TOXICOLOGICAL SCIENCES, 141(1), 2014, 44–58
doi: 10.1093/toxsci/kfu104
Advance Access Publication Date: June 3, 2014
Validation of a Genomics-Based Hypothetical Adverse
Outcome Pathway: 2,4-Dinitrotoluene Perturbs PPAR
Signaling Thus Impairing Energy Metabolism and
Exercise Endurance
Mitchell S. Wilbanks*,1 , Kurt A. Gust*, Sahar Atwa† , Imran Sunesara‡ ,
David Johnson*,§ , Choo Yaw Ang¶ , Sharon A. Meyer† , and Edward J. Perkins*
*Army Engineer Research and Development Center, Vicksburg, Mississippi 39180, † University of Louisiana at
Monroe, Monroe, Louisiana 71201, ‡ University of Mississippi Medical Center, Jackson, Mississippi 39216,
§
Conestoga-Rovers & Associates, Dallas, Texas 75234 and ¶ Badger Technical Services, San Antonio, Texas
71286
1
To whom correspondence should be addressed at United States Army Engineer Research and Development Center, 3909 Halls Ferry Road, Vicksburg,
MS. Fax: +601 634-4002. E-mail: [email protected].
ABSTRACT
2,4-dinitrotoluene (2,4-DNT) is a nitroaromatic used in industrial dyes and explosives manufacturing processes that is
found as a contaminant in the environment. Previous studies have implicated antagonism of PPAR␣ signaling as a principal
process affected by 2,4-DNT. Here, we test the hypothesis that 2,4-DNT-induced perturbations in PPAR␣ signaling and
resultant downstream deficits in energy metabolism, especially from lipids, cause organism-level impacts on exercise
endurance. PPAR nuclear activation bioassays demonstrated inhibition of PPAR␣ signaling by 2,4-DNT whereas PPAR␥
signaling increased. PPAR␣ (-/-) and wild-type (WT) female mice were exposed for 14 days to vehicle or 2,4-DNT (134
mg/kg/day) and performed a forced swim to exhaustion 1 day after the last dose. 2,4-DNT significantly decreased body
weights and swim times in WTs, but effects were significantly mitigated in PPAR␣ (-/-) mice. 2,4-DNT decreased transcript
expression for genes downstream in the PPAR␣ signaling pathway, principally genes involved in fatty acid transport. Results
indicate that PPAR␥ signaling increased resulting in enhanced cycling of lipid and carbohydrate substrates into
glycolytic/gluconeogenic pathways favoring energy production versus storage in 2,4-DNT-exposed WT and PPAR␣ (-/-) mice.
PPAR␣ (-/-) mice appear to have compensated for the loss of PPAR␣ by shifting energy metabolism to PPAR␣-independent
pathways resulting in lower sensitivity to 2,4-DNT when compared with WT mice. Our results validate 2,4-DNT-induced
perturbation of PPAR␣ signaling as the molecular initiating event for impaired energy metabolism, weight loss, and
decreased exercise performance.
Key words: dinitrotoluene; energy; adverse outcome pathway; PPAR
Nitroaromatic compounds are commonly used in manufacture
of both industrial compounds and explosives used by the military and commercially. 2,4-dinitrotoluene (2,4-DNT) is used as
an intermediate in production of polyurethane polymers, dyes,
and explosives (ATSDR, 1998). As a result of its extensive use in
ordnance and propellant production, 2,4-DNT has been found to
contaminate soil at artillery ranges and demilitarization areas
and surrounding waste water streams at munitions manufacturing sites leading to its listing on EPA’s Contaminate Candidate
List (Pontius, 1997).
Trinitrotoluene and/or dinitrotoluene (DNT) isomers have
been observed to decrease organism weights in birds (Quinn
Published by Oxford University Press on behalf of Toxicological Sciences 2014.
This work is written by (a) US Government employee(s) and is in the public domain in the US.
44
WILBANKS ET AL.
et al., 2007), lizards (McFarland et al., 2012), salamanders (Johnson et al., 2007), rats (Lee et al., 1978; Lent et al., 2012), and mice
(Ellis et al., 1979; Hong et al., 1985) along with increasing signs
of lethargy in birds (Johnson et al., 2005), lizards (McFarland
et al., 2012), and rats (Dilley et al., 1982; Lent et al., 2012; Levine
et al., 1984). These observations have been hypothesized to arise
via impacts on various components of energy metabolism as
observed via transcript expression assays (Deng et al., 2011;
Rawat et al., 2010). However, the direct linkage between these
molecular-level impacts and weight loss and other adverse effects at the individual level has not been tested.
Oxidation of the hemoglobin heme iron by the nitro group
of nitrotoluenes leads to formation of methemoglobin, decreased oxygen delivery (Kozuka et al., 1979), and ultimately resulting in cyanosis (McGee et al., 1942) which can impact energy metabolism via reduced oxidative phosphorylation. Recent transcriptomic data suggest that 2,4-DNT impacts energy metabolism via disregulation of glycolysis/gluconeogenesis
(Rawat et al., 2010) and peroxisome proliferator-activated receptor ␣ (PPAR␣) dependent lipid metabolism (Deng et al., 2011;
Wintz et al., 2006; Xu and Jing, 2012).
PPAR␣ is a ligand-activated transcriptional regulator of genes
involved in cellular fatty acid uptake, fatty acid activation, intracellular fatty acid transport, triglyceride storage, lipolysis, ketogenesis, and gluconeogenesis (Desvergne and Wahli, 1999; Mandard et al., 2004; Rakhshandehroo et al., 2010). Endogenous ligands such as long chain fatty acyl-CoAs and saturated fatty acids
are known to bind to and activate PPAR␣ leading to subsequent
decreased intracellular fatty acid concentrations and very lowdensity lipoprotein production in liver and plasma triglyceride
levels (Lefebvre et al., 2006). Antagonism of PPAR␣ may lead to
fatty liver disease as a result of perturbing beta-oxidation and
fatty acid cellular uptake. Although studies have shown that
plasticizers can activate PPAR␣ (Feige et al., 2010; Ito et al., 2012),
2,4-DNT has been shown to decrease transcriptional expression
of the PPAR␣ gene and alter lipid metabolism in fathead minnow
and rats (Deng et al., 2011; Wintz et al., 2006).
To understand whether these antagonistic effects on PPAR␣
expression and resulting energy metabolism translate to impaired individual performance, we used the Adverse Outcome
Pathway (AOP) framework where a molecular initiating event
(MIE) is linked to adverse effects relevant to risk assessment
(Ankley et al., 2010). We tested the validity of a hypothetical
AOP predicting that 2,4-DNT dosing would result in interference with PPAR␣ signaling, lipid metabolism, and ultimately
organism-level impacts on energy metabolism. PPAR␣ (-/-) and
wild-type (WT) mice were exposed to a sublethal dose at 1/10
of the published LD50 (ATSDR, 1998) of 2,4-DNT (134 mg/kg/day
for 14 days) or vehicle then given an exercise challenge (a forced
swim until exhaustion) 1 day after the last dose to determine if
2,4-DNT acted via PPAR␣ to cause decreased performance. We
compared the effects of 2,4-DNT in PPAR␣ (-/-) and WT mice
on metabolic pathways related to energy metabolism including
glucose metabolism, fatty acid metabolism and PPAR activation,
and response in liver tissue. We also examined corresponding
metabolic outputs of these pathways including glycogen content
in liver tissue and striated muscle in addition to serum glucose
and triglycerides. The overall design enabled us to examine if
2,4-DNT interacted directly with PPAR␣ to cause physiological
impacts on performance as postulated in the AOP.
45
MATERIALS AND METHODS
Chemicals. 2,4-DNT (97%, CAS number 121–14–2), dimethyl sulfoxide (DMSO), and corn oil were obtained from Sigma-Aldrich
(St. Louis, MO).
Animals and treatment. C57Bl/6N (WT) and PPAR␣ (-/-) (catalog
number TF3135) age-matched female mice (20 g) were purchased
from Taconic Farms (Hudson, NY) and housed and treated in accordance with the Guide for Use and Care of Animals (National
Research Council, 2011). Mice were housed in groups of 4 in
polycarbonate shoebox cages 1 week prior to study. Light cycle
was reversed with 7 a.m. to 7 p.m. as the dark phase such that
swim trials during the working day occurred during the mice’s
active time of day. Rodent chow (Harlan/Teklad 7012, Madison,
WI) and tap water were available ad libitum whereas oral gavage
administration occurred between 8 and 10 a.m. daily. Study protocols were pre-approved by the Institutional Animal Care and
Use Committees of the University of Louisiana at Monroe and
US Army Engineer Research and Development Center.
Determination of sublethal 2,4-DNT dose. A range-finding study
was conducted to determine a 2,4-DNT dose without overt systemic toxicity to mice. Groups of mice were weighed and randomly assigned to four treatment groups, each containing four
mice, including vehicle (5% DMSO in corn oil, vol/vol) or 2,4DNT (134, 402, 670 mg/kg/day in vehicle). The high dose of 670
mg/kg/day was selected based on 50% of the acute LD50 dose for
female Swiss and CF-1 mice for 2,4-DNT (ATSDR, 1998). All animals underwent daily acclimation swims of 15 min for 5 days
prior to 2,4-DNT dosing via gavage. During days 1–14 of 2,4-DNT
dosing, mice swam in tanks for 30 min each day, initiated 2 h after dosing. On day 15, animals were challenged with an exhaustive swim test. Mice were observed continuously for the first
hour after dosing, during the swim trial, and hourly thereafter
for 8 h. Moribund mice were euthanized with CO2 . All C57Bl/6
mice in the high dose treatment group died within 4 days of
daily dosing, indicating greater sensitivity of this strain to 2,4DNT than the reference Swiss or CF-1 mice. Further, it should
be noted that lethality of 2,4-DNT dosing in this study may have
also been increased due to enhanced bioavailability from the use
of 5% DMSO within the dosing medium unlike previous studies
which did not use DMSO (Ellis et al., 1978; Lee et al., 1975; Rickert and Long 1981; Vemot et al., 1977). Two animals in the 402
mg/kg/day treatment group lost ⬎10% of body weight by day 8
of dosing. Due to signs of toxicity, one of the two animals which
lost weight was euthanized. The ability to perform the short acclimation swim was compromised for all mice treated with 402
mg/kg/d on days 5–7. Due to animal death and toxicity in the
402 mg/kg/day treatment group and/or inability of treated mice
to perform short acclimation swims, the 402 mg/kg/d dose was
reduced to 268 mg/kg/day in remaining mice at day 8 and continued through the remainder of the bioassay. No overt toxicity
was observed in the 134 mg/kg/d treatment group.
Determining/optimizing approach for swim test. Based on the information gathered from the range-finding study, a pilot study
was carried out to determine optimal parameters for a definitive
swim endurance study where C57Bl/6N mice were acclimated
for 5 days, gavaged daily for 14 days with either 134 mg/kg 2,4DNT (n = 4), 268 mg/kg 2,4-DNT (n = 6), or vehicle control (n
= 4). Two extra mice were included in the 268 mg/kg group to
compensate for anticipated loss predicted by the range-finding
study. In order to shorten the impractically long swim times
46
TOXICOLOGICAL SCIENCES, 2014, Vol. 141, No. 1
(⬎2 h) observed in initial range-finding studies, a 5% of body
weight load was attached to the tail of each mouse during the
exhaustive swim (Jin and Wei, 2011; Takeda et al., 2011). Tail
weights were constructed by inserting the tail through flat washers (3/16” outside diameter, 1/16” inside diameter, weight 0.33
g) and affixing these at 1/3 the distance from the end of the tail
by threading the tail through a 1/8” length segment of silicone
tubing (Masterflex 96440–14) that acted as a distal bumper. Exhaustion was judged to be the time when animals sank and remained under the surface for 5 s (Takeda et al., 2011). After 5 s,
animals were retrieved from the tank and observed for clinical
symptoms before carbon dioxide asphyxia and necropsy.
In order to assess whether the acclimation swim protocol
was effective, stress levels were evaluated by measuring corticosterone. Levels were compared against unexposed WT mice
that had not swam (n = 3). Blood was collected via tail vein on
dosing day 8 in the range-finding and days 8 and 11 in the pilot
study. Plasma was prepared, stored at −80◦ C and later assayed
for corticosterone concentration using a corticosterone enzyme
immunoassay kit (product number 500655; Cayman Chemical,
Ann Arbor, MI). Corticosterone levels of mice after acclimation
swims and 11 days of 134 mg/kg/day 2,4-DNT treatment (151
± 51 pg/ml; mean ± SE, n = 4) were not elevated over that for
vehicle-exposed mice (217 ± 46) or unexposed mice that had not
swam (158 ± 55 pg/ml) (Supplementary table 1; p = 0.112). All
corticosterone levels were within the range for mice in the absence of imposed experimental stress (Pruett et al., 2009). Based
on a lack of overt effects or impacts on corticosterone levels, 134
mg/kg/d 2,4-DNT was selected as the optimal dose and a 5% of
body weight load for the exhaustive swim. Swim times for the
range-finding studies and the pilot study can be found in Supplementary table 2.
Definitive swim test. The definitive study consisted of four treatment conditions (n = 12 per treatment condition) including WT
or PPAR␣ (-/-) mice dosed orally with 134 mg/kg/day 2,4-DNT
or vehicle for 14 days. Swim trials were executed as described
above. Whole blood was collected from a nicked tail vein using
capillary tubes on dosing day 11. Cervical dislocation was utilized for euthanasia in the definitive study as a replacement for
CO2 -based asphyxia to eliminate confounding serum lactate levels, as previously noted by Translavina et al. (2010). On day 15,
after the exhaustive swim trial, whole blood was collected from
the heart by thoracotomy which was performed immediately after cervical dislocation. Liver was excised, weighed and a section cut from the left lateral lobe was placed in neutral buffered
formalin for paraffin embedment and histopathological assessment. Remaining liver was minced and stored in RNAlater solution (Ambion-Life Technologies, Grand Island, NY). Spleen was
excised and weighed.
Hematology. Serum was prepared from whole blood collected immediately after the exhaustive swim trial following manufacturer’s recommendations for lactate (L-Lactate Assay kitproduct
number 700510, Cayman Chemical) and triglycerides and glucose using a Triglyceride (product number 10010303, Cayman
Chemical) and a Glucose Assay Kit (product number 10009582,
Cayman Chemical), respectively. All assays were performed using the manufacturer’s recommended protocols. Whole blood
was analyzed for hematocrit and reticulocytes. Tubes were spun
in a capillary centrifuge and hematocrit measured. A portion
of the whole blood was stained using reticulocyte stain R4132
(Sigma Aldrich, St. Louis, MO) and wedge preparations were prepared. Reticulocytes were determined as percentage of 2000 red
blood cells counted per animal. Hematology procedures were
adopted from Sharp and LaRegina (1998).
Liver and muscle glycogen. Liver and soleus striated muscle from
rear legs disarticulated at the knee were placed in neutral
buffered formalin for 24 h then stored in 70% ethanol. Glycogen content was assessed on liver and striated muscle sections
(5 ␮m) stained with Periodic acid-Schiff (PAS) base ± ␣-amylase
(Vacca, 1985).
Total RNA extraction. Total RNA was extracted from about 30 mg
of liver tissue that was stored in RNAlater solution (AmbionLife Technologies, Grand Island, NY) following manufacturer’s
instructions. Tissues were homogenized in lysis buffer with a
FAST Prep-24 instrument (MP Biomedicals, Santa Ana, CA) before RNA isolation with RNeasy kits (Qiagen, Valencia, CA). Total RNA concentrations were measured using a NanoDrop ND1000 Spectrophotometer (NanoDrop technologies, Wilmington,
DE). The integrity of total RNA was assessed on an Agilent 2100
Bioanalyzer (Palo Alto, CA). Criteria for acceptable RNA integrity
included an RNA integrity number ≥ 7.5 from the Agilent 2100
Bioanalyzer and a 260/280 spectrophotometric reading ⬎ 2.0.
Gene expression assays. To select a subset of replicates of sufficient number for statistical power whereas optimizing resources, a random number generator was used to select 6 of
12 overall replicate mice from each of the four experimental
groups for gene expression analysis. Total RNA collected from
liver tissues of each replicate was used to synthesize cDNA using SABiosciences (Valencia, CA) RT2 First Strand Kit. The cDNA
was used in each of the following qPCR arrays that analyzed
84 genes involved in fatty acid metabolism (catalog number
PAMM-007, SABiosciences), glucose metabolism (catalog number PAMM-006, SABiosciences), or PPAR activation and response
(catalog number PAMM-149Z, SABiosciences) following the manufacturer’s protocols. The qPCR assays were conducted using
an Applied Biosystems 7900HT (Grand Island, NY) instrument.
Gene expression data are publicly available at the Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/info/
linking.html) under series accession GSE50170.
Canonical pathway analysis. Significantly enriched canonical
pathways were identified for each experimental treatment using
Ingenuity Pathway Analysis (IPA) software (Ingenuity Systems,
Redwood City, CA) based on all differentially expressed genes
(p⬍0.05 and ≥1.5-fold difference). The same set of differentially
expressed genes for each treatment condition was mapped onto
the Kyoto Encyclopedia of Genes and Genomes (KEGG; Kanehisa,
2000) PPAR signaling pathway (map03320) because PPAR signaling and/or activation pathways were found to be significantly
impacted using IPA.
PPAR nuclear activation and inhibition bioassays. To determine the
effect of 2,4-DNT on PPAR signaling, nuclear receptor reporter
assays were conducted (PPAR␣, PPAR␥ , PPAR␦, and RXR␣ human cell-based assays, Indigo Biosciences, State College, PA).
Cell viability was measured in tandem with the nuclear receptor reporter assays using the live cell multiplex assay (Indigo Biosciences). All bioassays were conducted according to
the manufacturer’s specifications. Briefly, PPAR␣, PPAR␥ , PPAR␦,
or RXR␣−luciferase reporter cells were thawed in cell recovery medium for 10 min at 37◦ C. Cells were distributed into 96well plates, then dosed immediately with 2,4-DNT in compound
screening medium at 0.01, 0.1, 1, or 10 mg/l (n = 3) diluted from a
WILBANKS ET AL.
2,4-DNT stock solution in DMSO. Final DMSO concentration in all
wells was 0.05%. Cells were also incubated with their respective
agonists as a positive control (PPAR␣: 100nM GW590735; PPAR␥ :
1000nM rosiglitazone; PPAR␦: 30nM GW0742; RXR␣: 1000nM 9-cis
retinoic acid), a negative control (PPAR␣: 100nM GW6471; PPAR␥ :
30nM GSK0660; PPAR␦: 1000nM GW9662), or vehicle control.
After the 24 h exposure to 2,4-DNT, the cells were rinsed
with live cell multiplex assay buffer and then incubated with
live cell multiplex assay media (containing calcein-AM) for 45
min at 37◦ C and 5% CO2 . Fluorescence of calcein (the cleaved
product indicative of cell survival) was measured at 492 nm
excitation/513 nm emission with a spectrophotometer (Tecan
Safire v2.20, Research Triangle Park, NC). The media containing calcein was discarded and replaced with luciferase detection
reagent. Luminescence (relative luminescence units) was measured with a microtiter plate luminometer (Dynex MLX 1000,
Chantilly, VA) after a 15-minute incubation with the luciferase
detection reagent. Finally, to determine the antagonistic activity of 2,4-DNT, cells were co-incubated with 2,4-DNT at 100 mg/l
and the respective receptor agonists at the concentrations listed
above.
Statistical analyses. Body and organ weights, clinical, and swim
results are presented as mean ± standard error. SigmaStat version 3.1 software (Systat Software, Incorporated, Chicago, IL) was
used to analyze organ weights, body weights, hematocrit, reticulocytes, triglycerides, and glucose data with two-way analysis of
variance (ANOVA) performed with post-hoc comparisons made
using the Tukey method. A Student’s t-test was used to compare groups [PPAR␣ (-/-) vehicle vs. WT-2,4-DNT and PPAR␣ (/-) 2,4-DNT vs. WT vehicle] which were not specifically tested
in the ANOVA post-hoc test. A square root transformation with
3/8 continuity factor was performed on reticulocyte data before
analysis to meet the assumptions of the test. Statistical significance was indicated by p⬍0.05.
Generalized linear models with Gaussian distribution and
identity link functions were used to test treatment effects on
times to exhaustion in the swimming trials using Stata Statistical Software (Stata Statistical Software: Release 12. College Station, TX: StataCorp LP). A multiple imputation method was used
to impute swim time for one missing observation, WT vehicle (Allison, 2001). Huber-White robust standard errors and 95%
confidence intervals were calculated. Estimated means were calculated for each group and differences in means were reported
(Huber, 1967; White, 1980). Alpha was adjusted for multiple comparisons using Bonferroni correction. Sensitivity analysis was
conducted using ANOVA and Kruskal-Wallis rank test and found
to give similar qualitative results.
Gene expression assay data were analyzed using RT2 Profiler
PCR Array Data Analysis software version 3.5 (SABiosciences).
The 2(⌬ ⌬ Ct) method was used to determine the fold change in
transcript expression relative to WT vehicle-exposed animals.
When comparisons were made between PPAR␣ (-/-) vehicle and
2,4-DNT-exposed mice, expression values were made relative to
PPAR␣ (-/-) vehicle-exposed mice. Expression values were analyzed for treatment differences using Student’s t-test and were
considered significantly different at p⬍0.05 including a 1.5-fold
cutoff. Significance for canonical pathways using IPA was based
on a right-tailed Fisher exact test.
All cell culture-base nuclear receptor and cell viability experiments were performed in at least triplicate. Mean and standard
deviation were calculated for each chemical treatment. Z’ values
were determined according to Zhang et al. (1999). Statistical significance between chemical treatment and vehicle control was
47
determined using one-way ANOVA with Holm-Sidak post-hoc
test (p⬍0.05) (SigmaStat, version 3.1).
RESULTS
Toxicity Measurements
One animal death in the C57/BL6 vehicle treatment group occurred due to a gavage accident. Animals within the WT 2,4DNT treatment group lost significantly more weight than animals within the other three treatment groups (Table 1). Additionally, the WT animals lost significantly more weight than
PPAR␣ (-/-) animals in each vehicle and especially the 2,4-DNT
exposure conditions. The WT 2,4-DNT treatment group and both
the PPAR␣ (-/-) vehicle and PPAR␣ (-/-) 2,4-DNT treatment groups
displayed statistically significant increases in liver weight in
comparison to the WT vehicle treatment group (Table 1). No significant changes were observed in spleen weights for any treatment group.
Clinical Chemistry
Hematocrit values did not vary across treatment conditions (Table 1). The PPAR␣ (-/-) 2,4-DNT treatment group had a significant increase in the number of reticulocytes in peripheral blood
compared against the PPAR␣ (-/-) vehicle and WT vehicle treatment groups. Reticulocyte counts were not affected in WT controls or WT animals exposed to the 134 mg/kg/day 2,4-DNT dosing. Collectively, these results demonstrated equivalent erythrocyte levels for all groups at the time of swimming and, hence,
indicated that effects on exercise performance would not be affected by impaired oxygen delivery to tissues. No significant differences in serum triglyceride concentration were seen between
the 2,4-DNT-treated WT or PPAR␣ (-/-) group compared with
their respective vehicle controls (Table 1). However, the PPAR␣ (/-) 2,4-DNT treatment group had significantly higher (53%) concentrations of triglycerides in serum than WT 2,4-DNT-treated
animals. No significant differences in serum glucose concentrations were observed between the 2,4-DNT-treated WT or PPAR␣
(-/-) animals and their respective vehicle controls. Significantly
higher serum glucose concentrations were observed in the WT
2,4-DNT treatment group where measurements were 36% and
42.5% greater compared with the PPAR␣ (-/-) 2,4-DNT and PPAR␣
(-/-) vehicle treatment groups, respectively.
Swim to Exhaustion
The most noticeable difference between swim times of any two
treatment conditions was between WT 2,4-DNT and WT vehicle exposures where the 2,4-DNT treatment in the WT mice reduced average swim time by 82% (Fig. 1). The WT vehicle treatment group swam 1.26 times longer than the PPAR␣ (-/-) vehicle
treatment group however there was no statistically significant
difference among these conditions. Swim times of PPAR␣ (-/-)
mice were much less affected by 2,4-DNT where PPAR␣ (-/-) 2,4DNT-treated mice swam significantly longer (2.9 times) than WT
2,4-DNT-treated mice. Comparatively, 2,4-DNT dosing in PPAR␣
(-/-) mice reduced swim time by only one-third.
Liver and Striated Muscle Glycogen
PAS base staining of liver sections from mice after exhaustive exercise revealed PAS-positive material, purple-magenta in
color possibly indicating polysaccharide, that was seen to bridge
centrilobular regions when viewed at low power (Figs. 2A–D).
Higher power revealed red punctate staining of centrilobular
hepatocytes (Figs. 2E–H) that was sensitive to glycogen-digesting
enzyme, ␣-amylase (Figs. 2I and J). PAS staining appeared more
48
TOXICOLOGICAL SCIENCES, 2014, Vol. 141, No. 1
TABLE 1. Weights and Clinical Parameters of WT and PPAR␣ (-/-) Mice Exposed to 134 mg/kg/d 2,4-DNT or Vehicle for 14 Days
Parameter
Body weight gain (g)
Relative liver weight (%)
Relative spleen weight
(%)
Serum lactate (mM)
Hematocrit (%)
Reticulocytes (%)
Serum triglycerides
(mg/dl)
Serum glucose (mg/dl)
WT vehicle
WT 2,4-DNT
PPAR␣ (-/-) vehicle
PPAR␣ (-/-) 2,4-DNT
− 0.35 ± 0.23
3.86 ± 0.09
0.37 ± 0.03
− 0.96 ± 0.15
4.40 ± 0.06
0.34 ± 0.02
− 0.11 ± 0.21c
4.19 ± 0.09a
0.38 ± 0.02
0.46 ± 0.26b
4.47 ± 0.13d
0.36 ± 0.01
5.60 ± 0.9
54.63 ± 0.58
1.35 ± 0.26
155.67 ± 22.33
5.54 ± 0.5
53.37 ± 0.32
1.58 ± 0.18
132.47 ± 13.79
5.36 ± 0.7
53.78 ± 0.36
1.34 ± 0.14
166.78 ± 17.16
4.95 ± 0.8
54.48 ± 0.52
2.12 ± 0.22d
201.91 ± 24.38b
157.56 ± 20.89
188.99 ± 16.41
132.63 ± 19.60c
139.14 ± 7.50b
Note. Values shown are mean ± SEM. All values represented as “different” in the following text represent statistically significant difference, p⬍0.05. Bold values are
2,4-DNT treatments different than vehicle control. All pair-wise tests among experimental treatments are also provided (significant differences are highlighted by
italics) with test descriptions denoted by the following superscripts:
a
PPAR␣ (-/-) vehicle different than WT vehicle.
b
PPAR␣ (-/-) 2,4-DNT different than WT 2,4- DNT.
c PPAR␣ (-/-) vehicle different than WT-2,4-DNT.
d PPAR␣ (-/-) 2,4-DNT different than WT vehicle.
FIG. 1. Swim times of WT and PPAR␣ (-/-) mice exposed to 134 mg/kg/d 2,4-DNT or vehicle for 14 days. Open circles are individual times for each animal. Filled circles
are swim time means for each treatment with error bars representing SEM. Brackets are used to show comparisons between treatment groups. Numbers to the left in
each bracket are the coefficient used in the model with the robust standard error in parenthesis. Numbers to the right of each bracket are the 95% confidence interval.
P values are shown below each bracket.
intense with livers from WT compared with PPAR␣ (-/-) mice, but
was unaffected by 2,4-DNT. Intracellular vesicular bodies were
seen in livers of 2,4-DNT-exposed PPAR␣ (-/-) animals (arrows,
Fig. 2H), but the structures were smaller and PAS-negative, unlike those noted earlier in livers of rats from acute exposure to
400 mg/kg 2,4-DNT (Deng et al., 2011). ␣-Amylase-sensitive PAS
staining of striated muscle was likewise unaffected by 2,4-DNT
in neither PPAR␣ (-/-) nor WT mice (Supplementary fig. 2). Muoio
et al. (2002) examined glycogen levels in exercised or starved WT
and PPAR␣ (-/-) mice and found that that liver glycogen levels
was a more reliable indicator of exhaustion than glycogen levels
in muscle tissues (Muoio et al., 2002). These results indicate that
WILBANKS ET AL.
49
FIG. 2. Histopathology of livers from WT (C57Bl/6N) and PPAR␣ (-/-) mice exposed to 134 mg/kg/day 2,4-DNT or vehicle for 14 days. PAS base staining revealed
red punctate staining of centrilobular hepatocytes (A–H). Slides pre-exposed with glycogen-digesting enzyme, ␣-amylase, before PAS staining (I and J). Intracellular
vesicular bodies were seen in 2,4-DNT-exposed PPA␣ (-/-) livers (arrows (H)).
50
TOXICOLOGICAL SCIENCES, 2014, Vol. 141, No. 1
WILBANKS ET AL.
51
FIG. 3. Transcript expression within KEGG canonical PPAR signaling pathway derived from liver tissue of mice. Expression profiles represent: (A) WT 2,4-DNT-exposed,
(B) PPAR␣ (-/-) 2,4-DNT-exposed, or (C) PPAR␣ (-/-) vehicle-exposed experimental treatments with fold change values calculated relative to WT, vehicle-exposed animals.
Blue highlighted boxes represent transcripts having significantly decreased expression relative to WT vehicle-exposed animals. Red highlighted boxes represent
transcripts having significantly increased expression relative to WT vehicle-exposed animals. Gene names are listed in Table 1.
the observed deficit in endurance exercise was not likely a consequence of 2,4-DNT impairment of mobilization of stored liver
glucose, in the form of glycogen, as an energy source.
Transcript Expression Relative to WT Vehicle-Treated Mice
Many transcripts coding for genes involved in fatty acid
metabolism had significantly decreased expression in WT 2,4DNT, PPAR␣ (-/-) vehicle, or PPAR␣ (-/-) 2,4-DNT-treated mice in
comparison to WT vehicle-treated mice (Table 2). Of these genes,
expression decreased by −1.5- to −3-fold except for Cyp4a10,
Decr1, Decr2, Ppara, Scd-1, and Slc27a2 genes which were decreased to a greater degree. Both the PPAR␣ (-/-) 2,4-DNT and
PPAR␣ (-/-) vehicle treatment groups had 4.5-fold decreased
expression of Decr2 compared with WT vehicle-treated mice.
Cyp4a10 expression was decreased dramatically in PPAR␣ (-/-)
vehicle-treated (−200-fold) and PPAR␣ (-/-) 2,4-DNT-treated animals (−76-fold). Although the WT 2,4-DNT group did not have
differentially expressed PPAR transcription factors, PPAR␣ (-/) animals had not only the expected decreased expression of
PPAR␣ (−1100- and −2350-fold expression in vehicle-exposed
and 2,4-DNT-exposed, respectively) but had significantly increased PPAR␥ (4- and 4.4-fold expression in vehicle-exposed
and 2,4-DNT-exposed treatments, respectively) in comparison to
WT vehicle-treated animals. PPAR␣ (-/-) vehicle-treated animals
also had increased retinoid x receptor alpha expression (1.5-fold)
when compared with vehicle-treated WT mice. No significant effects (p⬍0.05 including a 1.5-fold cutoff) were observed in other
transcripts associated with PPAR (␣, ␦/␤, or ␥ ) function including
steroid receptor coregulators (Ncoa3 (Src3), Ncoa6); PPAR coactivators (Ppargc1a, Ppargc1b, Pprc1); or retinoid X receptors (Rxrb
and Rxrg) relative to WT control-treated mice (GEO accession
number GSE50170). A notable finding was that 2,4-DNT induced
expression of 10 genes involved in either fatty acid or glucose
metabolism observed in WT mice that were not affected in the
PPAR␣ (-/-) mice, whether vehicle or 2,4-DNT treated (Table 2).
Transcript Expression Relative to PPAR␣ (-/-) Vehicle-Treated Mice
Transcripts for two genes (Acot2 and Acot6) involved in fatty acid
metabolism via fatty acid oxidation had significantly increased
expression in the PPAR␣ (-/-) 2,4-DNT treatment group relative to
the PPAR␣ (-/-) vehicle treatment (Table 2). Only Cyp7a1, involved
in fatty acid metabolism via cholesterol metabolism, was found
to have significantly decreased expression in the PPAR␣ (-/-) 2,4DNT-treated group. Three genes involved in glucose metabolism
(G6pc, Pcx, and Pdp2) had decreased expression in PPAR␣ (-/) 2,4-DNT-exposed mice relative to PPAR␣ (-/-) vehicle-treated
mice. Each of these genes participates in glucose metabolism
via unique pathways; G6pc in gluconeogenesis, Pcx in the tricarboxylic (TCA) cycle, and Pdp in glucose regulation.
Impacts on PPAR Signaling Pathway
Differentially expressed gene transcripts from each experimental condition relative to WT controls were mapped onto the
PPAR signaling pathway specific for mouse in the KEGG database
(mmu03320) to determine effects of 2,4-DNT on energy production associated with PPAR signaling (Fig. 3). WT 2,4-DNTexposed animals showed decreased transcript expression for
several genes involved in fatty acid transport, both upstream
and downstream of PPAR signaling pathways as well as decreased expression of a transcript upstream of the gluconeo-
52
TOXICOLOGICAL SCIENCES, 2014, Vol. 141, No. 1
TABLE 2. Differentially Expressed Transcripts Involved in Fatty Acid or Glucose Metabolism and PPAR Activation in Liver Tissue of WT and
PPAR␣ (-/-) Mice Exposed to 134 mg/kg/day 2,4-DNT
Values represent fold change relative to WT, vehicle-exposed mice where gray highlights represent significant down-regulation (p⬍0.05 and ⬎1.5-fold change) and boxed
values represent significant increased expression (p⬍0.05 and ⬎1.5-fold change). Values in the last column represent fold change relative to PPAR␣ (-/-), vehicle-exposed
mice where gray highlights and boxed values represent significant down-regulation (p⬍0.05 and ⬎1.5-fold change) and unhighlighted, unboxed values represent
significant increased expression (p⬍0.05 and ⬎1.5-fold change).
genesis pathway (Fig. 3A). 2,4-DNT had no effect on PPAR␣ expression in WT animals (p = 0.58). PPAR␣ (-/-) 2,4-DNT-exposed
animals showed decreased transcript expression for genes involved in ketogenesis, lipogenesis, fatty acid transport (both upstream and downstream of PPAR signaling), fatty acid oxidation,
and adipocyte differentiation pathways (Fig. 3B). Expression in
the PPAR␣ (-/-) 2,4-DNT-exposed and PPAR␣ (-/-) vehicle-exposed
mice (Fig. 3C) were largely equivalent to the WT vehicle-exposed
animals (Fig. 3). Notable exceptions comparing the PPAR␣ (-/-)
2,4-DNT versus PPAR␣ (-/-) vehicle treatments included: an elimination of increased expression for retinoid X receptor ␣ (RXR␣),
elimination of decreased expression for acetyl-CoA acyltransferase 1 (thiolase B), and decreased expression of fatty acid desaturase 2 (delta-6-desaturase) and angiopoietin-like 4 (PGAR)
transcripts.
Canonical Pathway Analysis
An investigation of overall canonical pathway enrichment indicated 13 common pathways (Fig. 4) among experimental conditions. The common pathways span three major functional
categories: lipid metabolism, signaling pathways, and amino
acid metabolism. The majority of enriched pathways were in-
WILBANKS ET AL.
53
FIG. 4. Effects of 2,4-DNT exposure on canonical pathways in liver tissue of WT 134 mg/kg/day 2,4-DNT exposed or PPAR␣ (-/-) vehicle or 134 mg/kg 2,4-DNT mice for
14 days. Names of significantly regulated canonical pathways are listed for WT or PPAR␣ (-/-) 2,4-DNT-exposed mice and those pathways shared in all three treatment
conditions in relevance to WT vehicle-exposed mice.
volved in lipid metabolism (8 of 13) and three of the four enriched signaling pathways are known upstream regulators of
lipid metabolism (Fig. 4). These results indicate similarity in the
effects of the PPAR␣ (-/-) and the effects of 2,4-DNT exposure
in WT animals. Additionally, there were unique responses by
WT animals exposed to 2,4-DNT that may serve as indicators
for poor performance in the swim trial. Specifically, WT 2,4DNT-treated animals, which had the poorest performance in the
swim test (Fig. 1), had six pathways uniquely enriched within
biological functions including: carbohydrate metabolism, lipid
metabolism, and organ development (Fig. 4). As a class, pathways involved in carbohydrate metabolism were the most affected, where all genes involved in the pathways had significantly decreased expression (Table 2).
PPAR Nuclear Activation and Inhibition Bioassays
The 2,4-DNT exposures caused complex effects on PPAR nuclear
signaling in the in vitro assessments (Fig. 5). In the activation
assays, 2,4-DNT caused significant impairment of nuclear signaling for both PPAR␣ and PPAR␦ whereas the 2,4-DNT caused
significantly increased PPAR␥ nuclear signaling at the highest
exposure concentration (10 mg/l). The nuclear receptor inhibition assays indicated that 2,4-DNT was not a strong competitor
for the binding sites of PPAR␣ or PPAR␥ , compared with their respective agonists; however, significant inhibition was observed
for PPAR␦ signaling. Finally, neither exposure to 2,4-DNT or positive controls affected cellular survival (Fig. 5).
DISCUSSION
Exposure to nitrotoluenes at high doses has been observed
to cause hemolysis, anemia, and methemoglobinemia (ATSDR,
1998) which can impair oxygen available for use by an exposed
animal and therefore potentially impair exercise performance.
This study used a dose of 2,4-DNT that is not expected to have
altered oxygen transport in WT 2,4-DNT-exposed animals. The
only relevant effect noted regarding red blood cells was an increase in reticulocytes of 2,4-DNT- PPAR␣ (-/-) mice; however,
oxygen delivery was unlikely to be impaired because hematocrit
was restored to normal at the time of the endurance challenge
(Table 1).
Further evidence that the administered dose of 2,4-DNT did
not alter the extent of aerobic energy metabolism was the ab-
sence of a change in serum lactate, a marker of aerobic energy
pathways being depleted (Table 1). Therefore, our range-finding
and pilot studies allowed us to design a definitive experiment
to specifically interrogate the impacts of 2,4-DNT exposure on
PPAR signaling in the absence of confounding changes in systemic oxygen transport and/or aerobic respiration.
AOP
We tested if PPAR␣ played a significant role in facilitating 2,4DNT-based impacts on energy metabolism and subsequent effects on exercise performance. We used the results of this study
to create an AOP (Ankley et al., 2010) for effects associated with
2,4-DNT exposure specifically highlighting impacts on overall
PPAR signaling connected to effects observed across multiple
levels of biological organization (Fig. 6). The goal of this study
was not to focus on the already well-described hematoxicity,
but instead explore direct impacts on PPAR signaling and downstream effects on fatty acid and glucose metabolism that affect
the energy substrates needed for exercise performance. The AOP
(Fig. 6) serves as the organizational outline for the following discussion.
MIE (Impaired PPAR␣ signaling)
The present study provides two lines of evidence supporting the
hypothesis based on previous studies (Deng et al., 2011; Rawat
et al., 2010; Wintz et al., 2006) that 2,4-DNT-induced interference
with PPAR␣ signaling impacts energy metabolism and therefore
represents the MIE for our proposed AOP (Fig. 6). First, 2,4-DNT
significantly decreased PPAR␣ signaling when tested with in vitro
rat/human PPAR nuclear activation assays (Fig. 5). Although 2,4DNT was not a strong competitor against the known agonist for
PPAR␣ nuclear signaling when both substrates were available at
high relative concentrations, 2,4-DNT did reduce basal PPAR␣
nuclear signaling in the activation assay (Fig. 5). PPAR␣ is a primary transcriptional regulator for lipid metabolism (Desvergne
and Wahli, 1999; Lefebvre et al., 2006), therefore the observation of reduced transcriptional expression for genes involved
in lipid metabolism (Table 2, Fig. 3) was expected in addition
to the resultant impacts on lipid metabolism and downstream
energy homeostasis (Supplementary fig. 1). Although 2,4-DNT
did not directly affect PPAR␣ expression in WT 2,4-DNT-exposed
mice (Table 2), the PPAR␣ activation pathway was significantly
enriched (p = 0.00085) having significantly decreased transcrip-
54
TOXICOLOGICAL SCIENCES, 2014, Vol. 141, No. 1
FIG. 5. Nuclear receptor activation assay conducted in Chinese hamster ovary cells.*p⬍0.05.
tional expression (Fig. 3A, Table 2). Future studies investigating
PPAR␣ and PPAR␥ protein levels in tissue are needed to further
develop a detailed mechanistic description of the MIE of impaired PPAR signaling in order to transition this AOP into a quantitative model of effects.
The second line of evidence supporting the hypothesized MIE
was a direct connection between PPAR␣ signaling and the apical
organism-level outcome where swim performance by PPAR␣ (/-) mice was significantly less impacted by 2,4-DNT than in WT
mice. This observation indicates that the interaction between
2,4-DNT and PPAR␣ signaling is a primary source for adverse impacts on exercise performance at the whole-organism level. Together, these observations demonstrate the importance of 2,4DNT-mediated impairment of the PPAR␣ signaling pathway in
propagating apical impacts on exercise endurance on the individual.
Molecular Response (Impaired Lipid Metabolism)
2,4-DNT reduced transcriptional expression of genes for which
PPAR␣ acts as a positive transcriptional regulator in WT mice
(Fig. 3A). The majority of the affected transcripts in WT 2,4-DNT-
exposed mice were involved in the transport of fatty acid (the
PPAR ligand) into the cell representing six of the nine total differentially expressed transcripts involved in lipid metabolism (Table 2). Specifically, transcripts for fatty acid binding proteins 2,
3, and 7 (Fabp2, 3, and 7), which function as inter- and intracellular transporters (Storch and Thumser, 2000) and mediators
of PPAR␣ and ␥ nuclear signaling (Wolfrum et al., 2001), had decreased expression. Additionally, Cd36 (fatcd36), a known fatty
acid transporter (Baillie et al., 1996) and solute carrier family 27
(fatty acid transporter) members 2 and 5 (SLC27A2 and 5) which
work alone or in conjunction with Cd36 to transport fatty acids
into cells (Stahl, 2004), showed decreased expression (Table 2).
Given that fatty acid metabolism is an inducible pathway where
increasing substrate levels activate PPAR signaling and gene expression for enzymes catalyzing fatty acid metabolism, the decreased availability of fatty acid transport proteins would be expected to lead to decreased expression of genes downstream in
the PPAR␣ regulatory network. The observation of decreased expression within the PPAR␣ signaling pathway in WT mice (Fig.
3A) is consistent with this expected effect thus revealing a feedback response where reduced levels of fatty acid transport pro-
WILBANKS ET AL.
55
FIG. 6. A proposed AOP associated with 2,4-DNT exposure. MIE = molecular initiating event.
teins would negatively affect downstream transcription of their
upstream gene transcripts (Table 2).
Other transcriptional effects of 2,4-DNT on genes involved
in WT lipid metabolism (Table 2) included decreased expression
for acyl-CoA thioesterase 2 (Acot2), a gene involved in fatty acid
oxidation (Hunt and Alexson, 2002), and glycerol-3-phosphate
dehydrogenase 1 and 2 (Gdph1 and 2), genes that facilitate glycerol catabolism (Patsouris et al., 2004). Both Acot and Gdph are
positively regulated by PPAR␣ (Dongol et al., 2007; Hunt and Alexson 2002; Patsouris et al., 2004) providing an additional indication of 2,4-DNT impairment of PPAR␣ signaling pathways and
decreased expression of downstream targets (Fig. 3A). Overall,
impaired PPAR␣ signaling (Fig. 5) due to 2,4-DNT exposure resulted in decreased expression of genes necessary for transporting fatty acids into cells (Fig. 3A) as well as for genes that convert
fatty acids and glycerol to cellular energy (Table 2). These results
indicate the reduced potential for 2,4-DNT-exposed WT animals
to utilize lipid resources in the maintenance of a balanced cellular energy budget.
PPAR␣ (-/-), vehicle-exposed mice also showed reduced expression of fatty acid transport proteins relative to WT vehicleexposed mice (Table 2). However, the PPAR␣ knockdown primarily affected Cd36 and Slc27A transcription whereas Fabps
were also affected in 2,4-DNT-exposed WT mice. The most pronounced difference between PPAR␣ (-/-) mice and WT mice exposed to 2,4-DNT was the decreased expression of genes in-
volved in fatty acid oxidation in PPAR␣ (-/-) mice (Table 2). These
results indicate that 2,4-DNT acts by inhibiting PPAR␣ signaling
uniquely in WT mice compared with PPAR␣ (-/-) mice with directed impacts primarily on genes involved in fatty acid transport as opposed to fatty acid oxidation.
Cellular Response (Impaired Lipid-Based Energy Production)
Knockdown of PPAR␣ in combination with 2,4-DNT exposure
caused elevated concentrations of serum triglycerides compared
with WT mice (Table 1) and PPAR␣ (-/-) vehicle-exposed mice.
The decrease in expression of fatty acid oxidation genes in
PPAR␣ (-/-) mice (Table 2) was reflected in a higher triglyceride
level in serum consistent with the reduced potential to metabolize fatty acid moieties (Table 1).
Organ Response (Effects on Liver)
2,4-DNT dosing produced altered PPAR␣ signaling with significantly increased liver weight in WT animals in agreement with
other published observations in rat (Deng et al., 2011; Lee et al.,
1978) and fathead minnow (Wintz et al., 2006). Wintz et al. (2006)
reported that increased liver weights in fathead minnow exposed to 2,4-DNT in a 10-day study were due to increased phospholipid accumulation hypothesized to stem from the decreased
transcriptional expression of genes involved in lipid catabolism.
Similarly, in the present study intracellular vesicular bodies, potentially lipid vesicles, were observed in liver tissue of 2,4-DNT-
56
TOXICOLOGICAL SCIENCES, 2014, Vol. 141, No. 1
treated PPAR␣ (-/-) animals (Fig. 2). Although 2,4-DNT effects on
glycogen and lipid were not observed in tissue sections of liver
and striated muscle of exercised mice, future studies are needed
in order to quantitatively measure glycogen and lipid in liver and
muscle of both exercised and non-exercised animals in order to
further develop the proposed AOP.
Individual Response (Negative Energy Budget)
The 2,4-DNT-induced impairment of PPAR␣ signaling and subsequent reduced transcript expression for genes facilitating lipid
metabolism has implications for mice in maintaining a positive
energy budget as lipid represents a principal energy currency in
animal physiology. In support of this hypothesis, 2,4-DNT impairment of lipid metabolism leading to a negative energy budget has been observed in carp exposed to 2,4-DNT for 15 days
where the fish had decreased amounts of lipid but not total protein stored in liver tissue (Xu and Jing, 2012). The reduced potential to utilize lipid substrates likely contributed to a negatively
impacted whole-organism energy budget as manifested by significantly reduced body weights observed in WT mice exposed
to 2,4-DNT (Table 1). 2,4-DNT has also been shown to affect oxidative phosphorylation which may be linked to the observed
effects on lipid metabolism (Xu and Jing, 2012).
Molecular Response (Compensatory PPAR␥ Signaling)
Increases in PPAR␥ expression in response to PPAR␣ impairment
was observed in the PPAR␣ (-/-) vehicle exposure (Fig. 3C, Table 2). PPAR␥ plays a role related to PPAR␣ in lipid metabolism,
but also has a unique role in cycling lipid and carbohydrate substrates into glycolytic/gluconeogenic pathways (Fig. 3, Millward
et al., 2010; Powell et al., 2007). We posit that our observations
in the PPAR␣ (-/-) model demonstrate evidence of a compensatory response to ameliorate the decreased potential for fatty
acid metabolism and subsequent energy production given the
loss of PPAR␣ signaling thereby necessitating increased PPAR␥
signaling to sustain energy needed by the animal (Fig. 3B). This
was further confirmed by the nuclear activation bioassays where
PPAR␥ was activated by 2,4-DNT (Fig. 5), thus further mitigating
the effects of 2,4-DNT on energy metabolism pathways.
Molecular Response (Glycolysis vs. Gluconeogenesis)
2,4-DNT dosing caused significantly decreased transcriptional
expression of genes involved in glucose metabolism in WT mice
and eliminated constitutively increased transcriptional expression of genes involved in glucose metabolism in PPAR␣ (-/-)
mice (Table 2). These results are consistent with previous observation that DNTs affect the transcriptional expression of
genes involved in the glycolysis/gluconeogenesis pathway in
bobwhite quail (Rawat et al., 2010) and fathead minnow (Wintz
et al., 2006). Specifically, the current study demonstrated that
transcriptional expression of three genes critical within the
glycolysis/gluconeogenesis pathway had decreased expression
in 2,4-DNT-exposed WT mice (Table 2, Supplementary fig. 1A):
phosphoenolpyruvate carboxykinase 1 (Pck1), pyruvate carboxylase (Pcx), and fructose biphosphatase 1 (Fbp1). Of particular
note, Pck1 catalyzes the rate-limiting reaction for gluconeogenesis (Gómez-Valadés et al., 2008). Decreased transcriptional expression of these genes is consistent with a shift away from
gluconeogenesis in favor of increased glycolysis (Supplementary
fig. 1A, Voet et al., 1999). As evidence of this metabolic effect in an
alternative species, carp exposed to 2,4-DNT had decreased glucokinase, an enzyme involved in glycolytic and gluconeogenic
pathways, along with decreased glycogen stores in liver after 7
and 15 days (Xu and Jing, 2012). Similar to the observations in
PPAR␣ (-/-) mice described in the previous section, the 2,4-DNT
exposure in the WT mice is potentiating this pathway that controls central energy metabolism in favor of energy production
versus energy storage.
Although the overall glycolysis/ gluconeogenesis pathwaylevel response to 2,4-DNT exposure is similar to the response
observed in PPAR␣ (-/-) mice, the molecular mechanisms facilitating energy use are largely dissimilar. The principal difference
is that WT mice exposed to 2,4-DNT decreased expression of
molecular targets to elicit lowered gluconeogenic flux compared
with the approach of PPAR␣ (-/-) mice which increased transcriptional expression of PPAR␥ and PPAR␥ pathway signaling elements to achieve increased energy production from glucose (Table 2, Figs. 3 and 6, Supplementary fig. 1). When compared with
WT mice, PPAR␣ (-/-) vehicle-exposed animals had increased expression of genes involved in glycolysis (phosphoglucomutase,
Pgm3 and pyruvate kinase-liver and red blood cell, Pklr; Supplementary fig. 1B), a gene involved in glycogen regulation (glycogen synthase kinase 3 beta, Gsk3b; Supplementary fig. 1B), and a
gene involved in the pentose phosphate pathway (phosphoribosyl pyrophosphate synthetase 2, Prps2, Table 2). PPAR␣ (-/-) mice
appear to have compensated for the loss of PPAR␣ by activation
of glycolysis/gluconeogenesis pathway in order to maintain energy metabolism compared with WT animals forced to abruptly
respond to 2,4-DNT-induced PPAR␣ signaling impairment.
Organ Response (Blood Glucose Levels)
As apical evidence of glycolytic rerouting described above,
serum glucose concentration in the PPAR␣ (-/-) vehicle and
PPAR␣ (-/-) 2,4-DNT-treated groups were significantly decreased
compared with the WT 2,4-DNT-treated group (Table 1).
Individual Response (Weight Loss and Decreased Exercise Performance)
2,4-DNT-treated WT animals were the only experimental group
to have weight loss significantly greater than WT controls over
the 14-day study (Table 1). Weight loss has been previously associated with 2,4-DNT exposure in rats and mice (Hong et al.,
1985; Lee et al., 1978, 1985; Lent et al., 2012), salamanders (Johnson et al., 2007), and northern bobwhite quail (Johnson et al.,
2005). Observed weight loss suggested energy budget issues in
nitrotoluene exposures and our study has provided the first evidence that 2,4-DNT-induced effects on energy metabolism can
impair exercise performance (Fig. 1). Further, the investigation of
PPAR␣ (-/-) mice has implicated PPAR␣ signaling as a significant
pathway by which 2,4-DNT disrupts energy metabolism causing impacts on both body weight and exercise performance (Table 1 and Fig. 1). These observations demonstrate important fitness implications where animals exposed to 2,4-DNT are likely
to have compromised performance in fight or flight responses
in nature leading to decreased individual survival and therefore
potential negative impacts on sustaining local populations.
SUPPLEMENTARY DATA
Supplementary data are available online at http://toxsci.
oxfordjournals.org/.
FUNDING
U.S. Army Environmental Quality and Installations applied research program. Permission was granted by the Chief of the U.S.
Army Corps of Engineers to publish this information.
WILBANKS ET AL.
ACKNOWLEDGMENT
The authors thank Dr. Tanwir Habib for his technical assistance
with bioinformatic software.
REFERENCES
Agency for Toxic Substances and Disease Registry
(ATSDR). (1998) Toxicological profile for 2,4 and 2,6-dinitrotoluene.
U.S. Department of Health & Human Services, Public Health Service, Agency for Toxic Substances and Disease Registry,
Atlanta, GA.
Allison, P. D. (2001) Missing data. Sage University Papers Series
on Quantitative Application in the Social Sciences, 07–136. Sage,
Thousand Oaks, CA.
Ankley, G. T., Bennet, R. S., Erickson, R. J., Hoff, D. J., Hornung,
M. W., Johnson, R. D., Mount, D. R., Nichols, J. W., Russom, C.
L., Schmieder, P. K., et al. (2010) Adverse outcome pathways:
A conceptual framework to support ecotoxicology research
and risk assessment. Environ. Toxicol. Chem. 29, 730–741.
Baillie, A. G., Coburn, C. T. and Abumrad, N. A. (1996) Reversible
binding of long-chain fatty acids to purified FAT, the adipose
CD36 homolog. J. Membr. Biol. 153, 75–81.
Deng, Y., Meyer, S. A., Guan, X., Escalon, B. L., Ai, J., Wilbanks, M.
S., Welti, R., Garcia-Reyero, N. and Perkins, E. J. (2011) Analysis of common and specific mechanisms of liver function
affected by nitrotoluene compounds. PLoS One 6, e14662.
Desvergne, B. and Wahli, W. (1999) Peroxisome proliferatoractivatd receptors: nuclear control of metabolism. Endocr.
Rev. 20, 649–688.
Dilley, J. V., Tyson, C. A. and Spanggord, R. J. (1982) Short-term
oral toxicity of 2,4,6-trinitrotoluene in mice, rats, and dogs. J.
Toxicol. Environ. Health 9, 565–585.
Dongol, B., Shah, Y., Kim, I., Gonzalez, F. J. and Hunt, M. C. (2007)
The acyl-CoA thioesterase I is regulated by PPARalpha and
HNF4alpha via a distal response element in the promoter. J.
Lipid Res. 48, 1781–1791.
Ellis, H. V., Hagensen, J. H. and Hodgson, J. R. (1979) Mammalian
toxicity of munitions compounds. Phase III: Effects of lifetime exposure. Part I: 2,4-Dinitrotoluene. Final Report NO. 7. U.S. Army Medical Bioengineering Research and Development Laboratory, Fort Detrick, MD. Order No. ADA077692.
Ellis, H. V., Hodgson, J. R., Hwang, S. W., Halfpap, L. M., Kelton, D. O., Andersen, B. S., VanGoethem, D. L. and Lee, C.C.
(1978) Mammalian toxicity of munitions compounds. Phase I.
Acute oral toxicity, primary skin and eye irritation, dermal sensitization, disposition and metabolism and Ames tests of additional compounds. Progress Report No. 6. Midwest Research Institute. Kansas City, MO. Contract no. DAMD 17–74-C-4073, AD
A069 444.
Feige, J. N., Gerber, A., Casals-Casas, C., Yang, Q., Winkler, C.,
Bedu, E., Bueno, M., Gelman, L., Auwerx, J., Gonzalez, F.
J., et al. (2010) The pollutant diethylhexyl phthalate regulates hepatic energy metabolism via species-specific PPAR␣dependent mechanisms. Environ. Health Perspect. 118, 234–
241.
Gómez-Valadés, A. G., Méndez-Lucas, A., Vidal-Alabró, A.,
Blasco, F. X., Chillon, M., Bartrons, R., Bermúdez, J. and
Perales, J. C. (2008) Pck1 Gene silencing in the liver improves
glycemia control, insulin sensitivity, and dyslipidemia in
db/db mice. Diabetes 57, 2199–2210.
Hong, C. B., Ellis, H. V., Lee, C. C., Sprinz, H., Dacre, J. C. and Glennon, J. P. (1985) Subchronic and chronic toxicity studies of
2,4-dinitrotoluene. Part III: CD-1 Mice. J. Am. Coll. Toxicol. 4,
57
257–269.
Huber, P. J. (1967) The behavior of maximum likelihood estimates under nonstandard conditions. In Proceedings of the Fifth Berkeley
Symposium on Mathematical Statistics and Probability. Vol.
1, pp. 221–233. University of California Press, Berkeley, CA.
Hunt, M. C. and Alexson, S. E. (2002) The role Acyl-CoA
thioesterases play in mediating intracellular lipid
metabolism. Prog. Lipid Res. 41, 99–130.
Ito, Y., Nakamura, T., Yanagiba, Y., Ramdhan, D. H., Yamagishi, N., Naito, H., Kamijima, M., Gonzalez, F. J. and Nakajima, T. (2012) Plasticizers may activate human hepatic
peroxisomeproliferator-activated receptor less than that of
a mouse but may activate constitutive androstane receptor
in liver. PPAR Res. 2012, 201284, doi:10.1155/2012/201284.
Jin, H. M. and Wei, P. (2011) Anti-fatigue properties of tartary
buckwheat extracts in mice. Int. J. Mol. Sci. 12, 4770–4780.
Johnson, M. S., Michie, M. W., Bazar, M. A. and Gogal, R. M., Jr
(2005) Influence of oral 2,4-dinitrotoluene exposure to the
northern bobwhite (Colinus virginianus). Inter. J. Toxicol. 24,
264–274.
Johnson, M. S., Suski, J. and Bazar, M. A. (2007) Toxicological responses of red-backed salamanders (Plethodon cinereus) to
subchronic soil exposures of 2,4-dinitrotoluene. Environ. Pollut. 147, 604–608.
Kanehisa, M. (2000). Post-genome Informatics. Oxford University
Press, New York, NY.
Kozuka, H. Mori-M. and Naurse, Y. (1979) Studies on the
metabolism and toxicity of dinitrotoluenes. Toxicological
study of 2,4-dinitrotolue (2.4-DNT) in long-term feeding. J.
Toxicol. Sci. 4, 221–228.
Lee, C. C., Dilley, J. V., Hodgson, J. R., Helton, D. O., Wiegand, W.
J., Roberts, D. N., Andersen, B. S., Halfpap, L. M., Kurtz, L. D.
and West, N. (1975) Mammalian toxicity of munition compounds:
Phase I. Acute oral toxicity, primary skin and eye irritation, dermal
sensitization, and disposition and metabolism. Report No. 1. Contract DAMD17–74-c-4073; Midwest Research Institute, Project no.
3900-B.
Lee, C. C., Ellis, H. V., Kowalski, J. J., Hodgson, J. R. and Hwang,
S. W. (1978) Mammalian toxicity of munitions compounds. Phase
II: Effects of multiple doses. Part II: 2,4-Dinitrotoluene. Progress
Report No. 3. U.S. Army Medical Bioengineering Research
and Development Laboratory, Fort Detrick, MD, Order No.
ADA061715.
Lee, C. C., Hong, C. B., Ellis, H. V., Dacre, J. C. and Glennon,
J. P. (1985) Subchronic and chronic toxicity studies of 2,4dinitrotoluene. Part II: CD rats. J. Am. Coll. Toxicol. 4, 243–256.
Lefebvre, P., Chinetti, G., Fruchart, J. and Staels, B. (2006) Sorting out the role of PPAR␣ in energy metabolism and vascular
homeostasis. J. Clin. Invest. 116, 571–580.
Lent, E. M., Crouse, L. C. B., Quinn, M. J., Jr and Wallace, S. M.
(2012) Comparison of the repeated dose toxicity of isomers
of dinitrotoluene. Int. J. Toxicol. 31, 143–157.
Levine, B. S., Furedi, E. M. and Gordon, D. E. (1984) Subchronic
toxicity of trinitrotoluene in Fischer 344 rats. Toxicology 32,
253–265.
Mandard, S., Muller, M. and Kersten, S. (2004) Peroxisome
proliferator-activated receptor ␣ target genes. Cell. Mol. Life
Sci. 61, 393–416.
McFarland, C. A., Talent, L. G., Quinn, M. J., Bazar, M. A.,
Wilbanks, M. S., Nisanian, M., Gogal, R. M., Jr, Johsnon, M.
S., Perkins, E. J. and Gust, K. A. (2012) Multiple environmental stressors elicit complex interactive effect in the western
fence lizard (Sceloporus occidentalis). Ecotoxicology 21, 2372–
2390.
58
TOXICOLOGICAL SCIENCES, 2014, Vol. 141, No. 1
McGee, L. C., McCausland, A., Plume, C. A. and Marlett, N. C.
(1942) Metabolic disturbances in workers exposed to dinitrotoluene. Am. J. Dig. Dis. Nutr. 9, 329–332.
Millward, C. A., Desantis, D., Hsieh, C. W., Heaney, J. D., Pisano, S.,
Olswang, Y., Reshef, L., Beidelschies, M., Puchowicz, M. and
Croniger, C. M. (2010) Phosphoenolpyruvate carboxykinase
(Pck1) helps regulate the triglyceride/fatty acid cycle and development of insulin resistance in mice. J. Lipid Res. 51, 1452–
1463.
Muoio, D. M., MacLean, P. S., Lang, D. B., Houmard, J. A., Way, J.
M., Winegar, D. A., Corton, J. C., Dohm, G. L. and Kraus, W.
E. (2002) Fatty acid homeostasis and induction of lipid regulatory genes in skeletal muscles of peroxisome proliferatoractivated receptor (PPAR) alpha knock-out mice. Evidence for
compensatory regulation by PPAR delta. J. Biol. Chem. 277,
26089–26097.
National Research Council (US) Committee for the Update
of the Guide for the Care and Use of Laboratory Animals. (2011) Guide for the Care and Use of Laboratory Animals. 8th ed. National Academies Press, Washington DC.
Patsouris, D., Mandard, S., Voshol, P. J., Escher, P., Tan, N. S.,
Havekes, L. M., Koenig, W., März, W., Tafuri, S., Wahli, W.,
et al. (2004) PPARalpha governs glycerol metabolism. J. Clin.
Invest. 114, 94–103.
Pontius, N. L. (1997) SDWA mandates drinking water contaminant candidate list, Opflow 23, 8–9.
Powell, E., Kuhn, P. and Xu, W. (2007) Nuclear receptor
cofactors in PPARgamma-mediated adipogenesis and
adipocyte energy metabolism. PPAR Res. 2007, 53843,
doi:10.1155/2007/53843.
Pruett, S. B., Fan, R., Zheng, Q. and Schwab, C. (2009) Patterns of
immunotoxicity associated with chronic as compared with
acute exposure to chemical or physical stressors and their
relevance with regard to the role of stress and with regard to
immunotoxicity testing. Toxicol. Sci. 109, 265–275.
Quinn, M. J., Jr, Bazar, M. A., McFarland, C. A., Perkins, E. J., Gust,
K. A., Gogal, R. M. and Johnson, M. S. (2007) Effects of subchronic exposure to 2,6-dinitrotoluene in the northern bobwhite (Colinus virginianus). Environ. Toxicol. Chem. 26, 2202–
2207.
Rakhshandehroo, M., Knoch, B., Muller, M. and Kersten, S. (2010) Peroxisome proliferator-activated receptor alpha target genes. PPAR Res. 2010, 612089,
doi:10.1155/2010/612089.
Rawat, A., Gust, K. A., Deng, Y., Garcia-Reyero, N., Quinn, M. J.,
Jr, Johnson, M. S., Indest, K. J., Elasari, M. O., Johnson, M. S.
and Perkins, E. J. (2010) From raw materials to validated systems: the construction of a genomic library and microarray
to interpret systematic perturbations in Northern bobwhite.
Physiol. Genomics 42, 219–235.
Rickert, D. E. and Long, D. M. (1981) Metabolism and excretion of
2,4-dinitrotoluene in male and female Fischer 344 rats after
different doses. Drug Metab. Dispos. 9, 226–232.
Sharp, P. E. and LaRegina, M. C. (1998) The Laboratory Rat. CRC
Press, Boca Raton, FL.
Stahl, A. (2004) A current review of fatty acid transport proteins
(SLC27). Pflugers Arch. 447, 722–727.
Storch, J. and Thumser, A. E. (2000) The fatty acid transport
function of fatty acid-binding proteins. Biochim. Biophys. Acta
1486, 28–44.
Takeda, K., Machida, M., Kohara, A., Omi, N. and Tohru, T. (2011)
Effects of citrulline supplementation on fatigue and exercise
performance in mice. J. Nutr. Sci. Vitaminol (Tokyo) 57, 246–250.
Translavina, R. P., King, E. J., Loar, A. S., Riedel, E. R., Garvey, M. S.,
Ricart-Arbona, R., Wolf, F. R. and Couto, S. S. (2010) Euthanasia by CO2 inhalation affects potassium levels in mice. J. Am.
Assoc. Lab. Anim. Sci. 49, 316–322.
Vacca, L. L. (1985) Laboratory manual of histochemistry. Raven
Press, New York, pp. 182–186.
Vemot, E. H., MacEwen, J. D., Haun, C. C. and Kinkead, E. R.
(1977) Acute toxicity and skin corrosion data for some organic and inorganic compounds and aqueous solutions. Toxicol. Appl.Pharmacol. 42, 417–423.
Voet, D., Voet, J. G. and Pratt, C. W. (1999) Fundamentals of Biochemistry. John Wiley & Sons, New York, NY.
White, H. (1980) A heteroskedaticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica 48, 817–830.
Wintz, H., Yoo, L. J., Loguinov, A., Wu, Y., Steevens, J. A., Holland,
R. D., Beger, R. D., Perkins, E. J., Hughes, O. and Vulpe, C. D.
(2006) Gene expression profiles in fathead minnow exposed
to 2,4-DNT correlation with toxicity in mammals. Toxicol. Sci.
94, 71–82.
Wolfrum, C., Borrmann, C. M., Borchers, T. and Spener, F.
(2001) Fatty acids and hypolipidemic drugs regulate peroxisome proliferator-activated receptors alpha - and gammamediated gene expression via liver fatty acid binding protein:
a signaling path to the nucleus. Proc. Natl Acad. Sci. U.S.A. 98,
2323–2328.
Xu, J. and Jing, N. (2012) Effects of 2,4-dinitrotoluene exposure
on enzyme activity, energy reserves and condition factors
in common carp (Cyprinus carpio). J. Hazard. Mater. 203–204,
299–307.
Zhang, J. H., Chung, T. D. and Oldenburg, K. R. (1999) A simple
statistical parameter for use in evaluation and validation of
high throughput screening assays. J. Biomol. Screen. 4, 67–73.