Center for Bioorganic Chemistry, Research Triangle Institute,
Research Triangle Park, North Carolina
1,3-Diphenyl-1-triazene (DPT) is used in the synthesis of polymers
and dyes, and has been found as an impurity in the color additives D&C
Red 33 and FD&C Yellow 5. [14C]DPT, randomly labeled in
the phenyl rings, was used to investigate its disposition in rodents.
Dermal doses to rats and mice (2 and 20 mg/cm2) were poorly
absorbed (
7%) in 72 h of exposure. Oral doses of DPT (20 mg/kg)
to male rats and mice were well absorbed and excreted mainly in the
urine, with exhalation of volatile organics accounting for about 1% of
the dose. The sole volatile component present in breath was benzene,
and all of the metabolites present in urine were composed of those
known for the differential metabolism of benzene and for aniline in the
two species. Benzene and aniline were detected in the blood of rats
administered oral doses of DPT, and relatively high circulating levels
of their metabolites were measured as early as 15 min postdosing.
Metabolites of these two carcinogens were also formed in human liver
slices, indicating a carcinogenic potential for DPT in humans.
 |
Introduction |
1,3-Diphenyl-1-triazene (diazoaminobenzene;
DPT)2
is used as a synthetic intermediate, complexing agent and polymer
additive. DPT has been found as an impurity in the
color additive, D&C Red 33 (Bailey, 1985
), which is permitted for
ingested and externally applied drugs and for cosmetics such as
lipsticks. It's also found as an impurity in FD&C Yellow 5 (Palmer and
Mathews, 1986), which is used in food products. The use of both of
these additives provides potential exposures to this compound by the
oral or dermal route. No epidemiological or case studies investigating
the cancer risks of humans exposed to DPT are available, nor have the
toxicological effects of chronic exposure to humans been reported. Some
information on the effects after acute exposures are available, however
(Aldrich Chemical Company Inc., 1995; National Library of Medicine,
1995). Effects in humans from acute dermal exposure include eye and
skin irritations and central cyanosis unresponsive to oxygen therapy. Irritation of the mucous membranes and upper respiratory tract, as well
as dyspnea and tachypnea, has been reported from inhalation exposure.
Nausea and vomiting have been reported from ingestion of the compound.
Methemoglobinemia has been reported after exposure to DPT by
inhalation, skin absorption, or ingestion. Only limited studies of
metabolism of 1,3-triazenes are found in the literature. The objective
of the present study was to determine the fate of DPT after i.v., oral,
and dermal administration to rats and mice, and to relate that
disposition to the prediction of the risks associated with its exposure.
| |
Materials and Methods |
Chemicals.
Nonradiolabeled DPT (95%) was purchased from ACROS Organics
(Pittsburgh, PA) or Aldrich Chemical Co., Inc. (Milwaukee, WI). The
identity of nonradiolabeled DPT was confirmed by NMR spectrometry. [14C]DPT, randomly radiolabeled with carbon-14
in the phenyl rings, was obtained from Wizard Laboratories, Inc. (West
Sacramento, CA) at a specific activity of 37.9 mCi/mmol. The
radiochemical purity of [14C]DPT was
established using a Waters Associates high performance liquid
chromatograph equipped with two model 510 pumps, a Rheodyne 7125 injector, and a Supelcosil LC-18-DB (Supelco, Inc., Bellefonte, PA)
analytical column (4.6 × 250 mm, 5 µm). An isocratic mobile phase of 70% acetonitrile in water was used at a flow rate of 1.0 ml/min. The column effluent was monitored by a Ramona 5-LS radioactivity detector with a solid scintillator-packed flow cell. After injection of [14C]DPT, the column
effluent was collected in fractions and radioactivity eluting in each
fraction was measured by liquid scintillation spectrometry. The
radiochemical purity of the [14C]DPT ranged
from 95 to 97% during the course of the studies. 2-Aminophenol,
4-acetamidophenol, phenol, 2-acetamidophenol, hydroquinone, and muconic
acid were purchased from Aldrich. Distilled deionized water was used in
HPLC mobile phases. Emulphor EL-620 (Emulphor) was obtained from GAF
Corporation (New York, NY).
Animal Studies.
The highest dose levels for oral (20 mg/kg) and dermal (20 mg/cm2) studies corresponded to about 1/10 the
LD50 for DPT in rodents, with the i.v. doses 1/10
the oral. Adult Fischer male and female 344 (F-344) rats and female
B6C3F1 mice were purchased from Charles River Laboratories, Inc.
(Raleigh, NC) and furnished Purina Rodent Chow (no. 5002) and water ad
libitum. Animals were quarantined at least 1 week before use, and were
housed individually after dosing. At dosing, rats were 65 to 73 days
old (males 196-249 g, females 135-144 g) and mice were 63 to 72 days
old (males 20-27 g). Intravenous dose formulations for rats and mice
contained 18 to 20 and 9 to 11 µCi [14C]DPT,
respectively, per dose and were made by dissolving
[14C]DPT and nonradiolabeled DPT in
water/Emulphor, 9:1 (v/v) to allow delivery of the target dose in a
total of 1 ml/kg (rat) and 2 ml/kg (mouse). Intravenous doses were
administered into a lateral tail vein. Dermal dose formulations
contained 22 to 31 µCi radiolabel per dose for rats and 13 to 17 µCi per dose for mice, an appropriate amount of unlabeled DPT and
acetone, for a total volume of 50 to 80 µl per dose for rats and 25 to 50 µl for mice. DPT was applied to an area of skin (2 cm2 for rats, 1 cm2 for
mice) on the backs of the animals from which hair had been clipped the
previous day. Each animal was inspected to make sure the skin had not
been nicked before use. After dosing, a hemispherical dome of wire mesh
(histology tissue capsule) was glued over the dose area with
cyanoacrylate adhesive to serve as a nonocclusive protective appliance.
The oral dose formulation for rats contained 10 to 24 µCi radiolabel
per dose for rats and 9 to 11 µCi per dose for mice, an appropriate
amount of unlabeled DPT in water/Emulphor, 8:2 (v/v) to allow delivery
of the target dose in a total of 5 ml/kg b.wt.
During experiments, rodents were housed in individual glass metabolism
chambers that permitted separate collection of urine, feces, and carbon
dioxide and volatile organic compounds in breath. Radiolabeled
components in breath were trapped as described previously (Mathews et
al., 1991), using 1 N NaOH to trap
14CO2 and ethanol to trap
volatile organic compounds. Animals were anesthetized at the end of
each experiment before sacrifice. Rats were sacrificed by overdosing
with sodium pentobarbital (300 mg/kg intracardially) and mice were
sacrificed by cervical dislocation.
At the conclusion of the dermal studies, the skin at the dose site was
completely excised and thoroughly rinsed with ethanol, then gently
wiped with cotton gauzes soaked with soapy water. The rinsing solutions
and gauzes were collected and analyzed for radiochemical content. The
washed dose site skin was digested in 2 N ethanolic sodium hydroxide.
Radioactivity Analysis.
Radioactivity was determined using a Packard Tricarb 2200CA Liquid
Scintillation Analyzer (Packard Instrument Company, Downers Grove, IL).
Gauzes from the dermal studies and aliquots of urine (0.1-0.2 g),
dermal rinse solutions (0.1-0.2 g) and breath traps (ca. 1 g)
were added directly to vials containing scintillation cocktail (Ultima
Gold, Packard Instrument Company). Samples of tissues (0.1-0.5 g),
feces (0.1-0.2 g), and blood (0.1-0.2 g) were digested in 2 ml of
Soluene-350 at room temperature overnight. After digestion, samples
requiring bleaching were decolorized with 125 µl of 70% perchloric
acid and 300 µl of 30% hydrogen peroxide before addition of 10 to 12 ml of scintillation cocktail (Ultima Gold, Packard Instrument Company).
Characterization of Urinary, Blood, and Volatile Breath
Metabolites from Rats.
Urinary and blood metabolite profiles were obtained using a Zorbax ODS
(Rockland Technologies, Inc., Newport, DE) or a Supelcosil LC-18-DB
analytical column (each 4.6 × 250 mm, 5 µm). Urine samples were
first filtered through a 0.45 µm Millex-HV filter. Blood samples (150 µl) were mixed with 1 ml acetone then centrifuged at 1300g
for 10 min. The supernatant was removed. The pellet was extracted again
with 0.5 ml acetone. The supernatants were combined and evaporated to
dryness on a Savant Speed-Vac (Holbrook, NY), then reconstituted in
HPLC mobile phase. Metabolites were eluted using a linear gradient,
changing from 10 to 90% methanol in water (v/v, each solvent
containing 35 mM tetrabutylammonium hydrogen sulfate) over a 35-min
period. The flow rate was 1 ml/min and the column was maintained at
40°C. Volatile components in breath traps were analyzed directly
using a Zorbax ODS analytical column and an isocratic mobile phase
consisting of 60% methanol in water (v/v). The flow rate was 1 ml/min.
Column effluents were monitored by an Applied Biosystems 757 absorbance
detector at a wavelength of 254 nm and by a Ramona 5-LS flow through
radioactivity detector equipped with a 600-µl solid scintillate flow cell.
Assignment of one urinary metabolite as 4-acetamidophenyl sulfate and
another as phenyl sulfate was made by treatment of urine with sulfatase
(Sigma Chemical Co., St. Louis, MO), followed by demonstration of
coelution of the resulting analytes in the incubation solution with
those of standards of 4-acetamidophenol and phenol, respectively.
Sulfatase incubations included 100 µl of urine combined with
tris(hydroxymethyl)aminomethane (TRIZMA) buffer (pH 7.6, 100 µl) and
sulfatase (100 µl, Sigma; prepared from Aerobacter aerogenes, 18.9 U/ml) for 3 h at 37°C. Four other urinary
metabolites were identified by coelution with urinary metabolites of
benzene (Mathews et al., 1998). The radioactivity eluting in each
fraction was measured by liquid scintillation spectrometry, and the
pmol metabolite/g blood values were calculated using the specific
activity of the radiolabel.
Characterization of Urinary and Volatile Breath Metabolites in
Mouse.
Urinary metabolites were resolved using a Supelcosil LC-18-DB
analytical column (Supelco, Inc., Bellefonte, PA). Metabolites were
eluted using a linear gradient, changing from 10 to 90% methanol in
water (each solvent containing 35 mM tetrabutylammonium hydrogen sulfate) over 35 min. The urinary metabolites were also eluted using a
linear gradient, changing from 98 to 20% 20 mM aqueous ammonium
acetate in methanol over 15 min with the remaining balance of solvent
being methanol. The flow rate was 0.5 ml/min and the column was
maintained at 40°C in both of chromatographic systems. Volatile
components in breath were analyzed using a Zorbax ODS analytical column
and an isocratic mobile phase consisting of 60% methanol in water. The
flow rate was 1 ml/min. Column effluents were monitored by an Applied
Biosystems 757 absorbance detector at a wavelength of 254 nm and by a
Ramona 5-LS flow through radioactivity detector equipped with a
600-µl solid scintillate flow cell.
Assignment of urinary metabolites as 4-acetamidophenyl sulfate,
2-aminophenol sulfate, or 2-acetamidophenol sulfate was made by
treatment of isolated metabolites with sulfatase, followed by
demonstration of coelution of the resulting analytes in the incubation
solution with those of standards of 4-acetamidophenol, 2-aminophenol,
or 2-acetamidophenol, respectively. An aliquot of the isolated
metabolite was incubated with sulfatase (prepared from Aerobacter
aerogenes) in tris(hydroxymethyl)aminomethane buffer for 1 h
at 37°C. Identification of hydroquinone glucuronide and
4-acetamidophenol glucuronide was made by incubation of isolated metabolites with 
-glucuronidase (Sigma Chemical Co.) followed by
coelution of the resulting analyte with hydroquinone and
4-acetamidophenol, respectively. An aliquot of the isolated metabolite
was incubated with -glucuronidase for 1 h at 37°C. One other
urinary metabolite was identified by coelution with urinary metabolites
of benzene (Mathews et al., 1998).
Liver Slices.
Human liver slices were received from the International Institute for
the Advancement of Medicine (Exton, PA). The donor was a 45-year-old
black male that died from a gunshot wound to the leg. He had renal
cancer and used cocaine. The medications received during his hospital
stay were desmopressin, Cefadyl and dopamine. Suitable conversion of
coumarin to 7-hydroxycourmarin and its glucuronide by the batch of
slices was demonstrated by the supplier. Human liver slices were
incubated for 5 h with DPT (1 mM in Krebs-Henseleit buffer, pH
7.4) as described previously (Mathews et al., 1996). One slice,
250 µm in thickness and 8 mm in diameter, was used per incubation.
| |
Results |
Metabolism and Disposition in Rats.
Radioactivity associated with oral doses of DPT (20 mg/kg) to male rats
were mainly excreted in the urine, with 76% of the dose excreted in
the first 24 h post dosing and 81% after 72 h (Table
1). Fecal elimination accounted for 16%
of the dose. Exhalation of radiolabeled volatile organic compounds and
carbon dioxide accounted for only 1.4 and 0.05% of the dose,
respectively. The routes and rates of excretion after i.v.
administration (2 mg/kg) were similar to that of the oral, but less was
excreted in the feces (8% of the dose) and as radiolabeled volatile
organic compounds (0.7%) in breath.
A single radiolabeled product was present as volatile organic chemicals
in the exhaled breath of rats. Analysis by reversed phase HPLC
determined that it coeluted with benzene (data not shown). The profile
of metabolites appearing in urine collected over the 24 h after
oral dosing is shown in Fig. 1. Five of
the major urinary components were metabolites common to those of
benzene, and were identified by coelution with urinary metabolites
identified in a separate study of the metabolism of benzene (Mathews et
al., 1998). These included hydroquinone glucuronide, muconic acid, "prephenylmercapturic acid" (the nonaromatic product of the oxirane ring-opening of benzene oxide with the thiol), phenol glucuronide, and
phenyl sulfate. The major urinary metabolite, however, was 4-acetamidophenyl sulfate, which comprised 32 and 25% of the oral and
i.v. doses, respectively (Table 2).
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 1.
HPLC radiochromatograms of urine collected
0 to 8 h (top) and 8 to 24 h (bottom) after an oral dose of
[14C]DPT (20 mg/kg) to a male F-344 rat.
Reversed phase HPLC was performed as described in Materials and
Methods. See the footnote to Table 2 for the identity of
metabolites A to F.
|
|
The pattern of DPT metabolites recovered in excreta after oral
administration of DPT (20 mg/kg) to rats changed markedly with pretreatment with 1-aminobenzotriazole (ABT), a mechanism-based inhibitor of cytochrome P-450 (CYP) enzymes in rats (Ortiz de Montellano and Mathews, 1981). Urinary excretion of radiolabel during
the first 8 h post dosing was decreased from about 50% in
nonpretreated rats to about 12% in ABT-treated rats (Table 1). In the
24 h post dosing, exhalation of radiolabeled volatile metabolites
increased from 1% of the dose in nonpretreated rats to 12% of the
dose in rats administered ABT. Virtually all of this volatile
radioactivity was present as benzene (data not shown). Concomitant with
the increase in exhalation of benzene was a ca. 60% decrease in the
excretion of its metabolites after treatment with ABT (Table 2).
The timecourse of the appearance of DPT and its metabolites in blood
after oral administration of DPT (20 mg/kg) to male and female rats is
shown in Fig. 2. The parent compound was
quickly metabolized and its levels reached only a fraction of that of its major metabolites. The carcinogens benzene and aniline, presumably precursors of the remaining metabolites, were detected at all time
points in the 6-h experiment. Their major metabolites, phenyl sulfate,
hydroquinone glucuronide (Mathews et al., 1998), and 4-acetamidophenyl
sulfate (Kao et al., 1978; McCarthy et al., 1985) composed the majority
of the circulating equivalents from the earliest time point (15 min) to
the end of the experiment.
Only about 1% of the radioactivity remained in the tissues sampled
(adipose, blood, kidney, liver, muscle, skin, and spleen) from both
male and female rats 24 h after oral administration of
[14C]DPT (data not shown). No tissue developed
marked accumulations of radiolabel, and tissue/blood ratios of
radioactivity were all 1.
Metabolism and Disposition in Mice.
Single oral and i.v. doses of DPT (20 and 2 mg/kg, respectively) to
male B6C3F1 mice were mainly excreted in the urine, with ca. 70% of
the oral dose and ca. 60% of the i.v. dose excreted by that route in
the first 72 h post dosing (Table
3). About 20% of the dose was recovered
in feces; however, a portion of this may have been due to contamination
with urine. Thus, it is probable that this does not indicate a major
difference in the routes of excretion between the species. Only a trace
of the radioactivity was excreted as
14CO2 (oral dose), and 1%
of the radioactivity was recovered as volatile organic chemicals after
each route of administration.
The radiolabel excreted as volatile organic chemicals in the breath of
mice was found to be present as benzene after each route of
administration (data not shown). As was the case in rat, virtually all
of the metabolites identified in the urine of mice were those known to
be produced by the metabolism of benzene (Mathews et al., 1998) and of
aniline (McCarthy et al., 1985) in that species. These included the
benzene metabolites hydroquinone glucuronide (13% of urinary
radioactivity), muconic acid (4%), and phenol (5%), as well as the
metabolites resulting from the N-acetylation and/or
ortho- or para-hydroxylation of aniline (Table
4). Aniline metabolites included
4-acetamidophenol (1%), its glucuronide (24%) and sulfate (5%), and
2-acetamidophenol (5%) and 2-aminophenyl sulfate (2%).
Dermal Absorption.
The excretion of radiolabel after dermal administration of DPT to rats
and mice at 2 and 20 mg/cm2 skin is shown in
Table 5. The rate of absorption was very
slow, with only 6 to 7% of the lower dose level absorbed in 72 h.
Application of a 10-fold higher dose did not result in a proportionally
higher mass of material absorbed and, in fact, only slightly more
material (and as with oral and i.v. administration, absorbed a lower
percentage of dose) was absorbed at the higher dose. The majority of
the dose was excreted in urine. The profile of metabolites was very similar to those found after oral dosing (data not shown).
View this table:
[in this window]
[in a new window]
|
TABLE 5
Disposition of radioactivity after dermal administrationa of
[14C]DPT to male B6C3F1 mice and F344 rats
(N = 4)
|
|
Metabolism by Human Liver Slices.
DPT was not extensively metabolized after a 5-h incubation (1 mM DPT)
with human liver slices. At 5 h about 90% of the radioactivity in
the incubation was associated with the medium, but only 1 to 2% of
that was present as metabolites (data not shown). The predominant metabolite was hydroquinone glucuronide (0.37 ± 0.08% of the
radioactivity in the medium), followed by aniline (0.32 ± 0.13%), phenyl sulfate (0.19 ± 0.07%), and 4-acetamidophenyl
sulfate (0.13 ± 0.03%); less than 0.1% was present as any of
the other metabolites. However, slices from the same batch afforded
high turnover of coumarin in the supplier's assay, and in this
laboratory converted 25 to 30% of incubated 
-methylstyrene to
metabolic products (data not shown).
| |
Discussion |
Oral doses of DPT were well absorbed, readily metabolized, and
excreted predominantly in the urine; in contrast, dermal doses of DPT
were poorly absorbed. All of the metabolites present in urine after
oral dosing were those common to the metabolism of benzene (Mathews et
al., 1998), which was detected as the sole radiolabeled component in
breath, or aniline (Kao et al., 1978; McCarthy et al., 1985),
suggesting that DPT is metabolized initially by scission of the
triazeno linkage to yield those intermediate products. The nature and
rank order of the amounts of metabolites produced in rat and mouse also
support this sequence of reductive cleavage of DPT, followed by
oxidation and conjugation by the two species. Whereas 90% of the
hydroxylation of aniline in rat occurs in the para-position,
the ratio of para- to ortho-hydroxylation is
about 2 in mouse (McCarthy et al., 1985), consistent with the findings
in the present work. However, the degree of subsequent conjugation of
benzene metabolites in mouse was lesser in animals treated with DPT
than with benzene, due possibly to competition for phase 2 metabolism
with aniline-derived metabolites.
In separate experiments using electron spin resonance, we have
demonstrated that phenyl radicals are formed in incubations of DPT with
either CYP reductase and its cofactor, NADPH, or anaerobically with
cecal contents (Mathews, 1998; Mathews et al., 1999). Taken together,
the data suggest that DPT is metabolized by the pathway proposed in
Fig. 3. In this scheme, DPT is cleaved by
reduction to initially form phenyl diazenyl radical and aniline. Phenyl diazenyl radical is known to decompose to form benzene and nitrogen gas
(Galli, 1988). Reductions can be catalyzed by gut microflora (Scheline,
1973) and by mammalian P-450 and/or its reductase (Hernandez et al.,
1967a,b; Fujita and Peisach, 1977). Pretreatment with ABT
resulted in a large increase in the amount of benzene exhaled unchanged, indicating that P-450 itself was not catalyzing the cleavage
of DPT. The presence of significant amounts of radiolabel in feces
after i.v. administration suggests that DPT and/or its metabolites are
excreted in the bile, and that gut microflora and/or P-450 reductase
could have access to the triazene after either oral or i.v.
administration. Formation of benzene and aniline metabolites in human
liver slices indicates that cleavage of DPT can occur in the liver.
Taken together, the data did not allow determination of the relative
contributions of these two reductive pathways in rats, but the presence
of a majority of the metabolites as triazene cleavage products in the
blood minutes after oral administration indicates that this process was
very fast.
This work presents the first comprehensive report of the metabolism of
the linear 1,3-triazeno functional group. If the proposed mechanism is
common to triazenes, it may be expected that their metabolism would
yield as intermediates a radical and an amine. Other 1-aryltriazenes
have been studied as potential antitumor drugs and one, dacarbazine, is
of therapeutic value against several tumors, particularly malignant
melanoma (Carvalho et al., 1998). These agents are thought to act as
alkylating agents that methylate DNA (Mizuno et al., 1975), and it has
been postulated that diazomethane is the proximate alkylating agent.
The work presented here suggests an alternate mechanism by which methyl
diazenyl radicals may be formed by reduction of these triazenes in the
hypoxic environment of these tumors to effect methylation of DNA bases.
Similarly, the triazo functionality of the anti-HIV drug zidovudine
(AZT) is also metabolized by reduction, converting the azido group to an amine in a reaction that is at least partially catalyzed by CYP
reductase (Rajaonarison et al., 1993; Eagling et al., 1994; Veal and
Back, 1995).
In the present case the metabolism of DPT yielded the known carcinogens
benzene (Snyder and Kalf, 1994) and aniline (National Cancer Institute,
1978). Additionally, aniline produces methemoglobinemia (Kiese, 1966),
consistent with that toxic effect observed with DPT. The production of
metabolites of aniline and benzene by human liver slices, as well as by
rats and mice, indicates a toxic/carcinogenic potential for DPT in humans.
This work was performed under National Institute of
Environmental Health Sciences Contract No. N01-ES-75407.
Abbreviations used are:
DPT, 1,3-diphenyl-1-triazene;
CYP, cytochrome P-450;
F-344, Fischer 344;
ABT, 1-aminobenzotriazole.
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Top
Abstract
Introduction
), aniline (
), and benzene (
). b, aniline
metabolite 4-acetamidophenyl sulfate (
), and benzene metabolites
hydroquinone glucuronide (
) and phenyl sulfate (
).
