fnbmwq

METHODS 22, 49–52 (2000)

d

Monoclonal Antibody-Based Enzyme Immunoassayfor Mercury(II) Determination

Alexander Marx and Bertold Hock1

oi:10.1006/meth.2000.1035, available online at http://www.idealibrary.com on

Department of Botany, Technische Universitat Munchen at Weihenstephan,Alte Akademie 12, D-85350 Freising, Germany

A monoclonal antibody (K3C6) was developed against Hg(II)and applied in different enzyme immunoassay (EIA) formats todetermine the test system with the highest sensitivity. A detectionlimit of 1.0 mg/L Hg(II) could be achieved with a competitiveormat in contrast to a detection limit of 2.1 mg/L Hg(II) with aoncompetitive EIA. A competitive displacement EIA yielded theest detection limit of 0.4 mg/L Hg(II) and was well suited toeasuring real samples. For this purpose different water samplesere diluted at least 1:10 to avoid matrix effects and subse-uently spiked with 1 mg/L HgCl2. Recovery of the spiked samples

was between 80 and 120%. © 2000 Academic Press

The toxicity of mercury depends on the metal speciesand the route of uptake into the body (1). It passesbiological membranes very easily, can bind to enzymes,and can disrupt vital functions. Accumulation takesplace in the body because of its highly lipophilic nature.Besides organic mercury compounds, which are themost toxic, there are three inorganic species that canbe found as environmental contaminants: elementaryHg, Hg(I), and Hg(II). Vapor of elementary Hg showsthe highest lipophilic properties and is therefore verytoxic, followed by Hg(II) and Hg(I). For toxicity assess-ment it is consequently not only interesting to deter-mine the total amount of mercury but also the differentspecies.

The total amount of mercury is conventionally ana-lyzed by physical methods, e.g., atomic absorptionspectroscopy. Most of these methods are not suited forspeciation of mercury at a low-cost level. In contrastimmunoassays are simple and inexpensive tools formetal speciation. Large numbers of samples can be

1 To whom correspondence should be addressed. E-mail:hock@weihenstephan.de.

1046-2023/00 $35.00Copyright © 2000 by Academic PressAll rights of reproduction in any form reserved.

screened for the occurrence of a metal species withinhours. Several antibodies (Abs) were already developedfor the detection of heavy metals (2–7). Basically thereare two approaches to generating Abs: (1) The heavymetal is bound to a chelator, e.g., EDTA; Abs raisedagainst this complex do not recognize the metal itselfbut the entire structure. (2) Abs are produced directlyagainst the heavy metal attached to a suitable im-munogen; the advantage of this method is that the Abrecognizes the free metal and not a cagelike structureas in the first approach.

We have recently developed the monoclonal antibody(MAb) K3C6, which is directed against Hg(II) (8). ThisMAb was used to improve assay conditions suited forthe measurement of water samples.

MATERIALS AND METHODS

Materials

Buffers and Solutions

Carbonate buffer, 50 mmol/L, pH 9.6.Phosphate-buffered saline (PBS), 40 mmol/L, pH 7.2

(10.15 mol/L NaCl).PBS washing buffer: PBS diluted 1:10, containing

0.5 ml/L Tween 20.Substrate solutions for horseradish peroxidase

(POD): (I) tetramethylbenzidine (TMB) 1.2 mmol/L,H3PO4 8.0 mmol/L, penicillin G 12.0 mg/L, dimethylsulfoxide (DMSO) 10.0% (v/v); (II) hydrogen peroxide 3mmol/L, Na2HPO4 z 12 H2O 3.5 mmol/L, NaH2PO4 z1H2O 132.0 mmol/L; one part of solution (I) and two

parts of solution (II) were mixed shortly before use.

Stopping solution for POD: 1 mol/L H2SO4.

49

wmdK

wwst

C

F

50 MARX AND HOCK

Mercury(II) StandardsHgCl2 1 g/L was dissolved in 2% (v/v) HNO3 (stock

solution) to prevent loss of mercury during storage (9,10). The stock solution was stable for 1 month. It wasfurther diluted with PBS for the preparation of Hg(II)standards. Safety precautions, e.g., wearing gloves,were employed in handling mercury(II).

BSA–GSH ConjugateBovine serum albumin–glutathione (BSA–GSH)

conjugates were provided by Professors A. Rigo and M.Scarpa (University of Padova, Padua, Italy) and pre-pared as described (8).

AntibodiesThe MAb K3C6 (IgG1), directed against mercury(II),

has been described recently (8). It was purified byprotein A affinity chromatography from protein-freecell culture supernatant. Labeled goat anti-mouse Abswere purchased from Pierce and Sigma.

MethodsNoncompetitive Enzyme Immunoassay against

Mercury(II)BSA–GSH conjugate (10 mg/mL) was adsorbed to the

surface of a 96-well microtiter plate (200 ml/well, over-night, 4°C). After blocking with 300 mL 1% gelatin for90 min, Hg(II) standards ranging from 0.1 to 1000 mg/L

ere added to the wells (200 ml/well). During the 30-in incubation period Hg(II) was bound to the sulfhy-

ryl groups of the BSA–GSH conjugate. The MAb3C6 (200 ml/well, diluted 1:5000, 1-h incubation time)

was added to each well, followed by incubation of 200mL POD-labeled second anti-mouse Ab for 1 h(1:10,000). The plates were washed with washingbuffer between steps. All dilutions were made in PBSexcept for the coating reaction (carbonate buffer). Fi-nally the enzyme–substrate reaction was carried out,

FIG. 1. Influence of pH on the absorption in the EIA at 100 mg/LHgCl2. Absorptions were normalized by setting the absorption at pH7.2 to 100%.

ts1

resulting in the development of a blue color propor-tional to the amount of bound Hg(II). The reaction wasstopped with H2SO4, yielding a yellow color, and ab-sorption was measured at 450 nm with an EIA reader(Titertek Multiscan II, Flow Laboratories). Data anal-ysis was carried out with a commercial EIA softwarepackage (EIA3, Flow Laboratories). The assay was re-peated at least three times to calculate the mean testrange.

Competitive EIABSA–GSH-coated plates (10 mg/mL) were blocked

with 1% gelatin. Then 200 mL HgCl2 (100 mg/L) wasincubated for 30 min. Subsequently 100 mL of K3C6(1:2500) and 100 mL HgCl2 in the range 0.1–100 mg/L

ere incubated for 1 h. A second Ab labeled with PODas used in the last step for the detection of bound

pecific Ab (see above). The plates were washed threeimes between incubation steps.

ompetitive Displacement EIA for Hg(II)Microtiter plates coated with BSA–GSH (10 mg/L)

and blocked with 1% gelatin were incubated with 200mL HgCl2 (100 mg/L) for 30 min. Subsequently, 200 mLof the MAb K3C6 (1:5000) was incubated for 1 h. ThenHgCl2 standards (0.01–100 mg/L) were added and in-cubated for 30 min followed by a labeled second Ab asmentioned above. After each incubation step the mi-crotiter plates were washed three times.

Influence of pH and Salt Concentration on Hg(II)Coupling EfficiencyThe influence of the pH and salinity of the buffer on

the efficiency of coupling of Hg(II) to the BSA–GSHconjugates was checked with the noncompetitiveHg(II)-EIA. Instead of PBS (40 mmol/L, pH 7.2) dis-tilled water or a phosphate buffer (pH 7.2) with differ-ent saline concentration (10–200 mmol/L) was used forthe salinity experiments. For the pH experiments

IG. 2. Influence of different salinities on the Hg(II) EIA. Absorp-ions, which were achieved with 100 mg/L HgCl

2 in different PBSolutions, were normalized. Absorption at 40 mmol/L PBS was set to00%. The coefficient of variation was below 11%.

a

0

Fd

51Hg(II) DETERMINATION BY EIA

phosphate buffer (40 mmol/L) with different pH values(2–11) was applied. Hg(II) standard solutions were pre-pared with the different buffers and added to the BSA–GSH conjugates, which were adsorbed to a microtiterplate. All other steps were done according to the non-competitive EIA. Absorption of all variants was nor-malized by transformation to %A according to theequation

%A 5A 2 A0

AC 2 AC0z 100 [1]

where A is the absorption at a certain pH or saltconcentration with 100 mg/L HgCl2; A 0, absorption at acertain pH or salt concentration without HgCl2; AC,bsorption at pH 7.2, 40 mmol/L with 100 mg/L HgCl2;

and AC0, absorption at pH 7.2, 40 mmol/L withoutHgCl2.

RESULTS AND DISCUSSION

The MAb K3C6 was first applied in a noncompetitiveEIA for the detection of Hg(II). BSA–GSH conjugatewas adsorbed to the surface of a microtiter plate. Thefree thiol groups of the conjugate were able to bindHg(II) very tightly from Hg standard solutions (11, 12).Then the MAb K3C6 was bound to the conjugate. Theamount of bound MAb was detected with a secondenzyme-labeled MAb. This assay format yielded a testrange between 2.1 and 21.9 mg/L Hg(II) and a middle ofthe test (IC50) of 7.3 6 2.96 mg/L Hg(II).

To improve the sensitivity of the assay, the influenceof pH and salinity on the coupling of Hg(II) to theBSA–GSH conjugate was investigated. Standard solu-tions (100 mg/L HgCl2) were prepared in PBS of differ-ent pH and applied in the EIA. All other steps of theassay were carried out with the standard PBS (pH 7.2,

FIG. 3. Competitive and competitive displacement Hg(II) EIA (ab-sorption range, 0.2–1.6).

tH

40 mmol/L). The highest absorption was achieved witha coupling buffer at pH 7.2. Decreasing or increasingthe pH resulted in a strong reduction of the absorptionas shown in Fig. 1. This effect could be due to struc-tural changes in the BSA–GSH conjugate (13). Thiolgroups, which are responsible for the binding of Hg(II),may no longer be available for the coupling reactionafter this conversion.

We also observed an influence of salinity on theefficiency of coupling of Hg(II) to the BSA–GSH conju-gate. Hg(II) standards (100 mg/L HgCl2) were preparedin phosphate buffer at different salt concentrations(10–200 mmol/L PBS) or distilled water, respectively.As shown in Fig. 2 there was a significant decrease inabsorption with salt concentrations in the range 0–20mmol/L in comparison to the control (40 mmol/L). Also,with high salinity there was a decrease in absorptionbut to a much lesser extent than with low salinity.

The results of both the pH and salinity experimentsclearly demonstrated that the Hg EIA is very sensitivetoward changes in assay conditions. PBS at pH 7.2 anda concentration of 40 mmol/L proved to be optimal forthe Hg(II) binding reaction. Measurement of real sam-ples by the noncompetitive EIA therefore requires ad-justment of pH and salinity.

For further improvement of assay sensitivity a com-petitive EIA (14) was applied. In this assay format, freeHg(II) and immobilized Hg(II) compete for free bindingsites of the MAb K3C6. In the absence of free Hg(II) theavailable Ab binds to the mercury immobilized on theBSA–GSH conjugate. After incubation of the second Aba strong absorption signal can be detected in this case.If there is an excess of free Hg(II) it will block thespecific Ab binding sites and no signal is expected. Atest range of 1.0–4.9 mg/L Hg(II) and an IC50 of 2.1 6.65 mg/L Hg(II) were achieved (Fig. 3). Further in-

crease in sensitivity was achieved when the MAb wasbound to the BSA–GSH–Hg conjugate before freeHg(II) was added (14). In this competitive displace-

IG. 4. Determination of Hg(II) in water of different sources: (1)istilled water, (2) tap water, (3) Lake Mensa, (4) effluent of sewage

reatment plant. Each sample was diluted and spiked with 1 mg/LgCl2.

11

111

1

1

52 MARX AND HOCK

ment EIA free Hg(II) was added after a washing stepand displaced the bound Ab. If there was an excess ofHg(II) the Ab was completely displaced. This EIA for-mat yielded a test range of 0.4–2.4 mg/L Hg(II) and anIC50 of 0.8 6 0.32 mg/L Hg(II) as shown in Fig. 3. It isobvious that the highest sensitivity was obtained withthe competitive displacement EIA. It yielded a twofoldenhancement of sensitivity compared with the compet-itive EIA. This effect can be explained by the highersample volumes containing Hg(II).

Different water samples (tap water, distilled water,river water, and effluent of a sewage treatment plant)were analyzed with the competitive displacement EIA.Each sample was diluted 1:10, 1:100, and 1:1000 andsubsequently spiked with 1 mg/L HgCl2. Recovery ofthe added Hg(II) was high for tap water (119%) anddistilled water (81%) at a dilution of 1:10, as shown inFig. 4. Higher dilution (1:100 or 1:1000) does not sig-nificantly improve the results. The water sample of asmall lake, however, yielded recoveries of 46% at adilution of 1:10, 85% at a dilution of 1:100, and 102% ata dilution of 1:1000. Similar results could be obtainedwith the effluent of a sewage treatment plant (Fig. 4).The increasing recovery with higher dilutions may bedue to matrix effects. For example, unsuitable pH orsalinity of the samples can be responsible for underes-timation of the Hg EIA. Usually antigen–Ab reactionsdepend strongly on physiological conditions. ExtremepH or salinity has been reported to interfere with theantigen–Ab reaction (15, 16). Also, organic substancescan influence EIA performance as they bind Hg(II) veryrapidly (17). Spiking of samples with Hg(II) can resultin the complete binding of mercury to the organic mat-ter. In this case Hg(II) is no longer available for theHg(II) EIA.

CONCLUSION

Antibodies directed against heavy metals are of con-siderable interest for speciation and availability stud-ies. The MAb K3C6 against Hg(II) was applied in dif-ferent assay formats to achieve high sensitivitycombined with suitable assay conditions.

The competitive displacement EIA yielded the bestdetection limit [0.4 mg/L Hg(II)] and was well suited tothe measurement of real samples. Samples have to bediluted at least 1:10 to avoid matrix effects, whichresult from organic substances or extreme pH or salin-

ity. If the dilution of the samples is taken into accountthe detection limit of the competitive displacement EIA

1

is 4 mg/L Hg(II). Recovery of spiked samples was be-tween 80 and 120%.

Further work is required to check in which mannerextreme conditions (e.g., pH, salt concentration) influ-ence the displacement reaction in the competitive dis-placement EIA.

ACKNOWLEDGMENTS

We are grateful to Professors Adelio Rigo and Marina Scarpa(University of Padova, Padua, Italy) who prepared several coatingconjugates. We also thank Sabine Schapermeier for her skillful tech-nical assistance. The work was supported by the EC (EV5V-CT94-0357).

REFERENCES

1. Kaiser, G., and Tolg, G. (1980) in The Handbook of Environmen-tal Chemistry (Hutzinger, O., Ed.), Vol. 3, Part A, pp. 1–58,Springer-Verlag, Berlin.

2. Blake, D. A., Chakrabarti, P., Khosraviani, M., Hatcher, F. M.,Westhoff, C. M., Goebel, P., Wylie, D. E., and Blake, R. C. (1996)J. Biol. Chem. 271, 27677–27685.

3. Lindgarde, F., and Zettervall, O. (1974) Scand. J. Immunol. 3,277–285.

4. Reardan, D. T., Meares, C. F., Goodwin, D. A., McTigue, M.,David, G. S., Stone, M. R., Leung, J. P., Bartholomew, R. M., andFrincke, J. M. (1985) Nature 316, 265–268.

5. Stewart, J. D., Roberts, V. A., Crowder, M. W., Getzoff, E. D., andBenkovic, S. J. (1994) J. Am. Chem. Soc. 116, 415–416.

6. Wylie, D. E., Lu, D., Carlson, L. D., Carlson, R., Babacan, K. F.,Schuster, S. M., and Wagner, F. W. (1992) Proc. Natl. Acad. Sci.USA 89, 4104–4108.

7. Baker, B. L., and Hultquist, D. E. (1978) J. Biol. Chem. 253,8444–8451.

8. Marx, A., and Hock, B. (1998) Anal. Lett. 31, 1633–1650.9. Coyne, R. V., and Collins, J. A. (1972) Anal. Chem. 44, 1093–

1096.0. Lo, J. M., and Wal, C. M. (1975) Anal. Chem. 47, 1869–1870.1. Casas, J. S., and Jones, M. M. (1980) J. Inorg. Nucl. Chem. 42,

99–102.2. Sarkar, B. (1983) Life Chem. Rep. 1, 165–207.3. T. Peters, T., Jr. (1985) Adv. Protein Chem. 37, 161–245.4. Marx, A., Krotz, E., and Hock, B. (1998) Anal. Lett. 31, 1651–

1661.5. Yarmush, M. L., Antonsen, K. P., Sundaram, S., and Yarmush,

D. M. (1992) Biotechnol. Prog. 8, 168–178.6. Phillips, T. M. (1992) in Journal of Chromatography Library

(Heftmann, E., Ed.), Vol. 51A, pp. A309–A338, Elsevier, Amster-dam.

7. Babcan, J., and Sevc, J. (1994) Ekologia (Bratislava) 13, 199–205.