(A) Real-time analysis of actin dynamics during purse-string wound closure in Caco-2 cells (also see Movie 1; see supplemental material online at www.gastrojournal.org). Closure of oligocellular epithelial wounds was analyzed in Caco-2 cells stably expressing EGFP-β-actin. Within 2 minutes after wounding, actin polymerization occurred at the wound edge (arrow). Residual EGFP fluorescence of dead cells retained within the wound is apparent at time points up to 8 minutes (note top of wound). By 8 minutes, a continuous ring of actin (arrow) was present and the wound rounded, indicating the development of circumferential tension. This was followed by contraction of the actomyosin ring as the wound closed. Most activity occurred within the first 32 minutes and wound closure was complete before 56 minutes (arrow). The wound-healing process was associated with stretching and flattening of some cells, resulting in a gradual shift of some areas out of the plane of focus. Bar = 20 μm. (B) Wound closure occurs rapidly following an initial lag. Wound area was measured at intervals throughout the healing process and normalized to the initial wound area. Wound area began to decrease rapidly 6–8 minutes after injury. Wound closure was complete within 30–45 minutes. Data are representative of more than 20 similar experiments.

(A) Activated rho is present at the wound edge 2 minutes after wounding. Staining for activated rho (red in merged image) revealed its presence in multiple foci along the wound edge within 2 minutes of wounding, both in areas with (insets) and without (arrows) early actin accumulation (green in merged image). In areas with actin accumulation, actin and activated rho colocalized (insets). Cytoplasmic-activated rho was also present away from the developing actin ring (top of field). Bars = 20 μm in low-magnification images and 7 μm in insets. (B) Activated rho is absent at the wound edge 8 minutes after wounding. Only small amounts of activated rho (red in merged image) remained at the wound edge by 8 minutes after wounding. The activated rho that was present did not colocalize with the actin ring (green in merged image). Damaged cells within the wound stained strongly (*) but nonspecifically. Controls using an irrelevant GST fusion protein in place of the GST-rhotekin rho binding domain fusion protein were negative under these imaging conditions. Bars = 20 μm in low-magnification images and 7 μm in insets.

(A) ROCK is present at the wound edge 2 minutes after wounding. Within 2 minutes after wounding, ROCK (red in merged image) was present in a discontinuous punctate pattern at the wound edge, both in association with (insets) and separate from (arrows) actin (green in merged image). ROCK was also detected in punctate cytoplasmic foci throughout the monolayer in cells remote from and bordering the wound. Bars = 20 μm in low-magnification images and 7 μm in insets. Quantitative analysis of ROCK and actin intensity was performed serially over the entire wound edge. To allow comparison of multiple areas, the location of the developing actin ring was set as zero with negative values closer to the center of the wound and positive values further away from the center of the wound. ROCK (red) and actin (green) intensity values are shown. A peak of ROCK is present just ahead of the developing actin ring (black arrow). ROCK then increases further away from the peak of actin, indicating detection of cytosolic ROCK not associated with the wound edge. (B) ROCK is lost from the wound edge by 8 minutes after wounding. Unlike early time points, ROCK (red in merged image) accumulation at the wound edge was not apparent by 8 minutes after wounding despite continued punctate cytoplasmic staining. The actin ring (green in merged image) is clearly visible. Bars = 20 μm in low-magnification images and 7 μm in insets. Quantitative analysis of ROCK and actin intensity was performed as previously described. ROCK (red) and actin (green) intensity values are shown. The peak of actin at 8 minutes is broader than at 2 minutes, indicating greater actin accumulation at this later time. Unlike earlier time points, no peak of ROCK is detected ahead of the actin ring (black arrow, compare with A). Like the earlier time point, cytosolic ROCK is detected behind the actin peak.

(A) Wound closure is disrupted by ROCK inhibition (also see Movie 2; see supplemental material online at www.gastrojournal.org). Closure of an oligocellular wound in the presence of Y-27632, a ROCK inhibitor, added before wounding was studied as in Figure 1. Foci of actin polymerization appeared early in regions along the wound edge (8 minutes, arrow), similar to that seen in untreated wounds, but were unstable and a complete ring did not form (12 minutes, arrow). Unlike untreated wounds, cells surrounding the wound did not exhibit any meaningful contraction. Residual EGFP fluorescence-containing dead cells retained within the wound are apparent at early and late time points (*, compare with Figure 1A). Wound edges did not round, and no significant wound closure occurred. Focal filopodial extensions formed (20 minutes, arrow) but were transient and did not contribute to wound closure. Bar = 20 μm. (B) Quantitative analysis of wound closure when ROCK is inhibited before or after ring assembly. Wounds treated with Y-27632 before wounding did not show any meaningful wound closure (black squares; control wound data are included for comparison, gray circles). In contrast, when Y-27632 was added after actin ring assembly, wound closure proceeded normally (black triangles and C). (C) ROCK inhibition after ring assembly does not disrupt wound closure. Wounds treated with Y-27632 after ring assembly (ie, 6 minutes after wounding) followed a similar pattern and time course to control wounds. Bar = 20 μm.

(A) MLCK recruitment and MLC phosphorylation are minimal 2 minutes after wounding. MLCK (red in merged image) was only focally present at the wound edge 2 minutes after wounding, and these foci were always at sites of actin recruitment (green in merged image). MLC phosphorylation (pMLC; blue in merged image) was minimal at these early time points. Dead/damaged cells within the wound stain nonspecifically (*). Bars = 20 μm in low-magnification images and 7 μm in insets. (B) MLCK recruitment to and MLC phosphorylation at the wound edge coincide with the initiation of contraction. By 8 minutes after wounding, a punctate array of MLCK (red in merged image) was prominent at the wound edge. This MLCK recruitment was accompanied by phosphorylation of MLC (pMLC; blue in merged image). Insets reveal increased staining of MLCK and pMLC at the wound edge in areas of actin accumulation (green in merged image) 8 minutes after wounding. As at 2 minutes, nonspecific staining of dead cells within the wound is evident (*). Bars = 20 μm in low-magnification images and 7 μm in insets.

(A) Model of PIK binding to active MLCK. This model of PIK binding to MLCK suggests that PIK can only bind to MLCK when it is in the active conformation. Thus, when MLCK is in the inactive conformation, PIK cannot bind and is washed away before detection with a secondary reagent. This property allows its use as a probe for activated MLCK. (B) PIK labels the perijunctional actomyosin ring. In steady-state monolayers, PIK binding is concentrated at intercellular junctions, a site of enhanced MLCK localization and MLC phosphorylation. Biotinylated PIK (red) and cell nuclei (blue) are shown. Bar = 20 μm. (C) PIK binds to activated MLCK. Caco-2 cells expressing the catalytic subunit of MLCK and firefly luciferase, both under the control of a tet-off expression system, were cultured in indicated doses of doxycycline. The cells were fixed and stained with biotinylated PIK and fluorescent streptavidin. Quantitative analysis of PIK binding showed that this correlated strongly with gene expression (luciferase, r² = 0.98), consistent with the model presented in A. (D) Activated MLCK is not present at the wound edge 2 minutes after wounding. Activated MLCK, as labeled with biotinylated PIK (red in merged image), was not present at early time points after wounding, when actin (green in merged image) ring assembly was incomplete. Exposures are matched between these images and E. Perijunctional staining (see B) is not visible because it was far less intense than staining at the wound edge (see E ). Some nonspecific staining of debris within the wound (*) is apparent. Bars = 20 μm in low-magnification images and 7 μm in insets. (E ) Activated MLCK is present at the wound edge at 8 minutes after wounding. Activated MLCK (red in merged image) was focally present at the wound edge 8 minutes after wounding and colocalized with actin (green in merged image). The bright nuclear staining (*) was occasionally seen regardless of wound status and likely reflects nonspecific interactions between DNA and the highly charged PIK peptide. Bars = 20 μm in low-magnification images and 7 μm in insets.

(A) Wound healing is delayed and ultimately inhibited by PIK, a specific MLCK inhibitor, in oligocellular cell wounds (also see Movie 3; see supplemental material online at www.gastrojournal.org). MLCK inhibition by PIK did not prevent actin ring assembly (arrows at 2 and 12 minutes) or early wound edge rounding. However, the ring then began to fragment (arrow at 56 minutes). The wound enlarged and appeared to spring back to the original area. Note faint residual cell debris within the wound. Bar = 20 μm. (B) Wound area diminishes briefly but then expands when MLCK is inhibited. In monolayers treated with PIK, wound area initially decreased similar to that of control wounds, but by 8 minutes the rates of closure diverge. With MLCK inhibition wound area decreased only slightly after 8 minutes and ultimately expanded to its original area. (C) PIK effectively inhibits MLCK. In vitro kinase reactions show that 250 μmol/L PIK inhibits >70% of MLC phosphorylation.

MLCK inhibition delays but does not prevent restoration of barrier function after single-cell wounds. In control cells, the magnitude of the conductance leak (g^leak) decreased exponentially over time after wounding. Between 2 and 8 minutes after wounding, g^leak decreased by 53% (from 0.41 ± 0.07 to 0.19 ± 0.03 microSiemens; n = 7; P < .01) in control experiments. A single-exponential fit revealed a time constant of 8.67 ± 1.40 min (n = 7) and was used to estimate g^leak during the 2-minute interval between wounding and the beginning of measurements. With increasing time after wounding, g^leak approached zero asymptotically and barrier function of the epithelium was similar to that before wounding. Monolayers incubated with PIK showed delayed repair kinetics. Between 2 and 8 minutes after wounding, g^leak decreased by only 36% (from 0.52 ± 0.09 to 0.32 ± 0.05 microSiemens; n = 9; P < .05 relative to control wounds). The time constant was 15.14 ± 2.21 minutes (n = 9; P < .05 relative to control wounds). Even 16 minutes after wounding, g^leak in PIK-treated wounds remained elevated relative to control wounds (0.18 ± 0.03 microSiemens with PIK vs 0.07 ± 0.02 microSiemens; n = 9; P < .01). Representative data using the ROCK inhibitor Y-27632 are provided for reference (n = 8; P > .05 vs control).9

MLC phosphorylation occurs at the edge of human oligocellular wounds. Colonic biopsy specimens from patients with active inflammatory bowel disease (both Crohn’s disease and ulcerative colitis) were immunohistochemically stained for phosphorylated MLC (brown). A light hematoxylin counterstain (blue) was then applied. (A) Nonwounded mucosa. As we have noted previously,26 MLC phosphorylation is enhanced at the site of the tight junction (arrows) in normal mucosa. No staining was detected when an irrelevant primary antiserum was used. In oligocellular wounds (B–D), MLC phosphorylation is markedly enhanced along the lateral edges of the wound (arrows). Bar = 5 μm. (B) As in the Caco-2 model, dead/damaged cells are retained within some wounds (*). Although from separate biopsy specimens in different patients, B–D have been arranged in a sequence consistent with progressive wound closure. Three nuclei of damaged or dead cells are apparent within the wound shown in B. (C) Nuclei of dead cells are not apparent, and MLC phosphorylation has extended over nearly the full height of the wound. (D) Ring contraction continues with sealing of the apical portion of the wound.

Model describing the roles of ROCK and MLCK during epithelial purse-string wound closure. (A) Schematic of initial wound within a previously intact epithelial sheet. A continuous multicellular actin ring is not present at the wound edge, and the profile of the wound is irregular. (B) The actin ring assembly phase, where ROCK, activated by rho, directs the recruitment and assembly of actin filaments to form a continuous ring at the wound edge. However, wound contraction has not been initiated and the wound profile remains irregular. (C) The contraction phase, where MLCK-mediated MLC phosphorylation triggers actomyosin contraction, providing the driving force for epithelial purse-string wound closure and rounding the wound edges.