Neural and liver tissues
Our first perfused tissue models.
While data from our collaborations remain confidential, results from our initial validation experiments demonstrate the platform’s potential. We compared our perfused solid tissue constructs (~3×3×3mm) with standard organoids, both cultured under identical conditions.
Excellent viability throughout the entire volume
Livegrid maintained viability throughout dense multi-mm3 neural constructs over two months in culture. The entire tissue volume was free of cleaved caspase-3, a marker of apoptosis, while standard organoids developed massive apoptosis (green), followed by complete cellular disintegration in their inner regions.
Accelerated differentiation
Our perfused tissues rapidly transitioned from pluripotency toward a neural identity, demonstrating unusually fast differentiation compared with standard organoids. After only two days of neural induction, PAX6, an early neuronal biomarker, was detected in 21.5% of cells in perfused tissues and was absent in organoids and non-perfused tissues. At the same time, the pluripotency biomarker NANOG demonstrated the opposite pattern.
Extensive morphological features
When cultured for longer periods, our perfused tissues exhibited more mature phenotypes. In particular, after two months, our perfused cortical tissues developed dense, intertwined networks of radial glial (BLBP, red) and neuronal (DCX, green) processes, interspersed with mature (NeuN-positive, white) neurons. The tissues also contained extended processes from both glutamatergic (not shown) and GABAergic (GAD67, white) populations and uniquely expressed MBP (yellow), a mature oligodendrocyte biomarker. None of these features was observed in identically cultured cortical organoids.
Long-term functionality
After two months in culture, our perfused liver tissues remained metabolically active, producing albumin and urea and expressing drug-metabolizing cytochrome P450 enzymes (CYP3A4, green) and other hepatic biomarkers. The tissues demonstrated excellent accuracy in predicting the human clearance of the very slowly metabolized drug theophylline, yielding a value of 0.8 mL/min/kg that closely matched the reported in vivo value of 0.77 mL/min/kg. After a three-day washout, a second application of the drug to the same tissues kept the prediction error within a twofold margin.
