Frischmann, M. | Search bioRxiv

bioRxiv2026

Human cerebral assembloids as a model to study the interaction of glioblastoma with resident glial cells

Glioblastoma (GB) is a highly invasive brain cancer characterized by extensive peritumoral glial reactivity. Although GB infiltration and recurrence remain major therapeutic challenges, the contribution of GB-induced gliosis to tumor progression is poorly understood. To investigate glia-GB interactions, we developed human iPSC-derived cerebral forebrain assembloids containing innate microglia and macroglia and transplanted them with GB cells. The model recapitulates key tumor features observed in patients, including distinct tumor regions, reactive gliosis, and altered tissue stiffness. Single-cell RNA sequencing revealed an unexpected enrichment of extracellular vesicles (EVs) in peritumoral reactive glia, which we validated in patient samples. EV inhibition impaired formation of the peritumoral niche, whereas EVs from GB-containing assembloids induced glial reactivity in GB-free assembloids. Furthermore, the modularity of our system enables incorporation of additional niche components, such as monocytes, allowing investigation of the monocyte-glia-tumor axis and providing a powerful platform to dissect mechanisms regulating the GB microenvironment.

neuroscience↗

bioRxiv2024

Scale invariance of mechanical properties in the developing mammalian retina

The computational capabilities of the human central nervous system arise from neuronal architectures. Neuronal tissues are built through force-driven physical remodeling over durations ranging from seconds to days. However, how cell-generated forces accumulate and relax to drive neurogenesis remains unknown due to the difficulty of applying stresses (i) in developing mammalian nervous tissue, (ii) directly within the cellular microenvironment, and (iii) for durations spanning seconds to hours. Previous studies have shown that non-neuronal tissues and cells in 2D culture remodel in scale-free manners. Whether and how this translates to developing neuronal tissues remains an open question. Here, we probed the mechanics of mammalian neuronal tissue on developmental timescales. We accessed developing mammalian neuronal tissue using retinal organoids, stem-cell-derived models of the retina. To probe mechanics at scales relevant for retinal development, we used magnetic droplets as long-term mechanical actuators. We recorded strain responses to applied stresses across four orders of magnitude in time, up to one hour. We found that dynamic creep compliance and tensile moduli follow a power law with an exponent consistent with a material just above the glass transition. This scale-free rheology represents an unprecedented description of nervous tissue mechanics on developmental timescales and opens the door to a biophysical understanding of the emergence of functional neuronal architectures.

bioengineering↗