An improved ELISA with linear sweep voltammetry detection

Journal of Immunological Methods, 149 (1992) 115-120

115

© 1992 Elsevier Science Publishers B.V. All rights reserved 0022-1759/92/$05.00

JIM 06259

An improved ELISA with linear sweep voltammetry detection Feng Tie Q

a,

Aihua Pan

b,

Binggen Ru

b,

Wenqing Wang

a,

Yinhua Hu

C

Department of Technical Physics, b Department of Biology, Peking University, Beijing 100871, People's Republic of China, and C Department of Chemistry, Northwest University, Xi'an 71()()69, People's Republic of China (Received 19 March 1991, revised received 15 September 1991, accepted 13 December 1991)

An improved ELISA combined with linear sweep voltammetry detection of p-nitrophenol generated by an enzyme has been investigated in this study. p-nitrophenol, produced from alkaline phosphatase catalysing p-nitrophenyl phosphate, yielded an oxidative peak at 1.06 V (vs. Ag/ AgCl) with a wax-impregnated tubular graphite anode. Without separation, the small three-electrode system was directly inserted in the well of an ELISA plate for detection. The detection limit for p-nitrophenol was 1 X 10- 6 M, lower than that obtained by measuring the absorbance of p-nitropheno1. The feasibility of utilizing linear sweep voltammetry as a detection scheme was demonstrated by determining metallothionein, granulocyte-colony stimulating factor and Xenopus laevis keratin using the above new system. The method was simple, reproducible and much more sensitive than traditional spectrophotometry. Key words: ELISA; linear sweep voltammetry; Alkaline phosphatase; p-Nitrophenol

Introduction The reactions of an antibody with its corresponding antigen in immunoassays are measured by the detection of labels bound to the immunoreactants. A number of labels have been employed, such as radioactive tracers (radioimmunoassays, RIA), enzymes (enzyme immunoassays, EIA), and electroactive, chemiluminescent

Correspondence to: A. Pan, Department of Biology, Peking University, Beijing 100871, People's Republic of China. Abbreviotions: APase, alkaline phosphatase; EIA, enzyme immunoassays; ELISA, enzyme-linked immunosorbent assays; LSV, linear sweep voltammetry; McAb, monoclonal antibody; PNP, p-nitrophenol; PNPP, p-nitrophenyl phosphate; RIA, radioimmunoassays.

and fluorescent labels. Of these, RIA are extremely useful for diagnostic purposes and basic research, and have some important applications. However, RIA have several inevitable disadvantages due to the radioactive label, the need for expensive equipment, potential health hazards, and the necessity for special treatment of radioactive waste. Enzymes are nonradioactive labels, with enormous catalytic power and high specificity for the substrate. Assays employing enzymes (EIA and ELISA) can, therefore, be highly sensitive. In ELISA, enzyme-generated products are detected by their absorbance. More sensitive ELISA detection systems can be obtained using fluorogenic substrates (Shaiev et aI., 1980; Thomas et aI., 1986). However, the detection capability of spectrophotometers is quite limited. Some trace substances in blood, such as digoxin and metal-

116

lothioneins (MTs) *, could not easily be determined using a general EIA or ELISA. Therefore, a complete change of detection procedure may be more effective in improving sensitivity and detectability. Enzyme-generated products are usually electroactive. When electrochemical detection is combined with EIA or ELISA, a detection limit as low as 10 - 21 mol can be obtained (Halsall et al., 1988) and sensitive electrochemical techniques have been applied to immunoassays during the past decade (Doyle et al., 1987). Some research work by Heineman and Halsall's group have concerned electrochemical EIA in which alkaline phosphatase (APase) and phenyl phosphate substrate were commonly used, the phenol produced being measured by liquid chromatography with amperometric detection (Wehmeyer et at., 1987; Buckley et at., 1989). Detection limits of pg/ml for digoxin, orosomucold and IgG were obtained (Wehmeyer et at., 1983, 1984, 1986; Doyle et al., 1984). Because the high oxidative potential of phenol caused a high background in addition to corroding the electrode surface, Tang (1988) suggested the use of p-aminophenyl phosphate as a substrate whose product, paminophenol, would be more easily quantified than phenol. A1wis and Wilson (I985, 1987) described an electrochemical EIA using glucose oxidase and electrochemical detection of H 2 0 2 • Various homogeneous electrochemical EIA using g1ucose-6-phosphate dehydrogenase have been described (Wright et at., 1986; Eggers et aI., 1982; Ngo et aI., 1985; Broyles et aI., 1986). Although amperometric detection following liquid-chromatography separation has high sensitivity and detectability, it has to be applied in conjunction with more complicated equipment. In this report, an improved ELISA combined with simple, rapid linear sweep voltammetry (LSV) is described.

Materials and methods MT from Cd-induced rabbit liver and its monoclonal antibody (McAb) were prepared as previously described (Pan et aI., 1991). Xenopus laevis keratin (XL keratin) and its McAb were gifts from Professor Haojian Yu of the Department of Biology of Peking University, and granulocyte colony-stimulating factor (G-CSF) and its McAb were obtained from the Department of Infectious Disease of the Union Hospital, Beijing. Anti-mouse IgG APase conjugate, PNPP and PNP were purchased from the Sigma Chemical company. The 40-well polystyrene ELISA plates used were made by the Tianjin Plastics Factory, China. The washing solution (PBS-T) was pH 7.4, 1/15 M KH 2 P04-Na 2 HP04 buffer containing 0.05% Tween 20. The buffer-substrate solution (BSS; pH 10.4, 0.05 M glycine buffer + 5.5 mM PNPP + 0.5 mM MgCI 2 ) was prepared as described by Linhardt and Walter (1963), and stored at 4°C.

ELISA procedure 100 JLI of MT (or G-CSF, XL keratin) diluted with pH 8.0, 0.02 M Tris-HCI buffer were added to each well. After incubation overnight at 4°C, the plates were washed three times with PBS-To Then 100 JLI of the appropriate McAb were added, and the plates further incubated for 60 min at 37°C. After washing, 100 JLI conjugate (1/400 dilution) were added to each well, followed by 50 min incubation at 37°C. Next, the plates were washed three times with PBS-T and twice with 0.05 M ammonium carbonate (pH 9.5). After the addition of 100 JLI of BSS to each well and incubation at 37°C for 10-20 min, the reaction was terminated with either (a) 50 JLI of 2 M NaOH solution, and the absorbance read on a Minireader II (Dynatech); or (b) 50 JLI of 0.2 M H 2 S04 solution, followed by PNP detection using LSV.

Preparation of graphite rod anode • MetallothioneiDs (abbreviated MTs) are a ubiquitous class of low molecular weight proteins consisting of a single polypeptide chain of 61 amino acid residues, 20 of which are cysteines that chelate seven bivalent cations, leaving neither free thiol groups nor disulphide bridles.

A spectroscopic graphite rod, made in Shanghai, China, was filed down to 2 mm in diameter. After cleaning, it was immersed in melted wax for 10 min. The surface of the rod was then polished with waterproof abrasive papers using the following order of grain size: no. 170, 600, 800, and

117

1000. Before each electrochemical run, the rod surface was washed and then dried with filter paper.

Electrochemical detection After the H 2 S04 termination, LSV was carried out with a Model JP3-1 oscillopolarographic analyzer (Shangdon 7th Electronic Equipment Factory, China) coupled to a Laser pp40 X-Y printer plotter (Hong Kong) and a three-electrode system: wax-impregnated graphite rod anode, platinum wire auxiliary and Ag/ AgCI reference electrode. The initial potential was 0.70 V, the final potential 1.30 V, and the scanning rate about 300 mv Is. The 2.5th order derivative of differential peak currents were recorded.

Preparation of samples Three groups of six female Swiss mice (each about 20 g) were studied. In group 1, each mouse was injected with 0.1 ml of 1mg/ml CdCl 2 and killed 26 h later; in group 2, each mouse was injected with 0.1 ml ZnS04 (0.4 mg Zn/mouse) and reinjected with 0.8 mg Zn/mouse 2 days later. 2 days after the second injection, the mice were killed; in group 3, six normal mice were killed. The liver of each mouse was separately homogenized with pH 8.0, 0.02 M Tris-HCI buffer (2 or 3 ml/g wet tissue), and then centrifuged at 12,000 X g at 4°C for 5 min. The supernatant was heated in boiling water for 60-90 s and centrifuged as above. For the ELISA procedure the supernatant was diluted ten times for coating.

Results and discussion PNP has an oxidizing wave at 1.08 V in 0.1 M H 2 S04 solution and Fig. 1 illustrates representative voltammograms. The peak current (i p) changes very little after four successive scan!!. The linear relationship between i p and PNP concentration is shown in Fig. 2. In the ELISA procedure, the enzyme catalytic reaction occurs in a buffer solution which influences the oxidizing wave. Three commonly used substrate buffer solutions were compared (Table I). The 0.05 M glycine buffer solution, pH 10.4, containing 0.05 mM MgCI 2 , required the least amount of H 2 S04 for termination, had the smallest peak potential (Ep) and a good wave shape. ELISA-LSV for antigens (MT, G-CSF, keratin) were performed. The i p showed a linear relationship with antigen concentration (Figs. 35). Table II compares the two detection methods. TIre ELlSA-LSV results were more sensitive, the detection limits for antigen being 1-2 orders of magnitude lower than those of the ELISA. The mouse liver MTs were first tested using ELISA-LSV. The voltammograms of the three groups are shown in Fig. 6. When mice were injected with Cd or Zn, the peak heights were greatly enhanced. The MT contents of mouse liver were obtained from the standard curve of MT and are shown in Table III. It is evident that the mouse liver MT content increased about 100 times after the injection of Cd or Zn. Hu and co-workers have combined the ELISA with simple and rapid linear sweep polarography (LSP) in order to detect the product generated by horseradish peroxidase catalyzin~ o-phenylen-

TABLE I COMPARISON OF SUBSTRATE BUFFER SOLUTIONS Buffer

Volume of H 2 S04 used for termination Ep(V) Wave form

pH 10.4, 0.05 M glycine buffer + 0.55 mM MgCI 2

pH 9.8,1 M diethanolamine +0.25 mM MgQ 2

pH 9.5, 0.05 M ammonium carbonate

3O#,lofO.2M.

60 #,1 of2 M 1.18 Poor

70 #,1 of 0.2 M

1.04 Good

1.15 Fair

118 MT (rabbit) ng/1oo IJI 10

20

30

40

50

60

Ip

10IJA

T

IJA

60 50

20

40 30

10

20 1 30

1.20

1.10

100

0.90

10

0.80

E(V vs. Ag/AgCI)

2345678

Fig. 1. Characteristic linear sweep voltammograms for PNP in 0.1 M H 2 S04 , Voltammograms were recorded using 2.5th order derivative of current over four successive time scans. Interval t = 15 s.

Ip lolA

~

A

50

~I

40'

10:/

20

30~

20r 10f I

2

3 PNP

4

5 x 10·'

I

4

M

Fig, 3, Representative standard curve of metallothionein, Dilution of anti-mouse IgG-APase conjugate 1/400,

ediamine and H 202 substrates (Hu et al., 1991). They assayed a number of samples and showed that the ELISA-LSP results were more reliable than those of ELISA. The use of p-nitrophenyl phosphate (PNPP) as a substrate for APase in ELISA is popular because its product, p-nitrophenol (PNP), absorbs strongly at 405 nm (Tijssen, 1985), Our laboratory has developed an ELISA which uses a monoclonal antibody for the determination of MT (Pan et aI., 1991). In this paper, the ELISA was improved by using anode linear sweep voltammetry

B

/

2

MT(rabblt) ng/100 IJI

r. 09970 6

,

8 10x10· 6

PNP M

Fig, 2, Plots of peak current versus PNP concentration in 0,1 M H 2 SO4 , Scan rate: (A) 310 mV Is; (B) 550 mV Is,

TABLE II COMPARISON OF PEAK CURRENT AND LIGHT ABSORBANCE MT(ng) • Absorbance ip ("A) G-CSF (ng) Absorbance ip ("A) Keratin (ng) Absorbance ip ("A)

b

62,5 0,05 68,0

31.2 0,03 54,0

62,5 0,07,0,03

• Antigen amount/IOO ".1. b Given by Minireader II (Dynatech),

7,8 0,01 36,0

31.2 0,05,0,04 16,16

28,25 156 0,10 57,0

15,6 0,02 40,0

78 0,04 33.0

39 0,02 16.8

3,9 0,03 26,0 15,6 0,03,0,03 9,6,10 19 0,01 10.8

1.9 0,01 21.2

0,48 0,02 17,6

7,8 0,02,0,02 5,6,6,0 9,7 0.01 8.0

Blank 0,01 10,0 Blank 0,00,0,01 3,2,3,2

2,4 0.00 6.4

Blank 0.00 3.2

119 Ip

IIA

10 20 30 40 50 60 70 80

!

KeratIn ng/l001l1

Fig. 4. Plots of peak current versus keratin concentration. McAb dilution 1/400, anti-mouse IgG-APase conjugate dilution 1/400.

for the detection of PNP. When the enzyme-catalytic reaction had been terminated by addition of "2S04 solution, the three electrode system with a wax-impregnated tubular graphite anode could be directly inserted into the well of microtitre plates to detect PNP. Using the above conditions PNPP has no oxidizing peak. Determinations of MT, G-CSF and XL-keratin were performed by attaching the antigens directly to the solid phase and using monoclonal antibodies as detectors with anti-mouse IgG-APase conjugate as the second antibody. When both LSV and spectrophotometry are used for detection, the detection limits of the ELISA-LSV can be two orders of magnitude lower than those of ELISA. The detecting procedure can be performed within half a minute with

1.30 1.20 1.10 1.00 0:90 0.80 E(V vs. Ag/AgCll Zn-Induced mouse lIver MT, 1/30 dilution coating

,

!

,

1.30 1.20 1.10 1.00 0.900.80 E(Vvs. AgjAgCI) normal mouse liver MT 1/30 dilution

1.30 1.20 1.10 1.00 0.900.80 E(Vvs AgAgCI) Cd-Inducing mouse IiverMT,1/40 dilution cootlng

Fig. 6. Linear sweep voltammograms for mouse liver MT. The broken line represents the blank.

a simple and cheap linear sweep voltammetric analyzer. In the preparation of samples, the first centrifugation step removes heavy particles. Because MT is quite heat-stable, heating and the second centrifugation step can remove high molecular weight proteins. In fact, the MT contents of the samples are higher than the values given in Table III. The supernatants contain other low molecular weight proteins in addition to MT. When TABLE III MT CONTENT OF MOUSE UVER Group no. (2) Zninduced

(3) Normal

induced 20.0 3.71 %

18.9 4.93%

0.18 10%

(1) Cd,

1.30

,

I

1.20

1.10

,

1.00

,

,

0.90 0.60

E(V vs. Ag/AgCil

Fig. 5. Linear sweep voltammograms for granulocyte CSF. (1) 62.5 ng, (2) 31.2 ng, (3) 15.6 ng, (4) 7.8 ng.

MT (J.l.g/g wet tissue) Coefficient of variance (n =6)

120

coating the supernatants, MT competes with other proteins for attachment to the plate. As a result, only part of the MT coats the surface of the plate well. From the results of Cd/haemoglobin affinity assays, the MT contents are about ten times larger than indicated in Table III. However, this does not detract from the ELISA-LSV. Immunoassay procedures such as the double antibody sandwich assay can be used to solve these problems. ELISA combined with LSV is more sensitive and precise than the traditional ELISA procedure. References A1wis, W.u. and Wilson, G.S. (1985) Rapid sub-picomole electrochemical enzyme Immunoassay for immunoglobin G. Anal. Chern. 57, 2754. A1wis, W.U. and Wilson, G.S. (1987) Rapid heterogeneous competitive electrochemical immunoassay for IgG in the picomole range. Anal. Chern. 59, 2786. Broyles, C.A and Rechnitz, G.A. (1986) Drug antibody measurement by homogeneous enzyme immunoassay with amperometric detection. Anal. Chern. 58, 1241. Buckley, E., Smyth, M.R., Heineman, W.R. and Halsall, H.B. (1989) Some recent developments in electrochemical immunoassays. Anal. Proc. (London) 26, 5. Doyle, M.J., Halsall, H.B. and Heineman, W.R. (1984) Enzyme-linked immunoabsorbent assay with electrochemical detection for aI-acid glycoprotein. Anal. Chern. 56, 2355. Doyle, M.J., Wehmeyer, K.R., Heineman, W.R. and Halsall, H.B. (1987) Immunoassay by differential pulse polarography and anodic stripping voltammetry. In: T.T. Ngo (Ed.), Electrochem. Sens. Immunol. Anal. Plenum, New York, p. 87. Eggers, H.M., Halsall, H.B. and Heineman, W.R. (1982) Enzyme immunoassay with f1ow-amperometric detection of NADH. C1in. Chern. 28, 1848. Halsall, H.B., Heineman, W.R., Jenkins, S.H., et al. (1988) Electrochemical enzyme immunoassay. J. Res. Natl. Bur. Stand. 93, 491. Hu, Y., Song, J. and Kang, X. (1991a) A new method for clinical determination of hemoglobin(Hb) in plasma, serum and blood by linear sweep polarography. In: 1991 Pittsburgh Conference, Chicago, p. 809. Hu, Y., Liu, J.and Song. J., (l991b) Electrochemical enzyme immunoassay for detection of tobacco mosaic virus (TMV)

and its antibody. In: 1991 Pittsburgh Conference, Chicago, p. 1069. Hu, Y., Zheng, L., Song, J., et al. (199lc) Electrochemical enzyme immunoassay for the detection of human alphafetoprotein(AFP). In: 1991 Pittsburgh Conference, Chicago, p. 1068. Linhardt, K and Walter, K (1963) Phosphatase. In: H.U. Bergmeyer (Ed.), Methods of Enzymatic Analysis. Academic Press, New York, p. 783. Ngo, T.T., Bovaird, I.H. and Lenhoff, H.M. (1985) Separation-free amperometric enzyme immunoassay. Appl. Biotech. 11, 63. Pan, A and Ru, B. (1991) Preparation of anti-rabbit metallothionein monoclonal antibody and ELISA for the detection of metallothionein. Submitted. Shalev, A, Greenberg, A.H. and McAlpine, P.J. (1980) Detection of attograms of antigen by a high-sensitiviy enzyme-linked immunoabsorbent assay (HS-ELlSA) using a f1uorogenic substrate. J. Immunol. Methods 38, 125. Tang, H.T., Lunte, C.E., Halsall, H.B. and Heineman, W.R. (1988) p-Aminophenyl phosphate: an improved substrate for electrochemical enzyme immunoassay. Anal. Chim. Acta 214, 187. Thomas, D.G., Linton, H.J. and Garvey, J.S. (1986) Fluorometric ELISA for the detection and Quantitation of metalothionein. J. Immunol. Methods 89, 239. Tijssen, P. (1985) In: R.H. Burdon and P.H. Van Knippenberg (Eds.), Laboratory Techniques in Biochemistry and Molecular Biology, Practice and Theory of Enzyme Immunoassays. Elsevier, Amsterdam, p. 196. Wehmeyer, K.R., Doyle, M.J., Wright, D.S., Eggers, H.M. Halsall, H.B. and Heineman, W.R. (1983) Liquid chromatography with electrochemical detection of phenol and NADH for enzyme immunoassay. 1. LiQ. Chromatoar. 6, 2141. Wehmeyer, K.R., Halsall, H.B. and Heineman, W.R. (1985) Heterogeneous enzyme immunoassay with electrochemical detection: competitive and 'sandwich'-type immunoassays. Clin. Chern. 31, 1546. Wehmeyer, KR., Hasall, H.B., Heineman, W.R. et al. (1986) Competitive heterogeneous enzyme immunoassay for digoxin with electrochemical detection. Anal. ClIem. 58, 135. Wehmeyer, KR., Doyle, M.J., Halsall, H.B. and Heineman, W.R. (1987) Heterogeneous enzyme immunoassay with amperometric detection. In: T.T. Ngo (Ed.), Electrochem. Sens. Immunol. Anal. Plenum, New York, p. 309. Wright, D.S., Halsall, H.B. and Heineman, W.R. (1986) Digoxin homogeneous enzyme immunoassay using highperformance liquid chromatographic column switching with amperometric detection. Anal. Chem. 58, 2995.