Detection of residual chloramphenicol, florfenicol, and thiamphenicol in yellowtail fish muscles by capillary gas chromatography-mass spectrometry

Article abstract of DOI:10.1021/jf950343u

Chloramphenicol (CAP), florfenicol (FF), and thiamphenicol (TAP) were extracted from yellowtail muscles with ethyl acetate, and the extract was evaporated. The residues was dissolved with sodium chloride solution and partitioned with n-hexane to remove lipids, and then the drugs were extracted with ethyl acetate. After evaporation of ethyl acetate extract, the residue was dissolved with n-hexane followed by ethyl ether and applied to a Sep-Pak Florisil cartridge successively. The drugs were eluted from the cartridge with methanol-ethyl ether (3:7), and eluate was evaporated to dryness. After acetonitrile and BSA were added to the residues, the drugs were derivatized at 50 °C for 10 min. After the solvent was evaporated, the residue was dissolved with ethyl acetate and applied to gas chromatography-mass spectrometry. The drugs were separated by a capillary column coated with 5% phenyl methyl silicone, and SIM was performed at m/z 208 and 225 for CAP, at mlz 257 for FF, and at mlz 242, 257, and 330 for TAP. Recoveries of each drug from yellowtail muscle fortified at 0.1 ppm were more than 65%, and detection limits were 5 ppb.

Full text of DOI:10.1021/jf950343u

1280
J. Agric. Food Chem. 1996, 44, 1280−1284
Detection of Resid u a l Ch lor a m p h en icol, F lor fen icol, a n d
Th ia m p h en icol in Yellow ta il F ish Mu scles by Ca p illa r y Ga s
Ch r om a togr a p h y-Ma ss Sp ectr om etr y
Tomoko Nagata*,† and Hisao Oka‡
Public Food Analysis Laboratory of Chiba Prefecture, 205-8 Shinmei-cho Chuo-ku, Chiba City 260, J apan,
and Aichi Prefectural Institute of Public Health, 7-6 Nagare Tsuji-machi, Kita-ku, Nagoya 462, J apan
Chloramphenicol (CAP), florfenicol (FF), and thiamphenicol (TAP) were extracted from yellowtail
muscles with ethyl acetate, and the extract was evaporated. The residues was dissolved with sodium
chloride solution and partitioned with n-hexane to remove lipids, and then the drugs were extracted
with ethyl acetate. After evaporation of ethyl acetate extract, the residue was dissolved with
n-hexane followed by ethyl ether and applied to a Sep-Pak Florisil cartridge successively. The drugs
were eluted from the cartridge with methanol-ethyl ether (3:7), and eluate was evaporated to
dryness. After acetonitrile and BSA were added to the residues, the drugs were derivatized at 50
°C for 10 min. After the solvent was evaporated, the residue was dissolved with ethyl acetate and
applied to gas chromatography-mass spectrometry. The drugs were separated by a capillary column
coated with 5% phenyl methyl silicone, and SIM was performed at m/z 208 and 225 for CAP, at m/z
257 for FF, and at m/z 242, 257, and 330 for TAP. Recoveries of each drug from yellowtail muscle
fortified at 0.1 ppm were more than 65%, and detection limits were 5 ppb.
Keyw or d s: Chloramphenicol; thiamphenicol; florfenicol; antibacterials; yellowtail
INTRODUCTION
mode, residual CAP in meat was determined as low as
5 ppb and CAP in milk by GC-MS with negative
chemical ionization mode was determined at 0.2 µg/L
(Keukens et al., 1992). By GC-MS in methane negative
chemical ionization, CAP in milk was confirmed at a
low level of 0.5 ng/mL (Kijak, 1994). By gas chroma-
tography-high-resolution mass spectrometry, CAP resi-
due in egg was confirmed at a low level of 0.5 ppb.
However, few reports of the confirmation of residual
TAP in tissues have been published. Residual TAP in
rat tissues has been determined by GC-MS at a
detection limit of 0.1 ppm (Plomp and Maes, 1976).
In this paper, residual CAP, FF, and TAP in yellowtail
muscle were determined by GC-MS simultaneously.
The drugs were extracted from muscle and cleaned up
by liquid-liquid followed by solid-liquid partition as
described earlier (Nagata and Saeki, 1992). Then the
drugs were derivatized with trimethylsilyl reagent.
Selected ion monitoring mass spectrometry (SIM-MS)
was employed to quantify and confirm residual CAP,
FF, and TAP in yellowtail muscles at low levels of 5 ppb.
Chloramphenicol (CAP) and thiamphenicol (TAP)
have been used to treat many kinds of animal diseases
as effective broad spectrum antibacterials. Recently, in
J apan, florfenicol (FF), the structure of which is quite
similar to those of the above two drugs, has been
permitted for use on cultured yellowtails. The struc-
tures of these three antibacterials are shown in Figure
1.
Many various determination methods of CAP, TAP,
and FF by gas chromatography (GC) or high-perfor-
mance liquid chromatography (HPLC) have been re-
ported so far. Reviews of these three antibacterials have
been already published (Allen, 1985; Nagata, 1995). GC
and HPLC are useful for the separation and quantita-
tion of residual antibacterials. However, interfering
substances from animal tissues render unequivocal
identification by GC and HPLC using retention and/or
fixed wavelength. Mass spectrometric analysis, which
provides molecular weight and characteristic fragmen-
tation patterns, provides confirmation of antibacterials
with retention and spectral characteristics. Several
determination methods of CAP residues in urine, plasma,
meat, egg, and milk by gas chromatography-mass
spectrometry (GC-MS) (Gazzaniga et al., 1973; J anssen
and Vanderhaeghe, 1973; Nakagawa et al., 1975; Keu-
kens et al., 1992; Kijak, 1994; Borner et al., 1995) or
liquid chromatography-mass spectrometry (Bories et
al., 1983; Delepine and Sanders, 1992; Ramsey et al.,
1989) have been reported so far. The detection limits
of these determination methods were between 0.2 ppb
and 0.5 ppm. Recently, CAP has been confirmed at very
low residual levels. By GC-MS in the electron impact
EXPERIMENTAL PROCEDURES
Ch em ica ls. N,O-Bis(trimethylsilyl)acetamide (BSA) was
obtained from GL Science Inc. (Tokyo, J apan). Chlorampheni-
col and thiamphenicol were from Sigma Chemical Co. (St.
Louis, MO). Florfenicol was a gift from Takeda Yakuhin
Kogyo Co. (Tokyo, J apan). Each stock solution was prepared
by dissolving 10 mg of each drug in 100 mL of methanol at
100 µg/mL. The working standard solution was prepared at
1 µg/mL in methanol using each 1 mL stock solution. A Sep-
Pak Florisil cartridge (Waters Co., Milford, MA) was precon-
ditioned with 5 mL of n-hexane followed by 5 mL of ethyl ether
prior to use. Other chemicals were the same as described
before (Nagata and Saeki, 1992).
In str u m en ta tion . A Hewlett-Packard HP 5972 Series II
mass spectrometry system and a Vectra 486/60 XM data
system were used. The following operating conditions were
used: ionization energy, 70 eV; electron multiplier, 1800 V;
emission current, 300 µA; scan range, 150-500. All data for
* Author to whom correspondence should be ad-
dressed (fax 043 246 9912).
† Public Food Analysis Laboratory of Chiba Prefec-
ture.
‡ Aichi Prefectural Institute of Public Health.
S0021-8561(95)00343-8 CCC: $12.00
© 1996 American Chemical Society

Detection of CAP, FF, and TAP by GC
−MS
J. Agric. Food Chem., Vol. 44, No. 5, 1996 1281
F igu r e 1. Chemical structures of CAP, FF, and TAP.
Ta ble 1. Recover y of CAP , F F , a n d TAP Ad d ed to
Yellow ta il Mu scles
CAP
TAP
FF
m/z 257
m/z 208 m/z 225
m/z 242
m/z 257
m/z 330
no.
of found recov found recov found recov found recov found recov found recov
expt (µg) (µg) (µg) (µg) (µg) (µg)
%
%
%
%
%
%
2 µg Addeda
1
2
3
4
5
1.41 70.5 1.30 65.0 1.28 64.0 1.32 66.0 1.30 65.0 1.29 64.5
1.45 72.5 1.64 82.0 1.35 67.5 1.34 67.0 1.35 67.5 1.30 65.0
1.43 71.5 1.60 80.0 1.24 62.0 1.40 70.0 1.45 72.5 1.31 65.5
1.30 65.0 1.47 73.5 1.29 64.5 1.39 69.5 1.42 71.0 1.32 66.0
1.45 72.5 1.53 76.5 1.32 66.0 1.41 70.5 1.44 72.0 1.30 65.0
mean 1.41 70.5 1.51 75.5 1.30 65.0 1.37 68.5 1.39 69.5 1.31 65.5
SD 0.060
CV 4.25
0.132
8.73
0.042
3.25
0.042
3.04
0.064
4.57
0.016
1.27
F igu r e 2. Effect of reaction time on derivatization of CAP
(b), FF (2), and TAP (4).
1 µg Addeda
1
2
3
4
5
0.692 69.2 0.633 63.3 0.622 62.2 0.644 64.4 0.671 67.1 0.644 64.4
0.714 71.4 0.661 66.1 0.693 69.3 0.621 62.1 0.642 64.2 0.691 69.1
0.603 60.3 0.695 69.5 0.621 62.1 0.662 66.2 0.692 69.2 0.682 68.2
0.641 64.1 0.630 63.0 0.621 62.1 0.633 63.3 0.649 64.9 0.659 65.9
0.620 62.0 0.690 69.0 0.664 66.4 0.693 69.3 0.608 60.8 0.641 64.1
quantification were collected in the SIM mode at m/z 208, 225,
242, 257, and 330 at each dwell time of 100 ms. A Hewlett-
Packard HP 5890 Series II gas chromatograph provided with
a fused silica capillary column coated with 5% phenyl methyl
silicone chemically bonded and cross-linked phase (HP-5MS,
0.25 mm i.d. × 30 m, 0.25 µm film, Hewlett-Packard, Avondale,
PA) was used. The following conditions were used: column
temperature, 250 °C; injection temperature, 265 °C; detection
temperature, 280 °C; carrier gas, He at 1 mL/min; inlet
pressure, 40 psi for 1 min, 40 psi to 20 psi, 99 psi/min; injection
mode, splitless.
Sa m p le P r ep a r a tion . The sample preparation was the
same as described before (Nagata and Saeki, 1992), with some
minor modifications; both evaporation procedures of ethyl
acetate extracts were performed at 50 °C. The drugs retained
to the Sep-Pak Florisil cartridge were eluted with 5 and 3 mL
of methanol-ethyl ether (3:7), successively. Eluate was
evaporated to dryness on a rotary evaporator at 35 °C.
A 10 g muscle sample was homogenized with 50 mL of ethyl
acetate and centrifuged at 2300g for 10 min, and the super-
natant was transferred to a round-bottom flask. The residue
mean 0.654 65.4 0.662 66.2 0.652 65.2 0.651 65.1 0.652 65.2 0.663 66.3
SD 0.047
CV 7.24
0.031
4.62
0.031
4.75
0.028
4.32
0.032
4.85
0.022
3.38
a
1 or 2 µg of each drug was added to 10 g of muscles.
was homogenized with another 50 mL of ethyl acetate and
centrifuged, and the supernatant was collected into the flask.
The extract was concentrated to 2-3 mL under vacuum on
rotary evaporator at 50 °C.
The residual solution was transferred with 25 mL of 3%
sodium chloride solution to a separatory funnel quantitatively.
Twenty-five milliliters of n-hexane was added to the separatory
funnel, shaken vigorously for 5 min, and allowed to stand until
layers were separated. The lower phase was transferred to
another separatory funnel, and the drugs were extracted with
40 mL of ethyl acetate twice. Each upper phase was collected
F igu r e 3. Mass spectrum of TMS derivative of CAP. Mass condition: ionization energy, 70 eV; electron multiplier, 1800 V;
emission current, 300 µA; scan range, 150-500.

1282 J. Agric. Food Chem., Vol. 44, No. 5, 1996
Nagata and Oka
F igu r e 4. Mass spectrum of TMS derivative of FF. The mass condition was the same as described in Figure 3.
F igu r e 5. Mass spectrum of TMS derivative of TAP. The mass condition was the same as described in Figure 3.
in a round-bottom flask, and the ethyl acetate layer was
evaporated to dryness under vacuum on a rotary evaporator
at 50 °C.
RESULTS AND DISCUSSION
The study on sample preparation has been described
in detail elsewhere (Nagata, 1995). Evaporation of ethyl
acetate extracts and eluate from the Sep-Pak Florisil
cartridge was performed at 50 and 35 °C, respectively,
resulting in satisfactory evaporating speed. Elution of
the three drugs was performed with 5 mL followed by
3 mL of methanol-ethyl ether (3:7) solution, resulting
in a minor improvement of the recoveries of the drugs,
especially TAP.
The residue was dissolved with 5 mL of n-hexane in an
ultrasonic bath and poured into a Sep-Pak Florisil cartridge,
and the cartridge was washed. The flask was rinsed with 5
mL of ethyl ether in the ultrasonic bath, and the rinsing was
poured into the cartridge and then the cartridge was washed.
The drugs were eluted with 5 and 3 mL of methanol-ethyl
ether (3:7), successively, and both eluates were collected and
then evaporated to dryness under vacuum on a rotary evapo-
rator at 35 °C.
Der iva tiza tion . The evaporated residue of the sample
solution was dissolved with 0.5 mL of acetonitrile and 0.4 mL
of BSA. The drugs were derivatized for 10 min at 50 °C and
excess solvents were evaporated to dryness using a gentle
stream of nitrogen. The derivatives were dissolved with 1 mL
of ethyl acetate, and 2 µL of this solution was injected into
the GC-MS system.
It was found that more than 8 min was required to
complete silylation of the three drugs, as shown in
Figure 2. On derivatization solvents, such as acetoni-
trile, ethyl acetate, and pyridine, which were used to
dissolve the drug residues, did not show any significant
difference.

Detection of CAP, FF, and TAP by GC
−MS
J. Agric. Food Chem., Vol. 44, No. 5, 1996 1283
F igu r e 6. SIM of blank yellowtail muscle extract. CAP at m/z 208 and 225 was scanned from 3.5 to 5 min (1, 2). FF and TAP
at m/z 257 were scanned from 5 to 10 min (4). TAP at m/z 242 and 330 was scanned from 7 to 10 min (3, 5).
F igu r e 7. SIM of drug-fortified yellowtail muscle extract. Two micrograms of drug was added to 10 g of yellowtail muscle. Scan
conditions were the same as described in Figure 6.
Mass spectra of derivatives of CAP, FF, and TAP
standards are shown in Figures 3-5. The silylated
molecular ion of each drug was not observed. With BSA
derivatization, two positions of CAP and TAP, as
described before (Nakagawa et al., 1975; Plomp and
Maes, 1976), and one position of FF were considered to
be trimethylsilylated. In CAP, m/z 453 and 451, in FF,
m/z 416 and 414, and in TAP, m/z 396 and 394,
contributed to M - CH3 (15) in each drug. In CAP, m/z
363 and 361, and in TAP, m/z 396 and 394, were
considered to be M - (CH3 + TMS - OH) (105) in both
drugs. For SIM analysis, the base peak m/z 225 and
additional peak m/z 208 for CAP, the base peak m/z 257
for FF, and the base peak m/z 242 and additional peaks
m/z 257 and 330 for TAP were selected. SIM chromato-
grams of blank and fortified yellowtail muscle extracts
are shown in Figures 6 and 7. As shown in Figure 6,
there were no interfering peaks observed in the blank
yellowtail muscle extract chromatogram. CAP, FF, and
TAP peaks were observed at 4.11, 5.68, and 8.21 min,
respectively, as shown in Figure 7.
The standard curves of CAP at m/z 225, of FF at m/z
257, and of TAP at m/z 242 and 257 were linear over
the range 0.05-2 µg/mL with correlation coefficients of
0.9987, 0.9987, 0.9982, and 0.9965, respectively. The
standard curves of CAP at m/z 208 and of TAP at m/z
330 were both linear at 1-2 µg/mL with correlation
coefficients of 0.9975 and 0.9979, respectively. The
detection limit, defined as the level of each drug in
animal tissue that produces a signal-to-noise ratio of 3,
was 5 ppb for CAP, FF, and TAP.
Recovery studies were performed by adding 1.0 or 2.0
mL of standard solution containing 1 µg/mL of CAP, FF,
or TAP to 10 g of minced yellowtail muscles. The
average recoveries of 0.1 and 0.2 ppm were more than
65%, as shown in Table 1. There were no significant
differences statistically between the recoveries at m/z

1284 J. Agric. Food Chem., Vol. 44, No. 5, 1996
Nagata and Oka
strategy for the regulatory control of residues of chloram-
phenicol in meat: preliminary studies in milk. J . Assoc.
Off. Anal. Chem. 1992, 75, 245-256.
Kijak, P. J . Confirmation of chloramphenicol residues in bovine
milk by gas chromatography/mass spectrometry. J . Assoc.
Off. Anal. Chem. 1995, 77, 34-40.
Nagata, T. Chemical analysis of chloramphenicols. In Chemi-
cal Analysis of Antibiotics in Agriculture; Oka, H., Naka-
zawa, H., Harada, K., Macneil, J . D., Eds.; AOAC Interna-
tional: Arlington, VA, 1995; pp 207-234.
Nagata, T.; Saeki, M. Simultaneous determination of thiam-
phenicol, florfenicol, and chloramphenicol residues in muscles
of animals and cultured fish by liquid chromatography. J .
Liq. Chromatogr. 1992, 15, 2045-2056.
Nakagawa, T.; Masada, M.; Uno, T. Gas chromatographic
determination and gas chromatographic-mass spectrometric
analysis of chloramphenicol, thiamphenicol and their me-
tabolites. J . Chromatogr. 1975, 111, 355-364.
Plomp, T. A.; Maes, R. A. A. A rapid gas chromatographic
determination of thiamphenicol in plasma and amniotic
fluid. J . Chromatogr. 1976, 121, 243-250.
Ramsey, E. D.; Games, D. E.; Startin, J . R.; Crews, C.; Gilbert,
J . Detection of residues of chloramphenicol in crude extracts
of fish and milk by tandem mass spectrometry. Biomed.
Environ. Mass Spectrom. 1989, 18, 5-11.
208 and 225 in CAP or among m/z 242, 257, and 330 in
TAP at the 95% confidence level.
This determination method is useful for the accurate
and specific detection of residual drugs.
LITERATURE CITED
Allen, E. H. Review of chromatographic methods for chloram-
phenicol residues in milk, eggs, and tissues from food-
producing animals. J . Assoc. Off. Anal. Chem. 1985, 68,
990-999.
Bories, G. S. F.; Peleran, J . C.; Wal, J . M. Liquid chromato-
graphic determination and mass spectrometric confirmation
of chloramphenicol residues in animal tissues. J . Assoc. Off.
Anal. Chem. 1983, 66, 1521-1528.
Borner, S.; Fry, H.; Balizs; Kroker, R. Confirmation of chloram-
phenicol residues in egg by gas chromatography/high-
resolution mass spectrometry and comparison of quantita-
tion with gas chromatography-electron capture detection.
J . Assoc. Off. Anal. Chem. 1995, 78, 1153-1160.
Delepine, B.; Sanders, P. Determination of chloramphenicol
in muscle using a particle beam interface for combining
liquid chromatography with negative-ion chemical ionization
mass spectrometry. J . Chromatogr. 1992, 582, 113-121.
Gazzaniga, A.; Pezzotti, E.; Ramusino, A. C. A rapid gas
chromatographic method for the determination of thiam-
phenicol in body fluids and tissues. J . Chromatogr. 1973,
81, 71-77.
Received for review J une 6, 1995. Revised manuscript received
J anuary 23, 1996. Accepted March 18, 1996.X
J anssen, G.; Vanderhaeghe, H. Preparation of trimethylsilyl
derivatives of chloramphenicol for gas-liquid chromatogra-
phy. J . Chromatogr. 1973, 82, 297-306.
Keukens, H. J .; Aerts, M. M. L.; Traag, W. A.; Nouws, J . F.
M.; Ruig, W. G.; Beek, W. M. J .; Hartog, J . M. P. Analytical
J F950343U
X Abstract published in Advance ACS Abstracts, May
1, 1996.