Development of an enzyme-linked immunosorbent assay with monoclonal antibody for quantification of homovanillic in human urine samples

Departments of
¹ Pharmacy,
² Pharmacology, and
³ Chemical Pathology, Faculty of Medicine, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong.
^a Author for correspondence. Fax 852-2603-5295; e-mail yeepingho{at}cuhk.edu.hk.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
A monoclonal antibody to homovanillic acid (HVA) was prepared by synthesis of a HVA-protein conjugate (HVA-ovalbumin) as an immunogen, immunization of mice, and the subsequent hybridization technique. Monoclonal antibodies were screened on the basis of sensitivity, specificity, and accuracy. An indirect ELISA was developed for quantification of HVA in human urine. The assay was characterized and shown to have high specificity, with cross-reactivities to vanillylmandelic acid and normetanephrine at 0.18% and <0.1%, respectively. The assay coefficients of variation were <10% within the working range of 0.5–40 mg/L. Initial results from testing urine samples of patients with neuroblastoma and other diseases were validated by HPLC, suggesting that this ELISA method is a reliable and convenient system for quantification of HVA in urine and can be used in the mass screening of neuroblastoma in infants.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Neuroblastoma is the most common solid tumor of childhood with an incidence of ~1 in 7000 in children under the age of 5 (1)(2). It arises from cells of the sympathetic nervous system and has the characteristic of secreting dopamine (DA)¹ and its chief metabolite, homovanillic acid (HVA), in excess (3)(4)(5). Improvements in methodologies for measuring concentrations of HVA and/or DA have led to an increasing use of these compounds as markers in diagnosis of neuroblastoma in infants. The routine measurement of HVA and DA involves analytical methods such as gas chromatography–mass spectrometry (6)(7) and HPLC with electrochemical detection (8)(9)(10). Although regarded as accurate and reliable in HVA quantification, HPLC with electrochemical detection and gas chromatography–mass spectrometry are unsuitable as screening methods because they are time-consuming, technically complicated, and expensive. Moreover, they involve sample pretreatment, and rates of requesting repeat tests are high if the compounds are not resolved by these two methods. An immunoassay that uses monoclonal antibodies against HVA or vanillylmandelic acid (VMA) has been described previously for the screening of neuroblastoma (11). Large-scale mass screening of neuroblastomas in Austria, using this assay, has shown that an immunoassay method can be used for neuroblastoma screening, but it requires a backup analytical technique such as HPLC with electrochemical detection or gas chromatography–mass spectrometry to exclude false-positive samples (12). More recently, a competitive enzyme immunoassay utilizing a monoclonal antibody to HVA has been reported (13). We present here a sensitive and specific indirect competitive enzyme immunoassay for the quantification of HVA in urine samples, using a different strategy in immunogen synthesis, which produces a monoclonal antibody against a possibly different epitope of HVA-immunogen conjugate. The sensitivity, specificity, and validity of the assay are reported.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
apparatus
Indirect solid-phase ELISAs were performed with 96-well microtiter plates (Microtest III, Falcon 3912; Becton Dickinson), and results were read by a microplate reader (Dynatech MR5000). Thin-layer chromatography plates (silica gel F-254, aluminum backing) were from Alltech Associates, Inc. Dialysis membrane (Spectra/Por^®; M_r cutoff, 12 000–14 000) was from Spectrum (Spectrum Medical Industries, Inc.).

chemicals
HVA, VMA, 3,4-dihydroxyphenylacetic acid, 4-hydroxyphenylacetic acid, 3,4-dihydroxymandelic acid, DA, normetanephrine, metanephrine, epinephrine (adrenaline), vanillic acid, vanillin, phenylacetic acid, dry 1,2-dimethoxyethane, ethyldimethylaminopropyl carbodiimide, N-hydroxysuccinimide, diphenyldiazomethane, keyhole limpet hemocyanin (KLH), ovalbumin (OVA, chicken egg), human serum albumin, pyridine, and goat anti-mouse IgG peroxidase were from Sigma Chemical Co. Benzophenone hydrazone and methylbromoacetate were from Aldrich Chemical Co. Acetonitrile (HiperSolv^TM for HPLC), methanol (HiperSolv for HPLC), and flash column chromatography silica (22–40 µm) were from BDH Limited. Dichloromethane was from Mallinckrodt Baker Inc. All other chemicals used were of reagent grade and were from Sigma.

hva-protein carrier synthesis
A previously reported synthetic scheme (Fig. 1 ) was followed for the production of HVA-immunogen conjugates, HVA-OVA and HVA-KLH (14). In brief, benzhydryl-protected HVA (Fig. 1 , I) was synthesized by mixing diphenyldiazomethane in dichloromethane (57.4 mL of a 200 g/L solution) and HVA (5 g, 27.4 mmol) in 100 mL of a mixture of 800 mL/L dichloromethane and 200 mL/L methanol. HVA diester (II) was produced by dissolving I in dry dimethoxyethane (111.8 mL) and treated with sodium hydride (600 g/L dispersion, 0.98 g, 24.6 mmol), and methylbromoacetate (4.1 g, 2.5 mL) was subsequently added. Two portions of aqueous sodium hydroxide (1 mol/L, 17 mL) were added to the mixture of II and tetrahydrofuran (170 mL) with water to give benzhydryl-protected HVA derivative (III). Benzhydryl-protected immunogen HVA-OVA or HVA-KLH (IV) was then synthesized by treating III (dissolved in acetonitrile) with N-hydroxysuccinimide (97 mg, 0.8 mmol) and dimethylaminopropyl carbodiimide hydrochloride (0.4 g, 2.1 mmol) and mixed with 0.5 g of OVA or KLH in 50 mL of 10 g/L NaCl. The protecting benzhydryl ester group of compound IV (HVA-OVA or HVA-KLH) was removed by treating IV with aqueous sodium hydroxide (1 mol/L), cooled, and sufficient aqueous citric acid (1 mol/L) was added to adjust the pH to ~13. The mixture was then dialyzed and lyophilized to give V (immunogen HVA-OVA or HVA-KLH).

View larger version (9K): [in this window] [in a new window]   Figure 1. Synthesis of HVA derivatives I, II, and III and HVA-protein conjugates IV and V.

monoclonal antibody production
Ten male C57 mice (25–30 g) were initially immunized intraperitoneally with 1.4 mg of immunogen (Fig. 1 , V) dissolved in 2 mL of normal saline and 2 mL of Freund's complete adjuvant (water-in-oil emulsification). Freund's incomplete adjuvant was used for four subsequent booster injections at 4-week intervals. The final booster of 1.4 mg of HVA-OVA in 2 mL of normal saline only was injected 3 days before the harvesting of lymphoid cells. Spleen cells of the mice with the highest affinity antiserum for HVA were selected and fused with NS1 myeloma cells as described (15). After cloning, cells were cryopreserved and subcultured in flasks; the supernatant was collected and kept at 0–4 °C for subsequent use in the indirect ELISA.

indirect competitive elisa
Microtiter plates (96-well) were coated with HVA-KLH (10 mg/L) in coating buffer (0.05 mol/L phosphate buffer, pH 7.2) and incubated at 4 °C overnight. To each well washed with phosphate-buffered saline-Tween (9.0 g of NaCl in 74 mL of 0.6 mol/L NaH₂PO₄, 176 mL of 0.6 mol/L Na₂HPO₄, and 0.5 mL of Tween 20, made up to 1 liter with deionized distilled water, pH 7.2), we added 50 µL of urine sample and HVA calibrator (HVA in phosphate buffer, pH 7.2) or phosphate-buffered saline-Tween buffer, premixed with 50 µL of culture supernatant (1:100 dilution) containing monoclonal antibody against HVA, and incubated overnight at 4 °C. After the plates were incubated 1 h at room temperature, 100 µL of goat anti-mouse IgG peroxidase (1:1000 dilution) was added to each well, and the plates were incubated for 1 h at room temperature. One hundred microliters of o-phenylenediamine (10 mg in 25 mL of citrate-phosphate buffer with 0.3 g/L H₂O₂) were added to each washed well, and the wells were incubated at room temperature for 10–15 min, according to color development. To stop the reaction, 250 g/L H₂SO₄ was added, and the plates were read by an automatic plate reader at 490 nm.

calculations
A linear log-logit transformation method was used for the fitting of the calibration curve (16). The ratio B/Bo (B and Bo represent the bound enzyme activity measured in the presence and absence of competitor, respectively) was expressed as a function of the logarithm of the antigen concentration. All measurements for samples and calibrators were duplicated and quadruplicated for determining Bo values. Nonspecific binding was obtained by using an incubation mixture in which the specific monoclonal antibody was replaced by an equal volume of buffer. In all cases, nonspecific binding did not exceed 0.2% of the total enzyme activity introduced in the assay. The minimum detectable concentration was calculated as the concentration of the competitor inducing a significant decrease (3 SD) in Bo. The precision profile of the calibration curve was obtained by eight measurements of each HVA concentration in control urine samples with added HVA and expressed in terms of the CV vs HVA concentration (17).

assay characterization
The specificity of the established indirect ELISA was studied by examining its capability to detect compounds of analogous structure to HVA. By establishing the corresponding calibration curves for each of HVA, its analogs, or protein carriers, the concentrations for each at a B/Bo value of 50% (EC₅₀) were determined. The cross-reactivities were thus expressed in terms of percentage of ratio of the EC₅₀ of HVA over those of other compounds. The optimum dilution of hybridoma cell culture supernatant that contained monoclonal antibody against HVA was determined. The method by which the calibration curve for each assay was constructed was also determined. Intra- and interassay variations were determined by measurements of a urine sample eight times.

human urine samples
Random urine samples were obtained from the Prince of Wales Hospital (Shatin, Hong Kong). Urine samples from 12 patients, ages from newborn to 13 years, with neuroblastoma and at various clinical disease stages were obtained at diagnosis. Seventeen urine samples from patients, ages from newborn to 16 years, suffering from various diseases (excluding any malignant diseases) were also used. All samples were analyzed by ELISA and HPLC, and the results were compared.

hplc analysis
A previously reported reversed-phase HPLC method was used for the preparation and analysis of urine samples (18). Briefly, ethyl acetate (3 mL) was added to a 3-mL urine sample, and the organic phase was separated and collected. The process was repeated twice with 2 mL of ethyl acetate. The extracts were pooled and evaporated to dryness under a stream of dry nitrogen; the residue was redissolved in 0.5 mL of water and filtered by 0.25 µm filters. Samples were kept frozen until analysis. A Hewlett-Packard Series 1050 System and HPLC 3D Chem Station software (Hewlett-Packard Co.) were used. An HP 1050 Series Diode-Array Detector was used for acquisition of signals at 280 and 285 nm, as well as spectra. A spherisorb ODS2 5U reversed-phase column (Jones Chromatography Ltd.) was used for the separation of catecholamines in samples. A 45-min gradient from 0% to 60% of buffer A (a mixture of 600 mL/L acetonitrile and 400 mL/L water) and from 100% to 40% of buffer B (0.1 mol/L KH₂PO₄, pH 2.5) was utilized. Peak identification was performed on the criteria of retention time and comparison of the spectrum with those of standard compounds.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
synthesis of hva-ova and hva-klh
Pure benzyl-protected HVA (I) was obtained as a light yellow oil with an infrared (IR) spectrum of 3400 and 3475 broad (OH), 1720 (CO), and 1590 (Ar) cm^-1. Diester (II) was obtained as a viscous light yellow oil with an IR spectrum of 1755 (methyl ester), 1730 (benzhydryl ester), and 1590 (Ar) cm^-1. Pure HVA derivative (III) was obtained as recrystallized and nonrecrystallized products with IR spectra of 1750 (methyl ester) and 1590 (Ar), and 1725 (benzhydryl ester) and 1590 (Ar) cm^-1, respectively. The reactions for the final steps of the synthesis of HVA-OVA and HVA-KLH were monitored by thin-layer chromatography.

monoclonal antibody specificity
The monoclonal antibody used in the ELISA was selected on the basis of sensitivity, specificity, and accuracy. Three representative clones, 1G, 4A, and 9F, yielded a high-titer ELISA endpoint at a dilution of 1:3125 after coating of microtiter plates with HVA-KLH but contained no activity against KLH. Specificities of all three clones were examined, and the results are shown in Table 1 . Clone 1G was the best of the three and was selected for further studies. Tissue culture supernatants from clones 1G, 4A, and 9F all showed low specificity to VMA, which is abundant in control and patient urine samples. The EC₅₀s of 1G, 4A, and 9F at a 1:100 dilution were 3.6, 155, and 160 g/L, respectively, making 1G the most sensitive clone of the three. The tissue culture supernatant from clone 1G was collected twice a week and pooled. When an antibody dilution curve was used , there was no difference between (a) the tissue culture supernatants that were collected at different times, 1 week apart over a period of 3 months from the initiation of the culture; (b) one tissue culture supernatant that was stored under different conditions for up to 1 year, including storage at 4 °C, -20 °C and freeze-dried; and (c) one tissue culture supernatant either subjected to fractionation by IgG cut with sodium sulfate, or not (results not shown).

calibration curve
A routine ELISA calibration curve at 1:100 dilution of supernatant containing monoclonal antibody (1G) is shown in Fig. 2 . A typical fivefold dilution curve for urine samples, using the same dilution of supernatant, was shown to be parallel to the calibration curve (Fig. 2 ).

View larger version (18K): [in this window] [in a new window]   Figure 2. Typical calibration curve () for HVA assayed with the indirect ELISA and typical dilution curve () for the urine samples.

precision and cross-reactivities
The day-to-day assay CVs (n = 8) were consistently <10% within the range of 0.5–40 mg/L HVA (2.7 to 220 µmol/L, Fig. 3 ); this range was thus defined as the working range of the assay. The minimum detectable concentration was 0.3 mg/L (1.6 µmol/L). The assay cross-reactivities are shown in Table 1 .

View larger version (15K): [in this window] [in a new window]   Figure 3. The precision profile of the indirect ELISA illustrates nonuniform error, expressed as CV (%), in the concentration estimate as a function of HVA concentration. The working range of the indirect ELISA was established as 0.5–40 mg/L HVA (2.7–220 µmol/L), where imprecision was <10%.

comparison with an hplc method
The present ELISA method was used to assay 29 urine samples that were also analyzed by HPLC (for details, see Materials and Methods). From the 12 patients diagnosed with neuroblastoma, 3 showed clearly increased urinary HVA by both methods, compared with a reference interval of 2–8 mg/L for the age range of patients whose urine samples were analyzed (19). Linear regression analysis of ELISA and HPLC results gave a line with a slope of 1.06, an intercept of -0.50 (S_yx = 4.63), and r of 0.95 (Fig. 4 ).

View larger version (11K): [in this window] [in a new window]   Figure 4. HVA in 29 random urine samples from patients with neuroblastoma () and other diseases (), assayed by ELISA vs HPLC. The correlation coefficient was 0.95.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
Catecholamines are analogous in structure, and both HVA and VMA have high concentrations in control and patient urine samples (20). Thus the development of quantitative immunoassays for HVA or VMA requires very specific monoclonal antibodies against each of the metabolites. In 1989, a urinary mass screening system using monoclonal antibodies against HVA and VMA was first reported (11). Preparation of the antigens involved the Mannich reaction, which linked protein carrier human serum albumin directly to VMA or HVA (21). The monoclonal antibodies generated were shown to have cross-reactivities against HVA and VMA, although the extent was not detailed. Only one other immunoassay using a monoclonal antibody against HVA has been reported (13). In this recent report, the antigen HVA-KLH was synthesized by linking KLH to the R₄ group of HVA (Table 1 ). The monoclonal antibody against HVA selected for the ELISA assay was shown to have cross-reactivity to VMA and normetanephrine at 0.5% and 0.8%, respectively.

In our study, we used a different approach, as reported previously (14). Derivatives from each synthetic step were purified and characterized before coupling onto a protein carrier, OVA or KLH, at the final step (Fig. 1 ). Esterification of HVA gave I (Fig. 1 ), in which the carboxyl group was protected from participating in subsequent reactions. The phenol hydroxyl was then alkylated with methyl bromoacetate to give diester II. The methyl ester was successfully deprotected without concomitant hydrolysis of benzhydryl ester-protected HVA (III), and was suitable for subsequent coupling to protein carrier OVA or KLH. These reactions gave good yields, and structural assignment for compounds I, II, III was supported by IR spectroscopy, melting point determination, and thin-layer chromatography.

The most common functional group among all the groups in catecholamine structures is R₁. It is desirable to direct the specificity of monoclonal antibodies towards R₃ and/or R₄. Antigens in which HVA was coupled to protein carrier OVA or KLH via R₁ in theory would generate more-specific monoclonal antibodies against it. It is also a strategy for generating monoclonal antibodies against other catecholamines with high specificity. The specificity study confirmed that monoclonal antibodies generated in our study indeed had high specificity against HVA, and cross-reactivity to other structural analogs, VMA in particular, was negligible. The ELISA method developed was accurate and reliable for quantification of HVA in urine samples, as was confirmed by HPLC analysis. The good correlation of the immunoassay with HPLC can be attributed to the design of the antibody and the subsequent improved specificity. The advantages of our ELISA method also include large amount of monoclonal antibody readily available from culture medium of hybridoma cells instead of from mouse ascites, as reported (13). After testing more urine samples, particularly those from infants suffering from neuroblastoma, we believe that this indirect ELISA method can be used effectively in the mass screening of neuroblastoma in infants.

View larger version (10K): [in this window] [in a new window]   Figure .

	Acknowledgments

 
This study was supported by Direct Grant No. MD 94031 (to Y.-P. Ho) and a postdoctoral fellowship (to R. Z. Shi) from The Chinese University of Hong Kong, which we gratefully acknowledge. We also thank Chung Shun Ho at the Department of Chemical Pathology, Prince of Wales Hospital, Hong Kong, for providing patient urine samples.

	Footnotes

 
¹ Nonstandard abbreviations: DA, dopamine; HVA, homovanillic acid; VMA, vanillylmandelic acid; KLH, keyhole limpet hemocyanin; OVA, ovalbumin (chicken egg); IR, infrared; and EC₅₀, concentration of the inhibitor at a B/Bo value of 50% in an inhibition curve.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Young JL, Ries LG, Silverberg E, Horm JW, Miller RW. Cancer incidence, survival and mortality for children under 15 years of age. Cancer 1986;58:598-602. [Medline] [Order article via Infotrieve]
2. Bernstein ML, Leclerc JM, Bunin G, Brisson L, Robison L, Shuster J, et al. A population-based study of neuroblastoma incidence, survival, and mortality in North America. J Clin Oncol 1992;10:323-329. [Abstract]
3. Robinson R, Smith P. Urinary phenols in stress. Nature 1960;186:240.
4. Greenberg RE, Gardner LI. Catecholamine metabolism in a functional neural tumour. J Clin Investig 1960;39:1729.
5. Smellie JM, Sandler M. Secreting intrathoracic ganglioneuroma. Proc R Soc Med 1961;54:327.
6. Davis BA, Durden DA, Boulton AA. Simultaneous analysis of twelve biogenic amine metabolites in plasma, cerebrospinal fluid and urine by capillary column gas chromatography-high resolution mass spectrometry with selected-ion monitoring. J Chromatogr 1986;374:227-238. [ISI][Medline] [Order article via Infotrieve]
7. McGill AC, Seviour JA, Dale G, Craft AW. Screening for neuroblastoma by gas chromatography/mass spectrometry in the northern region. Ann Clin Biochem 1988;25:132-133.
8. Fitzgibbon M, FitzGerald RJ, Tormey W, O'Meara A, Kenny D. Reference values for urinary HMMA, HVA, noradrenaline, adrenaline and dopamine excretion in children using random urine samples and HPLC with electrochemical detection. Ann Clin Biochem 1992;29:400-404.
9. Henderson MJ, Heney D, McGinlay JM, Lewis I, Bailey C. Measurement of dopamine, HVA and HMMA in untimed urine samples: establishment of age-related reference data in children. Ann Clin Biochem 1992;29:162-167.
10. Mashige F, Ohkubo A, Matsushima Y, Takano M, Tsuchiya E, Kanazawa H, et al. High-performance liquid chromatographic determination of catecholamine metabolites and 5-hydroxyindoleacetic acid in human urine using a mixed-mode column and an eight-channel electrode electrochemical detector. J Chromatogr B 1994;658:63-68. [ISI][Medline] [Order article via Infotrieve]
11. Yokomori K, Hori T, Tsuchida Y, Kuroda M, Yoshioka M. A new urinary mass screening system for neuroblastoma in infancy by use of monoclonal antibodies against VMA and HVA. J Pediatr Surg 1989;24:391-394. [ISI][Medline] [Order article via Infotrieve]
12. Kerbl R, Urban CE, Ambros PF, Lackner H, Ladenstein R, Spuller E, et al. Screening for neuroblastoma in late infancy by use of EIA (enzyme-linked immunoassay) method: 115000 screened infants in Austria. Eur J Cancer 1996;32A:2298-2305.
13. Taran F, Frobert Y, Creminon C, Grassi J, Olichon D, Mioskowski C, et al. Competitive enzyme immunoassay with monoclonal antibody for homovanillic acid measurement in human urine samples. Clin Chem 1997;43:363-368. [Abstract/Free Full Text]
14. Gallacher G, Ho Y-P. Homovanillic acid: synthesis of derivatives and production and characterisation of antibodies. Biogenic Amines 1992;9:99-114.
15. Grassi J, Frobert Y, Lamourette P, Lagoutte B. Screening of monoclonal antibodies using antigens labeled with acetylcholines-terase: application to the peripheral proteins of photosystem 1. Anal Biochem 1988;138:436-450.
16. Rodbard D, Lewald JE. Computer analysis of radioligand assay and radioimmunoassay data. Acta Endocrinol 1970;64:79-103.
17. Ekins RP, Newman B. Theoretical aspects of saturation analysis. Acta Endocrinol 1970;147:11-36.
18. Krustulovic AM, Zakaria M, Lohse K, Bertani-Dziedzic L. Diagnosis of neural crest tumours by reversed-phase high-performance liquid chromatographic determination of urinary catecholamine metabolites. J Chromatogr 1979;186:733-748. [ISI][Medline] [Order article via Infotrieve]
19. Lentner C, ed. Geigy scientific tables. Vol. 1. Units of measurements, body fluids, composition of the body, nutrition, 8th ed. Basel, Switzerland: Ciba-Geigy Ltd, 1981:55, 99..
20. Rosano TG, Swift TA, Hayes LW. Advances in catecholamine and metabolite measurements for diagnosis of pheochromocytoma. Clin Chem 1991;37:1854-1867. [Abstract/Free Full Text]
21. Yoshioka M, Aso C, Amano J, Tamura Z, Sugi M, Kuroda M. Preparation of monoclonal antibodies to vanillylmandelic acid and homovanillic acid. Biogenic Amines 1987;4:229-235.