|Home | About | Journals | Submit | Contact Us | Français|
Burn injury increases the risk of morbidity and mortality by promoting severe hemodynamic shock and risk for local or systemic infection. Graft failure due to poor wound healing or infection remains a significant problem for burn subjects. The mechanisms by which local burn injury compromises the epithelial antimicrobial barrier function in the burn margin, containing the elements necessary for healing of the burn site, and in distal unburned skin, which serves as potential donor tissue, are largely unknown. The objective of this study was to establish defects in epidermal barrier function in human donor skin and burn margin, in order to identify potential mechanisms that may lead to graft failure and/or impaired burn wound healing. In the present study, we established that epidermal lipids and respective lipid synthesis enzymes were significantly reduced in both donor skin and burn margin. We further identified diverse changes in the gene expression and protein production of several candidate skin antimicrobial peptides (AMPs) in both donor skin and burn margin. These results also parallel changes in cutaneous AMP activity against common burn wound pathogens, aberrant production of epidermal proteases known to regulate barrier permeability and AMP activity, and greater production of pro-inflammatory cytokines known to be induced by AMPs. These findings suggest that impaired epidermal lipid and AMP regulation could contribute to graft failure and infectious complications in subjects with burn or other traumatic injury.
Wound healing after burn injury ideally results in the establishment of new granulation tissue and/or wound closure, without complications such as infection and with minimal scarring. Improved outcomes for severely burned patients have been accredited to medical advances in several areas, including fluid resuscitation, local burn wound care, and infection control practices1–3. Despite these important advances, the majority of all deaths in burn subjects with a >40% total body surface area (TBSA) injury are related to sepsis from a burn wound infection or other infectious complications4–7. Burn subjects with significant deep partial-thickness and/or full-thickness burns necessitate skin grafting, due to the extensive damage or complete destruction of the underlying fascia8,9, in order to minimize invasive infection and extreme fluid shifts. Autologous grafts from distal, unburned skin often exhibit functional defects and potentially global deterioration after grafting. Despite these clinical observations, the mechanisms behind the deterioration of primary skin grafts after burn injury have not been extensively investigated. However, impaired epidermal barrier function likely serves as a mechanism for graft loss and wound healing complications (e.g. infection) in burn subjects.
Normal epidermal barrier function and wound healing responses are stimulated, in part, through the production of epidermal antimicrobial peptides (AMPs), which are a highly conserved component of the innate immune system involved in tissue repair and microbial defense10,11, and epidermal lipids, which generate an essentially impermeable barrier to the external environment12. Direct perturbations in AMP or lipid production, or in those molecules which regulate their activity (i.e. enzymes or proteases), can lead to defects in barrier permeability, impaired wound healing, and pathogen resistance. Moreover, the pathways that produce and regulate the antimicrobial barrier of the skin are intimately coupled to pathways that modulate epidermal permeability barrier function, including epidermal lipid synthesis13–15. The epidermal serine proteases, Kallikrein [KLK] 5 and 7, regulate epidermal desquamation as a part of the normal skin barrier homeostasis16 and cleave inactive AMP pro-forms into bioactive peptides17. The human AMPs cathelicidin (CAMP) and β-defensin-1 and -2 provide a first line of defense against infection and directly participate in immune and wound healing responses10,11. Production of several epidermal AMPs, including CAMP and β-defensins (hBDs), is augmented following epithelial barrier disruption, injury, or microbial stimuli13,14,18,19. Other clinically relevant inducible AMPs in the epidermis include psoriasin (S100A7)20 and ribonuclease 7 (RNAse7)21.
Following human burn injury, AMPs are reduced or absent from acute burn wounds and blister fluid compared to non-burn wounds22,23. In other studies of acute burn wounds, changes in the localization of AMPs were observed in the remaining keratin layers and in surviving dermal and subcutaneous layers, including hair follicles and sweat glands, but not in vascular endothelium or adipocytes24,25. Previous studies in human acutely burned skin and burn blister fluid revealed significant hBD reduction compared to unburned subjects22–24. Furthermore, previous studies in mice and humans have demonstrated that burn-induced myeloid cells are capable of suppressing keratinocyte AMP production26,27. In contrast to AMPs, few studies have attempted to characterize the effects of burn injury on lipid composition in surrounding tissue (burn margin) or distal skin (potential donor site). Using an established acute scald-burn mouse model, we recently determined that the epidermal antimicrobial and lipid barrier function at both the burn margin and the potential donor site is compromised compared to skin from unburned mice, leading to greater water loss from the skin and delayed barrier recovery28. Similar to the burn margin, distal unburned skin from burned mice displayed greater cutaneous protease activity and altered AMP production, which paralleled a diminished ability to inhibit the growth of several skin microbes28. This suggests that the compromised epidermal barrier from donor skin in human burn subjects may contribute to graft failure, wound infection, sepsis, or other infectious complications. Furthermore, dysregulation of AMPs and lipids following burn injury likely alters the cutaneous microbiome at potential donor sites, as bacteria that comprise the skin microbiota likely stimulate or are influenced by AMPs, lipids, and other innate immune molecules29,30.
In this study, we investigate the impact of burn injury on epidermal lipid regulation and AMP responses in human donor skin and burn margin, in order to identify potential mechanisms that may lead to graft failure and/or impaired burn wound healing.
All protocols were approved by the Institutional Review Board at Loyola University Chicago, Health Sciences Campus. A standing approval for discarded skin was used to collect the tissue samples. Burn patients admitted to the burn intensive care unit (BICU) with the following were excluded from the study: age < 18 years, pre-existing clinically-evident infection, pre-existing skin disease (i.e. eczema, psoriasis, rosacea, etc), previous transplant recipient, recent major traumatic injury (i.e. broken bones, fall, surgery, etc.) <4 months prior to the burn injury, pre-existing immunodeficiency, and/or a history of disseminated cancer. The following clinical characteristics and outcomes were obtained from the electronic medical records and entered into a database: age, gender, % total body surface area (%TBSA) injured, inhalation injury, burn injury mechanism, sepsis and/or multisystem organ dysfunction (MODS), pneumonia, urinary tract infection, graft failure, wound infection, and mortality. Injury severity was determined based on %TBSA with partial and/or full thickness burns, which was recorded in the patient’s medical records. Per the BICU standard protocol, initial fluid resuscitation was conducted according to the Parkland formula (4 mL / kg / % TBSA with half given during the first 8 hours following injury and the remaining half given over the next 16 hours). Discarded skin samples from burn patients undergoing routine excision/debridement and skin grafting were collected in the operating room. The burn margin (partial thickness) was taken from the skin adjacent to the excised area of the burn and not directly in contact with the thermal source. When the burn wounds were excised, a 5–10 mm margin of grossly normal appearing skin was excised simultaneously with the wound. The wound itself was debrided to viable tissue to allow for optimal wound healing in the patients, thus yielding viable tissue at the burn margin that was excised. Donor skin (partial thickness) was taken from a site distal to the original injury (autograft site), per standard surgical protocols. Although the same burn patient may have had multiple surgeries and thus contributed multiple samples to the study, none of the patients required repeat use of a specific donor site. Control skin samples were collected in the operating room from patients undergoing non-oncologic, elective surgeries (such as breast reduction or panniculectomy). Skin samples were either snap-frozen in liquid nitrogen, placed in Trizol® (Life Technologies), embedded in optimal cutting temperature (OCT; Tissue-Tek® O.C.T. ™ Compound), or placed in acetic acid, and were stored appropriately prior to tissue processing. For the subsequent experiments described, the N for each group was variable due to the amount of sample obtained at the operation, and thus, it was not feasible to use every sample for every assay. Sub-groups of the patient samples were utilized at the time of each assay performance and were randomly selected for individual assays while quantities were available.
ELISA was performed for interleukin (IL)-6, IL-8, hBD1, hBD2 (PeproTech), s100A7 (CircuLex), RNASE7 (CUSABIO), and LL37 (Hycult Biotech) according to each manufacturer’s protocol. Protein concentrations of skin homogenates were determined using a bicinchoninic acid (BCA) assay kit (Thermo Scientific). The ELISA results were then normalized to total protein concentrations. These analyses were performed in duplicate.
Skin samples were homogenized directly in Trizol (Life Technologies), and RNA was extracted per the manufacturer’s instructions. RNA was reverse transcribed using iScript (Bio-Rad) according to the manufacturer’s protocol. cDNA samples were analyzed to detect the expression of acetyl-CoA carboxylase alpha (ACC, Hs01046047_m1), fatty acid synthase (FAS, Hs01005622_m1), HMG-CoA reductase (HMGCR, Hs00168352_m1), serine palmitoyl transferase (SPTLC1, Hs00272311_m1), β-defensin 1 (DEFB1, Hs00608345_m1), β-defensin 4a (DEFB4A, Hs00175474_m1), cathelicidin antimicrobial peptide (CAMP, Hs00189038_m1), ribonuclease 7 (RNASE7, Hs00261482_m1), psoriasin (s100A7, Hs00161488_m1), kallikrein-5 (klk5, Hs01548153_m1), kallikrein-7 (klk7, Hs00192503_m1), IL-6 (Hs00985639_m1), IL-8 (Hs00174103_m1), and IL-22 receptor (IL-22R, Hs00364814_m1). Quantitative real-time PCR was performed using the TaqMan Gene Expression pre-mix (Life Technologies) and Applied Biosystems 7000 Sequence Detection System (Foster City, CA). To normalize the mRNA expression levels, the relative expression of beta-2-microglobulin (B2M, Hs99999907_m1) was analyzed in parallel. Results were analyzed using the 2(−ΔΔCt) method. Fold-change relative to the control group was calculated (default level set at 1). Each sample from each cohort was analyzed in duplicate.
Nile Red staining was performed as previously described to assess epidermal lipids28. In brief, skin was embedded in OCT compound, stored at −80°C, and subsequently sectioned to 8 μm samples using a cryostat. For Nile Red staining, tissue sections were expanded using half-strength Sorensen-Walbum buffer for 20 minutes. A stock solution containing 0.05% (wt/vol) Nile Red in acetone was stored at 4°C, protected from light. Prior to staining, the stock solution was diluted to 2.5 μg/mL with 75:25 (vol/vol) glycerol:water and briskly vortexed. A single drop of the glycerol-dye solution was applied to each tissue section and immediately covered with a coverslip. Samples were protected from light and incubated at 37°C for 1 hour. Sections were examined with an Evos Digital Microscope using a 20x objective. These analyses were performed in duplicate.
IHC was performed on the skin using primary antibodies for KLK5 (R&D Systems), KLK7 (R&D Systems), and appropriate secondary antibodies as previously described28. In brief, tissue samples were embedded in OCT and sectioned at 8 μm using a cryostat. Sections were fixed in acetone, incubated in blocking buffer, then incubated overnight at 4°C with the appropriate dilution of primary antibody in blocking buffer. The following day, sections were washed and incubated at room temperature with the appropriate secondary antibody conjugated to Cy3 or Alexa-fluor 456. Nuclei were stained using ProLong® Antifade Gold with DAPI (Life Technologies). Images were obtained with an Evos Digital Inverted Microscope using a 20x objective and analyzed in a blinded manner. These analyses were performed in duplicate.
Protein extraction was done by homogenizing skin and filtering the supernatant. The protein concentrations of the samples were then analyzed using a standard BCA protein assay (Thermo Scientific Pierce). Protease activity was assessed using the EnzChek Protease Assay kit (E6638 green fluorescence, Molecular Probes Inc, Life Technologies), according to the manufacturer’s instructions and as previously described17,28,31. In brief, quantitative measurements were assessed on whole protein extracts from homogenized skin samples. Trypsin (Life Technologies) with and without the protease inhibitor Dispase II (Roche) were used as controls. Protease activity was measured using pH-insensitive green fluorescent BODIPY-FL-conjugated casein as a substrate in a total reaction volume of 200 μl. All enzymes and extracts were diluted in PBS (pH 7.4). After the addition of 100 μl of substrate solution and 100 μl (up to 10 μg) of sample into 96-well plates, the reaction mixtures were incubated for 24 hours at room temperature while protected from light. Fluorescence was measured using a Synergy 2 microplate reader (BioTek Instruments, Winooski, VT, USA) at an excitation wavelength of 485 nm and an emission wavelength of 530 nm using the Gen 5 program. These analyses were performed in duplicate.
Qualitative protease activity was assessed on sectioned skin samples prepared in OCT (8 μm sections), as previously described28. In brief, skin sections were washed and incubated with the BODIPY-FL-conjugated casein substrate from the EnzChek Protease Assay kit (E6638 green fluorescence, Molecular Probes Inc, Life Technologies) at a concentration of 2 ug/ml for one hour at 37°C, protected from light. Slides were again washed, and the nuclei were stained using ProLong® Antifade Gold with DAPI (Life Technologies). Sections were examined with an Evos Digital Microscope using a 20x objective. All samples were analyzed in a blinded manner, and all experiments were performed in duplicate.
Skin samples were placed in 1N acetic acid for at least 24 hours at 4°C, homogenized, lyophilized, and then dissolved in water. HPLC peptide separation was performed using a C18 column (8 μM, ST 4.6/250; Thermo Scientific #25008–254630). Equilibration was done in 0.1% trifluoroacetic acid (TFA). Peptides were eluted with an acetonitrile gradient (10–100% over 40 minutes at 2 ml/min with fractions collected from 10–30 minutes)32. All fractions (2 ml each) were lyophilized and reconstituted in 50 μl of sterile water, and then combined into one sample (all fractions from one patient were combined into a single sample). Combined samples were again lyophilized and re-suspended in 20 μl of sterile water. Samples were then normalized for peptide content, as determined by an O.D. reading at 214 nm.
Recombined HPLC fractions were evaluated using a radial diffusion assay (RDA), as previously described28,29. Antimicrobial activity was assessed against Escherichia coli (E. coli, ATCC), Pseudomonas aeruginosa (P. aeruginosa, clinical isolate), Staphylococcus aureus (S. aureus; strain SA113; ATCC), and Group A Streptococcus (GAS; M49 strain, NZ141). Bacteria from glycerol stocks were inoculated and grown in Tryptic Soy Broth (TSB), Todd Hewitt Broth (THB), or Luria Broth (LB) at 37°C overnight, as appropriate. The overnight cultures were then diluted 1:100 and grown to the exponential phase. Thin plates (1 mm) with 1% SeaKem® GTG® Agarose (Cambrex Corporation, East Rutherford, NJ) in 0.5% tryptone containing ~4 × 105 cells/ml of bacteria were used for the assay. Sterile water and synthetic catestatin (100 μM, GeneScript) were used as the negative and positive controls, respectively. The zone of inhibition of bacterial growth (area) was quantified using ImageJ Software (National Institutes of Health, Bethesda, MD). Samples were used to perform these experiments in duplicate.
All quantitative data are described as mean ± standard error of the mean (SEM). Comparisons between control vs. margin and control vs. donor were performed using a Mann-Whitney test for all human samples. P values ≤0.05 were considered statistically significant. We did not use a One-way ANOVA, as we were not comparing donor vs. margin. For analysis of the in vitro model data, a two-way ANOVA with a Bonferroni post-test was used. A p ≤0.05 was considered statistically significant. Data analyses were calculated using GraphPad Prism (GraphPad Software v5, La Jolla, CA, USA).
Samples from 43 patients admitted to the BICU were evaluated in this study. Patients used for the analysis had characteristics that are representative of the general population treated in the BICU (Table 1). Ages of the patients ranged from 20 to 87 years (mean of 47, median of 46), burn sizes ranged from 1% to 52% TBSA (mean of 18%, median of 12.5%). On average, the burn skin samples were collected during routine surgeries (excision, debridement, and grafting) within 5 days post-burn (details provided in Table 2). Of the 43 patients studied, 28% (n = 12 patients) developed pneumonia, 42% (n = 18 patients) suffered a wound infection of the donor or burn site, and 25% (n = 11 patients) were treated for blood culture positive sepsis. Even with the high rate of complications seen in our study group, only 5 patients (11%) succumbed to their illness. The control samples were taken from healthy patients with no chronic diseases or skin disorders, with a mean age of 49 years. (Table 1)
The epidermal permeability barrier relies on the production of extracellular lipids (e.g. cholesterol, fatty acids, and ceramides) in order to minimize water loss, promote transport of electrolytes, and limit penetration of chemicals and pathogenic microorganisms12. Thus, we first assessed epidermal lipid production between our cohorts using Nile Red staining. We observed a robust decrease in epidermal lipid staining in donor skin and burn margin, relative to controls (Fig. 1A). We then assessed the gene expression of key epidermal lipid synthesis enzymes: acetyl-CoA carboxylase (ACC), fatty acid synthase (FAS), hydroxymethylglutaryl-CoA reductase (HMGCR), and serine palmitoyl transferase (SPTL)12,33,34. Both ACC (Fig. 1B) and FAS (Fig. 1C) gene expression were significantly lower in both donor skin (p < 0.001) and burn margin (p < 0.05) compared to controls. Although no significant change was observed with HMGCR (Fig. 1D), SPTL gene expression was significantly lower in both donor skin (p < 0.001) and burn margin (p < 0.05) relative to controls.
Because the lipid and AMP barriers are co-dependent15, we next assessed the gene expression of several candidate AMPs known to play a significant role in epidermal barrier function and microbial defense, and compared their expression in donor skin or burn margin relative to control skin. Both hBD1 (DefB1) and hBD2 (DefB4A) gene expression were significantly elevated in margin specimens (p < 0.025 and p < 0.001 respectively) relative to controls (Fig. 2A and B); no change was observed between controls and donor skin. Cathelicidin (CAMP) was not significantly different between cohorts, although there was a 4-fold increase in margin samples (p = 0.06) (Fig. 2C). However, psoriasin (S100A7) gene expression was significantly elevated in the margin specimens (p < 0.001) relative to controls (Fig. 2D). In contrast to the other candidate AMPs, RNASE7 gene expression was significantly reduced in both donor skin and burn margin relative to controls (p < 0.025) (Fig. 2E).
To next determine whether these changes in gene expression parallel changes in AMP protein production in donor skin or burn margin, we assessed skin AMP protein levels by ELISA. hBD1 protein levels were significantly increased in both donor and margin specimens (p < 0.001 and p < 0.025) (Fig. 3A); whereas hBD2 was significantly elevated in burn margin only (p < 0.025) relative to controls. The donor specimens contained nearly identical levels of hBD2 as controls (Fig. 3B). In contrast, protein levels of LL-37, the bioactive peptide of CAMP, were significantly reduced in burn margin as compared to controls (p < 0.005) (Fig. 3C). No differences in psoriasin levels were observed relative to controls (Fig. 3D); however, RNASE7 protein levels were increased in burn margin by ~50% (p < 0.05) (Fig. 3E).
We next evaluated whether burn margin or donor skin exhibits normal AMP activity against several skin pathogens and/or pathogens associated with the development of sepsis1,35. Both donor skin and burn margin exhibited a significant decrease in AMP activity against E.coli growth compared to peptides isolated from control skin (p < 0.003) (Fig. 4A). In contrast, we observed a significant increase in AMP activity against P. aeruginosa (p < 0.025) (Fig. 4B) in burn margin and S. aureus (p < 0.05) (Fig. 4C) in both donor skin and burn margin relative to controls. However, no significant changes in AMP activity against Group A Streptococcus were observed between control and donor skin or burn margin.
Proteolytic processing of AMPs into bioactive peptides is critical for their antimicrobial and inflammatory capacity. Two epidermal proteases involved in both AMP processing and epidermal barrier function include Kallikrein-5 [KLK5] and KLK716,17,28. Thus, we next evaluated the gene expression, protein abundance and localization of KLK 5 and 7, as alterations in these proteases may likely elicit skin barrier defects after burn injury. Although no significant changes were observed in KLK5 or KLK7 gene expression (Fig. 5A and B), we did observe more robust staining of KLK5 and KLK7 in burn margins compared to controls, primarily in the epidermis (Fig. 5C). No differences were observed between donor skin and control specimens when staining for KLK5 or KLK7. When total protease activity was evaluated, we observed a significant increase in the burn margin samples with the quantitative assessment (Fig. 5D) (p < 0.001) and a robust increase in epidermal protease activity in donor skin and burn margin by IHC (Fig. 5E) when compared to controls.
With evidence of altered AMPs in donor skin and burn margin, we next evaluated the same sites for pro-inflammatory cytokines induced by AMPs in the skin 36,37. We first assessed the gene expression of IL-6 and IL-8, and observed that IL6 expression was only elevated in the margin samples (Fig. 6A) (p < 0.025), whereas IL8 expression was significantly elevated in both donor and margin relative to controls (p < 0.025 and p < 0.001) (Fig. 6B). When protein concentrations were evaluated, no changes were seen with IL-6 protein levels (Fig. 6C). However, we observed a significant elevation of IL-8 in both donor and margin relative to controls (Fig. 6D) (p < 0.025). Nevertheless, we did not observe any robust changes in inflammatory cell numbers in the donor skin by hematoxylin and eosin staining (data not shown). These data suggest that AMP dysregulation may correlate with higher levels of pro-inflammatory mediators in donor skin and burn margin.
Graft failure at both the donor site and/or burn site is a persistent challenge related to post-burn management. Furthermore, burn wound infection at both the donor site and burn site remain a frequent and serious complication of major burn injury, and are associated with over 50% of all deaths related to burn injury1,35,38. Despite this fact, no studies have examined potential mechanisms by which burn injury at a local site impacts epidermal barrier function at the donor site of human burn subjects. The complex pathophysiology of burn injuries results in various local and systemic effects largely related to the acute inflammatory response, thus affecting several organ systems39 and here we highlight its systemic effects on the skin as a whole. In the current study, we established that epidermal lipid and AMP responses are impaired in both donor skin and burn margin from human burn patients. We further identified an increase in epidermal protease activity and pro-inflammatory cytokine production in both donor skin and burn margin. These alterations in epidermal barrier function may be a source of graft failure, burn wound infections, and/or subsequent infectious complications in burn patients.
Epidermal lipids are critical for maintaining the permeability barrier, modulating wound healing responses, and anti-bacterial activity. Despite this fact, few studies have attempted to investigate their role in relation to burn injury outcomes. Using an established mouse model, we recently determined that the burn margin and distal, unburned skin (i.e. donor skin) exhibited greater epidermal barrier permeability as indicated by a higher pH, greater transepidermal water loss, and reduced lipid synthesis enzyme expression up to 96 hours post-burn28. In the current study, we were able to confirm that the robust changes in epidermal lipid production and lipid synthesis enzyme production after burn injury is conserved in both mouse and human models. The dramatic alterations in epidermal lipid regulation reported in our mouse and human burn models sets the foundation for future studies designed to elucidate the role of lipids and their regulatory mechanisms in relation to burn wound healing.
Although these changes are often part of the “normal” response to thermal injury and result in acceptable outcomes, it may be possible to further optimize the wounds and donor sites in order to prevent graft failure and improve wound healing. By enhancing the restoration of epidermal lipid homeostasis, the time required for the skin to regain its fundamental barrier functions may be significantly shortened. Several studies using various wound dressings have attempted to mimic the natural barrier function of the skin and demonstrated with variable success improvements in wound healing, patient experiences, and outcomes40. However, the optimal dressing remains a topic of significant controversy, and the identification of novel mechanisms for potential intervention could prove quite beneficial.
One potential mechanism for impaired epidermal lipid metabolism is through the peroxisome proliferator-activated receptors (PPAR- alpha, beta/ delta or gamma), which are ligand-activated nuclear hormone receptors. All three PPAR subtypes are expressed in mice and humans, and exert variable effects on epidermal barrier repair, sebocyte differentiation/lipogenesis, and keratinocyte differentiation/proliferation/lipogenesis41,42. Furthermore, a deficiency in epidermal PPAR-beta/delta was found to delay skin barrier recovery after tape stripping and increase skin inflammation43. The observed increase in pro-inflammatory cytokines in donor skin and burn margin, in conjunction with our lipid analyses in mouse and human, indicate that burn injury may be influencing PPAR-dependent epidermal lipid regulation to impair barrier function and modulate local inflammation, which warrants further investigation. Because topical lipids have been used for centuries to improve wound healing outcomes44, application of natural epidermal lipids (e.g. ceramides, fatty acids, cholesterol) to donor skin prior to grafting may improve burn wound healing outcomes by improving epidermal barrier function and reducing inflammation before the graft is placed, thus grafting a less damaged area of skin.
AMPs are gene-encoded peptides that comprise a highly conserved component of the innate immune system, and contribute to direct microbial destruction and tissue repair pathways10,11. Several AMPs are expressed in human skin under inflammatory conditions, including burn injury, and exert detrimental effects on epithelial function if their production is in excess. Increased AMP production likely generates a more inflammatory microenvironment through the production of pro-inflammatory cytokines (e.g. IL-6 and IL-8), which can impede normal wound re-epithelialization and promote an oxidative milieu. In parallel, serum levels of IL-6 and IL-8 increase after burn injury, and higher levels are associated with greater morbidity and mortality45–48. In contrast, reduced AMP production may increase the invasive capacity of pathogenic microbes and slow tissue repair pathways10,11. Previous studies in human burn margin or burn fluid have demonstrated a reduction in hBD levels23,25 and changes in keratinocyte localization of AMPs24,49. However, no studies have evaluated AMP production or quantitatively assessed AMP protein levels in potential donor sites. Our observed increase in hBD1, hBD2, and RNAse 7, and parallel decrease in LL-37, in both donor skin and burn margin may be due to differences in the skin microbiome after burn injury. Several skin pathogens are known to induce keratinocyte AMPs and pro-inflammatory cytokines through activation of Toll-like receptors (TLRs)14,50–53. Furthermore, resident commensal microbes (e.g. Staphylococcus epidermidis; Propionibacterium acnes) can produce their own AMPs, augment the normal production of AMPs via keratinocytes, and help maintain inflammatory homeostasis by limiting excess cytokine release after epidermal injury54–56. Several skin pathologies and non-healing wounds, reveal an imbalance of this microbiota, even in the absence of clinical infection30,57,58. As such, studies are underway to characterize the skin microbiome in donor skin and burn margin after burn injury.
Certain metabolic and hemodynamic mechanisms, which are disrupted after burn injury, may also contribute to the observed AMP dysregulation. For example, it has been established that a pH gradient exists in human skin59. We speculate that the observed increase in pH in mouse donor and margin skin after burn injury28 can exert several deleterious effects on epidermal barrier function by increasing serine protease activity (i.e. KLKs), as observed in our studies, and by promoting an imbalance of electrolytes in both the donor skin and burn margin caused by the burn injury itself. The expression and activity of KLKs in the skin can be distinctly stimulated by several factors, such as cell differentiation, vitamin D, calcium, and retinoic acid60. Alterations in these variables are observed after burn injury61, and may augment epidermal protease activity, leading to local AMP inactivation or processing into more inflammatory peptides62. The local ionic milieu may also explain the dramatic reduction in E. coli antibacterial activity, as psoriasin kills E. coli by Zn2+ sequestration20, and plasma levels of zinc have been shown to significantly decrease after burn injury63. AMP inactivation or loss of function has also been associated with physiological concentrations of ionic factors, and carbonate was identified as a critical ionic factor in mammals that can impart the bactericidal activity of AMPs at physiologic salt concentrations64. Future studies designed to assess these parameters will help identify alternative mechanisms for the divergent AMP responses in the skin after burn injury.
In conclusion, our data indicate that a burn injury perturbs epidermal barrier function by modulating epidermal lipid and AMP responses at the primary burn site and at the donor site. These findings may promote the development of new therapies to improve local burn wound management and healing, and may have potential implications for tissue flap reconstructions and repeatedly exploited autograft sites by facilitating pre-operative optimization of the wounds and donor sites. Diminished epidermal barrier function due to impaired lipid production and AMP responses, or due to alterations in the molecules which regulate their activity, likely imparts detrimental consequences on wound healing responses, graft survival, hemodynamic stability, and microbial defense. Our data suggest that improving epidermal barrier function at both the primary burn site (i.e. burn margin) and potential donor sites prior to grafting would be expected to improve graft survival and reduce secondary infectious complications. Furthermore, controlling the wound microenvironment may have systemic implications, as the burn wound serves as the foundation for all of the downstream inflammatory/repair responses and complications. We are currently investigating potential mechanisms for impaired epidermal barrier function in human burn subjects, which may help elucidate the optimal timeframe for novel therapeutic intervention.
Financial Support: Research reported in this publication was supported by the National Institutes of Health (NIH) grant NIH T32 GM008750 (RLG) and the Dr. Ralph and Marian C. Falk Medical Research Trust (KAR and RLG). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. Any opinions, findings, conclusions, or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the NIH/NIGMS.
Conflict of Interest: The authors have declared that no conflict of interest exists.