*3.3. The Role of Suspension in EFM Sca*ff*olds*

To further elucidate how mechanical nuances can influence cell behavior, we investigated EFMs lacking support in a central region, i.e., EFMs suspended across an annulus (Figure 1B). This architecture resembles tissue engineered constructs in which a support material may be missing or only partially present [12,13], such as artificial trachea constructs. PS disks were used to suspend EFMs of different thicknesses (i.e., 50, 100 and 200 μm) over two inner gap diameters (i.e., 3 and 10 mm). Compared to 50 μm PS-supported EFMS, cells cultured on 50 μm EFMs supported across a 10 mm annular gap displayed Feret diameters similar in magnitude, but cells on the 3 mm gap scaffold had a significantly higher mean Feret diameter than either support (Figure 5). According to FEM, the 3 mm gap scaffold would have an indentation modulus between that of the 10 mm gap scaffold and the PS support.

**Figure 5.** Glioblastoma cell Feret diameter on supported (PS) and suspended (10 mm, 3 mm) 50 μm EFMs. Scaffolds are arranged left to right by increasing indentation modulus. A total of 935 cells were analyzed across at least two independent experiments. Levels connected by a star (\*) are statistically significant from each other (*p* < 0.05).

To gain additional information on these responses, we investigated the effect of EFM thickness coupled to annular diameter (Figure 6). As EFM thickness increased from 50 to 200 μm, Feret diameter decreased by ~37% for a fixed 3 mm gap diameter. At the larger 10 mm gap, this decrease was much less pronounced (~18%). In addition, there was a statistically significant difference in Feret diameter as a function of gap diameter for the 50 μm thick and 100 μm thick EFMs. This difference was abrogated at the 200 μm EFM thickness. Thus, increasing gap diameter correlates with a decreased indentation modulus and an observed decrease in Feret diameter. Interestingly, increased EFM thickness correlates with increased indentation moduli, but in this case caused a decreased Feret diameter. This highlights the mechanical complexity of EFM scaffolds, especially in the more nuanced scenario of a suspended EFM. These results are likely influenced by the role of curvature, which could be pronounced at the length scale of a cell. Alternatively, these responses may reflect internal residual tension that is introduced by deformation of the EFM in the central region.

**Figure 6.** Glioblastoma cell Feret diameter measured on supports with 3 and 10 mm diameter annular gaps supporting EFMs of different thicknesses. A total of 1097 cells were analyzed across at least two independent experiments. Levels connected by a star (\*) are statistically significant from each other (*p* < 0.05).
