(a) Sketch of AFM set-up used for indentation measurements including a bead (4.5 μm diameter) glued at the end of a tipless cantilever. The cantilever is moved up and down by a piezo system. The bending of the cantilever is measured by a laser beam detected by a photodetector. And the AFM cantilever bending is linearly proportional to the force applied on the cell. (b) Differential interference-contrast optical image of the bead cantilever pushed onto a cultured DRG neuron (scale bar, 20 μm). (c) Examples of force–indentation curves for WT and STOML3−/− sensory neurons as fitted for double contributions: in the curve two regimes with different elastic properties can be identified (black marker, first point for first elastic component E1, and pink marker second point for second elastic component E2) (Methods, Supplementary Figs 2 and 3). (d) Quantitative comparison of elasticity modulus E1 and E2 values obtained by the fitting procedure (WT: n=110 neurons from 4 cultures, 4 animals; STOML3−/−: n=120 neurons from 4 cultures, 4 animals; two-way ANOVA with post hoc Bonferroni’s test, P<0.05). See also Supplementary Fig. 4. (e) Representative force–distance curves (approach in grey, retraction in blue) acquired at 5 μm s−1. On the retraction curve, after first detachment force several step-like structures are clearly discernible (indicated by arrows), corresponding to the sequential detachment of membrane tethers from the cantilever. (f) Tether force as a function of pulling velocity in WT and STOML3−/− sensory neurons (WT: n=20 neurons from 2 cultures, 2 animals; STOML3−/−: n=19 neurons from 2 cultures, 2 animals). Solid lines are the linear fit of force–velocity curves used to estimate the static tether force at zero F0 and the effective viscosity ηeff (Methods, Supplementary Table 2) (R-value is 0.6 for WT and 0.97 for STOML3−/−). ****P<0.0001; Error bars indicate s.e.m.