Growing the blood-brain barrier in a dish
The brain's border wall keeps drugs out of the brain — a headache for medicine and a marvel of biology. Human stem-cell models now let us grow that wall in the lab. Here's how it works and what it's revealing.
Every sip of coffee that sharpens your morning crosses a checkpoint. So does every antibiotic, antidepressant, and experimental Alzheimer’s drug — or fails to. The checkpoint is the blood-brain barrier (BBB): a lining of tightly joined cells along the brain’s blood vessels that decides, molecule by molecule, what gets in.
The barrier is magnificent at its job. It is also medicine’s most stubborn gatekeeper — the reason 98 percent of small-molecule drugs never make it into the brain in useful amounts.
Why we needed a model
You cannot easily poke a living human BBB to study it, and mouse barriers are not human barriers — species differences in transporters and tight junctions matter enormously for drug predictions. The fix arrived with induced pluripotent stem cells (iPSCs): adult cells reprogrammed back into a stem-like state, then steered into becoming the endothelial cells of the brain’s vessels.
In a dish, these cells build a working facsimile of the barrier — tight junctions, transporters, selective permeability and all. We can grow them on soft or stiff surfaces (stiffness itself changes barrier quality, as we showed in Cellular and Molecular Bioengineering), screen nuclear receptor drugs for transport effects, and ask what happens when inflammation arrives.
What the dish reveals
Our Nature Communications work put reactive astrocytes on the other side of that model barrier and showed how they pass inflammation through the TNF–STAT3 axis. More recently, we found that the protein IQGAP2 helps regulate immune dynamics at the barrier — one of those discoveries that only becomes visible when you can hold the system still and watch.
The model has limits — it lacks blood flow, and single-cell accuracy is still improving — but as a bridge between petri dish and patient, it has already changed how brain-drug questions get asked.
What’s next
Combine the barrier model with organoids (self-organizing clusters of brain cells) and you get the field’s current frontier: vascularized human brain tissue in a dish, where we can watch immune signals, drug transport, and — our contribution — RNA-level interventions play out in human cells.
Want to know how we grow brain cells to feed these models? That story is coming soon — subscribe via RSS.