Hapten synthesis and development of polyclonal antibody-based multi-sulfonamide immunoassays

Article abstract of DOI:10.1021/jf060868u

This paper reports the synthesis of five sulfonamide derivatives, the production of broad-specificity polyclonal antibodies for immunoassay of sulfonamides, and the analysis of milk samples by developed assay. The three-step synthesis procedure reported in most of the literature was adopted and modified in this study. In the procedure, the purification of the intermediate was avoided and the time of synthesis was shortened from >20 to 6-9 h with improved yields. This method is generally applicable to the synthesis of haptens containing the common structure of sulfonamides. Three haptens were coupled to keyhole limpet hemocyanin, and polyclonal antibodies were obtained from rabbits immunized with these conjugates. Using the antibodies obtained, from one of these was developed an enzyme-linked immunosorbent assay (ELISA) based on the competition between free sulfonamides and the hapten-horseradish peroxidase (HRP) conjugates. The hapten-HRP conjugate giving the best competitive results and 11 structurally different sulfonamides showed 50% inhibition at concentrations of <100 ng mL^-1. After removal of the protein with acetone, milk samples were analyzed by ELISA directly; a matrix effect could be avoided when a 1:20 dilution with phosphate-buffered saline was used, and 104-131% recoveries of spiked samples were obtained. The developed immunoassay is suitable to determine sulfisozole, sulfathiazole, sulfameter, sulfamethoxypyridazine, sulfapyridine, and sulfamethizole below the maximum residue limit in milk (100 ng mL^-1 of total sulfonamides) rapidly and reliably.

Full text of DOI:10.1021/jf060868u

J. Agric. Food Chem. 2006, 54, 4499−4505 4499
Hapten Synthesis and Development of Polyclonal
Antibody-Based Multi-Sulfonamide Immunoassays
HONGYAN ZHANG, ZHENJUAN DUAN, LEI WANG, YAN ZHANG, AND
SHUO WANG*
Tianjin Key Laboratory of Food Nutrition and Safety, Faculty of Food Engineering and
Biotechnology, Tianjin University of Science and Technology, Tianjin 300222,
People’s Republic of China
This paper reports the synthesis of five sulfonamide derivatives, the production of broad-specificity
polyclonal antibodies for immunoassay of sulfonamides, and the analysis of milk samples by developed
assay. The three-step synthesis procedure reported in most of the literature was adopted and modified
in this study. In the procedure, the purification of the intermediate was avoided and the time of synthesis
was shortened from >20 to 6-9 h with improved yields. This method is generally applicable to the
synthesis of haptens containing the common structure of sulfonamides. Three haptens were coupled
to keyhole limpet hemocyanin, and polyclonal antibodies were obtained from rabbits immunized with
these conjugates. Using the antibodies obtained, from one of these was developed an enzyme-
linked immunosorbent assay (ELISA) based on the competition between free sulfonamides and the
hapten-horseradish peroxidase (HRP) conjugates. The hapten-HRP conjugate giving the best
competitive results and 11 structurally different sulfonamides showed 50% inhibition at concentrations
of <100 ng mL-1. After removal of the protein with acetone, milk samples were analyzed by ELISA
directly; a matrix effect could be avoided when a 1:20 dilution with phosphate-buffered saline was
used, and 104-131% recoveries of spiked samples were obtained. The developed immunoassay is
suitable to determine sulfisozole, sulfathiazole, sulfameter, sulfamethoxypyridazine, sulfapyridine, and
sulfamethizole below the maximum residue limit in milk (100 ng mL-1 of total sulfonamides) rapidly
and reliably.
KEYWORDS: Sulfonamides; milk; ELISA; hapten; polyclonal antibodies
INTRODUCTION
immunosorbent assay (ELISA) has been developed for detecting
several of the sulfonamides. Antibodies for the detection of
individual sulfonamides were produced by hapten conjugates,
which were prepared by modification at the N4-position, leaving
the specific structure of the sulfonamides unchanged. Different
immunoassay kits for the detection of sulfonamide residues in
food have been marketed using these antibodies. The broad-
specificity antibodies obtained by using sulfonamide derivatives,
differing at the N1-position, leaving the common structure of
the sulfonamides exposed, were expected to provide a general
assay for groups of sulfonamides. More effort is now being
directed to developing such broad-specificity antibodies for
detecting groups of sulfonamides.
In the past 20 years, several authors (5-9, 14) have designed
and synthesized several sulfonamide haptens that all had
carboxyl groups at the N1-position and kept the common
structure of sulfonamides unchanged. These haptens were mostly
synthesized according to a three-step procedure, and the
synthetic route was time-consuming, laborious, and low in yield.
In this paper, this three-step procedure was modified with
improvement in the performances.
Sulfonamides are widely used for therapeutic and prophylactic
purposes in humans and other animals and sometimes as
additives in animal feed (1). These synthetic compounds are
derivatives of sulfanilamide with an aromatic amino group at
the N4-position and differing in the substitution at the N1-
position. The backbone and the structures of sulfonamides
studied in this work are shown in Figure 1.
If the proper withdrawal periods are not observed before
slaughtering or milking of the medicated animals, meat or milk
from these animals may be contaminated with residual sulfon-
amides (2). The presence of sulfonamide residues in food is
considered harmful to consumers. In the European Union,
Canada, and the United States, the maximum residue limit
(MRL) of total sulfonamides in edible tissues is 100 µg kg-1
(3, 4), and it is 20 µg kg-1 in Japan (5).
Currently, Integral Production Chain Systems demand faster
on-site (farmhouses) and/or on-line (slaughterhouses) test
systems. Immunoassays are capable of detecting low amounts
of residues in many samples rapidly. An enzyme-linked
Sheth and Sporns (6) immunized rabbits with a sulfathiazole
derivative (N1-[4-(carboxymethyl)-2-thiazolyl]-sulfonamide) linked
* Corresponding author [telephone (86 22) 6060 1456; fax (86 22) 6060
1332; e-mail s.wang@tust.edu.cn].
10.1021/jf060868u CCC: $33.50
© 2006 American Chemical Society
Published on Web 06/03/2006

4500 J. Agric. Food Chem., Vol. 54, No. 13, 2006
Zhang et al.
Figure 1. Structures of the sulfonamides studied in this work.
to keyhole limpet hemocyanin (KLH). The polyclonal antibodies
recognized nine sulfonamides showing 50% inhibition at a con-
centration of <5 µg mL-1. Assil et al. (10) synthesized another
sulfonamide derivative (N1-[4-methyl-5-[2-4-carboxyethyl-1-
hydroxyphenyl]-azo-2-pyridyl]-sulfanilamide). The purified poly-
clonal antibody obtained showed 50% inhibition with seven
sulfonamides at concentrations of <10 µg mL-1. The first
published study on broad-specificity monoclonal antibodies
was from Muldoon et al. (11). After immunization with an
N-sulfanilyl-4-aminobenzoic acid-protein conjugate, only one
monoclonal antibody was obtained that recognized eight sul-
fonamides at levels of <10 µg mL-1. Haasnoot et al. (7) used
the sulfonamide derivatives of Sheth and Sporns (6) and Assil
et al. (10) to induce monoclonal antibodies, and one antibody
showed 50% inhibition with 18 tested sulfonamides at values
of <10 µg mL-1 and with eight at a concentration of <0.1 µg
mL-1
.
Korpimaki et al. (12) used protein engineering to improve
the broad specificity of sulfonamide antibodies. The new mutants

Multi-Sulfonamide Immunoassays
J. Agric. Food Chem., Vol. 54, No. 13, 2006 4501
recognized different sulfonamides with affinities sufficient for
the detection of all 13 tested sulfonamides below 100 ng mL-1
.
The aim of this study was the production of broad-specificity
polyclonal antibodies using various synthesized hapten-KLHs
as immunogens. With such antibodies a sensitive, broad-
specificity immunoassay was developed that could detect several
sulfonamides at relevant levels.
MATERIALS AND METHODS
Reagents. N-Acetylsulfanilyl chloride, 4-aminobenzoic acid, 2-amino-
4-thiazoleacetic acid, 6-aminonicotinic acid, and 4-aminobutanoic acid
were obtained from Fluka-Sigma-Aldrich Co. Pyridine and ethanol were
of analytical grade, dried, and purified before use. Horseradish
peroxidase (HRP), KLH, bovine serum albumin (BSA), reagent grade
3,3′,5,5′-tetramethylbenzidine (TMB), fish skin gelation (FG), and
Freund’s complete and incomplete adjutants were purchased from Sigma
(St. Louis, MO). Protein A-Sepharose 4B was purchased from
Amersham (Uppsala, Sweden). HPLC grade methanol were obtained
from Merck (Darmstadt, Germany). Reagent grade 1-(3-dimethyl-
aminopropyl)-3-ethyl carbodiimide hydrochloride) (EDC) and hydrogen
peroxide were from Sigma.
Solutions. Phosphate-buffered saline (PBS; 10 mmol L-1 sodium
phosphate, 137 mmol L-1 NaCl, 2.7 mmol L-1 KCl, pH 7.5), PBS
with 0.05% Tween 20 (PBS-T), coating buffer (CB; 50 mmol L-1
sodium carbonate buffer, pH 9.6), and TMB substrate solution [prepared
by adding 3.3 mg of TMB in 250 µL of DMSO to 25 mL of phosphate-
citrate buffer (0.1 mol L-1 citric acid + 0.2 mol L-1 Na2HPO4; pH
4.3) containing 3.25 µL of a 30% H2O2 solution] were used.
1
Instrumentation. H nuclear magnetic resonance (NMR) spectra
Figure 2. Chemical structures of D1−D5 haptens synthesized.
were recorded on Bruker AV-300 and Bruker AV-400 spectrometers,
and the chemical shifts were expressed in parts per million (δ scale)
using tetramethylsilane as an internal standard. Infrared spectroscopy
(IR) spectra were performed using a Bruker Vector 22 spectrometer.
The ESI-MS spectra were obtained on a Thermo-Finnegan LCQ
Advantage HPLC with mass spectrometry (HPLC-MS) instrument.
Thin-layer chromatography (TLC) was performed on silica gel G
detected by an ultraviolet-visible (UV) detector. For column chroma-
tography, silica gel 200-300 mesh was used. Polystyrene 96-well
microwell plates were from Nunc (Roskilde, Denmark), and the
microplate washer was from Bio-Rad (Hercules, CA). Immunoassay
absorbance was read with a Multiscan Spectrum purchased from
Thermo (Labsystems, Vantaa, Finland) in dual wavelength mode (450-
650 nm).
Na2SO4. The extract was evaporated under reduced pressure, giving a
yellow powder (0.6946 g, 59%), the (2-amino-1,3-thiazol-4-yl)acetic
acid ethyl ester: TLC [dichloromethane/MeOH (30:1)] Rf ) 0.45; H
NMR (300 MHz, CDCl3) δ 6.341 (s, 1H, CHar-S,), 5.166 (br, 2H, NH2),
4.180 (q, 2H, J ) 7.2 Hz, OCH2CH3), 3.560 (s, 2H, COCH2), 1.270 (t,
3H, J ) 7.2 Hz, CH3); MS (ESI) [found, m/z 187.12 [M + H]+; calcd
for C7H10N2O2S: M, 186]; IR (KBr, cm-1) 3420, 3282, 3144, 2990,
1717, 1626, 1525, 1474, 1448, 1423, 1371, 1337, 1311, 1265, 1172,
1130, 1033, 694.
1
Synthesis of [2-(4-Aminobenzenesulfonylamino)-1,3-thiazol-4-yl]ace-
tic Acid (D1). The above-described general procedure was applied to
prepare hapten D1; 0.3963 g of (2-amino-1,3-thiazol-4-yl)acetic acid
ethyl ester (2.12 mmol) was used. The residue was a brown-red oil
and then purified in silica gel columns using methanol/chloroform/ice
acetic acid (1:20:0.04) as eluant, giving product D1 (0.1654 g, 24.9%
yield, yellow powder): TLC [methanol/chloroform/ice acetic acid (1:
Haptens Synthesis. The structures of the haptens designed are shown
in Figure 2.
General Procedure for the Preparation of Haptens D1-D5. N-
Acetylsulfanilyl chloride (0.5328 g, 2.28 mmol) was dissolved in dry
pyridine (5 mL), and a solution of acid or ester including an amino
group (2.12 mmol) in 6 mL of dry pyridine was dropwise added under
N2 atmosphere with stirring. After addition, the reaction mixture was
refluxed at 90 °C for 2-3 h, and the reaction was monitored by TLC.
Then 20 mL of 2 M NaOH was added, and the pyridine was removed
under reduced pressure. The reaction mixture was refluxed for an
additional 2-3 h, and the reaction was monitored by TLC. Afterward,
the mixture was cooled to room temperature, and the pH was adjusted
to 4.0 with concentrated HCl. The mixture was extracted with ethyl
acetate (3 × 30 mL). The organic phase was dried over anhydrous
Na2SO4 and evaporated under reduced pressure. The residue was further
purified on a silica gel column (200-300 mesh), and the elution
conditions were established according to the TLC results.
Synthesis of Hapten [2-(4-Aminobenzenesulfonylamino)-1,3-thiazol-
4-yl]acetic Acid (D1)]. Synthesis of (2-Amino-1,3-thiazol-4-yl)acetic
Acid Ethyl Ester (M1). Ethanol (18.4 mL) and concentrated HCl (3.7
mL) were added to 1.00 g of 2-amino-4-thiazoleacetic acid (6.3 mmol)
with stirring. The mixture was heated under reflux for 2-3 h with
stirring, after which the solvent was removed under reduced pressure.
The residue was dissolved in 19 mL of distilled water and adjusted to
pH 8.0 with 6.5 mol L-1 NH4OH. The mixture was extracted with ethyl
acetate (3 × 20 mL), and the organic phase was dried over anhydrous
1
10:0.02)]: Rf ) 0.35; H NMR (300 MHz, DMSO-d6) δ 12.392 (br,
2H, COOH, SO2NH), 7.457 (d × t, 2H, J ) 8.7 Hz, 2.3 Hz, CHar),
6.584 (d × t, 2H, J ) 8.7 Hz, 2.3 Hz, CHar), 6.533 (s, 2H, NH2), 3.534
(s, 1H, CH-S), 1.924 (s, 2H, CH2COOH); MS (ESI) [found: m/z 314.12
[M + H]+, calcd for C11H11N3O4S2, M, 313.35]; IR (KBr, cm-1) 3476,
3383, 3108, 2925, 1717, 1697, 1624, 1595, 1541, 1501, 1419, 1340,
1294, 1269, 1241, 1186, 1138, 1086, 1023, 885, 829, 738, 715, 691,
620, 572, 552.
Synthesis of Hapten 6-(4-Aminobenzenesulfonylamino)nicotinic Acid
(D2). Synthesis of 6-Aminonicotinic Acid Ethyl Ester (M2). M2 was
prepared as the route of M1 preparation, and a white powder was
obtained (0.3583 g, 75%): TLC [dichloromethane/methanol (16:1)] Rf
1
) 0.66; H NMR (300 MHz, CDCl3) δ 8.734 (d, 1H, J ) 2.1 Hz,
CHar-N), 8.023 (dd, 1H, J ) 2.1, 9.0 Hz, CHar-CCO), 6.475 (d, 1H, J )
9.0 Hz, CHar-CNH ), 4.954 (br, 2H, NH2), 4.341 (q, 2H, J ) 7.2 Hz,
2
CH3CH2O), 1.372 (t, 3H, J ) 7.2 Hz, CH3CH2O); MS (ESI) [found:
m/z 166.37 [M]+, calcd for C8H10N2O2, M, 166]; IR (KBr, cm-1) 3414,
3321, 3142, 2977, 1692, 1652, 1602, 1515, 1444, 1370, 1277, 1138,
950, 836, 783.
Synthesis of 6-(4-Aminobenzenesulfonylamino)nicotinic Acid (D2).
The above-described general procedure was applied to prepare the
hapten D2. For that, 0.3534 g of 6-aminonicotinic acid ethyl ester (2.12

4502 J. Agric. Food Chem., Vol. 54, No. 13, 2006
Zhang et al.
mmol) was used. The residue was a yellow powder. Further purification
in silica gel columns using methanol/chloroform/ice acetic acid (1:20:
0.04) as eluant finally gave the product D2 (0.1564 g, 25.2% yield,
yellowish powder): TLC [methanol/chloroform/ice acetic acid (5:5:
Antibody Production. Antibodies were produced in rabbits as
described by Wang et al. (13). Six white rabbits (two for each
immunogen) were immunized by intradermal and intramuscular injec-
tions of the immunogens D1-KLH, D3-KLH, and D4-KLH. IgG
from the antisera was purified by Protein A-Sepharose 4B affinity
chromatography.
Screening of Antisera. Antisera titers, antibody coating quantity,
and enzyme conjugate dilution factors were optimized to give absorb-
ance values ranging from 0.7 to 1.2 in the absence of analytes using a
checkerboard titration. The antisera titers were tested for D1-KLH on
D1-OA coating, for D3-KLH on D3-OA coating, and for D4-KLH
on D4-OA coating using antigen-coated indirect ELISA format, and
the antisera were diluted from 1/1000 to 1/200000 to microtiter plates
coated with 1 µg per well of the hapten-OA conjugate. Similarly,
antibody coating quantity and enzyme conjugate dilution factors were
tested on an antibody-coated direct ELISA format using enzyme
conjugate dilutions from 1/1000 to 1/100000 and plates coated with
0.5, 1, or 1.5 µg per well of the antibodies
Indirect ELISA. Flat-bottom polystyrene ELISA plates were coated
with OA-hapten (D1-D5) conjugates at 1 µg per well in 100 µL of
CB and incubated overnight at room temperature. Plates were then
washed three times with 10 mmol L-1 PBS-T, and unbound active sites
were blocked with 200 µL of 1% BSA/PBS per well for 1 h. After the
plate had been washed four times, 100 µL of the appropriate sera
dilution in PBS was added per well and incubated for 1 h at room
temperature. After four washings, plates were incubated for 1 h with
peroxidase-labeled goat anti-rabbit immunoglobulins diluted 1:10000
in PBS (100 µL per well). After five washings, the HRP tracer activity
was measured by adding 150 µL per well of TMB substrate solution.
The enzymatic reaction was stopped after 30 min by adding 2.5 mol
L-1 H2SO4 (50 µL per well), and the absorbance was read in dual-
wavelength mode (450 nm as test and 650 nm as reference).
Direct ELISA. Polystyrene ELISA plates were coated with purified
antibodies at 0.5, 1, or 1.5 µg per well in 100 µL of CB, incubated
overnight at room temperature. Plates were then washed three times
with 10 mmol L-1 PBS-T, and unbound active sites were blocked with
200 µL of 1% BSA/PBS per well for 1 h. After the plate had been
washed four times, for competitive assays, 100 µL of standards in PBS
and 100 µL of HRP-haptens in PBS were then added to each well
and incubated for 1 h at room temperature. After five washings, the
HRP tracer activity was measured by adding 150 µL per well of TMB
substrate solution. The enzymatic reaction was stopped after 30 min
by adding 2.5 mol L-1 H2SO4 (50 µL per well), and the absorbance
was read in dual-wavelength mode (450 nm as test and 650 nm as
reference).
1
0.04)] Rf ) 0.76; H NMR (400 MHz, DMSO-d6) δ 12.167 (br, 1H,
COOH), 8.581 (s, 1H, SO2NH), 8.548 (d, 1H, J ) 4.4 Hz, CHar-N),
8.057 (d × d, 1H, J ) 8.8 Hz, 2.4 Hz, CHar-CCO), 7.537 (d, 2H, J )
8.8 Hz, CHar), 7.070 (d, 1H, J ) 8.8 Hz, CHar-CNH), 6.532 (d, 2H, J )
8.8 Hz, CHar), 6.029 (s, 2H, NH2); MS (ESI) [found: m/z 294.17 [M
+ H]+, calcd for C12H11N3O4S, M, 293]; IR (KBr, cm-1) 3474, 3375,
3234, 1707, 1635, 1595, 1538, 1505, 1383, 1298, 1276, 1139, 1058,
959, 837, 770, 681, 666, 583, 554.
Synthesis of Hapten 4-(4-Aminobenzenesulfonylamino)benzoic Acid
(D3). The above-described general procedure was applied to prepare
the hapten D3. For that, 0.2900 g of 4-aminobenzoic acid (2.12 mmol)
was used. The residue was a yellow oil, and further purification in
silica gel columns using methanol/chloroform/ice acetic acid (1:40:
0.08) as eluant finally gave the product D3 (0.4445 g, 71.8% yield,
yellow powder): TLC [methanol/chloroform/ice acetic acid (1:20:0.04)]
1
Rf ) 0.25; H NMR (400 MHz, DMSO-d6) δ 7.803 (d, 2H, J ) 8.0
Hz, CHar), 7.479 (d, 2H, J ) 8.0 Hz, CHar), 7.162 (d, 2H, J ) 8.0 Hz,
CHar), 6.575 (d, 2H, J ) 8.4 Hz, CHar), 6.071 (s, 2H, NH2); MS (ESI)
[found: m/z 606.86 [2M + Na]+, calcd for C13H12N2O4S, M, 292]; IR
(KBr, cm-1) 3494, 3464, 3391, 3373, 3167, 2930, 1698, 1633, 1608,
1595, 1505, 1460, 1407, 1324, 1291,1235, 1187, 1148, 1092, 1018,
912, 834, 769, 718, 678, 658, 578, 548.
Synthesis of Hapten 6-(6-Aminobenzenesulfonylamino)hexanoic Acid
(D4). The above-described general procedure was applied to prepare
the hapten D4. For that, 0.2777 g of 6-aminohexanoic acid (2.12 mmol)
was used. The residue was a yellow oil and purified in silica gel columns
using methanol/chloroform/ice acetic acid (1:20:0.04 and 1:10:0.02)
as eluant, finally giving the product D4 (0.2158 g, 35.6% yield, milky
powder): TLC [methanol/chloroform/ice acetic acid (1:10:0.02)] Rf )
0.61; 1H NMR (300 MHz, DMSO-d6) δ 11.957 (br, 1H, COOH), 7.403
(d × t, 2H, J ) 8.7 Hz, 2.4 Hz, CHar), 7.065 (t, 1H, J ) 6.0 Hz,
SO2NH), 6.607 (d × t, 2H, J ) 8.7 Hz, 2.3 Hz, CHar), 5.923 (s, 2H,
NH2), 2.519 (d × d, 2H, J ) 6.8 Hz, 6.3 Hz, NHCH2), 2.156 (t, 2H,
J ) 7.4 Hz, CH2COOH), 1.471-1.159 (m, 6H, CH2CH2CH2); MS(ESI)
[found: m/z 595.10 [2M + Na]+, calcd for C12H18N2O4S, M, 286];
IR(KBr, cm-1) 3413, 3326, 3126, 2937, 2859, 1699, 1627, 1596, 1502,
1467, 1429, 1304, 1250, 1231, 1197, 1143, 1095, 1036, 1013, 985,
936, 846, 678, 581, 563, 546.
Synthesis of Hapten 6-(4-Aminobenzenesulfonylamino)butanoic Acid
(D5). The above-described general procedure was applied to prepare
the hapten D5. For that, 0.2184 g of 4-aminobutanoic acid (2.12 mmol)
was used. The residue was a yellow oil and purified in silica gel columns
using methanol/chloroform/ice acetic acid (1:20:0.04, 1:10:0.02 and
1:5:0.01) as eluant, finally giving the product D5 (0.1538 g, 28.1%
yield, yellowish powder): TLC [methanol/chloroform/ice acetic acid
(1:10:0.02)] Rf ) 0.26; 1H NMR (300 MHz, DMSO-d6) δ 12.109 (br,
1H, COOH), 7.423 (d, 2H, J ) 8.7 Hz, CHar), 7.151 (t, 1H, J ) 5.7
Hz, SO2NH), 6.630 (d, 2H, J ) 8.7 Hz, CHar), 5.957 (s, 2H, NH2),
2.683 (d × d, 2H, J ) 6.0 Hz, 6.8 Hz, 6.3 Hz, NHCH2), 2.226 (t, 2H,
J ) 7.2 Hz, CH2COOH), 1.594 (m, 2H, CH2CH2CH2); MS (ESI)
[found: m/z 258.87 M + H]+, calcd for C10H14N2O4S, M, 258]; IR
(KBr, cm-1) 3413, 3336, 3127, 2950, 2622, 1907, 1699, 1633, 1596,
1501, 1432, 1305, 1261, 1216, 1188, 1147, 1090, 1012, 965, 896, 841,
682, 577, 548, 509.
Preparation of Enzyme and Protein Conjugates. The sulfonamide
derivatives were attached to the carrier proteins (KLH, OA) or HRP
using a mixed anhydride method using EDC without isolation of
intermediates. Haptens D1, D3, and D4 were coupled to KLH for use
as immunogens. All five haptens were coupled to OA as coating
proteins, coupled to HRP as enzyme tracers, respectively.
The carboxylic acid of the hapten (0.025 mmol) and 0.05 mmol of
EDC were dissolved in 1 mL of dry DMF. Then mixture was added
dropwise with mixing to 2 mL of solution of 10 mg of protein (KLH,
OA, or HRP) in 130 mmol L-1 sodium bicarbonate (pH 8.1). After
4-6 h of mixing at room temperature on a mixing wheal, another 4.7
mg of EDC was added. After mixing overnight at 4 °C on the mixing
wheel, the mixture was dialyzed against PBS.
Sample Preparation. Milk samples were bought from local markets.
Before the spike and recovery studies, each test sample was verified
to contain sulfonamides at <10 ng mL-1 by HPLC. For a spiking study,
20 mL milk samples were spiked with a single sulfonamide dissolved
in methanol at different levels, thoroughly mixed, and then centrifugated
for 10 min at 3000 rcf. After the fat layer had been discarded, 20 mL
of acetone was added and shaken by hand, then centrifugated for 10
min at 3000 rcf. The upper clear liquid was diluted with PBS and
analyzed.
The sample extracts were appropriately diluted before ELISA
analysis, and 1% BSA, 0.5% FG, or 0.05% Tween 20 was added to
the PBS to avoid the sample matrix effect.
Instrumentation for HPLC Analysis. The test sample was verified
using a Shimadzu (Tokyo, Japan) HPLC equipped with an LC-10AT
vp pump with a Hamilton injector (25 µL loop), a DGU-12A online
degasser, and a CTO-10AS vp column oven. A C18 reversed-phase
column (15 cm × 4.6 mm i.d., 5 µm) was used. The analyses were
performed at 270 nm, and the mobile phase was methanol/water (28:
72) (water pH value was adjusted to 3.5 before it was mixed with
methanol) at a flow rate of 1.0 mL min-1. The temperature of the
column oven was 35 °C.
RESULTS AND DISCUSSION
Hapten Synthesis. Five different sulfonamide haptens (D1-
D5) were previously reported for the detection of sulfonamides

Multi-Sulfonamide Immunoassays
J. Agric. Food Chem., Vol. 54, No. 13, 2006 4503
Table 1. Titers of the Antisera during the Immunization
bleed
D1−KLH−1
D1−KLH−2
D3−KLH−1
D3−KLH−2
D4−KLH−1
D4−KLH−2
first
second
third
6000
died
2000
1000
1000
1000
64000
80000
died
64000
died
3000
8000
20000
20000
160000
200000
100000
fourth
Table 2. Optimization of the Coating Antibody Quantity and Enzyme Conjugate Dilution Factors
D3−KLH−2
D4−KLH−2
D1−KLH−2
0.5
µ
g/well
1.0
µ
g/well
1.5
µ
g/well
0.5
µ
g/well
1.0
µ
g/well
1.5
µ
g/well
0.5
µ
g/well
1.0
µ
g/well
1.5
µ
g/well
D1−
D2−
D3−
D4−
D5−
HRP
HRP
HRP
HRP
HRP
<1000
2000
10000
2000
<1000
8000
40000
8000
<1000
10000
40000
10000
<1000
<1000
<1000
<1000
1000
<1000
<1000
<1000
3000
<1000
<1000
<1000
3000
<1000
<1000
<1000
<1000
<1000
<1000
<1000
<1000
<1000
<1000
<1000
<1000
<1000
<1000
<1000
<1000
<1000
<1000
<1000
<1000
(5-8, 11, 12). They conjugated the haptens with carrier protein
through the R group and produced group-specific antibodies.
In previous works, these haptens were synthesized using three
steps, namely, esterification of carboxylic acid compounds,
condensation with N-acetylsulfanilyl chloride, and then basic
hydrolysis to give the target molecule product. This reported
method has two disadvantages: First, overall reaction times of
D1-D3 described by Pastor-Navarro et al. were over 22, 44,
and 31 h, respectively. The second step was the most time-
consuming step and required >20 h, with reaction first in an
ice bath and then at room temperature. Second, the procedure
was more laborious because the product in each step had to be
separated and purified.
In this work, we adopted and integrated the three-step
procedure reported in most of the literature. Because the third
step also took place in basic conditions, we modified the above
method; namely, the pyridine in the product of the second step
was evaporated, NaOH solution was added, and the mixture
was refluxed to hydrolyze. This route only needed 6-9 h under
the condition of reflux, which saved reaction time.
The raw materials of haptens D3, D4, and D5 could dissolve
in pyridine, so they were used to synthesize these haptens
directly and the step of esterification was left out. The synthesis
of haptens D1 and D2 using this method gave poor yield,
perharps due to the poor solubility of the raw materials
(2-amino-4-thiazoleacetic acid and 6-aminonicotinic acid).
Therefore, these raw materials were esterified first, and
then the esters were used to synthesize these haptens. The yields
of the synthesized sulfonamide derivatives [2-(4-amino-
benzenesulfonylamino)-1,3-thiazol-4-yl]acetic acid (D1), 6-(4-
aminobenzenesulfonylamino)nicotinic acid (D2), 4-(4-amino-
benzenesulfonylamino)benzoic acid (D3), 6-(6-aminobenzene-
sulfonylamino)hexanoic acid (D4), and 6-(4-aminobenzene-
sulfonylamino)butanoic acid (D5) were 24.9, 25.2, 71.8, 35.6,
and 28.1%, respectively. Although the final chromatography was
laborious, this step assured the purification of the end product
and led to the good quality of the antibodies.
Antibody Production and Screening of Antisera. Table 1
shows the optimal results of the antisera titers (the dilution of
a serum containing a specific antibody at which the solution
retains the minimum level of activity needed to neutralize or
precipitate an antigen) during the immunization. The titers from
the first three bleeds for D1-KLH-2, D3-KLH-2, and D4-
KLH-2 were 1000, 100000, and 20000, respectively.
in Table 2. This shows that the antibody of D1-KLH-2 did
not recognize all five enzyme conjugates and that the antibody
of D4-KLH-2 could recognize only the D4-HRP conjugate,
but the dilution factors were too low. The antibody of D3-
KLH-2 recognized D2-HRP, D3-HRP, and D4-HRP with
relatively high dilution factors. To balance the absorbance value
and save the antibody, we selected the coating antibody with a
coating quantity of 1.0 µg per well, and the corresponding
enzyme conjugate dilution factors were 8000, 40000, and 8000
for D2-HRP, D3-HRP, and D4-HRP, respectively.
Analytical Characteristics of the Sulfonamide ELISA. The
broad specificity of the developed ELISA was evaluated by
determining the cross-reactivity with a set of sulfonamides. The
concentrations of analytes giving 50% inhibition of color
development (IC50) and cross-reactivity were determined for
different sulfonamides in buffer solution, and the values obtained
are given in Table 3. The results showed that direct ELISA
developed with the antibody against D3-KLH and the D2-
HRP conjugate showed high sensitivity and good broad
specificity. Using this combination, the IC50 values of sulfisox-
azole, sulfathiazole, sufameter, sulfapyridine, sulfamethoxy-
pyridazine, sulfachlorpyridazine, sulfamethizole, sulfamerazine,
sulfadoxine, sulfadiazine, and sulfaquinoxaline were below 100
ng mL-1 in buffer solution. These results were generally more
sensitive than those of polyclonal antibodies reported previously
by other researchers: Sheth and Sporns (6) reported IC50 values
of 9 sulfonamides ofwere <5 µg mL-1; Assil et al. (10) reported
IC50 values of 7 sulfonamides of <10 µg mL-1), better than
those for monoclonal antibodies reported by Muldoon et al. (11)
(IC50 values of 8 sulfonamides were <10 µg mL-1); Haasnoot
et al. (7) reported IC50 values of 18 tested sulfonamides of <10
µg mL-1 and of 8 sulfonamides of <0.1 µg mL-1 and equal to
those reported by Korpima¨ki et al. (12) (could detect all 13 tested
sulfonamides below 100 ng mL-1). The cross-reactivities for
the 11 sulfonamides were between 1.4 and 100%.
The intra-assay and interassay reproducibilities were deter-
mined to study the assay precision. The variations of percent
inhibition for 20, 10, 5, 2.5, 1.25, 0.63, and 0.31 ng mL-1 of
sulfisozole tested three times on the same day were 3.8, 3.1,
4.2, 7.5, 15.8, 24.4, and 31.1%, respectively. The assay of the
same material run over 6 months gave deviations from the means
of 5.9, 9.2, 11.3, 11.9, 29.5, and 35.7%. The average detection
limit for buffer curves was defined as the concentration of
sulfonamides causing 15% inhibition of color development.
Using this criterion, the limit of detections (LOD) for sulfisozole,
sulfathiazole, sufameter, sulfamethoxypyridazine, sulfapyridine,
The results of optimization of antibody coating quantity and
enzyme conjugate dilution factors using direct ELISA are shown

4504 J. Agric. Food Chem., Vol. 54, No. 13, 2006
Table 3. Assay Sensitivities of Antibody Obtained from D3
Zhang et al.
KLH-Immunized Rabbit for 16 Sulfonamides Using Different Antigen−HRP Conjugates
−
D2
−
HRP
cross-reactivity
(%)
D3
IC50 (ng mL
−
HRP
cross-reactivity
(%)
D4
−
HRP
cross-reactivity
(%)
-1
-1
-1
IC50 (ng mL
)
)
IC50 (ng mL
)
sulfisozole
sulfathiazole
sufameter
sulfamethoxypyridazine
sulfapyridine
sulfamethizole
sulfachlorpyridazine
sulfamerazine
sulfadoxine
1.3
2
3.2
8.4
10
12.8
14
19
31
32
90
130
135
150
190
>200
100
65
41
15
13
10
9
7
4
4
1.4
1
12
15
25
20
25
130
55
85
150
150
175
200
175
200
80
100
80
48
60
48
9
22
14
8
8
7
6
7
18
25
19
32
28
100
72
95
56
64
14
26
12
<9
9
<9
<9
<9
<9
<9
<9
125
68
155
>200
190
>200
>200
>200
>200
>200
>200
sulfadiazine
sulfaquinoxaline
sulfamethazine
sulfamethoxazole
sulfamoxole
sulfadimethoxine
sulfisoxazole
1
0.9
0.7
<0.7
6
15
<6
>200
Figure 4. Standard curves of sulfisozole.
Table 4. Recoveries of Sulfonamides Spiked at Three Concentrations
in Milk Samples
fortification
level (ng mL
mean
(ng mL
±
SD
)
recovery
(%)
CV
(%)
-1
-1
analyte
)
sulfisozole
50
100
200
52.47
±
6.27
7.53
19.68
104.9
111.9
119.0
11.9
6.73
8.27
111.9
237.9
±
±
sulfathiazole
sufameter
50
100
200
57.96
±
11.20
6.22
16.91
115.9
113.6
104.0
19.3
5.48
8.12
113.6
208.1
±
±
50
100
200
55.60
±
8.95
8.48
15.88
111.2
124.4
130.9
16.1
6.82
6.06
Figure 3. Standard curves of six sulfonamides detected in milk below
the MRL.
124.4
261.9
±
±
sulfamethoxy-
pyridazine
50
100
200
58.90
±
7.01
6.06
13.29
117.8
119.2
123.5
11.9
5.08
5.38
sulfamethizole, sulfachlorpyridazine, sulfamerazine, sulfadoxine,
sulfadiazine, sulfaquinoxaline, sulfamethazine, sulfamethox-
azole, sulfamoxole, sulfadimethoxine, and sulfisoxazole were
0.1, 0.2, 0.4, 1.5, 1.6, 2.1, 3.0, 6.2, 6.8, 7.0, 10, 22, 23, 27, 27,
and 80 ng mL-1, respectively.
119.2
247.0
±
±
sulfapyridine
50
100
200
62.07
±
5.74
5.11
14.52
124.1
128.2
130.4
9.25
3.99
5.57
128.2
260.8
±
±
sulfamethizole
50
100
200
53.74
±
12.8
8.37
9.91
107.5
116.9
112.8
23.8
7.16
4.39
Milk Sample Analysis. In the course of sulfonamide analysis
in milk, a nine-point standard curve was included in each ELISA
plate to estimate analyte concentrations. To gain basic informa-
tion on matrix effect, conformity of standard curves generated
in PBS was compared with that of curves obtained using milk
matrices. The inhibition curves generated in PBS were consistent
with those prepared in 10, 20, or 100 times diluted milk matrices,
but the OD value curves differed significantly. After removal
of the protein with acetone, the milk samples were diluted 10,
20, or 50 times with PBS, 1% BSA in PBS, 0.5% FG in PBS,
or 0.05% Tween 20 in PBS, respectively. The dilution of 20
times with PBS was found to be good to reduce matrix effects
for the detection of sulfisozole, sulfathiazole, sufameter, sulfa-
methoxypyridazine, sulfapyridine, sulfamethizole, and only these
116.9
225.5
±
±
six sulfonamides could be detected below MRL after dilution.
The standard curves of these six sulfonamides are shown in
Figure 3, and the standard curves of sulfisozole in PBS or milk
samples are shown in Figure 4. The results of the recoveries
of the six sulfonamides added to milk at 50, 100 (the maximum
residue level of sulfonamides in milk), and 200 ng mL-1 against
a calibration curve set up in buffer are shown in Table 4.
The average recoveries (n ) 4) obtained from the matrices
were found in the range of 104-131%. The interassay coef-
ficient of variation (CV) for 100 ng mL-1 diluted milk, which

Multi-Sulfonamide Immunoassays
J. Agric. Food Chem., Vol. 54, No. 13, 2006 4505
(3) EU Regulation 508/1999, L60(9-3-1999), 16-52.
(4) Food and Drug Regulation. Can. Gaz., Part II 1991, 125 (Table
3, Division 15, Part B), 1478-1480.
(5) Pastor-Navarro, N., et al. Specific polyclonal-based immuno-
assays for sulfathiazole. Anal. Bioanal. Chem. 2004, 379, 1088-
1099.
(6) Sheth, H. B.; Sporns, P. Development of a single ELISA for
detection of sulfonamides. J. Agric. Food Chem. 1991, 39,
1696-1700.
(7) Haasnoot, W.; Cazemier, G.; Du, P. J.; Kemmers-Voncken, A.;
Bienenmann-Ploum, M.; Verheijen, R. Sulfonamide antibodies:
from specific polyclonals to generic monoclonals. Food Agric.
Immunol. 2000, 12, 15-30.
(8) Haasnoot, W.; du Pre´, J.; Cazemier, G.; Kemmers-Voncken, A.;
Verheijen, R.; Janssen, B. J. M. Monoclonal antibodies against
a sulfathiazole derivative for the immunochemical detection of
sulfonamides. Food Agric. Immunol. 2000, 12, 127-138.
(9) Li, J. S.; Li, X. W. Synthesis and structure elucidation of
sulfonamide haptens. J. China Agric. UniV. 1999, 4, 109-
113.
(10) Assil, H. I.; Sheth, H.; Sporns, P. An ELISA for sulfonamide
detection using affinity-purified polyclonal antibodies. Food Res.
Int. 1992, 25, 343-353.
(11) Muldoon, M. T.; Font, I. A.; Beier, R. C.; Holtzapple, C. K.;
Young, C. R.; Stanker, L. H. Development of a cross-reactive
monoclonal antibody to sulfonamide antibiotics: evidence for
structural conformation-selective hapten recognition. Food Agric.
Immunol. 1999, 11, 117-134.
was the concentration of interest, was below 10%; intra-assay
variabilities were below 15%. The competitive direct ELISA
presented here could be used to determine sulfisozole, sulfa-
thiazole, sufameter, sulfamethoxypyridazine, sulfapyridine, and
sulfamethizole in milk simultaneously.
Real milk samples were bought from local markets for
sulfonamide screening and determination. Analysis of the 60
food samples using the direct ELISA resulted in 9 samples being
positive (sulfonamide concentration was greater than the LOD
of the assay). The confirmation step with the HPLC analysis
also resulted in detection of the same positive samples,
demonstrating that the developed assay can be used for real
sample analysis.
Conclusion. This synthetic method of sulfonamide haptens
avoided the purification of the intermediate, saved most of the
reaction time, and achieved higher yield. It is a good method
for the synthesis of haptens that expose the common structures
of sulfonamides. The developed ELISA method was capable
of detecting 11 of the tested 16 sulfonamides within a
concentration range of 1.3-90 ng mL-1 in buffer. After the
deposition of protein using acetone, milk samples were analyzed
by ELISA directly; the matrix effect could be avoided when
1:20 dilution with phosphate-buffered saline was used. The
developed immunoassay is suitable to determine sulfisozole,
sulfathiazole, sufameter, sulfamethoxypyridazine, sulfapyridine,
and sulfamethizole in milk rapidly and reliably.
(12) Korpima¨ki, T.; Brockmann, E.-V.; Kuronen, O.; Sarate, M.;
Lamminma¨ki, U.; Tuomola, M.; et al. Engineering of a broad
specificity antibody for simultaneous detection of 13 sulfon-
amides at the maximum residue level. J. Agric. Food Chem.
2004, 52, 40-47.
(13) Wang, S.; Allan, R. D.; Skerritt, J. H.; Kennedy, I. R. Develop-
ment of a class-specific competitive ELISA for the benzoyl-
phenylurea insecticides. J. Agric. Food Chem. 1998, 46, 3330-
3338.
ABBREVIATIONS USED
ELISA, enzyme-linked immunosorbent assay; HPLC, high-
performance liquid chromatography; IR, infrared spectroscopy
spectra; MS, mass spectrometry; TLC, thin-layer chromatog-
raphy; UV, ultraviolet-visible; MRL, maximum residue limit;
LOD, limit of detection; HRP, horseradish peroxidase; KLH,
keyhole limpet hemocyanin; BSA, bovine serum albumin; TMB,
3,3′,5,5′-tetramethylbenzidine; EDC, 1-(3-dimethylamino-
propyl)-3-ethyl carbodiimide hydrochloride; IC50, concentration
of analyte giving 50% inhibition of color development; PBS,
phosphate-buffered saline; PBST, phosphate-buffered saline with
0.05% Tween 20; CB, coating buffer.
(14) Li, J. S.; Li, X. W.; Yuan, J. X.; Wang, X. Determination of
sulfonamides in swine meat by immunoaffinity chromatography.
J. AOAC Int. 2000, 83, 831-836.
Received for review March 28, 2006. Revised manuscript received May
5, 2006. Accepted May 8, 2006. We are grateful for financial support
from the New Century Talent Program of the Ministry of Education
of the People’s Republic of China (Project NECT-04-0243), the Scientific
Research of Refund of Tianjin University of Science and Technology,
China (Project 20050204), and the Fok Ying Tung Education Founda-
tion (101073).
LITERATURE CITED
(1) Long, A. R.; Hsieh, L. C.; Malbrough, M. S.; Short C. R.; Barker,
S. A. Multiresidue method for the determination of sulfonamides
in pork tissue. J. Agric. Food Chem. 1990, 38, 423-426.
(2) Franco, D. A.; Webb, J.; Taylor, C. E. Antibiotic and sulfonamide
residues in meat: implication for human health. J. Food Prot.
1990, 53, 178-185.
JF060868U