Nov 3, 1989 - sidad de Extremadura, 06071 Badajoz, Spain. 2To whomcorrespondenceshouldbe addressed. Nonstandard abbreviations: FIA-EC, flow- ...
CLIN. CHEM. 36/4, 662-665 (1990)
Competitive Heterogeneous Enzyme Immunoassay for Theophylline by Flow-Injection Analysis with Electrochemical Detection of p-Aminophenol E. PInlIla Gil,1
Hua 1. Tang, II. BrianHalsall,2WilliamR. Helneman,2and A. SanchezMIsIego1
A competitive enzyme-linked immunoabsorbent
assay based on the flow-injection amperometnc detection of p-aminophenol has been investigated with use of the materials and general procedure of a commercial kit for the determination of theophylline in human serum. The antibody is immobilized on glass beads, and the enzyme label is alkaline phosphatase (EC 3.1.3.1). The high currents generated during the electrochemical detection allowed a rapid (35 mm) and simple determination of theophylline throughout its therapeutic range (10-20 mg/L) and also in the subtherapeutic range (detection limit of about 80 Lg/L).
AdditIonalKeyphrases:alkaline phosphatase . electrochernical immunoassay
. p-aminophenyl phosphate
Enzymes, when used as labels in iinmunoassay methods, allow the development of diverse assay protocols, with adaptability to both homogeneous and heterogeneous assays. Moreover, they provide a signal amplification that is especially useful at low analyte concentrations. The techmque used to detect the enzyme product is a critical factor in the development of an enzyme iminunoassay method. Most routine clinical methods are based on the formation of colored or fluorescent products, which are detected spectrophotometrically. The application of electrochemical detection in enzyme iinmunoassays is under active investigation (1-8), given the potential for combining low detection limits with good selectivity. Both potentiometric and amperometric methods have been developed for use with homogeneous and heterogeneous assays (6). The coupling of immunoassay with liquid chromatography with electrochemical detection, or with flow-injection analysis with electrochemical detection (FIA-EC), has allowed the development of extremely sensitive assays (9, 1O). Although homogeneous procedures are faster and simpler than heterogeneous assays, the wide variety of compounds present in biological samples limits the use of electrochemical detection for this assay format. Because a high potential is required for the detection of enzymegenerated NADH, certain electroactive substances in biological samples, e.g., ascorbic acid, uric acid, and acetaminophen, can contribute to a high background signal. Proteins can also foul the electrodes by adsorption. Electrode protection (by use of a precolumn) to prevent fouling, the isolation of analytical signal (via a chromatographic col-
Biomedical Chemistry Research Center, Department of ChemUniversity of Cincinnati, Cincinnati, OH 45221-0172. ‘Departamento deQuiinica Analitica y Electroquimica, Universidad de Extremadura, 06071 Badajoz, Spain. 2To whom correspondenceshouldbe addressed. Nonstandard abbreviations: FIA-EC, flow-injection analysis with electrochemical detection; PAPP, p-aminophenyl phosphate; and EAE, 2-(ethylamino)ethanol. Received November 3, 1989; acceptedJanuary 30, 1990. istry,
662 CLINICAL CHEMISTRY, Vol. 36, No. 4,
1990
and the use of low electrode potential have been investigated as remedies for these problems (7, 8). On the other hand, use of heterogeneous assays, which require an additional separation step to dispose of all the sample material except that bound to the antibody, would also avoid electrochemical interferences from the extraneous sample components. A recently developed commercial heterogeneous enzyme immunoassay method is the EZBEADTM kit (Immunotech Corp., Boston, MA 02134) for determining the bronchodilater drug theophylline in serum. This heterogeneous solidphase competitive-binding enzyme immunoassay is based on competition between theophylline in the patient’s sample and a theophylline-alkaline phosphatase (EC 3.1.3.1) conjugate in the reagent for a limited number of binding sites on a bead coated with a highly specific anti-theophylline antibody. The amount of theophylline in the patient’s serum is inversely proportional to the amount of conjugate bound to the bead, which is determined by the amount of phenol enzymatically generated from phenyl phosphate. As marketed, the phenol produced in the method is detected spectrophotometrically by the colored complex formed when potassium ferricyanide (stopping reagent) is added. Here we assess the feasibility of using FIA-EC in the EZ-BEAD procedure for a theophylline assay by using p-aminophenyl phosphate (PAPP) as the enzyme substrate to generate p-aminophenol. The latter compound is detected by its oxidation current at a glassy carbon electrode in the FIA-EC system. The electrode reaction monitored is as follows: umn),
H2N-(j-_OH
p-aminophenol
-
HN==O=O
+
p-quinone
imine
2W
+
2e
Materials and Methods Reagents and Instrumentation EZ-BEAD enzyme immunoassay kits for theophylline were purchased from Immunotech Corp. 2-(Ethylaimno)ethanol (EAE) was obtained from Aldrich Chemical Co. (Milwaukee, WI 53201). Because concentrated EAE is a toxic material (ORL-MAM LD50: 1200 mg/kg), caution should be taken to ensure safe handling and disposal (11). The PAPP was synthesized from p-nitrophenyl phosphate (Boehringer Mannheim Biochemicals, Indianapolis, IN 46250), by catalytic hydrogenation in a hydrogen shaker (Parr Instruments Co. Inc., Moline, IL), with use of palladium on activated carbon (Aldrich) as a catalyst (12). The p-aminophenol was purchased from Sigma Chemical Co. (St. Louis, MO 63178); abnormal control serum was from Ciba Corning Diagnostics Corp. (h-vine, CA 92714). Glass disposable test tubes (12 x 74 mm) were a product of Fisher Scientific (Pittsburgh, PA 15219). Electronic digital pipets (from Rainin Instrument Co., Inc., Woburn,
the solutions. Electrochemical analyses were performed with flow-amperometric equipment from Bioanalytical Systems (BAS, West Lafayette, IN 47906). The amperometer was a BAS Model LC-4A. The thin-layer electrochemical cell had a glassy carbon working electrode, an Ag/AgCl (3 molJL KC1) reference electrode, and a stainless-steel auxiliary electrode. The injection volume was 20 L, and the tubing from the injection to the detector was 20 cm long. To prepare 1 molJL EAE solution, used as the mobile phase, we added 100 mL of the concentrated EAE to 900 mL of doubly-distilled water and adjusted to pH 9.8. We used a flow rate of 1.0 mL/min. PAPP solution (5 mmollL) was prepared in 1 mol/L EAE solution. p-Aminophenol was detected at +0.1 V. All the potentials are referred to an Ag/AgCl reference electrode.
120
MA 01801) were used to transfer
Procedure We modified the EZ-BEAD kit procedure, pipetting 25 tL of the kit’s standards or patients’ serum samples and 200 L of the kit’s alkaline phosphatase-theophylline conjugate sequentially into a series of 12 x 75 mm test tubes. An antibody-coated bead was added to each tube, then incubated for 30 mm at room temperature. Each bead was washed three times with water to remove any unbound material and placed in a test tube containing 1 mL of PAPP solution instead of the kit’s phenyl phosphate solution. After the recommended incubation time of 15 mm, we injected 20 L of this solution into the flow-injection system and recorded the oxidation current. if necessary (because of limitations in the upper current rate of the amperometer), we diluted the incubation solution with 1 mol/L EAE solution just before injection. Serum samples from patients receiving theophylline therapy were analyzed by this method and also by a fluorescence polarization immunoassay procedure, Abbott TDx (13), used at the Toxicology Laboratory of University Hospital, Cincinnati.
Results and Discussion Electrochemical Detection of p-Ammnophenol Alkaline phosphatase is a commonly used enzyme in electrochemical enzyme immunoassay, particularly with phenyl phosphate as substrate. The generated phenol is measured by its oxidation current at +0.80 V (4, 6). However, there are essentially two problems with the electrochemical detection of phenol: the high oxidation potential, which results in high background noise, and fouling of the electrode surface, owing to the electropolymerization of phenolic radicals at concentrations exceeding about iO moJiL. We have previously (8) described the use of PAPP as a better substrate than phenyl phosphate, because its product, p-aminophenol, is more easily oxidizable (at about +0.1 V) than is phenol and does not foul the electrode, even
a
100
b 80 C C
60 C
U
0
40
4 0
20
d
0
-5
5
0
20
15
Time (mm) Fig. 1. Decay of p-aminophenol signal in different buffers and pHs by FIA-EC at a glassy carbon electrode: (a) 1 mol/L, pH 9.8, EAE; (b) 0.1 mol/L pH 9.0, ammonium carbonate + 5 x 10 mol/L oxalic acid; (C) 0.1 mol/L, pH 9.0, ammonium carbonate; (c 0.1 mol/L, pH
9.8, ammonium carbonate with use of short incubation periods. Adding oxalic acid, which reportedly prevents air oxidation of ascorbic acid (14), to the ammonium carbonate solution substantially slows the decay rate of p-aininophenol (Figure lb). However, as seen in Figure la, EAE (1 mol/L, pH 9.8) is an even better buffer to minimize the air oxidation of p-aminophenol. In fact, although not shown, there is no significant decay of the p-aminophenol signal in this medium for at least an hour at room temperature. At the same time, EAE, because of its phosphate-acceptor ability, reportedly is one of the best buffers for enhancing activity of alkaline phosphatase (15). Using the analysis of Cornish-Bowden and Eisenthal (16), we calculated Km and V for the theophylline-alkaline phosphatase conjugate in EAE and ammoniuin carbonate at pH 9.0. Km and V,,, were 17.39 mmol/L and 50.63 nA/mm, respectively, in EAE, and 1.03 mmolIL and 6.97 nA/mm, respectively, in ammonium carbonate. The use of EAE allows a faster reaction rate, although Km is increased in this buffer. We chose a PAPP concentration of 5 mmol/L for the immunoassay procedure. The rate of nonenzymatic hydrolysis of PAPP is slow, only about 0.29 nmol/L per minute. Fresh substrate solutions were prepared daily. Figure 2 shows hydrodynamic voltammograms of paminophenol in EAE, and of EAE only, at the glassy carbon electrode in the flow amperometric system. For p-aminophenol, the region at which the current reaches a plateau is at +0.1 V; moreover, because EAE becomes electroactive above +0.2 V, the working potential chosen was +0.1 V. #{149} A standard calibration plot of p-aminophenol as mea-
‘00 BJ
L
at high concentration.
However, an important problem encountered with paminophenol is its susceptibility to air oxidation, which is faster at the high pH (9.0-10.0) essential for optimum activity of alkaline phosphatase. We therefore assessedthe decay rate of the p-anunophenol signal in different buffers and pHs (Figure 1). In ammomum carbonate (Figure ld), as reported previously (8), p-aminophenol signal decays so quickly at pH 9.8 (optimum for alkaline phosphatase activity) that the assay must be performed at pH 9.0 (Figure lc),
10
1200
-0.3
C.2
-0%
0 Ct #{163}i.at,00. 000001
0.2 (%
03
0
0.3
02
-0%
0 01.00.5.
Q
02 S1..0a1
0.3
Ca
05
CS
M
Fig. 2. FIA-EC hydrodynamic voltammograms: (A) 106 mol/L p. aminophenol in EAE, 1 mol/L, pH 9.8; (8) EAE only, 1 mol/L, pH 9.8 CLINICAL CHEMISTRY, Vol. 36, No. 4, 1990 663
with the flow-injection system shows a very wide linear response range for peak current vs p-aminophenol concentration, from 5 x iO to iO mol/L (y = -0.58nA + 49.56n.A p.mol1Lx, r = 0.999, S = 0.15n.A mol’L, n 00 15). The detection limit was about iO mol/L, defined as the p-aminophenol concentration for which the signal is twice the blank signal. Ten repeated injections of a pam.inophenol standard of 10-6 molJL gave reproducible peaks with a coefficient of variation (CV) of 0.8%. sured
Electrochemical
100
80 60 0 U
40
C,
Enzyme Immunoassay
20
The enzymatically generated p-aminophenol was easily monitored by electrochemical oxidation, giving very large currents after 15-mm incubation. The responses over 20 miii for three theophylline concentrations were linear (Figure 3). As expected, the presence of more theophyllune results in a smaller slope, because less conjugate is bound to the bead. A calibration plot was then constructed according to the manufacturers instructions (Figure 4). The y-axis values are equal to 100 C/C0, where C and C0 are the currents for the standard and 0.0 mgfL standard, respectively. Each point is the average of two separate determinations. The points lie on the straight line, y = 100.4 44.78x (r = 0.999. n = 5). The reproducibility of the results, i.e., interassay precision, is shown in Table 1 for each standard used. -
0
10
Theophylline Concentration (mg/L) Fig. 4. Standard calibration curve for theophylline standards in human serum, plotted on the semilogarithmic graph paper supplied with the EZ-BEAD enzyme immunoassay kit Assay conditions as in Fig.3. IncubationtIme: 15 mm
Table 1. Interassay PrecIsion of Theophylline Immunoassay with Electrochemical Oetectlon Signal C/C0,%5 Theophyllln. concn, mg/I. 2.5
5.0 10.0 20.0 40.0
Comparison of Methods: Patients’ Samples Serum samples from patients receiving theophylline therapy at University Hospital were analyzed by the FIAEC method and the Abbott TDx, but not on the same days. A good correlation was obtained within and below the therapeutic range, with a linear-regression coefficient (r) of 0.98 (Figure 5). Results for the single sample that was analyzed in the toxic range (>20 mg/L) also gave a good correlation, differing by 3.5% between procedures.
n
=
#{149}
71.63 56.86 42.04
SD 4.8
CV, %
3.9
5.4 3.9
5.8
2.2
7.0
2.9 1.4
26.98
5.3
each. C, signalfor standard;C,,,signal forD mg/L standard.
4
20 -J 0’
Shortened Assay The high concentrations of p-anunophenol generated during 15 miii of substrate incubation meant that a shorter procedure was possible, involving only 5 mm of substrate incubation. A calibration plot of p-aminophenol oxidation current vs theophylline concentration is shown in Figure 6, in the typical nonlinear form of competitive immunoassays. The currents shown range from 150 to 720 nA, high enough to suggest that the incubation time could be shortened yet
Mean 82.81
E
15
C 0
a
10
c 01
U
c 0
U
5
0 5
0
10
EEIA Concentration
15
20
(mg/I)
Fig.5. Comparison of theophylline concentration for patients serum samples by Abbott TDx (fluorescence polarization immunoassay, FPIA) and FIA-EC (electrochemical enzyme immunoassay, EEIA): y 1.Olx- 0.41 (n 23)
2000
=
1500
further if an automated precision.
C
1000
system was used to maintain
U
Detection Umit
0 01
a-
500
0
0
4
8
12
16
20
24
IncubationTime (mm)
Fig.3. Plot of peak current vs substrateincubationtime for theophyllinestandardsin human serum: (a) 0.0, (b) 10.0, and (C) 20.0 mg/L Substrate:5 mmol/L PAPP. Pro-injectiondilutionfivefold for all cases 664
CLINICAL CHEMISTRY, Vol. 36, No. 4, 1990
The detection limit of the manufacturer’s prescribed procedure was claimed to be about 1.1 mg/L. Electrochemical detection, in the modified procedure, had the same detection limit, well below the therapeutic concentration range (10-20 mg/L). The detection limit in the shortened electrochemical assay was about 1.2 mg/L. The high sensitivity of electrochemical detection also permitted us to dilute both the theophyllmne standards (with abnormal
800
600
Spanish Ministry of Education and Science and from the Autonomic Govermnent of Extremadura (Spain). The useful comments of George C. Barone III and Dr. Robert Thompson were gratefully appreciated. This work was supported by NSF grant CHE 8217045.
500
References
400
noassay for serum proteins by differential pulse anodic stripping
700
1. Doyle MJ, Halsall HB, Heineman WR. Heterogeneous immu-
voltammetry. Anal Chem 1982;54:2318-22. 2. Wehmeyer KR, Halsall HB, Heineman WR. Electrochemical investigation ofhapten-antibody interactions by differential pulse 200 polarography. Clin Chem i982;28:1968-72. 3. Weber SG, Purdy WC. Homogeneousvolt.ammetric immunoassay. A preliminary study. Anal Lett i979;12:1-9. 100 -10 0 10 20 30 40 50 & Doyle MJ, Halsall HB, Heineman WR. Enzyme linked immunoabsorbent assay with electrochemicaldetectionfor alpha-i acid Theophylline Concentration (mg/L) glycoprotein.Anal Chem 1984;56:2355-60. Fig.6. Direct plotof peak currentvs concentration oftheophyllmne 5. Smyth MR. Buckley E, RodriguezFJ, O’Kennedy R. Applicastandards in the shortenedassay procedure tion of adsorptive stripping voltammetry for a studyofthe reaction Incubationtime: 5 mm of mouse IgG with anti-mouse IgG. Analyst 1988;1i3:31-3. 6. Heineman WR, Halsall HB. Strategies for electrochemical immunoassay. Anal Chem i985;57:l32iA-31A. 7. Eggers HM, Halsall HB, Heineman WR. Enzyme immunoassay control serum) and the theophylline-alkaline phosphatase with flow amperometric detection of NADH. Clin Chem 1982; solution (with Ths-saline buffer, 0.1 mol/L, pH 10.4). Two 28:1848-51. calibration plots were constructed and compared for 108. Tang lIT, Lunte CE, Halsall HB, Heineman WR. p-Arniand 100-fold dilutions of both reagents. A lower detection nophenyl phosphate: an improved substrate for electrochemical limit (about 80 4t&gfL)was obtained with the 100-fold enzyme immunoassay. Anal Chim Acta i988;214:187-95. dilution procedure. This demonstrates that assays could be 9. Jenkins SH, Heineman WR, Halsall HB. Extending the detection limit of solid-phase electrochemical enzyme immunoassay to performed not only at lower cost (about half the cost per the attomole level. Anal Biochem 1988;168:292-9. test vs the manufacturers assay), but also with a substan10. Xu Y, HalsallHB, Heineman WR. Solid-phase electrochemical tially smaller sample volume (0.25 tL) than is usually used enzyme immunoassay with attomole detection limit by flow injecfor this assay (25 ML). tion analysis. J Pharm Biomed Anal 1990; in press. 11. The Sigma-Aldrich library of chemical safety data. Milwaukee, WI: Sigma-Aldrich Corp., 1988. In conclusion, electrochemical detection based on the 12. De Riemer LII, Meares CF. Synthesis of mono and dinucleoxidation current of p-aminophenol applied to the EZotides photoaffinity probes of ribonucleic acid polymerase. BioBEAD enzyme immunoassay procedure for theophylline chemistry 198i;20:1606-12. analysis in serum is feasible and practical. Because this is 13. Li TM, Benovic U, Buckler RT, Burd JF. Homogeneous a heterogeneous procedure, there is no interference by substrate-labeled fluorescent immunoassay for theophylline in electroactive substances, or electrode fouling; therefore, serum. Clin Chem i981;27:22-6. 14. Koupparis MA, Anagnostopoulou P, Malmstadt HV. Autoelectrode protection is not required. The high sensitivity of mated flow izjection pseudotitration of strong and weak acids, the electrochemical detection provides a faster procedure ascorbic acid, and calcium, and catalytic pseudotitrations of amithan does the manufacturer’s and also allows for determinopolycarboxylic acids by use of a microcomputer-controlled ananations of the analyte in the ,ug/L range. 300
We thank Dr. Amadeo Peaceand Mr. Michael Hassan from the Toxicology Laboratory of University Hospital for providing the patients’ samples, and Dr. Hans Ziinmer for permitting use of the hydrogen shaker. E.P.G. acknowledges support received from the
lyzer. Talanta 1985;32:41i-7. 15. McComb RB, Bowers GN, Posem S. In: Alkaline phosphatase. New York: Plenum Press, 1979:258-60. 16. Cornish-Bowden A, Eisenthal R. Statistical considerations in the estimation of enzyme kinetic parameters by the direct linear plot and other methods. Biochem J 1974;i39:721-30.
CLINICAL CHEMISTRY, Vol. 36, No. 4, 1990 665
