Xenograft tissue slice tandem co-cultures are a highly specific model to selectively analyze drug inhibitory effects on glioblastoma invasion

Test systems enabling preclinical assessment of drug effects in relevant models are essential for optimizing the selection of candidate therapeutics before their further clinical translation. Xenograft tissue slice tandem co-culture (XTCC) models were developed as ex vivo systems for visualizing glioblastoma (GBM) tumor growth and invasion into the complex host tissue structures of the brain. Work here tested the XTCC model for delineating specific drug effects, in particular inhibition of invasion as major issue in GBM. The established chemotherapeutic Temozolomide (TMZ) and three promising candidates - two histone deacetylase inhibitors, Vorinostat and Entinostat, and the neuropeptide Apamin - were tested. XTCCs were generated by placing G55T2 or U87-MG cell-derived tumor xenograft tissue slices onto murine cortical brain slices. Upon drug treatment, effects on growth, invasion, proliferation, and apoptosis were analyzed by immunohistochemistry. Differences in invasion capacity were seen between the two cell lines. Profound invasion-inhibitory effects of 100 μM TMZ were accurately monitored and substantially higher than inhibition of the bulk tumor mass. Likewise, the extent of single-cell invasion into the normal brain tissue was massively inhibited by Vorinostat and especially by Entinostat, indicating that histone deacetylase inhibitor treatment is particularly efficient in inhibiting GBM cell invasion. Despite the absence of inhibitory effects of Apamin in 2D cell culture, G55T2 XTCCs revealed ∼70% reduced GBM invasion, associated with a substantial inhibition of proliferation as indicated by loss of Ki-67 positivity. Taken together, we show the suitability of the XTCC models for monitoring tumor growth and, in particular anti-invasive effects of drugs.