ToxSci Advance Access originally published online on May 5, 2009
Toxicological Sciences 2009 110(1):168-180; doi:10.1093/toxsci/kfp091
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Neurotoxicogenomic Investigations to Assess Mechanisms of Action of the Munitions Constituents RDX and 2,6-DNT in Northern Bobwhite (Colinus virginianus)
* U.S. Army Corps of Engineers, Environmental Laboratory, EP-P, Vicksburg, MS 39180
The Johns Hopkins University, School of Medicine, Baltimore, MD 21287
U.S. Army Center for Health Promotion and Preventative Medicine, Edgewood, MD 21010
SpecPro Inc., ERDC-USACE-EL-EP-P, Vicksburg, MS 39180
1 To whom correspondence should be addressed. Fax: 601-634-4002. E-mail: kurt.a.gust{at}usace.army.mil.
Received February 9, 2009; accepted April 20, 2009
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Munitions constituents (MCs) including hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX), 2,4,6-trinitrotoluene (TNT), and TNT derivatives are recognized to elicit aberrant neuromuscular responses in many species. The onset of seizures resulting in death was observed in the avian model Northern bobwhite after oral dosing with RDX beginning at 8 mg/kg/day in subacute (14 days) exposures, whereas affective doses of the TNT derivative, 2,6-dinitrotoluene (2,6-DNT), caused gastrointestinal impacts, lethargy, and emaciation in subacute and subchronic (60 days) exposures. To assess and contrast the potential neurotoxicogenomic effects of these MCs, a Northern bobwhite microarray was developed consisting of 4119 complementary DNA (cDNA) features enriched for differentially-expressed brain transcripts from exposures to RDX and 2,6-DNT. RDX affected hundreds of genes in brain tissue, whereas 2,6-DNT affected few (
17), indicating that 2,6-DNT exposure had relatively little impact on the brain in comparison to RDX. Birds exhibiting RDX-induced seizures accumulated over 20x more RDX in brain tissues in comparison to non-seizing birds even within a common dose. In parallel, expression patterns were unrelated among seizing and non-seizing birds exposed to equivalent RDX doses. In birds experiencing seizures, genes related to neuronal electrophysiology and signal transduction were significantly affected. Comparative toxicology revealed strong similarity in acute exposure effects between RDX and the organochlorine insecticide dichlorodiphenyltrichloroethane (DDT) regarding both molecular mechanisms and putative mode of action. In a manner similar to DDT, we hypothesize that RDX elicits seizures by inhibition of neuronal cell repolarization postaction potential leading to heightened neuronal excitability and seizures facilitated by multiple molecular mechanisms. Key Words: munitions constituents; genomics; mechanisms of action; seizures.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
In order to protect wild avian species that inhabit military installations where energetic compounds may be present, it is essential to establish the toxicity of these chemicals and their breakdown products. Environmental contamination by munitions constituents (MCs) primarily occurs in soils at munitions manufacturing plants, load and pack operations, firing ranges, and demilitarization areas (Jenkins et al., 2001
). Given their behavior, ground foraging birds have the potential to be exposed to soils contaminated with MCs through accidental and intentional ingestion of soil and grit to assist in digestion (Brennan, 1999). The Northern bobwhite (Colinus virginianus) was chosen as a model species for this research because of its characteristic behavior as a ground forager, widespread geographical distribution, availability via commercial breeders, and demonstrated amenability to laboratory study (Baker et al., 2004; Gogal et al., 2002; Johnson et al., 2005). Additionally, although Northern bobwhite distributions in nature have been historically widespread, Northern bobwhite has recently become a species of concern whose populations have been declining in many states (Sauer et al., 2004).
Characterization of MC toxicity and the mechanisms underlying potential toxic effects are key to establishing critical parameters involved in field risk assessments for wildlife. Neurological impacts of MCs including hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX), 2,4,6-trinitrotoluene (TNT), and other nitroaromatics have been observed in a variety of species (Goldberg et al., 1992; Levine et al., 1983; Talmage et al., 1999) including avian species (Johnson et al., 2007; Quinn et al., 2009). One of the key strengths of toxicogenomic investigations is the ability to assay thousands of biologically relevant molecular targets that may be directly or indirectly impacted by toxicant exposure. This characteristic has gained recognition for identification of mechanisms of toxic action (Ankley et al., 2006; Perkins et al., 2007) as well as providing diagnostics of chemical exposure (Poynton et al., 2007).
The purpose of this study was to develop and use a novel complementary DNA (cDNA) microarray to assess the neurotoxicogenomic effects of the MC compounds RDX and 2,6-dinitrotoluene (2,6-DNT) to determine mechanisms of action in the Northern bobwhite avian model. Microarray construction and genomic inquiry were conducted utilizing RNA samples collected in experiments investigating RDX and 2,6-DNT toxicity in Northern bobwhite (Johnson et al., 2007; Quinn et al., 2007, 2009). In these studies, affective doses of RDX were observed to cause a progression of tremors, tonic seizures, and clonic seizures followed by death in Northern bobwhite. In contrast, 2,6-DNT caused impaired gastrointestinal function, altered blood chemistry, anemia, dehydration, edema in visceral organs, lethargy, emaciation, and death. Although no overt neurological effects were observed in 2,6-DNT exposures, we were interested in determining if the integrative properties of central nervous system (CNS) may act as a remote monitor for stresses occurring in various organ systems. We tested the hypothesis that exposure to RDX or 2,6-DNT would alter gene expression profiles in Northern bobwhite brain tissue in a manner indicative of potential modes through which RDX or 2,6-DNT caused toxicity.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Detailed animal husbandry methods, exposure techniques, and clinical toxicology are described in Johnson et al. (2007)
, Quinn et al. (2007), and Quinn et al. (2009). A summary of experimental exposures is provided below. All protocols were conducted consistent with Good Laboratory Practices and approved by the Institutional Animal Care and Use Committee at the U.S. Army Center for Health Promotion and Preventative Medicine.
Summary of exposures.
A progression of four experimental exposures was conducted investigating the toxicological effects of RDX and 2,6-DNT in Northern bobwhite. The experiments included a 14-day 2,6-DNT exposure, a 14-day RDX high-dose exposure, a 14-day RDX low-dose exposure, and a 60-day 2,6-DNT exposure. Birds were dosed daily by gavage using corn oil as a carrier for all exposures. In the subacute (14 days) 2,6-DNT exposure, Northern bobwhite were dosed with 0, 50, 100, 190, or 350 mg/kg/day 2,6-DNT. In the 14-day RDX high-dose experiment, doses included 0, 20, 80, 125, or 180 mg/kg/day RDX. In both experiments, each treatment group included seven birds (at least three of each sex per treatment). All RDX doses in the 14-day high-dose experiment elicited seizures within 3 days of experiment initiation (Johnson et al., 2007). RDX induced seizures in Northern bobwhite were irreversible and terminal (Johnson, personal observation); therefore, birds were humanely euthanized upon seizure onset. Brain tissues collected only from freshly-euthanized birds (all dead birds were eliminated from the study) from the 14-day RDX high-dose and 14-day 2,6-DNT exposures were utilized as the source of messenger RNA (mRNA) transcripts for cDNA library construction (see below). Since RDX-induced seizures in all birds in the high-dose RDX exposure, a 14-day RDX low-dose exposure administering RDX at 0, 0.5, 3, 8, 12, and 17 mg/kg/day in 12 replicate male birds at each dose was conducted to investigate sublethal effects of RDX. Males were chosen due to higher relative sensitivity to RDX compared with females (Johnson et al., 2007). Finally, a subchronic (60 days) 2,6-DNT experiment was conducted exposing Northern bobwhite to 2,6-DNT at 0, 5, 10, 40, or 60 mg/kg/day with each treatment including 12 male and 12 female birds. Excised brain tissues were immediately fixed in RNAlater (Ambion, Austin, TX) after euthanasia following manufacturer's recommendations and stored at –80°C. Brain tissues samples were homogenized using a liquid nitrogen–chilled mortar and pestle and utilized for the activities listed below.
Analytical chemistry for tissue.
High-performance liquid chromatography (HPLC) analysis was used to assess the concentrations of RDX and three prominent anaerobic metabolic breakdown products of RDX (hexahydro-1-nitroso-3,5-dinitro-1,3,5-triazine, hexahydro-1,3-dinitroso-5-nitro-1,3,5-triazine, and hexahydro-1,3,5-trinitroso-1,3,5-triazine) in all biological replicates. Detailed description of analytical methods is presented in Johnson et al. (2007), in addition to analytical results for the 14-day RDX high-dose experiment.
RNA extraction.
RNeasy Mini RNA extraction kits (Qiagen Inc., Valencia, CA) were used for RNA extractions. RNA quality assessment methods and criteria used to establish quality-assured RNA samples are presented in the Supplementary Methods. Total RNA was utilized for cDNA library construction, microarray analysis, and reverse-transcription quantitative real-time polymerase chain reaction (RT-qPCR).
cDNA library construction.
Suppression-subtractive hybridization was used (Diatchenko et al., 1996) to develop brain cDNA libraries for Northern bobwhite exposed to 2,6-DNT or RDX. RNA samples from all doses of the 14-day 2,6-DNT exposure were pooled, and all doses from birds killed upon the onset of seizures in the 14-day RDX high-dose exposure were pooled to create individual 2,6-DNT-exposed and RDX-exposed samples, respectively. Each was forward and reverse subtracted against a pooled control sample to create libraries enriched for upregulated and downregulated transcripts (see Supplementary Methods for details). The aggregate cDNA library consisted of 1536 forward RDX, 1056 reverse RDX, 1152 forward 2,6-DNT, and 1824 reverse 2,6-DNT clones, respectively. Portions of each library were sequenced to assess clone redundancy. Redundancy estimates (see Supplementary Methods) for the forward RDX, reverse RDX, forward 2,6-DNT, and reverse 2,6-DNT libraries were 81.5, 14.8, 29.2, and 34.1%, respectively. Due to high redundancy within the forward RDX cDNA library, 2 cDNA representatives of 36 unique contiguous overlapping sequences (contigs) and 15 individual sequences having no significant overlap with other sequences (singletons) for a total of 87 cDNAs were selected to represent the forward RDX library on the cDNA microarray. The revised cDNA library (reduced forward cDNA component) represented a total of 4119 cDNA transcripts derived from Northern bobwhite brain tissue.
Microarray construction.
Microarrays were constructed by spotting the revised cDNA library, ArrayControl Spot cDNAs (Ambion), and printing buffer blanks onto Ultra GAPS-coated microarray slides (Corning Inc., Corning, NY) in duplicate (see Supplementary Methods explaining ArrayControl Spot strategy and printing procedures). Spotted cDNAs were immobilized on the chip surface by applying 300 mJ of ultraviolet radiation in a Stratagene UV Stratalinker 2400 (Agilent Technologies, Waldbronn, Germany). All slides were stored in a darkened desiccator thereafter and pretreated as described in the Supplementary Methods prior to microarray hybridizations.
Microarray experimental designs.
Interwoven loop and balanced interwoven loop designs, which were identified to be the most statistically robust designs available for two-color microarray analysis (Churchill, 2002), were utilized in five independent microarray experiments (Fig. 1). Experiment 1 investigated the 14-day RDX high-dose exposure. In that study, all RDX-dosed quail accumulated 
20 mg/kg RDX in brain tissues (Johnson et al., 2007) and exhibited seizures. Transcript expression was compared among three male and three female controls versus three male and three female RDX-seized quail incorporating two dye swaps per biological replicate totaling 24 microarrays (Fig. 1A). Experiment 2 investigated the 14-day RDX low-dose study (Quinn et al., 2009) where a complete dose-response range was achieved (Fig. 2A). Two response classes (RDX-seized birds [RS] and RDX–non-seizure birds [RN]) within the 12 mg/kg/day RDX treatment were tested against controls including four biological replicates totaling 24 microarrays (Fig. 1B). Experiment 3 investigated the effects of the 14-day 2,6-DNT exposure in females including controls, 60, and 100 mg/kg/day treatments, and experiment 4 investigated effects in males among controls and the 60 mg/kg/day treatment (Figs. 1C and D). Both experiments 3 and 4 included three biological replicates per treatment and utilized 11 and 8 microarrays, respectively. Experiment 5 investigated the 60-day 2,6-DNT exposure comparing controls and the highest affective 2,6-DNT dose (60 mg/kg/day) in males and females with all groups including three biological replicates (Fig. 1E) totaling 24 microarrays. Assignment of biological replicates for all microarray experiments was completely randomized using a random number table.
|
|
Microarray hybridization and analysis.
Two-color microarray hybridizations including ArrayControl RNA Spikes (Ambion; Supplementary Table 1) were conducted as described in the Supplementary Methods. Microarrays were scanned at 5 µm using a VersArray Chipreader (Bio-Rad, Hercules, CA). Grids were manually fit to microarray spots for each image file, and both positive and negative controls were flagged prior to spot analysis. Data for which the spot intensity was less than the background intensity plus 2x the background SD for
1 signal channel were excluded from analysis. Summary data consisted of Lowess normalized Cy3:A647 ratio values calculated using background-normalized median signal intensities.
To broaden signal detection, each microarray was scanned at high and low laser power to resolve low-intensity spots and reduce signal saturation, respectively (Skibbe et al., 2006). Independent statistical analyses were run on high- and low-intensity scan data, and significant targets were summed among analyses. Statistical analyses were conducted using Bayesian analysis of gene expression levels software version 3.62 (Townsend and Hartl, 2002). Nonoverlapping 97.5% confidence intervals (CI) among treatments were established a priori as statistically significant differences in expression. For increased stringency, only targets significant in duplicated technical replicates were considered true significant targets. Due to the semi-redundant nature of the cDNA clone–based microarray (where one gene may be represented by multiple spotted cDNA fragments), the majority of targets (50%) for a given gene were required to be significant and differentially expressed in the same relative direction to be considered truly differentially expressed. For example, cytochrome C oxidase subunit 1 was identified as significant in only a small fraction of the 277 known targets on the microarray and displayed mixed, upregulation and downregulation in the RS. We do acknowledge that within a given gene family, variability among domains may lead to subgroups of the gene population that may be affected by the chemical stressor. However, we do stress that interpretation of such significant microarray results should be made with caution.
Heat map and hierarchical clustering analyses were generated for each microarray experiment using MultiExperiment Viewer (MeV) 4.0 software (TM4 Development Group, Boston, MA). All significant targets for each respective microarray experiment are represented. Treatment effects on gene expression were calculated relative to control expression values unless otherwise noted. Hierarchical clustering was performed in MeV utilizing Euclidean distance and average linkage clustering to arrange each of the treatments and the transcript targets in order of relatedness. Additionally, nonmetric multidimensional scaling analysis was conducted using PRIMER 6 (PRIMER-E Ltd, Ivybridge, UK) to examine gene expression relationships among control, RN, and RS in the 14-day RDX low-dose experiment.
Bioinformatics.
Approximately 2100 total cDNAs represented on the microarray were sequenced as a result of the redundancy test, additional cDNA library sequencing, and, finally, sequencing of all significant targets identified by microarray analyses that were not previously sequenced. Sequences were vector trimmed, quality assessed, and assembled into clusters of contigs and singletons as described in Pirooznia et al. (2007). Identities of cDNAs were established by searching all six potential reading frames of each unique sequence against the National Center for Biotechnology Information (NCBI) nonredundant protein database using blastx (www.ncbi.nlm.nih.gov). An E value of 10–5 was designated as a significant match between the quail transcripts and the best NCBI database match.
Sequence annotation and gene ontology.
Gene ontology (GO) was assembled for the sum of all unique gene identities that were identified to have undergone significant differential expression in the high- and low-dose RDX exposures. For redundant microarray targets, 50% of common targets were required to be significant and differentially expressed in the same relative direction (either upregulated or downregulated) to be included in the GO analysis. Putative functions for unique sequences were assigned employing blast2go (Conesa and Gotz, 2008), GOTM (Zhang et al., 2004), GOfetcher (Pirooznia et al., 2008), and GOstat (Beissbarth and Speed, 2004) to extract the GO hierarchical terms of their homologous genes from the protein databases. GO and functional classification was conducted following Pirooznia et al. (2007).
Real-time PCR.
The accuracy of the RDX microarray results was assessed by RT-qPCR utilizing 22 primer sets to investigate 17 unique transcript targets (Supplementary Table 2). Transcript expression levels were examined using cDNA from DNAse-treated (Qiagen) total RNA (see Supplementary Methods for molecular methods) from the RDX low-dose experiment in brain tissues of control birds, birds experiencing RDX-induced seizures, and RN selected from the 8 and 12 mg/kg/day treatments. Each condition included eight biological replicates. An Applied Biosystems 3900HT sequence analyzer and SDS 2.2 software were utilized to resolve real-time PCR data and the 
CT method used to quantify results. The 18S ribosomal RNA was unaffected by RDX treatment and was therefore considered an appropriate normalizer. One-way ANOVA was used to test for differences among controls, RDX-seized and RDX non-seizure birds within each primer set using Sigma Stat 3.1.1 software (Systat Software Inc., San Jose, CA). We used a weight of evidence (WOE) approach to assess the correspondence among the microarray and real-time PCR results. The WOE was established as a match among results if the majority (50%) of the microarray targets (if represented by multiple cDNAs) matched the results of RT-qPCR.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
RDX elicited seizures in Northern bobwhite in two independent 14-day exposures (Johnson et al., 2007
; Quinn et al., 2009). Conversely, 2,6-DNT caused no observable neurotoxic effects amid gastrointestinal impacts, anemia, lethargy, emaciation, and death in independent 14- and 60-day exposures (Johnson et al., 2007; Quinn et al. 2007). Onset of RDX-induced seizures occurred in a dose-dependent manner (Fig. 2A) in the 14-day RDX low-dose exposure. RS accumulated over 20x more RDX in their brain tissues compared with RN birds within a common dose and 43x more RDX overall (25.0 mg/kg [mean, n = 16] vs. 0.58 mg/kg [mean, n = 14] wet weight tissue concentrations, respectively, Fig. 2B).
Bioinformatics for Microarray and Microarray Performance
Of the approximately 2100 total sequences assessed, 1717 were quality assured, of which 947 had significant blastx matches. All quality-assured sequences were submitted to GenBank dbEST (accession numbers FL684479–FL685629 and GE469687–GE470231). The microarrays performed well in all assays, where 72–82% of the transcript targets contributed to the statistical analyses (Table 1). Specific microarray performance information including false-positive and false-negative rates is presented in the Supplementary Results and Supplementary Table 3.
|
Microarray Results Summary
RDX elicited a greater number of differentially expressed transcripts in brain tissue than did 2,6-DNT (Table 1). In the 14-day RDX high-dose experiment, high- and low-intensity scans yielded 118 and 27 targets, respectively, that were significant in duplicate technical replicates. Of these, 26 targets were common among scans for a total of 119 unique significant targets (2.9% of the total 4119 quail transcripts). In the 14-day RDX low-dose exposure, high- and low-intensity scan data yielded 237 and 191 targets, respectively, that were significant in duplicate. Of these, 143 targets were common for a total of 285 significant targets (6.9% of the total quail cDNA targets). Of the significant targets identified among the 14-day RDX high-dose and 14-day RDX low-dose experiments, 27 cDNA targets (22.7% of 119) were common. A total of 43 and 61 cDNAs representing unique "gene" identities were observed within each list of significant cDNA targets for the RDX high-dose and RDX low-dose experiments, respectively. Of these unique gene identities, 17 (39.4% of 43) were common among high- and low-dose gene lists (Table 2). In the individual male and female 14-day 2,6-DNT microarray analyses, summation of high- and low-intensity microarray scans yielded a total of only 5 and 17 differentially expressed targets, respectively. The 60-day 2,6-DNT exposure resolved only nine significant targets. Only two significant targets were common between two of the three microarray analyses in the 2,6-DNT exposures (Supplementary Table 4).
|
Effect of RDX on Gene Expression in Northern Bobwhite Brain
The expression patterns of RDX-treated birds indicated that there was some clustering based on sex (Supplementary Figure 1); however, the effects of RDX exposures on transcription among males and females appeared conserved between sexes overall. In the 14-day RDX low-dose experiment, expression patterns clustered discretely by experimental condition (Supplementary Figure 2), demonstrating distinct differences between RS and RN within the 12 mg/kg/day dose. Results of multidimensional scaling analysis of expression values for all 285 significantly affected transcript targets, and all biological replicates indicated that RS clustered separately from RN and controls (Fig. 3); however, one RS appeared to be an outlier. Control and RDX–non-seizing birds clustered together suggesting little difference in overall transcript expression among unexposed-control birds and RN.
|
GO of Transcripts Affected by RDX
We examined the GO terms associated with differentially expressed genes in order to assess possible functional impacts of RDX. Many differentially expressed genes had GO functions associated with cell electrophysiology and cell signaling in the CNS function (Fig. 4). Investigation of the functional terms encompassing the greatest number of unique gene identities across hierarchical levels 2–6 indicates a prevalence of metal ion binding, ion transport, and specifically, ion transport across membranes. The differentially expressed genes are involved in a variety of functional pathways including calcium signaling pathway, cell communication, phosphatidylinositol signaling, and Parkinson’s disease (Supplementary Table 5). GOs for the complete sequence set represented on the microarray are available at http://mcbc.usm.edu/pirooznia/Colinus-virginianus/ontology/.
|
RDX Impact on Genes Involved in CNS Structure/Function
In the 14-day RDX high-dose and 14-day RDX low-dose experiments, 41 and 34 significant cDNA clones representing 15 and 14 unique genes, respectively, were identified to be related to CNS structure and/or function (Supplementary Tables 6 and 7). Of these unique genes, eight were common among high- and low-dose experiments (Table 2). The four genes similar to aspartate aminotransferase, glial fibrillary acidic protein
, Na+, K+-adenosine triphosphatase (ATPase) β-1 subunit, and synaptosomal-associated protein 25 (SNAP-25) were upregulated in both microarray experiments, visinin-like 1 (VSNL1) was downregulated in both experiments, while the three genes calmodulin 2 (CALM2), chimaerin 1, and ornithine decarboxylase antizyme 1 (OAZ1) had a mixed response among experiments that was sex dependent.
WOE for RDX Effects
The WOE approach indicated a reasonable agreement (70–85%) between methods where 17 of 20 and 14 of 20 RT-qPCR results confirmed microarray result for RN and RS, respectively, in the 14-day RDX low-dose exposure (Table 3). Confirmed expressed targets included: downregulation of hemoglobin alpha 1 and upregulation of ferritin, OAZ1, and gene trap locus 3. The results of the SYBR Green-based RT-qPCR analysis were observed to be less sensitive than the microarray analysis. Comparison of 97.5% CI around means for the control, RDX-seized, and RDX–non-seizure conditions for 14 RT-qPCR primer sets versus 97.5% CI around means for corresponding microarray targets indicated that the CIs for RT-qPCR were larger than CIs for the microarray counterparts in 27 of 42 tests. Therefore, real-time PCR results were considered a parallel line of evidence to microarray results instead of a validation test.
|
Effect of 2,6-DNT on Gene Expression in Northern Bobwhite Brain
Few significant transcripts (5–17) were identified (see Supplementary Results) in brains of Northern bobwhite exposed for 14 or 60 days (Table 1, Supplementary Table 4) indicating that, unlike RDX (119 and 285 differentially expressed transcripts), the brain is not a major target organ for toxicity in 2,6-DNT exposures. No significant clustering was observed when differentially expressed genes were examined across all replicates in 14-day 2,6-DNT and 60-day 2,6-DNT exposures (Supplementary Figure 3), suggesting that neither dose nor sex had significant impact on differential expression.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
RDX caused terminal seizures in Northern bobwhite beginning at the 8 mg/kg/day dose (Fig. 2A, Quinn et al., 2009
). Birds exhibiting RDX-induced seizures bioaccumulated over 20x more RDX in brain tissue than non-seizing counterparts within a common dose level (Fig. 2B). Paralleling the observed differential bioaccumulation of RDX and overall transcript expression patterns among RS versus RN were markedly different appearing to be closely related to RDX tissue concentrations (Fig. 3, Supplementary Figures 1 and 2). Conversely, 2,6-DNT exposures caused no overt signs of neurotoxicity in Northern bobwhite, even at doses causing altered blood chemistry, gastrointestinal impacts, edema in viscera, and even mortality at high doses (Johnson et al., 2007; Quinn et al. 2007). Though 2,6-DNT elicited these pronounced effects, there was marginal indication of these stresses in the brain at the transcriptional level of organization (Table 1, Supplementary Table 4). We discuss the implications of each of these results below.
RDX Bioaccumulation and Seizures
Tissue-RDX levels were directly correlated with seizures and impacts on gene expression. High concentrations of RDX accumulated in brain tissues (Fig. 2B, Johnson et al., 2007) and liver tissues (Johnson et al., 2007) of RDX-seized Northern bobwhite. Non-seizing birds accumulated a fraction of the RDX accumulated in brain tissue of seizing birds (Fig. 2B). Investigations of RDX and RDX metabolite concentrations in liver tissues from the RDX low-dose experiment indicated that non-seizing birds accumulated equivalently low RDX concentrations in the liver compared with brain and that RDX metabolism in non-seizing birds was extensive (Fig. 5). These results indicate that the blood-brain barrier (BBB) did not effectively exclude RDX from brain tissues, and therefore, concentrations of RDX in the brain were parallel to systemic RDX concentrations. Failure of RDX metabolism and excretion mechanisms (likely in the liver and/or kidneys) may have facilitated the accumulation of the critical brain tissue residue of RDX that causes terminal seizures.
|
Neurotoxicogenomics of RDX
Analysis of two independent subacute (14 days) RDX exposures indicated that affective doses of RDX induced the differential expression of hundreds of transcript targets in Northern bobwhite brain tissue (Table 1). A comparison of significantly differentially expressed gene lists for the two RDX experiments (Supplementary Tables 6 and 7) indicated that, although there were several targets in common (Table 2), the majority were unique within each experiment. RDX dose levels in the high and low exposures may have contributed the most to the difference in differentially expressed transcripts among the experiments as the high-dose RDX experiment elicited seizures in all birds within 3 days (Johnson et al., 2007
), whereas many birds resisted seizures beyond day 10 in the RDX low-dose experiment (Quinn et al., 2009). A number of other factors including use of a non-inbred model species, differing experimental designs, and technical variation among microarray experiments may have also contributed to observed differences among experiments (Churchill, 2002). Regardless, the sum of unique targets identified across experiments provides insight into the broad genomic response of Northern bobwhite to RDX, while targets identified in common among experiments by multiple lines of evidence (i.e., multiple microarray experiments and RT-qPCR) provide greater confidence for establishing conserved toxicological responses and a robust assessment for potential mechanisms of action. We assessed the potential roles of the differentially expressed transcripts to reconstruct the toxicological and molecular events by which RDX may have caused terminal seizures. We focused primarily on targets having the most lines of evidence indicative of differential expression (Tables 2 and 3). The acute toxicology of RDX in Northern bobwhite was elicitation of seizures followed by death. Either as the result of direct effects of RDX on molecular targets and/or a secondary response, a number of genes related to ion binding and ion transport across membranes, which is critical to the electrophysiology and signal transduction in neurons, were significantly affected in brains of quail experiencing RDX-induced seizures.
Insights into RDX Mode of Action
Although RDX accumulated to relatively high levels in brain tissues of RS, histopathology revealed no abnormalities or physical damage in Northern bobwhite brains (Johnson et al., 2007; Quinn et al., 2009). Comparative toxicology indicated that the seizure response to RDX bears a marked resemblance to the mammalian seizure response to acute levels of organochlorine insecticides, specifically dichlorodiphenyltrichloroethane (DDT). DDT elicited tremors followed by tonic and clonic convulsions leaving no pathologic trace in CNS tissues (Ecobichon, 2001). Acute doses of RDX caused equivalent effects in both Northern Bobwhite (Johnson et al., 2007; Quinn et al., 2009) and rats (Bannon et al., 2009; Talmage et al. 1999). The mode of action (MOA) underlying DDT neurotoxicity involved a delayed repolarization of neuron postaction potential yielding hyperexcitability (Ecobichon, 2001). Seizures were observed to be sparked by magnified sensitivity to touch and/or sound, which was also observed in Northern bobwhite exposed to RDX (Johnson, personal communication). At least four simultaneously acting mechanisms were observed to underlie the DDT seizure response (Ecobichon, 2001). These mechanisms included reduced K+ transport across the neuronal membrane, slowed closing of Na+ channels postaction potential, inhibition of neuronal ATPases, particularly Na+, K+-ATPase and Ca2+-ATPase, and inhibition of calmodulin control of Ca2+ at axonal nodes affecting neurotransmitter release. All these mechanisms contributed to inhibited repolarization of neurons increasing overall neuronal excitability resulting in seizures. Our observations indicate that the MOA for seizures caused by RDX is likely similar to that of DDT, and below, we investigated if the mechanisms underlying this MOA might also be similar.
Mechanisms Underlying RDX-Induced Seizures
Genomic results indicated that transcripts related to three of the four mechanisms described above were affected in RS but not in RN. Transcripts encoding Na+, K+-ATPase and CALM2 were each differentially expressed in RDX-seized animals (Table 2) in addition to a transcript encoding OAZ1, which regulates Kir (K+) channels in neurons (Kilpeläinen, 2002). Na+, K+-ATPase is necessary to drive active transport of Na+ and K+ against concentration gradients to reestablish repolarization of neurons (Catterall, 1988). Na+, K+-ATPase targets were upregulated possibly in compensation for impaired neuronal repolarization. Calmodulin and Ca2+ antagonists inhibit seizures induced by convulsant agents (Solà et al., 1999), demonstrating the hyperexcitability that neurons experience when Ca2+ pools become unconstrained. Consistent with observations of seizures and possible hyperexcitability, calmodulin expression was downregulated. Ornithine decarboxylase and its antizyme (OAZ) expressed in rat neurons are involved in gating and rectification of inward rectifying Kir (K+) channels (Kilpeläinen, 2002), which can influence membrane potential and hence neuronal excitability. OAZ1 was upregulated in Northern bobwhite that may have been involved in K+ channel regulation to lessen neuronal excitability.
In addition to the mechanisms listed above, genes involved in multiple neurophysiological pathways were affected consistent with a potential broad-scale physiological response to counteract the hyperexcitation caused by RDX (Fig. 6). Cholinergic neurons are stimulated by the neurotransmitter acetylcholine eliciting the fight or flight response of the sympathetic nervous system. VSNL1 modulates 4β2 acetylcholine receptors increasing membrane-surface expression levels and agonist sensitivity in response to intracellular Ca2+ levels (Lin et al., 2002). VSNL1 was downregulated, possibly in response to unbound Ca2+ levels, to counteract hyperstimulation of cholinergic neurons. Aspartate and glutamate are also excitatory neurotransmitters in the CNS (Clements et al., 1986). Astrocytes can release the cognate ketoacid (-ketoisocaproate [KIC]) to neurons, which have aspartate aminotransferase gene (AspAT) activity that reaminates the KIC to leucine (Yudkoff et al., 2005). This process consumes glutamate functionally buffering glutamate concentrations to mitigate hyperexcitation. Upregulation of cAspAT may indicate that additional cAspAT was required to maintain glutamate-buffering capacity. Both SNAP-25 and heat-shock protein (HSP) (HSP70/72) mRNA expression in rat brains have been noted to increase after seizures (Martí et al., 1999), which was also observed in our study. SNAP-25 is an essential component for neurotransmitter release (Martí et al., 1999), and HSP90 was observed to associate with voltage-dependent ClC-2 chloride channels facilitating channel opening (Hinzpeter et al., 2006), each of which may inhibit neuronal repolarization. The sum of these transcriptomic events highlights the integrated response of CNS tissue to RDX-induced seizures (Fig. 6).
|
Neurotoxicogenomics of 2,6-DNT
Both the subacute and the subchronic exposures to daily oral doses of 2,6-DNT in Northern bobwhite caused minimal differential expression of transcripts in brain tissues (Supplementary Table 4). The BBB has been observed to effectively exclude 2,6-DNT in mice (Faust, 1998
), and likewise, concentrations were below detection limits in Northern bobwhite brain tissues (Johnson et al., 2007). Correspondingly, few transcripts related to nervous system function were differentially expressed (Supplementary Table 4). The robust representation of clones derived from 2,6-DNT exposures on the microarray indicates that the low number significantly differentially expressed transcripts in 2,6-DNT exposures is not an artifact of minimal representative transcript coverage (see Supplementary Discussion regarding cDNA library). The exclusion of 2,6-DNT from brain tissue, the lack of observable neurotoxic effects, and minimal differential expression of transcripts in brain tissue indicate that 2,6-DNT is an unlikely neurotoxicant. Regarding the utility of CNS transcriptomics for identifying markers of systemic stress, hemoglobin -subunit transcription was upregulated (Supplementary Table 4) likely in response to anemia and reduced hemoglobin concentrations in blood caused by 2,6-DNT exposure (Quinn et al., 2007). However, due to minimal differential expression of transcripts, even at the threshold of lethality, CNS transcriptomics did not provide a robust indicator of systemic effects.
Future Investigations
In this study, we utilized genomic tools to develop a hypothetical MOA by which RDX elicited seizures. Functional testing of the hypothetical MOA using electrophysiological and neurobiochemical investigations is required to validate or disprove these hypotheses. Validated mechanisms can be incorporated into risk assessment protocols for Army ranges. Additionally, investigation of RDX pharmacokinetics is necessary to explain the cause of impaired RDX metabolism and excretion that lead to extensive RDX accumulation and resultant seizures in susceptible individuals. Liver and kidney transcriptomics may reveal how RDX may impair these functions.
|
SUPPLEMENTARY DATA |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
The supplementary description of Materials and Methods, Results, Discussion, and supplementary figures can be viewed at http://toxsci.oxfordjournals.org/.
|
FUNDING |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
U.S. Army Environmental Quality and Installations basic research program. Permission was granted by the Chief of the U.S. Army Corps of Engineers to publish this information.
|
ACKNOWLEDGMENTS |
|---|
We thank Dr Desmond Bannon, Dr Lawrence Williams, Dr Valerie Adams, and all anonymous peer reviewers for their insights that enhanced this article.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Ankley GT, Daston GP, Degitz SJ, Denslow ND, Hoke RA, Kennedy SW, Miracle AL, Perkins EJ, Snape J, Tillitt DE, et al. Toxicogenomics in regulatory ecotoxicology. Environ. Sci. Technol. (2006) 40:4055–4065.[Medline]
Baker LM, Larsen CT, Sriranganathan N, Jones DE, Johnson MS, Gogal RM Jr. Effects of energetic compounds on the northern bobwhite quail and biotransformation applications of the intestinal flora. Bull. Environ. Contam. Toxicol. (2004) 72:1–6.[CrossRef][Web of Science][Medline]
Bannon DI, Dillman JF, Hable MA, Phillips CS, Perkins EJ. Global gene expression in rat brain and liver after oral exposure to the explosive hexahydro-1,3,5-triazine (RDX). Chem. Res. Toxicol. (2009) 22(4):620–625.[CrossRef][Web of Science][Medline]
Beissbarth T, Speed TP. GOstat: Find statistically overrepresented Gene Ontologies within a group of genes. Bioinformatics (2004) 20(9):1464–1465.
Brennan LA. Northern Bobwhite (Colinus virginianus). In: The Birds of North America, No. 397—Poole A, Gill F, eds. (1999) Philadelphia, PA: The Birds of North America, Inc. 1028–1029.
Catterall WA. Structure and function of voltage-sensitive ion channels. Science (1988) 242:50–61.
Churchill GA. Fundamentals of experimental design for cDNA microarrays. Nat. Genet. (2002) 32:490–495.[CrossRef][Web of Science][Medline]
Clements JR, Madl JE, Johnson RL, Larson AA, Beitz AJ. Localization of glutamate, glutaminase, aspartate and aspartate aminotransferase in the rat midbrain periaqueductal gray. Exp. Brain Res. (1986) 67:594–602.[Web of Science]
Conesa A, Gotz S. Blast2GO: A comprehensive suite for functional analysis in plant genomics. Int. J. Plant Genomics (2008) 2008:619832.[Medline]
Diatchenko L, Lau YF, Campbell AP, Chenchik A, Moqadam F, Huang B, Lukyanov S, Lukyanov K, Gurskaya N, Sverdlov ED, et al. Suppression subtractive hybridization: A method for generating differentially regulated or tissue-specific cDNA probes and libraries. Proc. Natl. Acad. Sci. U S A (1996) 93:6025–6030.
Ecobichon DJ. Toxic effects of pesticides. In: Casarett and Doull's Toxicology: The Basic Science of Poisons—Klaassen CD, ed. (2001) 6th ed. New York: McGraw-Hill Medical Publishing Division. 763–810.
Faust RA. (1998) Toxicity Summary for 2,6-Dinitrotoluene. Risk Assessment Information System, Oak Ridge National Laboratories, Oak Ridge, TN, pp. 1–11. Available at http://rais.ornl.gov/tox/profiles/2_6_dinitrotoluene_f_V1.shtml. Accessed August, 2009.
Gogal RM Jr, Larsen CT, Prater MR, Duncan RB, Ward DL, Johnson MS, Holladay SD. Influence of dietary exposure to TNT derivatives on the immune system of the northern bobwhite quail (Colinus virginianus). Environ. Toxicol. Chem. (2002) 21:81–86.[CrossRef][Web of Science][Medline]
Goldberg DJ, Green ST, Nathwani D, McMenamin J, Hamlet N, Kennedy DH. RDX intoxication causing seizures and widespread petechial rash mimicking meningococcaemia. J. R. Soc. Med. (1992) 85:181.[Web of Science][Medline]
Hinzpeter A, Lipecka J, Brouillard F, Baudoin-Legros M, Dadlez M, Edelman A, Fritsch J. Association between Hsp90 and the ClC-2 chloride channel upregulates channel function. Am. J. Physiol. Cell Physiol. (2006) 290:C45–C56.
Jenkins TF, Pennington JC, Ranney TA, Berry TE Jr, Miyares PH, Walsh ME, Hewitt AD, Perron NM, Parker LV, Hayes CA, et al. Characterization of Explosives Contamination at Military Firing Ranges. ERDC TR-01–5. Final/Technical Report (2001) Hanover, NH: Engineer Research and Development Center, U.S. Army Corps of Engineers.
Johnson MS, Michie MW, Bazar MA, Gogal RM Jr. Influence of oral 2,4-dinitrotoluene exposure to the Northern Bobwhite (Colinus virginianus). Int. J. Toxicol. (2005) 24:265–274.
Johnson MS, Quinn MJ Jr, Bazar MA, Gust KA, Escalon BL, Perkins EJ. Subacute toxicity of oral 2,6-dinitrotoluene and 1,3,5-trinitro-1,3,5-triazine (RDX) exposure to the Northern Bobwhite (Colinus virginianus). Environ. Toxicol. Chem. (2007) 26:1481–1487.[CrossRef][Web of Science][Medline]
Kilpeläinen P. Ornithine decarboxylase expression and regulation in rat brain and in transgenic mice (2002) Dissertation, Department of Biochemistry, University of Oulu, Oulu, Finland. ISBN 951-42-6631-5.
Levine BS, Rust J, Burns JM, Lish PM. Determination of the Chronic Mammalian Toxicological Effects of RDX: (Twenty-Four Month Chronic Toxicity/Carcinogenicity Study of Hexahydro-1,3,5- Trinitro-1,3,5-Triazine (RDX) in the Fischer 344 Rat (1983) ADA 160774. IIT Research Institute, Chicago, IL; US Army Medical Research and Development Command: Fort Detrick, Frederick, MD.
Lin L, Jeanclos EM, Treuil M, Braunewell KH, Gundelfinger ED, Anand R. The calcium sensor protein visinin-like protein-1 modulates the surface expression and agonist sensitivity of the 4β2 nicotinic acetylcholine receptor. J. Biol. Chem. (2002) 277:41872–41878.
Martí E, Blasi J, Gomez de Aranda I, Ribera R, Blanco R, Ferrer I. Selective early induction of synaptosomal-associated protein (molecular weight 25,000) following systemic administration of kainate at convulsant doses in the rat. Neuroscience (1999) 90:1421–1432.[CrossRef][Web of Science][Medline]
Perkins EJ, Gust KA, Steevens J. Does toxicogenomics have any practical use in ecotoxicology? Assessing population level impacts of contaminants. Integrated Environ. Assess. Manag. (2007) 3(4):559–566.[CrossRef]
Pirooznia M, Gong P, Guan X, Inouye LS, Yang K, Perkins EJ, Deng Y. Cloning, analysis and functional annotation of expressed sequence tags from the Earthworm Eisenia fetida. BMC Bioinformatics (2007) 8(Suppl. 7):S7.
Pirooznia M, Habib T, Perkins EJ, Deng Y. GOfetcher: A database with complex searching facility for gene ontology. Bioinformatics (2008) 24(21):2561–2563.
Poynton HC, Varshavsky JR, Chang B, Cavigiolio G, Chan S, Holman PS, Loguinov AV, Bauer DJ, Komachi K, Theil EC, et al. Daphnia magna ecotoxicogenomics provides mechanistic insights into metal toxicity. Environ. Sci. Technol. (2007) 41:1044–1050.[Medline]
Quinn MJ Jr, Bazar MA, McFarland CA, Perkins EJ, Gust KA, Gogal RM, Johnson MS. Effects of subchronic exposure to 2,6-dinitrotoluene in the Northern Bobwhite (Colinus virginianus). Environ. Toxicol. Chem. (2007) 26:2202–2207.[CrossRef][Web of Science][Medline]
Quinn MJ Jr, Bazar MA, McFarland CA, Perkins EJ, Gust KA, Johnson MS. Sublethal effects of subacute exposure to RDX (1,3,5-trinitro-1,3,5-triazine) in the Northern Bobwhite, Colinus virginianus. Environ. Toxicol. Chem. (2009) Jan 27:1. [Epub ahead of print].
Sauer JR, Link WA, Nichols JD, Royle JA. Using the North American breeding bird survey as a tool for conservation: A critique of Bart et al (2004). J. Wildl. Manage. (2004) 69:1321–1326.[CrossRef]
Skibbe DS, Wang X, Zhao X, Borsuk LA, Nettleton D, Schnable PS. Scanning microarrays at multiple intensities enhances discovery of differentially expressed genes. Bioinformatics (2006) 22:1863–1870.
Solà C, Barrón S, Tusell JM, Serratosa J. The Ca2+/calmodulin signaling system in the neural response to excitability. Involvement of neuronal and glial cells. Prog. Neurobiol. (1999) 58:207–232.[CrossRef][Web of Science][Medline]
Talmage SC, Opresko DM, Maxwell CJ, Welsh CJE, Cretella FM, Reno PH, Daniel FB. Nitroaromatic munition compounds: Environmental effects and screening values. Rev. Environ. Contam. Toxicol. (1999) 161:1–156.[Medline]
Townsend JP, Hartl DL. Bayesian analysis of gene expression levels: Statistical quantification of relative mRNA level across multiple strains or treatments. Genome Biol. (2002) 3. research0071.1–0071.16.
Yudkoff M, Daikhin Y, Nissim I, Horyn O, Luhovyy B, Lazarow A, Nissim I. Brain amino acid requirements and toxicity: The example of leucine. J. Nutr. (2005) 135:1531S–1538S.
Zhang B, Schmoyer D, Kirov S, Snoddy J. GOTree Machine (GOTM): A web-based platform for interpreting sets of interesting genes using Gene Ontology hierarchies. BMC Bioinformatics (2004) 5:16.[CrossRef][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||