J BlOLUMlN CHEMILUMIN 1995; 10: 1-7

A New Highly Sensitive Chemiluminescent Assay of Alkaline Phosphatase Using Lucigenin and its Application t o Enzyme lmmunoassay

Mitsuko Kitamura, Masako Maeda and Akio Tsuji School of Pharmaceutical Sciences, Showa University, Hatanodai, Shinagawa-ku, Tokyo 142, Japan

The chemiluminescent reaction of lucigenin w i th various reducing sugars and reducing compounds has been studied. It was found that dihydroxyacetone gave the most intense chemiluminescence (CL). We have developed highly sensitive chemiluminescent methods for alkaline phosphatase (ALP) based on the production of dihydroxyacetone using NADP’ or glycerol-3-phosphate as substrate. The detection limits for ALP using each substrate were 1.25 x lo-” mol/assay and 2.5 x lo-” mol/assay, and the coefficient of variation (n = 7) was 2.8% and 3.7%. respectively. We have also applied the method using NADP’ as substrate in enzyme immunoassays (EIA) for cholecystokinin (CCK) and human chorionic gonadotropin (hCG). CCK-8 (octapeptide sulphated form of a carboxy terminal fragment of CCK) concentrations released from alimentary canal o f rat were assayed using the chemiluminescent EIA (CLEIA) and a fluorimetric EIA (ALP label). The correlation between CCK-8 values obtained by these methods was y = 1.04~ + 18.21, r=0.946, n=28. hCG values in serum and in urine were measured. The correlation between hCG values in serum samples obtained using the CLEIA and a time-resolved fluoroimmunoassay (TR-FIA), and in urine samples obtained using the CLEIA and the fluorimetric EIA using ALP were satisfactory. The correlations were y = 1 .OOx - 0.04, r = 0.997 (n = 51) and y = 1 .OOx - 0.03, r = 0.999 (n = lo), respectively.

Keywords: chemiluminescence; alkaline phosphatase; enzyme immunoassay; lucigenin; chemiluminescent enzyme immunoassay

INTRO D U CTlO N

Chemiluminescent (CL) assay techniques are attractive in clinical chemistry because of their sensitivity, rapidity of reaction and wide dynamic range. In order to increase the sensitivity of enzyme immunoassay (EIA), we have applied chemi- luminescent reactions to the detection of enzyme labelled antigens and antibodies. Lucigenin ( N , N’-dimethyL9,9’-bisacridiniurn dinitrate) (1) is one of the classical organic CL reagents, and is commonly used for analytical purposes.

Chemiluminescence of lucigenin is produced by addition of either hydrogen peroxide or organic reducing compounds in alkaline solution.

Nieman et al. (2-5) reported the chemiluminescent assay of biologically important reducing com- pounds, such as glucose, glucuronic acid, hyaluronic acid, ascorbic acid and creatinine. Previously, we have developed a flow injection analysis of various sugar and substances with a lucigenin-based CL assay (6) . We also developed a CL assays of invertase using sucrose (7) as substrate, and of ALP using ascorbic acid

CCC 0884-3996/95/010001-07 8 1995 by John Wiley & Sons, Ltd

Received 28 August 1993 Revised 19 October 1993

2 M. KITAMURA, M. MAEDA AND A. TSUJI

ALP

pH9.8 Method 1 : Glycerol-3-phosphate I) Glycerol + H3PO4

ALP

pH9.8 Method2 : NADP' I=) NAD' + H,PO,

Glycerol dehydrogenase FH20H Glycerol + NAD+ 7) F=O + NADH

CHzOH Dihydroxyacetone

@ Lucigenin Q Figure 1. Chemiluminescent assay of ALP using lucigenin

2-O-phosphate (8) as substrate and lucigenin as the CL reagent.

We have now developed highly sensitive methods for the assay of ALP using NADP' or glycerol-3-phosphate as substrate and lucigenin as CL reagent (Fig. l), and investigated their application in EIAs for CCK and hCG.

MATERIALS AND METHODS

Materials

Alkaline phosphatase (ALP, EC 3.1.3. l), NADP', glycerol-3-phosphate, NAD', glycerol dehydro- genase (Glycerol DH, EC 1.1.1.6) were purchased from Boehringer Mannheim-Yamanouchi Co. (Tokyo). Glycerol and lucigenin were purchased from Sigma Chemical Co (St Louis, USA) and Tokyo Kasei Co. (Tokyo), respectively. Other chemicals used were of analytical reagent grade. Cholecystokinin (CCK-octapeptide (26-33) (sulphate-form) and anti-CCK antibody were purchased from Peptide Institute Inc. and Otsuka Pharmaceutical Co. Ltd, respectively. Anti-rabbit IgG, hCG, anti-hCG antibody and its ALP-conjugate were gift from Tosoh Co. Ltd (Tokyo).

Instruments

Luminescence reader (Aloka, Co., LTD:BLR-301) and MLR-100 MICRO LUMINO READER

(Corona Electric Co., Ltd) were used for measuring chemiluminescence intensity.

M e t hods

CL assay using lucigenin. To a tube containing 50 pL of test compound solution in 0.1 ml/L Tris-HC1 buffer (pH 9.8), add lOOpL of 0.5mol/L KOH and 8 x YO lucigenin containing 0.005% Triton X-100 and then measure the CL intensity for the 10-s interval from 10 to 20s after addition of the CL reagent solution. Fifty compounds shown in Table 1 were tested by th s procedure.

CL assay for ALP. To each well of a microtitre plate containing 25pL of ALP standard solution or sample solution in 0.1 mol/L Tris-HC1 buffer (pH 9.8) containing 0.25 mmol/L MgC12, add 25 pL of either 0.1 mmol/L NADP' or 0.5 mol/L glycerol-3-phosphate in the same buffer and incubate for 2 h at 37°C with stirring. When using NADP' as substrate, also add to each well 25pL of 0.1 mol/L glycerol and 1 U/mL glycerol DH in 0.1 mol/L glycine-KOH buffer (pH 9.0), and incubate for 30min at 37°C with stirring. When using glycerol-3-phosphate as substrate, also add to each well 25pL of 5mmol/L NAD' and 1 U/ml glycerol DH in 0.1 mol/L glycine-KOH buffer (pH 9.0) containing 25 mmol/L MgC12, and incubate for 15min at 37°C with stirring. After incubation, stop the enzymatic reaction by addition of 1OpL of 0.25mol/L H2S04, and then measure the CL intensity from 10s after addition

CH EM I LU M I N ESCENT ALKALINE PHOSPHATE KI NASE 3

Table I. Relative chemiluminescence of various compounds with lucigenin

Compound CL intensity Compound CL intensity Compound CL intensity

Glucose Ribose Ara bi nose Xylose Lyxose Mannose R hamnose Galactose Fucose Methylglucoside Glucosamine Mannosamine Galactosamine 2-Deoxyglucose 2- Deoxygalactose Glucuronic acid Galacturonic acid

1.0 1.3 0.2 1.5 0.7 0.2

0.5' 0.1

1 .o 2.8 1 .I

-

-

-

-

3.0 0.7

N- Acetylglucosamine N- Acetylgalactosamine Gluconic acid Di hydroxyacetone Fructose Sorbose Cortisol 1 1 - Deoxycortisol Tetrahydrocortisol Dopa Dopamine Adrenalin Noradrenalin Serotonin Glycolaldehyde Glyoxal Glyceraldehyde

- - -

1596 20.3 12.0

1299 1249 1209 10.4 10.2 30.4 12.0 0.1

2.2 108

192

Sucrose Maltose Lactose Melibiose Gentiobiose Tre halose Cellobiose Chondrosine Raffinose Ascorbic acid Sorbitol lnositol Glutathione Dithiothreitol Pyruvate Uric acid

- 1.8

3.7 0.1

1.8 0.1

-

-

- 737 - - 0.2 3.2

0.7 -

of each 50pL of 1 mol/L KOH and 8 x YO lucigenin containing 0.005% Triton X-100 with a Corona MLR-100.

CLEIA for CCK. To each well of a microtitre plate coated with second antibody (affinity purified goat anti-rabbit IgG) (lOpg/mL), add 50pL of CCK standard or sample solution follows by 25pL of anti-CCK antibody and CCK-ALP conjugate (x 1000) in 0.1 mol/L Tris-HC1 buffer (pH 7.0) containing 0.1% gelatin and 0.05% NaN3, mix well, and incubate for 72h at 4°C. After incubation, the well is washed three times with saline containing 0.1% Tween 20 and 0.05% NaN3. To each well, add lOOpL of 0.025mmol/L NADP' in 0.1 mol/L Tris-HC1 buffer (pH 9.8) containing 0.25 mmol/L MgC12, and incubate for 2 h at 37°C with stirring. After incubation, add to each well 25pL of 0.15mol/L glycerol and 1.5 U/ml glycerol DH in 0.1 mol/L glycine-KOH buffer (pH 9.0), incubate for 30min at 37°C with stirring. Stop the enzymatic reaction by adding 1OpL of 1 mol/L H2S04, and then measure the CL intensity according to the procedure described above.

CLEIA for hCG. To each well of a microtitre-plate coated with anti-hCG antibody (lOpg/mL), add lOOpL of hCG standard or sample solution, and incubate for 1 h at 37°C with stirring. After incubation, the well is washed three times with saline containing 0.1% Tween 20 and 0.05%

NaN3. To each well, add 1OOpL of ALP conjugate (~4000) in 0.1 mol/L Tris-HC1 buffer (pH 7.0) containing 0.1% gelatin and 0.05% NaN3, incubate for 1 h at 37°C with stirring, and then wash the well as described above. After washing, bound enzyme activity in each well was determined as described above.

RESULTS

We examined the chemiluminescent reaction of biological compounds such as various sugars, reducing compounds and related substances with lucigenin. This CL reaction was applied to the chemiluminescent assay of enzyme activity for ALP, and used in an EIA based on an ALP label.

Optimization of assay conditions

In order to establish CL assays for various com- pounds (e.g. glucose, ascorbic acid and dihydroxy- acetone) with lucigenin, various assay conditions were examined. Constant chemiluminescent intensity was obtained using 0.5 mol/L potassium chloride, 4 x lucigenin, and 0.005% Triton X-100. The CL assay for ALP using lucigenin was also optimized. The concentration of glycerol- 3-phosphate and NADP' as substrate were examined in the range of 0.2 to 2mmol/L and 0.05 to 0.5 mmol/L, respectively, and 0.5 mmol/L

4 M. KITAMURA, M. MAEDA AND A. TSUJI

(glycerol-3-phosphate) and 0.1 mmol/L (NADP') were found to be optimal concentrations. The optimal concentration of magnesium chloride (MgC12) in 0.1 mol/L Tris-HC1 buffer (pH 9.8) was 0.25mmol/L MgC12. The optimal MgC1, concentration in 0.1 mol/L glycine-KOH buffe; (pH 9.0) for the assay using glucose-3-phosphate as substrate was 25mmol/L. For the assay using NADP' as substrate MgC12 was not added as it was detrimental to the chemiluminescent reaction. Optimal glycerol dehydrogenase concen- tration (0.5 to 5U/mL) was 1 U/mL in both CL assays. The optimal concentration of NAD' (2 to 20mmol/L) and glycerol (0.05 to 0.5mol/L) were 5 mmol/L and 0.1 mol/L, respectively. The reaction temperature was examined in the range 4-45"C, and 37°C was found to be optimal. As a result of examining the reaction time with ALP in the range 60 to 120min, 120rnin was chosen for the assay. The reaction time to produce dihydroxyacetone was varied in the range 5 to 60min for glycerol-3-phosphate and 15 to 120min in the case of NADP' reaction times of 15 min (glycerol-3-phosphate) and 30 rnin (NADP') gave a constant CL intensity.

Relative chemiluminescence of various compounds with lucigenin

As shown in Table 1, glucose (aldose class) and fructose (ketose class) gave intense CL, but sugars blocked at the C-1 position such as methylglucoside did not produce CL. Substances such as glycer- aldehyde and ascorbic acid also gave an intense CL. Steroids (e.g. cortisol) and related steroids having an a-hydroxycarbonyl group also gave an intense CL. Among the compounds tested, lucigenin gave the most intense CL with dihydroxyacetone and this was exploited in the ALP assay.

CL assay for ALP

The principle of the CL assay system for ALP is shown in Fig. 1. NADP' and glycerol-3-phos-

[count]

'05 1 lo4-

lo3:

102:

Figure 2. Standard curves of ALP using Methods 1 and 2

phate were used as substrates in Method 1 and 2, respectively. In the first reaction step, these substrates are dephosphorylated by ALP to produce NAD' or glycerol. In the second reaction step, generated NAD' or glycerol reacts in a glycerol dehydrogenase (glycerol DH) catalysed reaction to yield dihydroxyacetone. In the last step, dihydroxyacetone was measured by a CL assay using alkaline lucigenin solution. Standard curves for ALP using NADP' or glycerol-3- phosphate as a substrate are illustrated in Fig. 2. The range, detection limit, precision (coefficient of variation) and Km value for ALP obtained using each substrate are shown in Table 2. The detection limits of ALP were 1.25 x mol/assay (NADP' substrate) and 2.5 x mol/assay (glycerol-3-phosphate as substrate).

Applications to CLEIA

CLEIA for CCK-8. An EIA for CCK-8 was developed using ALP as the label which was detected by the CL assay using NADP' as substrate. The measurable range was 0.2 to

Table 2. Detection limits of ALP by each methods using lucigenin CL reactions

Substrate Range (mol/L)

Detection limit CV% Krn value (mol/assay) (n = 7) ( mmol/ L)

Glycerol - 3- phosphate 1 x 10-14-5 x lo-'' 2.5 x 1 0 - l ~ 3.7% 0.42 NADP+ 5 x 1 0-1 5-1 I o-" I .25 10-l 2.8% 0.27

C H EM I LU MI N ESC ENT ALKALI N E PH 0s PH ATE K I NAS E 5

0 100 200 300 CLElA [ pg I mL I

Figure 3. Correlation between CLEIA and fluorimetric EIA of CCK-8

50 pg/assay and the detection limit was 0.2 pg/ assay, corresponding to 1.7 x mol/assay. The detection limit of radioimmunoassay (RIA) (9,lO) usin the same antibody was 1.5 pg/assay (1.3 x lO-"mol/assay); therefore the detection limit of the CL method is 7.5 times more sensitive than the RIA method. The intra- and inter-assay precision for the standard curve were 4.5% (n = 11) and 8.8% (n = 9), respectively. CCK released from the alimentary canal of rat was assayed by the CLEIA and a fluorimetric EIA using 4-methylumbelliferyl phosphate (4-MUP). The correlation between CCK values obtained by the two methods was y = 1.04x+ 18.21, Y = 0.946, n = 28 (Fig. 3).

CLEIA for hCG. The measurable range was from 0.005 to 20 mIU/mL, the detection limit was

Y

4 l o 2 z , 10

10 O

y=l.OO~-Q.04 k0.997 v,.. (n=51) , _

l o o 10' lo2 lo3 lo4 10' l o 6 CLEIA [ mlU I mL ]

Figure 4. Correlation between CLEIA and TR-FIA of hCG in serum sample

0.005 mIU/mL, and the accuracy of the standard curve ranged from 3.9% to 6.9% (n = 11). In comparison to other methods, the detection limits of a fluorimetric EIA using ALP (1 1) as a label and a time-resolved fluoroimmunoassay using Eu-chelate as label (TR-FIA) (12) were 0.09 mIU/ mL and 0.3 mIU/mL, respectively. Therefore the CLEIA was 18 times and 60 times more sensitive than these methods. hCG concentration in serum and in urine were measured by CLEIA. The dilution curves were linear, and the intra- and inter-assay precision were acceptable as shown in Table 3 (intra-assay (n = 7), 4.2% to 11.7% and inter-assay (n = 7), 4.6% to 11.3%). The hCG values of same serum samples were assayed by TR-FIA (12) and the CLEIA. The results obtained are shown in Fig. 4. The correlation

Table 3. Intra-assay and inter-assay precision of the hCG assay

Sample Intra-assay (n = 7) Inter-assay (n = 7)

Average SD cv % Average SD cv % (ml U/mL) (mlU/mL)

0.589 1.18 3.24 3.47 6.53

75.6 446

6777

0.051 0.08 0.1 6 0.30 0.76 6.60

26.0 282

8.6 6.8 4.8 8.6

11.7 8.7 5.7 4.2

0.574 1.13 3.08 3.50 6.86

78.5 449

671 0

0.050 0.05 0.22 0.35 0.45 4.40

51 .O 423

8.9 4.6 7.3

10.0 6.5 5.7

11.3 6.3

6 M. KITAMURA, M. MAEDA AND A. TSUJl

CLEIA [ mlU mL I Fig. 5. Correlation between CLEIA and fluorimetric EIA of hCG in urine samples

between hCG values obtained by the two methods was satisfactory ( y = 1 . 0 0 ~ - 0.04, r = 0.997, n = 51). There were 46 samples with no measurable hCG by the TR-FIA method (less than 1 .O mIU/mL), but the CLEIA detected hCG in these samples (the range was from 0.08 to 1.45mIU/mL). A fluorimetric EIA using ALP (1 1) as a label is used as our routine method for measuring hCG in urine samples. Urine samples were reassayed by CLEIA and fluorimetric EIA. As shown in Fig. 5, the correlation obtained was satisfactory ( y = 1 .00~ - 0.03, r = 0.999, n = 10). There were 12 samples having no measurable hCG by the routine method (less than 0.5mIU/ mL), but CLEIA detected hCG in these samples (the range was from 0.06 to 0.77 mIU/mL).

DISCUSSION

ALP has been widely used as the label enzyme in EIA and it generally assayed by spectrophotometry or fluorophotometry. Recently, Bronstein and Schaap have developed adamantyl 1,2-dioxetane aryl phosphates (e.g. AMPPD) as novel CL substrates for ALP (13-16), and Miska et al. (17,18) have reported D-luciferin-O-phosphate as a BL substrate for ALP. We have also developed CL assays for ALP using NADP+/ethanol/alcohol dehydrogenase (19) or an enhanced CL reaction using enzyme cycling (20). The sensitivity of the present method based on a lucigenin reaction is slightly lower than those of the AMPDD method and NADPf-enhanced enzyme cycling method.

However, the cost of reagents is lower than the AMPPD method and assay procedure is more simple than the NADP'-enhanced enzyme cycling method.

The CL assay for ALP NADP' substrate was applied in EIAs for CCK-8 and hCG. Cholecysto- kinin (CCK) is known to play an important role in regulating pancreatic secretion, gallbladder con- traction, and intestinal motility. However, physio- logical and pathophysiological behaviour of endogenous CCK remains partially unexplained because there were no sensitive analytical methods for this analyte. We have developed a highly sensitive CLEIA system using lucigenin for CCK-8 employing the region-specific CCK-33 antiserum. Human chorionic gonadotropin (hCG) is a glyco- protein hormone, employed for the diagnosis of pregnancy. It is produced by the placenta and is produced by malignant tumours such as cancer of the chorion. We have applied CL assay of ALP using NADP' as substrate to a sensitive determina- tion of hCG in serum for the purpose of early detection of malignant tumours.

In conclusion, the results indicate that micro amounts of various biological substances, such as dihydroxyacetone, can be determined by a CL assay based on the lucigenin reaction. The CL assay of ALP based on the CL reaction of lucigenin with dihydroxyacetone is sensitive, and the reagents used in the proposed methods are inexpensive. Further studies on the application of this system are in progress in our laboratory.

Acknowledgements

The authors thank staff of the Central Clinical Laboratory of Showa University Hospital for serum and urine samples. This work was supported in part by Grants in Aid from the Ministry of Education, Science and Culture of Japan.

REFERENCES

1. Gleu K, Petsch W. Die Chemiluminescenz der Diemethyl-diacridylium-salze. Angew Chem

2. Veazey RL, Nieman TA. Chemiluminescent determination of clinically important organic reductions. Anal Chem 1979;51:2092-5.

3. Steen RA, Nieman TA. Lucigenin chemilumi- nescence and bioluminescence determination of heparin. Anal Chim Acta 1983;155:123-9.

1935;48:57-72.

CHEMILUMINESCENT ALKALINE PHOSPHATE KINASE 7

4. Klopf LL, Nieman TA. Determination of con- jugated glucuronic acid by combining enzymatic hydrolysis with lucigenin chemi- luminescence. Anal Chem 1985;57:46-51.

5. Veazy RL, Nekimken H, Nieman TA. Chemi- luminescent reaction of lucigenin with reducing sugars. Talanta 1984;31:603-6.

7. Maed M, Tsuji A. Chemiluminescence flow injection analysis of biological compounds based on the reaction with lucigenin. Anal Sci

7. Shimizu S, Arakawa H, Maeda M, Tsuji A. Chemiluminescent assay for invertase activity by using lucigenin and its application to chemiluminescent enzyme immunoassay. BNSKAK(J Anal Chem Japan) 1988;37:123-7.

8. Maeda M, Tsuji A, Yang KH, Kamada S. Chemiluminescence assay of alkaline phosphatase using ascorbic acid 2-phosphate as substrate and its application to chemi- luminescent enzyme immunoassay. In: Stanley PE, Kricka LJ, editors. Bioluminescence and chemiluminescence: current status. Chichester: John Wiley, 1990: 119-22.

9. Himeno S, Tarui S, Kanayama S, et al. Plasma cholecystokinin responses after ingestion of liquid meal and intraduodenal infusion of fat, amino acid or hydrochloric acid in man: analysis with region specific radioimmuno- assay. Am J Gastroenterol 1983;78:703-7.

10. Ohgo S, Takemura J, Oki Y, et al. Radio- immunoassay of cholecystokinin in plasma. Clin Chem 1988;34:1579-1584.

11. Saito H, Okuyama D, Kozuka K, Chiba H, Yanaihara T. Study on the assay of serum LH, FSH, PRL, hCG, P-hCG by automatic enzyme immunoassay system (AIA- 1200). Hormone to Rinsho 1992;40:97-103.

12. Kim P, Harri S, Ilkka H, Erkki S, Tim0 L, Veli H, Pirjo T, Ulf-Hakan S. Time-resolved fluoroimmunoassay of human choriogonado- tropin. Clin Chem 1983;29:60-4.

1986;2: 183-6.

13. Bronstein I, Edwards B, Voyta JC. 1,2-Dioxetanes: novel chemiluminescent enzyme substrates. applications to immuno- assays. J Biolumin Chemilumin 1989;4:

14. Bronstein I, Voyta JC, Thorpe GHG, Kricka LJ, Armstrong G. Chemiluminescent assay of alka- line phosphatase applied in an ultrasensitive enzyme immunoassay of thyrotropin. Clin Chem 1989;35: 144-6.

15. Schaap AP, Chen TS, Handley RS, DeSilva R, Giri PP. Chemical and enzymatic triggering of 1,2-dioxetanes. Fluoride-induced chemilumi- nescence from (tert-butyldimethylsily1)oxy- substituted dioxetanes. Tetrahedron Lett

16. Schaap AP, Sandison MD, Haandley RS. Chemical and enzymatic triggering of 1 ,Zdioxetanes. Alkaline phosphatase-catalyzed chemiluminescence from an aryl phosphate- substituted dioxetanes. Tetrahedron Lett

17. Miska W, Geiger R. Synthesis and characteriza- tion of lucigenin derivatives for use in billumines- cence enhanced enzyme immunoassay. J Clin Chem Clin Biochem 1987;25:23-30.

18. Geiger R, Miska W. Bioluminescence enhanced enzyme immunoassay new ultra- sensitive detection systems for enzyme immunoassay, 11. J Clin Chem Clin Biochem

19. Maeda M, Arakawa H, Tsuji A. Chemi- luminescent assay of various enzyme activities and its application to enzyme immunoassays. J Biolumin Chemilumin

20. Kawamoto M, Arakawa A, Maeda M, Tsuji A. Sensitive chemiluminescence assay of reduced nicotinamide adenine dinucleotide using enzyme cycling and its applications. BNSKAK (J Anal Chem Japan) 1991;40:

99-1 11.

1987;28:1155-8.

1987;28:1159-62.

1987;25:31-8.

1989;4: 140-8.

5 3 7-42.