JOURNAL OF

CHROMATOGRAPHY B:
BIOMEDICALAPPLICATIONS
ELSEVIER Journal of Chromatography B, 681 (1996) 233-239

Simultaneous radiometric and fluorimetric detection of lauric acid

metabolites using high-performance liquid chromatography
following esterification with 4-bromomethyl-6,7-dimethoxycoumarin
in human and rat liver microsomes
Y. Amet*, F. Berthou, J.F. Menez
Laboratoires de Biochimie-Nutrition, Equipe d'Accueil DGRT EAD 948, Facult# de M#decine, B.P. 815, 29285 Brest, France
Received 17 November 1995; revised 17 January 1996; accepted 18 January 1996

Abstract

The formation of (oJ-1)-hydroxylauric acid from lauric acid (LA) can be used as an indicator of the activity of
cytochrome P450 2El (CYP2E1) in rat and human liver microsomes. A high-performance liquid chromatographic (HPLC)
method that is capable of identifying and measuring the two main metabolites of lauric acid, (~o-1)- and ~o-OH-LA, has been
developed and used in the study of rat and human liver microsomes. Measurement of the enzymatic activities, based on the
esterification of the metabolites and substrate with the fluorescent agent, 4-bromomethyl-6,7-dimethoxycoumarin, is
described using both radiometric and fluorimetric detection methods. Extraction efficiencies of metabolites and residual
substrate were calculated using radioactivity and were greater than 85%. The assay is accurate and reproducible and has a
detection limit of 75 pg (0.37 pmol). Additionally, a strong correlation between the two techniques was found in both human
(r=0.945, n=15, p<0.01) and rat (r=0.949, n=18, p<0.01) livers, for the (o~-l)-hydroxylauric acid.

Keywords: Lauric acid; Hydroxylauric acid; 4-bromomethyl-6,7-dimethoxycoumarin

1. Introduction lapping patterns of substrate specificities. In recent

years, particular interest has focused on the ethanol-
Microsomal cytochrome P450s ' constitute a inducible cytochrome P450 2El (CYP2E1) which
superfamily of heme-thiolate enzymes which cata- has been shown to be involved in the activation of
lyze the monooxidation of a wide variety of lipo- many low-molecular-mass toxins and carcinogens
philic compounds, such as fatty acids. The various [I]. CYP2E! also has the capacity to metabolize
cytochrome P450 isoforms exhibit distinct but over- other endo- and xenobiotic compounds such as
ketone bodies, acetaminophen, chlorzoxazone and
ethanol [2]. Because of the importance of CYP2E1
*Corresponding author. in alcohol-related diseases, a number of compounds
'Cytochrome P450 (EC 1.14.14.1): P450 or CYP. The name
cytochrome must be abandoned according to the Nomenclature have been investigated as substrate probes for this
Committee of the International Union of Biochemistry, the enzyme, such as 4-nitrophenol [3,4] or chlorzox-
appropriate name being 'heme-thiolateprotein'. azone [5,6], which have been used as markers of

0378-4347/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved

P l l S0378-4347(96)00043-6
234 Y. Amet et al. / J. Chrornatogr. B 681 (1996) 233-239

CYP2E1 enzymatic activity in a number of animal and (o~-1)-OH-LA)] were given to us by Dr. Salatin
species. (Strasbourg, France). [1-14C]Lauric acid (50 mCi/
The NADPH- and oxygen-dependent to-oxidation mmol) was purchased from Amersham (Amersham,
of fatty acids, such as lauric acid (LA), has been UK). 18-Crown-6-ether and 4-bromomethyl-6,7-di-
shown to be catalyzed by the microsomal P450 methoxycoumarin were obtained from Aldrich (St.
family and leads to the formation of both to- and Quentin Fallavier, France) and NADPH (nicotin-
(to- 1)-hydroxylated products. The (to-1)-hydroxy- amide adenine diphosphate hydrogen) was from
lation of lauric acid, a medium-chain-length fatty Sigma (St. Louis, MO, USA). HPLC-grade acetoni-
acid, was recently described as being due to CYP2E 1 trile was purchased from SDS (Peypin, France). All
in rat and human liver microsomes [7-9]. The ratio other chemicals and solvents were of reagent grade.
between the two metabolites was shown to be Deionised water and acetonitrile were further filtered
dependent on environmental parameters, such as before use.
drug administration (hypolipidaemic drugs) and
phthalate esters, which enhanced the to-hydroxylated 2.2. Preparation of rat and human liver
metabolite, or starvation, diabetes and alcoholism, microsomes
which increased the (to-1)-hydroxylated compound.
A number of reversed-phase HPLC methods have Human liver samples were obtained in accordance
been developed for the detection of to- and (to-1)- with French legal considerations and approval of the
hydroxylated metabolites of fatty acids. In most of ethical committee.
these studies [10-13], radioactive lauric acid was Male Sprague-Dawley rats were maintained on a
used as the substrate in order to obtain sensitive water and standard diet ad libitum at 22±2°C with a
detection. More recently, a number of reports using 12 h light-dark cycle.
HPLC in conjunction with non-specific detection Rat and human liver samples were frozen immedi-
methods, mainly pre-column derivatization of the ately after removal and the microsomal fraction was
carboxylic group, have been published. These au- prepared according to a previously described method
thors used either the diastereoisomeric compound, [20]. The subcellular fraction, in a 100 mM phos-
methoxyfluoromethyl phenylacetate [14], 4-bromo- phate buffer, pH 7.4, containing 20% glycerol, was
methyl-6,7-dimethoxycoumarin [ 15,16], the UV-ab- stored at - 8 0 ° C until use. Microsomal protein
sorbing compound, p-bromophenacyl bromide [17], content was determined using the Bradford method
1-pyrenyldiazomethane (PDAM) [18] or p-(9-an- (Biorad, Munich, Germany).
throyloxy)phenacylbromide (panacyl bromide) [19].
The aim of this paper was to describe and compare 2.3. Assay of lauric acid metabolism
the formation and the detection of the metabolites of
lauric acid using both [1-14C]lauric acid, as the A 2-ml reaction mixture containing microsomal
substrate probe, and a pre-column ester derivatiza- proteins (0.3 rag), lauric acid (1 #Ci, 0.1 mM) in
tion of the carboxylic function with a simultaneous 0.12 M phosphate buffer, pH 7.4, and 5 mM MgC12,
fluorescent marker, 4-bromomethyl-6,7-dimethox- was incubated at 37°C for 10 min. The reaction was
ycoumarin. This method gives a fluorimetric and initiated by the addition of 1 mM NADPH. After
radiometric analysis of lauric acid metabolites. terminating the reaction with 0.8 ml of 10% HzSO 4,
the metabolites were extracted twice with 6 ml of
diethyl ether, dried on NazSO 4 and evaporated under
2. Experimental a nitrogen stream. Fluorescent derivatives were
prepared by adding 0.1 ml of a solution of 18-crown-
2.1. Chemicals 6-ether in acetonitrile (2.5 mg/ml), 2 mg of dried
potassium carbonate and 0.1 ml of a solution of
Lauric acid was obtained from Fluka (Buchs, 4-bromomethyl-6,7-dimethoxycoumarin in acetone
Switzerland), while the metabolites [11- and 12- (10 mg/ml) to the dried residue. Following vigorous
hydroxylauric acid (11-OH-LA and 12-OH-LA or w- shaking, the derivatization was performed at 70°C
 Y. Amet et al. / J. Chromatogr. B 681 (1996) 233-239 235

for 60 min in darkness. The mixture was then cooled Several mobile phases were investigated in order to
and diluted with acetonitrile before HPLC analysis. establish optimum separation and to acheive the
highest analytical sensitivity possible for lauric acid
2.4. HPLC analysis and both radiometric and metabolites. The best results were obtained under the
fluorimetric detections conditions reported in Section 2.4, using acetonitrile
rather than methanol.
The o~- and (o~-l)-hydroxylauric acids and re- The w- and (to-1)-hydroxylauric acid metabolites
sidual lauric acid were separated by HPLC using a were identified on the basis of their retention times
reversed-phase Nucleosil C~8 column (5 /zm, 250X and their radiometric and fluorimetric labelling. The
4.6 mm I.D.; specific area, 300 m2/g; pore size, 120 retention times of the esters of ll-OH-LA, 12-OH-
A; carbon yield, 13%, Interchim, Montluqon, LA and LA were 20.5, 22 and 41 min, respectively.
France). The mobile phase (1% acetic acid in water- The two metabolites formed in the presence of
acetonitrile) programme began isocratically with a NADPH were clearly baseline-separated during the
45:55 mixture (v/v) for 20 min at a flow-rate of 1.0 isocratic step of the gradient elution and were not
ml/min, followed by a 3-min linear gradient to 5:95 detected in the absence of NADPH (data not shown).
(v/v) water and acetonitrile, and a hold for 15 min Blank samples (without substrate or without
before returning to the initial conditions. NADPH) from rat and human liver microsomes
The chromatograph was equipped with a fluorime- showed no endogenous peaks interfering with the
ter (300-400 and 417-700 nm excitation and emis- assays. The peaks shown near the beginning of the
sion wavelengths, respectively; Fluoromonitor III, chromatogram when using fluorimetric detection
LDC-Milton-Roy, Riviera Beach, FL, USA), inter- were due to the fluorescent marker.
faced with a Flo-one Beta radiometric detector
(Packard, Meriden, CT, USA). Emission and fluores- 3.2. Excitation and emission spectra from
cence were recorded by a TSP FL 200 spectro- fluorescent compounds
fluorometer (Thermo Separation Products, San Jose,
CA, USA). Peak areas were calculated, using both Excitation and emission scan profiles are illus-
radiometric and fluorimetric data, from the percent- trated in Fig. 2. Each of the three compounds (I1-
age of metabolite area to the total product area OH-LA, 12-OH-LA and LA) had the same profiles,
(metabolite + residual substrate). Data were expressed both in emission and excitation scans (excitation
as nmol/min/mg of protein. Statistical analysis was range 300-420 nm, emission range 350-500 nm).
performed using the Student's t-test. The maxima were 340 and 440 nm for excitation and
emission respectively and these conditions were used
in our experiments.
3. Results and discussion
3.3. Validation of the method
3.1. Elution profiles of lauric acid and its
metabolites by HPLC using both fluorimetric and Calibration curves using fluorescent standards
radiometric detections gave excellent linearity, the intercepts of the regres-
sion lines were close to zero and the correlation
Fig. 1 shows the lauric acid metabolite HPLC coefficients were greater than 0.995. The limit of
profiles generated by enzymes from a microsomal detection for lauric acid (signal-to-noise ratio of 3)
fraction of rat (A and B) and human (C and D) was 75 pg (0.37 pmol) and this value was in
livers, following derivatization with the fluorescent accordance with previous data [15], where the au-
marker. Detection of the peaks was performed using thors had a detection limit for fatty acids of about 0.5
a simultaneous fluorimetric (A and C) and radiomet- pmol.
ric (B and D) method. The extraction efficiency of lauric acid and its
Fluorescence detection was used because it pro- metabolites was determined by radioactivity. The
vides a greater specificity than UV absorbance. extraction solvent, diethyl ether, was chosen as it
236 Y. Arnet et al. / J. Chromatogr, B 681 (1996) 2 3 3 - 2 3 9

INU ll)- Fluorimetrictrace LA dpm (B)- Radiometric trace LA

7475 300.

4~0 200'
! 10H-LA
i I-OH-LA
i i 120H-LA

2425 I00

-IN
20 40 mln 20 40 mln

15000 (C)- Fluorimetrictrace LA dpm (D)- Radiometrictrace LFI

11225 300

'/451 200

II-OH-LA I I-OH-LA

3675

20
12-OH-LA

40 min
100

_25 2'0 40 min

Fig. 1. Lauric acid metabolite HPLC profiles following esterification with 4-bromomethyl-6,7-dimethoxycoumarin: (A) and (B) incubation
of 0.3 mg of microsomal protein from control rat liver; (C) and (D) incubation of 0.3 mg of microsomal protein from human liver FH3.
Detection was carried out by either fluorimetry (left) or radiometry (right).
 Y. Amet et al. / J. Chromatogr. B 681 (1996) 233-239 237

j . IB - Emission spectrum
IR Excitation spectrum
i

-~ l(excttation=340 rim)
,m
(emission = 440 rim)
g) =3
o ; p l | "q
• ". ! 2-OH-LA
a)
2 .
,,
,. ; ... x...
~p
..,;
0 "-- 2 Olt-tA
rail
L,-

~/( ', ,;-Off-I.,

" ~ i , _ ° '~',
! "j~j ~'~".
F ~, %.•
• 0' a al • i•
"" •

'" , , ~ LA
I "' • / I
, -- ' , Zeo 420 460 Soo nm
300 340 380 420
Fig. 2. Excitation (A) and emission (B) spectra of lauric acid metabolites following esterification with 4-bromomethyl-6,7-dimethoxy-
coumarin.

produced no potentially interfering peaks and led to pound by linear regression analysis of the metabolite
an extraction efficiency of 87.5+7.2% (n= 12), with peak concentration versus the ratio of lauric acid
suitable reproducibility. Additionally, interferences metabolite to internal standard peak areas. The
from fatty acids extracted from the microsomal results showed good r values (y=0.0508x+0.125,
fractions were not observed. Under these conditions, r=0.985 and y=0.0897x+0.140, r=0.958, for l l -
it was found that the coumarin esters of lauric acid OH and 12-OH-LA, respectively), suggesting that
and its metabolites are stable for at least two months this internal standard could be used to calibrate the
when stored in the dark at 4°C and diluted in method and also that it was possible to calculate the
acetonitrile rather than in methanol. enzymatic activities by the area ratio method.
The reproducibility of the esterification and fluo-
rescence or radiolabelling detections was evaluated 3.4. Comparison o f the fluorimetric and
by duplicate injections of three separate incubations radiometric detection o f lauric acid metabolism in
of rat control microsomal preparations. The repro- rat and human liver microsomes
ducibility for the two metabolites was excellent with
a coefficient of variation (C.V.) of 5.8+1.9% and This method of fluorescent derivatization was
6.2 + 1.7% for radiometric and fluorimetric detection, carried out in fifteen human and eight control rat
respectively. liver microsomal fractions. The o~- and (m-1)-hy-
Calibration of the method was carried out using droxylase activities of lauric acid are shown in Table
caproic acid, which had a retention time of 30 rain 1. The rate of metabolite formation was calculated
under the above conditions, as the internal standard. based on the percentage of either the fluorescent
A calibration graph was constructed for each corn- compound or the radioactivity of the given metabo-
238 Y. Amet et al. / J. Chromatogr. B 681 (1996) 233-239

Table 1
to- and (to-I)-Hydroxylase activities of lauric acid from microsomes of human (n-15) and control rat (n=8) livers
(w-1)-OH-LA to-OH-LA
Fluorimetric Radiometric Fluorimetric Radiometric
Rat 1.51 + 0 . 3 1 1.42±0.28 1.66-+0.27 1.42±0.38
Human 1.66+0.80 1.31+-_0.69 1.07-+0.65 0.88 -+0.30
Detection of the peaks was performed using either fluorimetric or radiometric detection. Results are expressed as nmol/min/mg of
microsomal protein_+standarddeviation.

lite relative to the total fluorescence or radioactivity esters obtained with PDAM and panacyl bromide.
eluting from the HPLC column. The enzymatic These fluorescent markers permitted a less sensitive
activities were significantly correlated between the detection than the one described in Section 2.4 and
two detection methods and are in accordance with the limit of detection was greater than with 4-bromo-
results previously described [7,8]. methyl-6,7-dimethoxycoumarin (150 and 200 pg for
Correlation coefficients (r) were calculated using the panacyl bromide and PDAM esters, respectively),
either eighteen control and C Y P 2 E l - i n d u c e d rats or results which were only slightly higher than those
fifteen human livers. Detection of the peaks was previously described [18,19].
performed with the fluorimetric and radiometric
methods. The results are illustrated in Fig. 3 and the
correlation coefficients were approximately 0.95 4. Conclusion
(p<0.001).
When the use of radiolabelled substrate is not
3.5. C o m p a r i s o n w i t h o t h e r f l u o r e s c e n t d e r i v a t i v e s possible or is prohibited, derivatization of fatty acids
with fluorescent or UV-labelling probes is necessary
In order to validate the present method, other and therefore the method must provide great sen-
fluorescent derivatives were tested and compared to sitivity, a weak detection limit and suitable repro-
the method described in Section 2, i.e. fluorescent ducibility. In our study, 4-bromomethyl-6,7-di-

(A) P,~Uver (B)- H u m a n ~

,~Lo | 11-OH-LA •
0 12-OH-LA r 0.912 3.o o 12-OH-LA r-0.~6

° ! x.J

o.o , i ,~ o.o ,' 'i

bdkmmrk dam~ (umd/m~/mS) bmoam~ 6mec~ (mumoUm~m~
Fig. 3. Correlationplots of radiometric and fluorimetric detections of 11-OH- and 12-OH-LA in human liver microsomes (A; n = 15) and
control or CYP2El-induced rat liver microsomal samples (B; n= 18).
 Y. Amet et al. / J. Chromatogr. B 681 (1996) 233-239 239

methoxycoumarin appears to be a useful fluorescent [7] Y. Amet, F. Berthou, T. Goasduff, J.E Salaiin, L. Le Breton
marker. The derivatization reaction is rapid and and J.F. M6nez, Biochem. Biophys. Res. Commun., 203
(1994) 1168.
provides a good yield. Fluorescence detection is [8] Y. Amet, F. Berthou, S. Baird, Y. Dr6ano, J.E Bail and J.F.
approximately as sensitive as radiometric detection M6nez, Biochem. Pharmacol., 50 (1995) 1775.
and is more sensitive than UV detection. [9] S.E. Clarke, S.J. Baldwin, J.C. Bloomer, A.D. Ayrton, R.S.
In conclusion, these results indicate that 4-bromo- Sozio and R.J. Chenery, Chem. Res. Toxicol., 7 (1994) 836.
methyl-6,7-dimethoxycoumarin is an excellent de- [10] G.G. Gibson, T.C. Orton and EP. Tamburini, Biochem. J.,
203 (1982) 161.
rivative for the detection of fatty acids and hydroxy- [11] R. Okita and C. Chance, Biochem. Biophys. Res. Commun.,
lated metabolites in biological samples following 121 (1984) 304.
HPLC analysis. [12] EE Tamburini, H.A. Masson, S.K. Bains, R.J. Makowski, B.
Morris and G.G. Gibson, Eur. J. Biochem., 139 (1984) 235.
[13] J.P. Hardwick, B.J. Song, E. Huberman and F.J. Gonzalez, J.
Biol. Chem., 262 (1987) 801.
References [14] R. Azerad, J.L. Boucher, P. Dansette and M. Delaforge, J.
Chromatogr., 498 (1990) 293.
[1] F.E Guengerich, D.H. Kim and M. Iwasaki, Chem. Res. [15] R. Farinotti, E Siard, J. Bourson, S. Kirkiacharian, B. Valeur
Toxicol., 4 (1991) 168. and G. Mahuzier, J. Chromatogr., 269 (1983) 81.
[2] M. Morimato, A.L. Hagbjork, A.A. Nanji, M. Ingelman- [16] H. Dirven, A. de Bruijn, P. Essink and F.J. Jongeneelen, J.
Sundberg, K.O. Lindros, EC. Fu, E. Albano and S.W. Chromatogr., 564 (1991) 266.
French, Alcohol, 10 (1993) 459. [17] T. Aoyama and R. Sato, Anal. Biochem., 170 (1988) 73.
[3] D.R. Koop, Mol. Pharmacol., 29 (1986) 399. [18] S.R. Mason, L.C. Ward and P.E. Reilly, J. Liq. Chromatogr.,
[4] W. Tassaneeyakul, M.E. Veronese, D.J. Birkett, F.J. Gonzalez 17 (1994) 619.
and J.O. Miners, Biochem. Pharmacol., 46 (1993) 1975. [19] S. Salari, M. Yeung, S. Douglas and W. Morozowich, Anal.
[5] E Peter, R. B6cker, E Beaune, M. Iwasaki, F.E Guengerich Biochem., 165 (1987) 220.
and C.S. Yang, Chem. Res. Toxicol., 3 (1990) 566. [20] F. Berthou, D. Ratanasavanh, C. Rich~, D. Picart, T. Voirin
16] V. Carri~re, T. Goasduff, D. Ratanasavanh, F. Morel, J.C. and A. Guillouzo, Xenobiotica, 19 (1989) 401.
Gautier, A. Guillouzo, E Beaune and F. Berthou, Chem. Res.
Toxicol., 6 (1993) 852.