Temporal patterns of gene expression in murine cutaneous burn wound healing
Robert J. Feezor1,*,
Heather N. Paddock1,*,
Henry V. Baker2,
Juan C. Varela3,
Joyce Barreda1,
Lyle L. Moldawer1,
Gregory S. Schultz3 and
David W. Mozingo1
1 Department of Surgery
2 Department of Molecular Genetics and Microbiology, University of Florida
3 Institute for Wound Research, Department of Obstetrics and Gynecology and Department of Surgery, Gainesville, Florida 32610
 |
ABSTRACT
|
---|
The global changes in gene expression in injured murine skin were characterized following a second-degree scald burn. Dorsal skin was harvested from uninjured and from burned mice at 2 h and at 3 and 14 days following immersion in 65°C water for 45 s. Gene expression was surveyed using an Affymetrix U74Av2 GeneChip, and patterns of gene expression were analyzed using hierarchical clustering and supervised analysis. Burn injury produced significant alterations in the expression of a number of genes, with the greatest changes seen 3 and 14 days after the scald burn. Using a supervised analysis with a false discovery rate of 1% or 5%, differences in the expression of 192 or 1,116 genes, respectively, discriminated among the unburned skin and the three time points after the burn injury. Gene expression was primarily a transient and time-dependent upregulation. The expression of only 24 of the 192 discriminating genes was downregulated after the burn injury. No gene exhibited a sustained increase in expression over the entire 14 days following the burn injury. Gene ontologies revealed an integrated upregulation of inflammatory and protease genes at acute time intervals, and a diminution of cytoskeletal and muscle contractile genes at 3 or 14 days after the injury. Following a second-degree scald burn, global patterns of gene expression in the burn wound change dramatically over several weeks in a time-dependent manner, and these changes can be categorized based on the biological relevance of the genes.
microarray; heat shock protein; chemokine; muscle contraction; actin cytoskeleton; sarcomere; immune response
 |
INTRODUCTION
|
---|
THE HEALING DERMAL WOUND UNDERGOES sequential phases of inflammation, repair, and remodeling. Models of incisional wounds have been primarily used to describe the multiphase process of wound healing and the cellular events involved therein, including platelet degranulation, leukocyte migration, and epithelial cell migration. Examination of the fluid from incisional wounds has begun to disclose the continuous network of intercellular communication in a healing wound, much of which involves the production of proinflammatory cytokines. It is now believed that these inflammatory mediators are responsible for initiating and regulating the process of wound healing (1, 3, 5, 23, 30, 32).
Burn wounds differ from incisional wounds in both the direction and cellular mechanism of injury. Burn wounds are characterized by heat-induced tissue coagulation at the time of injury (35), and in contrast to incisional wounds, the predominant direction of tissue injury in thermal wounds is horizontal, not vertical (26). These fundamental differences result in distinct healing mechanisms of incisional vs. thermal wounds. Vertical injuries such as those seen with surgical incisions heal rapidly by blood clot formation, re-epithelialization, and fibroblast proliferation, whereas cutaneous burn wounds heal more slowly, in part due to the edema, extensive necrosis, and relative hypoxia of the burn wound (26). Furthermore, it is well established that large burn injuries induce systemic immune dysfunction (2, 14, 17, 20, 28). Evidence of this dysregulation can be seen locally at the wound itself or systemically by alterations in the concentrations of circulating inflammatory mediators and leukocyte phenotypes (22, 24).
The depth of thermal injury is the most important determinant of healing capacity and mechanism in burn wound patients. Whereas superficial burns involve only the topmost layers of skin, partial thickness and deep partial thickness burn wounds involve destruction of the underlying dermal elements. In the most extreme cases of full thickness burn injuries, all layers of the skin are destroyed.
Advances in the field of burn wound healing remain limited in part by an incomplete understanding of the fundamental cellular mechanisms driving the normal healing process. Several investigators have attempted to characterize the physiological response to thermal injury in terms of systemic inflammatory cytokine production, local wound milieu, or more recently with the advent of microarray technology, the gene expression profiles of the healing wound. We sought to determine sequential changes in global patterns of gene expression during normal healing of partial thickness burn wounds in a murine model. A better appreciation for the gene expression patterns of normal burn wound healing may elucidate the underlying forces that dictate the cellular and protein changes during wound healing. Such knowledge may eventually be used to affect therapeutic interventions in terms of preventing overzealous wound healing (as seen with hypertrophic scars) or incomplete or dysregulated wound healing.
Microarrays have emerged as a powerful new technology to measure genomic regulation and can potentially elucidate the driving genetic machinery dictating phenotypic responses of individual tissues or cell populations. Whereas former efforts to uncover the genetic regulatory events controlling the production of inflammatory mediators were restricted in their ability to measure only a few genes at a time, microarray technology allows the simultaneous assessment of the expression pattern of large portions of the genome. Using an established model for inducing partial thickness scald injury to mice (6), we compared the gene expression patterns of skin from mice subjected to the same thermal injury after 2 h, 3 days, and 14 days of recovery.
 |
METHODS
|
---|
Murine partial thickness burn.
All protocols were reviewed and approved by the University of Florida Institutional Animal Care and Use Committee. Male C57BL/6 mice, 68 wk of age and weighing 2025 g each, were obtained from Jackson Laboratories (Bar Harbor, ME). Anesthesia was induced using 2.54% inhaled isoflurane that was administered continuously via a nose cone and adjusted throughout the procedure to maintain an appropriate depth of anesthesia. After induction of general anesthesia, buprenorphine (0.05 mg/kg body wt) was administered subcutaneously to control postprocedure pain and discomfort. The dorsum of the mouse was cleansed and shaved, and each mouse was given a preinjury fluid bolus consisting of 0.08 ml/g body wt of normal saline (0.9% sodium chloride) intraperitoneally. A 2 cm x 3 cm scald wound was created on the prepped dorsal surface by placing the mouse in a water-tight container with a cut-out measuring 2 cm x 3 cm, which was then submerged in a 65°C water bath for 45 s. Immediately thereafter, the burn wound was immersed in an iced water bath for 45 s to immediately stop the burning process (6). A second-degree burn injury was confirmed histologically (Fig. 1). Mice were allowed to recover in a warm, dry cage, given water and food (standard mouse chow) ad libitum, and monitored regularly for signs of distress. Additional doses of buprenorphine at 0.05 mg/kg of body wt were administered as needed over the first 48 h. Mice were weighed daily, and wounds were inspected daily for signs of infection.

View larger version (112K):
[in this window]
[in a new window]
|
Fig. 1. Histological confirmation of a second-degree burn in mouse skin. Mice were subjected to a 65°C scald burn for 45 s. After euthanasia at 3 days, the burn and adjacent healthy skin was excised, fixed in buffered formalin, and stained with hematoxylin and eosin. Original magnification was 100x. Original inset magnification was 400x. The vertical line indicates the demarcation between burned and healthy skin. The burned skin is characterized by the presence of a superficial loss of cellularity of the involved hair follicles (arrow A), epithelial blistering and sloughing (arrow B), and changes in collagen organization due to thermal denaturation (arrow C).
|
|
In preliminary studies (data not shown), the water temperature was varied between 65 and 70°C and immersion times ranged between 15 and 60 s to obtain a reproducible second-degree burn in the animals. A second-degree burn was confirmed by the appearance of a blister, the presence of edema, the absence of any tissue damage to the underlying fascia and muscle tissue, and complete healing with the absence of scarring or wound contraction. Histological documentation confirmed the presence of a second-degree burn associated with the superficial loss of cellularity of the involved hair follicles, epithelial blistering and sloughing, and changes in collagen organization due to thermal denaturation (Fig. 1).
In addition, no effort was made to control for the effect of the hair growth cycle on the depth of the burn injury and the gene expression profiles. In all cases, when the animals were killed, the presence of a second-degree scald burn was confirmed by gross visualization. Age- and sex-matched animals, irrespective of their place in the hair growth cycle, were studied. By employing animals in which the hair growth cycle was chosen randomly, the effect of this variable on gene expression was presumably minimized. The changes in gene expression observed over the 4-wk period could therefore be attributed to the burn injury, and not potentially, to any concerted responses secondary to synchrony in the hair growth cycle.
Animals were killed immediately prior to the burn injury or at 2 h, 3 days, and 14 days postburn injury. An overdose of pentobarbital (50 mg/kg) was administered intraperitoneally, and the animals were subsequently euthanized by cervical dislocation. The dorsal burn wound was surgically excised to the fascia, taking care to remove only burn tissue and not adjacent, healthy skin. Immediately thereafter, the tissue was (
1 min) snap frozen in liquid nitrogen and stored at -80°C for RNA isolation. The control animals were given general anesthetic of similar duration as those in the burn groups, then subsequently killed without induction of the burn injury. A similarly sized sample of shaved skin was excised and processed in a manner identical to that of the experimental group. Three mice were included in each treatment group and the control group.
RNA isolation.
Total RNA was isolated from murine burn wound skin using TRI Reagent (Molecular Research Center, Cincinnati, OH). Briefly, each frozen sample (weight 70150 mg) was thoroughly homogenized in 2 ml TRI Reagent. Total RNA was extracted with BCP Phase Separation Reagent (Molecular Research Center), precipitated with isopropanol, washed with ethanol, and resuspended in RNase-free water. The RNA from each sample was then purified over a commercially available column (RNeasy; Qiagen, Valencia, CA).
Quality of the RNA was confirmed by a spectrophotometric absorbance ratio (A260/A280), a 1% ethidium bromide agarose gel, and RT-PCR for a housekeeping gene (Cu-Zn superoxide dismutase) (38).
Biotinylated cRNA preparation.
Biotinylated cRNA was synthesized from 20 µg of total cellular RNA according to the protocol as published by Affymetrix, with few modifications. Briefly, the mRNA was selected by hybridizing to T7-(dT)24 primer (Genset Oligos; Genset, La Jolla, CA) and subsequently reverse transcribed to cDNA. In vitro transcription was then performed incorporating biotinylated ribonucleotides using an Enzo BioArray HighYield RNA Transcript Labeling Kit (Enzo Life Sciences, Farmingdale, NY). The resultant labeled cRNA was fragmented and hybridized onto Affymetrix U74Av2 oligonucleotide arrays. The microarrays were hybridized for 16 h at 45°C, stained and washed according to an Affymetrix protocol (EukGE-WS2v4) using an Affymetrix fluidics station, and scanned with an Affymetrix scanner. RNA samples were processed in parallel and samples were not pooled.
Data acquisition, normalization, and clustering.
The level of hybridization was measured with Affymetrix Microarray Analysis Suite version 5.0 (MAS 5.0). The mean signal intensity for all probe sets on each array was set to 500, thereby allowing array-to-array comparisons. Probe sets whose signal intensities (which presumably relate to gene expression levels) were not detected above the background noise on any of the arrays in the experiment were removed from the data set. This process of noise filtering reduced the number of probe sets from 12,488 to 7,275.
Unsupervised cluster analysis was performed after removing signals from genes that did not vary much between arrays. To accomplish this analysis, probe sets were ranked by their coefficient of variation (the ratio of standard deviation to mean) across the data set of arrays, and the signals of the top half of the genes in rank order were normalized across all arrays to a mean of 0 and a standard deviation of 1 (n = 3,638 genes). Hierarchical cluster analysis was performed and displayed using software developed by Eisen et al. (13).
Supervised analysis was performed specifically to detect differences among the four experimental groups (control or unburned mice, and mice subjected to thermal injury and killed at 2 h postburn, 3 days postburn, and 14 days postburn) using significance analysis of microarrays (SAM) (41). When the SAM algorithm is used for the analysis of experiments with multiple classes, a modified t-test is used to identify genes with statistically significant differences in expression among the groups. The percentage of genes identified by chance using this method is called the false discovery rate (FDR) and is estimated using permutations of the data set. The genes identified by SAM were then subjected to hierarchical cluster analysis after variance normalization as described above.
Last, the mean fold change of expression of burned to unburned (control) gene expression was calculated using the mean of the three replicates for each time group. Subsequently, MAPPFinder (Gladstone Institutes, University of California, San Francisco) was used to identify classes of genes according to the Gene Ontology (GO) Consortium (4) that were represented among the significant genes. Genes were included in pathway analysis if they were identified by SAM (using the FDR of 5%) and if their mean expression was at least twofold up- or downregulated from baseline. Using MAPPFinder, a statistical probability (z score) ranks the likelihood that a specific gene ontology is represented in the gene list of interest more often than would occur by chance alone (11). A z score threshold of 4.0 was used as the cutoff.
 |
RESULTS
|
---|
Three mice were evaluated in each of four groups: control (uninjured) and three burn (experimental) groups. All experimental mice were subjected to the same thermal injury as described above, but killed 2 h, 3 days, or 14 days after injury. RNA was harvested from injured skin and processed for microarray analysis in the manner described above. The resulting data set was filtered to remove data from probe sets that were not detected above background on any of the arrays. A variation filter was applied to remove signals from one-half of the probe sets that varied the least in the experiment. The signals of the remaining probe sets were then subjected to unsupervised hierarchical cluster analysis as shown in Fig. 2. The primary node of separation was between control and 2-h mouse skin samples, and the 3- and 14-day samples, implying that the global changes in gene expression in the early injury response (2 h) were not as great as the changes seen at the later time points. The results also suggest that by 14 days after the burn injury, gene expression patterns had not returned to those seen in the unburned skin.

View larger version (32K):
[in this window]
[in a new window]
|
Fig. 2. Unsupervised cluster analysis of the 50% of the genes within the filtered data set whose expression changed the most among all 12 microarrays. The primary node of separation was between the acute time periods (uninjured and 2 h after injury) and the prolonged periods (72 h postburn and 14 days postburn). Clustering and display accomplished using software developed by Eisen et al. (13); r1, r2, r3 = replicate number; r = Pearsons correlation coefficient.
|
|
Closer examination of Fig. 2 shows that the expression from one of the replicates from the control (replicate r3) and the day 3 (replicate r1) samples clustered separately from its experimental replicates. There was no a priori reason to discard these samples from subsequent analyses. One of the consequences of applying a variation filter is that it accentuates discrepancies due to uncontrolled experimental variables that contaminate the data set, as well as identifying the probe sets representing genes whose expression varies the greatest between the classes as a function of the experimental variable.
To further identify genes whose expression levels varied between treatment classes, we used a supervised learning method to identify probe sets representing genes with significant differences in expression. These probe sets were identified using SAM tuned such that the FDR among the probe sets identified as significant was 1%. The 192 probe sets so identified as significant were then subjected to hierarchical clustering as shown in Fig. 3. Once again, the primary node of division separated the gene expression patterns between the control and early (2 h) response from the prolonged recovery phases (days 3 and 14). In this analysis, there were no obvious outliers. Examination of this cluster revealed five distinct gene expression responses to the burn injury: one set whose expression was increased at each of the three times (2 h, 3 days, and 14 days), a group of genes whose expression was upregulated consistently at both 3 and 14 days after burn injury, and a set of genes whose expression was continuously downregulated after the burn injury.

View larger version (32K):
[in this window]
[in a new window]
|
Fig. 3. Supervised cluster analysis identified 192 genes with a false-positive rate of 1% (2 genes) whose expression levels discriminated among the four experimental groups: unburned skin (control), and burn skin harvests 2 h, 3 days, and 14 days after the injury. The clustering depicts five groups of genes. Analysis was performed using significance analysis of microarrays (SAM) (41) and displayed using Cluster and TreeView (13); r1r3, = replicate number; r = Pearsons correlation coefficient.
|
|
As shown in Fig. 3, there were 24 genes whose expression was sharply downregulated at all times after the burn injury, including three genes for procollagen and a secreted acidic cysteine-rich glycoprotein (another extracellular matrix protein). The expression of 52 genes was upregulated solely at 2 h after burn injury, and these included a heat shock protein, interferon-activated gene, the oncostatin receptor, and a gene induced by transforming growth factor-ß (TGF-ß). By contrast, the expression of 51 genes was upregulated solely at 3 days, whereas the expression of 41 genes was not upregulated until 14 days after the burn injury. Supplementary Table S1 (available at the Physiological Genomics web site)1
contains the list of all 192 probe sets allocated to the different bins and their relative changes in expression from baseline.
Examining the expression of some specific inflammatory genes revealed that the expression of the tumor necrosis factor (TNF) receptor 2 (p75) was upregulated after burn injury (2.4-fold), and this upregulation peaked at 3 days following the burn injury. The expression of several genes associated with innate immunity replicated this pattern of intense overexpression at 3 days seen with a gradual return to approximately baseline value by 14 days, including ferritin, lipopolysaccharide-binding protein (LBP), and TGF-ß. The expression of amyloid A3, primarily an hepatic acute phase reactant, was found to be upregulated at all times studied, but its most extreme upregulation also occurred at 3 days, at which time it was expressed 85-fold higher than seen in unburned skin.
Interestingly, examination of the expression patterns of genes encoding for chemokines and their receptors showed a strikingly different pattern. Two hours after the partial thickness burn, there was a downregulation in the expression of the CC chemokine receptor (CCR) type 5, and to a lesser extent, type 2. After 3 days, CCR-5 and CCR-2 expression were intensely upregulated, along with the expression of monocyte chemotactic protein (MCP) 2 and macrophage inflammatory protein (MIP) 1-
. The expression of all of these inflammatory genes returned to near baseline by 14 days.
The expression of the murine matrix metalloproteinases (MMP), namely MMP-3, -9, -12, and -13 were all downregulated to varying degrees as early as 2 h postinjury. MMP-9 and MMP-13 then showed a rebound overexpression at 3 days with sustained elevation as late as 2 wk after the injury. Expression of tissue inhibitor of metalloproteinases (TIMP) 2 was also upregulated nearly fourfold as early as 2 h and remains similarly elevated throughout the 14 days measured in this study.
To perform ontological analysis, supervised clustering was again performed on the same filtered data set, but setting the FDR to 5%, yielding 1,116 discriminatory genes. These probe sets are summarized in Supplementary Table S2. The mean expression values of each probe set for each experimental group were calculated, and a ratio of this mean to the control (uninjured) mean was calculated, thereby assessing the average relative change in expression from baseline in a group of genes known to be altered among the experimental groups.
MAPPFinder identified several gene ontologies to be uniquely regulated at each time point studied (Table 1). Not surprisingly, at 2 h, the predominant classes of genes upregulated included heat shock proteins (z score 6.91). Genes uniquely downregulated acutely were mostly metalloendopeptidases, intermediate filament, and cell-cell signaling genes (z scores 4.43, 5.24, and 4.88, respectively).
Three days after thermal injury, the genes responsible for mounting an innate immune response were upregulated, as well as genes encoding for chemokines and chemotaxis, consistent with pathway analysis as described above. Surprisingly, by 3 days, there was a dramatic downregulation of genes involved in muscle contraction (z score 10.12), actin cytoskeleton (z score 7.39), and sarcoplasmic reticulum (z score 6.17). These patterns of downregulated genes persisted to 14 days, at which time genes involved in muscle contraction and actin cytoskeleton were identified with z scores of 8.95 and 7.56, respectively.
 |
DISCUSSION
|
---|
The current results demonstrate that there are broad changes in gene expression in the skin that occur following a second-degree scald burn in the mouse. Furthermore, these changes in expression are sustained for prolonged periods, in excess of 14 days. In addition, the changes in expression affect not only components of the inflammatory response, but also include structural and regulatory genes. Dinh et al. (10) studied the global gene expression profiles from cultured keratinocytes exposed to thermal injury and showed that the predominant groups of genes that were upregulated initially included the heat shock proteins, growth arrest and repair genes, and DNA binding genes. Although the authors studied expression of only 588 genes, these acute alterations generally had returned to baseline by 24 h postburn (10). In contrast, as the technology supporting functional genomics has evolved, we studied over 12,000 genes and measured the gene expression alterations in an animal model up to 2 wk after the injury. Nevertheless, we found similar results. Setting a FDR of 5%, there were 260 genes whose expression were upregulated at 2 h after the injury and returned to baseline by 3 days. A majority of these were acute phase response or inflammation genes, including four heat shock proteins, four interferon-activated genes, and several genes induced by TGF-ß.
One of these acute phase responders, LBP, is thought to play a key role in the inflammatory response to invading organisms. Klein et al. (25) examined burn wounds in a rat model and found increased levels of LBP mRNA at 24 h, with subsequent decreases at 48 and 72 h, and speculated that the decline in LBP production may predispose to invasive bacterial infection irrespective of persistent IL-1ß production. By contrast, our data suggest the peak increase in LBP gene expression in the burn wound occurs at 72 h after the burn. Its expression was coregulated with several other known acute phase proteins, namely, amyloid and ferritin. Furthermore, we showed the greatest upregulation of cytokine receptors to occur at 3 days in that the expression of IL-4 receptor, TNF receptor type II, and the immunoglobulin G fixed component receptor were all upregulated more than twofold from unstimulated wounds by 3 days (Fig. 3).
Spies et al. (37) used oligonucleotide microarrays to examine the genomic alterations induced in a full thickness injury model in rats. They found a predominant downregulation of genes involved in metabolism and neural signaling shortly after injury with a return to near-normal levels by day 10 (37). They further found increased expression of cell stress-related genes, including heat shock proteins and inflammatory genes (37). This recapitulated the earlier studies by Fukuzuka et al. (16) from our laboratory, which showed that the inflammation-related genes were primarily upregulated during the initial 24 h after the injury (16).
The gene ontology analysis of the present studies further support the data of Spies et al. (37) and Fukuzuka et al. (16), as classes of genes involved in responses to stress and external stimuli were identified as uniquely upregulated at 2 h. Moreover, the increased expression of genes known to be involved in the response to biotic stimuli were upregulated at 3 days. Furthermore, by 3 days, chemotaxis and cell-signaling genes were also upregulated. By 2 wk, the inflammatory response had abated on the gene expression level, and the predominance of gene upregulation focused on cell adhesion, glycosaminoglycan binding, and extracellular matrix genes.
Looking at global gene expression in the viable tissue surrounding a full thickness scald burn, Spies et al. (37) found that the transcription profiles of genes from samples harvested at 24 and 240 h after the burn were more alike than those harvested at 2 and 6 h after the burn. This is analogous to data from the current study, with our acute (2 h) samples clustering with the unburned samples, but separate from the 3- and 14-day samples (Figs. 2 and 3). The alteration in gene expression may in fact be due to a shift in cellular predominance within the injured tissue that occurs between 2 h and 3 days after the injury. Alternatively, the change in genomic expression between acute and chronic times may simply reflect that the acute (2 h) time period is insufficient to effect a large-scale genomic alteration in this model.
Not unexpectedly, there were marked alterations in the expression values of the MMP and the protease inhibitors (TIMP). Several independent investigators have shown that the rate of epithelialization is partially dependent on the presence of MMPs (36, 37). In fact, chronic wounds have been shown to express exaggerated levels of proteases (9, 40). In the present study, the expression of MMP-3, -9, -12, and -13 were acutely downregulated at 2 h, but MMP-3, -9, and -13 expression were subsequently upregulated by 3 days. MMP-9 and -13 remained upregulated through 2 wk after the burn. Young and Grinnell (43) demonstrated the presence of MMP-9 in the burn fluid as early as 48 h postburn and MMP-3 four days after the burn.
Procollagen was among the genes whose expression was downregulated continuously after thermal injury. Hypertrophic scarring, which occurs as a complication of dysregulated normal wound healing, is known to be characterized by an accumulation of extracellular matrix proteins. Ghahary et al. showed that insulin-like growth factor I (IGF-I) induced the expression of procollagen
(I)-chain mRNA as early as 6 h after thermal injury (18), and overexpression of procollagen caused an accumulation of extracellular matrix proteins which clinical manifested as hypertrophic scarring (19). Furthermore, intralesional injections of Kenalog, a potent anti-inflammatory agent, have been shown to downregulate the production of procollagen mRNA in hypertrophic scars (42). Unlike human burn wounds, murine burn wounds rarely develop hypertrophic scarring. Although not evident as early as 2 h, these studies demonstrate an intense and sustained downregulation of genes known to be involved in actin formation and muscle contraction. This downregulation was evident even after 14 days.
The present studies demonstrate that there are global changes in the patterns of gene expression in murine skin following a second-degree scald burn. These changes are for the most part transient in nature and reflect in general a predominance toward increased expression. Future studies will be required to confirm these changes in expression prospectively and to evaluate how expression patterns may deviate in cases when patterns of healing are either delayed or accelerated.
 |
ACKNOWLEDGMENTS
|
---|
Grants
This work was supported by a supplement to National Institute of General Medical Sciences Grant R37-GM-40586-14, specifically for microarray analyses. H. N. Paddock and R. J. Feezor are supported by National Institute of General Medical Sciences Grant T32-GM-08721-05.
 |
FOOTNOTES
|
---|
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).
* H. N. Paddock and R. J. Feezor contributed equally to this work. 
Address for reprint requests and other correspondence: D. W. Mozingo, Burn Intensive Care Unit, Dept. of Surgery, PO Box 100286, 1600 SW Archer Road, Gainesville, FL 32610-0286 (E-mail: mozindw@surgery.ufl.edu).
10.1152/physiolgenomics.00101.2003.
1 The Supplementary Material for this article (Supplementary Tables S1 and S2) is available online at http://physiolgenomics.physiology.org/cgi/content/full/00101.2003/DC1. 
 |
References
|
---|
- Albina JE, Mills CD, Henry WL Jr, and Caldwell MD. Temporal expression of different pathways of 1-arginine metabolism in healing wounds. J Immunol 144: 38773880, 1990.[Abstract/Free Full Text]
- Alexander JW. Mechanism of immunologic suppression in burn injury. J Trauma 30: S70S75, 1990.[ISI][Medline]
- Arturson G. Pathophysiology of the burn wound. Ann Chir Gynaecol 69: 178190, 1980.[ISI][Medline]
- Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP, Dolinski K, Dwight SS, Eppig JT, Harris MA, Hill DP, Issel-Tarver L, Kasarskis A, Lewis S, Matese JC, Richardson JE, Ringwald M, Rubin GM, and Sherlock G. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 25: 2529, 2000.[CrossRef][ISI][Medline]
- Ashcroft GS, Lei K, Jin W, Longenecker G, Kulkarni AB, Greenwell-Wild T, Hale-Donze H, McGrady G, Song XY, and Wahl SM. Secretory leukocyte protease inhibitor mediates non-redundant functions necessary for normal wound healing. Nat Med 6: 11471153, 2000.[CrossRef][ISI][Medline]
- Cribbs RK, Luquette MH, and Besner GE. A standardized model of partial thickness scald burns in mice. J Surg Res 80: 6974, 1998.[CrossRef][ISI][Medline]
- Dahlquist KD, Salomonis N, Vranizan K, Lawlor SC, and Conklin BR. GenMAPP, a new tool for viewing and analyzing microarray data on biological pathways. Nat Genet 31: 1920, 2002.[CrossRef][ISI][Medline]
- Demling RH and LaLonde C. Early burn excision attenuates the postburn lung and systemic response to endotoxin. Surgery 108: 2835, 1990.[ISI][Medline]
- Di Colandrea T, Wang L, Wille J, DArmiento J, and Chada KK. Epidermal expression of collagenase delays wound-healing in transgenic mice. J Invest Dermatol 111: 10291033, 1998.[Abstract]
- Dinh HKB, Zhao B, Schuschereba ST, Merrill G, and Bowman PD. Gene expression profiling of the response to thermal injury in human cells. Physiol Genomics 7: 313, 2001.[Abstract/Free Full Text]
- Doniger SW, Salomonis N, Dahlquist KD, Vranizan K, Lawlor SC, and Conklin BR. MAPPFinder: using Gene Ontology and GenMAPP to create a global gene-expression profile from microarray data. Genome Biol 4: R7, 2003.[CrossRef][Medline]
- Drost AC, Burleson DG, Cioffi WG Jr, Mason AD Jr, and Pruitt BA, Jr. Plasma cytokines after thermal injury and their relationship to infection. Ann Surg 218: 7478, 1993.[ISI][Medline]
- Eisen MB, Spellman PT, Brown PO, and Botstein D. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA 95: 1486314868, 1998.[Abstract/Free Full Text]
- Faunce DE, Gregory MS, and Kovacs EJ. Effects of acute ethanol exposure on cellular immune responses in a murine model of thermal injury. J Leukoc Biol 62: 733740, 1997.[Abstract]
- Faunce DE, Llanas JN, Patel PJ, Gregory MS, Duffner LA, and Kovacs EJ. Neutrophil chemokine production in the skin following scald injury. Burns 25: 403410, 1999.[CrossRef][ISI][Medline]
- Fukuzuka K, Edwards CK III, Clare-Salzler M, Copeland EM III, Moldawer LL, and Mozingo DW. Glucocorticoid-induced, caspase-dependent organ apoptosis early after burn injury. Am J Physiol Regul Integr Comp Physiol 278: R1005R1018, 2000.[Abstract/Free Full Text]
- Gamelli RL, He LK, and Liu H. Marrow granulocyte-macrophage progenitor cell response to burn injury as modified by endotoxin and indomethacin. J Trauma 37: 339346, 1994.[ISI][Medline]
- Ghahary A, Shen YJ, Nedelec B, Scott PG, and Tredget EE. Enhanced expression of mRNA for insulin-like growth factor-1 in post-burn hypertrophic scar tissue and its fibrogenic role by dermal fibroblasts. Mol Cell Biochem 148: 2532, 1995.[ISI][Medline]
- Ghahary A, Shen YJ, Nedelec B, Wang R, Scott PG, and Tredget EE. Collagenase production is lower in post-burn hypertrophic scar fibroblasts than in normal fibroblasts and is reduced by insulin-like growth factor-1. J Invest Dermatol 106: 476481, 1996.[Abstract]
- Hansbrough JF, Zapata-Sirvent R, Peterson V, Wang X, Bender E, Claman H, and Boswick J. Characterization of the immunosuppressive effect of burned tissue in an animal model. J Surg Res 37: 383393, 1984.[ISI][Medline]
- Heldin CH and Westermark B. Role of platelet derived growth factor in vivo. In: The Molecular and Cellular Biology of Wound Repair (2nd ed.), edited by Clark RAF. New York: Plenum, 1996.
- Herndon DN, Wilmore DW, and Mason AD Jr. Development and analysis of a small animal model simulating the human postburn hypermetabolic response. J Surg Res 25: 394403, 1978.[ISI][Medline]
- Imaizumi T, Nomura K, Hashimoto I, and Sawada Y. Interleukin-1 levels in blister fluids of some skin diseases with blister formation. Clin Physiol Biochem 8: 179183, 1990.[ISI][Medline]
- Kataranovski M, Magic Z, and Pejnovic N. Early inflammatory cytokine and acute phase protein response under the stress of thermal injury in rats. Physiol Res 48: 473482, 1999.[ISI][Medline]
- Klein RD, Su GL, Aminlari A, Zhang H, Steinstraesser L, Alarcon WH, and Wang SC. Skin lipopolysaccharide-binding protein and IL-1beta production after thermal injury. J Burn Care Rehabil 21: 345352, 2000.[CrossRef][ISI][Medline]
- Linares HA. From wound to scar. Burns 22: 339352, 1996.[CrossRef][ISI][Medline]
- Mandrup-Poulsen T, Wogensen LD, Jensen M, Svensson P, Nilsson P, Emdal T, Molvig J, Dinarello CA, and Nerup J. Circulating interleukin-1 receptor antagonist concentrations are increased in adult patients with thermal injury. Crit Care Med 23: 2633, 1995.[ISI][Medline]
- Miller CL and Baker CC. Changes in lymphocyte activity after thermal injury. The role of suppressor cells. J Clin Invest 63: 202210, 1979.[ISI][Medline]
- Molloy RG, ORiordain M, Holzheimer R, Nestor M, Collins K, Mannick JA, and Rodrick ML. Mechanism of increased tumor necrosis factor production after thermal injury. Altered sensitivity to PGE2 and immunomodulation with indomethacin. J Immunol 151: 21422149, 1993.[Abstract/Free Full Text]
- Nguyen H, Lim J, Dresner ML, and Nixon B. Effect of local corticosteroids on early inflammatory function in surgical wound of rats. J Foot Ankle Surg 37: 313318, 1998.[Medline]
- Ogle CK, Mao JX, Wu JZ, Ogle JD, and Alexander JW. The 1994 Lindberg Award. The production of tumor necrosis factor, interleukin-1, interleukin-6, and prostaglandin E2 by isolated enterocytes and gut macrophages: effect of lipopolysaccharide and thermal injury. J Burn Care Rehabil 15: 470477, 1994.[Medline]
- Ono I, Gunji H, Zhang JZ, Maruyama K, and Kaneko F. Studies on cytokines related to wound healing in donor site wound fluid. J Dermatol Sci 10: 241245, 1995.[CrossRef][ISI][Medline]
- Rodriguez JL, Miller CG, Garner WL, Till GO, Guerrero P, Moore NP, Corridore M, Normolle DP, Smith DJ, and Remick DG. Correlation of the local and systemic cytokine response with clinical outcome following thermal injury. J Trauma 34: 684695, 1993.[ISI][Medline]
- Schmid P, Kunz S, Cerletti N, McMaster G, and Cox D. Injury induced expression of TGF-beta 1 mRNA is enhanced by exogenously applied TGF-beta S. Biochem Biophys Res Commun 194: 399406, 1993.[CrossRef][ISI][Medline]
- Shakespeare P. Burn wound healing and skin substitutes. Burns 27: 517522, 2001.[CrossRef][ISI][Medline]
- Singer AJ and Clark RA. Cutaneous wound healing. N Engl J Med 341: 738746, 1999.[Free Full Text]
- Spies M, Dasu MR, Svrakic N, Nesic O, Barrow RE, Perez-Polo JR, and Herndon DN. Gene expression analysis in burn wounds of rats. Am J Physiol Regul Integr Comp Physiol 283: R918R930, 2002; 10.1152/ajpregu.00170.2002.[Abstract/Free Full Text]
- Tannahill CL, Fukuzuka K, Marum T, Abouhamze Z, MacKay SL, Copeland EM III, and Moldawer LL. Discordant tumor necrosis factor-alpha superfamily gene expression in bacterial peritonitis and endotoxemic shock. Surgery 126: 349357, 1999.[CrossRef][ISI][Medline]
- Tchervenkov JI, Epstein MD, Silberstein EB, and Alexander JW. Early burn wound excision and skin grafting postburn trauma restores in vivo neutrophil delivery to inflammatory lesions. Arch Surg 123: 14771481, 1988.[Abstract]
- Trengove NJ, Stacey MC, MacAuley S, Bennett N, Gibson J, Burslem F, Murphy G, and Schultz G. Analysis of the acute and chronic wound environments: the role of proteases and their inhibitors. Wound Repair Regen 7: 442452, 1999.[CrossRef][ISI][Medline]
- Tusher VG, Tibshirani R, and Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA 98: 51165121, 2001.[Abstract/Free Full Text]
- Xu S, Guan B, and Bao W. Study on type I and type III procollagen mRNA and in situ collagen proteins expressions in hypertrophic scars after intralesional Kenalog treatment [in Chinese, English summary]. Zhonghua Zheng Xing Wai Ke Za Zhi 16: 236238, 2000.[Medline]
- Young PK and Grinnell F. Metalloproteinase activation cascade after burn injury: a longitudinal analysis of the human wound environment. J Invest Dermatol 103: 660664, 1994.[Abstract]