ToxSci Advance Access originally published online on March 3, 2007
Toxicological Sciences 2007 97(2):569-581; doi:10.1093/toxsci/kfm037
Gene Expression and Target Tissue Dose in the Rat Epidermis after Brief JP-8 and JP-8 Aromatic and Aliphatic Component Exposures
Boonshoft School of Medicine, Wright State University, Dayton, Ohio 45435
1 To whom correspondence should be addressed at Department of Pharmacology and Toxicology, Wright State University, 3640 Colonel Glenn Highway, Dayton, OH 45435. Fax: (937) 775-7221. E-mail: james.mcdougal{at}wright.edu.
Received February 21, 2007; accepted February 22, 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Exposures of jet propulsion fuel 8 (JP-8) to human and laboratory animal skin have resulted in skin irritation. JP-8 is a mixture of aromatic and aliphatic hydrocarbons, which in some cases have also been shown to be irritating to the skin. In an attempt to determine if aromatic or aliphatic components could mimic the JP-8–induced gene expression response, we exposed rats to JP-8, undecane (UND), tetradecane (TET), trimethylbenzene (TMB), and dimethylnaphthalene (DMN) for 1 h and examined the epidermis to characterize the gene expression response. We also measured the concentrations of the JP-8 components in the epidermis with gas chromatography/mass spectrometry after 1-h exposures to JP-8 and pure components to determine if differences in potency could be identified. Changes in gene expression, compared to sham treatment, were studied with microarray techniques and analyzed for changes in gene ontology categories. UND and TMB exposures caused the greatest number of changes in transcript levels compared to DMN and TET. When only the specific functional and signaling pathways that were changed by JP-8 were considered, these pathways were nearly all activated by the components, but to different extents. After pure component exposures, the epidermal concentrations of the components showed no significant differences, although the differences in magnitude of either total or pathway-specific gene expression differed by a factor of 10-fold. We conclude that no single component that we studied mimicked the gene expression resulting from the JP-8 exposure but that UND had the most similar responses. These data suggest that there are differences in potency between the four components studied.
Key Words: gene expression; skin irritation; JP-8 jet fuel; epidermis; cutaneous exposure; rat; aromatic hydrocarbons; aliphatic hydrocarbons.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Jet propulsion fuel 8 (JP-8) is a complex mixture of hydrocarbons that is used at the rate of about 5 billion gallons per year by the military (United States and North Atlantic Treaty Organization [NATO]) as a fuel for aircraft, ground vehicles, cooking stoves, and personnel heaters (Committee on Toxicology of National Research Council, 2003). Toxicity of JP-8 (NATO designation F-34) has been previously reviewed in general (Agency for Toxic Substances and Disease Registry, 1998), with emphasis on inhalation (Committee on Toxicology of National Research Council, 2003) and cutaneous exposures (McDougal and Rogers, 2004
). Because of concerns about increased skin irritation of JP-8 compared to the fuel it replaced (JP-4), several studies have characterized cutaneous physiological and molecular changes in a variety of laboratory animals after JP-8 exposures of various durations. JP-8 increases transepidermal water loss across the skin (an indicator of damage) and skin thickness (Kanikkannan et al., 2001; Monteiro-Riviere et al., 2001) that are characterized histologically as erythema, edema, neutrophilic infiltration, and ultimately hyperplasia (Baker et al., 1999; Hoekstra and Phillips, 1963; Monteiro-Riviere et al., 2001, 2004). Molecular studies in the skin have shown that JP-8 causes oxidative stress (Ramos et al., 2004; Rogers et al., 2001), induction of cytokines and chemokines (Allen et al., 2000; Gallucci et al., 2004; Kabbur et al., 2001; Ullrich and Lyons, 2000), and apoptosis (Chou et al., 2006; Rosenthal et al., 2001). JP-8 applied to the skin of rats once a day for 7 days induced the expression of numerous proinflammatory cytokine and chemokine mRNAs (Gallucci et al., 2004). A 1-h cutaneous exposure in rats promotes changes in epidermal structural protein, cell signaling, inflammatory mediator, growth factor, and enzyme gene expression in epidermis (McDougal et al., 2007).
JP-8 is a performance-based fuel, which means that hydrocarbon composition can vary due to the source of crude petroleum being refined and the refining process. JP-8 is related to kerosene and contains approximately 200 aromatic and aliphatic hydrocarbons, ranging from about 9 to 17 carbons, including numerous isomers (Committee on Toxicology of National Research Council, 2003). The composition of JP-8 is approximately 18% aromatic, and the rest are primarily C10–C14 aliphatics (Committee on Toxicology of National Research Council, 1996). Toxicology studies with JP-8 sponsored by the U.S. Air Force usually use a pooled fuel sample for which some of the largest aromatic and aliphatic components have been quantified (McDougal et al., 2000).
In general, aromatic compounds are more irritating to the skin than aliphatic compounds (Boman, 1996; Hoekstra and Phillips, 1963; Klauder and Brill, 1947). Within the aromatic category, methyl substitutions make benzene rings more irritating (Ahaghotu et al., 2005), and reactivity increases with the number of aromatic rings (Chou et al., 2003; Wade, 1995). However, aliphatics are irritating to the skin. Aliphatics (C8–C20) applied to the skin for several days resulted in irritancy that peaked around tetradecane (TET) (Brooks and Baumann, 1956; Brown and Box, 1970; Moloney and Teal, 1988). Moreover, straight chain hydrocarbons are more irritating than branched hydrocarbons with the same number of carbons (Brown and Box, 1970). Not surprisingly, mixtures of aliphatic and aromatic compounds, like jet fuel, are irritating to the skin (Baker et al., 1999; Koschier, 1999).
Aromatic and aliphatic hydrocarbon components of JP-8 show different rates of absorption (into the skin) and permeability (through the skin) with cutaneous contact. Studies with pig ear skin, rat skin, human skin, silastic membranes in diffusion cells, and isolated perfused pig skin show that permeability coefficients from JP-8 for aromatic components are greater than the coefficients of the aliphatic components (Kanikkannan et al., 2001; McDougal et al., 2000; Muhammad et al., 2004). Based on blood concentrations, human exposure studies with JP-8 demonstrate that aromatics penetrate better than aliphatics after a 30-min exposure (Kim et al., 2006). Although penetration of aromatics is faster than aliphatics, it is the aliphatic chemicals that remain in the skin in detectible concentrations. Six aliphatic components (C9–C14), but no aromatic components, were identified in rat skin at the end of 3.5-h static diffusion cell studies (McDougal et al., 2000). The mass of chemicals, as measured by tape stripping, in the upper epidermis of humans exposed to JP-8 for 30 min is greater for aliphatics than aromatics (Kim et al., 2006). The question is whether aromatics cause irritation on the way through the skin or whether the aliphatics cause irritation because they remain in the skin. It is unclear which hydrocarbon component or components are responsible for JP-8–induced skin irritation.
Cutaneous chemical toxicity results from two distinct components (tissue dose and potency) that can vary with the characteristics of the chemical. First, toxic responses for a specific chemical are generally proportional to tissue dose. Second, potency of different chemicals is proportional to the chemical reactivity and associated biological effects in tissues. The magnitude of the biological response to chemicals can easily differ by more than an order of magnitude, depending on the tissue concentration and mechanism of action. Aromatics are more effective at causing skin irritation than aliphatics, but aromatics are present in smaller amounts in both the fuel and the exposed skin than the aliphatics. Aliphatic components appear to partition into the skin resulting in greater exposure in the target tissue than aromatic components. Therefore, it is possible that either the aromatics, which are present in JP-8 at less than 20%, or the aliphatics, which make up the majority of the fuel, may be responsible for the JP-8–induced irritation. The purpose of this study was to compare the epidermal gene expression responses to two aromatic and two aliphatic components (Table 1) with those of JP-8 and investigate the potency of these JP-8 components for inducing gene expression in the epidermis.
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Rodent exposures.
Male Fischer rats (CDF/CrlBR, Charles River Laboratories, Wilmington, MA, 259–325 g) were housed one per cage and provided food and water ad libitum. On the study day, rats were exposed to chemicals as previously described (McDougal et al., 2007
). Briefly, they were anesthetized with isoflurane, their back was closely clipped of fur, and they were exposed to 500 µl (156 µl per cm2) of JP-8, or 99+% undecane (UND), 99+% TET, 98% 1,2,4-trimethylbenzene (TMB), or dimethylnaphthalene (DMN), mixture of isomers in a Hill Top Chamber held in place with a modified Lomir rat jacket. JP-8 was from Wright-Patterson AFB, OH, and all other chemicals were from Aldrich, St Louis, MO. Sham exposures with empty chambers were used to control for diurnal and occlusive effects. At the end of the 1-h exposure, the rats were euthanized, the exposure chamber and harness were removed, and residual chemical was wiped off the skin. Procedures involving rats were approved both by the Wright State University and Air Force Institutional Animal Care and Use Committee.
Skin collection and processing for gene expression.
Samples were taken after euthanasia with CO2 according to previously described methods (McDougal et al., 2007). Briefly, a permanent marker was used to outline the exposure site on the back of each animal, and the exposed skin sites were excised, trimmed, and completely submerged in ice-cold RNAlater (Ambion, Austin, TX) and stored at – 20°C overnight. The next morning, five 8-mm biopsy punches were taken, flattened between cover slips, mounted to a cryotome chuck with superglue, and immediately sectioned on a cryotome at – 32°C to provide 5-µ epidermal skin sections which were combined for gene array and real-time RT-PCR analysis.
Analysis of skin concentrations.
Skin samples were removed from the rat and processed in less than 2 min to minimize loss of the volatile components of JP-8. The exposure chamber was removed, the skin was cleaned with gauze, and the exposure area was marked with a laboratory marker. The skin was excised, the hypodermis was removed by scraping, and the 8-mm punch was taken from the skin at the center of the exposure site. The punch was superglued on the cryotome chuck and kept frozen at – 80°C until ready for sectioning as described above. Frozen skin slices were placed directly into ice-cold preweighed headspace vials. Vials were immediately capped and reweighed, and tissue weights were recorded. Average weight of the epidermal sections was 12.0 ± 2.8 mg. Samples were analyzed on an Agilent (Foster City, CA) 6890N gas chromatograph with a 7694 headspace sampler and a 5973 series mass spectrometric detector. Three milliliters of headspace was automatically injected on a 150 m x 0.25 mm Petrocol DH 150 (Supelco, Bellefonte, PA) capillary column with a 14.4:1 split ratio. The temperature program was an initial temperature of 150°C that was increased at 1.5°C/min to a final temperature of 235°C for 25 min. The identities of chemicals were confirmed using Chemstation software (Agilent) and a NIST 98.L database. In the pure component studies, the mass of hydrocarbon in the epidermal samples was determined from a standard curve of known amounts of each component. The lower limit of detection in these samples was approximately 7.6, 7.4, 8.8, and 10.1 µg for TET, UND, TMB, and DMN, respectively. In the JP-8 studies, a JP-8 standard curve was used to quantify TET, UND, TMB, and DMN based on proportional peak size relative to total area from the gas chromatograph (Dietzel et al., 2005).
Gene expression with microarrays.
Total RNA was extracted from the epidermis using TRIreagent (Molecular Research Center, Cincinnati, OH) and cleaned up using the RNeasy Mini Kit (Qiagen, Valencia, CA) according to the manufacturer's instructions. The concentration and quality of total RNA were determined using the Experion RNA Standard Sensitivity chips in the Experion Automated Electrophoresis System (Bio-Rad, Hercules, CA). Double-stranded cDNA was generated using a Two-Cycle cDNA synthesis kit (Affymetrix, Santa Clara, CA) according to instructions. Forty micrograms of biotinylated cRNA was fragmented and hybridized to rat genome 230 2.0 GeneChip arrays (Affymetrix) that contain over 31,000 genes. Chip hybridization, washing, and staining were performed according to the Affymetrix recommended protocols. After scanning, the digitalized image data were processed using GCOS software (Affymetrix) and analyzed with GeneSpring GX 7.3 (Agilent Technologies, Palo Alto, CA).
GeneSpring data entry.
After analysis of the report file for each chip (Table 2) and confirmation that the chip image was not smeared or distorted (one control gene chip was not used due to a smeared image), the Affymetrix "CEL" file for each sample was imported into GeneSpring using robust multichip average (RMA) normalization. During the RMA normalization, probe expression values less than 0.01 were set to 0.01, and expression values on each chip were computed by a three-step process: background correction, quantile normalization, and median polishing (Bolstad et al., 2003; Irizarry et al., 2003a,b). Gene expression in each treatment (JP-8, UND, TMB, DMN, and TET) was individually compared with the sham-treated control group. The ratio of signal to control (used for fold changes) and log of ratio of signal to control (used for clustering and statistics) were assigned to each gene.
|
Data analysis approach.
Of the 31,099 probe sets on the rat genome 230 2.0 microarray, only 3397 of these were changed twofold or more in at least one of the epidermal samples, and 2252 of these were statistically significant at p
0.05. The Welch ANOVA, which assumed that variances are not equal, was used to test the probes showing twofold change for statistical significance. For further data analysis, the three most similar (of the four analyzed) gene chips in each of the component treatment groups were selected using the GeneSpring "find similar samples" function with a weighting coefficient of 0.25 for the intensity. The box plot in Figure 1 demonstrates the results of normalization and quality control and reveals the overall similarity of the distributions. Post hoc (Student-Neuman-Keuls) analysis of all groups was carried out at p 0.05 to determine differences between groups. Categorization of gene changes by Gene Ontology (GO) was accomplished with GO SLIMS classification system built 14 February 2007 with annotations from Entrez Gene as the primary source. Hierarchical clustering was accomplished using GeneSpring to group the treatment conditions according to expression data chosen based on JP-8–induced inflammatory and growth-related pathways. The "average linkage" clustering algorithm was used with Pearson correlation as a similarity measure. The hierarchical relationship is shown in a tree structure.
|
Gene expression with real-time RT-PCR.
Total RNA was purified as described above. The RNA was quantified using the Experion StdSens RNA Analysis Kit (Bio-Rad). One hundred nanograms of total mRNA was then amplified using a Full Spectrum Complete Transcriptome RNA Amplification Kit (System Biosciences, Mountain View, CA) according to the manufacturer's protocol. Resultant cDNA from the amplification was quantified using the Fluorescent DNA Quantitation Kit (Bio-Rad). Real-time RT-PCR analysis of gene transcripts involved the use of the Applied Biosystems 7900HT Sequence Detection System, TaqMan chemistry, and Applied Biosystems TaqMan Gene Expression Assays according to manufacturer's recommended procedures. Data were analyzed using the comparative Ct method (also known as the
Ct method) to calculate relative changes in gene expression (Livak and Schmittgen, 2001). Fold change (the relative quantitation [RQ], value of the target gene) was calculated from the Ct. This fold change was normalized to the endogenous reference gene, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and was reported relative to the sham-treated samples. These calculations were repeated using each sham-treated sample as the calibrator, for a total of four different RQ values for each sample (RQ1, RQ2, RQ3, and RQ4). RQ values for each sample were averaged and the standard error was calculated, yielding the average fold change of the gene of interest. A two-tailed paired t-test of the control and treated sample groups for each gene of interest at each time point was performed to determine statistical significance at p 0.05.
Functional/pathway analysis.
As previously described (McDougal et al., 2007), lists of genes with significant changes in gene expression (increases and decreases) based on the microarray experiments were moved from GeneSpring into Ingenuity Pathways Analysis 4.0 (IPA) (Ingenuity Systems, http://www.ingenuity.com). Pathways analysis identified pathways from the IPA library of canonical pathways that were significantly changed by JP-8 treatment. The significance of the association between the JP-8 data set and the canonical pathway was determined using a ratio of the number of genes from the data set that map to the pathway divided by the total number of genes that map to the canonical pathway. Fischer's exact test was used to calculate p value. Genes changed by JP-8 treatment that were associated with specific pathways were grouped based on a primary relationship to either inflammation or growth. Pathway genes were categorized by location in the cell with IPA and imported back into GeneSpring for comparison with gene expression of the components.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Concentrations of the Hydrocarbons in the Skin
Molar concentrations of the aliphatic and aromatic chemicals in the skin at the end of the 1-h exposures were determined as described in the "Materials and Methods" section. Figure 2A shows the concentrations of the hydrocarbons after JP-8 exposure. The aliphatic chemicals (UND and TET) were present at higher concentrations than the aromatic chemicals (TMB and DMN), p
0.001. Figure 2B shows the concentrations of the hydrocarbons after exposure to the pure chemicals. UND, TMB, DMN, and TET were detectible in the epidermis at 1 h, but the differences were not significant by ANOVA (p = 0.25). The concentrations of the aliphatic components in the epidermis after JP-8 exposure (Fig. 2A) are nearly equivalent to the concentrations of the components when they are given as pure chemical (Fig. 2B).
|
Changes in Gene Expression
The total number of genes significantly changed at least twofold compared to sham treatment are shown in Table 3. JP-8, UND, and TMB caused changes in the largest number of genes followed by DMN and TET. One-way ANOVA post hoc analysis (Student-Neuman-Keuls) showed that the UND and TMB responses were similar, but JP-8, DMN, and TET were all different. In each treatment, there were more genes decreased than increased, ranging from 57 (UND) to 78% (TMB) of the total genes changed. Supplementary Tables 1–5 describe the fold change and statistical significance of each of the changed probe sets for JP-8 and each component treatment. Confirmation of changes in gene expression was accomplished with real-time PCR. Twenty-four genes were chosen for comparison with GeneChip expression based on their presence in the skin and involvement in the inflammatory process. In general, results showed that PCR values were generally higher than GeneChip values but mimicked the general behavior of the GeneChip probes (Supplementary Table 6).
|
GO Classification of Changed Genes
Categorization of the genes changed according to biological process (generic GO SLIMS) shows that JP-8 and the components displayed quantitative and qualitative differences in the responses (Fig. 3). Some genes related to development (with its subcategories morphogenesis, cell differentiation, and embryonic development) were changed by JP-8 and all components but TMB. UND and JP-8 caused very similar responses in the development category. Some genes in the metabolism categories (with its subcategories biosynthesis and catabolism) were changed by JP-8, TMB, and DMN. The majority (69–79%) of genes changed by TMB in the metabolism category were decreased. Genes in the behavior category were only changed by UND and TET. Cell death genes were significantly changed only by JP-8 and DMN, and genes related to cell growth were only changed by TET. None of the components exactly mimicked all the JP-8–induced changes in biological process, although there was significant overlap with all the components except TMB. Interestingly, TMB, which caused changes in the greatest total number of genes, primarily affected metabolic genes.
|
Functional Analysis of Changed Genes
Categorization of the genes changed by JP-8 according to IPA function, and the responses of the components in those categories are shown in Table 4. The three most significant changes in function due to JP-8 treatment were cell death, cellular growth and proliferation, and cellular movement. There were 169 gene transcripts specifically related to apoptosis (subcategory of cell death) that were changed by the JP-8 treatment and nearly equal numbers that had growth (153) and proliferation (160) functions. The three major functions were also significantly affected in each of the four component treatments, but the order of significance varied with the components and none of the components had exactly the same top three as JP-8. For example, with UND, the function that changed most significantly was cellular growth and proliferation and cell death was number four. The effects of the components on functional changes in gene expression were consistent, i.e., gene expression for each component in the functional categories (Table 4) was roughly proportional to the overall number of genes changed by that component (Table 3).
|
Signaling Pathways Affected by the Treatments
JP-8–induced fold changes in gene expression were imported into IPA to investigate epidermal signaling pathways that were affected by jet fuel. With 1111 transcripts changed, 530 of these had functional and pathway information in the Ingenuity Pathways Knowledge Base. These 530 genes mapped to 20 signaling pathways (Table 5) that were changed greater than would be expected by chance (p
0.05). These pathways, except xenobiotic metabolism, can be divided into categories of inflammation or growth, proliferation, and apoptosis as shown in Table 5. Changes in gene expression in each of the treatment groups for genes from the inflammation pathways are shown in Table 6, and changes in gene expression that are related to growth, proliferation, and apoptosis pathways are shown in Table 7. Inflammatory pathways and growth-related pathways are nearly equally represented, with interleukin-6 (IL-6) and epidermal growth factor pathways predominating in their respective categories. Twenty inflammatory genes from those pathways and 14 genes from the growth-related pathways characterize the JP-8 response and summarize it in a way that can be compared with the component gene responses.
|
|
|
Specific Inflammatory and Growth-Related Genes Affected by the Treatments
Cellular distribution of the characteristic JP-8–altered inflammatory and growth-related genes in the cell is shown in Figures 4A and 4C, respectively. Protein products of the inflammatory genes that were changed localize primarily in the extracellular space and plasma membrane. In contrast, protein products of the growth-related transcripts distribute primarily in the cytoplasm and nucleus. Ninety percent (18/20) of the genes involved in the inflammatory pathways were upregulated, and 64% (9/14) of the genes in the growth-related pathways were upregulated. Heatmap visualization of hierarchical clustering with both the inflammatory (Fig. 4B) and growth-related (Fig. 4D) genes shows the same clustering pattern of response by the components. Hierarchical clustering shows that UND responds most like JP-8, and TET responds least like JP-8 with both the inflammatory and growth-related genes. These clustering results with selected genes loosely mimic the results of the post hoc analysis of the complete genome (Table 3).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
We examined the changes in epidermal gene expression at the end of brief exposures to JP-8 and four representative components, in an attempt to understand whether differences in effects of hydrocarbon components in JP-8 might be responsible for the irritation of the mixture. Each of the five treatments was compared with a "sham" treatment, an empty application device placed on the skin for the same length of time. During each study, we also measured the concentration of the hydrocarbons in the epidermis to determine the amount of chemical in the target tissue. Classification and functional/pathway analysis of gene changes for each of the four components were compared with the JP-8 response. We addressed two main questions. First, is the response of any component similar to the response of the fuel mixture? Second, can we estimate the potency of each component as a function of the gene expression response?
Changes in JP-8–induced gene expression in the current study were similar to the 1-h changes with JP-8 in our previous time-course study (McDougal et al., 2007), where we found changes in pathways related to inflammation, apoptosis, and growth. Rapid improvements in gene annotations, chip technology, and software tools make direct gene-to-gene comparisons between different generation Affymetrix GeneChips nearly impossible, but more general functional and pathway comparisons are possible. Although an updated Affymetrix GeneChip (31,099 transcripts vs. 8798 transcripts) and an improved RMA normalization method were used, 9 of 10 signaling pathways previously identified as changed (McDougal et al., 2007) were also changed in this study. Furthermore, the larger number of transcripts on the chip and improved sensitivity with better normalization enabled 20 signaling pathways to be identified as changed in this study (Table 5). Also, the GO categorization in the current study (Fig. 3) showed more categories changed than the previous study; however, there is extensive overlap between the functional categories in these studies. Based on these comparisons and recognition of many important cytokine, chemokine, transcription factor, and growth factor genes that were changed, we conclude that the currently observed responses to JP-8 are consistent with our previous studies.
Categorization of the biological processes by GO (Fig. 3) showed diversity but some similarities between the components. The fact than none of the aliphatic or aromatic components exactly mimicked the responses of a mixture should not be surprising for several reasons. First, JP-8 is a complex mixture (many constituents with varying concentrations) having the potential for a large number of chemical effects and mechanisms of toxicity. Second, there is a great deal of controversy about how to assess the toxicity of mixtures using a mix of surrogate chemicals (Lampe et al., 1983; Mumtaz et al., 2002; Tinwell and Ashby, 2004), and it would be unrealistic to expect one component to mimic the effects of a complex mixture. UND, however, does cause changes in development, morphogenesis, and cell differentiation that are very similar to those caused by JP-8 (Fig. 3). TMB caused dramatic transcript decreases in metabolism categories that were not seen with any other treatments; however, decreases in transcripts related to metabolism were previously identified within 8 h after a 1-h JP-8 exposure (McDougal et al., 2007). GO categorization does not clearly identify a hydrocarbon component that has the same spectrum of responses as JP-8. Unique chemical or physical characteristics of each molecule appear to effect gene expression differently during exposure.
Another comparison approach is to identify significant JP-8 responses and focus on those functions, pathways, and genes when investigating changes due to component treatments. Table 4 shows some results of this approach based on the top three functions changed by JP-8. The number of genes changed related to these functions for each of the components was proportional to the overall response. This suggests that the mechanisms by which the JP-8 and the component chemicals cause changes in gene expression related to these functions are probably similar, and the differences are primarily in the magnitude of the response. When only specific pathways are considered, the difference in response may reflect the potency of each chemical.
Comparison of the individual component responses with the JP-8 response was facilitated by "distilling" the JP-8 responses to specific genes in inflammatory and growth-related pathways as described in the Materials and Methods section. Figure 4 illustrates important inflammatory and growth-related genes changed by JP-8 along with their cellular location and provides a visual comparison for each of the components. Twenty inflammatory genes that responded to JP-8 exposure and the corresponding fold change for each of the five treatments are listed in Table 6. Tumor necrosis factor (TNF) and IL-6 are major inflammatory cytokines that have many connections to other genes in the pathways (Fig. 4C). Both TNF and IL-6 transcripts were increased by JP-8 in our earlier study (McDougal et al., 2007). TNF is produced by keratinocytes and Langerhans cells in response to a variety of noxious stimuli (Corsini and Galli, 1998; Luster et al., 1999), resulting in the production of IL-6 and several other secondary inflammatory mediators (Luger, 1989; Luster et al., 1999). IL-6 is a secondary proinflammatory cytokine (produced by keratinocytes and stimulated by IL-1
and TNF) that activates a variety of local defense mechanisms aimed at limiting tissue injury (Sehgal, 1990). The levels of transcripts coding for IL-6 and TNF are increased by JP-8 treatment as well as by UND, TMB, and DMN treatments. These two transcripts were not increased by TET, and most of the other inflammatory genes increased by JP-8 were not significantly changed by TET (Table 6).
The chemokines, CCL2 and CCL7, show the same pattern as IL-6 and TNF, in which all components caused upregulation except TET. Chemokines are small molecules that are induced by the primary inflammatory proteins to be involved in activation and recruitment of immunocompetent and inflammatory cells to sites of damage (Foster, 2001; Rossi and Zlotnik, 2000). CCL2 (or MCP-1) is a chemoattractant for monocytes and lymphocytes (Moller et al., 2003), and CCL7 (or MCP-3) is a chemotactic for monocytes (Kondo et al., 2000). Additionally, CXCL12 and its receptor CXCR4 were only upregulated by JP-8 and UND. CXCL12 (or stromal cell–derived factor) is chemotactic for mononuclear cells (Bleul et al., 1996), and both CXCL12 and CXCR4 have previously been identified in keratinocytes and Langerhans cells (Chen et al., 2004; Tchou et al., 2001). These results show that UND-induced changes with this group of 20 inflammatory genes are most like JP-8–induced responses.
Based on these gene responses at the end of a 1-h exposure, it appears that TET does not participate in the inflammatory process to the same extent as JP-8, UND, TMB, and DMN. However, there is evidence that TET does cause irritation when applied to the skin for several days (Brooks and Baumann, 1956; Brown and Box, 1970; Moloney and Teal, 1988). Muhammad et al. (2005) also showed that TET produced greater erythema scores than other aliphatic hydrocarbons when applied to pigs in soaked cotton fabric for 24 or 96 h. However, our gene expression study suggests that this response may be due to a different mechanism than JP-8–induced irritation. In these 1-h exposures, before we see redness or other visible changes in the skin, we expect that the earliest event, or the trigger, of the irritant process will be seen as changes in gene expression. It is possible that prolonged levels of TET are required to cause the inflammatory response. It is interesting, however, that UND very closely mimicked the JP-8–induced changes in the inflammatory genes. In the soaked cotton pig study (Muhammad et al., 2005), neither JP-8 nor UND or any of the six aromatics tested (including DMN and TMB) caused visible erythema after 24 h of exposure. There may be many different mechanisms by which skin irritation can be initiated.
Tissue injury has been shown to rapidly activate repair processes (Cole et al., 2001; Icre et al., 2006). With chemical injury in the liver, repair has been suggested to be rapid and simultaneous with the process of injury (Rao et al., 1997). Many growth factors are mitogenic for epidermal and dermal skin cells (Tomic-Canic et al., 1998). JP-8 has been reported to have effects on both cell growth and apoptosis (Espinoza et al., 2004; McDougal et al., 2007; Rosenthal et al., 2001). In this study, TNF and insulin-like growth factor (IGF1, also known as somatomedin C) have the greatest number of connections with other changed genes that were selected based on growth-related pathway responses (Fig. 4C). In addition to the inflammatory effects of TNF discussed above, it also has important properties related to apoptosis and cell repair (Luster et al., 1999; Weisfelner and Gottlieb, 2003). Receptors for TNF are found on nearly all nucleated cells where they activate the extracellular signal-regulated kinase/mitogen-activated protein kinase (ERK/MAPK) pathway resulting in activation of nuclear transcription factors and expression of proteins related to stimulation of growth (Luster et al., 1999). In this study, TNF transcripts were increased by all treatments except TET. In the epidermis, IGF1 predominantly affects the promotion of cell growth through mitotic action and antiapoptotic activity (Edmondson et al., 2003). The IGF system promotes keratinocyte survival and protection from apoptosis after ultraviolet B irradiation (Kuhn et al., 1999). Only UND mimicked the changes in IGF1 with JP-8, TMB and DMN showed no change, and TET showed a twofold decrease (Table 7). When hydrocarbon component–induced changes in the 14 selected growth-related genes are considered as a whole, UND almost completely mimicked the magnitude of the JP-8–induced changes. The only discrepancy was lack of significant changes by UND in PIK3R1 and RXRA. In contrast, TET only had similar responses to JP-8 with two growth-related genes, JUN and RHOB. The comparisons of growth-related gene changes between the hydrocarbon components and JP-8 are consistent with the inflammation-related gene changes because hierarchical clustering shows the same relationship of the components.
In this study, we measured the amount of hydrocarbons in the skin after exposures to JP-8 and individual hydrocarbon components. Estimating the amount of chemical in the skin is a way to attempt to understand the local effects (such as changes in gene expression) of chemicals. The aliphatic components are absorbed into and remain in the skin to a greater extent than the aromatics during JP-8 exposures (Kim et al., 2006; McDougal et al., 2000; Riviere et al., 1999). Our study confirms these results in the epidermis and shows that the concentration of aliphatic components is five to six times greater than the concentration of the aromatic components (Fig. 2A). In contrast to the JP-8 exposures, Figure 2B shows that there is no statistically significant difference in component concentrations after a 1-h exposure to the pure chemical. The measured epidermal concentrations of the aliphatic components from the JP-8 exposure are higher than might be expected if the concentrations of the hydrocarbons in JP-8 were the only driving force. This can be explained by the effect of the remaining jet fuel components on UND and TET partitioning into the epidermis from the fuel mixture. The vehicle (in this case, JP-8) can have a pronounced effect on the absorption because it is the effective thermodynamic activity that is the driving force (Barry et al., 1985; Higuchi, 1960; Jepson and McDougal, 1999).
A central tenant of pharmacology and toxicology is that the target tissue dose is the determinant of the response (Eaton and Klaassen, 2001; Ross and Kenakin, 2001); therefore, we could expect the gene expression response for any chemical to be greater if the amount of chemical in the tissue was larger. We also expect that chemicals may differ in potency, and similar target tissue doses of two different chemicals might cause responses (homeostatic, beneficial, or toxic) that differ in magnitude. Gene expression studies are useful to help understand mechanisms of toxicity and have been shown to discriminate between categories of genotoxic compounds (Hu et al., 2004; van Delft et al., 2004). For example, the carcinogenic potency of polycyclic aromatic hydrocarbons (PAHs) in liver slices was investigated with gene expression studies where PAHs with low carcinogenic potency were distinguishable from those with high carcinogenic potency (Staal et al., 2007), and these authors were able to distinguish PAHs with low carcinogenic potency from those with high carcinogenic potency. Functional and pathway analyses of the changes in gene expression may be more useful in determining the character of the response to the chemical. In our study, the number of total (Table 3) and functional (Table 4) gene changes was higher for UND and TMB than they were for DMN and TET, although the molar concentrations in the epidermis were not significantly different (Fig. 2B). This suggests that DMN and TET are less potent than UND and TMB on a molar basis. As far as we know, this is the first study to attempt to relate in vivo changes in gene expression to measured target tissue concentration.
In summary, we have expanded and confirmed gene transcript changes in the epidermis due to a brief JP-8 exposure in comparison with the responses of two aliphatic and two aromatic JP-8 components. Chemical-induced changes in gene expression with the components showed consistent differences in magnitude, whether total gene expression or functional pathway–specific changes were investigated. UND and TMB caused the greatest number of gene changes, more than twice as many as DMN, and about 10-fold more than TET. GO categorization of the total number of changes showed that no component mimicked the JP-8 response, indicating that there are qualitative differences in gene expression. When only the genes related to specific pathways or functions changed by JP-8 were considered, we found that these pathways were nearly all activated by the components, but to different extents. Analysis of chemical concentrations in the skin after the component exposures indicated that the 10-fold differences in gene expression response were not due to different target tissue (epidermis) concentrations. UND and TMB appear to be more potent inducers of gene expression in the epidermis than DMN and TET, but both aliphatic and aromatic compounds cause responses that may result in irritation. We conclude that no one component will mimic the response of the complex JP-8 mixture, but in our study UND had the most similar responses. It also appears that TET causes minimal gene expression response after a 1-h exposure. This study does not allow us to state that either aromatics or aliphatics are responsible for JP-8–induced skin irritation. However, the data do suggest that there are differences in potency for total gene expression as well as specific functional and pathway responses.
|
SUPPLEMENTARY DATA |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Supplementary data are available online at http://toxsci.oxfordjournals.org/.
|
ACKNOWLEDGMENTS |
|---|
The authors thank Dr James Rogers (Battelle Biomedical Research Center) for reading the manuscript and providing valuable comments. We also acknowledge and appreciate the excellent support of the Dr Steven Berberich and the Wright State University Center for Genomics Research. The authors gratefully acknowledge grant support from the Air Force Office of Scientific Research (AFOSR/NL). Conflicts of interest: none declared.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Agency for Toxic Substances and Disease Registry. Toxicological Profile for Jet Fuels (JP-5 and JP-8) (1998) Atlanta, GA: U.S. Department of Health and Human Services.
Ahaghotu E, Babu RJ, Chatterjee A, Singh M. Effect of methyl substitution of benzene on the percutaneous absorption and skin irritation in hairless rats. Toxicol. Lett. (2005) 159(3):261–271.[CrossRef][ISI][Medline]
Allen DG, Riviere JE, Monteiro-Riviere NA. Identification of early biomarkers of inflammation produced by keratinocytes exposed to jet fuels jet A, JP-8, and JP-8(100). J. Biochem. Mol. Toxicol. (2000) 14(5):231–237.[CrossRef][ISI][Medline]
Baker W, Dodd D, McDougal JN, Miller TE. Repeated Dose Skin Irritation Study on Jet Fuels—A Histopathological Study. AFRL-HE-WP-TR-1999-0022 (1999) Wright Patterson AFB, OH: Air Force Research Laboratory.
Barry BW, Harrison SM, Dugard PH. Vapour and liquid diffusion of model penetrants through human skin; correlation with thermodynamic activity. J. Pharm. Pharmacol. (1985) 37:226–235.[ISI][Medline]
Bleul CC, Farzan M, Choe H, Parolin C, Clark-Lewis I, Sodroski J, Springer TA. The lymphocyte chemoattractant SDF-1 is a ligand for LESTR/fusin and blocks HIV-1 entry. Nature (1996) 382(6594):829–833.[CrossRef][Medline]
Bolstad BM, Irizarry RA, Astrand M, Speed TP. A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics (2003) 19(2):185–193.
Boman A. Irritants—organic solvents. In: The Irritant Contact Dermatitis Syndrome—van der Valk PGM, Maibach HI, eds. (1996) Boca Raton, FL: CRC Press. 95–104.
Brooks SC, Baumann CA. Skin sterols. X. Studies with certain hyperplasia-inducing hydrocarbons. Cancer Res. (1956) 16(4):357–363.
Brown VK, Box VL. Skin arginase activity as a measure of skin change under the influence of some alkanes and alkenes. Br. J. Dermatol. (1970) 82(6):606–612.[CrossRef][ISI][Medline]
Chen H, Zha J, Gowans RE, Camargo P, Nishitani J, McQuirter JL, Cole SW, Zack JA, Liu X. Alcohol enhances HIV type 1 infection in normal human oral keratinocytes by up-regulating cell-surface CXCR4 coreceptor. AIDS Res. Hum. Retroviruses (2004) 20(5):513–519.[CrossRef][ISI][Medline]
Chou CC, Riviere JE, Monteiro-Riviere NA. The cytotoxicity of jet fuel aromatic hydrocarbons and dose-related interleukin-8 release from human epidermal keratinocytes. Arch. Toxicol. (2003) 77(7):384–391.[CrossRef][ISI][Medline]
Chou CC, Yang JH, Chen SD, Monteiro-Riviere NA, Li HN, Chen JJ. Expression profiling of human epidermal keratinocyte response following 1-minute JP-8 exposure. Cutan. Ocul. Toxicol. (2006) 25(2):141–153.[ISI][Medline]
Cole J, Tsou R, Wallace K, Gibran N, Isik F. Early gene expression profile of human skin to injury using high-density cDNA microarrays. Wound Repair Regen. (2001) 9(5):360–370.[CrossRef][ISI][Medline]
Committee on Toxicology of National Research Council. Permissible Exposure Levels for Selected Military Fuel Vapors (1996) Washington, DC: National Academy Press.
Committee on Toxicology of National Research Council. Toxicologic Assessment of Jet-Propulsion Fuel 8 (2003) Washington, DC: National Academy Press.
Corsini E, Galli CL. Cytokines and irritant contact dermatitis. Toxicol. Lett. (1998) 102–103:277–282.
Dietzel KD, Campbell JL, Bartlett MG, Witten ML, Fisher JW. Validation of a gas chromatography/mass spectrometry method for the quantification of aerosolized Jet Propellant 8. J. Chromatogr. A. (2005) 1093(1–2):11–20.[CrossRef][ISI][Medline]
Eaton D, Klaassen C. Principles of toxicology. In: Casarett & Doull's Toxicology: The Basic Science of Poisons—Klaassen C, ed. (2001) 6th ed. New York: McGraw-Hill. 11–34.
Edmondson SR, Thumiger SP, Werther GA, Wraight CJ. Epidermal homeostasis: The role of the growth hormone and insulin-like growth factor systems. Endocr. Rev. (2003) 24(6):737–764.
Espinoza LA, Li P, Lee RY, Wang Y, Boulares AH, Clarke R, Smulson ME. Evaluation of gene expression profile of keratinocytes in response to JP-8 jet fuel. Toxicol. Appl. Pharmacol. (2004) 200(2):93–102.[CrossRef][ISI][Medline]
Foster JR. The functions of cytokines and their uses in toxicology. Int. J. Exp. Pathol. (2001) 82(3):171–192.[CrossRef][ISI][Medline]
Gallucci RM, O'Dell SK, Rabe D, Fechter LD. JP-8 jet fuel exposure induces inflammatory cytokines in rat skin. Int. Immunopharmacol. (2004) 4(9):1159–1169.[CrossRef][ISI][Medline]
Higuchi T. Physical chemical analysis of percutaneous absorption process from creams and ointments. J. Soc. Cosmet. Chem ll (1960) 2(2):85–97.
Hoekstra WG, Phillips PH. Effects of topically applied mineral oil fractions on the skin of guinea-pigs. J. Invest. Dermatol. (1963) 40:79–88.[ISI][Medline]
Hu T, Gibson DP, Carr GJ, Torontali SM, Tiesman JP, Chaney JG, Aardema MJ. Identification of a gene expression profile that discriminates indirect-acting genotoxins from direct-acting genotoxins. Mutat. Res. (2004) 549(1–2):5–27.[ISI][Medline]
Icre G, Wahli W, Michalik L. Functions of the peroxisome proliferator-activated receptor (PPAR) alpha and beta in skin homeostasis, epithelial repair, and morphogenesis. J. Investig. Dermatol. Symp. Proc. (2006) 11(1):30–35.[CrossRef][Medline]
Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, Speed TP. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res. (2003a) 31(4):e15.
Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U, Speed TP. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics (2003b) 4(2):249–264.[Abstract]
Jepson GW, McDougal JN. Predicting vehicle effects on the dermal absorption of halogenated methanes using physiologically based modeling. Toxicol. Sci. (1999) 48(2):180–188.
Kabbur MB, Rogers JV, Gunasekar PG, Garrett CM, Geiss KT, Brinkley WW, McDougal JN. Effect of JP-8 jet fuel on molecular and histological parameters related to acute skin irritation. Toxicol. Appl. Pharmacol. (2001) 175(1):83–88.[CrossRef][ISI][Medline]
Kanikkannan N, Patel R, Jackson T, Shaik MS, Singh M. Percutaneous absorption and skin irritation of JP-8 (jet fuel). Toxicology (2001) 161(1–2):1–11.[CrossRef][ISI][Medline]
Kim D, Andersen ME, Nylander-French LA. Dermal absorption and penetration of jet fuel components in humans. Toxicol. Lett. (2006) 165(1):11–21.[CrossRef][Medline]
Klauder JV, Brill FA. Correlation of boiling ranges of some petroleum solvents with irritant action on skin. Arch. Dermatol. (1947) 56:197–215.
Kondo A, Isaji S, Nishimura Y, Tanaka T. Transcriptional and post-transcriptional regulation of monocyte chemoattractant protein-3 gene expression in human endothelial cells by phorbol ester and cAMP signalling. Immunology (2000) 99(4):561–568.[CrossRef][ISI][Medline]
Koschier FJ. Toxicity of middle distillates from dermal exposure. Drug Chem. Toxicol. (1999) 22(1):155–164.[ISI][Medline]
Kuhn C, Hurwitz SA, Kumar MG, Cotton J, Spandau DF. Activation of the insulin-like growth factor-1 receptor promotes the survival of human keratinocytes following ultraviolet B irradiation. Int. J. Cancer (1999) 80(3):431–438.[CrossRef][ISI][Medline]
Lampe MA, Williams ML, Elias PM. Human epidermal lipids: Characterization and modulations during differentiation. J. Lipid Res. (1983) 24(2):131–140.[Abstract]
Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods (2001) 25(4):402–408.[CrossRef][ISI][Medline]
Luger TA. Epidermal cytokines. Acta Derm. Venereol. Suppl. (Stockh.) (1989) 151:61–76.[Medline]
Luster MI, Simeonova PP, Gallucci R, Matheson J. Tumor necrosis factor alpha and toxicology. Crit. Rev. Toxicol. (1999) 29(5):491–511.[CrossRef][ISI][Medline]
McDougal JN, Garrett CM, Amato CM, Berberich SJ. Effects of brief cutaneous JP-8 jet fuel exposures on time course of gene expression in the epidermis. Toxicol. Sci. (2007) 95(2):495–510.
McDougal JN, Pollard DL, Weisman W, Garrett CM, Miller TE. Assessment of skin absorption and penetration of JP-8 jet fuel and its components. Toxicol. Sci. (2000) 55(2):247–255.
McDougal JN, Rogers JV. Local and systemic toxicity of JP-8 from cutaneous exposures. Toxicol. Lett. (2004) 149(1–3):301–308.[CrossRef][ISI][Medline]
Moller AS, Ovstebo R, Westvik AB, Joo GB, Haug KB, Kierulf P. Effects of bacterial cell wall components (PAMPs) on the expression of monocyte chemoattractant protein-1 (MCP-1), macrophage inflammatory protein-1alpha (MIP-1alpha) and the chemokine receptor CCR2 by purified human blood monocytes. J. Endotoxin. Res. (2003) 9(6):349–360.[CrossRef][Medline]
Moloney SJ, Teal JJ. Alkane-induced edema formation and cutaneous barrier dysfunction. Arch. Dermatol. Res. (1988) 280(6):375–379.[CrossRef][ISI][Medline]
Monteiro-Riviere N, Inman A, Riviere J. Effects of short-term high-dose and low-dose dermal exposure to Jet A, JP-8 and JP-8 + 100 jet fuels. J. Appl. Toxicol. (2001) 21(6):485–494.[CrossRef][ISI][Medline]
Monteiro-Riviere NA, Inman AO, Riviere JE. Skin toxicity of jet fuels: Ultrastructural studies and the effects of substance P. Toxicol. Appl. Pharmacol. (2004) 195(3):339–347.[CrossRef][ISI][Medline]
Muhammad F, Brooks JD, Riviere JE. Comparative mixture effects of JP-8(100) additives on the dermal absorption and disposition of jet fuel hydrocarbons in different membrane model systems. Toxicol. Lett. (2004) 150(3):351–365.[CrossRef][ISI][Medline]
Muhammad F, Monteiro-Riviere NA, Riviere JE. Comparative in vivo toxicity of topical JP-8 jet fuel and its individual hydrocarbon components: Identification of tridecane and tetradecane as key constituents responsible for dermal irritation. Toxicol. Pathol. (2005) 33(2):258–266.[CrossRef][ISI][Medline]
Mumtaz MM, Tully DB, El-Masri HA, De Rosa CT. Gene induction studies and toxicity of chemical mixtures. Environ. Health Perspect. (2002) 110(Suppl. 6):947–956.[ISI][Medline]
Ramos G, Kazimi N, Nghiem DX, Walterscheid JP, Ullrich SE. Platelet activating factor receptor binding plays a critical role in jet fuel-induced immune suppression. Toxicol. Appl. Pharmacol. (2004) 195(3):331–338.[CrossRef][ISI][Medline]
Rao PS, Mangipudy RS, Mehendale HM. Tissue injury and repair as parallel and opposing responses to CCl4 hepatotoxicity: A novel dose-response. Toxicology (1997) 118(2–3):181–193.[CrossRef][ISI][Medline]
Riviere JE, Brooks JD, Monteiro-Riviere NA, Budsaba K, Smith CE. Dermal absorption and distribution of topically dosed jet fuels jet-A, JP-8, and JP-8(100). Toxicol. Appl. Pharmacol. (1999) 160(1):60–75.[CrossRef][ISI][Medline]
Rogers JV, Gunasekar PG, Garrett CM, Kabbur MB, McDougal JN. Detection of oxidative species and low-molecular-weight DNA in skin following dermal exposure with JP-8 jet fuel. J. Appl. Toxicol. (2001) 21(6):521–525.