Preparation of anti-tetracycline antibodies and development of an indirect heterologous competitive enzyme-linked immunosorbent assay to detect residues of tetracycline in milk

Article abstract of DOI:10.1021/jf062627s

Tetracycline (TC) is a broad-spectrum antibiotic used increasingly in animal husbandry to treat diseases or to promote growth as feed additives. To avoid using labor-intensive instrumental methods to detect residues of TC in food and food products, a simp

Full text of DOI:10.1021/jf062627s

J. Agric. Food Chem. 2007, 55, 211−218 211
Preparation of Anti-Tetracycline Antibodies and Development of
an Indirect Heterologous Competitive Enzyme-Linked
Immunosorbent Assay to Detect Residues of Tetracycline
in Milk
YULAN ZHANG, SHENGXIN LU, WEI LIU, CHENGBIAO ZHAO, AND RIMO XI*
The School of Chemistry and Chemical Engineering, Shandong University, Jinan,
Shandong 250100, People’s Republic of China
Tetracycline (TC) is a broad-spectrum antibiotic used increasingly in animal husbandry to treat diseases
or to promote growth as feed additives. To avoid using labor-intensive instrumental methods to detect
residues of TC in food and food products, a simple and convenient indirect heterologous competitive
enzyme-linked immunosorbent assay (ELISA) method for TC was developed using polyclonal antibody
prepared in this study. Three new immunogens, TC-o-tolidine-bovine serum albumin (BSA), TC-
4-aminobenzoic acid-cationized BSA (cBSA), and TC-1,1′-carbonyldiimidazole-cBSA, were synthe-
sized in this research to develop anti-TC antibodies. All antibodies raised in rabbits and coating
antigens synthesized were screened and characterized using homologous and heterologous ELISA
formats to select the best combination. An optimized ELISA gave an IC50 value of 3.92 µg/mL toward
TC in PBS buffer. The specificity of the assay was studied by measuring cross-reactivity of the antibody
with the structurally closely related compounds of chlortetracycline (112%) and oxytetracycline (<2%).
The recovery rates from the TC-fortified raw milk samples were in the range of 74-116%, while the
intra- and interassay coefficients of variation were <14.5 and <25.0, respectively.
KEYWORDS: Tetracycline; ELISA; drug residue; antibody; immunoassay
INTRODUCTION
Tetracyclines (TCs; Figure 1) are a group of broad-spectrum
antibiotics used for medical purposes as well as animal
husbandry (1). The TC antibiotics have a broad range of activity
against a variety of both Gram-positive and Gram-negative
bacteria. Furthermore, they are both easy to administer, effective
through oral dosing via water and feed, and inexpensive (1-
3). For these reasons, the TC antibiotics are extremely popular
as veterinary antibiotics in animal husbandry either for the
prevention and treatment of diseases or for feed additives to
promote growth. The TC antibiotics are licensed for use in a
variety of food-producing animals including cattle, pig, poultry,
and fish (4, 5). In 1998, the European Federation of Animal
Health carried out a survey on the veterinary use of antibiotics
in the 15 member States of the European Union (EU) and
Switzerland. It was found that TC antibiotics accounted for 65%
of all antibiotics and antibacterials consumed for therapeutic
and preventive use (4).
The use of TC antibiotics in animal husbandry, however, has
the potential to result in the presence of residues in tissues and
the increased emergence of resistant strains of pathogenic
bacteria that could have potential health risks to humans (5, 6).
Nowadays, antibiotic resistance has become a global threat
because existing antibiotics are becoming increasingly ineffec-
tive in combating microbial infections in humans (6). To ensure
Figure 1. Chemical structures of TC and related compounds evaluated
in this study.
the safety of food for consumers, more and more countries have
set MRLs (maximum residue levels) and withdrawal periods
* To whom correspondence should be addressed. Tel: +86(531)-
88361198. Fax: +86(531)88361198. E-mail: xirimo2000@yahoo.com.
10.1021/jf062627s CCC: $37.00
© 2007 American Chemical Society
Published on Web 12/16/2006

212 J. Agric. Food Chem., Vol. 55, No. 2, 2007
Zhang et al.
for TC antibiotics. In the United States, the MRLs of TC are 2
ppm in muscle, 6 ppm in liver, and 12 ppm in kidney (7).
In China, a withdrawal period of at least 4 days is required
before livestock can be slaughtered for food purposes (no. 278,
2003.5.22). To detect residues of TCs, suitable analytical
techniques have to be established. The microbiological assays
are usually used for the measurement of TCs in food because
they are easy to perform and inexpensive. However, these
methods are complicated, time-consuming, and lack specificity
(8). Instrumental analyses such as liquid chromatography-mass
spectrometry (LC-MS) (4, 6) and high-performance liquid
chromatography (HPLC) (9, 10) are the most widely used
methods to detect residues of TC in food and food products.
These methods are sensitive and highly specific but require
expensive equipment, large volume of solvents, derivatizing
treatment, and time-consuming sample cleanup process. There-
fore, they are not suitable to be used as routing screens and
field detection for TCs.
The enzyme-linked immunosorbent assay (ELISA) technique
has long been known as a rapid, sensitive, specific, and cost-
effective analytical method and has been used for diagnostic
and residue detection purposes for many years. The first step
to develop a high quality ELISA test kit to detect residues of
TC is to prepare a high-qualified immunogen of TC. TC is an
unstable substance under acidic and basic conditions and is
sensitive to humidity and heat. Besides, its complex chemical
structure also makes it difficult to synthesize qualified immu-
nogens.
There are no reports concerning the synthesis of TC immu-
nogen and preparation of anti-TC antibody so far in literature,
although commercially the ELISA kits to detect residues of TC
are available from Biopharm (Darmstadt, Germany). This paper
disclosed the detailed information related to synthesis of TC
immuogens and characterization of corresponding antibodies.
The ELISA procedure based on the antibody prepared was
developed and used to detect TC residues in milk.
Figure 2. Synthetic procedure for TC immunogen of TC-tolidine-BSA
through the tolidine method.
mM Na2HPO4, and 2.7 mM KCl. The wash buffer (PBST) was a PBS
buffer containing 0.05% Tween 20. Sodium borate buffer (pH 8.5)
consisted of 150 mM NaCl and 500 mM sodium borate. As a coating
buffer, 0.05 M carbonate buffer (15 mM Na2CO3 and 35 mM NaHCO3,
pH 9.6) was used. The blocking buffer was PBS + 1% OVA + 0.05%
(v/v) Tween 20. The substrate buffer was 0.1 M sodium acetate/citrate
buffer (pH 5.0). To prepare the substrate solution, 10 mg of OPD was
dissolved in 25 mL of sodium citrate buffer and this solution plus 5
µL of H2O2 [30% (w/w)]. The stopping solution was 2 N HCl.
Preparation of Cationized BSA (cBSA) and Cationized OVA
(cOVA). In this procedure, carboxylic acid groups of the carrier proteins
of BSA and OVA were converted into primary amine groups with an
excess of EDA. A solution of 1 g of BSA (15 µmol) and 56 mg of
EDC (300 µmol) in 20 mL of PBS (0.01 M, pH 7.4) was added slowly
into a solution of 18 mg of EDA (300 µmol) in 20 mL of PBS (0.01
M, pH 7.4) under stirring. The mixture solution was incubated
continuously for 2 h at room temperature and then dialyzed [molecular
weight cutoff (mwco), 12000-14000 Da] under stirring against PBS
(0.01 M, pH 7.4) to remove free EDA. Cationized BSA and OVA were
defined as cBSA and cOVA, respectively. The solution was lyophilized,
and the white solid (cBSA) obtained was stored at -20 °C before it
was used in the next reaction. The cOVA was prepared in a similar
method.
Preparation of Immunogens and Coating Antigens. Tolidine
Method (TC-Tolidine-BSA and TC-Tolidine-OVA). The immunogen TC-
tolidine-BSA and the coating antigen TC-tolidine-OVA were prepared
by the homobifunctional method as described in the literature (11). In
this procedure (Figure 2), 25 mg of tolidine (118 µmol) was dissolved
in 4.5 mL of 0.2 N HCl followed by the dropwise addition of 17.5 mg
of sodium nitrite (254 µmol) in 0.5 mL of distilled water in the dark
at 4 °C with constant stirring. The reaction mixture was allowed to
stand for 30 min. Subsequently, 1.0 mL of orange bis-diazotized tolidine
solution was slowly added to 2 mL of 0.5 M sodium borate solution
(pH 8.5, containing 0.15 M NaCl) containing 20 mg of BSA (0.294
µmol) and 10 mg of TC (22.5 µmol). The color of the mixture solution
immediately changed into purple. The reaction was incubated for 2 h
at 4 °C in the dark.
MATERIALS AND METHODS
Chemicals and Materials. TC, o-tolidine (tolidine), bovine serum
albumin (BSA), ovalbumin (OVA), 1-ethyl-3-(dimethylaminopropyl)-
carbodiimide hydrochloride (EDC), 4-aminobenzoic acid (ABA), 1,1′-
carbonyldiimidazole (CDI), and Freund’s complete and incomplete
adjuvants (cFA and iFA) were purchased from Sigma-Aldrich (St.
Louis, MO). N-Hydroxysuccinimide (NHS) was provided by Cxbio
Biotechnology Ltd. (Shanghai, China).
3,3′,5,5′-Tetramethylbenzidine (TMB) was purchased from Amresco
(Solon, OH). Goat anti-rabbit IgG-horseradish peroxidase conjugate
was provided by the Military Medical Institute (Beijing, China).
o-Phenylenediamine (OPD) was purchased from Xinjingke Biotech-
nology (Beijing, China). Acetone was HPLC grade obtained from
Tianjin Kermel Chemical Reagents Development Center (Tianjin,
China). Ethylenediamine dihydrochloride (EDA), hydrogen peroxide
(30%), and other reagents used were chemical grade from Guangmang
Chemical Co. (Jinan, China).
Instrumentation and Supplies. ELISA was performed in polysty-
rene 96 well microtiter plates (Bio Basic Inc.) and spectrophotometri-
cally read with an automatic microplate reader KHB ST-360 from
Shanghai Zhihua Medical Instrument Ltd. UV data were collected on
a U-4100 spectrophotometer from Hitachi Co. Centrifugation was
carried out with a refrigerated centrifuge (Biofuge Stratos, Heraeus).
Protein dialyses were performed using dialysis tubes from Aibo
Economic & Trade Co., Ltd. (Jinan, China). Rotary evaporation was
carried out with the rotary evaporator (RE-5203A) from Shanghai
Zhenjie Laboratory Instrument Co., Ltd.
The purple reaction mixture was dialyzed (mwco, 12000-14000 Da)
under stirring against PBS (0.01 M, pH 7.4) for 3 days with frequent
changes of the PBS solution to remove the uncoupled free hapten. The
precipitate was removed by centrifugation at 3000g, and the supernatant
was lyophilized to obtain a purple conjugate of TC-tolidine-BSA, which
was stored at -20 °C for future use. A conjugate of TC-tolidine-OVA
was prepared in a similar method. An UV absorbance spectrum was
employed to determine whether the linking had been a success (Figure
3).
Buffers. For the preparation of all buffers and reagents for the
immunoassays, ultrapure deionized water was used. Phosphate-buffered
saline (PBS, pH 7.4) consisted of 138 mM NaCl, 1.5 mM KH2PO4, 7
ABA Method (TC-ABA-cBSA and TC-ABA-cOVA). In this procedure
(Figure 4), 20 mg of ABA (145 µmol) was dissolved in 2.2 mL of 0.2

Detection of Residues of Tetracycline in Milk
J. Agric. Food Chem., Vol. 55, No. 2, 2007 213
Figure 5. UV absorbance: (a) ABA, (b) TC, (c) cBSA, and (d) TC-ABA-
Figure 3. UV absorbance: (a) tolidine, (b) TC, (c) BSA, and (d) TC-
cBSA.
tolidine-BSA.
Figure 6. Synthetic procedure for TC immunogen of TC-CDI-cBSA through
the CDI method.
ABA-cOVA conjugate was prepared in a similar method. An UV
absorbance spectrum was employed to determine the coupling result
(Figure 5).
CDI Method (TC-CDI-cBSA and TC-CDI-cOVA). In this procedure
(Figure 6), 40 mg of TC (90 µmol) was dissolved in 8 mL of dry
acetone. Subsequently, 29 mg of CDI (180 µmol) was quickly added
to the solution. The reaction mixture was allowed to stand at 37 °C for
3 h in the dark under dry nitrogen. Acetone was evaporated using a
rotary evaporator under vacuum, and the residue was redissolved in
15 mL of sodium borate (0.5 M, pH 8.5). To this solution, 200 mg of
cBSA (3 µmol) was added, followed by incubation at room temperature
for 48 h in the dark under dry nitrogen. The reaction mixture was
dialyzed (mwco, 12000-14000 Da) under stirring against PBS (0.01
M, pH 7.4) for 3 days with frequent changes of the PBS solution to
remove the uncoupled free hapten. The solution was lyophilized, and
the yellow TC-CDI-cBSA conjugate obtained was stored at -20 °C
for future use. A TC-CDI-cOVA conjugate was prepared in a similar
method. An UV absorbance method was employed to determine
whether the linking had been a success (Figure 7).
Immunization of Rabbits. Three immunogens of TC-tolidine-BSA,
TC-ABA-cBSA, and TC-CDI-cBSA were used to immunize rabbits
to prepare corresponding anti-TC antisera. For each immunogen, two
male New Zealand white rabbits were subcutaneously immunized at
multiple sites in the back. The initial immunization was subcutaneously
injected with 0.5 mg of conjugate in 0.5 mL of NaCl (0.9%) and 0.5
mL of Freund’s complete adjuvant. Subsequent booster injections [0.25
mg of conjugate in 0.5 mL of NaCl (0.9%) plus 0.5 mL of Freund’s
incomplete adjuvant] were performed 15 days later and then at 14 day
intervals. One week after each booster, serum titers were tested by
Figure 4. Synthetic procedure for TC immunogen of TC-ABA-cBSA
through the ABA method.
N HCl followed by the dropwise addition of 12 mg of sodium nitrite
(174 µmol) in 0.35 mL of distilled water in the dark at 4 °C with
constant stirring. The reaction mixture was allowed to stand for 1 h.
Subsequently, 1.5 mL of diazotized ABA solution was slowly added
to 10 mL of 0.5 M sodium borate (pH 8.5, containing 0.15 M NaCl)
containing 27 mg of TC (61 µmol). The mixture solution immediately
became purple. The reaction was incubated for 2 h at 4 °C in the dark
followed by addition of H3BO3 to adjust the pH to 7.4.
Subsequently, 136 mg of cBSA (2 µmol), 120 mg of EDC (626
µmol), and 36 mg of NHS (313 µmol) were added to the solution and
the reaction was allowed to stand at room temperature for 3 h with
constant stirring. The reaction mixture was dialyzed (mwco, 12000-
14000 Da) under stirring against PBS (0.01 M, pH 7.4) for 3 days
with frequent changes of the PBS solution to remove the uncoupled
free hapten. The solution was lyophilized, and the purple TC-ABA-
cBSA conjugate obtained was stored at -20 °C for future use. A TC-

214 J. Agric. Food Chem., Vol. 55, No. 2, 2007
Zhang et al.
DeVelopment of Indirect cELISA. The checker board procedure was
used to optimize the coating antigen and the primary antibody
concentrations. Three different coating antigens (TC-tolidine-OVA, TC-
ABA-cOVA, and TC-CDI-cOVA) were used in a competitive ELISA
for assessing the specificity of the serum from each animal to free TC.
To each well of a 96 well plate, 100 µL of 10 µg/mL of selected
coating antigen solution in bicarbonate buffer (0.05 M, pH 9.6) was
added and incubated overnight at 4 °C. The plate was washed with
wash buffer three times and blocked with 250 µL/well of blocking
buffer, followed by incubation for 1 h at room temperature. After the
blocking solution was removed and the plate was washed three times,
100 ng of primary antibody was added to each well followed by the
addition of buffer or competitor in buffer, and the plate was incubated
for 2 h. The plate was washed three times, and goat antirabbit IgG-
HRP (1:1000, 100 µL/well) was added, followed by incubation for 2
h at room temperature. The plate was washed three times, OPD substrate
solution was added (50 µL/well), and the plate was incubated for another
30 min at room temperature. The color development was halted by
adding stopping solution (50 µL/well), and absorbances were measured
at 492 nm. Absorbances were corrected by blank reading. Preimmune
withdrawn serum was used as a negative control. The result was
expressed in percent inhibition as follows: % inhibition ) %B/B0,
where B is the absorbance of the well containing competitor and B0 is
the absorbance of the well without competitor.
Figure 7. UV absorbance: (a) TC, (b) cBSA, and (c) TC-CDI-cBSA.
Table 1. Summary of Titersa of Antisera vs Three Coating Antigens
Specificity Determination. After coating antigens were optimized by
the above indirect competitive ELISA, the TC-ABA-cOVA was selected
as a coating antigen to determine cross-reactivity and establish the
standard calibration curve. Competitive immunoassays were performed
using various compounds structurally related to TC, to determine the
respective IC50 value and cross-reactivity. ELISA plates were coated
with coating antigen at 10 µg/mL (100 µL/well) by incubation overnight
at 4 °C and then washed three times. Plates were blocked (250 µL/
well) for 2 h at room temperature and washed three times again. For
the competition step, 50 µL/well of competitors (TC, chlortetracycline,
and oxytetracycline) were added at a concentration ranging from 0.01
to 200 µg/mL and coincubated with 50 µL/well of primary antibody
for 2 h. The secondary antibody and color development were the same
as described above. After reading the plate, the IC50 value was
determined by using the concentration of inhibitor that leads to a 50%
decrease of the maximum signal; the cross-reactivity (%) was calculated
as: (IC50,TC)/(IC50,compound) × 100 (12).
coating antigens
TC-ABA-cOVA
immunogens used
to raise antibodies
TC-tolidine-OVA
TC-CDI-cOVA
TC-tolidine-BSA
TC-ABA-cBSA
TC-CDI-cBSA
1:16000
1:200
1:1600
1:3200
1:64000
1:51200
1:3200
1:200
1:512000
a The titer of antiserum is defined as the antiserum dilution that gave 2.0 times
absorance of the control serum.
ELISA. The antiserum obtained after each booster was prepared by
allowing the blood to clot overnight at 4 °C, followed by centrifugation
to remove particulate material. Ten days after the last boost, all rabbits
were then exsanguinated by heart puncture under general anesthetic
and euthanized by lethal injection before recovery. The serum was
separated from blood cells by storage of the blood overnight at 4 °C
and centrifugation with 13000 rpm/min for 20 min. The crude serum
obtained was purified by saturated ammonium sulfate (SAS) precipita-
tion method [purified three times using 50, 33, and 33% (v/v) of SAS,
respectively], and sodium azide was added as a preservative at a final
concentration of 0.02% (w/w). The purified serum was then aliquotted
and stored at -70 °C.
Standard Curve Generation. The TC-ABA-cOVA (10 µg/mL) was
used as a coating antigen, and indirect competitive ELISA was
performed as described above. The selected antisera at 1:1000 dilution
were utilized as primary antibody and coincubated with TC. The
standard calibration curve with final TC concentrations of 0.2, 0.5, 1.0,
2.0, 5.0, and 10 µg/mL was run in PBST.
Matrix Effects Determination. The milk sample was defatted by
centrifugation at 4 °C (10000g, 15 min) and was artificially contami-
nated by adding drug standard solution in PBST. Competitive curves
with final TC concentrations of 0.01, 0.1, 0.2, 0.5, 1.0, 2.0, 5.0, 10,
and 20 µg/mL were run in PBST and in various dilutions (1:2, 1:10,
and 1:20) of the defatted milk with PBST in order to determine the
matrix effect of milk. IC50 and B0 values from each diluted curve were
obtained by comparing with IC50 and B0 values generated from the
PBST buffer.
Inter- and Intra-assay Variation Determination. The milk samples
were fortified by TC at final concentrations of 0.2, 0.5, 1.0, and 2.0 µg
/mL in milk diluted at 1:20 with PBST buffer. Interassay variation was
computed from the analysis of four replicates of each dilution carried
out on four different days. Intra-assay variation was measured by
analysis of five replicates of each dilution on a single day. Sample
recoveries were determined from a standard curve.
Antisera Evaluation. Antibody Titer Determination by Indirect
ELISA. The titer of the serum from each animal was determined by
measuring the binding of serial dilutions of the antisera to microplates
with each of the three different coating antigens (TC-tolidine-OVA,
TC-ABA-cOVA, and TC-CDI-cOVA).
The results are summarized in Table 1 and the detailed processes
are described below. Microplates were coated with three different
coating antigens (10 µg/mL, 100 µL/well) in PBS (0.01 M, pH 7.4) by
overnight incubation at 4 °C. Plates were washed with PBST three
times and blocked with 250 µL/well of blocking buffer, followed by
incubation for 2 h at room temperature. Plates were washed for three
times again, the appropriate dilution of the antisera was added, and
the plates were incubated for 2 h at room temperature. After another
washing step, 100 µL/well of a diluted (1:1000) goat anti-rabbit IgG-
HRP conjugate was added, followed by incubation for 1 h at room
temperature. Then, plates were washed three times, and 100 µL/well
of OPD substrate solution was added. The color development was halted
by adding stopping solution (2 N HCl, 100 µL/well) after 15 min of
incubation at room temperature, and absorbance was read at 492 nm.
Absorbances were corrected by blank reading. Preimmune withdrew
serum was used as a negative control, and the antibody titer was defined
as the reciprocal of the dilution that resulted in an absorbance value
that was twice that of the background.
RESULTS AND DISCUSSION
Hapten Conjugation. As a small molecule with a molecular
mass of 444.4, TC is not able to elicit the immune response of
an animal to produce the anti-TC antibody and is, therefore,
nonimmunogenic. To make it immunogenic, it must be conju-
gated to a carrier protein before immunization. Among protein

Detection of Residues of Tetracycline in Milk
J. Agric. Food Chem., Vol. 55, No. 2, 2007 215
carriers, BSA and OVA are two of the most commonly used
ones, and usually, they give satisfying results. BSA was treated
with an excess of ethylenediamine (EDA) as described previ-
ously (11, 13) to convert carboxylic acid groups into primary
amine groups to prepare cBSA. The cBSA prepared has the
advantage over BSA that more primary amino groups become
available on cBSA to couple with functional groups, such as
carboxylic groups, on hapten. Moreover, the use of cationized
carrier proteins can minimize cross-linking and increase their
pI values to generate more immune responses as compared to
their native forms (11).
nitrogen double bonds, working as two bridges in TC-tolidine-
BSA, connect two small conjugation systems together to form
a larger one (Figure 2). Consequently, we find a peak at a long
wavelength of 494 nm in UV spectrum for TC-tolidine-BSA
as the consequence of red shift in longer conjugation system
(Figure 3) as expected. The corresponding coating antigen TC-
tolidine-OVA gives a similar pattern in UV spectrometry.
In the conjugation reaction to synthesize the second TC
conjugate TC-ABA-cBSA, a linking agent ABA was diazotized
and linked with TC to introduce a carboxylic acid to form a
modified TC. The cBSA, then, was connected with the modified
TC to form a conjugate of TC-ABA-cBSA as shown in Figure
4. The carbodiimide of EDC was applied in the synthetic process
to activate the reaction. The UV spectra recorded from 250 to
600 nm for TC, cBSA, ABA, and TC-ABA-cBSA were shown
in Figure 5. An azo bond has been formed between TC and
ABA, which increase the length of the conjugation system in
TC-ABA-cBSA. The red shift in UV absorbance was observed
for the conjugate (417, 523 nm) as expected. The coating antigen
TC-ABA-OVA gives a similar UV pattern.
TC is quite stable in the dry state under normal storage
conditions, but high temperature and humidity both facilitate
TC degradation. TC changes color from light yellow to brown
and dark brown after exposure to the humidity or heat.
Moreover, TC is unstable in aqueous solution, especially in
strong acid and strong alkaline solution. In the acid solution,
because of the relative instability of hydroxyl group in C6 (see
Figure 1 for numbering of TC), strong acids dehydrate the
C-ring through the hydroxyl group in C6 in TC molecule,
forming 5-hydroxyanhydrotetracyclines. Furthermore, the dim-
ethylamino group in C4 can undergo a reversible epimerization
process in vitro and in vivo, producing a set of epimers with
different antibacterial activities (5, 14, 15). In the alkaline
solution, the C6 hydroxyl group easily loses the hydrogen,
forming an oxide. The oxide can attack the C11 carbon, forming
a lactone so that TC loses activity (16). On account of these
reasons, the preparation of TC immunogen with a high quality
was very difficult. To the best of our knowledge, the synthesis
of TC immunogen for the purpose to prepare anti-TC antibody
has not yet been reported in literature.
The third conjugate was synthesized using CDI as a linking
and activation agent. In the alkaline solution, because the
hydroxyl group in C6 is easy to lose the hydrogen to form the
oxide, we believe that the reaction should occur at this hydroxyl
group in the CDI method. The hydroxyl group in C6 as a
nucleophile attacks the carbonyl carbon of CDI, forming an
active imidazolyl carbamate. The imidazolyl carbamate is
attacked by the amine-containing cBSA, forming the conjugate
TC-CDI-cBSA. The synthetic pathway of this method is shown
in Figure 6, and the UV spectra of TC, cBSA, and TC-CDI-
cBSA, recorded from 250 to 500 nm, are shown in Figure 7. It
can be seen that the cBSA has a peak at 277 nm, and TC has
two peaks at 274 and 362 nm. The presence of carbonyl group
between TC and cBSA made the density of the electron cloud
of the conjugated system of TC fall, so that the requisite energy
of electron transition increased.
In this research, three conjugates of TC with carrier proteins
were prepared. The first conjugate (TC-tolidine-BSA) was
synthesized using tolidine, a homobifunctional cross-linking
agent, as a bridge to link TC and carrier protein BSA by a simple
one-step conjugation reaction. The two amino groups in tolidine
were diazotized to create the requisite bis-diazonium derivative
and then coupled simultaneously the para position of the
aromatic hydroxyl group of TC with the ortho position of
tyrosine residue in BSA, forming azo derivative (Figure 2).
The diazo reaction proceeded by electrophilic attack of the
diazonium group toward the electron-rich points on the target
molecules (11). It is well-known that phenolic compounds are
modified at positions ortho and para to the aromatic hydroxyl
group. However, as far as electron effects and steric effects are
concerned, the diazo reaction usually proceeds at the para
position of the aromatic hydroxyl group. When the para position
of the aromatic hydroxyl group has a substituent, the reaction
conducts at the ortho position (17). Both the para and the ortho
positions of the aromatic hydroxyl group of TC have no
substituent; thus, the diazo reaction is more likely to proceed
at the para position of TC according to the reason mentioned
above. For tyrosine side chains of BSA, only the ortho
modification is possible (Figure 2). To obtain evidence of
successful conjugation, UV absorbances recorded from 250 to
600 nm were measured for BSA, TC, tolidine, and TC-tolidine-
BSA conjugate as shown in Figure 3. BSA has an absorbance
peak at 277 nm coming from an aromatic group in the molecule.
Tolidine has one peak at 281 nm, a red shift as compared with
BSA’s 277 nm, because it has a longer conjugate system
consisting of two benzene rings. In the conjugate of TC-tolidine-
BSA, the conjugation system has been expended significantly
because tolidine, consisting of two benzene rings itself, links
two aromatic systems together to form the conjugate. The two
Thus, the blue shift occurred in UV spectrometry as expected.
A peak of TC at 362 nm had shifted to 320 nm of TC-CDI-
cBSA in UV spectrometry. The coating antigen TC-CDI-cOVA
gives a similar UV pattern.
Titers of the Antisera. The final antisera obtained from each
rabbit were purified with SAS and used for antibody charac-
terization. The titer of antibody from each rabbit was estimated
by the indirect ELISA using either homologous or heterologous
assays. Titers of the individual animals injected with the same
immunogen showed little differences. Most antisera showed
fairly high titers to their homologous antigens as compared to
heterologous antigens. The selected representative examples are
listed in Table 1. The results show that the antisera of number
6 rabbit from immunogen of TC-CDI-cBSA when using TC-
ABA-cOVA as coating antigen give the best combinational
results in terms of both affinity and specificity. We are going
to use this combination throughout the remainder of this report.
Table 1 suggests that all of the antisera obtained, more or less,
can distinguish the TC skeleton from hapten-protein conjugates.
Indirect Competitive ELISA: Recognition of Free TC.
The inhibition in an indirect cELISA is affected by the strength
of interaction between the antisera and the coating antigen. It
is well-known that heterologous assays often help improve the
immunoassay sensitivity (18-21) and overcome unwanted
cross-reactivity associated with the strong affinity of the
antibodies to the spacer arm that results in no or poor inhibition
by the analyte (22, 23). Heterology modifies equilibrium
conditions between coating antigens and analyte so that stronger

216 J. Agric. Food Chem., Vol. 55, No. 2, 2007
Zhang et al.
Figure 8. Representative inhibition curve of anti-TC antibody using TC
as the competitor and TC-ABA-cOVA as the coating antigen in PBS buffer
solution. Each point represents the average of five replicates.
Figure 9. ELISA curve for polyclonal anti-TC antibody using TC (
chlortetracycline ( ), and oxytetracycline ( ) as competitors in PBS buffer
solution. Each point represents the average of five replicates.
9),
1
b
structure (28). Moreover, because the hydroxyl group in C6 of
TC was used to conjugate protein, this coupling method
effectively prevents the formation of lactone and transformation
of TC structure. However, the antisera raised against TC-
tolidine-BSA and TC-ABA-cBSA have very low inhibition by
TC in both homologous assays and heterologous assays. The
low chemical stability of TC and the properties (such as length
and binging ability) of the spacer should be responsible for the
low quality of the antibodies. It is likely that the antibodies
produced can recognize only partial structure of TC of the
coating antigen and the cavity size of the antibody binding site
is not big enough to hold TC. The free TC molecule is, thus,
unable to displace the coating antigen from the antibodies (29).
Specificity. The specificity of the antibody was evaluated
by measuring inhibition curves using three structurally related
compounds (TC, chlortetracycline, and oxytetracycline) as
competitors and TC-ABA-cOVA as a coating antigen (Figure
9). The cross-reactivity studies were carried out by an indirect
competitive ELISA by adding various free competitors at
different concentrations (ranging from 0.01 to 200 µg/mL) to
compete with coating antigen to bind with the antibody to
estimate their respective IC50 value and then comparing this
value with that of TC. The IC50 value and cross-reactivity for
each compound were given in Table 2. These results demon-
strated that chlortetracycline showed high cross-reactivity
(112%) toward the antibody, whereas oxytetracycline showed
a low cross-reactivity (<2%). A chlorine atom at the position
of C7 is the only structural difference between TC and
chlortetracycline (Figure 1). This extra chlorine atom of
chlortetracycline does not make any significant change for the
affinity of chlortetracycline toward the anti-TC antibody,
indicating that the benzene ring in TC system is not an important
structural factor to determine the affinity. In the case of
oxytetracycline, however, an extra hydroxyl group linked at
position of C5 in the molecule makes it demonstrate very low
cross-reactivity toward anti-TC antibody, indicating that this
hydroxyl group, as a key structural factor, inhibits binding to
the TC antibody to result in low affinity. By combining these
results, we can conclude that a structural moiety consisting of
three nonaromatic rings in TC system is the main antigenic
determinant for antibodies developed in this research.
Table 2. IC50 and Percentage of Cross-Reactivity of TC Analogues
a
compound
TC
chlortetracycline
oxytetracycline
IC50
cross-reactivityb (%)
3.92
3.49
<200
100
112
<2
a IC50 values are in
µ
g/mL. b Cross-reactivity was determined by comparing
the concentration of analyte required to produce a B/Bo
expressed as a percentage relative to the figure for TC.
) 50%. Results were
recognition toward the analyte can be achieved (24, 25). Suitable
heterology can be carried out at different levels by hapten (use
of partial structure or change of key determinants), site, linker,
and spacer modification (composition, length, and conjugation
chemistry) (18).
In view of the above-mentioned reasons, in this study, in order
to set up a sensitive ELISA, all possible combinations between
coating antigens and antibodies were screened via the indirect
cELISA by serial concentrations (ranging from 0.01 to 100 µg/
mL) of TC dissolved in PBST, using the homologous or
heterologous assay. It was found that there was very low
inhibition by TC in homologous assays when an immunogen
and a coating antigen were prepared using a same linker.
In the heterologous assays, the combinations of TC-ABA-
cOVA or TC-tolidine-OVA as coating antigens with the
antibodies raised against immungen of TC-CDI-cBSA yielded
the best results in inhibitions, particularly the combination
between the TC-ABA-cOVA and the antibodies from rabbit
number 6. Figure 8 shows a TC inhibition curve obtained by
the indirect competitive ELISA with number 6 rabbit antibody.
The inhibition curve was obtained by five replicates. The limit
of detection (LOD, also called the least detectable dose)
estimated as the concentration of TC giving a 10% inhibition
(I10 value) of the maximum absorbance, and the working range
for ELISA, calculated as the concentration of TC providing a
20-80% inhibition (I20-I80 values) of the maximum signal,
were 0.01 and 0.1-100 µg/mL, respectively (26, 27). The value
of IC50 (Table 2) was 3.92 µg/mL. The result confirmed that
the heterology assay can really improve the sensitivity of TC
immunoassays. At the same time, the high sensitivity of the
antibody may also be due to the relatively short spacer arm
(five atoms) between TC and cBSA. Such a spacer arm can
help expose the determinant groups on the immunogen, prevent-
ing the determinant groups being masked by the protein tertiary
Matrix Effects and Inter- and Intra-assay Variation
Determination. It has been known that various substances
existing in complex matrixes can affect antigen-antibody
interaction in immunoassays. To reduce matrix effects, two

Detection of Residues of Tetracycline in Milk
J. Agric. Food Chem., Vol. 55, No. 2, 2007 217
ABBREVIATIONS USED
MRL, maximum residue level; BSA, bovine serum albumin;
OVA, ovalbumin; cBSA, cationized BSA; cOVA, cationized
OVA; EDC, 1-ethyl-3-(3-dimethyl aminopropyl)carbodiimide;
NHS, N-hydroxysuccinimide; IC50, concentration at 50% inhibi-
tion; Da, unit of molecular mass Dalton; PBS, phosphate-
buffered saline; PBST, phosphate-buffered saline Tween 20;
OPD, o-phenylendiamine; DMF, N,N-dimethyl formamide;
SAS, saturated ammonium sulfate; cFA, complete Freund’s
adjuvant; iFA, incomplete Freund’s adjuvant; ELISA, enzyme-
linked immunosorbent assay; CDI, 1,1′-carbonyldiimidazole;
TC, tetracycline; tolidine, o-tolidine; ABA, 4-aminobenzoic acid.
LITERATURE CITED
Figure 10. Standard calibration curve. Each point represents the average
of five replicates.
(1) Martindale. The Extra Pharmacopeia, 33rd ed.; Sweetman, S.
C., Ed.; Pharmaceutical Press: London, United Kingdom, 2002.
(2) Chopra, I.; Hawkey, P. W.; Hinton, M. Tetracyclines, molecu-
laran and clinical aspects. J. Antimicrob. Chemother. 1992, 29,
245-277.
(3) Kelly, M.; Tarbin, J. A.; Ashwin, H.; Sharman, M. Verification
of compliance with organic meat production standards by
detection of permitted and nonpermitted uses of veterinary
medicines (tetracycline antibiotics). J. Agric. Food Chem. 2006,
54, 1523-1529.
(4) Bogialli, S.; Curini, R.; Corcia, A. D.; Lagana`, A.; Rizzuti, G.
A rapid confirmatory method for analyzing tetracycline antibiot-
ics in bovine, swine, and poultry muscle tissues: matrix solid-
phase dispersion with heated water as extractant followed by
liquid chromatography-tandem mass spectrometry. J. Agric. Food
Chem. 2006, 54, 1561-1570.
Table 3. Inter- and Intra-assay Variation of Raw Milk Spiked with TC
interassaya
intra-assayb
level
(ppm)
measured
(ppm)
recovery
(%)
CV
(%)
measured
(ppm)
recovery
(%)
CV
(%)
n
n
0.2
0.5
1.0
2.0
4
4
4
4
0.16
±
±
±
±
0.04
80
92
100
116
25.0
17.4
15.0
17.7
5
5
5
5
0.17
±
±
±
±
0.02
85
74
85
11.8
10.8
14.1
14.5
0.46
1.00
2.32
0.08
0.15
0.41
0.37
0.85
2.13
0.04
0.12
0.31
107
a Interassay variation was determined by four replicates on four different days.
b Intra-assay variation was determined by five replicates on a single day.
common methods could be used. The first one is to clean up
the sample, which is time-consuming and laborious and may
influence assay reproduction and recovery. The second method
is dilution of the extract, which is simple and efficient and is
used in this research (30). As the dilution of raw milk increased
from 1:2 to 1:20, the absorbance gradually increased to approach
the PBST buffer values. The average B0 (antibody binding with
no competitor present) from five replicates for milk dilutions
at 1:2, 1:10, and 1:20 had absorbencies of 0.869, 0.942, and
1.007, respectively, as compared to 0.997 for antibody in PBST.
The IC50 values with TC as the competitor were 10.33, 5.50,
and 4.06 µg/mL, respectively, as compared with 3.81 µg/mL
for PBST. It can be seen that the matrix effect actually can be
ignored at a dilution of 1:20, although the difference in IC50
values between 1:20 dilution and PBST buffer shows a minor
difference indicating a certain level of matrix effect. Thus, a
1:20 dilution was used to perform the inter- and intra-assay
variation tests. The milk sample recovery rates were determined
from a standard curve (Figure 10) with TC concentrations of
0.2, 0.5, 1.0, 2.0, 5.0, and 10 µg/mL in PBST assay buffer. The
coefficient of variation for intra-assay was below 15% for intra-
assay and below 25% for interassay (Table 3). Recovery rates
were within 20% of theoretical values, indicating acceptable
accuracy.
(5) Aga, D. S.; Goldfish, R.; Kulshrestha, P. Application of ELISA
in determining the fate of tetracyclines in land-applied livestock
wastes. Analyst 2003, 128, 658-662.
(6) Aga, D. S.; O’Connor, S.; Ensley, S.; Payero, J. O.; Snow, D.;
Tarkalson, D. Determination of the persistence of tetracycline
antibiotics and their degradates in manure-amended soil using
enzyme-linked immunosorbent assay and liquid chromatography-
mass spectrometry. J. Agric. Food Chem. 2005, 53, 7165-7171.
(7) Moats, W. A. Determination of tetracycline antibiotics in beef
and pork tissues using ion-paired liquid chromatography. J.
Agric. Food Chem. 2000, 48, 2244-2248.
(8) Kurtittu, J.; Lo¨nnberg, S.; Virta, M.; Karp, M. A group-specific
microbiological test for the detection of tetracycline residues in
raw milk. J. Agric. Food Chem. 2000, 48, 3372-3377.
(9) Pena, A.; Pelantova, N.; Lino, C. M.; Silveira, M. I. N.; Solich,
P. Validation of an analytical methodology for determination of
oxytetracycline and tetracycline residues in honey by HPLC with
fluorescence detection. J. Agric. Food Chem. 2005, 53, 3784-
3788.
(10) Wan, G. H.; Cui, H.; Zheng, H. S.; Zhou, J.; Liu, L. J.; Yu, X.
F. Determination of tetracyclines residues in honey using high-
performance liquid chromatography with potassium permanga-
nate-sodium sulfite-â-cyclodextrin chemiluminescence detection.
J. Chromatogr. B 2005, 824, 57-64.
(11) Hermanson, G. T. Bioconjugate Techniques; Academic Press:
San Diego, 1996.
(12) Coillie, E. V.; Block, J. D.; Reybroeck, W. Development of an
indirect competitive ELISA for flumequine residues in raw milk
using chicken egg yolk antibodies. J. Agric. Food Chem. 2004,
52, 4975-4978.
In summary, the three new conjugates of TC with carrier
proteins have been prepared and were used to immunize rabbits
to prepare corresponding anti-TC antibodies. An indirect het-
erologous competitive ELISA has been developed based on an
anti-TC antibody prepared from an immunogen of TC-CDI-
BSA with TC-ABA-cOVA as the coating antigen. The ap-
plicability of the ELISA developed has been used to detect
spiked milk samples, and the results have shown that the assay
developed could be used as a screening method to detect TC
residues in milk.
(13) Holtzapple, C. K.; Buckley, S. A.; Stanker, L. H. Production
and characterization of monoclonal antibodies against sarafloxa-
cin and cross-reactivity studies of related fluoroquinolones. J.
Agric. Food Chem. 1997, 45, 1984-1990.
(14) Pena, A.; Palilis, L. P.; Lino, C. M.; Silveira, M. I.; Calokerinos,
A. C. Determination of tetracycline and its major degradation
products by chemiluminescence. Anal. Chim. Acta 2000, 405,
51-56.

218 J. Agric. Food Chem., Vol. 55, No. 2, 2007
Zhang et al.
(15) Wu, Y.; Fassihi, R. Stability of metronidazole, tetracycline HCl
and famotidine alone and in combination. Int. J. Pharm. 2005,
290, 1-13.
(16) Zheng, H. Medical Chemistry; People Health Press: Beijing,
2003; pp 265-267.
organophosphorus pesticides and effect of heterology in hapten
spacer arm length on immunoassay sensitivity. Anal. Chim. Acta
2003, 475, 85-96.
(25) Szurdoki, F.; Szekacs, A.; Le, H. M.; Hammock, B. D. Synthesis
of haptens and protein conjugates for the development of
immunoassays for the insect growth regulator fenoxycarb. J.
Agric. Food Chem. 2002, 50, 29-40.
(26) Xu, D. M.; Yu, X. Y.; Liu, Y. Q.; Feng, J. T.; Pan, L. G.; Liu,
X. J.; He, J.; Zhang, X. Development of an enzyme-linked
immunosorbent assay for podophyllotoxin. Int. Immunophar-
macol. 2005, 5, 1583-1592.
(27) Watanabe, E.; Eun, H.; Baba, K.; Arao, T.; Endo, S.; Ueji, M.;
Ishii, Y. Synthesis of haptens for development of antibodies to
alkylphenols and evaluation and optimization of a selected
antibody for ELISA development. J. Agric. Food Chem. 2005,
53, 7395-7403.
(28) Szurdoki, H. K. M.; Bekheit, M. P.; Marco, M. H.; Goodrow,
B. D. Hammock. In New Frontiers in Agricultural Immunoas-
says; Kurtz, D. A., Skerritt, J. H., Stanker, L., Eds.; AOAC:
Arlington, VA, 1995; p 39.
(17) Xing, Q. Y.; Xu, R. Basic Organic Chemistry; Advanced
Education Press: Beijing, 2005.
(18) Gueguen, F.; Boisde´, F.; Queffelec, A. L.; Haelters, J. P.;
Thouvenot, D.; Corbel, B.; Nodet, P. Hapten synthesis for the
development of a competitive inhibition enzyme-immunoassay
for thiram. J. Agric. Food Chem. 2000, 48, 4492-4499.
(19) Beier, R. C.; Creemer, L. C.; Ziprin, R. L.; Nisbet, D. J.
Production and Characterization of monoclonal antibodies against
the antibiotic tilmicosin. J. Agric. Food Chem. 2005, 53, 9679-
9688.
(20) Muldoon, M. T.; Holtzapple, C. K.; Deshpande, S. S.; Beier, R.
C.; Stanker, L. H. Development of a monoclonal antibody-based
cELISA for the analysis of sulfadimethoxine 1. Development
and characterization of monoclonal antibodies and molecular
modeling studies of antibody recognition. J. Agric. Food Chem.
2000, 48, 537-544.
(21) Lee, J. K.; Ahn, K. C.; Park, O. S.; Ko, Y. K.; Kim, D. W.
Development of an immunoassay for the residues of the herbicide
bensulfuron-methyl. J. Agric. Food Chem. 2002, 50, 1791-1803.
(22) Beier, R. C.; Ripley, L. H.; Young, C. R.; Kaiser, C. M.
Production, characterization, and cross-reactivity studies of
monoclonal antibodies against the coccidiostat nicarbazin. J.
Agric. Food Chem. 2001, 49, 4542-4552.
(23) Greirson, B. N.; Allen, D. G.; Gare, N. F.; Watson, I. M.
Development and application of an enzyme-linked immunosor-
bent assay for lupin alkaloids. J. Agric. Food Chem. 1991, 39,
2327-2331.
(29) Zheng, J.; Hammock, B. D. Development of polyclonal antibod-
ies for detection of protein modification by 1,2-naphthoquinone.
Chem. Res. Toxicol. 1996, 9, 904-909.
(30) Yang, Z. Y.; Kolosova, A. Y.; Shim, W. B.; Chung, D. H.
Development of monoclonal antibodies against pirimiphos-
methyl and their application to IC-ELISA. J. Agric. Food Chem.
2006, 54, 4551-4556.
Received for review September 14, 2006. Revised manuscript received
November 9, 2006. Accepted November 11, 2006.
(24) Kim, Y. J.; Cho, Y. A.; Lee, H. S.; Lee, Y. T.; Gee, S. J.;
Hammock, B. D. Synthesis of haptens for immunoassay of
JF062627S