Cell surface engineering using glycosylphosphatidylinositol anchored tissue inhibitor of matrix metalloproteinase-1 stimulates cutaneous wound healing Roghieh Djafarzadeh, PhD1; Claudius Conrad, MD, PhD2; Susan Notohamiprodjo, MD1; Stephanie Hipp1; Hanno Niess, MD3; Christiane J. Bruns, MD, PhD3; Peter J. Nelson, PhD1 1. Medical Clinic and Outpatient Clinic IV, 3. Department of Surgery, University of Munich, Munich, Germany, and 2. Department of Surgery and Harvard Stem Cell Institute, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts
Reprint requests: Prof. P. J. Nelson, Medizinische Klinik und Poliklinik IV, Campus Innenstadt, Ludwig-Maximilians-Universität München, Arbeitsgruppe Klinische Biochemie, Schillerstrasse 42, 80336, Munich, Germany. Tel: +89 218075 845; Fax: +89 218075 860; Email: [email protected] Manuscript received: November 8, 2012 Accepted in final form: October 10, 2013 DOI:10.1111/wrr.12132
ABSTRACT The balance between matrix metalloproteinases and their endogenous tissue inhibitors (TIMPs) is an important component in effective wound healing. The biologic action of these proteins is linked in part to the stoichiometry of TIMP/matrix metalloproteinases/surface protein interactions. We recently described the effect of a glycosylphosphatidylinositol (GPI) anchored version of TIMP-1 on dermal fibroblast biology. Here, cell proliferation assays, in vitro wound healing, electrical wound, and impedance measurements were used to characterize effects of TIMP-1-GPI treatment on primary human epidermal keratinocytes. TIMP-1-GPI stimulated keratinocyte proliferation, as well as mobilization and migration. In parallel, it suppressed the migration and matrix secretion of dermal myofibroblasts, and reduced their secretion of active TGF-β1. Topical application of TIMP-1-GPI in an in vivo excisional wound model increased the rate of wound healing. The agent positively influenced different aspects of wound healing depending on the cell type studied. TIMP-1-GPI counters potential negative effects of overactive myofibroblasts and enhances the mobilization and proliferation of keratinocytes essential for effective wound healing. The application of TIMP-1-GPI represents a novel and practical clinical solution for facilitating healing of difficult wounds.
Wound healing is a complex multistep process in which epidermal keratinocytes and fibroblasts play central roles. Reepithelialization of the wound relies on the mobilization and proliferation of keratinocytes: In parallel, fibroblast/ myofibroblast activation and proliferation within the wound are required for efficient wound healing. Overactivation of fibroblasts can lead to hypertrophic scar formation especially when keratinocyte-driven wound closure is delayed, or when myofibroblasts do not undergo efficient apoptosis.1 The matrix metalloproteinase (MMP) family is a large structurally related enzyme family. The actions of MMPs are linked to the processing of a wide range of extracellular proteins including extracellular matrix (ECM), cytokines, and growth factors. In this regard, these enzymes are thought to help control a diverse spectrum of biologic processes. The 24 MMPs found in humans include collagenases, gelatinases, stromelysins, and membrane-type MMPs.2 MMP activities are regulated at diverse levels including by proteolytic activation of precursor zymogens, and the presence of four endogenous inhibitors, the tissue inhibitors of metalloproteinases (TIMP-1, TIMP-2, TIMP-3, and TIMP-4).3–5 The net balance between MMP/TIMP expression is strongly associated with the turnover of ECM and its remodeling during normal tissue homeostasis and pathogenesis. MMP/TIMP activity also underlies the migratory and remodeling events that drive wound repair. While MMP activity is often elevated in wounds with delayed 70
or aberrant healing, targeting wound healing with small molecular antagonists has proven problematic in part because of the pleiotropic biology of MMPs and their inhibitors.6–8 We have previously shown that the bioactivity of TIMP-1 can be altered by changing its cellular distribution, specifically by focusing expression directly on to the cell surface.9–14 TIMP-1 is normally a secreted protein found within the ECM and only loosely associated with the cell surface through its interaction with other proteins. Modification human TIMP-1 protein by the addition of a glycosylphosphatidylinositol (GPI) anchor gives rise to a recombinant reagent can be transferred from one cell to another through a process called “cell painting” or “cell surface engineering.”15 Exogenously applied TIMP-1-GPI fusion protein is efficiently incorporated into virtually all cell surface membranes and effectively refocuses the actions of TIMP-1directly onto the cell surface. This can give rise to enhanced as well as unique biologic effects. We have recently shown that TIMP-1-GPI can have dramatic effects on the biology of primary human dermal fibroblasts.10 TIMP-1-GPI treatment was found to moderate processes linked to scar formation including reducing fibroblast/myofibroblast proliferation, ECM production, and enhancing their sensitivity to apoptosis.10 Here we have expanded on the characterization of effects of TIMP-1-GPI treatment on general aspects of wound healing and further report that treatment with the agent actively promotes the Wound Rep Reg (2014) 22 70–76 © 2014 by the Wound Healing Society
Djafarzadeh et al.
mobilization and proliferation of primary epidermal keratinocytes. In an in vivo excisional wound healing model, topical application led to enhanced wound closure.
MATERIALS AND METHODS Cellular reagents and tissue culture
Normal human epidermal keratinocytes (NHEK.f-c Promo Cell, Heidelberg, Germany, No. C-12001) were cultured in Keratinocyte Basal Medium (Calcium free) (Promo Cell, No. C-20211). Primary human dermal fibroblast cells (NHDF-c PromoCell, No. C-12300) were cultured in DULBECCO’S MEM with Glutamax-1 (Gibco BRL, Life Technologies GmbH, Eggenstein, Germany, No. 21885-025), supplemented with 10% heat-inactivated fetal calf serum (FCS). Recombinant TIMP-1-GPI protein was produced and purified as previously described.11 All experiments adhere to the Declaration of Helsinki Principles.
TIMP-1-GPI promotes wound healing
migration (ibidi GmbH, No. 80209), essentially a modified version of a scratch assay. A culture insert was used to provide two cell culture reservoirs separated by a 500 μm thick wall. Following confluent keratinocyte culture in both reservoirs, the silicone insert was removed leaving two defined cell patches separated by a zone defined by the original separation wall. Video microscopy was then used to evaluate cell migration over 24 hours. An ibidi Heating & Incubation Systems enabled live cell imaging with incubator conditions directly on the microscope. The recorded events were then quantified using Image J software (NIH freeware: http://rsbweb.nih.gov/ ij/). TIMP-1-GPI treatment lead to accelerated keratinocyte wound healing (see Video S1 provided in Supplement online). Western blot analysis
Western blot was used to detect fibronectin (DakoCytomation, Glostrup, Denmark, No. A0245) and β-actin (Acris, Hiddenhausen, Germany, No. ab8227) in dermal fibroblasts as previously described.9
Proliferation
Primary epidermal keratinocytes or fibroblasts 5 × 103 per 100 μL medium were cultured in 96-well microtiter plates for 24 hours under standard conditions to yield firmly attached and stably growing cells. After discarding supernatants, 50 μL of medium containing native TIMP-1-GPI, buffer or rhTIMP-1 were added and the cells were incubated from 24 to 72 hours for fibroblasts and 48 to 144 hours for keratinocytes, and 50 μL of 1 mg/mL solution of (3-[4,5-dimethylthiazol-2yl]-2,5 diphenyl tetrazolium bromide) assay (MTT) was added. Rapidly dividing cells exhibit high rates of MTT reduction. After 3-hour incubation at 37 °C, formazan crystals were dissolved by additional 100 μL isopropanol and 0.04 N HCl. Absorbance was measured at 570 nm using GENios Plus TECAN ELISA reader (Tecan, Männedorf, Switzerland). For each experiment, at least six wells were analyzed per experimental condition and time point.10 In vitro assays with primary keratinocytes cells
ECIS electrode arrays (8W1E) were obtained from (ibidi GmbH, Munich, Germany, No. 70001). Before starting ECIS measurements, 300 μL of complete medium was placed in each well and was allowed to equilibrate in the incubator for 24 hours. Three hundred microliters of keratinocyte cell suspension (5 × 104 cells per mL) were added to each well, resulting in a final surface concentration of 1.8 × 104 cells per cm2 and a volume of 300 μL of medium. After cell inoculation, the wells were incubated overnight, while attachment and spreading were followed by electrical impedance. Impedance levels were used to verify that confluence was achieved and maintained and that the cells exhibited normal levels of impedance fluctuations, indicative of healthy cell layers.16 The confluent cell layers were incubated for 1 day before wounding. The damaging pulse was delivered at 2.5 V at 40 kHz for 30 seconds. Medium was changed after wounding, and TIMP1-GPI, control rhTIMP-1 and vehicle were added to specific wells. Electrical impedence was then followed in real time for 48 hours. A second assay was used to disrupt an established confluent human primary keratinocyte monolayer using a commercial Wound Rep Reg (2014) 22 70–76 © 2014 by the Wound Healing Society
Fibroblast cell migration assay
Primary fibroblasts were cultured in FCS free medium for 12 hours before the onset of the experiment. The cells were then labeled with Calcein AM (5 μg/mL) (Molecular Probes, Life Technologies, Darmstadt, Germany; No. MW994.87) for 1 hour and detached from the culture flasks with 1.5 mM EDTA in 1 × phosphate-buffered saline (PBS) before migration. The assay medium (FCS free medium with 1% bovine serum albumin (BSA) (Gibco BRL, Life Technologies GmbH, No. 15260-037) was used to dilute the TIMP inhibitors. Six hundred microliters of assay medium containing 10 % FCS or without FCS was placed in the bottom chamber of a BD Fluoro Blok Insert System with a 3 μm pore size (Becton Dickinson Falcon and Company, San Jose, CA, No. 351156) and 50,000 calcein-labeled cells with and without TIMP-1GPI treatment with a final volume of 100 μL were placed in the wells and measured by detection of the fluorescence of cells migrated through the bottom chamber with an SPECTRA FLUOR plus TECAN reader at 485 nm excitation and 515 nm emission. Data represent the mean of n = 3 inserts. TGF-β enzyme-linked immunosorbent assay
Active and total TGF-β concentrations were measured using an enzyme-linked immunosorbent assay (ELISA) kit from (DB Biosciences, San Jose, CA , No. 559119) applied according to the manufacturer’s directions. Excisional wound healing model
C57BL/6 mice were purchased from Charles River Laboratories (Sulzfeld, Germany). The animal studies were approved by and conducted in accordance with the principles of the regulatory agency of the State of Bavaria, Germany (performed by Dr. Ralf Huss, Department of Pathology, University of Munich). The mice were anesthetized with ketavet (100 mg/kg mouse body weight) and xylazin (5 mg/kg mouse body weight) followed premedication with atropine sulfate. 71
1.0
Reepithelialization of cutaneous wounds depends on the proliferation, mobilization, and migration of keratinocytes.1 The effect of exogenously added TIMP-1-GPI on the proliferation of primary human keratinocytes in vitro was determined at 48, 96, and 144 hours. Treatment with increasing levels of TIMP1-GPI (2, 4, 6, 8, 10, 12, and 14 ng/mL) showed a statistically significant dose-dependent increase in proliferation relative to the vehicle control at all time points, while 14 ng/mL of control recombinant TIMP-1, GPI anchor control (phosphatidylserine), or denatured TIMP-1-GPI had no significant effect on proliferation (Figure 1). TIMP-1-GPI promotes keratinocyte migration in vitro
Two in vitro assays were used to model the mobilization and migration events that occur during wound healing. In the first assay, electric cell-substrate impedance was used to monitor keratinocyte mobilization and migration. Keratinocytes were grown on electrodes, and impedance levels were monitored to verify that cellular confluence was achieved and maintained, and that the cells exhibited normal levels of impedance fluctuations, indicative of healthy cell layers16,17 (Figure 2). Then cells in a defined region were then subjected to currents resulting in cell death, and impedance was further monitored to measure the rate of repair of the “wound”(Figure 2). Effects of TIMP-1-GPI treatment were evident at 4 ng/mL TIMP-1-GPI within a few hours of treatment and increased with dose. At 12 ng/mL TIMP-1-GPI, the “wound”/ keratinocyte monolayer was completely reestablished by approximately 12–15 hours (Figure 2). To directly visualize these effects, a second commercial assay system was used to disrupt an established confluent keratinocyte monolayer (ibidi). The method generates of a uniform area devoid of cells, and repopulation of this area in the keratinocyte cell sheet occurs as a consequence to the proliferation and migration of cells from the edge of the keratinocyte monolayer. These events were recorded by video microscopy and quantified using Image J software. TIMP-1GPI treatment again led to an immediate effect on keratinocyte migration/mobility, and complete closure was 72
+
0.8
+
+
0.6 + +
0.4
+ +
+
+ + +
+ +
+ +
+ + ++ + +
+ + +
+ + +
+ + ++ + +
+ ++ ++ +
0.2
ng 10 /mL ng /m L 12 ng /m L 14 ng /m L
/m L
ng
/m L 6
ng
/m L 4
PI
ng 2
8
TIMP-1-GPI
rh
TI
M
P1
Ve h
ng
ic le
/m L
0.0
RESULTS GPI-anchored TIMP-1 increases keratinocyte proliferation
+ +
+ +
+ +
+ +
14
After the dorsal surface of each mouse had been disinfected with an alcohol swab, a full-thickness excisional wound (0.8 cm in diameter) was created on the dorso-medial back of each animal, using a standard skin biopsy punch (Stiefel® Laboratorium GmbH, Offenbach, Germany). Immediately after surgery, as well as on days 3, 5, and 7 thereafter, rhTIMP-1 buffer, or rhTIMP-1, or TIMP-1-GPI was applied on the top of created wound. The measurement of wound size was performed on days 0, 3, 5, 7, and 10. The wound area (mm2) was calculated using the formula of an ellipse; the wound closure rate was expressed as a ratio of the wounded area at each time point divided by the area of the original wound at time 0. On postoperative day 14, the animals were euthanized and the wounds were excised, fixed in 4% neutralbuffered formalin solution, and embedded in paraffin.
Djafarzadeh et al.
Keratinocyte proliferation - MTT
TIMP-1-GPI promotes wound healing
Figure 1. Treatment with GPI-anchored TIMP-1 stimulates keratinocyte cell proliferation. To assess the effect of TIMP-1 surface engineering on proliferation of primary human keratinocytes, MTT assays were performed. TIMP-1-GPItreated keratinocytes show a dose-dependent increase in proliferation (at 48 [fine diagonal lines], 96 [wide diagonal], and 144 [white] hours, reflected in the three columns). Absorbance was then measured at 550 nm using GENios Plus TECAN ELISA reader. For each experiment, at least six wells were analyzed per experimental condition and time point. Each individual treatment result was compared with the corresponding vehicle control using paired Student’s t-test (+p < 0.01; ++p < 0.001; +++p < 0.0001). GPI, glycosylphosphatidylinositol; MTT, (3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide) assay; TIMP, endogenous tissue inhibitor.
seen by 12–13 hours at the highest concentration of TIMP-1GPI (12 ng/mL) tested. By contrast, the control samples were still not closed by the 24 hours time point (see Video S1 provided in online Supplement). The rapidity of the mobilization and migration response seen with TIMP-1-GPI treatment in both models suggests processes independent of the enhanced proliferation shown in Figure 1. The GPI-anchored TIMP-1 suppresses fibroblast migration, proliferation, and matrix production
Fibrosis and scar formation is characterized by fibroblast proliferation, differentiation into myofibroblasts, and excessive production and deposition of ECM.10,18 We recently demonstrated that TIMP-1-GPI treatment could inhibit fibroblast proliferation and lead to a more proapoptotic phenotype.10 The effect of TIMP-1-GPI on the proliferation of activated primary dermal fibroblast cells was verified here in parallel experiments to provide contrast with the epidermal keratinocyte results shown above. In contrast to that seen with keratinocytes, TIMP-1-GPI-treated primary dermal fibroblasts showed a statistically significant dose-dependent Wound Rep Reg (2014) 22 70–76 © 2014 by the Wound Healing Society
Djafarzadeh et al.
TIMP-1-GPI promotes wound healing
+ 0.40 0.35 0.30 0.25
+ + ++ + ++ + + ++ ++ + ++ + ++ + + ++ + + + +++ + + + ++ + ++ + ++ + + ++ ++ +
0.20 0.15 0.10 0.05
+
+ +
+ + + + + +
+
B
Migrated Fibroblasts × 103
TIMP-1-GPI + 0.4 ng/mL TNF-α
50 **** **
TN F Ve -α hi cl e
8 n 10 g/m n L 12 g/m ng L 14 /m ng L /m L
2
ic
le
ng 0 4 /mL ng / 6 mL ng 8 /mL n 10 g/m n L 12 g/m ng L 14 /m ng L /m L
0.00
Ve h
Figure 2. In vitro assay shows increased keratinocyte migration following application of TIMP-1-GPI. An electrical woundhealing assay using an ECIS electrode arrays measured resistive impedance over time and in response to induced damage to a monolayer. Keratinocytes treated with TIMP-1GPI showed rapidly increased impedance and thus enhanced repair of damage. GPI, glycosylphosphatidylinositol; TIMP, endogenous tissue inhibitor.
Fibroblast proliferation - MTT
A
TIMP-1-GPI
rhTIMP-1 14 ng/mL
N.s. N.s.
****
40
Treatment reduces levels of active TGF-β1
Fibroblast recruitment, proliferation, and production of ECM are strongly influenced by the profibrotic growth factor TGFβ.20 TGF-β1 also helps control the growth of keratinocytes.21,22 Specific ELISA was used to assess the production of total vs. active forms of TGF-β1 by primary dermal fibroblasts in serum-free media. TIMP-1-GPI treatment effectively reduced the levels of active TGF-β1 relative to the untreated control, suggesting a direct effect on proteolytic processing of the latent TGF-β1 complex by the surface bound TIMP-1 (TIMP1-GPI at 7 ng/mL p < 0.0002; 14 ng/mL p < 0.0001)10,13,14 (Figure 4). TIMP-1-GPI treatment accelerates excisional wound healing
The effect of TIMP-1-GPI on excisional dermal wound healing was then evaluated using C57BL/6 mice. After disinfection, a defined, full-thickness excisional wound (0.8 cm in Wound Rep Reg (2014) 22 70–76 © 2014 by the Wound Healing Society
**
30
**
20 10
ed
10 % ia 2 FC n 4 g/mS n L 6 g/m n L 8 g/m n 10 g/mL 12 ng/ L m 14 ng/ L ng mL /m L 8 ng 10 /m 12 ng/ L m 14 ng/ L ng mL /m L
0
M
reduction in proliferation (to TNF-α [0.4 ng/ml] [p < 0.01– 0.0001]) (Figure 3A). Addition of as little as 2 ng/mL of TIMP-1-GPI was also found to reduce serum-driven fibroblast migration (p < 0.004),and completely block migration of the cells at 12 ng/mL (p < 0.0001) (Figure 3B). By contrast, recombinant TIMP-1 alone (14 ng/mL) showed reduced effects on fibroblast cell migration. Fibronectin is an important component of the ECM and is secreted by activated fibroblasts.19 Dermal fibroblast cells grown in 15% FCS were treated with 14 ng/mL TIMP-1-GPI or control rhTIMP-1, and Western blot was used to quantify expression of fibronectin in whole cell extracts (Figure 3C). TIMP-1-GPI, but not an equivalent concentration of control recombinant TIMP-1, reduced expression of the glycoprotein. Similar results were previously shown at the steady state mRNA level.10
rhTIMP-1
TIMP-1-GPI
C
[14 ng/mL] Vehicle
Fibronectin β-actin Control
rhTIMP-1
TIMP-1-GPI
230 kDa 41 kDa
Figure 3. TIMP-1-GPI suppresses fibroblast proliferation, migration, and fibronectin production. (A) Exogenously added TIMP-1-GPI protein elicited a dose-dependent decrease in fibroblast proliferation at 24, 48, or 72 hours to cells stimulated with 1% FCS with or without 0.4 ng/mL TNF-α. See legend to Figure 1 for details. Each treatment result was compared with the corresponding TNF-α/FCS control value (e.g., 24, 48, and 72 hours) using pared paired Student’s t-test (+p < 0.01; ++p < 0.001; +++p < 0.0001). (B) TIMP-1-GPI shows a potent suppression of fibroblast migration in response to 10 % FCS. Addition of rhTIMP-1 at the same concentration showed limited effects on fibroblast migration. Each treatment result was compared with the FCS control value using paired Student’s t-test. (*p < 0.01; **p < 0.001; ***p < 0.0001). (C) Confluent fibroblasts were cultured in the presence or absence of TIMP-1-GPI, and fibronectin was quantified in cell extracts by Western blot. rhTIMP-1 at 14 ng/mL did not lead to pronounced decrease in fibronectin expression, while TIMP-1GPI at 14 ng/mL strongly reduced fibronectin levels. GPI, glycosylphosphatidylinositol; TIMP, endogenous tissue inhibitor.
73
TIMP-1-GPI promotes wound healing
Figure 4. Effect of TIMP-1-GPI on TGF-β1 production by activated dermal fibroblasts. Confluent primary dermal fibroblasts were cultured in the presence or absence of TIMP-1-GPI in serum free medium. Total vs. activated levels of TGF-β1 were measured using commercial ELISAs. The active form of TGF-β1 was significantly reduced by treatment with increasing levels of TIMP-1-GPI. Each active TGF-β1 value was compared with the vehicle control using paired Student’s t-test (*p < 0.01; **p < 0.001; ***p < 0.0001). ELISA, enzyme-linked immunosorbent assay; GPI, glycosylphosphatidylinositol; TGFβ1, transforming growth factor beta-1 (TGF-beta 1); TIMP, endogenous tissue inhibitor.
diameter) was created on the dorso-medial back of each animal using a standard skin biopsy punch. Immediately after surgery, as well as on days 3, 5, and 7 thereafter, 200 μL of vehicle (0.025% Triton × 100 in 1 × PBS), rhTIMP-1 (14 ng/ml in vehicle), or TIMP-1-GPI (14 ng/mL in vehicle) was applied to the wound surface (Figure 5A). Measurement of wound size was performed on days 0, 3, 5, 7, and 10. The resulting wound area (mm2) was calculated using the formula of an ellipse and the wound-closure rate expressed as a ratio of the wounded area at each time point, divided by the area of the original wound at time 0. On postoperative day 14, the animals were euthanized. The results show a significantly improved wound closure rate with TIMP-1-GPI treatment. Application of the agent showed an accelerated rate of wound closure as early as day 3 after surgery, under TIMP-1-GPI treatment relative to that achieved by rhTIMP-1 buffer (§p < 0.05) or rhTIMP-1 (**p < 0.05) application (both vs. control). This was evidenced by measurement of wound area (Figure 5).
DISCUSSION Cutaneous wound repair is a complex biologic process. Healing proceeds in a sequential manner that has been divided 74
Djafarzadeh et al.
into four overlapping phases: inflammation, granulation, reepithelialization, and tissue remodeling. The effector cells responsible for driving excisional wound healing include fibroblasts and keratinocytes. Reepithelialization of the wound is a critical phase that depends upon the speed with which keratinocytes can mobilize and effect wound closure. They must advance across the wound and proliferate at the edges. In the granulation phase, fibroblasts migrate into the wound, proliferate, and secrete ECM. When wound closure is not rapid enough, fibroblast/myofibroblast over proliferation within the wound can lead to excessive scar formation. MMP/TIMP biology underlies central aspects of tissue homeostasis and wound repair. The balance between MMPs and TIMPs helps regulate ECM turnover during normal tissue homeostasis and pathogenesis, and also helps control cell signaling through modification of growth factors.22,23 Where the balance between the proteolytic breakdown and deposition of ECM is disturbed, disorders such as abnormal wound healing may result. TIMP-1 is naturally a soluble protein that can be detected on the cell surface only through its interaction with surface bound molecules.24,25 Increase in the local concentration of TIMP proteins by the administration of recombinant protein or gene transfer has shown some efficacy in various animal models. Recombinant TIMP-1-GPI fusion protein is efficiently incorporated into cell surface membranes, and through this, shiftsTIMP-1 directly to the cell surface.9–14,26
Figure 5. Effect of TIMP-1-GPI protein on wound healing in vivo. A full-thickness excisional wound (0.8 cm in diameter) was created on the dorso-medial back of mice using a standard skin biopsy punch. Immediately after surgery, as well as on days 3, 5, and 7 thereafter, rhTIMP-1 buffer, or rhTIMP-1, or TIMP-1-GPI was applied on the top of created wound. After treated with rhTIMP-1 buffer, rhTIMP-1, or TIMP-1-GPI. Wounds were photographed at the times indicated and the volume of the wound area measured (see Materials and Methods section). Wound area is expressed as the mean ± SD. TIMP-1-GPI treatment relative to that achieved by rhTIMP-1 buffer (§p < 0.05) or rhTIMP-1 (**p < 0.05) application. *p < 0.05, #p < 0.05 (both vs. control). The data were analyzed using the paired Student’s t-test with p < 0.05 considered to be significant. GPI, glycosylphosphatidylinositol; TIMP, endogenous tissue inhibitor. Wound Rep Reg (2014) 22 70–76 © 2014 by the Wound Healing Society
Djafarzadeh et al.
This modification to the native protein results in enhanced as well as novel biologic effects. These effects have been previously shown to be independent of the lipid anchor and to require functional TIMP-1 protein.11–13 Here, TIMP-1-GPI treatment was shown to lead to an increase in the proliferation and migration of primary human keratinocytes, but in parallel suppresses the proliferation, migration, and ECM production of activated dermal myofibroblasts/fibroblasts.10 Topical application of the agent on excisional wounds in mice led to accelerated wound closure. While an effect on enhanced wound contraction cannot be ruled out, the in vivo experiments in general support the in vitro wound healing data. Because it is GPI-anchored protein, TIMP-1-GPI cannot penetrate deeply into tissues. Results suggest that only the top cell layers of treated tissues are influenced by topical application. The observation that TIMP-1-GPI treatment reduces fibrosis seems counterintuitive at first view. A general reduction in protease activity should lead to a more pronounced deposition of matrix components because of a reduction in the general degradation in the dynamics of tissue homeostasis (i.e., production balanced by degradation). However, the TIMP-1 proteins and their targets, the MMPs, have much broader biologic activities than ECM turnover. These include effects on cell growth and signal transduction, in part, through their modulation of diverse cytokine activities. I n addition, TIMP proteins have also been shown to have direct growth factor like characteristics that may impact the biology described here. Wound healing is a complex process linked to the balance of regulatory cytokines/growth factors present. Cell surface engineering using GPI-anchored TIMP-1 suppresses processing of growth factors including TGF-β1.10,13,14 TGF-β1 is rapidly induced in response to injury and has some of the broadest effects of any cytokine linked to wound healing.27 TGF-β1 has paradoxical actions in wound healing that are dependent on the cell type, its state of differentiation, and the context of action. While TGF-β1 is a potent inhibitor of keratinocyte proliferation and migration,28 it contributes to the fibrotic process by recruiting fibroblasts and stimulating their synthesis of collagens I, III, and V, proteoglycans, fibronectin, and other ECM components.27,29 Activated fibroblasts are a rich source of latent TGF-β1 in wounds. Active TGF-β1 must be released from its latent complex. TIMP-1-GPI treatment dramatically reduced the processing of the latent TGF-β1 to its active form presumably by a blockade of MMP or related enzymes.10,13,14,27 But it important to point out that TGF-β1 processing may represent only one aspect of the mode of action of this agent in the context of wound healing. Clearly, other soluble factors linked to efficient wound healing can be directly or indirectly influenced by treatment with TIMP-1GPI. Delayed wound healing represents an enormous resource consuming factor in medicine often requiring lengthy and expensive serial debridements, skin grafting, or vacuumassisted wound closure therapy, all of which show limited efficacy. We describe here an agent that shows remarkable efficacy in accelerated wound healing. Topical application with TIMP-1-GPI suppresses processes linked to dermal fibrosis and scarring, and at the same time dramatically enhances wound closure through activation and mobilization of epidermal keratinocytes. Thus, the agent may show efficacy as an adjuvant in excisional wound therapy. Wound Rep Reg (2014) 22 70–76 © 2014 by the Wound Healing Society
TIMP-1-GPI promotes wound healing
ACKNOWLEDGMENT Source of Funding: This work was supported by research grants from the German Research Foundation (DFG) NE 468/4-2. The in vivo wound healing assays were performed under the supervision and ethical license of Dr. Ralf Huss, LMU Pathology. The authors would also like to thank Dr. Ivan Ischenko for excellent technical assistance. Conflict of Interest: PJN holds patents covering the technology detailed here. No other conflicts of interest exist for any of the authors. REFERENCES 1. Broughton G, 2nd, Janis JE, Attinger CE. The basic science of wound healing. Plast Reconstr Surg 2006; 117 (7 Suppl.): 12S– 34S. 2. Fanjul-Fernandez M, Folgueras AR, Cabrera S, Lopez-Otin C. Matrix metalloproteinases: evolution, gene regulation and functional analysis in mouse models. Biochim Biophys Acta 2010; 1803: 3–19. 3. Overall CM, Lopez-Otin C. Strategies for MMP inhibition in cancer: innovations for the post-trial era. Nat Rev Cancer 2002; 2: 657–72. 4. Ra HJ, Parks WC. Control of matrix metalloproteinase catalytic activity. Matrix Biol 2007; 26: 587–96. 5. Moore CS, Crocker SJ. An alternate perspective on the roles of TIMPs and MMPs in pathology. Am J Pathol 2012; 180: 12– 6. 6. Mirastschijski U, Schnabel R, Claes J, Schneider W, Agren MS, Haaksma C et al. Matrix metalloproteinase inhibition delays wound healing and blocks the latent transforming growth factorbeta1-promoted myofibroblast formation and function. Wound Repair Regen 2010; 18: 223–34. 7. Fingleton B. Matrix metalloproteinases as valid clinical targets. Curr Pharm Des 2007; 13: 333–46. 8. Fingleton B. MMPs as therapeutic targets—still a viable option? Semin Cell Dev Biol 2008; 19: 61–8. 9. Djafarzadeh R, Milani V, Rieth N, von Luettichau I, Skrablin PS, Hofstetter M et al. TIMP-1-GPI in combination with hyperthermic treatment of melanoma increases sensitivity to FASmediated apoptosis. Cancer Immunol Immunother 2009; 58: 361–71. 10. Djafarzadeh R, Notohamiprodjo S, Rieth N, Hofstetter M, Noessner E, Nelson PJ. Treatment of dermal fibroblasts with GPI-anchored human TIMP-1 protein moderates processes linked to scar formation. J Invest Dermatol 2013; 133: 803– 11. 11. Djafarzadeh R, Mojaat A, Vicente AB, von Luttichau I, Nelson PJ. Exogenously added GPI-anchored tissue inhibitor of matrix metalloproteinase-1 (TIMP-1) displays enhanced and novel biological activities. Biol Chem 2004; 385: 655–63. 12. Djafarzadeh R, Noessner E, Engelmann H, Schendel DJ, Notohamiprodjo M, von Luettichau I et al. GPI-anchored TIMP-1 treatment renders renal cell carcinoma sensitive to FASmeditated killing. Oncogene 2006; 25: 1496–508. 13. Djafarzadeh R, Sauter M, Notohamiprodjo S, Noessner E, Goyal P, Siess W et al. Recombinant GPI-anchored TIMP-1 stimulates growth and migration of peritoneal mesothelial cells. PLoS ONE 2012; 7: e33963. 14. Notohamiprodjo S, Djafarzadeh R, Rieth N, Hofstetter M, Jaeckel C, Nelson PJ. Cell surface engineering of renal cell carcinoma with glycosylphosphatidylinositol-anchored TIMP-1
75
TIMP-1-GPI promotes wound healing
15. 16.
17. 18. 19.
20. 21.
22.
23.
24.
25.
76
blocks TGF-beta 1 activation and reduces regulatory ID gene expression. Biol Chem 2012; 393: 1463–70. Medof ME, Nagarajan S, Tykocinski ML. Cell-surface engineering with GPI-anchored proteins. FASEB J 1996; 10: 574–86. Keese CR, Wegener J, Walker SR, Giaever I. Electrical woundhealing assay for cells in vitro. Proc Natl Acad Sci USA 2004; 101: 1554–9. Giaever I, Keese CR. Micromotion of mammalian cells measured electrically. Proc Natl Acad Sci USA 1991; 88: 7896–900. Margadant C, Sonnenberg A. Integrin-TGF-beta crosstalk in fibrosis, cancer and wound healing. EMBO Rep 2010; 11: 97–105. Guo M, Toda K, Grinnell F. Activation of human keratinocyte migration on type I collagen and fibronectin. J Cell Sci 1990; 96 (Pt 2): 197–205. Roberts AB. Molecular and cell biology of TGF-beta. Miner Electrolyte Metab 1998; 24: 111–9. Yu Q, Stamenkovic I. Cell surface-localized matrix metalloproteinase-9 proteolytically activates TGF-beta and promotes tumor invasion and angiogenesis. Genes Dev 2000; 14: 163–76. Morrison CJ, Butler GS, Rodriguez D, Overall CM. Matrix metalloproteinase proteomics: substrates, targets, and therapy. Curr Opin Cell Biol 2009; 21: 645–53. van Goor H, Melenhorst WB, Turner AJ, Holgate ST. Adamalysins in biology and disease. J Pathol 2009; 219: 277– 86. Klier CM, Nelson EL, Cohen CD, Horuk R, Schlondorff D, Nelson PJ. Chemokine-induced secretion of gelatinase B in primary human monocytes. Biol Chem 2001; 382: 1405–10. Brew K, Dinakarpandian D, Nagase H. Tissue inhibitors of metalloproteinases: evolution, structure and function. Biochim Biophys Acta 2000; 1477: 267–83.
Djafarzadeh et al.
26. Raggi MC, Djafarzadeh R, Muenchmeier N, Hofstetter M, Jahn B, Rieth N et al. Peritumoral administration of GPI-anchored TIMP-1 inhibits colon carcinoma growth in Rag-2 gamma chain-deficient mice. Biol Chem 2009; 390: 893–7. 27. Verrecchia F, Chu ML, Mauviel A. Identification of novel TGFbeta/Smad gene targets in dermal fibroblasts using a combined cDNA microarray/promoter transactivation approach. J Biol Chem 2001; 276: 17058–62. 28. Sarret Y, Woodley DT, Grigsby K, Wynn K, O’Keefe EJ. Human keratinocyte locomotion: the effect of selected cytokines. J Invest Dermatol 1992; 98: 12–6. 29. Leask A, Abraham DJ. TGF-beta signaling and the fibrotic response. FASEB J 2004; 18: 816–27.
Supporting Information Additional Supporting Information may be found in the online version of this article at the publisher’s web-site: Video S1. Effect of TIMP-1-GI on human keratinocyte migration in the context of a “scratch” assay. An assay system (ibidi, Martinsried, Germany) was used to disrupt an established confluent keratinocyte monolayer leaving a uniform area devoid of cells. Repopulation as a consequence of proliferation and migration of cells from the edge of the monolayer was recorded by video microscopy. TIMP-1-GPI treatment had an immediate effect on keratinocyte migration/ mobility with complete closure seen by 12–13 hrs at the highest concentration of TIMP-1-GPI (12 ng/ml) tested. By contrast the control samples were not closed by the 24 hr time point.
Wound Rep Reg (2014) 22 70–76 © 2014 by the Wound Healing Society
Reprint requests: Prof. P. J. Nelson, Medizinische Klinik und Poliklinik IV, Campus Innenstadt, Ludwig-Maximilians-Universität München, Arbeitsgruppe Klinische Biochemie, Schillerstrasse 42, 80336, Munich, Germany. Tel: +89 218075 845; Fax: +89 218075 860; Email: [email protected] Manuscript received: November 8, 2012 Accepted in final form: October 10, 2013 DOI:10.1111/wrr.12132
ABSTRACT The balance between matrix metalloproteinases and their endogenous tissue inhibitors (TIMPs) is an important component in effective wound healing. The biologic action of these proteins is linked in part to the stoichiometry of TIMP/matrix metalloproteinases/surface protein interactions. We recently described the effect of a glycosylphosphatidylinositol (GPI) anchored version of TIMP-1 on dermal fibroblast biology. Here, cell proliferation assays, in vitro wound healing, electrical wound, and impedance measurements were used to characterize effects of TIMP-1-GPI treatment on primary human epidermal keratinocytes. TIMP-1-GPI stimulated keratinocyte proliferation, as well as mobilization and migration. In parallel, it suppressed the migration and matrix secretion of dermal myofibroblasts, and reduced their secretion of active TGF-β1. Topical application of TIMP-1-GPI in an in vivo excisional wound model increased the rate of wound healing. The agent positively influenced different aspects of wound healing depending on the cell type studied. TIMP-1-GPI counters potential negative effects of overactive myofibroblasts and enhances the mobilization and proliferation of keratinocytes essential for effective wound healing. The application of TIMP-1-GPI represents a novel and practical clinical solution for facilitating healing of difficult wounds.
Wound healing is a complex multistep process in which epidermal keratinocytes and fibroblasts play central roles. Reepithelialization of the wound relies on the mobilization and proliferation of keratinocytes: In parallel, fibroblast/ myofibroblast activation and proliferation within the wound are required for efficient wound healing. Overactivation of fibroblasts can lead to hypertrophic scar formation especially when keratinocyte-driven wound closure is delayed, or when myofibroblasts do not undergo efficient apoptosis.1 The matrix metalloproteinase (MMP) family is a large structurally related enzyme family. The actions of MMPs are linked to the processing of a wide range of extracellular proteins including extracellular matrix (ECM), cytokines, and growth factors. In this regard, these enzymes are thought to help control a diverse spectrum of biologic processes. The 24 MMPs found in humans include collagenases, gelatinases, stromelysins, and membrane-type MMPs.2 MMP activities are regulated at diverse levels including by proteolytic activation of precursor zymogens, and the presence of four endogenous inhibitors, the tissue inhibitors of metalloproteinases (TIMP-1, TIMP-2, TIMP-3, and TIMP-4).3–5 The net balance between MMP/TIMP expression is strongly associated with the turnover of ECM and its remodeling during normal tissue homeostasis and pathogenesis. MMP/TIMP activity also underlies the migratory and remodeling events that drive wound repair. While MMP activity is often elevated in wounds with delayed 70
or aberrant healing, targeting wound healing with small molecular antagonists has proven problematic in part because of the pleiotropic biology of MMPs and their inhibitors.6–8 We have previously shown that the bioactivity of TIMP-1 can be altered by changing its cellular distribution, specifically by focusing expression directly on to the cell surface.9–14 TIMP-1 is normally a secreted protein found within the ECM and only loosely associated with the cell surface through its interaction with other proteins. Modification human TIMP-1 protein by the addition of a glycosylphosphatidylinositol (GPI) anchor gives rise to a recombinant reagent can be transferred from one cell to another through a process called “cell painting” or “cell surface engineering.”15 Exogenously applied TIMP-1-GPI fusion protein is efficiently incorporated into virtually all cell surface membranes and effectively refocuses the actions of TIMP-1directly onto the cell surface. This can give rise to enhanced as well as unique biologic effects. We have recently shown that TIMP-1-GPI can have dramatic effects on the biology of primary human dermal fibroblasts.10 TIMP-1-GPI treatment was found to moderate processes linked to scar formation including reducing fibroblast/myofibroblast proliferation, ECM production, and enhancing their sensitivity to apoptosis.10 Here we have expanded on the characterization of effects of TIMP-1-GPI treatment on general aspects of wound healing and further report that treatment with the agent actively promotes the Wound Rep Reg (2014) 22 70–76 © 2014 by the Wound Healing Society
Djafarzadeh et al.
mobilization and proliferation of primary epidermal keratinocytes. In an in vivo excisional wound healing model, topical application led to enhanced wound closure.
MATERIALS AND METHODS Cellular reagents and tissue culture
Normal human epidermal keratinocytes (NHEK.f-c Promo Cell, Heidelberg, Germany, No. C-12001) were cultured in Keratinocyte Basal Medium (Calcium free) (Promo Cell, No. C-20211). Primary human dermal fibroblast cells (NHDF-c PromoCell, No. C-12300) were cultured in DULBECCO’S MEM with Glutamax-1 (Gibco BRL, Life Technologies GmbH, Eggenstein, Germany, No. 21885-025), supplemented with 10% heat-inactivated fetal calf serum (FCS). Recombinant TIMP-1-GPI protein was produced and purified as previously described.11 All experiments adhere to the Declaration of Helsinki Principles.
TIMP-1-GPI promotes wound healing
migration (ibidi GmbH, No. 80209), essentially a modified version of a scratch assay. A culture insert was used to provide two cell culture reservoirs separated by a 500 μm thick wall. Following confluent keratinocyte culture in both reservoirs, the silicone insert was removed leaving two defined cell patches separated by a zone defined by the original separation wall. Video microscopy was then used to evaluate cell migration over 24 hours. An ibidi Heating & Incubation Systems enabled live cell imaging with incubator conditions directly on the microscope. The recorded events were then quantified using Image J software (NIH freeware: http://rsbweb.nih.gov/ ij/). TIMP-1-GPI treatment lead to accelerated keratinocyte wound healing (see Video S1 provided in Supplement online). Western blot analysis
Western blot was used to detect fibronectin (DakoCytomation, Glostrup, Denmark, No. A0245) and β-actin (Acris, Hiddenhausen, Germany, No. ab8227) in dermal fibroblasts as previously described.9
Proliferation
Primary epidermal keratinocytes or fibroblasts 5 × 103 per 100 μL medium were cultured in 96-well microtiter plates for 24 hours under standard conditions to yield firmly attached and stably growing cells. After discarding supernatants, 50 μL of medium containing native TIMP-1-GPI, buffer or rhTIMP-1 were added and the cells were incubated from 24 to 72 hours for fibroblasts and 48 to 144 hours for keratinocytes, and 50 μL of 1 mg/mL solution of (3-[4,5-dimethylthiazol-2yl]-2,5 diphenyl tetrazolium bromide) assay (MTT) was added. Rapidly dividing cells exhibit high rates of MTT reduction. After 3-hour incubation at 37 °C, formazan crystals were dissolved by additional 100 μL isopropanol and 0.04 N HCl. Absorbance was measured at 570 nm using GENios Plus TECAN ELISA reader (Tecan, Männedorf, Switzerland). For each experiment, at least six wells were analyzed per experimental condition and time point.10 In vitro assays with primary keratinocytes cells
ECIS electrode arrays (8W1E) were obtained from (ibidi GmbH, Munich, Germany, No. 70001). Before starting ECIS measurements, 300 μL of complete medium was placed in each well and was allowed to equilibrate in the incubator for 24 hours. Three hundred microliters of keratinocyte cell suspension (5 × 104 cells per mL) were added to each well, resulting in a final surface concentration of 1.8 × 104 cells per cm2 and a volume of 300 μL of medium. After cell inoculation, the wells were incubated overnight, while attachment and spreading were followed by electrical impedance. Impedance levels were used to verify that confluence was achieved and maintained and that the cells exhibited normal levels of impedance fluctuations, indicative of healthy cell layers.16 The confluent cell layers were incubated for 1 day before wounding. The damaging pulse was delivered at 2.5 V at 40 kHz for 30 seconds. Medium was changed after wounding, and TIMP1-GPI, control rhTIMP-1 and vehicle were added to specific wells. Electrical impedence was then followed in real time for 48 hours. A second assay was used to disrupt an established confluent human primary keratinocyte monolayer using a commercial Wound Rep Reg (2014) 22 70–76 © 2014 by the Wound Healing Society
Fibroblast cell migration assay
Primary fibroblasts were cultured in FCS free medium for 12 hours before the onset of the experiment. The cells were then labeled with Calcein AM (5 μg/mL) (Molecular Probes, Life Technologies, Darmstadt, Germany; No. MW994.87) for 1 hour and detached from the culture flasks with 1.5 mM EDTA in 1 × phosphate-buffered saline (PBS) before migration. The assay medium (FCS free medium with 1% bovine serum albumin (BSA) (Gibco BRL, Life Technologies GmbH, No. 15260-037) was used to dilute the TIMP inhibitors. Six hundred microliters of assay medium containing 10 % FCS or without FCS was placed in the bottom chamber of a BD Fluoro Blok Insert System with a 3 μm pore size (Becton Dickinson Falcon and Company, San Jose, CA, No. 351156) and 50,000 calcein-labeled cells with and without TIMP-1GPI treatment with a final volume of 100 μL were placed in the wells and measured by detection of the fluorescence of cells migrated through the bottom chamber with an SPECTRA FLUOR plus TECAN reader at 485 nm excitation and 515 nm emission. Data represent the mean of n = 3 inserts. TGF-β enzyme-linked immunosorbent assay
Active and total TGF-β concentrations were measured using an enzyme-linked immunosorbent assay (ELISA) kit from (DB Biosciences, San Jose, CA , No. 559119) applied according to the manufacturer’s directions. Excisional wound healing model
C57BL/6 mice were purchased from Charles River Laboratories (Sulzfeld, Germany). The animal studies were approved by and conducted in accordance with the principles of the regulatory agency of the State of Bavaria, Germany (performed by Dr. Ralf Huss, Department of Pathology, University of Munich). The mice were anesthetized with ketavet (100 mg/kg mouse body weight) and xylazin (5 mg/kg mouse body weight) followed premedication with atropine sulfate. 71
1.0
Reepithelialization of cutaneous wounds depends on the proliferation, mobilization, and migration of keratinocytes.1 The effect of exogenously added TIMP-1-GPI on the proliferation of primary human keratinocytes in vitro was determined at 48, 96, and 144 hours. Treatment with increasing levels of TIMP1-GPI (2, 4, 6, 8, 10, 12, and 14 ng/mL) showed a statistically significant dose-dependent increase in proliferation relative to the vehicle control at all time points, while 14 ng/mL of control recombinant TIMP-1, GPI anchor control (phosphatidylserine), or denatured TIMP-1-GPI had no significant effect on proliferation (Figure 1). TIMP-1-GPI promotes keratinocyte migration in vitro
Two in vitro assays were used to model the mobilization and migration events that occur during wound healing. In the first assay, electric cell-substrate impedance was used to monitor keratinocyte mobilization and migration. Keratinocytes were grown on electrodes, and impedance levels were monitored to verify that cellular confluence was achieved and maintained, and that the cells exhibited normal levels of impedance fluctuations, indicative of healthy cell layers16,17 (Figure 2). Then cells in a defined region were then subjected to currents resulting in cell death, and impedance was further monitored to measure the rate of repair of the “wound”(Figure 2). Effects of TIMP-1-GPI treatment were evident at 4 ng/mL TIMP-1-GPI within a few hours of treatment and increased with dose. At 12 ng/mL TIMP-1-GPI, the “wound”/ keratinocyte monolayer was completely reestablished by approximately 12–15 hours (Figure 2). To directly visualize these effects, a second commercial assay system was used to disrupt an established confluent keratinocyte monolayer (ibidi). The method generates of a uniform area devoid of cells, and repopulation of this area in the keratinocyte cell sheet occurs as a consequence to the proliferation and migration of cells from the edge of the keratinocyte monolayer. These events were recorded by video microscopy and quantified using Image J software. TIMP-1GPI treatment again led to an immediate effect on keratinocyte migration/mobility, and complete closure was 72
+
0.8
+
+
0.6 + +
0.4
+ +
+
+ + +
+ +
+ +
+ + ++ + +
+ + +
+ + +
+ + ++ + +
+ ++ ++ +
0.2
ng 10 /mL ng /m L 12 ng /m L 14 ng /m L
/m L
ng
/m L 6
ng
/m L 4
PI
ng 2
8
TIMP-1-GPI
rh
TI
M
P1
Ve h
ng
ic le
/m L
0.0
RESULTS GPI-anchored TIMP-1 increases keratinocyte proliferation
+ +
+ +
+ +
+ +
14
After the dorsal surface of each mouse had been disinfected with an alcohol swab, a full-thickness excisional wound (0.8 cm in diameter) was created on the dorso-medial back of each animal, using a standard skin biopsy punch (Stiefel® Laboratorium GmbH, Offenbach, Germany). Immediately after surgery, as well as on days 3, 5, and 7 thereafter, rhTIMP-1 buffer, or rhTIMP-1, or TIMP-1-GPI was applied on the top of created wound. The measurement of wound size was performed on days 0, 3, 5, 7, and 10. The wound area (mm2) was calculated using the formula of an ellipse; the wound closure rate was expressed as a ratio of the wounded area at each time point divided by the area of the original wound at time 0. On postoperative day 14, the animals were euthanized and the wounds were excised, fixed in 4% neutralbuffered formalin solution, and embedded in paraffin.
Djafarzadeh et al.
Keratinocyte proliferation - MTT
TIMP-1-GPI promotes wound healing
Figure 1. Treatment with GPI-anchored TIMP-1 stimulates keratinocyte cell proliferation. To assess the effect of TIMP-1 surface engineering on proliferation of primary human keratinocytes, MTT assays were performed. TIMP-1-GPItreated keratinocytes show a dose-dependent increase in proliferation (at 48 [fine diagonal lines], 96 [wide diagonal], and 144 [white] hours, reflected in the three columns). Absorbance was then measured at 550 nm using GENios Plus TECAN ELISA reader. For each experiment, at least six wells were analyzed per experimental condition and time point. Each individual treatment result was compared with the corresponding vehicle control using paired Student’s t-test (+p < 0.01; ++p < 0.001; +++p < 0.0001). GPI, glycosylphosphatidylinositol; MTT, (3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide) assay; TIMP, endogenous tissue inhibitor.
seen by 12–13 hours at the highest concentration of TIMP-1GPI (12 ng/mL) tested. By contrast, the control samples were still not closed by the 24 hours time point (see Video S1 provided in online Supplement). The rapidity of the mobilization and migration response seen with TIMP-1-GPI treatment in both models suggests processes independent of the enhanced proliferation shown in Figure 1. The GPI-anchored TIMP-1 suppresses fibroblast migration, proliferation, and matrix production
Fibrosis and scar formation is characterized by fibroblast proliferation, differentiation into myofibroblasts, and excessive production and deposition of ECM.10,18 We recently demonstrated that TIMP-1-GPI treatment could inhibit fibroblast proliferation and lead to a more proapoptotic phenotype.10 The effect of TIMP-1-GPI on the proliferation of activated primary dermal fibroblast cells was verified here in parallel experiments to provide contrast with the epidermal keratinocyte results shown above. In contrast to that seen with keratinocytes, TIMP-1-GPI-treated primary dermal fibroblasts showed a statistically significant dose-dependent Wound Rep Reg (2014) 22 70–76 © 2014 by the Wound Healing Society
Djafarzadeh et al.
TIMP-1-GPI promotes wound healing
+ 0.40 0.35 0.30 0.25
+ + ++ + ++ + + ++ ++ + ++ + ++ + + ++ + + + +++ + + + ++ + ++ + ++ + + ++ ++ +
0.20 0.15 0.10 0.05
+
+ +
+ + + + + +
+
B
Migrated Fibroblasts × 103
TIMP-1-GPI + 0.4 ng/mL TNF-α
50 **** **
TN F Ve -α hi cl e
8 n 10 g/m n L 12 g/m ng L 14 /m ng L /m L
2
ic
le
ng 0 4 /mL ng / 6 mL ng 8 /mL n 10 g/m n L 12 g/m ng L 14 /m ng L /m L
0.00
Ve h
Figure 2. In vitro assay shows increased keratinocyte migration following application of TIMP-1-GPI. An electrical woundhealing assay using an ECIS electrode arrays measured resistive impedance over time and in response to induced damage to a monolayer. Keratinocytes treated with TIMP-1GPI showed rapidly increased impedance and thus enhanced repair of damage. GPI, glycosylphosphatidylinositol; TIMP, endogenous tissue inhibitor.
Fibroblast proliferation - MTT
A
TIMP-1-GPI
rhTIMP-1 14 ng/mL
N.s. N.s.
****
40
Treatment reduces levels of active TGF-β1
Fibroblast recruitment, proliferation, and production of ECM are strongly influenced by the profibrotic growth factor TGFβ.20 TGF-β1 also helps control the growth of keratinocytes.21,22 Specific ELISA was used to assess the production of total vs. active forms of TGF-β1 by primary dermal fibroblasts in serum-free media. TIMP-1-GPI treatment effectively reduced the levels of active TGF-β1 relative to the untreated control, suggesting a direct effect on proteolytic processing of the latent TGF-β1 complex by the surface bound TIMP-1 (TIMP1-GPI at 7 ng/mL p < 0.0002; 14 ng/mL p < 0.0001)10,13,14 (Figure 4). TIMP-1-GPI treatment accelerates excisional wound healing
The effect of TIMP-1-GPI on excisional dermal wound healing was then evaluated using C57BL/6 mice. After disinfection, a defined, full-thickness excisional wound (0.8 cm in Wound Rep Reg (2014) 22 70–76 © 2014 by the Wound Healing Society
**
30
**
20 10
ed
10 % ia 2 FC n 4 g/mS n L 6 g/m n L 8 g/m n 10 g/mL 12 ng/ L m 14 ng/ L ng mL /m L 8 ng 10 /m 12 ng/ L m 14 ng/ L ng mL /m L
0
M
reduction in proliferation (to TNF-α [0.4 ng/ml] [p < 0.01– 0.0001]) (Figure 3A). Addition of as little as 2 ng/mL of TIMP-1-GPI was also found to reduce serum-driven fibroblast migration (p < 0.004),and completely block migration of the cells at 12 ng/mL (p < 0.0001) (Figure 3B). By contrast, recombinant TIMP-1 alone (14 ng/mL) showed reduced effects on fibroblast cell migration. Fibronectin is an important component of the ECM and is secreted by activated fibroblasts.19 Dermal fibroblast cells grown in 15% FCS were treated with 14 ng/mL TIMP-1-GPI or control rhTIMP-1, and Western blot was used to quantify expression of fibronectin in whole cell extracts (Figure 3C). TIMP-1-GPI, but not an equivalent concentration of control recombinant TIMP-1, reduced expression of the glycoprotein. Similar results were previously shown at the steady state mRNA level.10
rhTIMP-1
TIMP-1-GPI
C
[14 ng/mL] Vehicle
Fibronectin β-actin Control
rhTIMP-1
TIMP-1-GPI
230 kDa 41 kDa
Figure 3. TIMP-1-GPI suppresses fibroblast proliferation, migration, and fibronectin production. (A) Exogenously added TIMP-1-GPI protein elicited a dose-dependent decrease in fibroblast proliferation at 24, 48, or 72 hours to cells stimulated with 1% FCS with or without 0.4 ng/mL TNF-α. See legend to Figure 1 for details. Each treatment result was compared with the corresponding TNF-α/FCS control value (e.g., 24, 48, and 72 hours) using pared paired Student’s t-test (+p < 0.01; ++p < 0.001; +++p < 0.0001). (B) TIMP-1-GPI shows a potent suppression of fibroblast migration in response to 10 % FCS. Addition of rhTIMP-1 at the same concentration showed limited effects on fibroblast migration. Each treatment result was compared with the FCS control value using paired Student’s t-test. (*p < 0.01; **p < 0.001; ***p < 0.0001). (C) Confluent fibroblasts were cultured in the presence or absence of TIMP-1-GPI, and fibronectin was quantified in cell extracts by Western blot. rhTIMP-1 at 14 ng/mL did not lead to pronounced decrease in fibronectin expression, while TIMP-1GPI at 14 ng/mL strongly reduced fibronectin levels. GPI, glycosylphosphatidylinositol; TIMP, endogenous tissue inhibitor.
73
TIMP-1-GPI promotes wound healing
Figure 4. Effect of TIMP-1-GPI on TGF-β1 production by activated dermal fibroblasts. Confluent primary dermal fibroblasts were cultured in the presence or absence of TIMP-1-GPI in serum free medium. Total vs. activated levels of TGF-β1 were measured using commercial ELISAs. The active form of TGF-β1 was significantly reduced by treatment with increasing levels of TIMP-1-GPI. Each active TGF-β1 value was compared with the vehicle control using paired Student’s t-test (*p < 0.01; **p < 0.001; ***p < 0.0001). ELISA, enzyme-linked immunosorbent assay; GPI, glycosylphosphatidylinositol; TGFβ1, transforming growth factor beta-1 (TGF-beta 1); TIMP, endogenous tissue inhibitor.
diameter) was created on the dorso-medial back of each animal using a standard skin biopsy punch. Immediately after surgery, as well as on days 3, 5, and 7 thereafter, 200 μL of vehicle (0.025% Triton × 100 in 1 × PBS), rhTIMP-1 (14 ng/ml in vehicle), or TIMP-1-GPI (14 ng/mL in vehicle) was applied to the wound surface (Figure 5A). Measurement of wound size was performed on days 0, 3, 5, 7, and 10. The resulting wound area (mm2) was calculated using the formula of an ellipse and the wound-closure rate expressed as a ratio of the wounded area at each time point, divided by the area of the original wound at time 0. On postoperative day 14, the animals were euthanized. The results show a significantly improved wound closure rate with TIMP-1-GPI treatment. Application of the agent showed an accelerated rate of wound closure as early as day 3 after surgery, under TIMP-1-GPI treatment relative to that achieved by rhTIMP-1 buffer (§p < 0.05) or rhTIMP-1 (**p < 0.05) application (both vs. control). This was evidenced by measurement of wound area (Figure 5).
DISCUSSION Cutaneous wound repair is a complex biologic process. Healing proceeds in a sequential manner that has been divided 74
Djafarzadeh et al.
into four overlapping phases: inflammation, granulation, reepithelialization, and tissue remodeling. The effector cells responsible for driving excisional wound healing include fibroblasts and keratinocytes. Reepithelialization of the wound is a critical phase that depends upon the speed with which keratinocytes can mobilize and effect wound closure. They must advance across the wound and proliferate at the edges. In the granulation phase, fibroblasts migrate into the wound, proliferate, and secrete ECM. When wound closure is not rapid enough, fibroblast/myofibroblast over proliferation within the wound can lead to excessive scar formation. MMP/TIMP biology underlies central aspects of tissue homeostasis and wound repair. The balance between MMPs and TIMPs helps regulate ECM turnover during normal tissue homeostasis and pathogenesis, and also helps control cell signaling through modification of growth factors.22,23 Where the balance between the proteolytic breakdown and deposition of ECM is disturbed, disorders such as abnormal wound healing may result. TIMP-1 is naturally a soluble protein that can be detected on the cell surface only through its interaction with surface bound molecules.24,25 Increase in the local concentration of TIMP proteins by the administration of recombinant protein or gene transfer has shown some efficacy in various animal models. Recombinant TIMP-1-GPI fusion protein is efficiently incorporated into cell surface membranes, and through this, shiftsTIMP-1 directly to the cell surface.9–14,26
Figure 5. Effect of TIMP-1-GPI protein on wound healing in vivo. A full-thickness excisional wound (0.8 cm in diameter) was created on the dorso-medial back of mice using a standard skin biopsy punch. Immediately after surgery, as well as on days 3, 5, and 7 thereafter, rhTIMP-1 buffer, or rhTIMP-1, or TIMP-1-GPI was applied on the top of created wound. After treated with rhTIMP-1 buffer, rhTIMP-1, or TIMP-1-GPI. Wounds were photographed at the times indicated and the volume of the wound area measured (see Materials and Methods section). Wound area is expressed as the mean ± SD. TIMP-1-GPI treatment relative to that achieved by rhTIMP-1 buffer (§p < 0.05) or rhTIMP-1 (**p < 0.05) application. *p < 0.05, #p < 0.05 (both vs. control). The data were analyzed using the paired Student’s t-test with p < 0.05 considered to be significant. GPI, glycosylphosphatidylinositol; TIMP, endogenous tissue inhibitor. Wound Rep Reg (2014) 22 70–76 © 2014 by the Wound Healing Society
Djafarzadeh et al.
This modification to the native protein results in enhanced as well as novel biologic effects. These effects have been previously shown to be independent of the lipid anchor and to require functional TIMP-1 protein.11–13 Here, TIMP-1-GPI treatment was shown to lead to an increase in the proliferation and migration of primary human keratinocytes, but in parallel suppresses the proliferation, migration, and ECM production of activated dermal myofibroblasts/fibroblasts.10 Topical application of the agent on excisional wounds in mice led to accelerated wound closure. While an effect on enhanced wound contraction cannot be ruled out, the in vivo experiments in general support the in vitro wound healing data. Because it is GPI-anchored protein, TIMP-1-GPI cannot penetrate deeply into tissues. Results suggest that only the top cell layers of treated tissues are influenced by topical application. The observation that TIMP-1-GPI treatment reduces fibrosis seems counterintuitive at first view. A general reduction in protease activity should lead to a more pronounced deposition of matrix components because of a reduction in the general degradation in the dynamics of tissue homeostasis (i.e., production balanced by degradation). However, the TIMP-1 proteins and their targets, the MMPs, have much broader biologic activities than ECM turnover. These include effects on cell growth and signal transduction, in part, through their modulation of diverse cytokine activities. I n addition, TIMP proteins have also been shown to have direct growth factor like characteristics that may impact the biology described here. Wound healing is a complex process linked to the balance of regulatory cytokines/growth factors present. Cell surface engineering using GPI-anchored TIMP-1 suppresses processing of growth factors including TGF-β1.10,13,14 TGF-β1 is rapidly induced in response to injury and has some of the broadest effects of any cytokine linked to wound healing.27 TGF-β1 has paradoxical actions in wound healing that are dependent on the cell type, its state of differentiation, and the context of action. While TGF-β1 is a potent inhibitor of keratinocyte proliferation and migration,28 it contributes to the fibrotic process by recruiting fibroblasts and stimulating their synthesis of collagens I, III, and V, proteoglycans, fibronectin, and other ECM components.27,29 Activated fibroblasts are a rich source of latent TGF-β1 in wounds. Active TGF-β1 must be released from its latent complex. TIMP-1-GPI treatment dramatically reduced the processing of the latent TGF-β1 to its active form presumably by a blockade of MMP or related enzymes.10,13,14,27 But it important to point out that TGF-β1 processing may represent only one aspect of the mode of action of this agent in the context of wound healing. Clearly, other soluble factors linked to efficient wound healing can be directly or indirectly influenced by treatment with TIMP-1GPI. Delayed wound healing represents an enormous resource consuming factor in medicine often requiring lengthy and expensive serial debridements, skin grafting, or vacuumassisted wound closure therapy, all of which show limited efficacy. We describe here an agent that shows remarkable efficacy in accelerated wound healing. Topical application with TIMP-1-GPI suppresses processes linked to dermal fibrosis and scarring, and at the same time dramatically enhances wound closure through activation and mobilization of epidermal keratinocytes. Thus, the agent may show efficacy as an adjuvant in excisional wound therapy. Wound Rep Reg (2014) 22 70–76 © 2014 by the Wound Healing Society
TIMP-1-GPI promotes wound healing
ACKNOWLEDGMENT Source of Funding: This work was supported by research grants from the German Research Foundation (DFG) NE 468/4-2. The in vivo wound healing assays were performed under the supervision and ethical license of Dr. Ralf Huss, LMU Pathology. The authors would also like to thank Dr. Ivan Ischenko for excellent technical assistance. Conflict of Interest: PJN holds patents covering the technology detailed here. No other conflicts of interest exist for any of the authors. REFERENCES 1. Broughton G, 2nd, Janis JE, Attinger CE. The basic science of wound healing. Plast Reconstr Surg 2006; 117 (7 Suppl.): 12S– 34S. 2. Fanjul-Fernandez M, Folgueras AR, Cabrera S, Lopez-Otin C. Matrix metalloproteinases: evolution, gene regulation and functional analysis in mouse models. Biochim Biophys Acta 2010; 1803: 3–19. 3. Overall CM, Lopez-Otin C. Strategies for MMP inhibition in cancer: innovations for the post-trial era. Nat Rev Cancer 2002; 2: 657–72. 4. Ra HJ, Parks WC. Control of matrix metalloproteinase catalytic activity. Matrix Biol 2007; 26: 587–96. 5. Moore CS, Crocker SJ. An alternate perspective on the roles of TIMPs and MMPs in pathology. Am J Pathol 2012; 180: 12– 6. 6. Mirastschijski U, Schnabel R, Claes J, Schneider W, Agren MS, Haaksma C et al. Matrix metalloproteinase inhibition delays wound healing and blocks the latent transforming growth factorbeta1-promoted myofibroblast formation and function. Wound Repair Regen 2010; 18: 223–34. 7. Fingleton B. Matrix metalloproteinases as valid clinical targets. Curr Pharm Des 2007; 13: 333–46. 8. Fingleton B. MMPs as therapeutic targets—still a viable option? Semin Cell Dev Biol 2008; 19: 61–8. 9. Djafarzadeh R, Milani V, Rieth N, von Luettichau I, Skrablin PS, Hofstetter M et al. TIMP-1-GPI in combination with hyperthermic treatment of melanoma increases sensitivity to FASmediated apoptosis. Cancer Immunol Immunother 2009; 58: 361–71. 10. Djafarzadeh R, Notohamiprodjo S, Rieth N, Hofstetter M, Noessner E, Nelson PJ. Treatment of dermal fibroblasts with GPI-anchored human TIMP-1 protein moderates processes linked to scar formation. J Invest Dermatol 2013; 133: 803– 11. 11. Djafarzadeh R, Mojaat A, Vicente AB, von Luttichau I, Nelson PJ. Exogenously added GPI-anchored tissue inhibitor of matrix metalloproteinase-1 (TIMP-1) displays enhanced and novel biological activities. Biol Chem 2004; 385: 655–63. 12. Djafarzadeh R, Noessner E, Engelmann H, Schendel DJ, Notohamiprodjo M, von Luettichau I et al. GPI-anchored TIMP-1 treatment renders renal cell carcinoma sensitive to FASmeditated killing. Oncogene 2006; 25: 1496–508. 13. Djafarzadeh R, Sauter M, Notohamiprodjo S, Noessner E, Goyal P, Siess W et al. Recombinant GPI-anchored TIMP-1 stimulates growth and migration of peritoneal mesothelial cells. PLoS ONE 2012; 7: e33963. 14. Notohamiprodjo S, Djafarzadeh R, Rieth N, Hofstetter M, Jaeckel C, Nelson PJ. Cell surface engineering of renal cell carcinoma with glycosylphosphatidylinositol-anchored TIMP-1
75
TIMP-1-GPI promotes wound healing
15. 16.
17. 18. 19.
20. 21.
22.
23.
24.
25.
76
blocks TGF-beta 1 activation and reduces regulatory ID gene expression. Biol Chem 2012; 393: 1463–70. Medof ME, Nagarajan S, Tykocinski ML. Cell-surface engineering with GPI-anchored proteins. FASEB J 1996; 10: 574–86. Keese CR, Wegener J, Walker SR, Giaever I. Electrical woundhealing assay for cells in vitro. Proc Natl Acad Sci USA 2004; 101: 1554–9. Giaever I, Keese CR. Micromotion of mammalian cells measured electrically. Proc Natl Acad Sci USA 1991; 88: 7896–900. Margadant C, Sonnenberg A. Integrin-TGF-beta crosstalk in fibrosis, cancer and wound healing. EMBO Rep 2010; 11: 97–105. Guo M, Toda K, Grinnell F. Activation of human keratinocyte migration on type I collagen and fibronectin. J Cell Sci 1990; 96 (Pt 2): 197–205. Roberts AB. Molecular and cell biology of TGF-beta. Miner Electrolyte Metab 1998; 24: 111–9. Yu Q, Stamenkovic I. Cell surface-localized matrix metalloproteinase-9 proteolytically activates TGF-beta and promotes tumor invasion and angiogenesis. Genes Dev 2000; 14: 163–76. Morrison CJ, Butler GS, Rodriguez D, Overall CM. Matrix metalloproteinase proteomics: substrates, targets, and therapy. Curr Opin Cell Biol 2009; 21: 645–53. van Goor H, Melenhorst WB, Turner AJ, Holgate ST. Adamalysins in biology and disease. J Pathol 2009; 219: 277– 86. Klier CM, Nelson EL, Cohen CD, Horuk R, Schlondorff D, Nelson PJ. Chemokine-induced secretion of gelatinase B in primary human monocytes. Biol Chem 2001; 382: 1405–10. Brew K, Dinakarpandian D, Nagase H. Tissue inhibitors of metalloproteinases: evolution, structure and function. Biochim Biophys Acta 2000; 1477: 267–83.
Djafarzadeh et al.
26. Raggi MC, Djafarzadeh R, Muenchmeier N, Hofstetter M, Jahn B, Rieth N et al. Peritumoral administration of GPI-anchored TIMP-1 inhibits colon carcinoma growth in Rag-2 gamma chain-deficient mice. Biol Chem 2009; 390: 893–7. 27. Verrecchia F, Chu ML, Mauviel A. Identification of novel TGFbeta/Smad gene targets in dermal fibroblasts using a combined cDNA microarray/promoter transactivation approach. J Biol Chem 2001; 276: 17058–62. 28. Sarret Y, Woodley DT, Grigsby K, Wynn K, O’Keefe EJ. Human keratinocyte locomotion: the effect of selected cytokines. J Invest Dermatol 1992; 98: 12–6. 29. Leask A, Abraham DJ. TGF-beta signaling and the fibrotic response. FASEB J 2004; 18: 816–27.
Supporting Information Additional Supporting Information may be found in the online version of this article at the publisher’s web-site: Video S1. Effect of TIMP-1-GI on human keratinocyte migration in the context of a “scratch” assay. An assay system (ibidi, Martinsried, Germany) was used to disrupt an established confluent keratinocyte monolayer leaving a uniform area devoid of cells. Repopulation as a consequence of proliferation and migration of cells from the edge of the monolayer was recorded by video microscopy. TIMP-1-GPI treatment had an immediate effect on keratinocyte migration/ mobility with complete closure seen by 12–13 hrs at the highest concentration of TIMP-1-GPI (12 ng/ml) tested. By contrast the control samples were not closed by the 24 hr time point.
Wound Rep Reg (2014) 22 70–76 © 2014 by the Wound Healing Society
