A Ca2+-Mediated Switch of Epiplakin from a Diffuse to Keratin-Bound State Affects Keratin Dynamics
Ratajczyk S, Drexler C, Windoffer R, Leube RE and Fuchs P
Keratins exert important structural but also cytoprotective functions. They have to be adaptable to support cellular homeostasis. Epiplakin (EPPK1) has been shown to decorate keratin filaments in epithelial cells and to play a protective role under stress, but the mechanism is still unclear. Using live-cell imaging of epithelial cells expressing fluorescently tagged EPPK1 and keratin, we report here an unexpected dynamic behavior of EPPK1 upon stress. EPPK1 was diffusely distributed throughout the cytoplasm and not associated with keratin filaments in living cells under standard culture conditions. However, ER-, oxidative and UV-stress, as well as cell fixation, induced a rapid association of EPPK1 with keratin filaments. This re-localization of EPPK1 was reversible and dependent on the elevation of cytoplasmic Ca2+ levels. Moreover, keratin filament association of EPPK1 led to significantly reduced keratin dynamics. Thus, we propose that EPPK1 stabilizes the keratin network in stress conditions, which involve increased cytoplasmic Ca2+
mScarlet - short EPPK1 re-localizes to keratin filaments after treatment with Tg. A shEpi/AK cell expressing mScarlet-tagged short EPPK1 (red) and EGFP-tagged HK13 (green) was imaged using time-lapse fluorescence microscopy before and after treatment with 200 nM Tg with an interval of 30 s. Time point of drug addition is indicated in the video. The video is a maximum intensity projection of acquired z-stacks. Display rate: 5 frames per second. Scale bar: 10 μm. This video corresponds to images shown in Fig. 2 A.
EPPK1 re-localizes reversibly to keratin filaments after irradiation with a UV- A laser. An shEpi/AK cell expressing mScarlet-tagged short EPPK1 (red) and EGFP- tagged HK13 (green) was partially irradiated with a UV-A pulse during live-cell fluorescence imaging using a 355 nm laser (passively Q-switched pulsed ablation laser 16 mW average power, irradiation area 50x50 pixels, 40% laser intensity, 10 ms irradiation time per pixel) in the area indicated by a dashed square. Images were taken every 90 s, and a z-stack was acquired for every time point. UV-A irradiation was applied once after three images were acquired. The start of irradiation is indicated in the video. Upon UV-A treatment of shEpi/AK cells, mScarlet-short EPPK1 (red) re-localized to HK13-EGFP (green) within 1 min and 30 s and returned to a completely diffuse state about 10 min after irradiation. Display rate: 5 frames per second. Scale bar: 10 µm. This video corresponds to images shown in Fig. 2 B.
Live-cell imaging of an shEpi-CaS cell treated with Tg. A shEpi-CaS cell overexpressing mScarlet-short EPPK1 (red) and the Ca2+ sensor GCaMP6m-XC (green) was imaged using time-lapse fluorescence microscopy before and after treatment with 200 nM Tg. The time point of drug addition is indicated in the video. Under standard conditions, EPPK1 shows a diffuse localization pattern and the Ca2+ sensor produces a faint fluorescent signal only, indicating low intracellular Ca2+ levels. 30 s after addition of Tg, short EPPK1 showed a filamentous localization pattern accompanied by an elevation of Ca2+ sensor fluorescence intensity, indicating a Ca2+ dependent EPPK1 translocation. The video is a maximum intensity projection of z-stacks acquired every 15 s. Display rate: 5 frames per second. Scale bar: 10 µm. This video corresponds to images shown in Fig. 3 B.
Live-cell imaging of an shEpi/AK cell before and after cell lysis shows re-localization to keratin filaments. An shEpi/AK cell expressing mScarlet-tagged short EPPK1 (red) and EGFP-tagged HK13 (green) was imaged using time-lapse fluorescence microscopy before and after lysis with a buffer containing 50 mM Tris HCl pH 7.4; 0.1% Triton X-100. The time of lysis buffer addition is indicated in the video. Before the addition of the lysis buffer, EPPK1 showed a diffuse localization in the cell, which changed to a keratin filament-associated pattern after addition of the lysis buffer. Non-keratin associated EPPK1 diffused out of the cell, hence the clear loss of EPPK1 fluorescence signals after cell lysis. The video is a maximum intensity projection of acquired z-stacks. Imaging interval: 1 min. Display rate: 5 frames per second. Scale bar: 10 µm. This video corresponds to images shown in Fig. S9 A
Live-cell imaging of an shEpi/AK cell before and after cell lysis reveals Ca2+ -dependent re-localization to keratin filaments. A shEpi/AK cell expressing mScarlet- tagged short EPPK1 (red) and EGFP-tagged HK13 (green) was imaged using time-lapse fluorescence microscopy before and after lysis with a buffer containing 50 mM TrisHCl pH 7.4, 0.1% Triton X-100 and 5 mM EDTA. Time of lysis buffer addition is indicated in the video. Before cell lysis, cells showed a non-keratin associated EPPK1 localization. After the addition of lysis buffer, no EPPK1 signal could be detected anymore, showing that the presence of EDTA inhibited Ca2+ -induced EPPK1 association with keratin filaments. The video is a maximum intensity projection of acquired z-stacks. Imaging interval: 1 min. Display rate: 5 frames per second. Scale bar: 10 µm. This video corresponds to images shown in Fig. S9 B..
EPPK1 association reduces the mean flow of keratins. An EpiKI/AK cell expressing mScarlet-tagged EPPK1 (red) and EGFP-tagged HK13 was imaged using time- lapse fluorescence microscopy before and after treatment with 50 nM Tg with an interval of 90 s. Only a representative peripheral part of the cell is shown. Tg was added after 21 min, which is indicated in the video. HK13-EGFP (inverted gray scale) flow is reduced in EpiKI/ AK cells after Tg-induced binding of EPPK1-mScarlet (red) to keratins. The video shows a maximum intensity projection of acquired z-stacks. Display rate: 5 frames per second. Scale bar: 10 µm. This video shows a representative cell used for the analysis presented in Fig. 4 A
Keratin flow is unaltered in Tg-treated EPPK‒/‒ cells. An EpiKO/AK 3F2 cell expressing HK13-EGFP was imaged using time-lapse fluorescence microscopy before and after treatment with 50 nM Tg with an interval of 90 s. Only a representative peripheral part of the cell is shown. Tg was added after 21 min, which is indicated in the video. HK13-EGFP (inverted gray scale) flow is not reduced in EpiKO/AK 3F2 cells after addition of Tg. The video shows a maximum intensity projection of acquired z-stacks. Display rate: 5 frames per second. Scale bar: 10 µm. This video shows a representative cell used for the analysis presented in Fig. 4 B.
Reintroduction of short EPPK1 into EPPK‒/‒ cells rescues the phenotype by reducing the mean flow of keratin. A shEpi-KO/AK 3F2 cell expressing mScarlet- tagged short EPPK1 (red) and EGFP-tagged HK13 was imaged using time-lapse fluorescence microscopy before and after treatment with 50 nM Tg with an interval of
90 s. Only a representative peripheral part of the cell is shown. Tg was added after 21 min, which is indicated in the video. HK13-EGFP (inverted gray scale) flow is reduced in shEpi- KO/AK 3F2 cells after Tg-induced binding of mScarlet-short EPPK1 to keratins. The video shows a maximum intensity projection of acquired z-stacks. Display rate: 5 frames per second. Scale bar: 10 µm. This video shows a representative cell used for the analysis
presented in Fig. 4 C.