Mechanical loading of desmosomes depends on the magnitude and orientation of external stress

Fig. 1. Desmoplakin (DP) tension sensors localize to desmosomes (DSMs). a DSM adhesion is mediated by the desmosomal cadherins desmocollin (Dsc) and desmoglein (Dsg), which engage the adapter proteins plakophilin (Pkp) and plakoglobin (Pg); DP forms the connection to intermediate filaments (IFs). The tension sensor module (TSM) was inserted into DP after the rod domain. b DP tension sensors (DPI-TS, DPII-TS) were generated along with controls lacking the IF-binding C-terminal region (DPI-ctrl, DPII-ctrl). The TSM comprises the F40 linker peptide (GPGGA)8 flanked by mTFP1/mEYFP in DPI, and YPet/mCherry in DPII. Tension reduces FRET in DP-TS but not in DP-ctrl constructs. c Live-cell imaging reveals normal subcellular localization of DPI-TS (expressed in MDCK cells) and DPII-TS (expressed in MEK-wt one day after DSM induction by Ca2+). Scale bar: 20 μm; in zoom: 4 μm. d DPII constructs localize to intercellular junctions in MEK-KO one day after DSM induction, similar to endogenous DP in MEK-wt. Immunostainings show that DPII constructs co-localize with Dsg1/2, and that full-length constructs but not DPII-ctrl mediate the recruitment of a coherent keratin 5 (K5) network. Images are summed…

Fig. 2. Desmoplakin tension is negligible under homeostatic conditions. a Donor intensity signals were masked and thresholded to generate a segmentation map of individual DSM puncta. For each punctum, a fluorescence lifetime was determined and the corresponding FRET efficiency calculated. FRET efficiencies for DPI-TS (yellow) and DPI-ctrl (blue) were indistinguishable in confluent MDCK monolayers. The median FRET efficiency per image is shown as a boxplot and reflects the underlying distributions of individual puncta values that were used to calculate the mean change in FRET efficiency as ΔE = Ectrl − ETS. b No FRET efficiency differences were observed at the edge of migrating MDCK cell sheets in cell–cell junctions categorized as having one or both cells on the colony edge. Similarly, treatment of confluent MDCK monolayers with the actin-destabilizing drug cytochalasin-D (see Supplementary Fig. 2c for immuno-staining) did not induce a significant change in FRET efficiency. c DPII constructs were expressed in MEK-wt monolayers and imaged at distinct time points after DSM induction (3–5 h, 1 day, and 3 days). MEK-KO were imaged one day after DSM induction. d In all conditions, the mean…

Fig. 3. DP is mechanically loaded on soft hydrogels and in response to external force. a Hydrogels of defined stiffness were located in the center of a glass-bottom microscopy dish. b No FRET efficiency differences between DPII-TS (yellow) and DPII-ctrl (blue) were observed for 25, 12, and 4 kPa substrates, but DPII-TS showed a significant decrease in FRET efficiency for 2 kPa substrates. c The FRET efficiency decrease for DPII-TS on 2 kPa substrates is specific to cells cultured on the hydrogel. FRET efficiencies in cells on the uncoated glass rim surrounding the hydrogel were indistinguishable for DPII-TS and DPII-ctrl. d Top row: Example FRET index measurements for individual puncta of DPI-TS in an MDCK monolayer before (Pre), during (Pull), and after (Post) the application of external mechanical stress using a glass micropipette (the red arrow indicates pulling direction). Bottom row: FRET indices of DPI-TS (yellow) and DPI-ctrl (blue) were indistinguishable before pulling, but decreased specifically for DPI-TS during force application; the DPI-TS decrease partially recovered after relaxation. e Top row: Example FRET index measurements for DPII-TS expressed in a monolayer of MEK-wt…

Fig. 4. DP loading is sensitive to the magnitude of the applied deformation. a Before pulling (Pre), the FRET index difference (green) between DPI-TS (yellow) and DPI-ctrl (blue) in MDCK was close to zero, confirming that no tension was transmitted across DPI without external stress. During pulling (Pull), tension increased with deformation. After pipette tip withdrawal (Post), tension partially relaxed. Evidence of residual tension was observed at higher recoil distances. Data are from the pulling experiment represented in Fig. 3d. b Mechanical tension on DPII depends on the degree of deformation, and relaxes after external stress is released. Data are from the pulling experiment represented in Fig. 3e. c FRET index change for DPI-TS in MEKs was small in magnitude, even at high recoil distances. P > 1450 total puncta from n ≥ 36 images measured in six independent experiments. d Introduction of the S2849G mutation in DPI-TS does not alter sensitivity to recoil distance in MEKs. P > 1250 total puncta from n ≥ 35 images measured in six independent experiments. DPI-ctrl data were the same as in c. a–d Puncta with recoil distances dr ≤ 10 μm were analyzed. The mean FRET index of DP-TS and…

Fig. 5. DPII loading is sensitive to the orientation and magnitude of the applied deformation. a Tension across DPI in MDCK cells did not depend on the recoil angle. Data are from the pulling experiment represented in Fig. 3d. b In MEK-wt, DPII tension increased for increasing recoil angles. Data are from the pulling experiment represented in Fig. 3e. c Puncta were classified as parallel (αr ≤ 45°) or perpendicular (αr > 45°) to pulling direction. d The effect of magnitude and orientation of the external deformation showed no apparent additive effect for DPI-TS in MDCK monolayers. e The effect of magnitude and orientation of the external deformation were additive for DPII-TS in MEK-wt. a–e The mean FRET index of DP-TS and DP-ctrl as well as the mean difference between DP-TS and the mean of all DP-ctrl puncta were calculated using the lmer-test and displayed with the 68% CI. Blue lines and corresponding shading indicate the mean with 95% CI of all DP-ctrl puncta. Yellow and green lines are guides to the eye. Source data are provided as a Source Data file