Talk Summary:
Central nervous system drugs fail in the clinic more than almost any other class, largely because rodent brains do not reproduce human neural, immune, and vascular biology at the same time. My lab builds iPSC-derived multi-region brain organoids (MRBOs) that recover close to 80 percent of the human fetal brain transcriptome and, unlike standard organoids, vascularize on their own. An endothelial-hematopoietic transition inside the tissue seeds microglia-like cells and assembles a rudimentary blood-brain barrier, so one model carries cortical and subcortical regions, resident immune cells, and a perfusable vascular interface together.

My core readouts are functional and molecular: multielectrode array electrophysiology for network activity, transcriptomics for cell-state and pathway response, and high-plex proteomics for secreted signaling. I will show how we run this across neuroinflammation, neurodegeneration, and traumatic brain injury programs, and how the data hold up as a human NAM under the FDA and NIST roadmap.

For this community, the frontier is convergence. A human, barrier-bearing, self-vascularized brain model is exactly the substrate molecular imaging has lacked: a place to test probe and tracer penetrance, target engagement, and spatial drug response in human tissue without animals. I will close by mapping where imaging modalities could plug into these organoids and what that opens up for CNS translation.

Author
Annie Kathuria
Johns Hopkins University