|Home | About | Journals | Submit | Contact Us | Français|
In the event of a radionuclear attack or nuclear accident, the skin would be the first barrier exposed to radiation, though skin injury can progress over days to years following exposure. Chronic oxidative stress has been implicated as being a potential contributor to the progression of delayed radiation-induced injury to skin and other organs. To examine the causative role of oxidative stress in delayed radiation-induced skin injury, including impaired wound healing, we tested a synthetic superoxide dismutase (SOD)/catalase mimetic, EUK-207, in a rat model of combined skin irradiation and wound injury. Administered systemically, beginning 48 h after irradiation, EUK-207 mitigated radiation dermatitis, suppressed indicators of tissue oxidative stress, and enhanced wound healing. Evaluation of gene expression in irradiated skin at 30 days after exposure revealed a significant upregulation of several key genes involved in detoxication of reactive oxygen and nitrogen species. This gene expression pattern was primarily reversed by EUK-207 therapy. These results demonstrate that oxidative stress plays a critical role in the progression of radiation-induced skin injury, and that the injury can be mitigated by appropriate antioxidant compounds administered 48 h after exposure.
Cutaneous radiation syndrome (CRS) may occur after total or partial body exposure to gamma radiation which penetrates deep into underlying tissue. It can also result from exposure to high-energy beta radiation which usually does not penetrate sufficiently deep to cause hematopoietic, gastrointestinal or neurovascular injury. CRS is an important concern for subjects exposed during a radiological accident or terrorist attack (Peter, 2005). Experience following the Chernobyl nuclear plant accident showed the impact of skin injury on patient prognosis. Almost half of the exposed individuals suffered from CRS and nearly 50% of these died with primary cause of death attributed to CRS (Mettler et al., 2007). Moreover, radiation dermatitis is a common consequence of radiation cancer therapy, and can be followed months later with atrophy, fibrosis, or telangiectasia (Ryan, 2012). It is well documented that cutaneous radiation exposure impairs wound healing (Jourdan et al.; Liu et al., 2005; Riedel et al., 2005; Schwentker et al., 1998; Wang et al., 2006).
Growing evidence links oxidative stress to the skin injury following acute radiation exposure (Robbins and Zhao, 2004; Zhao et al., 2007). We addressed the hypothesis that an antioxidant compound with appropriate properties would show benefits in both acute and chronic models of cutaneous radiation injury. We employed EUK-207, one of a class of synthetic compounds, salen Mn complexes, that mimic the antioxidant enzymes SOD and catalase, scavenging the reactive oxygen species (ROS) superoxide, O2−, and hydrogen peroxide, H2O2 (Doctrow et al., 2005; Doctrow et al., 2002), and reactive nitrogen species (Sharpe et al., 2002). Consistent with their catalase activity, salen Mn complexes are peroxidase mimetics (Doctrow et al., 2002), further broadening their potential to scavenge hydroperoxides and otherwise modulate the cellular redox environment. Such properties confer advantages over other antioxidants, such as noncatalytic or protein-based agents (Doctrow, 2003; Doctrow et al., 1997).
Prototype salen Mn complexes EUK-8, EUK-134 and EUK-189 are cytoprotective in various experimental systems (Doctrow et al., 2005; Doctrow et al., 2002; Halliwell and Gutteridge, 2007). EUK-207 is a newer generation cyclized salen Mn complex that has catalytic properties equivalent to those of EUK-134 and EUK-189, but greater stability and in vivo half-life (Doctrow et al., 2005; Liu et al., 2003; Rosenthal et al., 2011). The structure of EUK-207, with Mn bound to a poly-ether cyclized salen ligand, has been previously reported (Rosenthal et al., 2009; Liu et al., 2003). EUK-207 mitigates delayed radiation injury to the lung (Gao et al., 2012; Mahmood et al., 2011; Rosenthal et al., 2011) and kidney (Rosenthal et al., 2011) in rats, protects murine hearts from cardiac ischemia-reperfusion (Liesa et al., 2011), and improves age-associated cognitive impairment in mice (Clausen et al., 2010; Liu et al., 2003). In many of these efficacy models, salen Mn complexes are not only functionally protective, but also suppress oxidative modifications of proteins, lipids and nucleic acids (Clausen et al., 2010; Gonzalez et al., 1995; Jung et al., 2001; Liesa et al., 2011; Liu et al., 2003; Mahmood et al., 2011; Rong et al., 1999; Zhang et al., 2004). Salen Mn complexes are of further interest, as compared to other synthetic antioxidants, because of their "mitoprotective" properties in experimental models for mitochondrial injury (Doctrow et al., 2012; Doctrow et al., 2005; Liesa et al., 2011; Melov et al., 2001; Rosenthal et al., 2011).
We developed an animal model of combined radiation and wound injury to the skin where radiation-induced skin injury affected approximately 10% of total body surface, radiation did not penetrate deep into the tissues, and radiation exposure was accompanied by two full-thickness skin wounds (Jourdan et al., 2011). Rats treated under this combined injury protocol developed, in a radiation dose-dependent manner, acute radiation dermatitis spanning in severity from transient erythema to non-healing ulcers, as well as markedly impaired wound healing. Using this model, EUK-207 was tested as a potential mitigating drug on endpoints relevant to radiation dermatitis, skin wound healing, and chronic oxidative stress.
For this study, we employed a 30 Gy radiation dose that, without drug treatment, induced severe radiation dermatitis within 17–21 days after irradiation and ulcers that failed to heal over the 90 day observation period (Jourdan et al., 2011). Unanesthetized rats (n=48) were randomly divided into two experimental groups, irradiated, and given whole thickness wounds as described in Methods. Control animals (n=14) were sham irradiated, and wounded in the same manner. EUK-207 (1.8 mg/kg-day) or vehicle (water) was given by subcutaneous infusion beginning 48 hours after irradiation and continuing for up to 90 days. This delayed time was selected because, in a mass casualty scenario, therapies might be unavailable until long after radiation exposure. Radiation dermatitis was scored weekly as described in Methods.
The EUK-207-treated group showed reduced radiation dermatitis severity by 30 days post-irradiation (Figure 1a, 1b). Skin injury scores continued to improve, remaining significantly lower than in vehicle-treated rats (p<0.01; Figure 1b). At 90 days after irradiation, EUK-207-treated rats had only mild alopecia with multiple hairs growing in the center of radiation field. In contrast, vehicle-treated rats had permanent alopecia, persistent erythema, and non-healing ulcers (Figure 1b).
At 21 days after wounding, EUK-207 treated rats showed significantly smaller wounds than vehicle-treated (32% versus 58% of original wound size; p<0.05) (Figure 2a). Wounds in the EUK-207 treated group became completely healed within 35±4 days (data not shown). This is in contrast to the wounds in the vehicle-treated group which, like those in the originally-described model (Jourdan et al., 2011), did not fully heal during the duration of the study.
Since new blood vessels are an integral part of granulation tissue needed for the wound healing process, we evaluated blood vessel density as described in Methods. The number of blood vessels per optical field was decreased in irradiated skin, as early as 7 days following irradiation, as compared to in non-irradiated control animals (9.8±2.9 vs. 25.3±4.6; p<0.0001). The lower blood vessel density remained at the 14 day time point (9.4±3.7), and was further declined (3.3±2.3) at 30 days after irradiation. The blood vessel density in tissue samples taken from the wound edge 30 days after irradiation were dramatically increased with EUK-207 treatment (29.8±14.3 vs. 3.3±2.3; p<0.006) (Figure 2b and 2c).
Based on several histological indicators, the skin in the EUK-207 treated group displayed reduced injury and a more normalized phenotype (Figure 3a). These included reduction of dermal thickness at 30 and 90 days (p<0.005), increased epidermal thickness at 30 days (p<0.005), and restoration of hair growth signified by increased number of hair follicles (p<0.03; Figure 3b).
Chronic oxidative stress has been implicated in the progression of radiation-induced late effects, potentially via activation or repression of genes in important signaling pathways, but the specific genes involved in radiation-induced skin injury are ill-defined. We therefore obtained gene expression data on skin samples taken 30 days after irradiation. To evaluate gene expression patterns relevant to oxidative stress, mRNA was prepared from unirradiated and irradiated (control and EUK-207 treated) rats and analyzed by microarray as described in Methods. The data indicated that irradiation caused upregulation of 15 genes, including those involved in generation or detoxification of ROS, and downregulation of genes involved in excisional DNA repair (Xpa) or innate immunity (Mpo) (Table 1a). These changes were abrogated in skin from irradiated rats receiving EUK-207. In these rats, there was up-regulation of only three genes from the oxidative stress pathway, all changes distinct from those in the vehicletreated group (Table 1b).
Since the gene expression data demonstrated radiation-induced dysregulation of genes responsive to oxidative stress, we examined oxidative modifications of proteins and nucleic acids to confirm the occurrence of chronic oxidative stress in irradiated skin. Protein carbonylation is a commonly used biomarker of irreversible oxidative damage to proteins (Levine, 2002). Skin extracts from the irradiated rats showed increased protein carbonyls at 30 and 90 days after irradiation, with one prominent band (MW ~68Kd) likely corresponding to albumin, which, due to abundance, frequently appears as a major carbonylated band (Levine et al., 1994; Wehr and Levine, 2012). The cumulative density of all carbonylated bands was significantly attenuated by EUK-207 treatment (p<0.05; Figure 4a and 4b). Along with chronic oxidative damage to proteins, skin from irradiated rats also exhibited evidence for nucleic acid injury. Staining for oxidized 8-hydroxyguanosine (8-OHdG), a marker for DNA oxidation, was evident in irradiated skin even at 90 days after irradiation, and substantially decreased with EUK-207 treatment (Figure 4c and 4d).
It is well known that irradiation causes the immediate generation of short-lived ROS, but the role of chronically generated ROS in longer-term damage, though hypothesized and increasingly implicated (Zhao et al., 2007; Zhao and Robbins, 2009), is not well documented. Our study demonstrates that chronic oxidative stress occurs in irradiated skin, even a month or more after exposure. Furthermore, the dramatic reduction of skin injury by a synthetic SOD/catalase mimetic, EUK-207, concomitant with its abrogation of oxidative stress indicators, demonstrates that oxidative stress is causative in both chronic radiation dermatitis and impaired wound healing in irradiated skin. Because EUK-207 was initiated 48 hours after exposure, there is no question that the ROS targeted in our study are those generated well after the initial radiation insult. Increased localized expression of known antioxidant proteins was shown to decrease cutaneous radiation injury (Yan et al., 2008; Zhang et al., 2012). We further demonstrate, through mitigation by a synthetic antioxidant agent, that oxidative stress causes both the dermatitis and wound healing impairments characteristic of delayed radiation injury. Possibly, previously tested antioxidants either lacked the appropriate ROS specificity, did not have adequate bioavailability to the skin, or both. The required bioavailability may include access to the mitochondria, since mitochondrial dysfunction has been implicated in cellular sensitivity to radiation injury (Aykin-Burns et al., 2011; Greenberger and Epperly, 2004; Jiang et al., 2009). While we have not directly tested the role of "mito-protection" in this study, salen Mn complexes suppress oxidative mitochondrial injury in other experimental models (Doctrow et al., 2012; Doctrow et al., 2005; Liesa et al., 2011; Melov et al., 2001). Among the oxidative-stress responsive genes we found to be upregulated in irradiated, vehicle-treated skin is Sod2, for the mitochondrial antioxidant enzyme MnSOD (Table 1a). Sod2 is also upregulated in other models for radiation exposure, and increased SOD expression in the mitochondria is radioprotective (Epperly et al., 2007; Greenberger and Epperly, 2007; Wong et al., 1996). Our data also showed upregulation of genes for glutathione peroxidase 1 (Gpx1) and peroxidredoxin 5 (Prdx5). Peroxidredoxin 5 has been found in the mitochondria and is regarded as a potentially important scavenger of mitochondrial ROS, particularly hydroperoxides (Cox et al., 2010) Glutathione peroxidase 1, a cytosolic enzyme, has been reported to play a role in modulating the mitochondria, through redox changes, and its upregulation has been implicated in mitochondrial dysfunction (Handy et al., 2009). Overall, an upregulation of such mitochondrially associated antioxidant enzymes is potentially indicative of oxidative mitochondrial injury. If so, then prevention of their up regulation by EUK-207 (Table 1b) is consistent with a hypothesis that the compound's mitigating effects on cutaneous radiation injury are, at least in part, related to its ability to protect the mitochondria. However, its suppression of nearly all the changes in oxidative-stress responsive genes in irradiated skin indicates that EUK-207 is not acting exclusively at the mitochondria. This agrees with previous reports showing suppression of both mitochondrial (Hinerfeld et al., 2004; Liesa et al., 2011; Liu et al., 2003; Melov et al., 2001) and non-mitochondrial (Liesa et al., 2011; Peng et al., 2005; Rong et al., 1999; Zhang et al., 2004) oxidative modifications by salen Mn complexes. Indeed, unlike MitoQ and certain other antioxidants (Demianenko et al., 2010; Murphy and Smith, 2007), the salen Mn complexes were not designed for specific mitochondrial targeting. It is interesting to note that, in the irradiated rat skin, the gene for glutathione peroxidase 2 (Gpx2) was downregulated with vehicle, yet upregulated with EUK-207 treatment. The significance of this reversal in Gpx2 expression by EUK-207 is not yet apparent, but should be studied further. While there is little literature to address its potential role in radiation injury or wound healing, glutathione peroxidase 2 has been reported to have both a cytoplasmic and mitochondrial location in yeast (Ukai et al., 2011). In a recent study using diabetic mice (Tie et al., 2012), GI-PS, a polysaccharide derived from the medicinal fungi Ganoderma lucidum, accelerated wound healing while increasing total glutathione peroxidase activity in the skin. This enzymatic activity, of course, likely represents a combination of glutathione peroxidase 1, 2 and perhaps other forms. The GI-PS also increased skin MnSOD activity, not by changing the protein's expression but, instead, by suppressing its nitration, a known mechanism of its inactivation during oxidative stress. Based on the latter finding, Tie et al. attributed the wound healing properties of GI-PS to its known antioxidant properties and, in particular, to an inhibition of mitochondrial oxidative stress. In our study, it is also of interest that EUK-207 appeared to upregulate expression of peroxidredoxin 6 (Prdx6). Studies with Prdx6 knockout mice have indicated that peroxiredoxin 6 is essential for blood vessel integrity during skin wound healing, with vascular endothelial cells appearing to be highly sensitive to the loss of this novel peroxiredoxin (Kumin et al., 2006). Thus, in our model, the ability of EUK-207 to increase Prdx6 expression may relate to the compound's beneficial effects on wound healing and angiogenesis. From our perspective, these changes in gene expression are of interest as preliminary leads for future study, including a more detailed analysis of whether EUK-207 modulates protein levels or enzymatic activities of any of these potentially key antioxidant proteins during wound healing in irradiated skin.
Based on its other reported in vivo effects in radiation injury models (Mahmood et al., 2011; Rosenthal et al., 2011), mitigation of radiation dermatitis severity by EUK-207 was not unexpected. However, improved wound healing by an ROS scavenger was not necessarily predicted from literature describing a very complex association between ROS and normal skin wound healing. ROS, particularly H2O2, are believed to play key signaling roles to promote wound healing, and localized transfection of catalase delays healing in rodents (Roy et al., 2006; Sen and Roy, 2008). Yet, excess H2O2 impairs (Roy et al., 2006) and transfection with Sod2 improves wound healing (Luo et al., 2004), as does chronic administration of a mitochondrially-targeted antioxidant (Demianenko et al., 2010). Our observation that EUK-207 treatment resulted in increased angiogenesis in the irradiated wounded skin is particularly intriguing. ROS are believed to mediate mitogenesis stimulated by growth factors including the angiogenic factor VEGF (Roy et al., 2006; Ushio-Fukai and Alexander, 2004). And, interestingly, an endogenously generated oxidized lipid product promotes angiogenesis and wound healing in a VEGF-independent manner (West et al., 2010). Thus, one might expect an antioxidant to impair rather than, as we observe, facilitate angiogenesis. It is conceivable that, in our combined injury model, by promoting a more normalized skin phenotype (e.g. Figures 1 and and3),3), EUK-207 is inducing a microenvironment that facilitates more normal wound healing, enabling tissue remodeling processes including new blood vessel growth (Gurtner et al., 2008). In support of this, the basement membrane deposition of laminin 332 is impaired in our model (Jourdan et al., 2011) and is improved with EUK-207 treatment (data not shown). More broadly, EUK-207 may modify the microenvironment through suppression of oxidation-dependent events that might otherwise destroy microvasculature in irradiated tissue. Consistent with this, EUK-207 treatment preserves the microvasculature after lung irradiation (Gao et al., 2012). Thus, overall, the role of ROS, particularly H2O2, and of redox regulation in cutaneous wound healing and its associated angiogenesis is highly complex. Despite this complexity, our study supports the concept that selected ROS-scavenging agents such as EUK-207, having the appropriate specificity and given under the right circumstances, can mitigate radiation-induced skin injury, including facilitating wound healing.
Sixty two syngeneic male WAG/RijCmcr eight-week-old rats bred and housed in a moderate-security barrier were used for this study. Rats were monitored daily and maintained on a 12-hours light/dark cycle, with free food and water intake. At the times specified below, animals were euthanized using isoflurane inhalation. All animal research was approved by the Medical College of Wisconsin Institutional Animal Care and Use Committee.
We employed our previously published protocol (Jourdan et al., 2011). Unanesthetized rats were immobilized and irradiated with an x-ray beam with a steep dose gradient in the dorso-ventral direction (without injury to the internal organs). The source-to-skin distance was 37 mm and the dose rate was 0.68 Gy per minute. The irradiated area of skin corresponded to 10% of the total body surface. Animals were randomly divided into two experimental groups and either sham irradiated (n=14) or irradiated with a single dose of 30 Gy defined at the dermal layer (n=48). Within one hour following irradiation, all rats were anesthetized, and two full-thickness wounds were made on the back of each rat within the irradiation field using an 8 mm punch biopsy. The wounds were left uncovered and animals were housed individually to prevent damage to the wound site.
The custom synthesized EUK-207, (Liesa et al., 2011) was dissolved in ultrapure water, filter-sterilized and administered by subcutaneous Alzet infusion pumps (DURECT Corporation, Cupertino, CA) at 1.8 mg/kg/day beginning 48 hours after irradiation and continuing for up to 90 days. Sham-irradiated or irradiated-only animals received pumps filled with vehicle. The EUK-207 dose was about 4-fold lower than doses employed previously to mitigate radiation injury to rat lung (Mahmood et al., 2011) and kidney (Rosenthal et al., 2011). This lower dose was selected to eliminate the localized skin toxicities observed in those prior studies, while remaining well within the effective EUK- 207 dose range reported in other rodent models (e.g. Liu et al., 2003).
Photographs of animals were taken three times per week and coded. The skin injury score was assessed from coded images by two investigators in a masked fashion according to a previously published scoring system (Jourdan et al., 2011). Wound contraction was assessed at days 0, 3, 7, 14 and 21. Wound areas were not calculated at later time points due to heavy crusts covering the wounds and multiple erosions blending into the wound site. Wound area was determined as described (Jourdan et al., 2011) and wound contraction was calculated as follows:
Skin samples from four animals per experimental group (group 1: irradiated and vehicle treated; group 2: irradiated and EUK-207 treated) were harvested at 30 and 90 days after irradiation, fixed in 4% formaldehyde and embedded in paraffin. Tissue sections were stained with hematoxylin and eosin. Epidermal thickness was determined by measuring the epidermal layer and dermal thickness was determined by measuring the dermis from dermo-epidermal junction to the top of fatty layer three times in three consecutive optical fields in each skin sample using Scion image analysis software (National Institutes of Health). The number of hair follicles was determined by manually counting the hair follicles in three consecutive optical fields.
Skin samples from five animals per experimental group (group 1: irradiated and vehicle-treated; group 2: irradiated and EUK-207 treated; group 3: sham- irradiated and vehicle treated) were harvested at 30 days after irradiation. Samples were washed in ice-cold phosphate-buffered saline (Life Technologies, Carlsbad, CA) and immediately immersed in RNA Later (Life Technologies). The total RNA was prepared using RNeasy Fibrous tissue Mini Kit (Life Technologies), the RNA concentration was determined by 260:280 nm absorbance ratios, equal amounts of RNA from each sample in the experimental group were pooled and cDNA was synthesized from 1 µm of total RNA.
To determine the relative expression of genes associated with oxidative stress, quantitative real time RT-PCR analysis was performed with RT2 first strand cDNA kit (SABiosciences, Frederick, MD) and Rat Oxidative Stress and Antioxidant Defense RT2 Profiler™ PCR Array (SABiosciences). The samples were diluted in qPCR master mix and pipetted into 96- well array plates to evaluate expression of 84 oxidative stress related genes. RT-PCR was performed in technical duplicates using Applied Biosystems Step One Plus Real-Time PCR Systems (Applied Biosystems, Carlsbad, CA). Quality controls included in each plate confirmed the lack of DNA contamination and tested for successful PCR performance. For data analysis of PCR, the ΔΔCt method was used with algorithms provided by the manufacturer. Fold changes were then calculated and expressed as log-normalized ratios of values from irradiated versus sham-irradiated tissues or irradiated and EUK-207 treated tissues.
To detect the carbonyl groups we employed the Oxyblot™ Oxidized Protein Detection Kit (Chemicon International, Temecula, CA), which detects proteins containing 2,4-dinitrophenol (DNP)-derivatized carbonyl groups by immunoblotting. Skin tissue samples were rapidly frozen in liquid nitrogen. Equal amounts (100 mg) of skin tissue were each homogenized in ice cold 20 mM Tris HCl buffer containing Protease Inhibitor Cocktail (Sigma-Aldrich Corp., St. Louis, MO), incubated for 30 minutes on ice, and then centrifuged at 10,000 RPM for 20 min at 4°C. The supernatant was collected and protein concentration was determined by the Bio-Rad Protein Assay (Bio-Rad Laboratories, Hercules, CA). The 20 ug aliquots of protein extracts were analyzed according the manufacturer’s protocol. Following the Oxyblot procedure, the PVDF membrane was stripped and probed with specific anti-β actin antibodies (Santa Cruz Biotechnology Inc., Santa Cruz, CA). The protein bands on the membrane were detected using a chemiluminescence detection kit (Pierce, Rockford, IL). The intensity of the bands was quantified by densitometric analysis with NIH image J 1.43 and expressed as density relative to β actin loading control.
To further confirm the occurrence of oxidative stress in the irradiated skin at 90 days after irradiation, we evaluated DNA oxidation using immunohistochemistry. The paraffin embedded skin sections, were deparaffinated rehydrated and treated with sodium citrate buffer (10 mM sodium citrate, 0.05% Tween 20, pH 6.0), for 10 minutes at 95°C. After washing, samples were incubated with mouse anti-8-hydroxyguanosine IgG (Abcam Inc., Cambridge, MA) overnight at 4°C after biotinylation and a blocking procedure, which followed the manufacturer's protocol (Vector Laboratories Inc., Burlingame, CA). Mouse IgG2a (Abcam Inc.) served as an isotype control. After development with diaminobenzidine, the sections were counterstained with hematoxylin, coded and evaluated under a light microscope. Five consecutive images from each slide and the pixels corresponding to the stained region were measured using the NIH Image J program 1.43.
The skin and wound edge samples were embedded in OCT compound (Sakura, Japan) for immunofluorescence studies. Six-micron skin sections were incubated with anti-CD31 IgG (BD Biosciences, San Jose, CA) overnight at 4°C. The mouse IgG2a (Abcam Inc.) served as a negative control. The blood vessels were detected by FITC-conjugated goat F (ab’) 2 anti-mouse IgG (Santa Cruz Biotechnology Inc.). The slides were coded and evaluated under a fluorescent microscope. Blood vessels in five consecutive images from each slide were counted from four animals per experimental group.
For radiation dermatitis scores, differences in treatment group medians were assessed using a Wilcoxon-Mann-Whitney Rank Sum test. For histological analysis all values were expressed as mean ± SD and differences assessed using a two tailed Student’s t-test. For all other data, differences among treatment group means were assessed using a one-way ANOVA followed by a Student-Newman-Keuls post hoc test and data expressed as mean ± SD. For all analyses, p< 0.05 was considered to be statistically significant.
This work was funded by a pilot grant (ZL) under AI067734 (JEM) and Froedtert Hospital Skin Cancer Grant (ZL). Development of EUK-207 was funded in part by GM57770 (SRD).
Conflict of Interest: The authors state no conflict of interest. Dr. Doctrow is an inventor of patents describing EUK-207 and other salen Mn complexes, developed while she was employed at Eukarion, Inc. However, the company is no longer in business and she has retained no stock options or other rights.