Cornea
Inhibition of RAP1 Enhances Corneal Recovery Following Alkali Injury Ming Wai Poon,1–3 Limeng Yan,3 Dan Jiang,3 Peng Qin,3 Hung-fat Tse,1 Ian Y. Wong,3 David S. H. Wong,3 Vinay Tergaonkar,4 and Qizhou Lian1–3 1
Department of Medicine, Li Ka Shing Faculty of Medicine, University of Hong Kong, Hong Kong SAR, Hong Kong, China Shenzhen Institutes of Research and Innovation, University of Hong Kong, Hong Kong SAR, Hong Kong, China 3Department of Ophthalmology, Li Ka Shing Faculty of Medicine, University of Hong Kong, Hong Kong SAR, Hong Kong, China 4 Institute of Molecular and Cellular Biology, Biopolis, Singapore 2
Correspondence: Qizhou Lian, Department of Medicine and Department of Ophthalmology, University of Hong Kong, Hong Kong SAR; [email protected]. Vinay Tergaonkar, Department of Medicine and Department of Ophthalmology, University of Hong Kong, Hong Kong SAR; [email protected]. Submitted: July 18, 2014 Accepted: December 17, 2014 Citation: Poon MW, Yan L, Jiang D, et al. Inhibition of RAP1 enhances corneal recovery following alkali injury. Invest Ophthalmol Vis Sci. 2015;56:711–721. DOI:10.1167/ iovs.14-15268
PURPOSE. Recently, RAP1 (Telomeric Repeat Binding Factor 2, Interacting Protein [TERF2IP]) was discovered as a modulator that selectively regulates nuclear factor light chain kappa enhancer of activated B cells (NFjB) signaling. The roles of RAP1 in regulation of inflammation and angiogenesis for corneal recovery following corneal injury remain poorly understood. The effects of RAP1 deletion on corneal resurfacing and neovascularization in a corneal alkali burn mouse model were examined. METHODS. Corneal defects and neovascularization were induced in vivo by infliction of an alkali burn to the cornea with 1 N sodium hydroxide solution in RAP1 knock-out (RKO) and wild-type (RWT) mice. Corneal resurfacing was evaluated using slit-lamp biomicroscopy. Neovascularization following injury was evaluated by bright view biomicroscopy and immunofluorescence staining with the endothelial marker platelet endothelial cell adhesion molecule (PECAM). The cytokine profiles of corneal tissue involved in inflammation and neovascularization following injury was compared between RKO and RWT mice. Corneal epithelial cells were isolated for classic scratch wound healing assay and further testing with lipopolysaccharide stimulation. RESULTS. Resurfacing of the burned cornea was accelerated and angiogenesis was suppressed, faster recovery of corneal epithelial cells from classic scratch wound healing and superior tolerance of lipopolysaccharides challenge was observed in the RKO compared to RWT. Molecular investigation revealed that deletion of RAP1 reduced upregulation of inflammatory cytokine (IL1A), finely regulated the expression of angiogenic factor (VEGF), and antiangiogenic factor (PEDF), following injury for better corneal recovery. CONCLUSIONS. Deficiency of RAP1 facilitates corneal recovery after injury. Specificity of RAP1 inhibition may lead to design of specific inhibitors of NFjB in the treatment of ocular injuries. Keywords: RAP1, NFjB, corneal wound healing, corneal resurfacing, angiogenesis, cytokines
he failure of appropriate corneal repair following severe injury (such as chemical burn injuries) often leads to loss of vision. Alkali burn is one of the most severe corneal injuries because alkali agents are lipophilic. Such chemicals penetrate tissue more rapidly than acids to reach the corneal stroma and destroy tissues with a consequent severe inflammatory response.1,2 Nuclear factor kappa light chain enhancer of activated B cells (NFjB) signaling family factors have been implicated in the regulation of inflammation3 and angiogenesis4–6 in many cell types, including corneal epithelial cells.7–10 Given the multifaceted effects of NFjB activity on cardiac, bone and cancer cell fates,11 different molecular cascades that regulate NFjB activity may result in diverse biological outcomes of NFjB signaling in different cell types and at different time points. Prolonged activation of NFjB has a key role in aberrant inflammatory cytokine expression and neovascularization, and the pathogenesis of corneal wound healing.12,13 Inhibition of NFjB pathways may protect against inflammation-induced apoptosis and reduce corneal angiogenesis after corneal alkali injury.14,15 Nevertheless, the NFjB
pathway is vital in maintaining normal host defense and generating innate immune responses to pathogens and microbial products, so complete blockade of NFjB would likely produce unwanted side effects. An ideal NFjB inhibitor for drug and therapeutic development would ablate the proinflammatory actions, but preserve the autoregulatory, anti-inflammatory, and other restorative mechanisms to readjust these multiple interrelated processes.10 The use of transgenic animals is a helpful tool to elucidate the role and mechanisms of specific factor modulating NFjB activation in corneal recovery following severe alkali burn. Recently, we performed a genome-wide gain-of-function screen and unexpectedly identified a telomeric protein RAP1 (Telomeric Repeat Binding Factor 2, Interacting Protein [TERF2IP]) as an inhibitor kinase (IKK) adaptor that modulates NFjB in a dose-dependent manner.16–18 The RAP1-mutant mice are resistant to endotoxic shock and show defective NFjB signaling in response to lipopolysaccharide (LPS) stimulation. This results in reduced upregulation of inflammatory cytokines, such as TNFa and IL-1.16 Corneal repair following alkali injury is closely linked
Copyright 2015 The Association for Research in Vision and Ophthalmology, Inc. www.iovs.org j ISSN: 1552-5783
711
T
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
IOVS j February 2015 j Vol. 56 j No. 2 j 712
FIGURE 1. The effects of deletion of RAP1 on edema and corneal resurfacing post alkali burn. (A) Slit-lamp biomicroscopy was performed at various time points before and after injury (days 0, 3, 7, and 14) for bright field images (upper) and fluorescein images (lower) for RWT and RKO groups. Corneal defects as visualized by staining with fluorescein, cornea defects in green shows the extent of the ocular surface injury in the RWT and RKO groups. Representative photographs of corneal defects in a central corneal injury model for RWT and RKO were shown. Bright field images of (i–v) RWT and (x–xiv) RKO at various time points. The severity of corneal defeat in RKO group was less than that in RWT group. (edema, arrow; *opacity; RWT, N ¼ 4; RKO, N ¼ 4). Scale bars: 2 mm. Green fluorescein staining for corneal defeat in (vi–ix) RWT and (xv–xviii) RKO groups at various time
IOVS j February 2015 j Vol. 56 j No. 2 j 713
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
points. These fluorescein images are corresponding to the bright images. Corneal defeat recovery was significantly faster in the RKO group than RWT group. (B) The percentages of corneal defeat retained post alkali burn were calculated by the area defeat of a particular time point divided by the area of corneal defeat on postinjury day 0 (immediately after the burn was induced). The higher the percentage of defect retained in the particular time point, the slower the corneal resurfacing it means. No defects retained early on day 7 in RKO group, while RWT mice had an average of 67.69% 6 3.39% on day 7 (RWT, N ¼ 4; RKO, N ¼ 4).
to activation of NFjB-mediated inflammatory cytokines14,15 and angiogenic cytokines.19,20 We, therefore, set out to determine whether modulation of NFjB signaling by inhibition of RAP1 could promote corneal recovery following injury by regulating inflammatory cytokines and angiogenic factors. In this study, we compared RAP1 wild-type (RWT) and RAP1 knock-out (RKO) mice to investigate the effects on corneal wound healing following alkali injury, and expression levels of inflammatory cytokines/angiogenic factors after injury. Compared to wild type mice, RAP1-deficient mice displayed better corneal recovery following alkali injury that was accompanied by reduced upregulation of inflammatory factor (IL1A), and fine regulated levels of angiogenic factor (VEGF) as well as that of antiangiogenic factor (PEDF).
MATERIALS
AND
METHODS
Induction of Central Corneal Alkali Injury in RWT and RKO Mice All animal experiments were performed in accordance with relevant guidelines and regulations by the University of Hong Kong, and approved by the Committee on the Use of Live Animals in Teaching and Research (CULTAR). We declare our animal protocol is in compliance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. More details on mouse line16 and genotyping are described in the Supplementary Material. To compare the effects of RAP1 gene deletion on the healing of the cornea following alkali injury, a similar alkali burn was administered to the corneas of RWT and RKO mice. We assessed wound healing by corneal resurfacing and angiogenic response. Under general anesthesia, an ocular surface alkali burn was inflicted to both eyes of adult RWT (N ¼ 64) and RKO (N ¼ 64) mice by application of 3 lL 1 N sodium hydroxide solution to a confined circular area of approximately 2-mm diameter in the central corneal region.14
the corneal surface were performed after injury with fluorescein stain. This was monitored with the slit-lamp system. Eyes also were collected (at baseline, and 6 hours, and 1 and 14 days after injury; N ¼ 32) for total mRNA extraction and subsequent realtime PCR studies of inflammatory cytokines. For angiogenesis response, bright view microphotography examination was done. Angiogenic response was monitored at various postinjury time points (days 0, 7, and 14; N ¼ 8) with a light microscope (Nikon SM2800; Nikon, Tokyo, Japan) and photographed (Catalog No. Infinity 1-3C; Chinetek Scientific, Hong Kong, China). Corneas were dissected from the eye at each postinjury time point (days 3, 7, and 14; N ¼ 24) and flatmounted for platelet endothelial cell adhesion molecule (PECAM, Catalog No. sc18916; Santa Cruz Biotechnology, Santa Cruz, CA, USA) fluorescence staining. Eyes also were collected for total mRNA extraction and subsequent real-time PCR study of angiogenic cytokines. The Table shows the accession numbers and primers for mouse cytokines expression by real-time RT-PCR analysis. The PECAM immunofluorescence staining and imaging were performed for whole cornea and magnified parts of a particular cornea were monitored. More details are described in the Supplementary Material.
Wound Healing Scratch Assay and LPS Challenge in RWT and RKO Corneal Epithelial Monolayer Cultures Briefly, when cell culture reached confluence, a classic wound healing scratch assay (also called migration assay) was performed and different time points of wound recovery were recorded in addition to observed LPS challenge for wound healing. More details are described in the Supplementary Material.
Statistical Analysis A P value of 3 0
Reverse Primers 5 0 –>3 0
Inflammatory cytokine mIL1A
GACAGTATCAGCAACGTCAA
TCACTCTGGTAGGTGTAAGG
NM_001025250.3,NM_009505.4, NM_001025257.3 NM011340.3
CCCGACGAGATAGAGTACAT CCCGACTTCAGCAAGATTAC
CTGGCTTTGGTGAGGTTT CCCTCAGAACAAAGAGGAAAG
NM008084.2
CCACTCACGGCAAATTCA
AGTAGACTCCACGACATACTC
NM010554.4
Angiogenesis cytokines mVEGF mPEDF Housekeeping mGAPDH
The Table shows the accession numbers, oligonucleotide sequences of forward and reverse primers of selected candidates of inflammatory (IL1A), and angiogenesis cytokines (VEGF and PEDF) together with the house keeping probe (GAPDH). m, mouse.
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
IOVS j February 2015 j Vol. 56 j No. 2 j 714
FIGURE 2. Effects of deletion of RAP1 on healing of scratched wounds. (A) The RWT and RKO corneal epithelial cells cultured in 12-well plates were injured with a sterile 10-lL pipette tip at 0 hour. The wounded culture was allowed to re-epithelialize. The percentages of area change of a particular time point in reference to the area of the initial time point (0 hour; immediately after scratch) was monitored over 24 hours at 378C. Representative culture of the RWT at times 0, 6, 12, and 24 hours was displayed on the left, while the RKO culture at the same time points was displayed on the right. The RKO group exhibited significantly faster wound recovery. Compared to RWT, more rapid wound healing was observed in RKO mice at 24 hours to a nearly full recovery (RWT, N ¼ 2; RKO, N ¼ 2) Scale bars: 200 lm. (B) Effect of deletion of RAP1 on cell migration in
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
IOVS j February 2015 j Vol. 56 j No. 2 j 715
(A) is presented in a graph. The percentages of cells retained in the wound region are presented; the percentage of retained wound (RWT or RKO) was defined as the area of wound retained at the particular time point (6, 12, or 24 hours) divided by the area of the original wound at 0 hour in the microscopic view. RAP1 wild-type cell culture had 25.48% 6 6.37% wound retained, while RKO cell culture had 5.06% 6 0.93% wound retained on 24 hours (RWT, N ¼ 2; RKO, N ¼ 2).
RKO mice. Their genotypes were confirmed (Supplementary Fig. S1) as described previously.16 Slit-lamp biomicroscopy was performed at various time points before and after injury (days 0, 3, 7, and 14) for bright field images and fluorescein images (Figs. 1A, 1B). A slit-beam illumination was used to investigate the degree of edema and corneal opacity. Bright field images were captured. The results showed that corneal recovery was faster in RKO than RWT mice. Edema (Fig. 1A, arrow; upper panel of RWT) was very obvious on days 3 and 7 in RWT mice. No edema, but corneal opacity (Fig. 1A, asterisk; upper panel of RWT) was observed on day 14. In contrast, little edema (Fig. 1A, arrow; upper panel of RKO) was observed on days 3 and 7 in RKO mice. The corneas in these mice became transparent on day 14 (Fig. 1A, upper panel of RKO). Green fluorescein staining was used to evaluate corneal surface healing as determined by area of corneal defect14 which stained green. Corneal defeat is taken as 100% on postinjury day 0 (Fig. 1). The RWT and RKO mice were monitored for 14 days (Figs. 1A, 1B). Corneal defect recovery was significantly faster in the RKO group than the RWT group. In the RWT group, the corneal surface stained positive on days 3 and 7; negative fluorescein staining was observed on day 14 (Fig. 1A, lower panel in RWT). In contrast, the corneas of the RKO group were less positive to fluorescein staining on day 3, and negative on days 7 and 14 (Fig. 1A, lower panel in RKO). The percentages of epithelial defect retained in RKO mice were significantly reduced compared to RWT mice (Fig. 1B). The retained percentages for RWT and RKO were 100% and 100% on day 0, 88.60% 6 4.13% and 50.96% 6 6.23% on day 3 (*P < 0.01), 67.69% 6 3.39% and 0% on day 7 (*P < 0.01), and 0% and 0% on day 14 (Fig. 1B).
Deletion of RAP1 Enhances Cell Migration of Corneal Epithelial Cells Cells cultured from RWT and RKO mice stained positive in response to the corneal epithelial marker K3, but negative to the conjunctival epithelial cell marker K19 (Supplementary Fig. S2). This confirmed their identities as corneal epithelial cells. An in vitro scratch wound healing assay was used to assess the role of RAP1 deletion in migration.3,21,22 Confluent dishes of corneal epithelial cells isolated from corneas of RWT and RKO mice were wounded by scraping with a 10-lL pipette tip (Tip One; USA Scientific, Inc., Ocala, FL, USA), creating a space free of cells (Figs. 2A, 2B). The rate of wound closure was faster in RKO mice than RWT mice, evident 24 hours after scraping. The percentages retained for RWT and RKO were: 100% and 100% at 0 hours, 86.47% 6 19.13% and 56.71% 6 4.43% at 6 hours, 65.63% 6 9.32% and 43.52% 6 0.76% at 12 hours, and 25.48% 6 6.37% and 5.06 6 0.93% at 24 hours. Deletion of RAP1 facilitates migration of corneal epithelial cells upon wound scraping.
Absence of RAP1 Reduced LPS-Inhibited Wound Healing Lipopolysaccharides (LPS) provide a potent inflammatory stimulus and their presence has major effects on the determinants of inflammation and apoptosis.23 To evaluate the effects of RAP1 deletion on response to LPS insult, corneal
epithelial cell cultures were challenged with (100 ng/mL) LPS24,25 for 24 hours following wound scratching.24,25 The results revealed that RKO mice were more tolerant than RWT to LPS stimulation and demonstrated a faster wound healing response evident 24 hours after scraping (Fig. 3A). The percentages retained for RWT and RKO were 100% and 100% at 0 hours, 81.77% 6 13.91% and 87.33% 6 4.70% after 6 hours, 69.81% 6 3.05% and 66.58% 6 3.71% after 12 hours, and 66.63% 6 4.73% and 0% (*P < 0.01) after 24 hours (Fig. 3B). Cells in the RKO group had a superior recovery response to LPS stimulation than RWT.
Absence of RAP1 Reduced Inflammatory Cytokine IL1A After Alkali Injury The IL1A is known to be a master gene that upregulates inflammation and various cellular responses in corneal recovery.26 Expression of IL1A is upregulated and peaks at 6/ 24 hours after alkali injury,27,28 in part via the NFkB/RAP1 signaling pathway. We compared the IL1A expression dynamics in alkali-burned corneas in RWT to RKO mice, at each of the time points by real-time RT-PCR; the relative level of expression for RKO was presented as the ratio of RKO against RWT. The relative expression of RKO compared to RWT at the pre-injury time point was taken as the baseline (Fig. 4). Expression of IL1A was markedly less upregulated in the RKO group compared to the RWT group (Fig. 4, *P < 0.01). In RKO mice, IL1A expression was 8.33-fold lower after 6 hours compared to RWT mice following corneal alkali injury (*P < 0.01) and 4-fold lower after 24 hours (*P < 0.01). On recovery day 14 (1.23fold), in RKO, IL1A level returned to baseline level compared to RWT (1.34-fold, Fig. 4). Collectively, RAP1 deletion could finely regulate the expression of the inflammatory cytokine, IL1A at 6 and 24 hours after injury.
Reduced Angiogenic Response to Alkali Injury in Corneas of RKO Mice Alkaline was administered to the corneas of RWT and RKO mice to examine the effects of deletion of RAP1 (RKO) on corneal neovascularization. Bright view microscopy was performed over 14 days (after injury days 0, 7, and 14). The total area of thick and thin blood vessels was observed and measured for four corneas from either RWT or RKO mice on days 0, 7, and 14 after injury (Fig. 5A). A weaker corneal angiogenic response was observed in the RKO group compared to RWT. Corneal angiogenesis gradually increased from days 7 to 14 in the RWT group, but reduced in the RKO group at the same time points (Fig. 5A). The mean areas of corneal neovascularization of RWT and RKO were: 1.27 6 0.22 and 0.40 6 0.09 mm2 on day 7 (*P < 0.01), and 1.49 6 0.24 and 0.13 6 0.024 mm2 on day 14 (**P < 0.001, Fig. 5B). Thus, corneal neovascularization was significantly reduced by deletion of RAP1 (Fig. 5B).
Reduced Endothelial Vessels (PECAM-Positive) Sprouting From the Limbus to the Center of the Cornea in RKO Mice To confirm the angiogenic response in RWT and RKO mice corneas, corneas were stained with an endothelial cell marker, PECAM, to identify new vessels sprouting from the limbus
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
IOVS j February 2015 j Vol. 56 j No. 2 j 716
FIGURE 3. Effects of deletion of RAP1 on wound healing upon LPS stimulation. (A) The RWT and RKO corneal epithelial cells cultured in 12-well plates were injured with sterile 10-lL pipette tips at 0 hours, and cultured with (100 ng/mL) LPS supplement cultured medium. The wounded culture was allowed to re-epithelialize for 2 4hours at 378C. The percentages of area change of a particular time point in reference to the area of the initial time point (0 hour; immediately after scratch) was monitored over 24 hours at 378C. Representative culture of the RWT at time 0, 6, 12, and 24 hours was displayed on the left, while the representative RKO culture at the same time points was displayed on the right. Compared to RWT, more rapid wound healing was observed in RKO mice at 24 hours to a full recovery (RWT, N ¼ 2; RKO, N ¼ 2) Scale bars: 200 lm. (B) Effect of
IOVS j February 2015 j Vol. 56 j No. 2 j 717
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
deletion of RAP1 on LPS-challenge in (A) was presented in a graph. The percentages of cells retained in wound regions are represented for the time points. The percentage of retained wound (RWT or RKO) was defined as the area of wound retained at the particular time point (6, 12, or 24 hours) divided by the area of the original wound at 0 hours in the microscopic view. The RWT group had an average of 66.63% 6 4.73% wound retained on 24 hours (RWT, N ¼ 2; RKO, N ¼ 2; *P < 0.01).
region. These images have been merged to reconstruct a full view of a particular cornea (Figs. 6A, 6B). The total areas of PECAM-positive corneal vessels were measured from 4 corneas each of RWT and RKO mice and at each time point. Fewer PECAM-positive vessels were observed in RKO mice on day 7 (Fig. 6A) and day 14 (Fig. 6B) compared to RWT. In contrast, more PECAM-positive stained blood vessels were observed in the corneas of RWT mice at the same time points (Figs. 6A, 6B). The mean areas for RWT and RKO mice were 3.61 6 0.18 and 1.55 6 0.29 mm2 on day 7 (*P < 0.01, and 8.42 6 0.86 and 0.10 6 0.01 mm2 on day 14 (*P < 0.01, Fig. 6C). Neovascularization in RKO mice was significantly reduced compared to RWT at various time points following alkali injury. Magnified images (Fig. 6D, ~2.4 mm2 of the actual size from the cornea around the limbus) of corneas were captured from RWT and RKO mice. Representative images are presented, extracted from their corresponding images, in Figures 6A and 6B (asterisks). Progression of PECAM-positive stained blood vessels was observed in the corneas of RWT mice on days 3, 7, and 14 (Figs. 6D, 6E). Fewer PECAM-positive vessels were observed in RKO than RWT: PECAM-positive vessels were significantly reduced on day 14 (Figs. 6D, 6E). The mean areas of the PECAM-positive stain in the magnified images are shown in Figure 6E. The areas in RWT and RKO mice were 0.27 6 0.03 and 0.076 6 0.01 mm2 on day 3 (**P < 0.001), 0.63 6 0.1 and 0.31 6 0.08 mm2 on day 7 (*P 0.01, and 1.92 6 0.09 and 0.018 6 0.01 mm2 on day 14, respectively (**P < 0.001, Fig. 6E). Collectively, these data demonstrated that deficiency in RAP1 can reduce the angiogenic response to corneal alkali injury.
Fine Regulation of Angiogenic and Anti-Angiogenic Factors in RKO Mice Corneas Following Alkali Injury To evaluate the effects of RKO on angiogenesis responses in alkali-burned corneas, we performed real time RT-PCR on the angiogenic cytokines, VEGF, and PEDF (Fig. 7). Expression of
RKO was compared to RWT at each time point; the relative level of expression for RKO was presented as the ratio of RKO against RWT. The relative expression of RKO compared to RWT on the pre-injury time point was taken as the baseline (Figs. 7A, 7B). Compared to corneas of RWT mice, real time RT-PCR showed marked relatively downregulation (2.63-fold, *P < 0.01) of PEDF mRNA expression in RKO corneas 6 hours after injury, but relatively upregulation (2.15-fold, *P < 0.01) of VEGF in RKO. In addition, when PEDF expression was relative upregulated 2.20-fold (*P < 0.01) in RKO corneas on day 7, VEGF levels were relative downregulated 1.79-fold (*P < 0.01). Level of cytokine expression (PEDF and VEGF) returned toward normal (baseline) upon recovery on day 14 (1.41and 0.97-fold respectively) in RWT and RKO groups (Figs. 7A, 7B). These data demonstrates that deletion of RAP1 finely controls the expression levels of angiogenic and antiangiogenic factors for better corneal injury and recovery process compared to RWT.
DISCUSSION Alkali seriously damages the cornea and causes loss of vision in humans, and provokes inflammatory and angiogenic responses that result in a release of inflammatory cytokines and angiogenic factors, activation of NFjb pathway,14 and a cellular cascade including cell proliferation, and cell apoptosis in the corresponding corneal26 and endothelial cells.20,29 Prolonged activation of NFjB has a key role in aberrant inflammatory cytokine expression and neovascularization,12,13 and often results in pathogenic corneal wound healing with consequent vision loss.30 In alkali injury, inhibition of NFjB that can reduce corneal angiogenesis may serve as a therapeutic strategy.15 The NFjB inhibitors have shown anti-inflammatory effects and include emodin, besifloxacin, BOL-303242-X (mapracorat), thymosinb4, epigallocatechin gallate, Perilla frutescens leaf extract, and IKKb-targeting short interfering RNA.10,31 Nevertheless inhibi-
FIGURE 4. The effect of deletion of RAP1 on the relative mRNA expression level of IL1A of RKO compared to RWT in the corneas post alkali burn. The relative expression of IL1A in corneas was measured at various time points for RKO compared to RWT. The relative expression level at a particular time point is calculated as a ratio of expression level of the RKO group over which of the RWT group. The relative expression of RKO to RWT at time point pre-injury was taken as the baseline level. Compared to RWT corneas, the relative expression level of IL1A was significantly less upregulated in RKO group at 6 hours (~8.33-fold, *P < 0.01), day 1 (~4.07-fold, *P < 0.01), and returns to baseline on day 14 (~1.23-fold) after alkali injury. (RWT, N ¼ 16; RKO, N ¼ 16).
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
IOVS j February 2015 j Vol. 56 j No. 2 j 718
FIGURE 5. The effects of deletion of RAP1 on angiogenesis response after alkali injury. (A) Representative photographs of corneal neovascularization in a central corneal injury model. Dorsal and lateral images were presented for RWT and RKO corneas on days 0, 7, and 14 after injury time points. The RWT and RKO images are illustrated on the left and the right columns, respectively. The severity of corneal neovascularization in the RKO group was less than that in the RWT group (neovascularization was marked by arrows; RWT, N ¼ 4; RKO, N ¼ 4). Scale bars: 2 mm. (B) The mean total areas of new vessels were graphically presented. The total areas of new vessels from each cornea from the RWT and RKO groups were separately quantified by ImageJ (http://imagej.nih.gov/ij/; provided in the public domain by the National Institutes of Health, Bethesda, MD, USA) for various time points (days 7 and14). A significant decrease of angiogenesis response was observed in RKO group on day 7 (RWT and RKO; 1.27 6 0.22 and 0.40 6 0.09 mm2; *P < 0.01) and day 14 (RWT and RKO; 1.49 6 0.24 and 0.13 6 0.024 mm2; **P < 0.001) compared to RWT group (RWT, N ¼ 4; RKO, N ¼ 4).
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
IOVS j February 2015 j Vol. 56 j No. 2 j 719
FIGURE 6. (A–E) Deletion of RAP1 inhibits angiogenesis in the mouse cornea after alkali injury. (A) Alkali injury was administered on the central corneal region for RWT and RKO group mice in vivo. Corneas were dissected and collected on the various time points (days 7 and 14) for whole mount and stained with endothelial marker, PECAM. Representative PECAM-positive images of whole mount corneas of RWT (left) and RKO (right) on day 7 after injury (RWT, N ¼ 4; RKO, N ¼ 4). Scale bars: 1 mm. (B) Representative PECAM-positive images of whole mount corneas of RWT (left) and RKO (right) on day 14. Neovascularization was inhibited in PECAM-stained corneas from RKO on (A) day 7 and (B) day 14, compared to RWT corneas of the same time point (RWT, N ¼ 4; RKO, N ¼ 4). Scale bars: 1 mm. (C) Quantification of mean total areas of blood vessel sprouting in RWT and RKO mice. Total area of the PECAM-positive vessels very quantified for individual RWT and RKO cornea. Mean areas for RWT and RKO groups were shown in the graph. The sprouting of blood vessels was significantly reduced in RKO mice compared to RWT mice, less blood vessel sprouting in RKO at day 7 (RWT and RKO; 3.61 6 0.18 and 1.55 6 0.29 mm2, *P < 0.01) and day 14 (RWT and RKO and 8.42 6 0.86 and 0.10 6 0.01 mm2; *P < 0.01; RWT, N ¼ 8; RKO, N ¼ 8). (D) Magnified images of PECAM-positive vessels from the corneas (actual size 2.4 mm2 surrounding the limbus region). Representative photographs were extracted from sections marked with asterisks from Figures 6A and 6B on the various time points (days 3, 7, and 14); RWT (left column) and RKO (right column). Neovascularization was inhibited in PECAM-stained corneas from RKO on days 3, 7, and 14 (RWT, N ¼ 12; RKO, N ¼ 12; Scale bars: 1 mm) when compared to RWT. (E) Quantification of blood vessel sprouting in RWT and RKO mice on magnified (*) images from Figures 6A and 6B. The sprouting of blood vessels was significantly reduced in RKO mice compared to RWT mice, less blood vessel was sprouting in RKO in day 3 (RWT and RKO; 0.27 6 0.02 and 0.08 6 0.01 mm2 ; **P < 0.001), blood vessel sprouting in RKO in day 7 (RWT and RKO; 0.63 6 0.06 and 0.31 6 0.05 mm2; *P 0.01), and day 14 (RWT and RKO; 1.92 6 0.05 and 0.02 6 0.003 mm2; **P < 0.001; RWT, N ¼ 12; RKO, N ¼ 12).
tion of NFjB has shown variable results and its multifaceted and cell type specificity6,15,32–37 is associated with the fine control of healing of epithelial cells.38 A therapeutic strategy of fine modulation of NFjB signaling rather than complete blockage of NFjB, therefore, is important to preserve the autoregulatory, anti-inflammatory, and other restorative mechanisms and readjust these multiple interrelated processes.39,40 Most common NFjkB dimer is composed of p50 and p65 subunit and this is activated in many cancers. The SN50
peptide blocks this dimer and, hence, NFjkB activation response. This is a very generic mechanism of blocking NFjKB, much like inhibiting the upstream kinases, which, though effective in in vitro settings, can lead to many toxic effects in vivo.41 The Rap1 was identified as a telomeric protein and its role in inflammation is considered its noncanonical function. The Rap1 activated NFjkB by making it a strong chromatin remodeler. The NFjkB p65 subunit is modified by phosphorylation induced by IKKs, which are kinases that also
IOVS j February 2015 j Vol. 56 j No. 2 j 720
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
stress resistance,16 and is associated with fine regulation of inflammatory cytokines, angiogenic factors, and antiangiogenic factors. The IL1A is a prime proinflammatory cytokine arising from NFjB activation44 and the first to be released following injury for repair, proliferation, apoptosis, motility, differentiation, and necrosis in various cell types, including epithelial cells.26,45 Timely and tightly controlled angiogenic VEGF and antiangiogenic PEDF factors have important roles in corneal repairs following injuries46,47 that are associated with NFjB signaling regulation.20,29 This study provides in vivo evidence that a reduced RAP1 level promotes corneal repair following alkali-induced damage, and is associated with a reduced upregulation of inflammatory and angiogenic cytokines. Inhibition of RAP1 enhances corneal recovery and may serve as a novel therapeutic target that can fine regulate NFjB signaling-mediated inflammatory and angiogenic responses.
Acknowledgments
FIGURE 7. The effects of deletion of RAP1 on the relative mRNA expression of (A) PEDF and (B) VEGF in RKO corneas compared to RWT corneas after alkali injury. The relative expression of (A) PEDF and (B) VEGF in corneas was measured at various time points (preinjury, post injury 6 hours, and days 3, 7, and 14) for RKO compared to RWT. The relative expression at a particular time point is calculated as a ratio of expression of RKO over that of RWT at a particular time point. The relative expression of RKO to RWT on preinjury time point was taken as the baseline level. Compared to RWT, at 6 hours after injury in the RKO group, the relative expression level of PEDF (~2.63-fold, *P < 0.01) was significantly reduced, while that of VEGF was significantly upregulated (~2.15-fold, *P < 0.01). On day 7 post injury in RKO mice, dynamic changes to the relative expression levels of VEGF (~ 1.79-fold downregulated, *P < 0.01) and PEDF (~2.20-fold upregulated, *P < 0.01) were observed. (RWT, N ¼ 20; RKO, N ¼ 20).
mark IkB proteins for degradation. Only p65 that is modified by IKK in presence of Rap1 is capable of binding p300/CBP and open chromatin on specific promoters. Hence, inhibiting Rap1 is much more specific than generally inhibiting NFjkB by SN50, which will block NFjkB binding to all sites. Such specificity of Rap1 inhibition will lead to design of much more specific inhibitors of NFjkB and inflammation. This is apparent in many in vivo assays presented here and in the fact that Rap1 null mice are normal until challenged by stress-inducing agents. Like Rap1, the role of telomerase42 in regulating inflammation is now well appreciated and targeting telomerase is considered a worthwhile therapy.43 In this study we have shown that RAP1 is an effective modulator of NFjB signaling that, in turn, is required for cytokine regulation and subsequent fine healing of corneal injury. Our results demonstrated that inhibition of NFjB by inhibition of RAP1 facilitates corneal healing. Corneal resurfacing was associated with reduced upregulation of inflammatory cytokines, while in corneal angiogenesis, repressed induction of angiogenesis was associated with reduced upregulation of angiogenic factors, but increased upregulation of antiangiogenic factors. Inhibition of RAP1 facilitates cell migration in vitro and increases tolerance to bacterial challenge. Our findings are consistent with those described previously that inhibition of NFjB signaling facilitates corneal recovery, corneal resurfacing,15 reduced angiogenesis,14 and
Supported in part by grants from the RGC Research Grant (N_HKU 716/09, HKU3/CRF/11R, and T12-705/11 to [QL]); National Natural Science Foundation of China (31270967 [QL]); small project funding from The University of Hong Kong (201111159183 and 201211159030 [QL], and 201109176184 and 201309176129 [MWP]); and a core grant to IMCB from A*star (VT). Disclosure: M.W. Poon, None; L. Yan, None; D. Jiang, None; P. Qin, None; H.-F. Tse, None; I.Y. Wong, None; D.S.H. Wong, None; V. Tergaonkar, None; Q. Lian, None
References 1. Saika S, Kobata S, Hashizume N, Okada Y, Yamanaka O. Epithelial basement membrane in alkali-burned corneas in rats. Immunohistochemical study. Cornea. 1993;12:383–390. 2. Ishizaki M, Zhu G, Haseba T, Shafer SS, Kao WW. Expression of collagen I, smooth muscle alpha-actin, and vimentin during the healing of alkali-burned and lacerated corneas. Invest Ophthalmol Vis Sci. 1993;34:3320–3328. 3. Tong L, Tergaonkar V. Rho protein GTPases and their interactions with NFkappaB: crossroads of inflammation and matrix biology. Biosci Rep. 2014;34:e00115. 4. Ben-Neriah Y, Karin M. Inflammation meets cancer, with NFkappaB as the matchmaker. Nat Immunol. 2011;12:715–723. 5. Tabruyn SP, Griffioen AW. Molecular pathways of angiogenesis inhibition. Biochem Biophys Res Comm. 2007;355:1–5. 6. Tak PP, Firestein GS. NF-kappaB: a key role in inflammatory diseases. J Clin Invest. 2001;107:7–11. 7. Nakao S, Hata Y, Miura M, et al. Dexamethasone inhibits interleukin-1beta-induced corneal neovascularization: role of nuclear factor-kappaB-activated stromal cells in inflammatory angiogenesis. Am J Path. 2007;171:1058–1065. 8. Kubota M, Shimmura S, Kubota S, et al. Hydrogen and N-acetylL-cysteine rescue oxidative stress-induced angiogenesis in a mouse corneal alkali-burn model. Invest Ophthalmol Vis Sci. 2011;52:427–433. 9. Wang CY, Mayo MW, Korneluk RG, Goeddel DV, Baldwin AS Jr. NF-kappaB antiapoptosis: induction of TRAF1 and TRAF2 and c-IAP1 and c-IAP2 to suppress caspase-8 activation. Science. 1998;281:1680–1683. 10. Lan W, Petznick A, Heryati S, Rifada M, Tong L. Nuclear factorkappaB: central regulator in ocular surface inflammation and diseases. Ocular Surf. 2012;10:137–148. 11. Gordon JW, Shaw JA, Kirshenbaum LA. Multiple facets of NFkappaB in the heart: to be or not to NF-kappaB. Circ Res. 2011;108:1122–1132.
RAP1-NFjB Pathway Facilitates Corneal Wound Healing 12. Cooks T, Pateras IS, Tarcic O, et al. Mutant p53 prolongs NFkappaB activation and promotes chronic inflammation and inflammation-associated colorectal cancer. Cancer Cell. 2013; 23:634–646. 13. Kisseleva T, Song L, Vorontchikhina M, Feirt N, Kitajewski J, Schindler C. NF-kappaB regulation of endothelial cell function during LPS-induced toxemia and cancer. J Clin Invest. 2006; 116:2955–2963. 14. Saika S, Ikeda K, Yamanaka O, et al. Expression of Smad7 in mouse eyes accelerates healing of corneal tissue after exposure to alkali. Am J Path. 2005;166:1405–1418. 15. Saika S, Miyamoto T, Yamanaka O, et al. Therapeutic effect of topical administration of SN50, an inhibitor of nuclear factorkappaB, in treatment of corneal alkali burns in mice. Am J Path. 2005;166:1393–1403. 16. Teo H, Ghosh S, Luesch H, et al. Telomere-independent Rap1 is an IKK adaptor and regulates NF-kappaB-dependent gene expression. Nature Cell Biol. 2010;12:758–767. 17. Ghosh AS, Tergaonkar V. Telomeres and inflammation: Rap1 joins the ends? Cell Cycle. 2010;9:3834–3835. 18. Ghosh A, Saginc G, Leow SC, et al. Telomerase directly regulates NF-kappaB-dependent transcription. Nat Cell Biol. 2012;14:1270–1281. 19. Tong Q, Zheng L, Lin L, et al. VEGF is upregulated by hypoxiainduced mitogenic factor via the PI-3K/Akt-NF-kappaB signaling pathway. Resp Res. 2006;7:37. 20. Keskin D, Kalluri R. NF-kappaB-induced chromatin remodeling regulates angiogenesis. Blood. 2010;116:312–313. 21. Wells A. Tumor invasion: role of growth factor-induced cell motility. Adv Cancer Res. 2000;78:31–101. 22. Xu KP, Riggs A, Ding Y, Yu FS. Role of ErbB2 in corneal epithelial wound healing. Invest Ophthalmol Vis Sci. 2004;45: 4277–4283. 23. Liang H, Brignole-Baudouin F, Labbe A, Pauly A, Warnet JM, Baudouin C. LPS-stimulated inflammation and apoptosis in corneal injury models. Mol Vis. 2007;13:1169–1180. 24. Karlstetter M, Lippe E, Walczak Y, et al. Curcumin is a potent modulator of microglial gene expression and migration. J Neuroinflamm. 2011;8:125. 25. Kim SY, Lee JG, Cho WS, et al. Role of NADPH oxidase-2 in lipopolysaccharide-induced matrix metalloproteinase expression and cell migration. Immunol Cell Biol. 2010;88:197–204. 26. Wilson SE, Mohan RR, Ambrosio R Jr, Hong J, Lee J. The corneal wound healing response: cytokine-mediated interaction of the epithelium, stroma, and inflammatory cells. Prog Retin Eye Res. 2001;20:625–637. 27. Lausch RN, Chen SH, Tumpey TM, Su YH, Oakes JE. Early cytokine synthesis in the excised mouse cornea. J Interferon Cytokine Res. 1996;16:35–40. 28. Planck SR, Rich LF, Ansel JC, Huang XN, Rosenbaum JT. Trauma and alkali burns induce distinct patterns of cytokine gene expression in the rat cornea. Ocul Immunol Inflamm. 1997;5:95–100. 29. Aurora AB, Biyashev D, Mirochnik Y, et al. NF-kappaB balances vascular regression and angiogenesis via chromatin remodeling and NFAT displacement. Blood. 2010;116:475–484.
IOVS j February 2015 j Vol. 56 j No. 2 j 721 30. Zhang MC, Wang Y, Yang Y. The expression of nuclear factor kappa B in inflammation-induced rat corneal neovascularization. Ocul Immunol Inflamm. 2006;14:359–365. 31. Sosne G, Qiu P, Christopherson PL, Wheater MK. Thymosin beta 4 suppression of corneal NFkappaB: a potential antiinflammatory pathway. Exp Eye Res. 2007;84:663–669. 32. Seetharaman R, Mora AL, Nabozny G, Boothby M, Chen J. Essential role of T cell NF-kappa B activation in collageninduced arthritis. J Immunol. 1999;163:1577–1583. 33. Baldwin AS Jr. Series introduction: the transcription factor NFkappaB and human disease. J Clin Invest. 2001;107:3–6. 34. Tak PP, Gerlag DM, Aupperle KR, et al. Inhibitor of nuclear factor kappaB kinase beta is a key regulator of synovial inflammation. Arthritis Rheum. 2001;44:1897–1907. 35. Yamamoto Y, Gaynor RB. Therapeutic potential of inhibition of the NF-kappaB pathway in the treatment of inflammation and cancer. J Clin Invest. 2001;107:135–142. 36. Yamamoto Y, Gaynor RB. Role of the NF-kappaB pathway in the pathogenesis of human disease states. Curr Mol Med. 2001;1:287–296. 37. Jimi E, Aoki K, Saito H, et al. Selective inhibition of NF-kappa B blocks osteoclastogenesis and prevents inflammatory bone destruction in vivo. Nat Med. 2004;10:617–624. 38. Zhou R, Hu G, Gong AY, Chen XM. Binding of NF-kappaB p65 subunit to the promoter elements is involved in LPS-induced transactivation of miRNA genes in human biliary epithelial cells. Nucl Acids Res. 2010;38:3222–3232. 39. Krakauer T. Nuclear factor-kappaB: fine-tuning a central integrator of diverse biologic stimuli. Int Rev Immunol. 2008;27:286–292. 40. Beg AA, Sha WC, Bronson RT, Ghosh S, Baltimore D. Embryonic lethality and liver degeneration in mice lacking the RelA component of NF-kappa B. Nature. 1995;376:167– 170. 41. Dey A, Wong E, Kua N, Teo HL, Tergaonkar V, Lane D. Hexamethylene bisacetamide (HMBA) simultaneously targets AKT and MAPK pathway and represses NF kappaB activity: implications for cancer therapy. Cell Cycle. 2008;7:3759–3767. 42. Ghosh S, Karin M. Missing pieces in the NF-kappaB puzzle. Cell. 2002;109(suppl):S81–S96. 43. Li Y, Tergaonkar V. Noncanonical functions of telomerase: implications in telomerase-targeted cancer therapies. Cancer Res. 2014;74:1639–1644. 44. Sheedy FJ, Moore KJ. IL-1 signaling in atherosclerosis: sibling rivalry. Nat Immunol. 2013;14:1030–1032. 45. Papacleovoulou G, Critchley HO, Hillier SG, Mason JI. IL1alpha and IL4 signalling in human ovarian surface epithelial cells. J Endocr. 2011;211:273–283. 46. Azar DT. Corneal angiogenic privilege: angiogenic and antiangiogenic factors in corneal avascularity, vasculogenesis, and wound healing (an American Ophthalmological Society thesis). Trans Am Ophthalmol Soc. 2006;104:264– 302. 47. Pan Z, Fukuoka S, Karagianni N, Guaiquil VH, Rosenblatt MI. Vascular endothelial growth factor promotes anatomical and functional recovery of injured peripheral nerves in the avascular cornea. FASEB J. 2013;27:2756–2767.
Inhibition of RAP1 Enhances Corneal Recovery Following Alkali Injury Ming Wai Poon,1–3 Limeng Yan,3 Dan Jiang,3 Peng Qin,3 Hung-fat Tse,1 Ian Y. Wong,3 David S. H. Wong,3 Vinay Tergaonkar,4 and Qizhou Lian1–3 1
Department of Medicine, Li Ka Shing Faculty of Medicine, University of Hong Kong, Hong Kong SAR, Hong Kong, China Shenzhen Institutes of Research and Innovation, University of Hong Kong, Hong Kong SAR, Hong Kong, China 3Department of Ophthalmology, Li Ka Shing Faculty of Medicine, University of Hong Kong, Hong Kong SAR, Hong Kong, China 4 Institute of Molecular and Cellular Biology, Biopolis, Singapore 2
Correspondence: Qizhou Lian, Department of Medicine and Department of Ophthalmology, University of Hong Kong, Hong Kong SAR; [email protected]. Vinay Tergaonkar, Department of Medicine and Department of Ophthalmology, University of Hong Kong, Hong Kong SAR; [email protected]. Submitted: July 18, 2014 Accepted: December 17, 2014 Citation: Poon MW, Yan L, Jiang D, et al. Inhibition of RAP1 enhances corneal recovery following alkali injury. Invest Ophthalmol Vis Sci. 2015;56:711–721. DOI:10.1167/ iovs.14-15268
PURPOSE. Recently, RAP1 (Telomeric Repeat Binding Factor 2, Interacting Protein [TERF2IP]) was discovered as a modulator that selectively regulates nuclear factor light chain kappa enhancer of activated B cells (NFjB) signaling. The roles of RAP1 in regulation of inflammation and angiogenesis for corneal recovery following corneal injury remain poorly understood. The effects of RAP1 deletion on corneal resurfacing and neovascularization in a corneal alkali burn mouse model were examined. METHODS. Corneal defects and neovascularization were induced in vivo by infliction of an alkali burn to the cornea with 1 N sodium hydroxide solution in RAP1 knock-out (RKO) and wild-type (RWT) mice. Corneal resurfacing was evaluated using slit-lamp biomicroscopy. Neovascularization following injury was evaluated by bright view biomicroscopy and immunofluorescence staining with the endothelial marker platelet endothelial cell adhesion molecule (PECAM). The cytokine profiles of corneal tissue involved in inflammation and neovascularization following injury was compared between RKO and RWT mice. Corneal epithelial cells were isolated for classic scratch wound healing assay and further testing with lipopolysaccharide stimulation. RESULTS. Resurfacing of the burned cornea was accelerated and angiogenesis was suppressed, faster recovery of corneal epithelial cells from classic scratch wound healing and superior tolerance of lipopolysaccharides challenge was observed in the RKO compared to RWT. Molecular investigation revealed that deletion of RAP1 reduced upregulation of inflammatory cytokine (IL1A), finely regulated the expression of angiogenic factor (VEGF), and antiangiogenic factor (PEDF), following injury for better corneal recovery. CONCLUSIONS. Deficiency of RAP1 facilitates corneal recovery after injury. Specificity of RAP1 inhibition may lead to design of specific inhibitors of NFjB in the treatment of ocular injuries. Keywords: RAP1, NFjB, corneal wound healing, corneal resurfacing, angiogenesis, cytokines
he failure of appropriate corneal repair following severe injury (such as chemical burn injuries) often leads to loss of vision. Alkali burn is one of the most severe corneal injuries because alkali agents are lipophilic. Such chemicals penetrate tissue more rapidly than acids to reach the corneal stroma and destroy tissues with a consequent severe inflammatory response.1,2 Nuclear factor kappa light chain enhancer of activated B cells (NFjB) signaling family factors have been implicated in the regulation of inflammation3 and angiogenesis4–6 in many cell types, including corneal epithelial cells.7–10 Given the multifaceted effects of NFjB activity on cardiac, bone and cancer cell fates,11 different molecular cascades that regulate NFjB activity may result in diverse biological outcomes of NFjB signaling in different cell types and at different time points. Prolonged activation of NFjB has a key role in aberrant inflammatory cytokine expression and neovascularization, and the pathogenesis of corneal wound healing.12,13 Inhibition of NFjB pathways may protect against inflammation-induced apoptosis and reduce corneal angiogenesis after corneal alkali injury.14,15 Nevertheless, the NFjB
pathway is vital in maintaining normal host defense and generating innate immune responses to pathogens and microbial products, so complete blockade of NFjB would likely produce unwanted side effects. An ideal NFjB inhibitor for drug and therapeutic development would ablate the proinflammatory actions, but preserve the autoregulatory, anti-inflammatory, and other restorative mechanisms to readjust these multiple interrelated processes.10 The use of transgenic animals is a helpful tool to elucidate the role and mechanisms of specific factor modulating NFjB activation in corneal recovery following severe alkali burn. Recently, we performed a genome-wide gain-of-function screen and unexpectedly identified a telomeric protein RAP1 (Telomeric Repeat Binding Factor 2, Interacting Protein [TERF2IP]) as an inhibitor kinase (IKK) adaptor that modulates NFjB in a dose-dependent manner.16–18 The RAP1-mutant mice are resistant to endotoxic shock and show defective NFjB signaling in response to lipopolysaccharide (LPS) stimulation. This results in reduced upregulation of inflammatory cytokines, such as TNFa and IL-1.16 Corneal repair following alkali injury is closely linked
Copyright 2015 The Association for Research in Vision and Ophthalmology, Inc. www.iovs.org j ISSN: 1552-5783
711
T
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
IOVS j February 2015 j Vol. 56 j No. 2 j 712
FIGURE 1. The effects of deletion of RAP1 on edema and corneal resurfacing post alkali burn. (A) Slit-lamp biomicroscopy was performed at various time points before and after injury (days 0, 3, 7, and 14) for bright field images (upper) and fluorescein images (lower) for RWT and RKO groups. Corneal defects as visualized by staining with fluorescein, cornea defects in green shows the extent of the ocular surface injury in the RWT and RKO groups. Representative photographs of corneal defects in a central corneal injury model for RWT and RKO were shown. Bright field images of (i–v) RWT and (x–xiv) RKO at various time points. The severity of corneal defeat in RKO group was less than that in RWT group. (edema, arrow; *opacity; RWT, N ¼ 4; RKO, N ¼ 4). Scale bars: 2 mm. Green fluorescein staining for corneal defeat in (vi–ix) RWT and (xv–xviii) RKO groups at various time
IOVS j February 2015 j Vol. 56 j No. 2 j 713
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
points. These fluorescein images are corresponding to the bright images. Corneal defeat recovery was significantly faster in the RKO group than RWT group. (B) The percentages of corneal defeat retained post alkali burn were calculated by the area defeat of a particular time point divided by the area of corneal defeat on postinjury day 0 (immediately after the burn was induced). The higher the percentage of defect retained in the particular time point, the slower the corneal resurfacing it means. No defects retained early on day 7 in RKO group, while RWT mice had an average of 67.69% 6 3.39% on day 7 (RWT, N ¼ 4; RKO, N ¼ 4).
to activation of NFjB-mediated inflammatory cytokines14,15 and angiogenic cytokines.19,20 We, therefore, set out to determine whether modulation of NFjB signaling by inhibition of RAP1 could promote corneal recovery following injury by regulating inflammatory cytokines and angiogenic factors. In this study, we compared RAP1 wild-type (RWT) and RAP1 knock-out (RKO) mice to investigate the effects on corneal wound healing following alkali injury, and expression levels of inflammatory cytokines/angiogenic factors after injury. Compared to wild type mice, RAP1-deficient mice displayed better corneal recovery following alkali injury that was accompanied by reduced upregulation of inflammatory factor (IL1A), and fine regulated levels of angiogenic factor (VEGF) as well as that of antiangiogenic factor (PEDF).
MATERIALS
AND
METHODS
Induction of Central Corneal Alkali Injury in RWT and RKO Mice All animal experiments were performed in accordance with relevant guidelines and regulations by the University of Hong Kong, and approved by the Committee on the Use of Live Animals in Teaching and Research (CULTAR). We declare our animal protocol is in compliance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. More details on mouse line16 and genotyping are described in the Supplementary Material. To compare the effects of RAP1 gene deletion on the healing of the cornea following alkali injury, a similar alkali burn was administered to the corneas of RWT and RKO mice. We assessed wound healing by corneal resurfacing and angiogenic response. Under general anesthesia, an ocular surface alkali burn was inflicted to both eyes of adult RWT (N ¼ 64) and RKO (N ¼ 64) mice by application of 3 lL 1 N sodium hydroxide solution to a confined circular area of approximately 2-mm diameter in the central corneal region.14
the corneal surface were performed after injury with fluorescein stain. This was monitored with the slit-lamp system. Eyes also were collected (at baseline, and 6 hours, and 1 and 14 days after injury; N ¼ 32) for total mRNA extraction and subsequent realtime PCR studies of inflammatory cytokines. For angiogenesis response, bright view microphotography examination was done. Angiogenic response was monitored at various postinjury time points (days 0, 7, and 14; N ¼ 8) with a light microscope (Nikon SM2800; Nikon, Tokyo, Japan) and photographed (Catalog No. Infinity 1-3C; Chinetek Scientific, Hong Kong, China). Corneas were dissected from the eye at each postinjury time point (days 3, 7, and 14; N ¼ 24) and flatmounted for platelet endothelial cell adhesion molecule (PECAM, Catalog No. sc18916; Santa Cruz Biotechnology, Santa Cruz, CA, USA) fluorescence staining. Eyes also were collected for total mRNA extraction and subsequent real-time PCR study of angiogenic cytokines. The Table shows the accession numbers and primers for mouse cytokines expression by real-time RT-PCR analysis. The PECAM immunofluorescence staining and imaging were performed for whole cornea and magnified parts of a particular cornea were monitored. More details are described in the Supplementary Material.
Wound Healing Scratch Assay and LPS Challenge in RWT and RKO Corneal Epithelial Monolayer Cultures Briefly, when cell culture reached confluence, a classic wound healing scratch assay (also called migration assay) was performed and different time points of wound recovery were recorded in addition to observed LPS challenge for wound healing. More details are described in the Supplementary Material.
Statistical Analysis A P value of 3 0
Reverse Primers 5 0 –>3 0
Inflammatory cytokine mIL1A
GACAGTATCAGCAACGTCAA
TCACTCTGGTAGGTGTAAGG
NM_001025250.3,NM_009505.4, NM_001025257.3 NM011340.3
CCCGACGAGATAGAGTACAT CCCGACTTCAGCAAGATTAC
CTGGCTTTGGTGAGGTTT CCCTCAGAACAAAGAGGAAAG
NM008084.2
CCACTCACGGCAAATTCA
AGTAGACTCCACGACATACTC
NM010554.4
Angiogenesis cytokines mVEGF mPEDF Housekeeping mGAPDH
The Table shows the accession numbers, oligonucleotide sequences of forward and reverse primers of selected candidates of inflammatory (IL1A), and angiogenesis cytokines (VEGF and PEDF) together with the house keeping probe (GAPDH). m, mouse.
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
IOVS j February 2015 j Vol. 56 j No. 2 j 714
FIGURE 2. Effects of deletion of RAP1 on healing of scratched wounds. (A) The RWT and RKO corneal epithelial cells cultured in 12-well plates were injured with a sterile 10-lL pipette tip at 0 hour. The wounded culture was allowed to re-epithelialize. The percentages of area change of a particular time point in reference to the area of the initial time point (0 hour; immediately after scratch) was monitored over 24 hours at 378C. Representative culture of the RWT at times 0, 6, 12, and 24 hours was displayed on the left, while the RKO culture at the same time points was displayed on the right. The RKO group exhibited significantly faster wound recovery. Compared to RWT, more rapid wound healing was observed in RKO mice at 24 hours to a nearly full recovery (RWT, N ¼ 2; RKO, N ¼ 2) Scale bars: 200 lm. (B) Effect of deletion of RAP1 on cell migration in
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
IOVS j February 2015 j Vol. 56 j No. 2 j 715
(A) is presented in a graph. The percentages of cells retained in the wound region are presented; the percentage of retained wound (RWT or RKO) was defined as the area of wound retained at the particular time point (6, 12, or 24 hours) divided by the area of the original wound at 0 hour in the microscopic view. RAP1 wild-type cell culture had 25.48% 6 6.37% wound retained, while RKO cell culture had 5.06% 6 0.93% wound retained on 24 hours (RWT, N ¼ 2; RKO, N ¼ 2).
RKO mice. Their genotypes were confirmed (Supplementary Fig. S1) as described previously.16 Slit-lamp biomicroscopy was performed at various time points before and after injury (days 0, 3, 7, and 14) for bright field images and fluorescein images (Figs. 1A, 1B). A slit-beam illumination was used to investigate the degree of edema and corneal opacity. Bright field images were captured. The results showed that corneal recovery was faster in RKO than RWT mice. Edema (Fig. 1A, arrow; upper panel of RWT) was very obvious on days 3 and 7 in RWT mice. No edema, but corneal opacity (Fig. 1A, asterisk; upper panel of RWT) was observed on day 14. In contrast, little edema (Fig. 1A, arrow; upper panel of RKO) was observed on days 3 and 7 in RKO mice. The corneas in these mice became transparent on day 14 (Fig. 1A, upper panel of RKO). Green fluorescein staining was used to evaluate corneal surface healing as determined by area of corneal defect14 which stained green. Corneal defeat is taken as 100% on postinjury day 0 (Fig. 1). The RWT and RKO mice were monitored for 14 days (Figs. 1A, 1B). Corneal defect recovery was significantly faster in the RKO group than the RWT group. In the RWT group, the corneal surface stained positive on days 3 and 7; negative fluorescein staining was observed on day 14 (Fig. 1A, lower panel in RWT). In contrast, the corneas of the RKO group were less positive to fluorescein staining on day 3, and negative on days 7 and 14 (Fig. 1A, lower panel in RKO). The percentages of epithelial defect retained in RKO mice were significantly reduced compared to RWT mice (Fig. 1B). The retained percentages for RWT and RKO were 100% and 100% on day 0, 88.60% 6 4.13% and 50.96% 6 6.23% on day 3 (*P < 0.01), 67.69% 6 3.39% and 0% on day 7 (*P < 0.01), and 0% and 0% on day 14 (Fig. 1B).
Deletion of RAP1 Enhances Cell Migration of Corneal Epithelial Cells Cells cultured from RWT and RKO mice stained positive in response to the corneal epithelial marker K3, but negative to the conjunctival epithelial cell marker K19 (Supplementary Fig. S2). This confirmed their identities as corneal epithelial cells. An in vitro scratch wound healing assay was used to assess the role of RAP1 deletion in migration.3,21,22 Confluent dishes of corneal epithelial cells isolated from corneas of RWT and RKO mice were wounded by scraping with a 10-lL pipette tip (Tip One; USA Scientific, Inc., Ocala, FL, USA), creating a space free of cells (Figs. 2A, 2B). The rate of wound closure was faster in RKO mice than RWT mice, evident 24 hours after scraping. The percentages retained for RWT and RKO were: 100% and 100% at 0 hours, 86.47% 6 19.13% and 56.71% 6 4.43% at 6 hours, 65.63% 6 9.32% and 43.52% 6 0.76% at 12 hours, and 25.48% 6 6.37% and 5.06 6 0.93% at 24 hours. Deletion of RAP1 facilitates migration of corneal epithelial cells upon wound scraping.
Absence of RAP1 Reduced LPS-Inhibited Wound Healing Lipopolysaccharides (LPS) provide a potent inflammatory stimulus and their presence has major effects on the determinants of inflammation and apoptosis.23 To evaluate the effects of RAP1 deletion on response to LPS insult, corneal
epithelial cell cultures were challenged with (100 ng/mL) LPS24,25 for 24 hours following wound scratching.24,25 The results revealed that RKO mice were more tolerant than RWT to LPS stimulation and demonstrated a faster wound healing response evident 24 hours after scraping (Fig. 3A). The percentages retained for RWT and RKO were 100% and 100% at 0 hours, 81.77% 6 13.91% and 87.33% 6 4.70% after 6 hours, 69.81% 6 3.05% and 66.58% 6 3.71% after 12 hours, and 66.63% 6 4.73% and 0% (*P < 0.01) after 24 hours (Fig. 3B). Cells in the RKO group had a superior recovery response to LPS stimulation than RWT.
Absence of RAP1 Reduced Inflammatory Cytokine IL1A After Alkali Injury The IL1A is known to be a master gene that upregulates inflammation and various cellular responses in corneal recovery.26 Expression of IL1A is upregulated and peaks at 6/ 24 hours after alkali injury,27,28 in part via the NFkB/RAP1 signaling pathway. We compared the IL1A expression dynamics in alkali-burned corneas in RWT to RKO mice, at each of the time points by real-time RT-PCR; the relative level of expression for RKO was presented as the ratio of RKO against RWT. The relative expression of RKO compared to RWT at the pre-injury time point was taken as the baseline (Fig. 4). Expression of IL1A was markedly less upregulated in the RKO group compared to the RWT group (Fig. 4, *P < 0.01). In RKO mice, IL1A expression was 8.33-fold lower after 6 hours compared to RWT mice following corneal alkali injury (*P < 0.01) and 4-fold lower after 24 hours (*P < 0.01). On recovery day 14 (1.23fold), in RKO, IL1A level returned to baseline level compared to RWT (1.34-fold, Fig. 4). Collectively, RAP1 deletion could finely regulate the expression of the inflammatory cytokine, IL1A at 6 and 24 hours after injury.
Reduced Angiogenic Response to Alkali Injury in Corneas of RKO Mice Alkaline was administered to the corneas of RWT and RKO mice to examine the effects of deletion of RAP1 (RKO) on corneal neovascularization. Bright view microscopy was performed over 14 days (after injury days 0, 7, and 14). The total area of thick and thin blood vessels was observed and measured for four corneas from either RWT or RKO mice on days 0, 7, and 14 after injury (Fig. 5A). A weaker corneal angiogenic response was observed in the RKO group compared to RWT. Corneal angiogenesis gradually increased from days 7 to 14 in the RWT group, but reduced in the RKO group at the same time points (Fig. 5A). The mean areas of corneal neovascularization of RWT and RKO were: 1.27 6 0.22 and 0.40 6 0.09 mm2 on day 7 (*P < 0.01), and 1.49 6 0.24 and 0.13 6 0.024 mm2 on day 14 (**P < 0.001, Fig. 5B). Thus, corneal neovascularization was significantly reduced by deletion of RAP1 (Fig. 5B).
Reduced Endothelial Vessels (PECAM-Positive) Sprouting From the Limbus to the Center of the Cornea in RKO Mice To confirm the angiogenic response in RWT and RKO mice corneas, corneas were stained with an endothelial cell marker, PECAM, to identify new vessels sprouting from the limbus
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
IOVS j February 2015 j Vol. 56 j No. 2 j 716
FIGURE 3. Effects of deletion of RAP1 on wound healing upon LPS stimulation. (A) The RWT and RKO corneal epithelial cells cultured in 12-well plates were injured with sterile 10-lL pipette tips at 0 hours, and cultured with (100 ng/mL) LPS supplement cultured medium. The wounded culture was allowed to re-epithelialize for 2 4hours at 378C. The percentages of area change of a particular time point in reference to the area of the initial time point (0 hour; immediately after scratch) was monitored over 24 hours at 378C. Representative culture of the RWT at time 0, 6, 12, and 24 hours was displayed on the left, while the representative RKO culture at the same time points was displayed on the right. Compared to RWT, more rapid wound healing was observed in RKO mice at 24 hours to a full recovery (RWT, N ¼ 2; RKO, N ¼ 2) Scale bars: 200 lm. (B) Effect of
IOVS j February 2015 j Vol. 56 j No. 2 j 717
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
deletion of RAP1 on LPS-challenge in (A) was presented in a graph. The percentages of cells retained in wound regions are represented for the time points. The percentage of retained wound (RWT or RKO) was defined as the area of wound retained at the particular time point (6, 12, or 24 hours) divided by the area of the original wound at 0 hours in the microscopic view. The RWT group had an average of 66.63% 6 4.73% wound retained on 24 hours (RWT, N ¼ 2; RKO, N ¼ 2; *P < 0.01).
region. These images have been merged to reconstruct a full view of a particular cornea (Figs. 6A, 6B). The total areas of PECAM-positive corneal vessels were measured from 4 corneas each of RWT and RKO mice and at each time point. Fewer PECAM-positive vessels were observed in RKO mice on day 7 (Fig. 6A) and day 14 (Fig. 6B) compared to RWT. In contrast, more PECAM-positive stained blood vessels were observed in the corneas of RWT mice at the same time points (Figs. 6A, 6B). The mean areas for RWT and RKO mice were 3.61 6 0.18 and 1.55 6 0.29 mm2 on day 7 (*P < 0.01, and 8.42 6 0.86 and 0.10 6 0.01 mm2 on day 14 (*P < 0.01, Fig. 6C). Neovascularization in RKO mice was significantly reduced compared to RWT at various time points following alkali injury. Magnified images (Fig. 6D, ~2.4 mm2 of the actual size from the cornea around the limbus) of corneas were captured from RWT and RKO mice. Representative images are presented, extracted from their corresponding images, in Figures 6A and 6B (asterisks). Progression of PECAM-positive stained blood vessels was observed in the corneas of RWT mice on days 3, 7, and 14 (Figs. 6D, 6E). Fewer PECAM-positive vessels were observed in RKO than RWT: PECAM-positive vessels were significantly reduced on day 14 (Figs. 6D, 6E). The mean areas of the PECAM-positive stain in the magnified images are shown in Figure 6E. The areas in RWT and RKO mice were 0.27 6 0.03 and 0.076 6 0.01 mm2 on day 3 (**P < 0.001), 0.63 6 0.1 and 0.31 6 0.08 mm2 on day 7 (*P 0.01, and 1.92 6 0.09 and 0.018 6 0.01 mm2 on day 14, respectively (**P < 0.001, Fig. 6E). Collectively, these data demonstrated that deficiency in RAP1 can reduce the angiogenic response to corneal alkali injury.
Fine Regulation of Angiogenic and Anti-Angiogenic Factors in RKO Mice Corneas Following Alkali Injury To evaluate the effects of RKO on angiogenesis responses in alkali-burned corneas, we performed real time RT-PCR on the angiogenic cytokines, VEGF, and PEDF (Fig. 7). Expression of
RKO was compared to RWT at each time point; the relative level of expression for RKO was presented as the ratio of RKO against RWT. The relative expression of RKO compared to RWT on the pre-injury time point was taken as the baseline (Figs. 7A, 7B). Compared to corneas of RWT mice, real time RT-PCR showed marked relatively downregulation (2.63-fold, *P < 0.01) of PEDF mRNA expression in RKO corneas 6 hours after injury, but relatively upregulation (2.15-fold, *P < 0.01) of VEGF in RKO. In addition, when PEDF expression was relative upregulated 2.20-fold (*P < 0.01) in RKO corneas on day 7, VEGF levels were relative downregulated 1.79-fold (*P < 0.01). Level of cytokine expression (PEDF and VEGF) returned toward normal (baseline) upon recovery on day 14 (1.41and 0.97-fold respectively) in RWT and RKO groups (Figs. 7A, 7B). These data demonstrates that deletion of RAP1 finely controls the expression levels of angiogenic and antiangiogenic factors for better corneal injury and recovery process compared to RWT.
DISCUSSION Alkali seriously damages the cornea and causes loss of vision in humans, and provokes inflammatory and angiogenic responses that result in a release of inflammatory cytokines and angiogenic factors, activation of NFjb pathway,14 and a cellular cascade including cell proliferation, and cell apoptosis in the corresponding corneal26 and endothelial cells.20,29 Prolonged activation of NFjB has a key role in aberrant inflammatory cytokine expression and neovascularization,12,13 and often results in pathogenic corneal wound healing with consequent vision loss.30 In alkali injury, inhibition of NFjB that can reduce corneal angiogenesis may serve as a therapeutic strategy.15 The NFjB inhibitors have shown anti-inflammatory effects and include emodin, besifloxacin, BOL-303242-X (mapracorat), thymosinb4, epigallocatechin gallate, Perilla frutescens leaf extract, and IKKb-targeting short interfering RNA.10,31 Nevertheless inhibi-
FIGURE 4. The effect of deletion of RAP1 on the relative mRNA expression level of IL1A of RKO compared to RWT in the corneas post alkali burn. The relative expression of IL1A in corneas was measured at various time points for RKO compared to RWT. The relative expression level at a particular time point is calculated as a ratio of expression level of the RKO group over which of the RWT group. The relative expression of RKO to RWT at time point pre-injury was taken as the baseline level. Compared to RWT corneas, the relative expression level of IL1A was significantly less upregulated in RKO group at 6 hours (~8.33-fold, *P < 0.01), day 1 (~4.07-fold, *P < 0.01), and returns to baseline on day 14 (~1.23-fold) after alkali injury. (RWT, N ¼ 16; RKO, N ¼ 16).
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
IOVS j February 2015 j Vol. 56 j No. 2 j 718
FIGURE 5. The effects of deletion of RAP1 on angiogenesis response after alkali injury. (A) Representative photographs of corneal neovascularization in a central corneal injury model. Dorsal and lateral images were presented for RWT and RKO corneas on days 0, 7, and 14 after injury time points. The RWT and RKO images are illustrated on the left and the right columns, respectively. The severity of corneal neovascularization in the RKO group was less than that in the RWT group (neovascularization was marked by arrows; RWT, N ¼ 4; RKO, N ¼ 4). Scale bars: 2 mm. (B) The mean total areas of new vessels were graphically presented. The total areas of new vessels from each cornea from the RWT and RKO groups were separately quantified by ImageJ (http://imagej.nih.gov/ij/; provided in the public domain by the National Institutes of Health, Bethesda, MD, USA) for various time points (days 7 and14). A significant decrease of angiogenesis response was observed in RKO group on day 7 (RWT and RKO; 1.27 6 0.22 and 0.40 6 0.09 mm2; *P < 0.01) and day 14 (RWT and RKO; 1.49 6 0.24 and 0.13 6 0.024 mm2; **P < 0.001) compared to RWT group (RWT, N ¼ 4; RKO, N ¼ 4).
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
IOVS j February 2015 j Vol. 56 j No. 2 j 719
FIGURE 6. (A–E) Deletion of RAP1 inhibits angiogenesis in the mouse cornea after alkali injury. (A) Alkali injury was administered on the central corneal region for RWT and RKO group mice in vivo. Corneas were dissected and collected on the various time points (days 7 and 14) for whole mount and stained with endothelial marker, PECAM. Representative PECAM-positive images of whole mount corneas of RWT (left) and RKO (right) on day 7 after injury (RWT, N ¼ 4; RKO, N ¼ 4). Scale bars: 1 mm. (B) Representative PECAM-positive images of whole mount corneas of RWT (left) and RKO (right) on day 14. Neovascularization was inhibited in PECAM-stained corneas from RKO on (A) day 7 and (B) day 14, compared to RWT corneas of the same time point (RWT, N ¼ 4; RKO, N ¼ 4). Scale bars: 1 mm. (C) Quantification of mean total areas of blood vessel sprouting in RWT and RKO mice. Total area of the PECAM-positive vessels very quantified for individual RWT and RKO cornea. Mean areas for RWT and RKO groups were shown in the graph. The sprouting of blood vessels was significantly reduced in RKO mice compared to RWT mice, less blood vessel sprouting in RKO at day 7 (RWT and RKO; 3.61 6 0.18 and 1.55 6 0.29 mm2, *P < 0.01) and day 14 (RWT and RKO and 8.42 6 0.86 and 0.10 6 0.01 mm2; *P < 0.01; RWT, N ¼ 8; RKO, N ¼ 8). (D) Magnified images of PECAM-positive vessels from the corneas (actual size 2.4 mm2 surrounding the limbus region). Representative photographs were extracted from sections marked with asterisks from Figures 6A and 6B on the various time points (days 3, 7, and 14); RWT (left column) and RKO (right column). Neovascularization was inhibited in PECAM-stained corneas from RKO on days 3, 7, and 14 (RWT, N ¼ 12; RKO, N ¼ 12; Scale bars: 1 mm) when compared to RWT. (E) Quantification of blood vessel sprouting in RWT and RKO mice on magnified (*) images from Figures 6A and 6B. The sprouting of blood vessels was significantly reduced in RKO mice compared to RWT mice, less blood vessel was sprouting in RKO in day 3 (RWT and RKO; 0.27 6 0.02 and 0.08 6 0.01 mm2 ; **P < 0.001), blood vessel sprouting in RKO in day 7 (RWT and RKO; 0.63 6 0.06 and 0.31 6 0.05 mm2; *P 0.01), and day 14 (RWT and RKO; 1.92 6 0.05 and 0.02 6 0.003 mm2; **P < 0.001; RWT, N ¼ 12; RKO, N ¼ 12).
tion of NFjB has shown variable results and its multifaceted and cell type specificity6,15,32–37 is associated with the fine control of healing of epithelial cells.38 A therapeutic strategy of fine modulation of NFjB signaling rather than complete blockage of NFjB, therefore, is important to preserve the autoregulatory, anti-inflammatory, and other restorative mechanisms and readjust these multiple interrelated processes.39,40 Most common NFjkB dimer is composed of p50 and p65 subunit and this is activated in many cancers. The SN50
peptide blocks this dimer and, hence, NFjkB activation response. This is a very generic mechanism of blocking NFjKB, much like inhibiting the upstream kinases, which, though effective in in vitro settings, can lead to many toxic effects in vivo.41 The Rap1 was identified as a telomeric protein and its role in inflammation is considered its noncanonical function. The Rap1 activated NFjkB by making it a strong chromatin remodeler. The NFjkB p65 subunit is modified by phosphorylation induced by IKKs, which are kinases that also
IOVS j February 2015 j Vol. 56 j No. 2 j 720
RAP1-NFjB Pathway Facilitates Corneal Wound Healing
stress resistance,16 and is associated with fine regulation of inflammatory cytokines, angiogenic factors, and antiangiogenic factors. The IL1A is a prime proinflammatory cytokine arising from NFjB activation44 and the first to be released following injury for repair, proliferation, apoptosis, motility, differentiation, and necrosis in various cell types, including epithelial cells.26,45 Timely and tightly controlled angiogenic VEGF and antiangiogenic PEDF factors have important roles in corneal repairs following injuries46,47 that are associated with NFjB signaling regulation.20,29 This study provides in vivo evidence that a reduced RAP1 level promotes corneal repair following alkali-induced damage, and is associated with a reduced upregulation of inflammatory and angiogenic cytokines. Inhibition of RAP1 enhances corneal recovery and may serve as a novel therapeutic target that can fine regulate NFjB signaling-mediated inflammatory and angiogenic responses.
Acknowledgments
FIGURE 7. The effects of deletion of RAP1 on the relative mRNA expression of (A) PEDF and (B) VEGF in RKO corneas compared to RWT corneas after alkali injury. The relative expression of (A) PEDF and (B) VEGF in corneas was measured at various time points (preinjury, post injury 6 hours, and days 3, 7, and 14) for RKO compared to RWT. The relative expression at a particular time point is calculated as a ratio of expression of RKO over that of RWT at a particular time point. The relative expression of RKO to RWT on preinjury time point was taken as the baseline level. Compared to RWT, at 6 hours after injury in the RKO group, the relative expression level of PEDF (~2.63-fold, *P < 0.01) was significantly reduced, while that of VEGF was significantly upregulated (~2.15-fold, *P < 0.01). On day 7 post injury in RKO mice, dynamic changes to the relative expression levels of VEGF (~ 1.79-fold downregulated, *P < 0.01) and PEDF (~2.20-fold upregulated, *P < 0.01) were observed. (RWT, N ¼ 20; RKO, N ¼ 20).
mark IkB proteins for degradation. Only p65 that is modified by IKK in presence of Rap1 is capable of binding p300/CBP and open chromatin on specific promoters. Hence, inhibiting Rap1 is much more specific than generally inhibiting NFjkB by SN50, which will block NFjkB binding to all sites. Such specificity of Rap1 inhibition will lead to design of much more specific inhibitors of NFjkB and inflammation. This is apparent in many in vivo assays presented here and in the fact that Rap1 null mice are normal until challenged by stress-inducing agents. Like Rap1, the role of telomerase42 in regulating inflammation is now well appreciated and targeting telomerase is considered a worthwhile therapy.43 In this study we have shown that RAP1 is an effective modulator of NFjB signaling that, in turn, is required for cytokine regulation and subsequent fine healing of corneal injury. Our results demonstrated that inhibition of NFjB by inhibition of RAP1 facilitates corneal healing. Corneal resurfacing was associated with reduced upregulation of inflammatory cytokines, while in corneal angiogenesis, repressed induction of angiogenesis was associated with reduced upregulation of angiogenic factors, but increased upregulation of antiangiogenic factors. Inhibition of RAP1 facilitates cell migration in vitro and increases tolerance to bacterial challenge. Our findings are consistent with those described previously that inhibition of NFjB signaling facilitates corneal recovery, corneal resurfacing,15 reduced angiogenesis,14 and
Supported in part by grants from the RGC Research Grant (N_HKU 716/09, HKU3/CRF/11R, and T12-705/11 to [QL]); National Natural Science Foundation of China (31270967 [QL]); small project funding from The University of Hong Kong (201111159183 and 201211159030 [QL], and 201109176184 and 201309176129 [MWP]); and a core grant to IMCB from A*star (VT). Disclosure: M.W. Poon, None; L. Yan, None; D. Jiang, None; P. Qin, None; H.-F. Tse, None; I.Y. Wong, None; D.S.H. Wong, None; V. Tergaonkar, None; Q. Lian, None
References 1. Saika S, Kobata S, Hashizume N, Okada Y, Yamanaka O. Epithelial basement membrane in alkali-burned corneas in rats. Immunohistochemical study. Cornea. 1993;12:383–390. 2. Ishizaki M, Zhu G, Haseba T, Shafer SS, Kao WW. Expression of collagen I, smooth muscle alpha-actin, and vimentin during the healing of alkali-burned and lacerated corneas. Invest Ophthalmol Vis Sci. 1993;34:3320–3328. 3. Tong L, Tergaonkar V. Rho protein GTPases and their interactions with NFkappaB: crossroads of inflammation and matrix biology. Biosci Rep. 2014;34:e00115. 4. Ben-Neriah Y, Karin M. Inflammation meets cancer, with NFkappaB as the matchmaker. Nat Immunol. 2011;12:715–723. 5. Tabruyn SP, Griffioen AW. Molecular pathways of angiogenesis inhibition. Biochem Biophys Res Comm. 2007;355:1–5. 6. Tak PP, Firestein GS. NF-kappaB: a key role in inflammatory diseases. J Clin Invest. 2001;107:7–11. 7. Nakao S, Hata Y, Miura M, et al. Dexamethasone inhibits interleukin-1beta-induced corneal neovascularization: role of nuclear factor-kappaB-activated stromal cells in inflammatory angiogenesis. Am J Path. 2007;171:1058–1065. 8. Kubota M, Shimmura S, Kubota S, et al. Hydrogen and N-acetylL-cysteine rescue oxidative stress-induced angiogenesis in a mouse corneal alkali-burn model. Invest Ophthalmol Vis Sci. 2011;52:427–433. 9. Wang CY, Mayo MW, Korneluk RG, Goeddel DV, Baldwin AS Jr. NF-kappaB antiapoptosis: induction of TRAF1 and TRAF2 and c-IAP1 and c-IAP2 to suppress caspase-8 activation. Science. 1998;281:1680–1683. 10. Lan W, Petznick A, Heryati S, Rifada M, Tong L. Nuclear factorkappaB: central regulator in ocular surface inflammation and diseases. Ocular Surf. 2012;10:137–148. 11. Gordon JW, Shaw JA, Kirshenbaum LA. Multiple facets of NFkappaB in the heart: to be or not to NF-kappaB. Circ Res. 2011;108:1122–1132.
RAP1-NFjB Pathway Facilitates Corneal Wound Healing 12. Cooks T, Pateras IS, Tarcic O, et al. Mutant p53 prolongs NFkappaB activation and promotes chronic inflammation and inflammation-associated colorectal cancer. Cancer Cell. 2013; 23:634–646. 13. Kisseleva T, Song L, Vorontchikhina M, Feirt N, Kitajewski J, Schindler C. NF-kappaB regulation of endothelial cell function during LPS-induced toxemia and cancer. J Clin Invest. 2006; 116:2955–2963. 14. Saika S, Ikeda K, Yamanaka O, et al. Expression of Smad7 in mouse eyes accelerates healing of corneal tissue after exposure to alkali. Am J Path. 2005;166:1405–1418. 15. Saika S, Miyamoto T, Yamanaka O, et al. Therapeutic effect of topical administration of SN50, an inhibitor of nuclear factorkappaB, in treatment of corneal alkali burns in mice. Am J Path. 2005;166:1393–1403. 16. Teo H, Ghosh S, Luesch H, et al. Telomere-independent Rap1 is an IKK adaptor and regulates NF-kappaB-dependent gene expression. Nature Cell Biol. 2010;12:758–767. 17. Ghosh AS, Tergaonkar V. Telomeres and inflammation: Rap1 joins the ends? Cell Cycle. 2010;9:3834–3835. 18. Ghosh A, Saginc G, Leow SC, et al. Telomerase directly regulates NF-kappaB-dependent transcription. Nat Cell Biol. 2012;14:1270–1281. 19. Tong Q, Zheng L, Lin L, et al. VEGF is upregulated by hypoxiainduced mitogenic factor via the PI-3K/Akt-NF-kappaB signaling pathway. Resp Res. 2006;7:37. 20. Keskin D, Kalluri R. NF-kappaB-induced chromatin remodeling regulates angiogenesis. Blood. 2010;116:312–313. 21. Wells A. Tumor invasion: role of growth factor-induced cell motility. Adv Cancer Res. 2000;78:31–101. 22. Xu KP, Riggs A, Ding Y, Yu FS. Role of ErbB2 in corneal epithelial wound healing. Invest Ophthalmol Vis Sci. 2004;45: 4277–4283. 23. Liang H, Brignole-Baudouin F, Labbe A, Pauly A, Warnet JM, Baudouin C. LPS-stimulated inflammation and apoptosis in corneal injury models. Mol Vis. 2007;13:1169–1180. 24. Karlstetter M, Lippe E, Walczak Y, et al. Curcumin is a potent modulator of microglial gene expression and migration. J Neuroinflamm. 2011;8:125. 25. Kim SY, Lee JG, Cho WS, et al. Role of NADPH oxidase-2 in lipopolysaccharide-induced matrix metalloproteinase expression and cell migration. Immunol Cell Biol. 2010;88:197–204. 26. Wilson SE, Mohan RR, Ambrosio R Jr, Hong J, Lee J. The corneal wound healing response: cytokine-mediated interaction of the epithelium, stroma, and inflammatory cells. Prog Retin Eye Res. 2001;20:625–637. 27. Lausch RN, Chen SH, Tumpey TM, Su YH, Oakes JE. Early cytokine synthesis in the excised mouse cornea. J Interferon Cytokine Res. 1996;16:35–40. 28. Planck SR, Rich LF, Ansel JC, Huang XN, Rosenbaum JT. Trauma and alkali burns induce distinct patterns of cytokine gene expression in the rat cornea. Ocul Immunol Inflamm. 1997;5:95–100. 29. Aurora AB, Biyashev D, Mirochnik Y, et al. NF-kappaB balances vascular regression and angiogenesis via chromatin remodeling and NFAT displacement. Blood. 2010;116:475–484.
IOVS j February 2015 j Vol. 56 j No. 2 j 721 30. Zhang MC, Wang Y, Yang Y. The expression of nuclear factor kappa B in inflammation-induced rat corneal neovascularization. Ocul Immunol Inflamm. 2006;14:359–365. 31. Sosne G, Qiu P, Christopherson PL, Wheater MK. Thymosin beta 4 suppression of corneal NFkappaB: a potential antiinflammatory pathway. Exp Eye Res. 2007;84:663–669. 32. Seetharaman R, Mora AL, Nabozny G, Boothby M, Chen J. Essential role of T cell NF-kappa B activation in collageninduced arthritis. J Immunol. 1999;163:1577–1583. 33. Baldwin AS Jr. Series introduction: the transcription factor NFkappaB and human disease. J Clin Invest. 2001;107:3–6. 34. Tak PP, Gerlag DM, Aupperle KR, et al. Inhibitor of nuclear factor kappaB kinase beta is a key regulator of synovial inflammation. Arthritis Rheum. 2001;44:1897–1907. 35. Yamamoto Y, Gaynor RB. Therapeutic potential of inhibition of the NF-kappaB pathway in the treatment of inflammation and cancer. J Clin Invest. 2001;107:135–142. 36. Yamamoto Y, Gaynor RB. Role of the NF-kappaB pathway in the pathogenesis of human disease states. Curr Mol Med. 2001;1:287–296. 37. Jimi E, Aoki K, Saito H, et al. Selective inhibition of NF-kappa B blocks osteoclastogenesis and prevents inflammatory bone destruction in vivo. Nat Med. 2004;10:617–624. 38. Zhou R, Hu G, Gong AY, Chen XM. Binding of NF-kappaB p65 subunit to the promoter elements is involved in LPS-induced transactivation of miRNA genes in human biliary epithelial cells. Nucl Acids Res. 2010;38:3222–3232. 39. Krakauer T. Nuclear factor-kappaB: fine-tuning a central integrator of diverse biologic stimuli. Int Rev Immunol. 2008;27:286–292. 40. Beg AA, Sha WC, Bronson RT, Ghosh S, Baltimore D. Embryonic lethality and liver degeneration in mice lacking the RelA component of NF-kappa B. Nature. 1995;376:167– 170. 41. Dey A, Wong E, Kua N, Teo HL, Tergaonkar V, Lane D. Hexamethylene bisacetamide (HMBA) simultaneously targets AKT and MAPK pathway and represses NF kappaB activity: implications for cancer therapy. Cell Cycle. 2008;7:3759–3767. 42. Ghosh S, Karin M. Missing pieces in the NF-kappaB puzzle. Cell. 2002;109(suppl):S81–S96. 43. Li Y, Tergaonkar V. Noncanonical functions of telomerase: implications in telomerase-targeted cancer therapies. Cancer Res. 2014;74:1639–1644. 44. Sheedy FJ, Moore KJ. IL-1 signaling in atherosclerosis: sibling rivalry. Nat Immunol. 2013;14:1030–1032. 45. Papacleovoulou G, Critchley HO, Hillier SG, Mason JI. IL1alpha and IL4 signalling in human ovarian surface epithelial cells. J Endocr. 2011;211:273–283. 46. Azar DT. Corneal angiogenic privilege: angiogenic and antiangiogenic factors in corneal avascularity, vasculogenesis, and wound healing (an American Ophthalmological Society thesis). Trans Am Ophthalmol Soc. 2006;104:264– 302. 47. Pan Z, Fukuoka S, Karagianni N, Guaiquil VH, Rosenblatt MI. Vascular endothelial growth factor promotes anatomical and functional recovery of injured peripheral nerves in the avascular cornea. FASEB J. 2013;27:2756–2767.
