3.4 Normalized cell shape
To investigate the role of geometric features on the KIF network and to identify the appropriate size of patterned features, AK13-1 cells and HaCaT B10 cells spread on micropatterns with an average cell spreading area of 700 µm², 900 µm², 1100 µm² and 1600 µm² were analyzed. After cell attachment and spreading an increased number of malformed micropatterns could be observed (Fig. 10). These were mostly micropatterns occupied by one or multiple cells, suggesting that the printed proteins were peeled off by the cells, resulting in pattern degradation. A retreat of the micropattern boundaries is observed, which correlates with the cell boundaries (Fig. 10 A‘ and B‘). As the overall shape of the patterned cells was unchanged, these cases were also documented and the effects of pattern degradation on the distribution of KIFs assessed as negligible or insignificant.
Fully spread single cells were found often on small micropatterns (Fig. 11 A-A“‘) whereas an increase in micropattern size led to an increase in the number of cells per micropattern (Fig. 11 B-B“‘) or to incomplete cell spreading of single cells (Fig. 11 C-C“‘). Hence, less or no fully spread single cells could be found for many large micropatterns.First results showed that the human keratin 13 network of AK13-1 cells spread on micropatterns with an average cell spreading area of 700 µm² (Fig. 12 A-A‘) appeared extremely dense compared to fully spread cells on larger micropatterns (Fig. B-B‘) and on non-printed substrates (Fig. C-C‘). Same could be observed for the keratin 5 network of HaCaT B10 cells (no comparative data shown), which suggests the assumption that KIFs of cells spread on micropatterns seem in general compact compared to cells with unlimited cell spreading area on non-printed substrates.