Chronic venous leg ulcers represent the final outcome of lower extremity chronic venous insufficiency in most cases. It is a debilitating recurrent complication with an estimated cost of 1 billion dollars annually in the U.S.A. for its treatment. Moist wound healing providing dressings in combination with compression therapy represent the state-of-the-art ulcer treatment. Even though progress has been achieved, current efforts defining the highly aggressive chronic wound microenvironment promise further improvement of this treatment modality.
Chronic leg ulcers fail to progress through the normal pattern of wound repair involving inflammation, granulation tissue formation, and remodeling, but instead remain in a chronic inflammatory state with little signs of healing. There is increasing evidence that the persisting infiltration of neutrophils and macrophages (Rosner et al, 1995) in conjunction with elevated iron depositions in the ulcer tissue (Ackerman et al, 1988) play a major role in the generation of the prooxidant hostile microenvironment in chronic venous leg ulcers. Venous hypertension results in enhanced erythrocyte extravasation into the interstitium through widened interendothelial pores (Wenner et al, 1980). Iron is released from hemoglobin of degraded erythrocytes, ferritin, hemosiderin (Thomas et al, 1985;Biemond et al, 1988), aconitase, and other iron-sulfur proteins (Fridovich, 1995) via the attack of reactive oxygen species (ROS) and proteolytic enzymes occurring in nonhealing wounds. Subsequently, iron in combination with hydrogen peroxide released by activated neutrophils generates the highly toxic hydroxyl radical via the Fenton reaction (Halliwell and Gutteridge, 1989) Fe2+ + H2O2 
Fe3+ + OH
+ OH–. The resulting ferric iron (Fe3+) can be reduced to ferrous iron (Fe2+) in the Haber-Weiss reaction involving superoxide anions: Fe3+ + O2– Fe 2+ + O2, thus perpetuating the Fenton reaction and tissue damage. The hydroxyl radical initiates local lipid peroxidation of polyunsaturated fatty acids in cellular membranes resulting in an inflammatory response with continuous recruitment of neutrophils (Gutteridge et al, 1979;Weiss, 1989;Cheatle, 1991;Schaich, 1992;Falanga and Eaglstein, 1993).
In addition to direct damage of structural proteins of the extracellular matrix (Monboisse and Borel, 1992), the hydroxyl radical and other ROS enhance the synthesis and activity of matrix-degrading metalloproteinases (Wysocki et al, 1993;Weckroth et al, 1996;Saarialho-Kere, 1998) and serine proteases with concomitant inactivation of tissue inhibitors of matrix metalloproteinases (Mast and Schultz, 1996) and serine proteases (Grinnell et al, 1992;Rao et al, 1995;Grinnell and Zhu, 1996;Herrick et al, 1997) finally tilting the balance towards proteolysis. Apart from enhanced connective tissue degradation, serine proteases have been identified to degrade key growth factors in tissue repair such as platelet-derived growth factor (Wlaschek et al, 1997;Yager et al, 1997) and others.
The identification of factors delaying wound healing has been crucial for the development of a novel therapeutic concept that finally led to a new dressing concept called the selective pick-up principle.2 According to this optimized concept, we focused on the strategy to remove or eliminate the deleterious substances from the wound exudate by selectively acting biomolecules that are bound to the surface of traditional wound dressing materials instead of applying potentially beneficial agents to the wounds (Meyer-Ingold et al, 1998;Edwards et al, 1999). This selective sequestration is likely to result in a positive influence on wound healing or even in a reversal of this process from chronicity to healing. Furthermore, a more general benefit of selective pick-up dressings includes their ability to adjust the composition of wound fluids locally compared to the systemic action of drugs.
In a first attempt to disrupt the detrimental sequence of events leading to a prooxidant hostile microenvironment and impaired healing – and based on the herein reported finding of increased iron concentrations in the exudate of chronic venous leg ulcers – we have designed and developed a first prototype of a selective pick-up dressing for iron ions based on the iron chelator deferoxamine (DFO) covalently immobilized to a cellulose dressing.
The in vitro efficacy and functionality of the iron pick-up dressing was demonstrated by its iron-complexing properties, and its protective activity concerning the iron-driven upregulation of matrix-degrading metalloproteinase 1 (MMP-1) and lipid peroxidation in human dermal fibroblasts when subjected to different in vitro settings simulating the hostile microenvironment of venous leg ulcers.
Materials and methods
Cell culture
Dermal foreskin fibroblasts were established by outgrowth from biopsies of healthy human donors (Fleischmajer et al, 1981) at an age of 3–6 y and cultured in Dulbecco's modified Eagle's medium supplemented with glutamine (2 mM), penicillin (400 U per ml), streptomycin (50 mg per ml), and 10% fetal bovine serum in a humidified atmosphere of 5% CO2 and 95% air at 37°C. Cells were used at passages 5–15 corresponding to cumulative population doubling rates of 10–32 (Bayreuther et al, 1992).
Generation of ROS and the Fenton reaction
The Fenton reaction was simulated by exposure of confluent fibroblasts to concentrations of 220 
M Fe(III) citrate and 1 mM ascorbate in the presence or absence of H2O2 in serum-free medium. Non-toxic concentrations of dissolved DFO (220 M), equimolar concentrations of DFO immobilized to cellulose, and cellulose dressings without DFO were used for the incubation. By means of the transwell chamber system (Costar, Bodenheim, Germany) we avoided mechanical stress induced by direct contact between cells and the dressing. Cytotoxicity was analyzed using the MTT [3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide] assay (Green et al, 1984). The viability of cells was more than 80% of nontreated controls at the used concentrations of Fe(III) citrate, ascorbate, dissolved DFO, and cellulose-coupled DFO. After a 4 h exposure of fibroblasts to different treatment modalities, which was performed in medium with or without serum, fibroblasts were washed with phosphate-buffered saline and either fresh medium without serum was added to determine the MMP-1 production after 24 h using a MMP-1 specific enzyme-linked immunosorbent assay (ELISA), or fibroblasts were directly lyzed in water for the high pressure liquid chromatography (HPLC) based measurement of malondialdehyde (MDA), the major product of lipid peroxidation.
ELISA
An MMP-1 "sandwich" ELISA assay was performed according to the manufacturer's protocol (Amersham Pharmacia Biotech, Braunschweig, Germany) using precoated 96-well immunoplates, rabbit antihuman MMP-1 antibodies, and an antirabbit horseradish peroxidase conjugate. 3,3',5,5'-Tetramethylbenzidine was used as peroxidase substrate. Optical densities were determined at 450 nm using a microtiter plate reader LP 400 (Sanofi Diagnostics Pasteur, Freiburg, Germany). MMP-1 concentrations in the samples were determined against standard curves using Graph Pad Software (San Diego, CA).
Measurement of thiobarbituric-acid-reactive substances
Thio barbituric-acid-reactive substances were measured as described previously (Yu et al, 1986) with minor modifications. Briefly, to 0.5 ml of cell lysate, 20 l 5% butylated hydroxytoluene, 0.35 ml 20% trichloroacetic acid, and 0.5 ml 1.4% thiobarbituric acid were added. The mixtures were incubated at 95°C for 15 min. Following centrifugation, the samples were analyzed using HPLC on a 4 
250 mm LiChrospher 100 RP-18 column with fluorescence detection. The mobile phase consisted of 55% potassium phosphate (50 mM, pH 5.5) and 45% methanol. Standard solutions of MDA (0.1–1 M) were used for calibration. MDA concentrations were expressed per mg protein. Protein was determined using the Bradford assay (Biorad, Munich, Germany).
Collection of wound exudates and determination of total iron and protein content
Wound fluids were collected from chronic venous leg ulcers of nine patients after coverage of the wounds with a transparent polyurethane semiocclusive film dressing (Cutifilm, Beiersdorf, Hamburg, Germany). After 30 min wound exudate was aspirated through the film dressing using a 30 G needle mounted on a 2 ml syringe. Acute wound exudate was obtained by drainage from four patients who had undergone mastectomy and one patient after excision of a myosarcoma, and from four patients by punctation of blisters from bullous pemphigoid. In addition, pooled human plasma derived from 50 donors was analyzed. The fluids were centrifuged at 10,000 g, and supernatants were stored frozen at below -20°C before analysis. Only clear yellow exudates were analyzed for total iron levels using atomic absorption spectroscopy after microwave pressure disintegration with nitric acid. This method requires at least 200 l of fluid with a detection limit of 1 ppm. The total protein content of each fluid was determined using established procedures (Bradford, 1976).
Covalent coupling of DFO to cellulose dressing
DFO was covalently immobilized to a cellulosic support by using the 1,1'-carbonyldiimidazole (CDI) method (Hearn, 1987) yielding a stable urethane bond. Briefly, the procedure comprises two steps: (i) activation of the hydroxyl groups of cellulose support material (cotton gauze, viscose fabric, or cellulose film) was carried out with CDI in anhydrous acetone for 2 h according to an earlier published method (Feldhoff, 1992); (ii) reaction of the CDI-activated cellulose with DFO mesylate (Sigma, Deisenhofen, Germany) was performed as earlier described (Feldhoff, 1992) except that 25 mM sodium bicarbonate solution, pH 7.7, was used for 4 h. The extensively washed DFO-cellulose was air-dried at room temperature and sterilized by gamma irradiation at 30 kGy. Depending on the cellulose matrix used, between 20 and 100 mol DFO could be bound to 1 g of cellulose fabric. For the production of placebo-cellulose all steps were carried out identically without adding CDI and DFO.
Leakage studies on gamma-irradiated immobilized DFO
The stability of gamma-sterilized DFO-cellulose was examined by extracting 100 mg cellulose with 5 ml of either a methanol-water mixture (50:50 vol/vol) or physiologic solution (0.9% NaCl, 1 mM MgCl2, 1 mM CaCl2) for 16 h at room temperature. The resulting extracts were analyzed directly, in the case of methanol, or after solvent exchange by solid phase extraction, in the case of physiologic extracts, by means of liquid chromatography/mass spectrometry (LC/MS) with a detection limit of 1 g per ml for methanol extracts and 10 g per ml for physiologic extracts.
In a parallel set of experiments, dry pulverized DFO was gamma irradiated (30 kGy) and subsequently analyzed for changes in its mass profile by means of LC/MS.
A third approach was used for the detection of volatile breakdown products of DFO- and placebo-cellulose before and after gamma sterilization applying headspace gas chromatography/mass spectrometry with a detection limit of 1–50 ppm. Three hundred and fifty milligrams of cellulose were equilibrated in gas-tight 20 ml vials at 80°C for 1 h. Thereafter analysis of 1 ml of the resulting headspace atmosphere was performed.
Determination of the amount of DFO coupled to cellulose (iron pick-up assay) and measurement of the kinetics of iron binding
In order to determine the amount of functional DFO coupled to cellulose an indirect method was established. Accordingly, 15 mg of cellulose (placebo or verum) were incubated in 30 ml of a 150 M FeSO4 solution (pH 5.6) for 19 h at 25°C by mixing the content of tubes intensely on an overhead rotator (Heidolph REAX 2, Schwabach, Germany). The concentration of iron (Fe2+ and Fe3+) in the supernatants before and after addition of cellulose samples was determined by means of the ready-to-use ferrozine-based assay MPR3 (Roche Diagnostics, Mannheim, Germany). This assay does not detect DFO-bound iron. The amount of DFO coupled to cellulose was calculated on the basis of the difference in iron concentration (
iron concentration, M) as DFO chelates iron stoichiometrically. The resulting equation was: amount of immobilized DFO (mol per g) = iron binding capacity (mol per g) = iron concentration (M) 2 ml per g sample. The results were corrected by subtracting the amount of unspecifically bound iron determined with placebo-cellulose. The results were expressed in mol per cm2 by multiplying the iron binding capacity with the area weight (g per cm2).
The measurements of iron binding kinetics of gamma-sterilized DFO-cellulose were performed with slight modifications using an FeSO4 solution at a concentration of about 50 M and adding 40 ml iron solution per 15 mg cellulose sample (ratio DFO:Fe = 1.5:2.0). Samples were mixed intensely by overhead rotation at room temperature and iron concentrations were determined in the supernatants at defined time intervals.
In order to evaluate potential protein effects on the iron binding characteristics of DFO-cellulose a third setting was used: 50 mg (= 1.6 cm2) of gamma-sterilized cellulose dressing (placebo or verum) were added to 1.33 ml of a 49.6 M FeSO4 solution supplemented with 30 mg per ml bovine serum albumin (BSA) (albumin fraction V, Merck, Darmstadt, Germany). After a 24 h period of intense mixing at 25°C, the iron concentration in the supernatants was determined as described above.
Results
Iron is substantially increased in fluids derived from chronic venous leg ulcers compared to acute wounds or human plasma
Total iron levels in nine exudates from patients suffering from chronic venous leg ulcers ranged between 56.2 and 182 M with a mean of 110 M and standard deviation of 42 M. In order to compensate for different protein contents the values were related to grams of protein. Accordingly, iron levels in chronic wound fluids ranged between 1.86 and 5.7 mol per g protein with a mean of 3.71 mol per g protein and a standard deviation of 1.56 mol per g protein (Figure 1). Substantially lower iron levels were found in pooled human plasma (21.8 M Fe and 0.42 mol Fe per g protein, respectively), in acute wound exudates from blisters of four patients (35.8 
0.7 M Fe and 1.15 0.62 mol Fe per g protein, respectively), and in five drainage fluids (23.5 9.3 M Fe and 0.87 0.34 mol Fe per g protein, respectively).
Figure 1.
Total iron levels are significantly increased in chronic wound exudates compared to acute wound exudates and human plasma. Wound fluids were collected from chronic venous leg ulcers of nine patients, four patients who had undergone mastectomy, and one patient after surgical removal of a myosarcoma, and from blister fluids from four patients with bullous pemphigoid as detailed in Materials and Methods. Total iron content of each fluid and a sample of human plasma pooled from 50 donors was determined using atomic absorption spectrum analysis and expressed as mol Fe per g protein. Data are shown as mean values with standard deviation. Significance was calculated using the Wilcoxon, Mann and Whitney U test (p <0.02 – blister fluids; p <0.002 – drainage fluids, compared to chronic wound fluids).
Effective reduction of iron ions from iron solutions by the gamma-sterilized DFO-cellulose dressing
In order to assess the in vitro efficacy of the DFO-cellulose, the kinetics of iron uptake from a solution of 47 M FeSO4 by pieces of gamma-sterilized DFO-cellulose dressing was studied (DFO:Fe ratio 1.55:2.00) (Figure 2). The DFO dressing reduced the iron concentration in the BSA-free supernatant in a time-dependent manner by 50% after 6 h of incubation and 80% after 24 h. This corresponds to 93% of the total iron-binding capacity of the above specified cellulose. By contrast, no decrease in iron concentration was found up to 24 h for the same amount of control placebo dressing without DFO. When the DFO:Fe ratio was changed to 3.7:1, a 50% reduction was already achieved after 10 min, a 90% reduction after 30 min of incubation, and a 98% reduction of the initial iron concentration after 24 h of incubation. In order to demonstrate the ability of the iron pick-up dressing to remove iron ions also in the presence of protein, placebo or DFO-coupled cellulose was incubated in a BSA-containing FeSO4 solution with a protein:cellulose ratio of 25 mg per cm2. This test design mimics the conditions in chronic wound fluids concerning protein, total iron concentration, and the maximal volume of exudate produced in leg ulcers, which is 1.2 ml per cm2 wound area in 24 h (Thomas et al, 1996). After a 24 h period of incubation, the initial iron concentration was reduced in the presence of BSA alone by 23%, in the placebo-containing sample by 25%, and in case of the DFO-coupled dressing by 99.5%.
Figure 2.
Effective reduction of iron ions in iron solution by gamma-sterilized DFO-cellulose. In order to determine the iron binding capacity of active functional DFO-coupled cellulose and placebo-cellulose, we have used the iron pick-up assay as detailed in Materials and Methods. Briefly, 15 mg of cellulose sample (placebo or DFO verum) were added to 40 ml FeSO4 solution at an initial concentration of 47 M (ratio DFO:Fe = 1.5:2.0). The concentration of iron was determined in the supernatants before and after addition of cellulose samples as detailed in Materials and Methods. The results were expressed as mol per l. Three independent experiments were performed in triplicate. Each point represents the mean value with standard deviation.
DFO and DFO-cellulose withstand gamma irradiation
Gamma- sterilized DFO-cellulose was subjected to two different extraction solvents and these extracts were analyzed using the highly sensitive method of LC/MS. Neither DFO nor its breakdown products could be detected. Furthermore, no change in the chemical structure of gamma-irradiated dry pulverized DFO was detectable, which would be identified by a different mass profile. In addition, no volatile breakdown products of DFO-cellulose before and after gamma sterilization were found. The iron-binding capacity of DFO-cellulose decreased negligibly after gamma irradiation as determined by means of the iron pick-up assay (data not shown).
Dissolved and cellulose-coupled DFO reduced the iron-driven induction of interstitial collagenase (MMP-1)
We here established an in vitro model closely simulating the Fenton and Haber-Weiss reactions occurring in chronic venous leg ulcers. For this purpose human dermal fibroblasts were incubated with 220 M Fe(III) citrate and 1 mM ascorbate in the presence of DFO (220 M) dissolved in PBS, an equimolar amount of DFO coupled to cellulose, and control cellulose without DFO as detailed in Materials and Methods. We have used Fe(III) citrate concentrations in this in vitro model that relate to the upper range of Fe concentrations found in fluid from chronic wounds. Exposure of human dermal fibroblast monolayer cultures to Fe(III) citrate and ascorbate resulted in a 3-fold induction of the MMP-1 protein concentration compared to the mock-treated control. Exogenous addition of H2O2 did not further enhance this induction, most probably indicating that intracellular H2O2 levels are sufficient to drive the Fenton reaction (data not shown). This induction of MMP-1 could be suppressed by 75% in the presence of dissolved DFO at a concentration of 220 M (Figure 3a). Similar results were observed in the presence of 10%-serum-containing medium during the incubation with Fe(III) citrate (data not shown) indicating that the effect of Fe(III) citrate on MMP-1 induction is independent of serum. DFO-coupled cellulose suppressed the MMP-1 induction in human dermal fibroblasts by 80% of the control cellulose. Interestingly, addition of DFO-cellulose even at 2 h after initiating the Fenton reaction with Fe(III) citrate and ascorbate resulted in an almost complete inhibition of the MMP-1 induction indicating that iron-driven effects are at least in part responsible for the observed effects (Figure 3b).
Figure 3.
Dissolved and cellulose-coupled DFO suppresses the iron-driven induction of interstitial collagenase (MMP-1). Confluent monolayer cultures of human dermal fibroblasts were exposed for 4 h to a combination of Fe(III) citrate at a concentration of 220 M and ascorbate at a concentration of 1 mM in the absence and presence of dissolved DFO (220 M) (A), an equimolar amount of DFO immobilized to cellulose, or control cellulose without DFO (B) as detailed in Materials and Methods. Thereafter, cells were washed and after an incubation period of 24 h supernatants were collected and subjected to an MMP-1 specific ELISA. Three independent experiments have been performed in triplicate. Each bar represents the mean of a representative experiment with standard deviation. *p =0.0002 (A); *p =0.0247, #p =0.0026 (B) compared with cells exposed to Fe(III) citrate/ascorbate (Student's t test).
Dissolved and cellulose-coupled DFO reduced the iron-driven increase in lipid peroxidation
Using the in vitro setting with Fe(III) citrate and ascorbate to simulate conditions of chronic venous leg ulcers, we have studied the effect of dissolved and cellulose-coupled DFO on lipid peroxidation. The Fe(III) citrate, ascorbate driven Fenton reaction resulted in a 5-fold increase in MDA, the major product of lipid peroxidation, as determined by HPLC based methods. This increase in MDA levels could be significantly reduced in the presence of dissolved DFO (Figure 4a) or the pick-up dressing coupled with DFO (Figure 4b) clearly indicating the functionality of dissolved and cellulose-coupled DFO.
Figure 4.
Dissolved and cellulose-coupled DFO reduces the iron-driven increase in lipid peroxidation. Confluent monolayer cultures of human dermal fibroblasts were exposed to a combination of Fe(III) citrate at a concentration of 220 M and ascorbate at a concentration of 1 mM in the absence and presence of dissolved DFO (220 M) (A), an equimolar amount of DFO immobilized to cellulose, and control cellulose without DFO (B) as detailed in Materials and Methods. The major product of lipid peroxidation, MDA, was measured in cell lysates using established procedures as detailed in Materials and Methods. MDA concentrations are expressed per mg protein. Three independent experiments in triplicate were performed to determine MDA concentrations. Each bar represents the mean of a representative experiment with standard deviation. *p =0.0019 (A); *p =0.0135 (B) compared to Fe(III) citrate/ascorbate treated cells (Student's t test).
DISCUSSION
Venous leg ulcers are common and cause considerable morbidity in the adult population. As healing may be slow or may never be achieved, ulcers create persistent and substantial demands on resources. A wide variety of dressings have been used, but to date there is no dressing on the market that can claim to significantly enhance tissue repair. This may at least partly be due to the limited knowledge on the microenvironment of chronic wounds. In fact, the tremendous impact of the microenvironmental conditions on the outcome of wound healing has increasingly become apparent. Several comprehensive studies demonstrate that the proteolytic activity of matrix metalloproteinases and serine proteases is significantly upregulated in the exudate of chronic wounds (Grinnell et al, 1992;Wysocki et al, 1993;Rao et al, 1995;Grinnell and Zhu, 1996;Mast and Schultz, 1996;Weckroth et al, 1996;Herrick et al, 1997;Saarialho-Kere, 1998), thus contributing to the degradation of the extracellular matrix (Grinnell et al, 1992), growth factors (Wlaschek et al, 1997;Yager et al, 1997;Lauer et al, 2000) and potentially their receptors.
In this study we have first focused on the identification of potential deleterious factors that contribute to the prooxidant hostile microenvironment of chronic leg ulcers. Based on our findings with enhanced iron levels in the exudates of chronic wounds, we have designed and developed a novel iron chelating and reactive oxygen scavenging dressing with DFO-coupled cellulose and have subsequently evaluated this iron-selective pick-up dressing for its efficacy and functionality in several in vitro assays closely simulating prooxidant conditions in chronic venous leg ulcers.
We report here that total iron concentrations in exudates from chronic venous ulcers are significantly elevated compared to iron concentrations in acute wound fluids. This is not unexpected, as increased iron levels in ulcer tissue have been reported earlier (Ackerman et al, 1988). Moreover, it is known that ROS and proteases are able to release iron from different iron-containing proteins (Thomas et al, 1985;Biemond et al, 1988;Fridovich, 1995) most probably enhancing the transfer of iron from the tissue into the exudate. Due to low available amounts of ulcer fluid, only total iron contents have been measured. Using potentiometric stripping analysis that allows a distinction between different species of iron complexes,Liyanage et al (1996) found that in not further specified wounds almost no free iron ions can be detected. According to these authors, a considerable amount, 18%, of iron binds to citrate naturally occurring in plasma and presumably in wound fluid, too. Because the ferric iron binding constant of DFO (1031) is much higher than that of citrate (1010.5) (Königsberger et al, 2000), DFO should be able to preferentially extract ferric iron from citrate-coupled complexes. Furthermore, it is well documented that DFO is able to extract and subsequently bind iron ions from ferritin and in part also from transferrin (Keberle, 1964).
Increased iron levels in the exudate of chronic wounds and its iron-catalyzed toxicity with the generation of the highly aggressive hydroxyl radical may causally contribute to severe tissue damage most extensively observed in superficial layers of the ulcer base. The causal role of the iron-driven generation of hydroxyl radicals in venous ulceration and its persistence is substantiated by a clinical study with 133 enrolled patients that convincingly demonstrates that topical application of powder containing hydroxyl radical scavenging dimethylsulfoxide or the xanthine oxidase inhibitor allopurinol significantly stimulates the healing rate of venous leg ulcers (Salim, 1991). This author did not identify increased iron levels to be involved in the pathogenic sequence underlying impaired tissue repair, however.
In fact, the detrimental effects of increased free iron and its damaging effects have clearly been shown in primary and secondary hemochromatosis, thalassemia, and congenital hemolytic anemia, particularly in the heart. A significant positive correlation between the extent of cardiac iron deposits and cardial dysfunction and failure has been documented (Buja and Roberts, 1971;Zeimer et al, 1978;Motulsky, 1985). Furthermore, increased iron levels are thought to play a causal role in the progression of osteoarthritis (Okazaki et al, 1981;Blake et al, 1984), injury of the spinal cord (Zhang et al, 1996), and extrinsic skin aging (Bissett et al, 1990;Brenneisen et al, 1998). Interestingly, experimental and clinical studies demonstrate that iron chelation by DFO substantially protects the myocardial fibers against iron-dependent damage (Grisaru et al, 1986) and decreases necrosis in dorsally localized pig skin flaps (Weinstein et al, 1989). As iron in conjunction with H2O2 and O2– generated by activated neutrophils and – under hypoxic conditions – by xanthine oxidase most probably drives the toxicity underlying persistent ulceration, we have set out to design an iron chelating dressing with antioxidant properties potentially disrupting this detrimental sequence of events. Among several iron chelators, DFO is the most specific and potent, with a stability constant of 1031 (Keberle, 1964). Furthermore, DFO has been extensively studied in a variety of experimental and clinical settings with well-established efficacy and toxicity profiles (BeDell and Hulbert, 1999;Dollery, 1999).
We report here on the activation of functional groups of the cellulose dressing and successful covalent coupling of DFO to a cellulose dressing. As a result, approximately 2 mol DFO were covalently bound per square centimeter of cellulose dressing, providing an iron complexing capacity sufficient to cope with the iron levels found in chronic venous leg ulcers. Furthermore, it could be shown that DFO-cellulose is able to bind iron out of a BSA-containing iron solution as an artificial chronic wound fluid. As stability after gamma sterilization represents an important prerequisite for a dressing, leakage of gamma-sterilized DFO-coupled cellulose was analyzed. No leakage of DFO or its breakdown products from the dressing was detected. This is in line with data from the literature on the extraordinary stability of urethane bonds (Bethell and Ayers, 1981).
As connective tissue breakdown and lipid peroxidation represent pathogenic hallmarks preventing healing of chronic venous leg ulcers, we established a model that closely simulates the iron-driven Fenton reaction. Using this model system, we have screened dissolved and cellulose-bound DFO for their protecting properties from Fe(III) citrate/ascorbate driven upregulation of MMP-1, the major metalloproteinase in connective tissue breakdown, and enhanced lipid peroxidation. We found that exposure of fibroblasts to the Fenton reaction mimicking combination of 220 M Fe(III) citrate and 1 mM ascorbate resulted in a 4-fold induction of MMP-1 levels. This induction was almost completely suppressed by dissolved DFO at a concentration of 220 M or the equimolar amount of DFO covalently coupled to cellulose.
In addition, the Fe(III) citrate/ascorbate driven Fenton reaction resulted in a 5-fold increase in MDA, the major product of lipid peroxidation. The increase in MDA levels could be significantly reduced in the presence of dissolved DFO or DFO-coupled cellulose. Collectively, these data indicate that DFO in its dissolved or bound form effectively prevents the development of a prooxidant microenvironnment and, furthermore, protects from unfavorable consequences like MMP-1 and lipid peroxide induction. We anticipate that the protective effect of DFO by the above mentioned mechanisms limits tissue injury and the perpetuated recruitment of activated neutrophils characteristic for chronic venous leg ulcers. In fact, in a TPA (12-O-tetradecanoyl-phorbol-13-acetate) induced inflammation model, topical application of DFO prior to TPA treatment could almost completely inhibit the neutrophil influx into the interstitial tissue of mouse ears (Soybir et al, 1996).
Based on our promising results a double blind placebo controlled clinical study has been started to prove the principle and assess the therapeutic value for patients suffering from nonhealing wounds.
Notes
2 Meyer-Ingold W, Eichner W, Ettner N, Schink M: The composition of chronic wound fluid can be modulated by a dressing. Wound Rep Reg 7, A 281, 1999 (abstr.)
References
| 1. | Ackerman Z, Seidenbaum M, Loewenthal E & Rubinow A. Overload of iron in the skin of patients with varicose ulcers. Possible contributing role of iron accumulation in progression of the disease. Arch Dermatol (1988) 124: 1376–1378. | Article | PubMed | ISI | ChemPort | |
| 2. | Bayreuther K, Francz PI & Rodemann HP. Fibroblasts in normal and pathological terminal differentiation. Aging, Apoptosis Transformation Arch Geront Geriatr (1992) Suppl.3 47–74. | ChemPort | |
| 3. | BeDell LS & Hulbert MK eds. Physicians GenRx – Mosby's Complete Drug Reference (1999) 9th edn. St. Louis, MO: Mosby-Year Book. |
| 4. | Bethell GS & Ayers JS. Investigation of the activation of various insoluble polysaccharides with 1,1'-carbonyldiimidazole and of the properties of the activated matrices. J Chromat (1981) 219: 361–372. | ISI | ChemPort | |
| 5. | Biemond P, Swaak AJG, van Eijk HG & Koster JF. Superoxide dependent iron release from ferritin in inflammatory diseases. Free Radic Biol Med (1988) 4: 185–198. | Article | PubMed | ISI | ChemPort | |
| 6. | Bissett DL, Chatterjee R & Hannon DP. Photoprotection effect of superoxide scavenging antioxidants against ultraviolet radiation-induced chronic skin damage in the hairless mouse. Photodermatol Photoimmunol Photomed (1990) 7: 56–62. | PubMed | ISI | ChemPort | |
| 7. | Blake DR, Gallagher PH & Potter AR. The effect of synovial iron on the progression of rheumatoid disease. Arthritis Rheum (1984) 27: 495–501. | PubMed | ISI | ChemPort | |
| 8. | Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem (1976) 72: 248–254. | Article | PubMed | ISI | ChemPort | |
| 9. | Brenneisen P & Wenk J et al. Central role of ferrous/ferric iron in the ultraviolet B irradiation-mediated signaling pathway leading to increased interstitial collagenase (matrix-degrading metalloprotease (MMP-1) and stromelysin-1 (MMP-3) mRNA levels in cultured human dermal fibroblasts. J Biol Chem (1998) 273: 5279–5287. | Article | PubMed | ISI | ChemPort | |
| 10. | Buja LM & Roberts WC. Iron in the heart: ethiology and clinical significance. Am J Med (1971) 51: 209–221. | Article | PubMed | ISI | ChemPort | |
| 11. | Cheatle T. Venous ulceration and free radicals. Br J Dermatol (1991) 124: 508. | PubMed | ISI | ChemPort | |
| 12. | Dollery C ed. Therapeutic Drugs CD-ROM Database Release 1.0 (1999) London: Churchill Livingstone – Harcourt Brace Ltd. |
| 13. | Edwards JV, Batiste SL, Gibbins EM & Goheen SC. Synthesis and activity of NH2- and COOH-terminal elastase recognition sequences on cotton. J Peptide Res (1999) 54: 536–543. | ISI | ChemPort | |
| 14. | Falanga V & Eaglstein WH. The "trap" hypothesis of venous ulceration. Lancet (1993) 341: 1006–1008. | Article | PubMed | ISI | ChemPort | |
| 15. | Feldhoff PW. Comparison of coupling procedures for development of affinity membranes: opimization of the CDI method. Techniques in Protein Chemistry III (1992) London: Academic Press 151–160. | ChemPort | |
| 16. | Fleischmajer R, Perlish JS, Krieg T & Timpl R. Variability in collagen and fibronectin synthesis by scleroderma fibroblasts in primary culture. J Invest Dermatol (1981) 76: 400–403. | Article | PubMed | ISI | ChemPort | |
| 17. | Fridovich I. Superoxide radical and superoxide dismutases. Annu Rev Biochem (1995) 64: 97–112. | Article | PubMed | ISI | ChemPort | |
| 18. | Green LM, Reade JL & Ware CF. Rapid colorimetric assay for cell viability: application to the quantitation of cytotoxic and growth inhibitory lymphokines. J Immunol Methods (1984) 70: 257–268. | Article | PubMed | ISI | ChemPort | |
| 19. | Grinnell F & Zhu M. Fibronectin degradation in chronic wounds depends on the relative levels of elastase,  1-proteinase inhibitor, and 2-macroglobulin. J Invest Dermatol (1996) 106: 335–341. | Article | PubMed | ISI | ChemPort | |
| 20. | Grinnell F, Ho CH & Wysocki A. Degradation of fibronectin and vitronectin in chronic wound fluid: analysis by cell blotting, immunoblotting, and cell adhesion assays. J Invest Dermatol (1992) 98: 410–416. | Article | PubMed | ISI | ChemPort | |
| 21. | Grisaru D, Goldfarb AW & Gotsman MS. Deferoxamine improves left ventricular function in beta-thalassemia. Arch Intern Med (1986) 146: 2344–2349. | Article | PubMed | ISI | ChemPort | |
| 22. | Gutteridge JMC, Richmond R & Halliwell B. Inhibition of the iron-catalysed formation of hydroxyl radicals from superoxide and of lipid peroxidation by desferrioxamine. Biochem J (1979) 184: 469–472. | PubMed | ISI | ChemPort | |
| 23. | Halliwell B & Gutteridge JMC. Free Radicals in Biology and Medicine (1989) 2nd edn. Oxford: Clarendon Press. |
| 24. | Hearn MTW. 1,1'-Carbonyldiimidazole-mediated immobilization of enzymes and affinity ligands. Meth Enzymol (1987) 135: 102–117. | PubMed | ISI | ChemPort | |
| 25. | Herrick S, Ashcroft G, Ireland G, Horan M, McCollum C & Ferguson M. Up-regulation of elastase in acute wounds of healthy aged humans and chronic venous leg ulcers are associated with matrix degradation. Lab Invest (1997) 77: 281–288. | PubMed | ISI | ChemPort | |
| 26. | Keberle H. The biochemistry of desferrioxamine and its relation to iron metabolism. Ann N Y Acad Sci (1964) 119: 758–768. | PubMed | ISI | ChemPort | |
| 27. | Königsberger LC, Königsberger E, May PM & Hefter GT. Complexation of iron(III) and iron(II) by citrate. Implications for iron speciation in blood plasma. J Inorg Biochem (2000) 78: 175–184 10.1016/s0162-0134(99)00222-6. | PubMed | |
| 28. | Lauer G & Sollberg S et al. Expression and proteolysis of vascular endothelial growth factor is increased in chronic wounds. J Invest Dermatol (2000) 115: 12–18 10.1046/j.1523-1747.2000.00036.x. | Article | PubMed | ISI | ChemPort | |
| 29. | Liyanage JA, Taylor DM & Williams DR. Chemical speciation in wound fluid leading to enhanced bioavailability for healing. Chem Speciation Bioavailability (1996) 8: 45–52. | ISI | ChemPort | |
| 30. | Mast BA & Schultz GS. Interactions of cytokines, growth factors, and proteases in acute and chronic wounds. Wound Rep Reg (1996) 4: 411–420. |
| 31. | Meyer-Ingold W, Eichner W, Ettner N & Schink M. Wound coverings for removal of interfering factors from wound fluid. Patent Application EP 0 945 144 A2 (1998). |
| 32. | Monboisse JC & Borel JP. Oxidative damage to collagen. In: Emerit, I, Chance, B: eds.Free Radicals and Aging. Experimentia (Suppl 62) (1992) Basel: Birkhäuser Verlag pp 323–327. | ChemPort | |
| 33. | Motulsky AG. Hemochromatosis (iron storage disease). In: Wyngaarden, JB, Smith, LH, eds.Cecil's Textbook of Medicine (1985) 17th edn. Philadelphia: W.B. Saunders 1160–1163. |
| 34. | Okazaki I, Brinckerhoff CE & Sinclair JF. Iron increases collagenase production by rabbit synovial fibroblasts. J Lab Clin Med (1981) 97: 396–402. | PubMed | ISI | ChemPort | |
| 35. | Rao CN, Ladin DA, Liu YY, Chilukuri K, Hou ZZ & Woodley DT. 1-Antitrypsin is degraded and non-functional in chronic wounds but intact and functional in acute wounds: the inhibitor protects fibronectin from degradation by chronic wounds fluid enzymes. J Invest Dermatol (1995) 105: 572–578. | Article | PubMed | ISI | ChemPort | |
| 36. | Rosner K, Ross C, Karlsmark T, Petersen AA, Gottrup F & Vejlsgaard GL. Immunohistochemical characterization of the cutaneous cellular infiltrate in different areas of chronic leg ulcers. APMIS (1995) 103: 293–299. | PubMed | ISI | ChemPort | |
| 37. | Saarialho-Kere UK. Patterns of matrix metalloproteinase and TIMP expression in chronic ulcers. Arch Dermatol Res (1998) 290: S47–S54. | PubMed | ISI | ChemPort | |
| 38. | Salim AS. The role of oxygen-derived free radicals in the management of venous (varicose) ulceration: a new approach. World J Surg (1991) 15: 264–269. | Article | PubMed | ISI | ChemPort | |
| 39. | Schaich KM. Metals and lipid oxidation. Contemporary issues. Lipids (1992) 27: 209–218. | PubMed | ISI | ChemPort | |
| 40. | Soybir GR, Koyuneu H & Koksoy F. Protective effect of desferrioxamine against TPA caused inflammation in CD-1 mouse skin. Surg Oncol (1996) 5: 253–258. | PubMed | ISI | ChemPort | |
| 41. | Thomas CE, Morehouse LA & Aust SD. Ferritin and superoxide-dependent lipid peroxidation. J Mol Biol (1985) 260: 3275–3280. | ChemPort | |
| 42. | Thomas S, Fear M, Humphreys J, Disley L & Waring MJ. The effect of dressings on the production of exudate from venous leg ulcers. Wounds (1996) 8: 145–150. | ISI | |
| 43. | Weckroth M, Vaheri A, Lauharanta J, Sorsa T & Konttinen YT. Matrix metalloproteinases, gelatinase and collagenase, in chronic leg ulcers. J Invest Dermatol (1996) 106: 1119–1124. | Article | PubMed | ISI | ChemPort | |
| 44. | Weinstein GS, Maves MD & McCormack ML. Deferoxamine decreases necrosis in dorsally based pig skin flaps. Otolaryngol Head Neck Surg (1989) 101: 559–561. | PubMed | ISI | ChemPort | |
| 45. | Weiss SJ. Tissue destruction by neutrophils. N Engl J Med (1989) 320: 365–376. | PubMed | ISI | ChemPort | |
| 46. | Wenner A, Leu HJ, Spycher M & Brunner U. Ultrastructural changes of capillaries in chronic venous insufficiency. Expl Cell Biol (1980) 48: 1–14. |
| 47. | Wlaschek M, Peus D, Achterberg V, Meyer-Ingold W & Scharffetter-Kochanek K. Protease inhibitors protect growth factor activity in chronic wounds. Br J Dermatol (1997) 137: 646–663. | PubMed | ISI | ChemPort | |
| 48. | Wysocki AB, Staiano-Coico L & Grinnell F. Wound fluid from chronic leg ulcers contains elevated levels of metalloproteinases MMP-2 and MMP-9. J Invest Dermatol (1993) 101: 64–68. | Article | PubMed | ISI | ChemPort | |
| 49. | Yager DR, Chen SM, Ward SI, Olutoye OO, Diegelmann RF & Cohen IK. Ability of chronic wound fluids to degrade peptide growth factors is associated with increased levels of elastase activity and diminished levels of proteinase inhibitors. Wound Rep Reg (1997) 5: 23–32. |
| 50. | Yu LW, Latriano L, Duncan S, Hartwick RA & Witz G. High-performance liquid chromatography analysis of the thiobarbituric acid adducts of malonaldehyde and trans,trans-muconaldehyde. Anal Biochem (1986) 156: 326–333. | Article | PubMed | ISI | ChemPort | |
| 51. | Zeimer R, Belkin M & Leitersdorff E. A non-invasive method for the evaluation of tissue iron deposition in beta thalassemia major. J Lab Clin Med (1978) 91: 24–31. | PubMed | ISI | ChemPort | |
| 52. | Zhang JR, Scherch HM & Hall ED. Direct measurement of lipid hydroperoxides in iron-dependent spinal neuronal injury. J Neurochem (1996) 66: 355–361. | PubMed | ISI | ChemPort | |
Acknowledgments
We gratefully acknowledge the stimulating discussion with Cecilia Welling.
