Bulk RNA-seq and scRNA-seq confirmed the tissue-specific identity of the AFOs. SIAFOs exhibited crypt-villus structures and expressed markers of various intestinal cell types; they also showed functional brush border enzyme activity and could self-assemble into intestinal rings. KAFOs displayed tubuloid morphology, expressed renal epithelial markers, exhibited functional tight junctions and potassium ion channels. LAFOs expressed pulmonary markers and could be differentiated into proximal and distal airway phenotypes. Lung organoids derived from both AF and TF of CDH fetuses exhibited features of the disease, including altered surfactant gene expression and changes in cell populations. The entire workflow, from fluid sampling to organoid characterization, was completed within 4-6 weeks.

This study demonstrates a novel method for deriving autologous primary fetal organoids from easily accessible fetal fluids during ongoing pregnancies. The minimally invasive sampling method and rapid workflow makes the technique suitable for prenatal intervention, counseling, and personalized therapy. The AFOs offer significant advantages over iPS cell-derived organoids, requiring substantially less time and manipulation, thus preserving patient fidelity. The successful derivation of tissue-specific organoids from AF highlights the potential of this approach for modeling various fetal developmental processes and diseases. The generation of CDH organoids from AF and TF samples before and after FETO allows for the longitudinal study of disease progression and response to interventions. Future studies will focus on expanding the panel of AF-derived organoids to incorporate mesenchymal and vascular components, creating more complex in vitro models of prenatal development. The platform's potential to inform better patient stratification and design personalized prenatal therapies is considerable.

This study presents a novel method for deriving primary fetal epithelial organoids from amniotic and tracheal fluids collected during ongoing pregnancies. The resulting organoids demonstrate tissue-specific characteristics and can be used to model fetal development and disease. The technique offers considerable promise for advancing prenatal diagnostics and personalized therapies for congenital anomalies.

The study is limited to the epithelial compartment and doesn't fully capture the complexity of organogenesis, which often involves interactions with mesenchymal and vascular tissues. The sample size for certain analyses, particularly those involving CDH organoids, was constrained by the availability of patient samples. Future studies should address these limitations by incorporating additional cell types and expanding the patient cohort.