Download Acetaminophen Metabolism in Man, as Determined by High

CLIN. CHEM.
20/8. 1086-1096
(1974)
Acetaminophen Metabolism in Man, as Determined
by High-Resolution Liquid Chromatography1
J. E. Mrochek, S. Katz, W. H. Christie, and S. R. Dinsmore
Acetaminophen
is a commonly
used
analgesic,
avail-
able without prescription.
Several of its metabolites
have heretofore
been isolated from physiologic fluids
and analytically characterized.
In general, the separation methods are complicated,
usually requiring extensive sample pretreatment,
and do not measure the individual conjugated
metabolites.
High-resolution
anionexchange separation of urinary samples from subjects
receiving acetaminophen
reveals eight chromatograph-
ic peaks, representing seven metabolites and the free
drug itself. Metabolites separated include 2-methoxyacetaminophen, its glucuronide and sulfate conjugates,
the sulfate conjugate of 2-hydroxyacetaminophen, the
glucuronide and sulfate conjugates of acetaminophen,
S-(5-acetamido-2-hydroxyphenyl)cysteine,
and
S-(5-
acetamido-2-glucuronosidophenyl)cysteine.
Urinary and
serum concentrations of the drug and its seven metabolites were determined by high-resolution liquid chromatography as a function of time after two clinically normal
men ingested 1950 mg of the drug. Concentrations in
urine and serum are compared, and estimated urinary
excretion rates are reported for all metabolites except
S-(5-acetamido-2-hydroxyphenyl)cysteine.
Serum concentrations of the glucuronide were higher than concentrations of the free drug 2 h after the drug was ingested,
indicating
that solvent-extraction
procedures
for serum
will yield low estimates of total drug unless hydrolysis
precedes the extraction step.
AdditIonal
Keyphrases:
and metabolites
chromatography
in serum
detoxication
and urine
#{149}mass
spectrometry
cysteine
amido-2-glucuronosidophenyl)
#{149}drug
mechanisms
#{149}gas
#{149} toxicology
#{149}S-(
5-acet-
concentrations
A number
bolic
products
of drugs and the kinetand elimination
coupled
with clinical observations
of pharmacologic
activity
can be invaluable
aids to the clinical pharmacologist.
Complete
studies
including identificationand monitoring of metabolites
can be highly informative
regarding drug activity,
toxicity,
and mechanism
of ac-
tion. From
these
data,
new and more effective
drug
structuresmay be formulated. High-resolution liquid
chromatography isan idealanalyticaltool for studies
such as these because both unconjugated and conjugated metabolites
can be separated
and individual
‘Research
cal Sciences
Corporation’s
2Operated
sponsored
by the National
Institute
of General Medi.
and National
Cancer Institute
under Union Carbide
can be
for
Metabolites
of investigators
have studied the meta(AAP) in humans.
included the sulfate and glu-
of acetaminophen3
identified
ters were computed.
Prescott
et al. (7) reported
plas-
ma concentrations
of AAP as a function
of time and
total percent
excreted
(free + conjugated
AAP) in
24-h urine samples. Each of these studies
plete in one or more ways; sample treatment
is incom-
and separation
difficulties
precluded
analyses
of both serum
and urine, and few data are presented
on additional
metabolites
that are now known to be produced
in
humans (2).
Metabolism
of the homologous
drug acetophenetidin (phenacetin),
which is enzymatically
de-ethylated to acetaminophen
in man, has been more thoroughly
studied.
Metabolic
products
identified
include the glucuronide
and sulfate conjugates
of AAP,
2-hydroxyphenetidine,
2-hydroxyphe-
nacetin, S- (1-acetamido-4-hydroxyphenyl)cysteine,
and 3-methoxy-4-hydroxyacetanilide and itsglucuronide (2, 7-11). Considering the similarityin the two
drugs and the metabolic conversion of acetophenetidin to AAP, it isnot unreasonable to anticipatethat
the metabolites of AAP would be similar to those
found foracetophenetidin.
Using high-resolution anion-exchange chromatography in the study presented
here, we report urinary
excretion
rates for free AAP and seven metabolites,
and serum concentrations
of free AAP and three metabolites.
Urinary
excretion
was determined
periodically over the 24-h period after ingestion
of a total of
1950 mg of AAP by each of two clinically
normal
contract with the U. S. Atomic Energy Commission.
for the USAEC by the Union Carbide Corp.
Oak Ridge National
Laboratory,2
Received April 12, 1974; accepted
1058
metabolites
curonide
conjugates
of AAP (1) and S-(1-acetamido4-hydroxyphenyl)-cysteine
(2). Analytical
studies of
acetaminophen metabolism have generally focused
on the 24-h urinary
excretion
of free AAP and total
conjugated
AAP, the latter determined
after enzymatic or chemical
hydrolysis
(3-6). In one of the few
studies reporting analyticalresultsfor severalmetabolites,Cummings
et al. (1) listedurinary excretion
rates forthe freedrug and itsglucuronide and sulfate
conjugates; from these data, certain kinetic parame-
4-phenetidine,
Studies of the metabolism
ics of metabolite
formation
of all detectable
determined
in a single sample
without the need
hydrolysis or extensive sample pretreatment.
CLINICAL
CHEMISTRY
Oak Ridge, Tenn.
June 3, 1974.
VoL 2O No. 5 1974
37830.
‘ Trade
name, Tylenol;
also known
aminophenol,
and 4-hydroxyacetanilide.
as paracetamol,
N.acetyl-p-
men. Concentrations
in serum were also determined
periodically
during a 3-h period after drug ingestion.
Several new metabolites
of AAP in humans are reported. Metabolites
were identified
by gas chromatography, mass spectrometry,
and the two techniques
combined. Details of the identification
of one important new metabolite
of acetaminophen
are presented
here; complete details of studies on other new metabolites,including gas-chromatographic and mass-spectrometric data, will be presented in a subsequent
paper.
Materials and Methods
Chromatographic
System
The chromatographic
system used for these studies
Comparative
ultraviolet
spectrophotometric
data
were used to determine positional substitution
on the
aromatic ring where model compounds
were available. Ultraviolet
spectral data (maxima)
were reported for synthetic
S-(5-acetamido-2-hydroxyphenyl)cysteine in 0.1 mol/liter HC1 and 0.1 mol/liter
KOH by Focella et a!. (compound 18, ref. 18).
Drug-Ingestion
Two
Studies
clinically
normal
men
(both
nonsmokers;
Subject 1 age 41 and Subject 2 age 58 y) each ingest-
ed six tablets of AAP (Tylenol, McNeil, 325 mg/tablet) after overnight fasting (12 h). The total of 1950
mg ingested was equivalent
to 23 mg/kg of body
weight for each subject (Subject 1, 185 lb; Subject 2,
187 ib). The amount of drug ingested is three times
isgenerally similarto that described previously (12).
A more complete descriptionof the system used here
can be found in a laterpublication by Katz et al.(13).
Certain modifications were introduced into the sys-
the recommended
tem, which enabled the chromatographic
analysis of
acetaminophen
and related metabolites
to be completed in 21 h. Thus, including system regeneration,
the total cycle time was 24 h/sample. These modifications included the following: (a) reduction in column
length from 150 to 100 cm, (b) increasingthe column
temperature
from 60 to 70 #{176}C,
and (c) decreasing the
total amount of eluting buffer from 350 g to 225 g
while at the same time sharpening
the gradient. The
concentration
gradient of buffer eluent was formed
by using a nine-chamber
gradient box (Phoenix Precision Instrument
Co., Philadelphia,
Pa.) with 25 g of
pH 4.4 ammonium
acetate-acetic
acid buffer per
chamber (15 mmol/liter
in the first two chambers, 4
at 0.5,1, 2,and 3 h afterwards. The subjects concluded the fast after the 3-h blood and urine samples
mol/literin the second two chambers, and 6 mol/liter
in the finalfivechambers). This relativelysteep gra-
dient and higher temperature
enabled
elution of
three highly anionic sulfate conjugates in 21 h, allowing one sample per day to be analyzed. With use of
the previous gradient (12) and a 150-cm column at 60
#{176}C,
these strongly retained compounds required 42 h
to elute. Extensive band-broadening resulted from
such long retention on the column, making quantitation difficult.
The measured molar absorptivity of AAP at 254
nm (7950 litersmole’ cm’
was used in allquantifi-
cations,
and the metabolite
excretion was computed
in terms
of equivalent
AAP. Ultraviolet
spectra of
the individual metabolites were similar to that of
AAP, at leastwith respect to the main transitionat
244 nm. Thus, our use of a singlemolar absorptivity
in the calculationsisexpected to yield reasonably accurate results.
Identification
of Metabolites
Metabolites
of acetaminophen
fied by multiple
analytical
chromatography
or mass
urinary
products
isolated
have been identi-
techniques
(including
gas
spectrometry or both) on
by means
of preparative-
scale high-resolutionliquid chromatography
(14-17).
single dosage but within the rec-
ommended 24-h dosage, 1950 to 2600 mg (two tablets
taken three or four times daily). Blood samples were
collected immediately
before they took the drug and
were collected, consuming
a light lunch. Total urinary output was collected during each of the following time periods after ingesting the drug: 0-0.5,
0.5-1, 1-2, 2-3, 3-5, 5-7, 7-12, and 12-24 h. Urine
samples were filteredthrough a 0.2-zm (average pore
diameter)
filter (Nalge, Sybron Corp., Rochester,
N.
V. 14602) to remove particulatematter and then frozen at -60 #{176}C
untilchromatographic analysis.Blood
samples were allowed to clotat ambient temperature.
The clotwas centrifuged and the serum was ultrafiltered overnight (0 #{176}C)
through dialysistubing (Union
Carbide Corp., Food Products Division) with an air
overpressure of 104 kPa (15 psig).The ultrafiltered
serum samples were stored at -60 #{176}C
until analyzed
by liquidchromatography.
Results
Metabolite
Identification
Figure 1 depicts a typical urinary chromatogram
obtained for a patient receivingthe analgesicmedication, AAP. The anion-exchange system used to obtain this chromatogram employed a more gentle gradient (12) and did not incorporate
the changes men-
tioned above (Materials and Methods); therefore, the
time required for complete development
of this illustrativechromatogram was 42 h instead of the 21 h required for the data of our study. We have identified
seven metabolites of acetaminophen in urine samples
from human subjects receivingthe drug as shown in
Figure 1:
(I)S-(5-acetamido-2-hydroxyphenyl)cysteine
(IV) 2-methoxy-4-glucuronosidoacetanilide
(V) 4-glucuronosidoacetanilide
(VI) S- (5-acetamido-2-glucuronosidophenyl)cys-
teine
(VII) 2-methoxyacetaminophen
(VIII)acetaminophen
(IX) 2-hydroxyacetaminophen
CLINICAL
sulfate
sulfate
CHEMISTRY.
sulfate
Vol. 20. No. 8. 1974
1087
0
2
I
3
4
5
6
7
6
9
10 II
12
3
4
IS
C
IT
C
19 20
SI
22 23 24 25 25 27 26 29 30
rIME
It.,)
ANION
02
ce
31 32 33 34 35 31 37 31 39 40 41 42 43 44 s
CNCIIANGE
CIOMATOONAPW
11101 CONIOTIONS
.150cm
STAINLESS
STEEL
WITH
11-12,. DISK AMINEX 0-27 11(519.
TEMPENATIIWE. AMSIENT TO t.OC
AT 3 ELIlEIT
GNAOIENT, 0015 E
TO NOW AWNONIUN ACETATE. 111144.
ELUENT
TLOW NATE.
7ml9c.
COLUMIUPU(SSUNE
Fig. 1. Anion exchange chromatogram obtained with use of the two-detector
who had received acetaminophen
Shaded peaks are acetaminophen
100
r
60
N#p
H11C-II
Lc.I
>
I)
C-S.)
209
101
I-
_______
50
70
00
urine sample
from a subject
110
i
II’
io#{149}io tho
Fig. 2. Mass spectrum for a probe-inserted
acetamido-2-hvdroxvohenvl)cvsteine
210
230
sample
of S-(5-
Note that the conjugated metabolites are easily
20, No.8,
elsewhere.
Ultraviolet
spectra for compounds
I and VI
methanol were virtually identical, with maxima
the former observed at 244 and 299.5 nm and for
latter at 244 and 300.5 nm. Figure 2 illustrates
in
for
the
the
mass spectrum obtained for a probe-inserted sample
separated
from the unconjugated
metabolites,
thus
illustrating
a major advantage
for liquid-chromatographic analysis of physiologic body fluids.
The
conjugate
S-(5-acetamido-2-glucuronosidophenyl)cysteine
is an important
new metabolite
of
acetaminophen,
and identification
studies for it will
be discussed with reference to our data for S-(5-acetamido-2-hydroxypheny!)cysteine,
partially
characterized by Jagenburg
and Toczko (2) and completely
identified by means of the total synthesis of I by FoCLINICAL CHEMISTRY, Vol.
celia et a!. (18). Ultraviolet
absorption
maxima obtained for I in 0.1 mol/liter
HC1 and 0.1 mol/liter
NaOH agreed with those reported for the synthetic
compound in Table 2 of this reference. Identification
data on other metabo!ites,
including mass spectral
and gas-chromatographic
retention data, will be presented
U/E
1088
system for a typical
and its metaboiltes
>-
30
.27001111.5
1974
of I. Empirical formulas for all ions were confirmed
by high-resolution
mass spectrometry
except for m/e
226 (not observed in the high-resolution
spectrum).
However, a metastable
transition
observed in the
mass spectrum at m/e 171.7 (m*)4 related the ions
m/e 226 and 197 and suggested a loss of methylene
imine (CH2=NH).
Metabolite I had very low volatility in the mass spectrometer,
requiring high temperature (>200 #{176}C)
for a useful spectrum. No masses were
observed above those indicated in Figure 2. The identity of I was confirmed by the mass spectrum of the
trimethylsi!yl
derivative,
as shown in Figure 3. EmParent
m4
=
d2/p.
(p) and daughter
(d) ions are related
by the expression
CTLI1iO&l
33
110
13)
17)
170
133210
23)
27)270217
310
1077370
A4-
311
a)
4)0
43)
Fig. 3. Mass spectrum for the trimethylsilyl
acetamido-2-hydroxyphenyl)cysteine
47
470
o F
l7)5)0S33’77570
derivative
2O
pirical formulas for all ions pictured in Figure 3 were
confirmed by high-resolution
mass spectrometry
(see
Table 1). The molecular weight of 270 for I [558-288
(four trimethylsilyl
groups)] coupled with the presumed identity of the m/e 226 ion in Figure 2 (based
on the metastable
transition
in the low-resolution
mass spectrum) suggests partial thermal degradation
and loss of CO2 before volatilization
of the probe
sample. This isa reasonable assumption, considering
the probe temperature
required to volatilize the compound and its chemical structure.
Conjugated
compound
VI was most intractable
and difficult to study. A probe sample required even
higher temperatures
than were required for I, to obtain a mass spectrum
of reasonable
intensity.
The
spectrum was identical to that observed for I, except
that no m/e 226 ion was observed. High-resolution
spectrometry
verified
that
i,ick
17
170
I
1
170
1
210
77)
2T
270
27)
310 37)
7)
370
110 033
07)
070
410
47
QT
011)
the previously
7)
ii
110
iLL
rIL
13)
210
I Ih
27)
07)
ALl
270
47
1
1 1
310 77
L
77 370 77
411
Fig. 4. (a) Mass spectrum
of S-(5-acetamldo-2-glucuronoskloafter silylation
Underlined masses are characterIstic of the trimethylsilyl-derlvatlzed
giuclsophenyl)cysteine
nidemoiety
(b)
Mass
spectrum
of
S-(5-acetamido-2-glucuronosidoph#{232}nyl)cysteineafter derivatiz#{226}tionwith diazomethane
and
bis(trimethylsilyl)trifluoroacetamide
Interpretation of the data suggested the ions Illustrated above
and then silylated
with bis(trimethylsilyl)trifluoroa-
cetamide,
that
in hopes
esterification
of the carboxy!-
ic acid moieties with methyl groups would result in a
compound
that would volatilize in the mass spectrometer
without
thermal
fragmentation.
A comparifor m/e <500 is shown in
Figure 4. Previously
published
data have established
characteristic
masses for the glucuronic
acid moiety
of conjugated
urinary
constituents
(17). These masses are present
in the spectrum
of the trimethylsilyl
son of the two mass spectra
iden-
inlet)was quite complex and impossible to interpret.
was methylated
I
,
13D
(b)
Li
70
tified ions (Figure 2) were present. A mass spectrum
of the trimethylsilyl
derivative (gas chromatographic
The compound
110
of S.(50:117
mass
JI
70
with diazomethane
Table 1. High-Resolution Mass Spectral Data for the Trimethylsilyl Derivative of
S-(5-Acetamido-2-hyd roxyphenyl)-cysteine
Differences
from theoretical,
Measured
Empirical
mass
558.2254
543.1998
515.2085
486.1890
441.1919
341.1337
218.1050
formula
Suggested
C23H4N204SSi4#{176}
C22H43N2O4SSi4
C21H43N2O3SSi44
CH38N204SSi3
C19H27N2O2SSi3
C15H27N02SS12
+
+
CH,=Si(CH11),
COOTMSS
fragment
losses
millimass
1
+0.1
-2.3
I
+1.4
+3.0
+3.5
+3.6
I
TMSN=CHCOOTM5
C8H25N02S12
units
+1.7
CH
o=
N-TMS
“S-CH-Cl-I-COOTMS
OTMS
b Metastable
HNTMS
transition observed.
CLINICAL
CHEMISTRY,
Vol. 20, No.8.
1974
1089
derivative, as shown in Figure 4a. Methylation
of the
carboxylic
acid groups by means of diazomethane
would decrease those masses characteristic
of glucuronic acid by 58 mass units yielding m/e 407 and m/e
317 in place of the characteristic
glucuronide
ions
m/e 465 and 375, respectively
(19). As illustrated
in
Figure 4b, these were observed and the compound
was verified as a glucuronide.
Highest mass observed
for the trimethylsilyl-derivatized
conjugate was 691
and for the methylated
trimethylsilyl
derivative 681,
neither of which is nearly high enough for a fully derivatized glucuronide of compound I (mol wt, 950; mol
wt of comp. VI, 446 + 7 trimethylsilyl
groups). However, we did observe masses that we interpreted
as
verifying the aglycon as I (see Figure 4b). Unfortunately, the conjugate was resistant to hydrolysis by
fi-glucuronidase
(72-h
treatment
at 37 #{176}C),
and
acid
hydrolysis probably would have cleaved the sulfurcontaining side chain we were attempting
to identify.
Both I and VI showed a positive reaction with ninhydrin, indicating the presence of a free primary amine
functional group. From the nearly identical ultraviolet spectra for I and the glucuronide,
the mass spectra! data obtained
for the methylated
and tn-
methylsilylated conjugate, and its positive reaction
with ninhydnin,
we believe the metabolite
is S-(5acetamido-2-glucuronosidophenyl)cysteine.
The ultraviolet spectral data indicate that the position of
aromatic substitution
for the cysteine moiety is identical for compounds land VI.
Metabolite
inUrine
Excretion
The urinary
excretion
of the previously
listed
seven metabolites
of AAP was followed as a function
of time after two clinically normal subjects ingested
23 mg of AAP per kg of body weight (‘1950
mg).
This was the first time either subject had ingested
this drug. Figure 5 graphically illustrates the chromatographic
changes
in urinary
period. Note that excretion rates and cumulative
excretion of AAP and its glucuronide,
were initially
much higher for Subject 2 than for Subject 1. Simi-
TINIE. 11
CLINICAL CHEMISTRY. Vol. 20, No.8, 1974
concentra-
metabolites for both subjects during each collection
Fig. 5. Liquid-chromatographic
analyses
of urine samples
from
Subject
1, illustrating
concentration
changes undergone by acetaminophen
and
its metabolites
1090
metabolite
tions for Subject 1 as a function of time. Figures 6
and 7 compare the urinary excretion rates (in milligrams per hour) for AAP and six of its metabolites
(all computed as equivalent AAP) for Subjects 1 and
2, respectively;
Table 2 lists the total excretion of all
SUBJECT I
o ACETAMINOPHEN (AAP)
()
#{149}
4.GLUIJB0NOSI55ETANILIDE
ACETAMINOPHEN SULFATE()
#{149}
2-METHOXY 4-GLUCURONOSIDO*
tNt
A
o
2-METHOXV
ACETAMINOPHEN
SULFATE
O 2-HTOROXY ACETANINOPHEN SULFATE
#{149}
S(IACET
CYSTEINE I
‘C
4..
E
10
E
01
LI
I-
I-
4
4
11
z
z
0
0
I-
Iii
LII
C)
5)
5<
III
LI
x
5.
z
4
z
8
10
TIME, Is
TIME,),
Fig. 6. Rates of urinary excretion
metabolites after its oral ingestion
for acetaminophen
by Subject 1
and six
larly, peak excretion by Subject 2 occurred at 2.5 h
(collection period 1 h, the 2nd to 3rd hour), while it
was delayed until 4 h (collection period 2 h, the 3rd to
5th hour) for Subject 1.
An examination
of the excretion rates of AAP during the first half-hour
sampling
period indicated
Subject 1 excreted the free drug at a nate of 1.4 mg/h,
whereas the rate for Subject 2 averaged 8 mg/h. Similar differences were noted in a comparison of the excretion rates for the glucuronide
during this initial
time period,
with
2.4 mg/h
observed
for Subject
1 and
16.3 mg/h for Subject 2. However, during this same
period,
excretion
rates for the sulfate conjugate
(VIII)
were approximately
the same for both
subjects-12.4
mg/h for Subject 1 and 12.9 mg/h for
Subject
2. Complicating
the interpretation
effects of differences
data were possible
of the rate
in the rate of
Fig. 7. Rates of urinary excretion for acetamiriophen
metabolites after Its oral ingestion by Subject 2
drug absorption
and average
hourly
urine
and six
flow for
each subject. Urinary flow for Subject 1 averaged
only 35 ml/h, as compared with 100 ml/h for Subject
2.
Peak overlap in the chromatographic
S-(5-acetamido-2-hydroxyphenyl)cysteine
area where
eluted
(see Figure 5, 2.75 h) prevented accurate estimations
of its excretion
measurements
rate.
Crude
estimations
based
on
of the fluorescence of CeS+ suggested
that its excretion tended to follow that of compound
VI; however, this measurement
was complicated
by
the co-elution
of tryptophan
with I, resulting in a
non-gaussian peak shape and additional fluorescence
caused by Ce4 oxidation
of this amino acid. Collecting this chromatographic
peak and rechromatograph-
ing it on a high-resolution
(‘A INI(’AI
CHIMISTRY
cation-exchange
VnI
2fl
system
Nn R 1974
ini
Table 2. Comparison of Total Urinary Excretion Data for Acetaminophen and Its
Metabolites in Two Clinically Normal Male Adults
Amount
0.9
5.0
Vc
1.5
VIII
7.8
10.3
2
1.2
3.7
1
2.0
2
4.3
1
2.8
81.9
55.2
2
5.2
157.8
66.0
3.0-5.0
1
2
7.9
4.6
257.0
182.2
110.8
78.6
5.0-7.0
1
2
6.3
3.5
222.8
131.1
1
2.9
116.4
7.4
322.
1.4
3.5
428.8
157.6
ingestion (h)b
Subject
0-0.5
1
2
0.5-1.0
1.0-2.0
2.0-3.0
7.0-12.0
1
2
12.0-24.0
1
2
AAP
excreted,
mg11
IV
VII
N.D.
0.2
12.5
56.7
8.1
16.6
26.2
0.9
0.8
0.8
1.5
3.4
50.8
36.8
2.6
7.0
108.8
55.0
83.7
68.8
IX
VI
(mg)
0.02
0.05
10.7
25.9
0.5
34.4
1.3
1.8
2.0
1.4
1.5
0.4
94.0
4.8
3.9
3.4
3.5
4.2
1.1
100.8
5.2
1.8
184.5
9.5
0.6
5.1
7.9
10.8
7.8
6.3
3.7
12.3
4.0
5.3
4.4
9.3
5.5
7.4
3.7
19.7
5.3
164.0
248.7
427.8
288.2
8.4
12.8
21.9
14.8
11.2
11.5
7.4
17.5
360.5
219.0
18.5
11.2
196.8
531.2
756.9
280.2
10.1
27.2
38.8
14.4
4.8
3.2
3.6
4.2
53.8
167.1
208.1
102.4
7.0
7.9
29.0
4.6
2.9
6.3
31.5
1.7
3.9
10.8
17.4
4.5
9.9
8.9
40.6
6.1
Total, mg
1
2
25.6
37.1
1171.8
1127.4
572.8
572.2
66.6
44.3
70.0
25.1
48.5
35.3
96.8
30.4
2052.1
1871.8
Percent”
1
1.3
1.9
60.1
57.8
29.4
29.3
3.4
2.3
3.6
1.3
2.5
1.8
5.0
1.6
105.2
96.0
2
Percent”
0.3
0.7
All metabolites computed as equivalent acetaminophen.
Each subject ingested 1950 mg of acetaminophen,
equivalent to 23 mg/kg.
AAP = acetaminophen;
V = 4-glucuronosidoacetanilide;
VIII = acetaminophen
sulfate; IV = 2-methoxy-4-glucuronosidoacetanilide; VII = 2-methoxyacetaminophen
sulfate; IX = 2-hydroxyacetaminophen
sulfate; VI
S.(5-acetamido.2.glucuronosidophenyl) cysteine.
d Based
on 1950mg ofacetaminopheningested.
#{149}
Not detected.
(17) would enable accurate
quantitation
of this compound to be accomplished;
however, rechromatography of the fraction containing these compounds
was
not performed for this study.
It is interesting to note that excretion of the conjugate VI by Subject 1 was threefold
higher than was
observed for Subject 2 (see Table 2). This difference
in excretion by the two subjects was also qualitatively
observed
for I by following the Ce3+ detector
response in the two sample series.
Concentrations
of AAP and Its
Metabolites
in Serum
Figure 8 shows that neither I nor VI was detected
(absence of detector response at 2.75 h and 6.75 h) in
any of the serum samples from Subject 1, and this
was alsotrue for Subject 2. With the cerium oxidimetric detector the lower limit of detectabilityfor the
two cysteine-containing
metabolites was estimated to
be about 50 nanograms
per ml of serum. Thus, although we did detect traces of VII in the serum of
both subjects (see, for example, 1- and 2-h samples in
Figure 8 at about 16 h), neither I nor VI was detected
even though the urinary concentration
of the latter
generally exceeded that of VII (see Table 2). It is
possible that these two metabolites were bound
to
Blood samples (about 5 ml each) were collected before ingesting
AAP and 0.5, 1.0, 2.0, and 3.0 h afterwards. Figure 8 illustrates the concentration
changes
we observed for AAP and its metabolites
in serum
macromolecules
in the serum. Compounds
with molecular weights greter
than about 1000 would have
been removed from the samples by the ultrafiltration
step.
samples from Subject
1, by use of the cerate oxidimetric detector.
Generally,
this detector
was more sensi-
Figure 9 illustratescomparisons of urine and
serum concentrations of AAP and its two major metabolites,V and VIII, for Subjects 1 and 2. There is a
striking
similarity,
for both subjects,in the concen-
tive than ultraviolet
detection for the two cysteinecontaining
metabolites
by a factor of two to three
under our instrumental
conditions. It was also more
sensitive to the methoxy-substituted
metabolites
IV
and VII; however, it exhibited about the same sensitivity as ultraviolet
detection
for the other compounds.
102
CLINICAL
CHEMISTRY.
Vol.
20. No. 8.
1974
tration
changes
undergone
by AAP
in urine
and
serum. As reported in the discussionof urinary excre-
tion, a relatively high urinary excretion of AAP was
observed for Subject 2 (compared to Subject 1) in the
firstsample taken at 0.5 h. As illustratedin Figure 9,
URIC
ACID
NO
SO
70
NUVIC
ACID
NO
DO
ON
40
SO
DO
10
0
Do
40
I/I H
30
00
lO
Fig. 8. Cerium oxidimetric detector response for serum
samples from Subject 1
0
SO
The identity
of the compound eluting at
12 h In the 2-h sample is not known
40
30
20
0
10
0
SO
40
30
20
I0
0
SO
40
So
20
0
U
U
I
2
3
4
5
3
7
3
3
0
TIME.
II
2
3
4
IX
IN
Il
IS
20
II
N
a twofold-higher
concentration
of AAP in the urine of
Subject 2 was accompanied
by an equivalent
difference in the concentration
in serum (see also Table 3).
Concentrations
of AAP in serum remained fairly constant over the 3-h period for Subject 1, which does
not appear to fit known kinetics of drug elimination
at normal dose levels. Concentrations
in serum for
Subject 2 decreased
substantially
from the initial
highest concentration
of 12.4 ig/ml (Table 3). However, the concentrations
in the urine of both subjects
seem to more or lessmirror these relativechanges in
concentrations
in serum (Figure 9). The later peaking
of metabolite excretion rates in the urine (3 to 5 h period) for Subject 1 compared to that for Subject 2 (2
to 3 h period) seems to correlate with the difference
in the rates of elimination
for AAP from the sera of
the two subjects.
Note that appreciable
quantities
of conjugate
V.
even exceeding those of the free drug, were present in
the serum. Thus, it should be recognized that gas-liqi
II.JIrAI
(‘WIAISTRV
\inI
‘fl
Jn
R
1974
1002
Table 3. Comparisons of Serum Concentrations
of Acetaminophen and Its Metabolites in Two
Clinically Normal Men
5000
2000
Serum
Time
bOO
after
Ingestion”
0.5
drug
(h)
1.0
200
2
2
100
2.0
50
C)
2
3.0
0
U
,,,
2
20
II
H.
-----.
10.0
>...
Each subject
ingested
lent to 23 mg/kg.
5AAP = acetaminophen;
-
0
5.0
VIII
acetaminophen
phen sulfate.
#{149}
Not detectable.
/
IA
2.0
=
VS
VIII),
0.5
N.D.c
2
1.0
12.4
4.3
6.6
8.6
7.0
5.4
7.0
5.5
3.2
9.1
0.9
1.0
0.2
1
2
1
2
1
2
AAPS
1
E
0
,g/mI
6.2
Subject
500
concentration,
7.2
13.4
9.4
16.6
V
=
0.1
1.2
0.04
2.4
2.1
1.2
1.9
1.6
0.2
1950 mg of acetaminophen,
sulfate;
VIII
N.D.’
equiva-
4-glucuronosidoacetanilide;
VII = 2.methoxyacetamino-
1.0
0.5
A--
IUJECT
I
2
DUBJECT
URINARY
CONCENTRATION
SERUM CONCENTRATION
02
#{149}
----.---
0.1
4
0
0.5
1.0
1.5
2.0
2.5
3.0
4.0
TIME,
Fig. 9. Comparisons
for urinary and serum concentrations
of
acetaminophen
(AAP) and its glucuronide (t) and sulfate (VIII)
conjugates
uid chromatography in which extraction methods are
used (20, 21) for determination
of AAP in blood are
probably measuring only about half of the drug actu-
ally present. Sample hydrolysis performed before the
extraction
step may overcome this problem if additional complications
are not introduced
by such
harsh treatment
of the serum.
Discussion
Significance
of Compounds
/and
VI
Studies aimed at defining the mechanism by which
overdoses of AAP cause hepatic necrosis in rats and
mice, and the possible role that glutathione
may play
in preventing this liver damage, have been performed
by Mitchell et al. (22-25).
These workers found
markedly enhanced hepatic necrosis in AAP-loading
experiments
on rats and mice pretreated
with pheno-
barbital,
which stimulated the disappearance
of AAP
from tissue. In contrast, pretreatment
with piperonyl
butoxide,
tabolizing
a known inhibitor
of microsomal
drug-meenzymes,
decreased
the metabolism
of
AAP, delayed its disappearance
from tissues, and
dramatically
protected
against hepatic damage (22).
Using [3H]acetaminophen,
Jollow et a!. (23) reported
that amounts of covalently bound, radiolabeled
hepatic material
paralleled the severity of the histologically recognizable
necrosis, with maximal binding
preceding liver damage by at least 1 to 2 h. Based on
the data, these workers concluded that hepatic damIOQA
(I IMIfAI
(I.1IAIQTQV
I/,,I
Oil
0
107A
age is mediated by a toxic metabolite
of AAP rather
than by the drug itself.
Probably influenced by numerous reports of glutathione conjugation
with foreign aromatic or unsaturated compounds
and subsequent
excretion of these
detoxified
metabolites
as mercapturic
acids in animals (26-29), Mitchell et al. (25) examined the possibility that glutathione
may prevent AAP-induced
hepatic necrosis. They found that either glutathione
or
cysteine completely inhibited the binding of radiolabeled AAP to microsomal
protein
in vitro. Pretreating mice so as to artificially
deplete hepatic gluta-
thione
enhanced
AAP-induced
hepatic
necrosis,
whereas pretreatment
with cysteine, a glutathione
precursor,
prevented
the hepatic necrosis. Finally,
administration of AAP resulted in a dose-dependent
depletion of hepatic glutathione,
and significant covalent bonding of radiolabeled
AAP to hepatic macromolecules did not occur until at least 70% of the hepatic glutathion&was
depleted. These authors, on the
basis of their experimental findings in rats and mice,
suggested that a toxic metabolite of AAP
ed from covalent
binding
to hepatic
isprevent-
macromolecules
by preferentialreaction and subsequent detoxication
with glutathione.
In man, the importance
of glutathione
in preventing hepatic damage induced by AAP and other aryl
and unsaturated
drugs is unknown. Conjugation
of
drugs with glutathione
occurs, but enzyme activities
in the liver are reported to be lower in man than in
rats and mice (30). Warner and Lorincz were unable
to isolatea glutathione conjugate afteradministering
bromobenzene
to human
subjects (31). Jagenburg
and Toczko (2) reported the identificationof S-(1acetamido-4-hydroxyphenyl)cysteine
as a metabolite
of both AAP and phenacetin
in man; however,
there
was no evidence cited that indicated glutathione
participated
in the formation
of this metabolite.
We
crudely estimated
excretion of I by the two subjects
of this report to be no more than 1% of the ingested
dose, while that of conjugate VI was 5% and 1.6%, respectively, for Subjects 1 and 2. Jagenburg et al. (32)
reported urinary excretions for I of 0.4, 3.0, and 5.9%
by three healthy subjects after oral doses of 1.5 or 2 g
of AAP. These same authors reported the isolation of
another cysteine-containing
metabolite,
excreted in
amounts
ranging from 4.5 to 6.1% of the ingested
dose of AAP. The compound
which they isolated
yielded alanine after desulphurization
and subsequent acid hydrolysis. Based on the absence of reaction with ninhydrin
(I did react) and also on “the
well-known
fact that monohalogenated
benzenes
form
N-acetylcysteine
derivatives
(mercapturic
acids) in the animal body,” Jagenburg
et al. assumed
the metabolite
to be S-(1-acetamido-4-hydroxyphenyl)-mercapturic
acid.
It is possible that the metabolite
isolated by these
tin with nucleophiles
in acid solution. They reported
formation
of
4-hydroxy-3-methylthioacetanilide
when methionine
was the nucleophile
used and postulated that the intermediate
involved was N-acetylp-benzoquinoneimine.
They postulated
that S-(5-
authors is the same metabolite that we have identi-
cleophilic
sulfur group. Further
studies
and other questions
should add much
edge on metabolism
and detoxication
fied as S-(5-acetamido-2-glucuronosidophenyl)-cysteine; however, we found that both it and I show a
positive reaction
with ninhydrin.
This compound,
shown by us to be a glucuronide
by mass spectrometry, was completely
resistant
to hydrolysis
with figlucuronidase
after exposure for 72 h at 37 #{176}C
at a
pH of 5.04. This suggests the possibility of steric hindrance to attachment
of the enzyme or an intramolecular complex between the aglycon and the carboxylic acid group of glucuronic acid.
Our evidence for the identification
of cysteine as
the group attached to the aromatic ring is conclusive
only with regard to the attachment
of sulfur to the
ring (see discussion regarding the mass-spectrometric
studies on probe samples of VI). Based on our interpretation of mass spectral data obtained for VI after
treatment
with diazomethane
and bis(trimethylsilyl)trifluoroacetamide
(see Figure 4b) and its positive response
to ninhydrin,
we believe
only an alanine
moiety is attached
to the sulfur (giving
However, we did not observe a molecular
cysteine).
ion (M)
[and the characteristicM
-15(CH3) ion fortrimethylsilyl derivatives] for the high-molecular-weight
conjugate, so that some possibilityfor error stillexists.Considering the high molecular weight, the numpresent,
and the possibility
of an
bond, it is not surprising
that the
data failed to give a good structural
representation
of the total molecule.
Important
new questions can be answered about
the nature of a possible toxic metabolite
of AAP by
using high-resolution
liquid-chromatography
coupled
ber
of polar
groups
intramolecular
mass spectral
with additional studies of animals. Focella et a!.(18)
identified 4-hydroxy-3-methylthioacetanilide as a
new urinary metabolite of phenacetin in the dog. To
attempt to clarifythe mechanism by which the 3substituted metabolites might be formed, Calder et
al. (33) studied the reactions of N-hydroxyphenace-
acetamido-2-hydroxyphenyl)cysteine
might
be
formed via a reaction of cysteine with this same intermediate.
These results, coupled with our failure to
identify any mercapturic
acid metabolite,
may argue
against reaction of a presumed
toxic metabolite
of
AAP with glutathione
as suggested by Mitchell et al.
(25). The protective
role suggested by the studies of
these authors
on rats and mice may actually be
played by either cysteine or methionine
in human
metabolism.
Identifying
the precursor
of the new conjugated
metabolite
VI isvitalto further in vivo studies on the
nature of the presumed toxic metabolite. The precursor could be either I or V, the glucuronide
of AAP. If
it is the latter, this probably would rule out epoxide
formation (23) as preceding conjugation with the nuto clarify this
to our knowl-
of aromatic
compounds
in the human body, facilitating
the ftrmulation of safer and, perhaps, more effective drugs.
It is important
to note that recovery of the ingested drug dose averaged 100.6% for the two subjects in
this study. This again illustrates
the utilityof liquid-
chromatographic
a wide spectrum
techniques in drug studies to enable
of metabolic changes to be observed.
In this way, unexpected metabolic changes induced
by a drug can be observed, enabling mechanistic in-
ferences
regarding
its mode
of action
to be further
tested.
We gratefully
acknowledge
Jr., of the Analytical
Rainey,
the technical
Chemistry
assistance
of Dr. W. T.
Division at ORNL.
References
1. Cummings,
study
A. J., King,
of drug elimination:
metabolites
(1967).
in
man.
M. L., and
The excretion
Brit.
J.
Phormacol.
Martin, B. K., A kinetic
of paracetarnol and its
Chemother.
29,
150
2. Jagenburg,
0. R., and Toczko,
K., The metabolism
of acetophenetidine.
Isolation
and characterization
of S-(1-acetamido-4hydroxyphenyl).cysteine,
a metabolite
of acetophenetidine.
Bioc/oem. J. 92,639(1964).
3. Nelson, E., and Morioka,
T., Kinetics
of the metabolism
of acetaminophen by humans. J. Pharm. Sci. 53,864 (1963).
4. Levy, G., and
tions in man. V:
60, 608 (1971).
5. Shibasaki,
J.,
metabolism,
and
metabolism
of
Pharm. Bull. 16,
Regardh,
C-G., Drug biotransformation
interacAcetaminophen
and salicylic acid. J. Pharm.
Sci.
Koizumi,
T., and Tanaka,
T., Drug
excretion.
I. Some pharmacokinetic
acetanilide
and
4-hydrmtyacetanilide.
1661 (1968).
absorption,
aspects of
C/oem.
6. Shibasaki,
J., Konishi,
R., Takeda,
V., and Koizumi, T., Drug
absorption,
metabolism,
and excretion.
VII. Pharmacokinetics
on
formation
and excretion
of the conjugates
of N-acetyl-p-aminophenol in rabbits. Chem. Pharm. Bull. 19, 1800 (1971).
7. Prescott,
comparative
nol in man,
Pharmacol.
L. F., Sansur,
M., Levin, W., and Conney,
A. H., The
metabolism
of phenacetin
and N-acetyl-p-aminophewith particular
reference
to effects on the kidney. Gun.
Ther. 9,605 (1968).
CLINICAL
CHEMISTRY,
Vol. 20, No. 8, 1974
1095
8. Burtis, C. A., Butts, W. C., and Rainey,
the metabolites
W. T., Jr., Separation
by high-resolution
anion
Amer. J. Clin. Pathol. 53, 769 (1970).
of phenacetin
in urine
of
ex-
change chromatography.
9. Brodie, B. B., and Axeirod, J., Metabolic fate of acetophenetidme in man. J. Pharmacol. Exp. Ther. 97, 58 (1949).
10. SUch, H., HaUser, H., Pfleger, K., and RUdiger, W., Uber die
Ausscheidung eines noch nicht beschriebenen
Phenacetinmetaboliten beim Menschen
und bei der Ratte. Arch. Exp. Pathol. Pharmakol.
253, 25 (1966).
11. Klutch,
phenetidine,
9,63
A., Harfenist,
M., and Conney, H. H., 2-Hydroxyacetoa new metabolite
of acetophenetidine.
J. Med. Chem.
(1966).
12. Katz, S., Pitt,
W. W., Jr., and Jones, G., Jr., Sensitive fluorescence monitoring
of aromatic acids after anion-exchange
chromatography of body fluids. Clin. Chem. 19,817 (1973).
13. Katz, S., Pitt, W. W., Jr., and Mrochek, J. E., A two-detector
anion-exchange
chromatograph
for comparative serum and urine
analysis.
To be submitted to Anal. Biochem.
14. Mrochek, J. E., Butts, W. C., Rainey, W. T., Jr., and Burtis, C.
A., The separation
and identification
of urinary
constituents
by
use of multiple-analytical
techniques. Clin. Chem. 17, 72 (1971).
15. Mrochek, J. E., and Rainey, W. T., Jr., Identification
and biochemical significance of substituted
furans in human urine. Clin.
Chem. 18,821(1972).
16. Mrochek,
cal techniques
J. E., Dmsxnore,
S. R., and Waalkes, T. P., Analytiand identification
of degradation
products of tRNA: Application to urine samples from cancer patients. Submitted to J. Nat. Cancer Inst.
in the separation
17. Mrochek, J. E., and Rainey, W. T., Jr., Gas chromatography
and mass spectrometry
of some trimethylsilyl
derivative
of urinary
glucuronides.
Anal. Biochem. 57, 173 (1974).
18. Focella,
A., Heslin,
P., and
Teitel,
S., The
synthesis
of two
phenacetin metabolites.
Can. J. C/oem. 50, 2025 (1972).
19. Billets, S., Lietman, P. S., and Fenselau, C., Spectral analysis
of glucuronides. J. Med. Chem. 16,30(1973).
20. Grove, J., Gas-liquid chromatography
of N-acetyl-p-aminophenol (paracetamol)
in plasma and urine. J. Chromatogr.
59, 289
(1971).
1096
CLINICAL CHEMISTRY, Vol. 20, No. 8, 1974
21. Thomas,
B. H., and Coldwell, B. B., Estimation of phenacetin
in plasma and urine by gas-liquid chromatography. J. Pharm. Pharmacol. 24, 243 (1972).
22. Mitchell, J. R., Jollow, D. J., Potter, W. Z., et al., Acetaininophen-induced
hepatic necrosis. I. Role of drug metabolism.
J.
Pharmacol.
Exp. Ther. 187, 185 (1973).
23. Jollow, D. J., Mitchell, J. R., Potter, W. Z., et al., Acetaminophen-induced
hepatic necrosis. II.. Role of covalent binding in vivo.
J. Pharmacol. Exp. Ther. 187, 195 (1973).
24. Potter, W. Z., Davis, D. C., Mitchell, J. R., et al., Acetaminophen-induced
hepatic necrosis.
III. Cytochrome
P-450-mediated
covalent binding in vitro. J. Pharmacol.
Exp. Ther. 187, 203
and paracetamol
(1973).
25. Mitchell,
phen-induced
J. R., Jollow, D. J., Potter, W. Z., et al., Acetaininohepatic necrosis. IV. Protective role of glutathione.
J. Pharmacol.
Esp. Ther. 187, 211 (1973).
26. Boyland, E., and Chasseaud, L. F., Enzyme-catalysed
conjugations of glutathicoe
with unsaturated
compounds.
Biochem. J. 104,
95 (1967)..
27. Boyland, E., and Chasseaud, L. F., Enzymes catalysing conjugations of glutathione
with a,-unsaturated
carbonyl
compounds.
Biochem. J. 109, 651 (1968).
28. Clapp, J. J., and Young, L., Formation of mercapturic acids in
rats after the administration
of arailcyl esters. Biochem. J. 118, 765
(1970).
29. Giltham,
B., The reaction
of aralkyl
sulphate
esters
with gluts-
thione catalyzed by rat liver preparations.
Biochem. J. 121, 667
(1971).
30. Grover, P. L., and Sims, P., Conjugations
with glutathione:
Distribution
of glutathione
Biochem.
J. 90,603 (1964).
S-aryltransferase
in vertebrate
species.
31. Warner, A., and Lorincz, A. B., Mercapturic acid synthesis by
humans. Life Sci. 2, 504 (1963).
32. Jagenburg, R., Nagy, A., and R.odjer, S., Separation of p-acetaminophenol
metabolites
by gel filtration
on Sephadex
G-10.
Scand. J. Clin. Lab. Invest. 22, 11(1968).
33. Calder, I. C., Creek, M. J., and Williams, P. J., N-Hydroxyphenacetin as a precursor of 3-substituted
4-hydroxyacetanilide
metabolitesof phenacetin.
Chem.-Biol. Interactions
8,87 (1974).