ISSN 0003�6838, Applied Biochemistry and Microbiology, 2011, Vol. 47, No. 1, pp. 77–81. © Pleiades Publishing, Inc., 2011.Original Russian Text © M.M. Vdovenko, C.�F. Peng, C.�L. Xu, E.S. Vylegzhanina, A.A. Komarov, I.Yu. Sakharov, 2011, published in Prikladnaya Biokhimiya i Mikrobiologiya, 2011,Vol. 47, No. 1, pp. 84–89.

77

Illegal use of hormonal growth stimulators some�times occurs in highly intensive animal breeding,because this technology violation allows for a signifi�cant increase of production at minimum cost. Prod�ucts thus obtained are extremely dangerous for thehealth of humans, since the hormonal growth stimula�tors are carcinogenic and can disrupt sexual matura�tion and reproduction in humans [1–4]. Nowadaysthe use of hormonal growth stimulators in animalbreeding is prohibited in most countries. According tothe EC directive 96/23 meat and animal feed in thecountries of the European Union are monitored fortraces of prohibited anabolic growth stimulators, suchas stilbenes, steroids, lactones of resorcinic acid, andbeta�adrenomimetics. These substances belong togroup A of the most dangerous medicines. The mainobjective of the control of the preparations of thisgroup is the detection of any concentrations of the ille�gally used preparations.

The presence of hormonal growth stimulators infoods is most often detected by chromatographicmethods [5–7]. Notwithstanding the high sensitivity,reproducibility, and reliability of these methods, theyhave significant drawbacks, such as the necessity to useexpensive equipment operated by highly qualified spe�cialists; serial chromatographic analysis is labor�intensive and take a lot of time. Immunochemicalmethods are devoid of these drawbacks; therefore,much effort is put into the development of immu�nochemical assays for the detection of hormonalgrowth stimulators [8].

The aim of the present study was to develop anindirect competitive enzyme�linked immunosorbentassay for the analysis of hexestrol (HES), one of theanabolic preparations of the stilbene group, and toapply this method for the determination of hexestrol inthe meat of animals.

METHODS

Materials. A horseradish peroxidase preparation(RZ 3.0) from Sigma (United States) was used in thepresent work without further purification.

Hexestrol, 3,3',5,5'�tetramethylbenzidine (TMB)and caseine sodium salt were from Sigma (UnitedStates), 30% Н2О2 was from ChemMed (Russia), andTween 20 was from Ferak (Germany). All the salts usedwere of “analytical” or “chemically pure” grade. Theconcentration of hydrogen peroxide was assessed spec�trophotometrically by measuring the optical density ofthe solutions at 240 nm (ε240 = 43.6 М–1 cm–1) [9].

Immunoreagents. A conjugate of hexestrol andovalbumin (HES�OVA) was obtained as described pre�viously [10]. Briefly, 5 ml trimethylamine and 20 mlisobutylchloroformiate were added to 10 mg of 3�car�boxypropylhexestryl ester dissolved in 0.4 ml of 1 : 1(vol : vol) mixture of dioxane and dimethylsulfoxideand incubated for 1 h at 48°С (solution A). Simulta�neously, 20 mg of ovalbumin were dissolved in 0.8 mldistilled water and the solution was mixed with 0.8 mldimethylsulfoxide. This solution (solution B) was alsoincubated for 1 h at 48°С. The solution A was slowlydropwise added to the solution B and the mixture was

Enzyme Immunoassay for the Determination of Hexestrol in MeatM. M. Vdovenkoa, C.�F. Pengb, C.�L. Xub, E. S. Vylegzhaninac,

A. A. Komarovc, and I. Yu. Sakharova

a Faculty of Chemistry, Moscow State University, Moscow, 119991 Russia b School of Food Chemistry and Technology, South Yangtze University, 214122 China

c All�Russia State Center for Quality Control and Standardization of Veterinary Therapeutics and Animal Feed, Moscow, 123022 Russia

e�mail: sakharovivan@gmail.comReceived December 4, 2009

Abstract—An indirect competitive enzyme�linked immunosorbent assay (ELISA) of hexestrol (HES), ananabolic hormone forbidden for use in livestock farming, has been developed. Conditions of ELISA havebeen optimized by varying the concentrations of the coating conjugate (HES�ovalbumin), anti�HES antise�rum, casein, and Tween 20. In the absence of Tween 20 in the reaction mixture, the detection limit (IC10)equaled 0.01 ng/ml, IC50 equaled 0.17 ng/ml, and the working range (IC20–IC80) equaled 0.03–0.86 ng/ml,while, in the presence of 0.05% Tween 20, these values equaled 0.05, 2.9, and 0.26–32.0 ng/ml, respectively.Standard deviation of the analysis results did not exceed 5.4%. If ELISA was performed in the absence ofdetergents, the recovery value upon HES determination in spiked beef samples ranged from 74 to 147%.

DOI: 10.1134/S0003683811010182

78

APPLIED BIOCHEMISTRY AND MICROBIOLOGY Vol. 47 No. 1 2011

VDOVENKO et al.

incubated for 4 h at 48°С. The HES�OVA conjugateobtained was dialyzed against 10 mM K�phosphatebuffer, pH 7.4, containing 0.15 M NaCl (PBS) andstored at –20°С. The conjugate of HES and bovineserum albumin (BSA) was synthesized according to asimilar procedure.

Polyclonal antiserum to HES was obtained by sub�cutaneous immunization of rabbits (body mass 1.5–2.5 kg) with the HES�BSA conjugate during 6 months,as described previously [11]. The antibody titers weredetermined by ELISA. The specific antisera werestored in 50% (vol : vol) glycerol solutions at –20°С.

The conjugate of polyclonal sheep antirabbit IgGantibodies and horseradish peroxidase was synthesizedas described [12]: briefly, 5 mg of horseradish peroxi�dase (HRP) were dissolved in 0.5 ml distilled water andsodium periodate was added to the solution to the finalconcentration of 17 mM. HRP oxidation was per�formed for 20 min at room temperature in the dark,and after this, the reaction mixture was dialyzedagainst 1 mM sodium acetate buffer, pH 4.9, overnightat 4°С for the elimination of excess oxidizer and theproducts of its decomposition. The peroxidase solu�tion obtained was mixed with a solution of 0.5 mg ofpurified polyclonal sheep�antirabbit antibodies in75 μl of 50 mM sodium carbonate buffer, pH 9.6, andincubated at room temperature for 2 h with constantstirring. The Schiff bases formed were reduced by add�ing 10 μl of a freshly prepared 10% solution of NaBH4.The reaction was performed for 30 min at room tem�perature. The conjugate was dialyzed against PBS andstored in 50% (vol : vol) glycerol solution at –20°С.

Immunoenzymatic determination of hexestrol. TheHES�OVA conjugate was preadsorbed in the wells of96�well microplates with a high absorbing capacity(Costar, United States) by incubating 100 μl of theconjugate solution (0.25–2.00 mg/ml) in 50 mMbicarbonate buffer, pH 9.6, in the wells at 4°С over�night. After the removal of the unadsorbed conjugate100 μl of 0.1% casein solution in 50 mM phosphatebuffer, pH 7.4, containing 0.15 M NaCl (PBS) wereadded to each well and incubated for 1 h at 37°С.Excess casein was removed by four washes with PBS

containing 0.05% Tween 20 (PBST). After this, 50 μlof HES solution (0.003–600 ng/ml) and antiHESserum were added to the wells. The antiserum wasdiluted with PBS or PBST, and HES solutions werediluted with a 0.2% casein solution in 20% methanolin PBS or PBST. The competitive immunologicalreaction was performed for 1 h at 37°С. The unboundantibodies were washed away and 100 μl of the solutionof the HRP�secondary antibody conjugate (diluted 1 : 6000)in PBST were added to the wells and incubated for 1 hat 37°С. The working dilution of the conjugate usedwas determined in a separate experiment. After washingthe microplate, 100 μl of a freshly prepared substratesolution containing 0.42 mM TMB and 1.2 mM H2O2 in100 mM acetate buffer, pH 4.3, were added to eachwell. The enzymatic reaction was stopped with 50 μl of2 M H2SO4, and the optical density of the solutions inthe wells was measured at 450 nm.

Beef sample preparation. Five grams of beef notcontaining fat or connective tissue were ground, mixedwith 5 ml of 0.067 M phosphate buffer, pH 7.2, andhomogenized for 2 min in a knife homogenizer. Onegram of the homogenate obtained was mixed with 3 mlof methanol and sonicated (Elmasonic S, Germany)at the power of 80 W for 20 min, and then centrifugedfor 10 min at 4000 g. The pellet was extracted withmethanol again. The pellet obtained after the secondextraction was mixed with 3 ml of 1 : 3 water–metha�nol mixture and sonicated for 20 min. The homoge�nate was centrifuged for 10 min at 4000 g. The extrac�tion with the water–methanol mixture was performedtwice. After four rounds of extraction, the superna�tants from one meat sample were combined, mixedwith 2 ml hexane, and shaken intensively. The hexanelayer was removed and the hexane extraction wasrepeated. The methanol from the bottom (water–methanol) layer was removed by evaporation at 50°C.The aqueous solution was mixed with 5 ml of tert�butyl methyl ether and the extraction was performedfor 20 min with manual shaking once in 5 min. Thebottom layer was extracted with tert�butyl methylether again. The organic fractions were pooled andevaporated in a rotary evaporator (45°C). Before theanalysis, the dried sample was dissolved in 0.5 ml ofthe working buffer (PBS or PBST) and centrifuged for10 min at 7000 g. At the last stage of sample prepara�tion, the supernatant was diluted tenfold with theworking buffer. HES solutions with the concentrationscorresponding to IC25, IC50, and IC75 in the meatextract sample were prepared by mixing an ethanolsolution of HES (1 mg/ml) and beef extract.

HO OH

H5C2 H

H C2H5

+

ovalbumin

antigen

antibody

secondaryantibody

peroxidase

+

Fig. 1. The scheme of the indirect competitive enzyme�linked immunosorbent assay for hexestrol.

Fig. 2. The chemical structure of hexestrol.

APPLIED BIOCHEMISTRY AND MICROBIOLOGY Vol. 47 No. 1 2011

ENZYME IMMUNOASSAY FOR THE DETERMINATION OF HEXESTROL IN MEAT 79

RESULTS AND DISCUSSION

Although the major factors determining the qualityof immunoenzymatic assay are the characteristics ofthe immunoreagents (antibodies, coating conjugates,and enzyme conjugates) used, the conditions of theanalysis also have a significant influence on theparameters of this analytical method. Taking this intoaccount, we optimized the concentration of antiHESantibodies in the reaction medium on the first stage ofthe development of the indirect competitive ELISA(Fig. 1) for the detection of HES (the structural for�mula is given in Fig. 2) in the meat of animals.

As shown in Fig. 3, the use of a low dilution of thespecific antiserum (dilution 1 : 800) resulted in a cali�bration curve with a high background value of theoptic density (Fig. 3, curve 1). On the other hand, thevalue of the optic density did not depend on the con�centration of HES added to the reaction mediumwhen a highly diluted antiserum (dilution 1 : 6400) wasused, and, therefore, the concentration of HES couldnot be measured (Fig. 3, curve 4). Calibration curvesof high quality were obtained when the HES antiserumwas diluted 1 : 1600 or 1 : 3200 (Fig. 3, curves 2 and 3).

Detergents are used in conventional ELISA to pre�vent unspecific sorption on polystyrene plates. Inaddition, the sorption of a blocking protein is often

A450

3.0

2.5

2.0

1.5

1.0

0.5

0.01 0.1 1 10 100

1

2

3

[HES], ng/ml

A450

1.2

0.9

00.01 0.1 1 10 100

1

2

[HES], ng/ml

0.6

0.3

A450

2.5

2.0

1.5

1.0

0.5

0.01 0.1 1 10 100

1

2

3

[HES], ng/ml

4

A450

1.0

0.8

0.6

0.4

0.20.01 0.1 1 10 100

1

2

[HES], ng/ml

4

Fig. 3. Selection of the optimal dilution of the antiHESantiserum in the ELISA of hexestrol. The following dilu�tions were tested for the optimization: 1—1 : 800, 2—1 :1600, 3—1 : 3200, and 4—1 : 6400.

Fig. 4. Calibration curves for the immunoenzymatic assayfor hexestrol with 10% methanol in PBST (2) with or (1)without 0.1% casein as the working buffer.

Fig. 5. Selection of the optimal concentration of HES�OVA used as the coating conjugate in the ELISA of hex�estrol. The following concentrations (µg/ml) were used forthe optimization: 1–0.25, 2—0.5, 3—1.0, and 4—2.0.

Fig. 6. Calibration curves for the immunoenzymatic assayfor hexestrol with 10% methanol and 0.1% casein in PBST(2) with or (1) without 0.05% Tween 20 as the workingbuffer.

80

APPLIED BIOCHEMISTRY AND MICROBIOLOGY Vol. 47 No. 1 2011

VDOVENKO et al.

performed at a certain stage of the assay. On the firststage of our investigation, we used both of these proce�dures in the detection of HES, with Tween 20 as thedetergent and casein as the blocking protein. However,the curve 1 in Fig. 4 shows that the optical density cor�responding to the nonspecific reaction was relativelyhigh in this version of the analysis. Therefore, caseinwas used not only as blocking protein but also as one ofthe components of the reaction mixture on the stage ofcompetition. The addition of 0.1% casein into the

reaction mixture allowed for the successful suppres�sion of the nonspecific adsorption of the antibodies inthe microplate wells (Fig. 4, curve 2). It is also neces�sary to note that the introduction of casein into thereaction mixture caused a decrease of the efficiency ofthe specific immunological reaction.

In further experiments, the influence of the con�centration of the HES�OVA conjugate used as thecoating antigen on the analytical characteristics of theELISA was studied. As shown in Fig. 5, curve 1, thecalibration curve can not be constructed if the concen�tration of the coating antigen is low. Increasing theHES�OVA concentration allowed for the constructionof high�quality calibration curves for the determina�tion of HES (Fig. 5, curves 2 and 3). However, a fur�ther increase of HES�OVA concentration (to 2 μg/ml)had negative consequences, namely, a dramatic dete�rioration of the analytical characteristics of ELISA,such as an increase of the detection limit value and adramatic narrowing of the working range of HES con�centrations (Fig. 5, curve 4). The effect detected isobviously due to the formation of inactive HES�OVAaggregates due to the hydrophobic interaction of HESfragments of different molecules of the coating conju�gate. Comparison of such analytical characteristics asIC10, IC50 and the working range (IC20–IC80) for fourcalibration curves constructed using different HES�OVA concentrations showed that the best results wereobtained at a HES�OVA concentration of 0.5 μg/ml(Table 1).

It was found during the development of ELISA forthe detection of azinphosphomethyl in water samples[13] that the omission of Tween 20 from the reactionmixture at the competition stage caused an increase ofthe assay sensitivity. We have tried an analogousapproach in HES determination. As shown in Fig. 6,the omission of the detergent from the reaction mix�ture caused a shift of the calibration curve to the lower

Table 1. The influence of the coating conjugate (HES�OVA)concentration on the analytic parameters of the ELISA forhexestrol

HES�OVA, µg/ml IC10, ng/ml IC50, ng/ml Working

range, ng/ml

0.25 0.4 3.6 0.7–16

0.5 0.1 3.4 0.4–30

1.0 0.7 11.5 1.5–98

2.0 8.0 34.6 16–92

Table 2. The effect of Tween 20 in the reaction medium at thecompetition stage on the analytic parameters of the ELISA forhexestrol

Reaction medium composition

IC10, ng/ml

IC50, ng/ml

Working range, ng/ml

PBS with 0.1% casein, 10% methanol and 0.05% Tween 20

0.05 2.9 0.26–32.0

PBS with 0.1% casein and 10% methanol

0.01 0.17 0.03–0.86

Table 3. Hexestrol determination in beef samples by indirect competitive enzyme�linked immunosorbent assay

Meat sample

ELISA not involving Tween 20 ELISA involving Tween 20

HES concentration

recovery, %

HES concentration

recovery, %added, ng/ml found, ng/ml

(n = 3) added, ng/ml found, ng/ml (n = 3)

1 0.04 0.06 140 0.4 5.0 1163

0.17 0.20 118 2.9 18.5 638

0.55 0.57 104 18.5 100.9 545

2 0.04 0.06 140 0.4 5.0 1163

0.17 0.25 147 2.9 27.0 931

0.55 0.70 127 18.5 177.4 959

3 0.04 0.03 74 0.4 5.0 1163

0.17 0.14 82 2.9 39.3 1355

0.55 0.57 104 18.5 214.2 1158

APPLIED BIOCHEMISTRY AND MICROBIOLOGY Vol. 47 No. 1 2011

ENZYME IMMUNOASSAY FOR THE DETERMINATION OF HEXESTROL IN MEAT 81

range of HES concentrations, i.e., the detection limitof the anabolic hormone studied was decreased. Wecan also note with satisfaction that the omission of thedetergent from the reaction medium did not cause asignificant increase of the background value of theoptical density (Fig. 6). The IC10 and IC50 values forthe reaction mixture not containing Tween 20 equaled0.01 and 0.17 mg/ml, respectively, the working rangeof HES analysis was 0.03–0.86 ng/ml (Table 2). Thestandard deviation of the analysis results was in therange of 0.3–5.4%.

Comparison of the analytical parameters of HESdetermination by detergent�depleted ELISA and liq�uid chromatography with mass�spectrometric detec�tion (the detection limit for this method equaled0.8 ng/ml [14]) showed that the ELISA developed hada lower HES detection limit compared to the chro�matographic method. The comparison of the methoddeveloped to the previously reported immunochemicalmethods of analysis demonstrated that the analyticalparameters of the detergent�depleted ELISA are muchbetter than the parameters of the indirect ELISA (work�ing range 0.1–8.1 ng/ml, IC50 0.671 ng/ml) [15] andcomparable to those of direct ELISA (working range0.01–8.1 ng/ml, IC50 0.23 ng/ml) [11].

The analysis method developed was used for thedetection of HES added to samples of beef extract.HES concentration in the samples prepared equaled0.04, 0.17, and 0.55 ng/ml. If the reaction of specificbinding of the antibodies to HES and HES�OVA wasperformed in PBS, i.e., in the absence of detergents,the recovery values for HES in spiked meat samplesvaried from 74 to 147% (Table 3). It is necessary toemphasize that if Tween 20 was used in HES determi�nation by ELISA, the matrix effect of meat was so highthat the concentration of HES in meat sample (0.43,2.9, and 18.5 ng/ml) could not be measured correctly.The recovery values in this case ranged from 545 to1355% (Table 3).

Thus, we have developed an indirect competitiveenzyme�linked immunosorbent assay for hexestrol.We have demonstrated that the omission of detergentsfrom the reaction on the stage of competition allowsfor a significant decrease of the detection limit of thisanabolic hormone, as well as for the minimization ofthe matrix effect of meat in immunoenzymatic deter�mination of HES.

ACKNOWLEDGMENTS

The financial support for the present study was pro�vided by the Russian Foundation of Basic Research(08�03�92201 GFEN_a and 09�04�01639�a), theNational Natural Science Foundation of China(20675035), and the Prospective Program of Scienceand Technology Development in China(2006BAK02A09 and 2006BAK02A19).

REFERENCES

1. Metzler, M., Food Cosmetics Toxicol., 1981, vol. 19,pp. 611–615.

2. Hart, J.E., Food Chem. Toxicol., 1988, vol. 26, no. 3,pp. 227–232.

3. Byford, J.R., Shawa, L.E., Drewb, M.G.B., Pope, G.S.,Sauer, M.J., and Darbre, P.D., J. Steroid Biochem. Mol.Biol., 2002, vol. 80, no. 1, pp. 49–60.

4. Dobrydneva, Y., Williams, R.L., Katzenellenbogen, J.A.,Ratz, P.H., and Blackmore, P.F., Thromb. Res., 2003,vol. 110, no. 1, pp. 23–31.

5. Marcos, V., Perogordo, E., Espinosa, P., Pozuelo, M.,and Hooghuis, H., Anal. Chim. Acta, 2004, vol. 507,no. 2, pp. 219–227.

6. Inmaculada, G., Sarabia, L.A., Orti, C.M., andAldama, J.M., Anal. Chim. Acta, 2005, vol. 544, no. 1,pp. 26–35.

7. Xu, C.L., Peng, C.F., Liu, L., Wang, L.Y., Jin, Z.Y., andChu, X.G., J. Pharmac. Biomed. Analysis, 2006, vol. 41,no. 3, pp. 1029–1036.

8. Magliulo, M., Michelini, E., Simoni, P., Guardigli, M.,and Roda, A., Analyt. Bioanalyt. Chem, 2005, vol. 384,no. 1, pp. 27–30.

9. Kulmacz, R.J., Arch. Biochem. Biophys., 1986, vol. 249,no. 2, pp. 273–285.

10. Xu, C.L., Peng, C.F., Wang, L.Y., Hao, K., andJin, Z.Y., Food Agric. Immunol., 2006, vol. 17, no. 1,pp. 21–27.

11. Wang, L.Y., Peng, C.F., Chen, W., and Xu, C.L., FoodAgric. Immunol., 2008, vol. 19, no. 1, pp. 61–75.

12. Berlina, A.N., Zherdev, A.V., Dzantiev, B., andSakharov, I.Yu., Prikl. Biokhim. Mikrobiol., 2007,vol. 43, no. 5, pp. 614–620.

13. Mercader, J.V. and Montoya, A., J. Agric. Food Chem.,1999, vol. 47, no. 3, pp. 1285–1293.