No edge effect with NG10815 cells

I’ve repeatedly seen cytotoxic edge effects when I grow bEnd3 cells on the bottom of PET membrane transwells.  Here’s a post of images.  I’m getting closer to co-cultures of bEnd3 and NG108-15 cells on opposite sides of membranes and I’ve been investigating the optimal arrangement of the different cell types (bend3 on top or bottom?).  Here is an image of NG10815 on the bottom of PET transwells after 7 days of growth (passage 10, seeded at 50000 cells/cm²).  I pseudocolored the phase contrast image in blue.  The active membrane area occupies most of this image but the interface between the membrane and the sidewall is the arching black area toward the top of the image.

Unlike bEnd3 cells, NG10815 cells grew right across the membrane-sidewall interface and there were very few dead cells at the edge.  If you look carefully, the cells are slightly out of focus off the membrane because they are in a different vertical plane from the membrane.  Clumping doesn’t seem to be as much of an issue with this orientation, probably because the multilayers simply fall away from the cells attached to the membrane.

I’ve settled on growing endothelial cells on the top side and glial cells on the bottom side of transwell membranes.