Modeling prodromal mechanisms of Alzheimer’s disease in human Forebrain Organoids

Forebrain organoids (FOs) are 3D self-organizing neural constructs derived from stem cells that recapitulate key developmental and functional features of the human brain to investigate neurodegenerative disorders. Unlike transgenic animal models, FOs more faithfully reproduce the molecular and cellular landscape of the hCNS, enabling direct observation of alterations in patient-related conditions. A limitation of conventional FOs is the absence of intrinsic vascularization, which compromises oxygen and nutrient diffusion and leads to central necrosis and structural disintegration, particularly at long term stages. Moreover, considerable intra batch heterogeneity in shape, cytoarchitecture, and cell type composition remains a critical challenge for reproducibility. To address these limitations preserving physiological relevance, we developed a pivotal streamlined and highly reproducible protocol for generating human FOs under static conditions without orbital shakers, microfluidic devices, or bioreactors, which are inconsistent with the static, cerebrospinal fluid buffered environment of the adult brain. Guided by allometric scaling principles and after optimization of media composition and culture parameters, we introduced a transient increase in Fibroblast Growth Factor 2 (FGF2) concentration immediately prior to neural induction. This modification yielded uniform growing that led to viable FOs at 60 DIV. FGF2 acted both to enhance early neural activation and to sustain pluripotency associated networks, suggesting an imprinting-like effect that supported growth despite the absence of vascularization. Morphometric assessments showed a steady increase in hESCs-derived FOs area and volume over time, with a plateau between 30 and 60 DIV, indicating structural maturation rather than unchecked expansion. Weight values decreased between 30 and 60 DIV, possibly due to differences in internal density and lumen-like structures developing in cerebral organoids. Statistical analyses revealed low intra batch variability in morphometric parameters, underscoring protocol robustness. Immunofluorescence at 60 DIV revealed expression of canonical neural markers including TUJ1, MAP2, SOX2, PAX6, GFAP, and SYNAPSIN 1, indicating the coexistence of proliferative progenitor and differentiating neural cells. SOX2 and PAX6-positive regions corresponded to dense proliferative zones, while high MAP2 and SYN1 expression, detected in interventricle-like regions, marked emerging mature neuronal populations, consistent with established organoid radial maturation patterns. When applied to clinical case studies, the protocol revealed clear disease specific phenotypes. Healthy control (HC) FOs exhibited robust growth, regular morphology, and balanced marker expression, stabilizing between 30 and 60 DIV. In contrast, Alzheimer’s disease (AD) FOs showed an altered growth, reduced survival, and early molecular features of pathology, including increased p-tau and SIRT1, despite the absence of β-amyloid plaques at 60 DIV. Mild Cognitive Impairment (MCI) FOs demonstrated an intermediate and dynamic phenotype, with a 20 DIV growth similar to AD, a 30 DIV transient alignment with HC growth dynamics, and heightened expression of SOX2, HIF 1α, PSEN1/2, and GFAP at 60 DIV. These data suggest early activation of hypoxia-related proliferative pathways prior to advanced late stage neurodegenerative hallmarks. Finally, the optimized FGF2-based protocol enables the generation of human FOs that are structurally complex, long-time viable, and capable of recapitulating both normal neurodevelopment and early disease associated modifications. By removing the need for external mechanical stimulation and emphasizing physiological culture conditions, this approach provides a scalable, reproducible, and cost-efficient platform for modeling neurodevelopment and neurodegeneration, but also for bridging experimental biology with computational sciences, advancing precision medicine.