Figure S1. Fibronectin Coating Density on PAA Gels. Fluorescent Intensity of Polyacrylamide

Figure S1. Fibronectin coating density on PAA gels. Fluorescent intensity of polyacrylamide gels immunostained for fibronectin was compared under constant imaging conditions. No statistical difference in intensity was seen, n=15 per condition, and error bars represent SEM.

Figure S2. The cell cycle is not significantly altered on substrates of different rigidities. Barsindicate the results of a FACS analysis. The majority of the cells were in G1 phase after 24 hours growth, and this number did not change noticeably on more rigid substrates. Error bars indicate SEM, n = 3.

Figure S3. Cell spreading and stress fiber presence depend on substrate rigidity and gel thickness. Images are of basal planes of 3T3 fibroblasts stained for actin. Values on top indicate Young’s modulus of PAA gels. Values down the left side indicate thickness of each gel. Cells spread more and assemble more stress fibers with increasing stiffness on gels of 17 or 22 μmthickness, while all cells spread well and assemble stress fibers on 6 and 12 μm thick gels. Scale bar = 20μm.

Figure S4. Localization of the KASH domain at the nuclear envelope. Image shows the nucleusof a single NIH 3T3 fibroblast expressing GFP-KASH4. The KASH4 can be seen to clearly localize at the nuclear membrane. Scale bar is 10μm.

Figure S5: Minor and major axes of x-y nuclear cross-section do not change significantly with substrate rigidity, KASH4 overexpression, or blebbistatin treatment. (A) Minor axis,(B) Major axis of x-y cross-section of nucleus.(C) Nuclear aspect ratio (minor axis/major axis) remains roughly constant. Error bars represent SEM, n=30 cells. * indicates p<0.05 between two values, ** indicates p<0.05 from other values of same stiffness.

Figure S6. Basal stress fiber presence in 3T3 fibroblasts on glass.(A) Average number of basal stress fibers found beneath the nucleus increases with stiffness, though KASH4 overexpression eliminates this trend. * indicates p<0.05 from other control values. ** indicates p<0.05 from 0.4 kPa and 39 kPa, and from KASH4. *** indicates p<0.05 from KASH4 value. n=20. (B) Basal stress fibers counted beneath the nucleus in cells on glass. Cells treated with latrunculin-B for 30 mins showed only slightly fewer fibers, though this was not statistically different. (C) Representative basal actin image showing a count of zero basal stress fibers. Nucleus outlined in dashed oval, scale bar = 10μm. (D) Representative basal actin image showing a count of seven basal stress fibers. Nucleus outlined in dashed oval, scale bar = 10μm.

Figure S7. Actin cap presence in 3T3 fibroblasts on glass. (A)Probability of a clearly present (intact) actin cap versus an unclear or nonpresent (disrupted) actin cap, imaged with a laser scanning confocal microscope. (B)Probability of basal stress fiber presence in control and latrunculin-B treated fibroblasts.

Figure S8. Actin cap presence with DMSO control. Presence of a distinct actin cap versus adisrupted actin cap is shown forNIH 3T3 fibroblasts cultured on glass for 24 hours. Control cells were untreated and DMSO cellswere given 30 minute low dosage DMSO treatment prior to fixation. The amount of DMSO was the same used in the latrunculin-B experiments. Error bars indicate SEM, n=20.

Figure S9. Cell motility is significantly decreased on all substrates on overexpression of GFP-KASH4. The mean square displacement (MSD) (calculated over non-overlapping time intervals) is shown for control and KASH4 over-expressing cells. Error bars indicate SEM, n=8 cells per condition.