Spatiotemporal dynamics of PIEZO1 localization controls keratinocyte migration during wound healing

  1. Jesse R Holt
  2. Wei-Zheng Zeng
  3. Elizabeth L Evans
  4. Seung-Hyun Woo
  5. Shang Ma
  6. Hamid Abuwarda
  7. Meaghan Loud
  8. Ardem Patapoutian  Is a corresponding author
  9. Medha M Pathak  Is a corresponding author
  1. Departmentof Physiology & Biophysics, UC Irvine, United States
  2. Sue and Bill Gross Stem Cell Research Center, UC Irvine, United States
  3. Center for Complex Biological Systems, UC Irvine, United States
  4. Howard Hughes Medical Institute, Department of Neuroscience, The Scripps Research Institute, United States
  5. Department of Biomedical Engineering, UC Irvine, United States
5 figures, 1 table and 1 additional file

Figures

Figure 1 with 5 supplements
PIEZO1 is expressed in keratinocytes, produces Ca2+ flickers, and regulates skin wound healing.

(A) Representative images of LacZ stained Piezo1+/+ (left) and Piezo+/βGeo (right) skin sections from P2 (postnatal day 2) mice. Scale bar = 20 µm. epi, epidermis. hf, hair follicle. (B) qRT-PCR from primary neonatal keratinocytes of Piezo1 mRNA expression in Krt14+/+ Piezo1fl/fl (Con) and Krt14Cre/+;Piezo1fl/fl (conditional knockout [cKO]) mice. Data presented as the mean ± SEM. See also Figure 1—figure supplements 12. Data collected from two litters. (C) Representative examples of Ca2+ flickers recorded from ControlcKO (top) and Piezo1-cKO (bottom) keratinocytes. Traces show fluorescence ratio changes (ΔF/F0) from a Ca2+ flicker site, plotted against time. (D) Representative images of sites of Ca2+flickers (red dots) overlaid on images of keratinocytes isolated from ControlcKO (top) and littermate Piezo1-cKO (bottom) mice that are loaded with fluorescent Ca2+ indicator Cal-520AM. Gray line denotes cell boundary, Scale bar = 20 µm. (E) Cumming plot showing frequency of Ca2+ flickers in Piezo1-cKO and respective ControlcKO cells (p value calculated via Mann-Whitney test; Cohen’s d = −0.6175). n in E denotes the number of videos (i.e., unique fields of view imaged, each of which is composed of one or more cells). Videos were collected from four independent experiments from three litters. See also Figure 1—video 1. (F, G, H) Similar to C, D, E but using keratinocytes from ControlGoF (top) and Piezo1-GoF (bottom) mice (p value calculated via Mann-Whitney test; Cohen’s d = 0.6747). Videos were collected from four independent experiments from three litters. See also Figure 1—video 2. (I) Diagram of in vivo wound healing model (top) and representative wound images at days 1 and 6 (bottom). (J) Cumming plot showing wound area of Piezo1-cKO (left) and Piezo1-GoF (right) groups at day 6 relative to control (Con) littermates (p value calculated via two-sample t-test; Cohen’s d = −0.799 and 0.844, respectively). Control (Con) refers to the Cre-negative littermates in each group. n in J denotes number of wounds measured, with two wounds per animal. (K) Representative images of an in vitro scratch assay. White line represents the monolayer edge. Scale bar = 200 µm. (L) Cumming plot showing quantification of scratch wound closure in monolayers of keratinocytes isolated from: ControlcKO vs. Piezo1-cKO mice (left; p value calculated via two-sample t-test; Cohen’s d = 1.188; images from three independent experiments), ControlGoF vs. Piezo1-GoF mice (middle; p value calculated via two-sample t-test; Cohen’s d = −1.128; images from four independent experiments) or DMSO-treated vs 4 µM Yoda1-treated ControlcKO monolayers (right; p value calculated via Mann-Whitney test; Cohen’s d = −2.278; images from three independent experiments). n in L denotes the number of unique fields of view imaged. Data are normalized to the mean scratch closure of the corresponding control condition where one is the average closure distance of the control and 0 is no closure. See also Figure 1—figure supplement 3. (M) Schematic illustrating results from in vivo wound healing assay shown in I and J. Mice were wounded (left) and after 6 days, wounds of Piezo1-cKO mice healed more than Control, whereas wounds from Piezo1-GoF mice healed less. Vertical bars in upper Cumming plots denote mean  ±  s.d.

Figure 1—source data 1

PIEZO1 is expressed in keratinocytes, produces Ca2+ flickers, and regulates skin wound healing.

(Sheet 1) qRT-PCR data from primary neonatal keratinocytes of Piezo1 mRNA expression in Krt14+/+ Piezo1fl/fl (Con) and Krt14Cre/+;Piezo1fl/fl (conditional knockout [cKO]) mice seen in Figure 1B. (Sheet 2) Flicker fluorescence ratio changes (ΔF/F0) from a region of interest, plotted against time shown in Figure 1C from ControlcKO (left) and Piezo1-cKO (right) keratinocytes. (Sheet 3) Data plotted in Figure 1E showing frequency of Ca2+ flickers in Piezo1-cKO and respective ControlcKO cells. (Sheets 4, 5) Similar to Sheets 2, 3 but using keratinocytes from ControlGoF and Piezo1-GoF mice for data shown in Figure 1F,H. (Sheet 6) Data used for creating Cumming plot in Figure 1J showing wound area of Piezo1-cKO and Piezo1-GoF groups at day 6 relative to control (Con) littermates. (Sheet 7) Normalized plot data seen in Figure 1L showing scratch wound closure in monolayers of keratinocytes isolated from: ControlcKO vs. Piezo1-cKO mice, ControlGoF vs. Piezo1-GoF mice and DMSO-treated vs. 4 µM Yoda1-treated ControlcKO monolayers.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig1-data1-v2.xlsx
Figure 1—figure supplement 1
Piezo1-cKO mice develop normally.

(A) Images of P2 (postnatal day 2) pups from Krt14Cre;Piezo1fl/fl (conditional knockout [cKO]) and Cre- (Con) littermates illustrating normal development of neonatal pups. (B) Staining of keratin markers (green), K10 (top) and K14 (bottom), on skin sections derived from adult Krt14Cre;Piezo1fl/fl (cKO) and Cre- (Con) littermates, highlighting normal patterns of expression in control (left) and cKO (right). epi, epidermis. hf, hair follicle. Scale bar = 20 μm.

Figure 1—figure supplement 2
Piezo1 is the primary PIEZO channel expressed in mouse keratinocytes.

qRT-PCR from primary neonatal keratinocytes of Piezo1 and Piezo2 mRNA expression relative to Gapdh.

Figure 1—figure supplement 2—source data 1

Piezo1 is the primary PIEZO channel expressed in mouse keratinocytes.

qRT-PCR data shown in Figure 1—figure supplement 2 from primary neonatal keratinocytes of Piezo1 and Piezo2 mRNA expression relative to Gapdh.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig1-figsupp2-data1-v2.xlsx
Figure 1—figure supplement 3
Yoda1 inhibits scratch wound closure in control monolayers but has no effect on closure in Piezo1-cKO monolayers.

(A) Cumming plot showing quantification of scratch wound closure (as seen in Figure 1K) in monolayers of keratinocytes isolated from ControlcKO mice treated with DMSO, or 100 nM, 2 µM, or 4 µM Yoda1 (p values determined by Mann-Whitney test; Cohen’s d = 0.249,–1.586, –2.278 respectively; images from two independent experiments for 100 nM data and three independent experiments for DMSO, 2 and 4 µM Yoda1 data). (B) As for A but for keratinocytes isolated from Piezo1-cKO mice treated with DMSO or 4 µM Yoda1 (p values determined by Mann Whitney test; Cohen’s d = −0.151; images from two independent experiments). Data are normalized to the mean scratch closure of the DMSO-treated condition for each genotype where 1 is the average closure distance of the respective control condition and 0 is no closure. Bars in upper Cumming plots denote mean  ±  s.d.

Figure 1—figure supplement 3—source data 1

Yoda1 inhibits scratch wound closure in ControlcKO monolayers but has no effect on closure in Piezo1-cKO monolayers.

(Sheets 1,2) Normalized scratch closure data used to create (Sheet 1) control and (Sheet 2) conditional knockout (cKO) Cumming plots shown in Figure 1—figure supplement 3.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig1-figsupp3-data1-v2.xlsx
Figure 1—video 1
PIEZO1 Ca2+ flickers are reduced in Piezo1knockout keratinocytes.

A representative F/F0 ratio movie of Ca2+ flickers in ControlcKO (left) and Piezo1-cKO (right) keratinocytes. Related to Figure 1C–E.

Figure 1—video 2
PIEZO1 Ca2+ flickers are increased in Piezo1 gain-of-function keratinocytes.

A representative F/F0 ratio movie of Ca2+ flickers in ControlGoF (left) and Piezo1-GoF (right) keratinocytes. Related to Figure 1F–H.

Figure 2 with 5 supplements
PIEZO1 mediates speed and direction during single cell keratinocyte migration.

(A) Representative differential interference contrast (DIC) images from time-lapse series of individual migrating keratinocytes isolated from ControlcKO (top) and respective Piezo1-cKO mice (bottom). Thin white lines denote the cell boundary. Scale bar = 25 µm. (B) Cell trajectories derived from tracking single keratinocytes during time-lapse experiments. Trajectories are shown with cell position at time point 0 normalized to the origin. See also Figure 2—figure supplement 1. (C) Mean squared displacement (MSD) analysis of ControlcKO and Piezo1-cKO keratinocytes tracked in B. Average MSD is plotted as a function of time. Error bars (SEM) are smaller than symbols at some points. (D) Average direction autocorrelation measurement of Piezo1-cKO and ControlcKO keratinocytes plotted as a function of time interval. * denotes a statistically significant difference, and ns denotes ‘not statistically significant’. From left to right: p = 2.0307 × 10–4, 5.75675 × 10–14, 3.18447 × 10–15, 5.34662 × 10–10, 1.72352 × 10–4, 1.34648 × 10–5, 0.01951, 0.13381, 0.61758 as determined by Kruskal-Wallis test. Plotted error bars (SEM) are smaller than symbols. (E) Quantitation of the average instantaneous speed from individual Piezo1-cKO keratinocytes (left) and Piezo1-GoF keratinocytes (right) relative to the respective control cells are shown in a Cumming plot (Cohen’s d = 0.6 [Piezo1-cKO]; d = −0.362 [Piezo1-GoF]; p values calculated via Kolmogorov-Smirnov test). n in B–E denotes the number of individually migrating cells tracked. See also Figure 2—figure supplements 23 and Figure 2 videos 1 and 2. Data are from three independent experiments from two litters for conditional knockout (cKO) and six independent experiments from five litters for gain-of-function (GoF). Bars in upper Cumming plots denote mean  ±  s.d.

Figure 2—source data 1

PIEZO1 mediates speed and direction during single cell keratinocyte migration.

(Sheets 1, 2) X,Y coordinates used to plot cell trajectories derived from tracking single (Sheet 1) control and (Sheet 2) conditional knockout (cKO) keratinocytes plotted in Figure 2B. (Sheet 3) Average mean squared displacement (MSD) analysis of ControlcKO and Piezo1-cKO keratinocytes plotted in Figure 2C. (Sheet 4) MSD for individual cells used to create averages found in Sheet 3. (Sheet 5) Average direction autocorrelation measurement of Piezo1-cKO and ControlcKO keratinocytes plotted as a function of time interval in Figure 2D. (Sheet 6) Quantitation of the average instantaneous speed from individual Piezo1-cKO keratinocytes and Piezo1-GoF keratinocytes relative to the respective control cells plotted in Figure 2E.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig2-data1-v2.xlsx
Figure 2—figure supplement 1
Piezo1-cKO keratinocytes migrate further.

Individual trajectories seen in Figure 2B from Piezo1-cKO (right) and ControlcKO (left) keratinocytes.

Figure 2—figure supplement 1—source data 1

Piezo1-cKO keratinocytes migrate further.

(Sheets 1, 2) X,Y coordinates used to create trajectories seen in Figure 2—figure supplement 1 from (Sheet 2) Piezo1-cKO and (Sheet 1) ControlcKO keratinocytes.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig2-figsupp1-data1-v2.xlsx
Figure 2—figure supplement 2
Piezo1-GoF keratinocytes migrate straighter.

Individual trajectories from Piezo1-GoF (bottom) and ControlGoF (top) keratinocytes.

Figure 2—figure supplement 2—source data 1

Piezo1-GoF keratinocytes migrate straighter.

(Sheets 1,2) X,Y coordinates used to create trajectories from (Sheet 2) Piezo1-GoF and (Sheet 1) ControlGoF keratinocytes seen in Figure 2—figure supplement 2.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig2-figsupp2-data1-v2.xlsx
Figure 2—figure supplement 3
Single cell migration of Piezo1-GoF keratinocytes.

(A) Cell trajectories derived from tracking ControlGoF(left) and Piezo1-GoF (right) keratinocytes during time-lapse experiments. Trajectories are shown with cell position at time point 0 normalized to the origin. (B) Mean squared displacement (MSD) analysis of ControlGoF and Piezo1-GoF keratinocytes tracked in A. Average MSD is plotted as a function of time. Error bars (SEM) are smaller than symbols at some points. (C) Average direction autocorrelation for Piezo1-GoF and ControlGoF keratinocytes plotted as a function of time interval. * denotes a statistically significant difference, and ns denotes ‘not statistically significant’. From left to right: p = 7.922194 × 10–4, 0.3263, 0.8208, 0.14186, 0.02523, 1.03604 × 10–4, 1.42496 × 10–8, 2.18161 × 10–8, 1.71116 × 10–14, 3.13349 × 10–13, 3.0169 × 10–16, 2.17203 × 10–15, 3.3468 × 10–12, 2.83094 × 10–12, 1.17488 × 10–10, 3.54255 × 10–10, 1.21566 × 10–8, 7.6758 × 10–8, 8.58726 × 10–5, 3.76797 × 10–5 as determined by Kruskal-Wallis test. Related to Figure 2. Data are from six independent experiments from five litters. Data in B and C are presented as the mean ± SEM. See also Figure 2—video 2.

Figure 2—figure supplement 3—source data 1

Single cell migration of Piezo1-GoF keratinocytes.

(Sheets 1, 2) X,Y coordinates used to create trajectories in Figure 2—figure supplement 3A from tracking (Sheet 1) ControlGoF and (Sheet 2) Piezo1-GoF keratinocytes. (Sheet 3) Mean squared displacement (MSD) analysis of tracked ControlGoF and Piezo1-GoF keratinocytes plotted in Figure 2—figure supplement 3B. (Sheet 4) MSD for individual cells used to create averages found in Sheet 3. (Sheet 5) Average direction autocorrelation for Piezo1-GoF and ControlGoF keratinocytes plotted in Figure 2—figure supplement 3C.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig2-figsupp3-data1-v2.xlsx
Figure 2—video 1
Piezo1 knockout affects keratinocyte motility.

A representative differential interference contrast (DIC) time-lapse movie of single migrating ControlcKO (left) and Piezo1-cKO (right) keratinocytes. Related to Figures 2A–E3D and E.

Figure 2—video 2
Piezo1 gain-of-function (GoF) affects keratinocyte speed.

A representative differential interference contrast (DIC) time-lapse movie of single migrating ControlGoF (left) and Piezo1-GoF (right) keratinocytes. Related to Figures 2E, 3F and G and Figure 2—figure supplements 23.

Figure 3 with 1 supplement
PIEZO1 activity promotes cell polarization.

(A) Time-lapse series of representative differential interference contrast (DIC) (top) and total internal reflection fluorescence (TIRF) (bottom) images illustrating the location of PIEZO1-tdTomato protein during keratinocyte migration. White lines denote the cell boundary. Scale bar = 15 µm. Images representative of four independent experiments from two litters. See also Figure 3—video 1. (B) Representative overlays of cell outlines segmented from ControlcKO, Piezo1-cKO, ControlGoF, and Piezo1-GoF single cell migration time-lapse images and classified into 20 representative shape modes using VAMPIRE (visually aided morpho-phenotyping image recognition), a machine learning algorithm designed for quantification of cellular morphological phenotypes. (C) Dendrogram showing the level of correlation between shape modes identified by VAMPIRE. Cell shape modes are classified into the morphologically distinct categories of weakly polarized (pink), polarized (blue), or hyper-polarized (dark blue). (D) Bar plots showing the distribution of cell shape modes of ControlcKO (gray; n denotes the number of images analyzed from seven cells and two independent experiments) and Piezo1-cKO (purple; n denotes the number of segmented shapes from 12 cells and two independent experiments) cells. (E) Stacked bar graphs indicating the proportion of weakly polarized (W.P.) (pink), polarized (Pol.) (blue), or hyper-polarized (H.P.) (dark blue) shape modes in ControlcKO (left) and Piezo1-cKO (right) cells. (F, G) Similar to D and E respectively but showing the distribution of cell shape modes for Piezo1-GoF (green; n denotes the number of shapes from nine cells and three independent experiments) and respective ControlGoF (gray; n denotes the number of shapes from eight cells and three independent experiments) cells during single cell migration time-lapse experiments. See also Figure 2 videos 1 and 2. Bars in D-G denote frequency.

Figure 3—source data 1

PIEZO1 activity promotes cell polarization.

(Sheet 1) Relative frequency data used to create bar plots showing the distribution of cell shape modes of ControlcKO and Piezo1-cKO cells in Figure 3D. (Sheet 2) Cumulative frequency used to create stacked bar graphs seen in Figure 3E indicating the proportion of weakly polarized, polarized, or hyper-polarized shape modes in ControlcKO and Piezo1-cKO cells. (Sheet 3) Shape mode output from VAMPIRE (visually aided morpho-phenotyping image recognition) used to create shape mode averages in Sheets 1 and 2. (Sheets 4, 5, 6) Similar to Sheets 1, 2, and 3 respectively but using Piezo1-GoF and ControlGoF cells for graphs in Figure 3F,G.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig3-data1-v2.xlsx
Figure 3—video 1
PIEZO1-tdTomato is enriched at the rear of individually migrating keratinocytes.

Representative video showing differential interference contrast (DIC), total internal reflection fluorescence (TIRF) (for PIEZO1-tdTomato), and DIC+ TIRF merged videos from a migrating PIEZO1-tdTomato keratinocyte. Related to Figure 3A.

Figure 4 with 4 supplements
Dynamic PIEZO1 channel localization controls wound edge retraction.

(A) Representative differential interference contrast (DIC) (top) and total internal reflection fluorescence (TIRF) (bottom) image visualizing the location of PIEZO1-tdTomato (P1-tdT) protein in live, collectively migrating keratinocytes in an in vitro scratch assay. Gray lines denote the boundary of the cell monolayer. Gray arrows indicate regions of PIEZO1 enrichment. Scale bar = 20 μm. See also Figure 4—figure supplement 1. (B, C) Representative TIRF images taken from time-lapse image series of healing monolayers from PIEZO1-tdTomato keratinocytes highlighting fields of view in which PIEZO1-tdTomato is not enriched (B) and enriched (C) at the monolayer’s leading edge. Gray lines denote the boundary of the cell monolayer. Arrow indicates regions of enrichment. Scale bar = 20 μm. Blue dashed rectangles in D and E depict the regions used to generate kymographs in D and E. TIRF images were acquired every 10 min over a period of 16.6 hr. (D, E) Left panels: Representative kymographs illustrating PIEZO1-tdTomato puncta dynamics during the time-lapse series shown in B and C, respectively. Middle panel (for E only): Magenta line denotes periods of PIEZO1-tdTomato puncta enrichment at the wound edge, identified and tracked using the deep learning-based kymograph analysis software, Kymobutler. No Kymobutler tracks were detected in non-enriched regions (D). Right panels: Representative kymographs from binarized versions of DIC images corresponding to D and E, respectively, with the Kymobutler track output from the middle panel overlaid. The cell monolayer is represented in black, and white denotes cell-free space of the wounded area. Note the PIEZO1-tdTomato enrichment track correlates with periods of cell retraction. Scale bar = 10 μm. Time bar = 100 min. Black open circles on top represent the time-points of images shown in D and E. See also Figure 4—video 1 and Figure 4—video 2. (F) DIC (bottom) and TIRF (top) images during a time-lapse imaging series following scratch generation at 0 hr from a field of view showing sustained PIEZO1-tdTomato localization and marked monolayer retraction. Gray (top) and white (bottom) lines denote the boundary of the monolayer. Dotted line in the 12 hr image denotes the position of the monolayer at 6 hr; thin arrows indicate direction of monolayer movement during this period. Large gray arrow indicates region of PIEZO1 enrichment. Scale bar = 20 μm. See also Figure 4—video 3. (G) Plot showing 54 individual PIEZO1-tdTomato Kymobutler tracks from 25 kymographs collected from three independent experiments after normalizing the starting spatial and time coordinates of each track to the origin. (H) Schematic of a healing monolayer indicating distributed Piezo1 localization (red dots) following scratch generation (left), the development of areas of PIEZO1 enrichment (middle), and subsequent retraction of those areas (right).

Figure 4—source data 1

Dynamic PIEZO1 channel localization controls wound edge retraction.

(Sheet 1) PIEZO1-tdTomato Kymobutler tracks plotted in Figure 4G.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig4-data1-v2.xlsx
Figure 4—figure supplement 1
Absence of PIEZO1-tdTomato enrichment at wound edge immediately after scratch wound generation.

Representative differential interference contrast (DIC) (left) and total internal reflection fluorescence (TIRF) (right) images illustrating the location of PIEZO1-tdTomato protein immediately after wounding in a keratinocyte monolayer in an in vitro scratch assay. Black (left) and white (right) lines denote the cell boundary. This is the same field of view as in Figure 4A. Note the absence of PIEZO1-tdTomato enrichment at the wound edge as compared to 11 hr later in Figure 4A. Scale bar = 20 μm.

Figure 4—video 1
PIEZO1-tdTomato becomes enriched at regions at the edge of keratinocyte monolayers and elicits localized retraction.

Representative field of view showing time lapse of the differential interference contrast (DIC), total internal reflection fluorescence (TIRF) (PIEZO1-tdTomato), and DIC+ TIRF merged images from a monolayer of PIEZO1-tdTomato keratinocytes corresponding to Figure 4C. Arrows denote period of channel enrichment starting at ~7.5 hr. Note the PIEZO1-tdTomato enrichment during periods of retraction and absence of channel enrichment during periods of protrusion. Related to Figure 4C and E.

Figure 4—video 2
Lack of PIEZO1-tdTomato enrichment in advancing monolayers.

A representative field of view showing a time lapse of differential interference contrast (DIC), total internal reflection fluorescence (TIRF) (PIEZO1-tdTomato), and DIC+ TIRF merged images from a monolayer of PIEZO1-tdTomato keratinocytes corresponding to Figure 4B. Note lack of enrichment at the monolayer edge throughout video, as the wound edge advances. Related to Figure 4B and D.

Figure 4—video 3
Persistent PIEZO1-tdTomato enrichment at the wound edge elicits sustained retraction.

A representative field of view shows time lapse of differential interference contrast (DIC), total internal reflection fluorescence (TIRF) (PIEZO1-tdTomato), and DIC+TIRF merged images of a monolayer of PIEZO1-tdTomato keratinocytes. Note the enrichment of PIEZO1-tdTomato at the wound edge and the ensuing retraction. Related to Figure 4F.

Figure 5 with 11 supplements
PIEZO1 activity promotes cellular retraction and increases edge velocity.

(A) Image still from a differential interference contrast (DIC) time-lapse series of ControlcKO keratinocytes (left). Scale bar = 50 µm. The green line denotes the representative region of interest used for generating kymograph (right). The red dotted line denotes the addition of 4 µM Yoda1. The black arrows indicate retraction events, and the slope of the descending white lines denotes the speed of retraction. Dotted blue line denotes starting position of the cell edge. See also Figure 5—figure supplement 1, Figure 5—video 1 and Figure 5—video 2. Images representative of six independent experiments from cells isolated from three litters. (B) Similar to A but using Piezo1-cKO keratinocytes with the same annotations. Scale bar = 25 µm. See also Figure 5—figure supplement 1, Figure 5—video 3 and Figure 5—video 4. Images representative of seven independent experiments from cells isolated from four litters. (C) Representative overlays of cell boundary outlines detected and segmented from DIC time-lapse series of DMSO-treated ControlcKO, 4 µM Yoda1-treated ControlcKO, DMSO-treated Piezo1-cKO, ControlGoF, and Piezo1-GoF keratinocytes. Brackets separate Piezo1-cKO and Piezo1-GoF background cell types. Color of cell boundary outline indicates passage of time. Scale bar = 25 µm. (D) Violin plots showing the instantaneous protrusion (positive) and retraction (negative) velocities at each position of the segmented cell edges for each frame of time-lapse series videos. Plot shows the combined data from six DMSO and Yoda1 treated ControlcKO cells, respectively, nine Piezo1-cKO cells, five ControlGoF, and six Piezo1-GoF cells from two independent experiments each. p Values calculated using Mann-Whitney test; d values show Cohen’s d which is calculated as test condition minus the respective control condition. (E) Representative heatmaps corresponding to the cells shown in C illustrating cell edge velocities along segmented cell boundaries from time-lapse series. See also Figure 5—figure supplement 2. (F, G, H) Representative DIC kymograph taken at the wound edge and kymograph of the binarized version of the same video, where the cell monolayer is represented in black and cell-free space of the wounded area in white. Kymographs illustrate wound edge dynamics during an in vitro scratch assay performed in (F) control (DMSO)-treated monolayers, (G) control scratches treated with 4 μM Yoda1, and (H) Piezo1-cKO monolayers (right pair). Scale bar = 20 μm (vertical) and 100 min (horizontal). The data in F, G are representative four independent experiments from keratinocytes from three biological repeats for each condition. The data in H are representative of three independent experiments using keratinocytes from three biological repeats. See also Figure 5—figure supplements 34 and Figure 5—videos 57.

Figure 5—source code 1

Code used to analyze data shown in Figure 5.

See also Code availability.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig5-code1-v2.zip
Figure 5—source data 1

PIEZO1 activity promotes cellular retraction and increases edge velocity.

(Sheet 1) X,Y coordinates to create representative overlays of cell boundary outlines in Figure 5C detected and segmented from differential interference contrast (DIC) time-lapse series of DMSO-treated ControlcKO, 4 µM Yoda1-treated ControlcKO, DMSO-treated Piezo1-cKO, ControlGoF, and Piezo1-GoF keratinocytes. This dataset is available on Dryad.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig5-data1-v2.xlsx
Figure 5—source data 2

PIEZO1 activity promotes cellular retraction and increases edge velocity.

(Sheet 1) Data used to create violin plots in Figure 5D showing the instantaneous protrusion (positive) and retraction (negative) velocities at each position of the segmented cell edges for each frame of time-lapse series videos. This dataset is available on Dryad.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig5-data2-v2.xlsx
Figure 5—source data 3

PIEZO1 activity promotes cellular retraction and increases edge velocity.

(Sheets 1–5) Data used to create representative heatmaps corresponding to the cells shown in Figure 5E illustrating cell edge velocities along segmented cell boundaries from time-lapse series.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig5-data3-v2.xlsx
Figure 5—figure supplement 1
Piezo1 gain-of-function (GoF) keratinocytes have increased retraction.

(A) Image still from a differential interference contrast (DIC) time-lapse series of ControlGoF keratinocyte (left), the green line denotes a representative region of interest used for generating kymograph (right). Scale bar = 25 µm. The black arrows indicate retraction events, and the slope of the descending white lines denotes the speed of retraction. Dotted blue line denotes starting position of the cell edge. (B) Similar to Figure 5—figure supplement 1A but using Piezo1-GoF keratinocytes. Kymographs in Figure 5—figure supplement 1A,B representative of five independent experiments from cells isolated from four litters. Related to GoF cell boundaries shown in Figure 5C. See also Figure 5—video 4. (C) Quantification of the speed of retraction events marked by white lines in kymographs shown in Figure 5A and B, and Figure 5—figure supplement 1A,B (p values for conditional knockout [cKO]+ DMSO against cKO+ Yoda1 determined by two sample t-test while others are determined by Mann Whitney test; Cohen’s d = 2.26, –0.335, and 1.32, respectively). Vertical bars in upper Cumming plots denote mean  ±  s.d.

Figure 5—figure supplement 1—source data 1

Piezo1 gain-of-function keratinocytes have increased retraction.

(Sheet 1) Quantification of the speed of retraction events marked by white lines from kymographs shown in Figure 5A and B, and Figure 5—figure supplement 1A and B.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig5-figsupp1-data1-v2.xlsx
Figure 5—figure supplement 2
PIEZO1 activity increases edge retraction velocity.

(A) Heatmap of cell edge velocity along detected cell boundaries from the Piezo1-GoF and ControlGoF cells shown in Figure 5C and D. Related to Figure 5—video 4.

Figure 5—figure supplement 2—source data 1

PIEZO1 activity increases edge retraction velocity.

Data used to create heatmaps of cell edge velocity along detected cell boundaries from the (Sheet 2) Piezo1-GoF and (Sheet 1) ControlGoF cells shown in Figure 5C and D.

https://cdn.elifesciences.org/articles/65415/elife-65415-fig5-figsupp2-data1-v2.xlsx
Figure 5—figure supplement 3
Piezo1-GoF monolayers advance slower than controls.

(A) Representative differential interference contrast (DIC) kymograph of a region of interest at the wound edge and kymograph of the binarized version of the same video, where the cell monolayer is represented in black and cell-free space of the wounded area in white. Kymographs illustrate wound edge dynamics during an in vitro scratch assay performed in ControlGoF monolayers and (B) Piezo1-GoF monolayers. Images are representative of four independent experiments. Scale bar = 20 μm (vertical) and 100 min (horizontal). Related to Figure 5F–H.

Figure 5—figure supplement 4
Yoda1 shows no effect on Piezo1-cKO monolayer advancement.

(A) Representative differential interference contrast (DIC) kymograph of a region of interest at the wound edge and kymograph of the binarized version of the same video, where the cell monolayer is represented in black and cell-free space of the wounded area in white. Kymographs illustrate wound edge dynamics during an in vitro scratch assay performed in Piezo1-cKO monolayers treated with 4 µM Yoda1. Images are representative of two independent experiments. Scale bar = 20 μm (vertical) and 100 min (horizontal). Related to Figure 5F–H.

Figure 5—video 1
PIEZO1 agonist Yoda1 increases retraction in keratinocytes.

A representative differential interference contrast (DIC) time-lapse movie showing cellular dynamics of ControlcKO keratinocytes in response to treatment with 4 µM Yoda1. Imaging was performed in control medium for 30 min, after which Yoda1 was added, the cell was incubated for 5–7 min, and imaging was resumed. Related to Figure 5A and C–E and Figure 5—figure supplement 1C.

Figure 5—video 2
PIEZO1 agonist Yoda1 can cause widespread retraction in keratinocytes.

A representative differential interference contrast (DIC) time-lapse movie of ControlcKO keratinocytes showing marked cellular retraction in response to treatment with 4 µM Yoda1. Imaging was performed in control medium for 55 min, after which Yoda1 was added, the cell was incubated for 5–7 min, and imaging was resumed. Related to Figure 5A and C–E.

Figure 5—video 3
Yoda1 does not increase retraction in Piezo1 knockout keratinocytes.

A representative differential interference contrast (DIC) time-lapse video showing cellular dynamics of Piezo1-cKO keratinocytes in control medium for 55 min, after which 4 µM Yoda1 was added to imaging solution, the cell was incubated for 5–7 min, and imaging was resumed. Related to Figure 5B–D and Figure 5—figure supplement 1C.

Figure 5—video 4
Piezo1-GoF keratinocytes show increased retraction.

A representative differential interference contrast (DIC) time-lapse video showing cellular dynamics of migrating ControlGoF (left) and Piezo1-GoF keratinocytes (right). Related to Figure 5C, Figure 5—figure supplements 1C and 2.

Figure 5—video 5
PIEZO1 agonist Yoda1 inhibits migration of keratinocyte monolayers and increases wound edge retraction.

A representative differential interference contrast (DIC) time-lapse series showing healing monolayers of keratinocytes during an overnight in vitro scratch assay under control (left) and +4 µM Yoda1 (right) conditions. Related to Figures 1L, 5F and G.

Figure 5—video 6
Piezo1-cKO keratinocyte monolayers heal faster than littermate control.

A representative differential interference contrast (DIC) time-lapse series showing healing monolayers of ControlcKO (left) and Piezo1-cKO (right) keratinocytes during an overnight in vitro scratch assay. Related to Figures 1L and 5H.

Figure 5—video 7
Piezo1-GoF keratinocyte monolayers heal slower than littermate control.

A representative differential interference contrast (DIC) time-lapse series showing healing monolayers of ControlGoF (left) and Piezo1-GoF (right) keratinocytes during an overnight in vitro scratch assay. Related to Figures 1L and 5F–H and Figure 5—figure supplement 3.

Tables

Key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Genetic reagent(mouse)Krt14Cre;Piezo1fl/fl(Piezo1-cKO)This paperGenerated by breeding Piezo1fl/fl mice (Jax stock 029213) with K14Cre mice(The Jackson Laboratory, stock 004782)
Genetic reagent(mouse)Krt14Cre;Piezo1cx/+and Krt14Cre;Piezo1cx/cx(Piezo1-GoF)This paperGenerated by breeding Piezo1cx/cx mice (Ma et al., 2018) with K14Cre mice.
Genetic reagent(mouse)Piezo1-tdTomatoJAX; Ranade et al., 2014029214(RRID:IMSR_JAX:029214)
Genetic reagent(mouse)Piezo1 LacZ reporter miceJAX026948(RRID:IMSR_JAX:026948)
Biological sample (mouse)Murine keratinocytesUC Irvine,The Scripps Research InstituteFreshly isolated from P0–P5 mouse pups
Antibodyanti-Keratin 14 (Rabbit polyclonal)CovanceCat#PRB-155P(RRID:AB_292096)IF (1:1000)
Antibodyanti-Keratin K10 (Rabbit polyclonal)CovanceCat#PRB-159P(RRID:AB_291580)IF (1:1000)
Antibodyanti-Rabbit Alexa Fluor 488 (Goat polyclonal)InvitrogenCat#A11008(RRID:AB_143165)IF (1:1000)
Sequence-based reagentPiezo1Thermo FisherPCR primersTaqman Assay ID: Mm01241570_g1
Sequence-based reagentPiezo2Thermo FisherPCR primersTaqman Assay ID: Mm01262433_m1
Sequence-based reagentGapdhThermo FisherPCR primersTaqman Assay ID: Mm99999915_g1
Sequence-based reagentKrt14Thermo FisherPCR primersTaqman Assay ID: Mm00516876_m1
Commercial assay or kitSuperScript IIIInvitrogen (Thermo Fisher)Cat#12574026Synthesizing cDNA
Commercial assay or kitRNeasy kitQiagenIsolating RNA
Chemical compound, drugCal-520 AMAAT Bioquest IncCat#21130
Chemical compound, drugYoda1TOCRIS558610
Software, algorithmOrigin ProOriginlabOriginPro(RRID:SCR_014212)
Software, algorithmVAMPIREPhillip et al., 2021Version 1.0 (RRID: SCR_021721)https://github.com/kukionfr/VAMPIRE_open
Software, algorithmADAPTBarry et al., 2015(RRID: SCR_006769)https://github.com/djpbarry/Adapt
Software, algorithmDiPERGorelik and Gautreau, 2014(RRID:SCR_021720)
Software, algorithmFlikaEllefsen et al., 2019Version 0.2.17 (RRID: SCR_021719)https://flika-org.github.io/ https://github.com/kyleellefsen/detect_puffs
Software, algorithmCell TrackerPiccinini et al., 2016Version 1.1 (RRID:SCR_021718)http://celltracker.website/index.html
Software, algorithmCellposeStringer et al., 2021Version 0.06 (RRID:SCR_021716)https://github.com/MouseLand/cellpose
Software, algorithmilastikBerg et al., 2019Version 1.4.0b13 (RRID:SCR_015246)https://www.ilastik.org/
Software, algorithmWolfram MathematicaWolframWolfram Mathematica 12
Software, algorithmFIJISchindelin et al., 2012Version 1.53 c(RRID:SCR_002285)https://imagej.net/software/fiji/
Software, algorithmPythonhttps://www.python.org/Version 3.7.0(RRID:SCR_008394)
Software, algorithmKymobutlerJakobs et al., 2019Version: V1v0v2.wl (RRID:SCR_021717)https://github.com/alexlib/KymoButler-1
OtherDAPI stainInvitrogenD1306(RRID:AB_2629482)1:50,000

Additional files

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Jesse R Holt
  2. Wei-Zheng Zeng
  3. Elizabeth L Evans
  4. Seung-Hyun Woo
  5. Shang Ma
  6. Hamid Abuwarda
  7. Meaghan Loud
  8. Ardem Patapoutian
  9. Medha M Pathak
(2021)
Spatiotemporal dynamics of PIEZO1 localization controls keratinocyte migration during wound healing
eLife 10:e65415.
https://doi.org/10.7554/eLife.65415