- Email: [email protected]
A rate turbidimetric immunoassay for theophylline using biotin-avidin system
ELSEVIER
Clinica Chimica Acta 218 (1993) 59-71
A rate turbidimetric immunoassay for theophylline using biotin-avidin system Chan Oh*, Julie Kim, Betty Kearns, Anthony Cheng, Tom Dobashi Beckman Instruments. Brea. California. USA
(Received9 November 1992; ~evisionreceived22 March 1993; accepted 22 March 1993)
Abstract
We describe a biotin-avidin based rate turbidimetric homogeneous immunoassay for the determination of serum theophylline with a measuring range of 0-40 #g/ml. The assay features an avidin-biotin-labeled theophylline conjugate and a monoclonal antibody. The rate of change of turbidity caused by the antibody-conjugate complexing was monitored on the Beckman Synchron CX® 4 System at 340 nm and 37°C. Theophylline in a sample inhibited the complexation, and the extent of inhibition allowed the quantitation of theophylline in the sample. The within-run coefficient of variation (CV) was < 3,7% and the between-run CV was 5,2-9.2% depending on 1he theophylline concentration. Linear regression analysis showed a good correlation with an established fluorescence polarization immunoas~ay (r = 0.9866, n = 94). Key words: Theophylline; Turbidimetric immunoassay; Biotin.avidin
I. Introduction
Theophylline, 1,3.dimethylxanthine, is a potent smooth-muscle relaxant especially affecting bronchial smooth muscle. It is used extensively for the prevention and treatment of asthma, and for the treatment ofapnea and bradycardia in infants [!,2]. Monitoring of serum levels is important for the safe and effective use of theophylline because it has a narrow therapeutic range, and patients vary considerably in their * Corresponding author. 0009.8981/93/$06.00 © 1993 ElsevierScience Publishers B.V. All rights reserved. SSDI 0009.8981(93)05558-V
60
C Oh et al. / Clin. Chim. Acta 218 (1993) $9-71
ability to clear the drug. Serum concentrations of 5-15 ~g/ml theophylline reportedly control apneic spells in neonates without causing apparent side effects [3]. In most patients, theophylline serum concentrations of 10-20 ~g/ml effectively suppress chronic asthmatic and other bronchospastic symptoms [3]. Above 20 ~gtml, toxic effects such as nausea, headache, vomiting, diarrhea, irritability, and insomnia have been reported [4]. Severe toxicity is characterized by seizures and cardiac arrhythmias. The elimination kinetics of the drug are influenced by age, smoking, and liver function. Methods available for theophylline quantitation are gas chromatography [5,6], high performance liquid chromatographic techniques [7-9], and a variety of isotopic [10] and non-isotopic immunoassays. Non-isotopic immunoassays include the homogeneous enzyme immunoassay [ 11-15], nephelometric immunoassay [! 5-19], fluorescence polarization immunoassay (FPIA) [13,15,20-22], homogeneous substrate-labeled fluorescent immunoassay [23], apoenzyme reactivation immuno. assay system (ARIS) [24], particle enhanced turbidimetric inhibition assay [20,25], ligand displacement assay [26], and latex particle counting [27]. The biotin-avidin rate turbidimetric method described in this report represents a new homogeneous immunoturbidimetric method for the measurement of theophylline in serum. In addition to being specific, fast and reliable, the method allows the use of neat sample and small sample volume (3 ~!). Fig. ! illustrates the principle of the biotin-avidin assay.
2. Biotin-avidin based rate turbidimetric immunoassay The use of biotin-avidin for nephelometric inhibition immunoassay for small molecules was previously described by Oh and Sternberg [16]. The authors used biotinylated hapten and avidin in place of the conventional pseudo-antigen or developer antigen, such as polyhaptenylated apoferritin, Since each avidin molecule can bind up to four biotin molecules, the developer antigen thus takes the ~'orm, avidin-(biotin-hapten)4. When this tetrahaptenylated avidin conjugate (avidin conjugate) was mixed with an antibody specific to the hapten, immunoprecipitin reaction occurred at high rate. The fast immunoprecipitin reaction and the high scatter intensity from the immunocomplex formation appeared to be associated with the positive charge of the abundant lysine termini of avidin. This charge assisted immunoprecipitin reaction is specific and characterizes the unique feature of the biotin-avidin methodology. Under similar condition, streptavidin or charge neutralized avidin failed to produce the immunoprccipitin reaction with high rate. When the spacer length between biotin and avidin was about 27 A, steric hindrance between the Fab portion of the antibody and avidin permitted only two antibodies, one at each end of the avidin molecule, to bind to the avidin conjugate [16]. The binding led to the formation of the linear polymer, (antibody..haptenbiotin..avidin..biotin.hapten)n, as shown in Fig, !. In an assay medium containing polyethylene glycol, the linear polymer resulted in light scattering, and the rate of change of light scattering could be monitored on a nephelometer or turbidimeter. The hapten, when present in the sample, competed with the hapten-biotin of the avidin conjugate for the antibody and led to reduction of rate signal. Thus, the rate
C Oh et aL/Clin. Chim. Acta 218 (1993) 59-71
61
Without Theophylline in Sample: Formation of Complexes Resulting in Turbidity Chango
With Theophylline in Sample: Inhibition of Formation of Complexes Resulting in Oecrease in Turbidity Change
L •
Biotin
Drug
e.....~ Biotin-Drug
~
Avidin
Antibody Conjugate
Fig. I. Principle of biotin-avidin immunoturbidimetric assay.
of nephelometric or turbi~Jimetricresponse was inversely proportional to the concentration of the hapten in the specimen. The biotin-avidin assay presented in this report for the quantitation of theophylline is based on the turbidimetric rate measurement. Unlike nephelometry, turbidimetric measurement allows the use of neat sample without interference from excess light scattering. A theophylline monoclonal antibody solution (6-75 #1) was
62
C. Oh et al. / Ciin. Chim. Acta 218 (1993) 59-71
introduced into a reaction cell which contained the assay buffer and was maintained at 37°C. A steady optical absorbance level was observed until the neat sample, ranging from 3 to 25/zl, was added when a higher and steady optical absorbance baseline was reached. Subsequent addition of the avidin..biotin-theophyilineconjugate to the above mixture initiated the formation of the immunocomplex. The rate of turbidity change caused by the complex was then monitored spectrophotometricallyat 340 nm using a fixed read window of 8-64 s with reading to begin from 0-16 s after the trigger reagent was introduced. Theophyiline in the sample inhibited the complex formation, and a dose response curve was constructed for the quantitation of theophylline. 3. Materials and methods
3.1. Chemicals N-Hydroxysuccinimide (NHS), !,1 '-carbonyldiimidazole (CDI), 1,5-pentanediamine, 4-aminobutyric acid, dimethylaminocinnamaldehyde spray reagent were purchased from Aldrich (Milwaukee, USA). Ninhydrin was obtained from Calbiochem (San Diego, CA, USA). Biotin was available from Schweizerhall (Piscataway, N J, USA). Avidin was obtained from Boehringer (Mannheim, FRG). Theophylline for calibrator preparations was purchased from USP (Md, USA). 3.2. Buffers Buffer A. Phosphate-buffered saline (PBS) solution containing 0,01 mol/I sodium phosphate, 0.15 mol/I NaCI and 0.1% sodium azide, pH 7.0. Buffer B. Citrate.buffered saline (CBS) solution containing 0.025 mmol/I sodium citrate, 5 mmol/I disodium EDTA, 0.15 mmol/I NaCl and 0. I% sodium azide, pH 6.0. Buffer C, Phosphate buffer containing 0. I mmol/I sodium phosphate, 0.15 mol/I NaCI, 0.1% sodium azide and 2,6% polyethylene glycol, pH 7.0. 3.3. Serum specimens Serum specimens for correlation study were provided by Damon Laboratories (Rancho Cucamonga, CA, USA). 3.4. Theophylline calibrators Theophylline calibrators were prepared in a human serum pool, which was determined to be theophylline-free based on results from a commercially available fluorescence polarization kit. A stock solution containing 4 mg/ml of theophylline in 50 mmol/i NaOH was prepared. The stock solution was then diluted in the serum matrix to give theophyllinecalibrators consisting of six levels at 0, 2.5, 5, 10, 20 and 40 ~g/ml. 3.5. Antibody reagent Theophylline monoclonal antibody was produced by Beckman Instruments (Irvine, CA, USA). The antibody was diluted in phosphate buffered saline (buffer A) for assays.
63
C Oh et at./Clin. Chim. Acta 218 t1993) 59-71
O
H3C
O ,
N
0
H-N~N-H
0 H
H
O
CH5
Biotin-theophylline derivotive Fig, 2. Chemical structure of biotin-theophylline derivative
3.6. Biotin-theophyiline derivative The structure of the biotin-theophyilinederivative used in this study is shown in Fig. 2, and was synthesized by coupling theophylline-8-butyramidopentylamineto biotinamidobutyric acid as described below. Theophylline-8-butyramidopentylamine. Theophylline-8-butyric acid (14.1 ram), synthesized according to the method of Cook [28], was activated with CDI (16.5 ram) and NHS (16.7 ram) in 180 ml of dimethylformamide (DMF) at 70-75°C for 30 rain and at room temperature for 16 h. A solution of 1,5.diaminopentane (38.1 nun) in 100 ml DMF was introduced and the resulting mixture was stirred at room temperature for an additional 15 h. A white solid was formed and filtered. After the filtrate was evaporated, the resulting oil was chromatographed on a silica gel column. Elution with methanol/chloroform/ammonium hydroxide (25:71:4, by vol.) gave theophylline-g-butyramidopentylamine.The product was pure based on TLC using the same solvent system, and positive to ninhydrin spray for the presence of amino group (yield 19°/o, m.p. 125-126°C). Biotinamidobutyric acid. A solution of 4.aminobutyric acid (22 mm) in 75 ml of 0.1 molll NaHCO~ was added to 100 ml of DMF containing biotin-N-hydroxysuccinimide (20 ram), prepared as described by Jasiewicz [29]. After overnight stirring, the solid formed was filtered, washed with ethyl acetate, and dried to give 3.61 g of product (yield 50c%,m.p. 219-220°C). The product was pure as judged by TLC using dichloro.methane/methanol/acetic acid (67:32:1, by vol.) as the eluent, and positive to dimethylaminocinnamaldehydespray for the presence of biotin. Biotin-theophyiline derivative (Fig. 2). A solution of biotinamidobutyric acid (2.57 ram), CDi (2.83 ram) and NHS (2.~;', ram) in 50 ml of DMF was stirred at 70-75°C for 30 rain and at room temperature for 15 h. Theophyiline-8-butyramidopentylamine (2.57 ram) dissolved in 17 mi of DMF was then added, and stirring continued for an additional 15 h. The precipitate formed was filtered, washed with ethyl acetate, and recrystallized twice from hot methanol to afford 1.26 g of product (yield 74%). TLC revealed one spot using 25% methanol in chloroform as the cluent. The spot was positive to dimethylaminocinnamaldehyde spray indicating the presence of biotin. Avidin-biotin-theophylline conjugate. The avidin-biotin-theophyllineco:ljugate was prepared by mixing a solution of avidin (I I I rag) in 9 ml of 0. ! rnmol~!phosphate buffer (pH 7.4) and a solution of the biotin-theophylline derivative (5 rag) in 1 mi
64
C Oh el aL /Clin. C'him. Acto 218 (1993) 59-71
0"
/
.
30 °
0 O O/
i
o
O
y : 1.011X - 0 4 r = 0.~7
10
N~94
v
0
10
2O
30
FPIA (Thoo, pglrnl)
FiB, 3, Correlation of serum theophylline values obtained by biotin.uvidin immunoturbidimetric ussay and FPIA methods.
of methanol/dimethylsulfoxide (2: I, v/v), After I h of reaction followed by dialysis with the citrate buttered saline (CBS, buffer B), the conjugate concentrate was diluted with CBS for immunoassays, 3.7. FPIA reagents and assays A commercially available FPIA theophylline assay kit was employed for the serum sample correlation study, The reagents were handled and the assays run according to procedure supplied by the manufacturer, 3.8. Assay protocol on Beckman Synchron CX ® 4 System Reagents and buffers for the biotin-avidin based theophylline assay were introduced and mixed automatically by the Beckman Synchron CX ® 4 System employing
40
C. Oh el al. / Clin, Chim, Acta 218 (1993) 59-71
65
Ablorbance change/rain 0.3 --I-
0.2S
- - O - - - - - - OAY 0 2 --"¢~--
0.2
DAY 3 DAY 5
~ D A Y ?
0.15
0.1
0.06
/
0
I
I
5
10
15
20
25
30
35
40
Theophylllne (/~g/mi) Fig. 4. Stability of avidin.,biotin-theophylline conjugate at 4S°C.
the User Defined Reagents (UDR) mode. The assay protocol and addition sequence were as follows: (a) introduction of reaction buffer (buffer C, 230 ~l) and antibody (45 ~d), (b) addition of neat sample (3/d), i c) addition of avidin..biotin-theophylline conjugate (43/~l) 16 s after step (b) and (d) measurement of rate of change of turbidimetric response at 340 nm for 36 s at 37°C immediately after step (c).
C Oh et al./Clin. Chim. Acta 218 (1993) 59-71
66
Table ! Precisionevaluation Sample pool
Mean (t~g/ml)
S.D.
C.V.
(pg/ml)
(%)
4.7 15.5 34.9
0.173 0.521 1.012
3.7 3.4 2.9
A
4.6
B
15.4
C
35.2
0.422 0.801 2.091
9.2 5.2 5.9
Within-run precision ( n = 20)
A B C Between.run precision {n = 20)
4. Results 4. I. T h e o p h y l l i n e d o s e - r e s p o n s e c u r v e
A typical calibration curve for the biotin-avidin based theophylline assay on the Beckman Synchron CX ® 4 System is displayed in Fig. 4 as represented by the Day 0 curve. The assay covered a measuring range o f 0 - 4 0 ~g/ml, corresponding to a turbidimetric rate change of 0.286-0.069 absorbance units/rain at 340 nm, respectively.
4.2. Precision Within- and between-run precision of the assay were evaluated using three levels of commercial controls which contained mean theophylline concentrations of 4.6, Table 2 Cross.reactivity
Sample' containing theophylline and compound studied Compound
Theophylline observed (v ~g/ml)
% cross,reactivity (100 * 0 ' - 9.2)/x)
10 50 100 100 100 20 200 200
9.2 11.7 13.8 11.9 10.6 13.I 9.7 9.4
--
13
I 1,4
16.9
100
I !.0
1.8
Cone,
(x/~B/ml) b Caffeine Theobromin¢ 3-Methyixanthine 1,3-Dimethyluric acid I-Methyluric acid Uric acid
8-Chlorotheophylline Dyphyllinc
5.0 4.6 2.7 1.4 19.5 0.3 O.I
a Sampleswere prepared by mixingequal volumes of 20 Isg/mlof theophylline in a normal human serum pool and twice the listed concentration of the compound in the same matrix. b Sample containing theophylline only.
67
C Oh et al./Clin. Chim, Acta 218 (1993) 59-71
i 5.4 and 35.2 pg/ml of theophylline (Table !). The within-run precision (CV < 4%) was determined by running 20 replicates of each sample within the same run. Ten separate runs of the three controls gave the results for the between-run CV, with each control assayed in duplicate per run. The between-run CVs were 9.2%, 5.2% and 5.9%, respectively. 4.3. Specificity
The cross-reactivity of the theophylline metabolites and structurally related compounds was studied (Table 2). The cross-reactivity o f a compound was assessed by running as known in the theophylline assay a sample containing both the compound at or greater than its pharmacological concentration and theophyiline at 9.2 ;tg/ml. Cross-reactivity of a compound is defined as the percent increase of the theophylline concentration divided by the concentration of the compound in the sample containing 9.2 pg/ml o f theophylline. As can be seen from Table 2, 1,3dimethyluric acid and 8-chlorotheophylline exhibited some interference in the theophylline assay. The cross reaction observed with 1,3-dimethyluric acid poses no problem for serum theophylline measurements because methylated uric acids are not present in the serum of patients receiving theophylline therapy. Drugs such as dimenhydrinate, which contain 8-chlorotheophylline, could increase the apparent
concentration of theophylline. No significant cross-reactivity was observed for caffeine and 3-methylxanthine, the latter being one of the major metabolites of theophylline.
Table 3 Analytical recovery Serum sample
Theophylline supplemented
Theophylline target
Theophyiline observed
pool
(p8)
(/~g/mi)
(pg/ml)
A
0 4.9 9,7 19,4 29,I 38.8 0 4.9 9.7 29,I 38,8
C
0
% recovery
9,2 17,6 25,3 32.3
0 4,9 9,3 18,7 26.8 34.7
-103,6 100,6 106,0 105.9 107.3
-9.5 13.9 22.I 29.6
4.9 9. I 13.8 23.2 30.6
95.7 . 99.3 105,0 103.5
4,7
0
--
4.9
9.7 29. I
14,4 18.7 26.6
15.2 20.0 28.2
9.9
--
105,6 107.1 105.9
38,8
33.9
35.3
104.1
68
C Oh et aL ~C/in. Chim, Acta 218 (1993) 59-71
4.4. Correlation with FPIA A group of 94 serum samples covering a range of approx. 1-30/~g/ml was assayed by FPIA and by the avidin-biodn rate turbidimetric method on the Beckman Synchron CX ® 4 System. A good correlation between the two methods was observed (Fig. 3). The linear regression analysis resulted in the following correlation parameters: Synchron = 1.0106 * TDX - 0.38, r = 0.9866.
4.5. Analytical recovery Three serum samples containing 0, 4.9 and 9.9 /~g/ml of theophylline were supplemented with known amounts of theophylline, and then reassayed. Good agreement between expected and recovered theophylline concentrations were obtained (Table 3).
4.6. Sample dilution linearity Three serum samples containing 22.4, 26.5 and 28.1/~g/ml of theophyiline were serially diluted to 1:6, 1:8 and 1:8, respectively, with normal human serum, and the resulting samples were assayed for dilutional linearity. The observed theophylline concentrations were 100 4- 10% of the expected concentrations in the range of 3-30 t~g/ml, and were considered by us as acceptable.
4.7. Interference study Interference to triglycerides, hemoglobin and bilirubin were tested by assaying three serum pools which contained mean theophylline concentrations of 8.78, 19.3 and 33. I ~8/ml, and were supplemented with different levels of the interferant of interest. No significant interference was detected: triglycerides (500 mg/dl), hemoglobin (500 mg/dl) and bilirubin (30 mg/dl).
4.8. Stability of avidin.biotin.theophylline conjugate The avidin-biotin-theophylline conjugate was heat stressed at 45°C for up to 7 days. No significant change in both the immunoreactivity (94% retention) and dose-response curve was observed as compared to the conjugate kept at 4°C (Fig. 4). Thus, the conjugate is very stable to heat stressing.
S. Discussion We have described a biotin-avidin rate immunoturbidimetric procedure for the quantitation of theophylline on the Beckman Synchron CX ® 4 System and its performance characteristics. The biotin-avidin based assay is homogeneous and involves the mixing of reagent components and sample in a reaction buffer at 37°C. A fast immunoprecipitin reaction was obtained, which permitted the use of sl~ort and fixed read window for the measurement of the turbidity change at 340 nm. Measurement could begin immediately after the trigger reagent was introduced and completed in approx. 36 s for each sample covering the entire assay range. Unlike nephelometry rate turbidimetry allows the use of neat serum samples without interference from background light scattering. Because larger sample size can be
C Oh et al./Clin. Chim. Acta 218 (1993) 39-71
69
utilized, rate turbidimetry may enhance the detection of those analytes with low serum concentration. With the Beckman Synchron CX®, as much as 25 ~1 sample size can be used. We selected the rate measurement over the end-point method to eliminate the sample to sample variation in background turbidity. The biotin-avidin theophylline assay protocol described in this report employed the sequential saturation method as the sample was first allowed to incubate with the antibody until the change in background optical absorbance was stabilized and before the conjugate was introduced. The sequential saturation method not only increased the assay sensitivity as expected, but also ensured that only the specific rate was measured. Since the avidin..biotin-theophyilineconjugate exhibited substantially less turbidity than the antibody component and gave no non-specific rate in the reaction buffer, conjugate triggering was preferred. The biotin-theophyllinederivative (Fig. 2) selected for this study has a spacer arm of 21 atoms equivalent to approx. 27 .~ long between theophyllineand biotin moiety [16]. This spacing led to the formation of essentially linear repeating units o£ the form [16]: [(antibody)(theophyiline-biotin)(avidin)(biotin-theophylline)],. The linear polymer formation is ideal for inhibition assays, since in an inhibition assay, the binding of a single hapten molecule to an antibody binding site effectively terminates the polymer chain. This leads to an improved sensitivity and a greater economy of antibody than can be obtained with the conventional polyhaptenylated pseudo-antigens. The avidin..biotin-theophyllineconjugate can be readily prepared for large scale manufacturing. The conjugation procedure is easy to perform and reproducible. The procedure involves merely the mixing of avidin and the biotinylated drug derivative followed by the removal of the excess organic components by dialysis. Each lot of the avidin conjugates prepared contained stoichiometric quantity of hapten molecules per molecule of avidin, and a simple ultraviolet spectrum allowed the quantitation ¢f the conjugate. This feature obviates the cumbersome cross-titering of the conjugate with a given batch of antibody. The conjugate is very stable to heat stressing. After seven days at 45°C, the dose response curve remained basically unchanged and 94% of immunoreactivity was retained, As for the biotintheophylline derivative (Fig. 2), large quantity can be prepared from the readily available commercial materials. The biotin-avidin theophylline assay performed well in all the areas described in this report. Reproducible results were obtained over the assay range of 0-40/~g/ml. The within-run CV was < 3.7%, and the between-run CV values ranged from 5.2% (> 15 ~g/ml) to 9.2% (< 5/~g/ml). Correlation with a well established FPIA gave a correlation coefficient of 0.9866, a slope of 1.01 and an intercept of -0.4. Dilution of samples through the assay range gave the expected linear response to dilution. Interference study indicated negligible effects at concentration of up to 500 mg/dl for triglycerides, 500 mg/dl for hemoglobin and 30 rag/d! for bilirubin. Cross-reactivity to caffeine was insignificant. In summary, the biotin-avidin method represents the use of a new class of conjugate for immunoprecipitin reaction, and the resulting turbidity change can easily be measured spectrophotometrically. The method is simple, rapid, reliable, and applicable to other drug assays.
70
C Oh et al./Clin. Chim. Acta 218 fl993) 59-71
6. References i 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 '~.0 21 22 23 24
Uaury R, Shapiro DL, Smith B, Warshaw JB. Treatment ofsevere apnoea irt prematures with orally administered theophylline. Pediatrics 1975:55:595-598. Weinberger M, Riegelman S. Rational use of theophylline for bronchodilatation. N Engl J Med 1974;291:151-152. Hendeles L, Weinberger MM. Theophyiline therapeutic use and serum concentration monitoring. In: Taylor WJ, Finn AI, eds, Individualizing drug therapy: Practical applications of drug monitoring. New York: Gross Townsend Franck Inc. 1981;1:31-36. Aranda JV et al. Pharmacologic considerations in the therapy of neonatal apnea. Ther Pediatr Clin Am 1981;28:113-133. Shah VP, Riegelman S. GLC determination of theophylline in biological fluids, J Pharmacol Sci 1974;63:1283-1285. $chwemer HA. Ludden TM. Wallace JE, Determination of theophylline in plasma by electron capture gas chromatography, Anal Chem 1976:48:1875-1878. Thompson RD. Nagasawa HT, Jenne JW. Determination of theophylline and its metabolites in human urine and serum by high-pressure liquid chromatography, J Lab Clin Med 1974;84:584-593. Midha KK, Sved S, Hossie RD. McGiiveray IJ, High performance liquid chromatography and mass spectrometric identification of dimethylxanthine metabolites of caffeine in human plasma. Biomed Mass Spectrom 1977;4:i72-177. Pickard CE. Stewart AD, Hartley R, Lucock MD. A rapid HPLC method for monitoring plasma levels of caffeine and theophylline using solid phase extraction columns, Ann Clin Biochem 1986;23:440-446. NeeseAL. Soyka LF, Development of a radioimmunoassay for theophyllinc. Clin Phamacol Ther 1977;21:633-641, OellerichM. Sybrecht GW, Haeckel R. Monitoring of serum theophylline ~oncentrations by a fully mechanized enzyme immunoassay (EMIT). J Clin Chem Clin Biochem 1979:17:299,-302. Chang J, Gotcher S, Gushaw JB. Homogeneous enzyme immunoassay for theophylline in serum and plasma, Clin Chem 1982:28:361-367, Vasiliades J, Halsted T, giteley "L Cox RS, Determination of theophylline, phenytoin, phenobarbital, gentamicin, and tobramycin by fluorescence polarization immunoassay and enzyme immunoassay, J Clin Lab Autom 1984:4:315~320. Singh P, Hu MW, Gushaw JB, UIIman EF, Speoific antibodies to theophylline for use in a homogeneous enzyme immunoassay, J Immunoassay 1980;1:309-322, Yatscoff RW, Hayter J, Competitive nephelometric inhibition immunoassay of theophylline with use of monoclonal reagents: comparison with fluorescence polarization and EMIT procedures. Clin Chem 1983;29:1857-1858. Oh CS. Sternberg J. Nephelometri¢ inhibition immultoas~y for small molecules, In: Ngo TT, ed. Nonisotopi¢ immunoassay. New York: Plenum Press, 1988:457-476, Sternberg, J. Monitoring the precipitin reaction t~sing rate neph¢1ometry. Am Clin Prod Rev 1984:24-26. Deaton C. Perry C. Decker M. A nephelometric inhibition immunoassay system for therapeutic monitoring of anticonvulsants, aminoglyeosides and theopl,ylline. Clin Chem 1982:28:161I, Cambiaso CL. Riccomi HA, Masson PL, Heremans JF. Automated nephelometri¢ immunoassay !1. its application to the determination of hapten, J Immune Methods 1974:5:293-302, Lawrence JL, Elser RC. Comparison of particle.enhanced turbidimetric inhibition immunoassay and fluorescence polarization immunoassay using monoclo,:al antibodies for theophylline. Ther Drug Monitoring 1986;8:228-231. Jolley ME, Stoupe SD, Schwenzer KS, et al. Fluorescence polarization immunoassay, III. An automated system for therapeutic drug determination, Clin Chem 1981;27:i575-1579, Kraws LM, Sweet RV, Fluorescent immunoassay of theophylline, Clin Chem 1981;27:1085. Li TM, iknovi¢ JL, Buckler TT, Bard JF, Homogeneous substrate-labelled fluorescent immunoassay for theophylline in serum, Clin Chem 1981:27:22-26, Rupchock P, Sommer R, Greenquist A, Tyhack R, Walter B, Zipp A, A Dry-reagent strips used for detgrmination of theophyUine with a centrifugal analyser, Clin Chem 1985;31:737-740,
C Oh et al./Clin. Chim. Acta 218 (1993) 59-71
71
25 OoteUi G, Mitchell M. Particle-enhanced turbidimetric inhibition immunoassay for theophylline with a centrifugal analyser. Clin Chem 1985:31:1085-1086. 26 Hinds JA, Pincombe CF, Kanowski RK, Sanderson JC, Duffy P. Ligand displacement immunoassay: A novel enzyme immunoassay demonstrated for measuring theophylline in serum. Clin Chem 1984;30:1174-1 |78. 27 Poncelet SM, Limet JN, Noel JP, Kayaert MC, Galanti L, Collet-Cassart D. Immunoassay of theophylline by latex particle counting. J Immunoassay 1990;I 1:77-88. 28 Cook CE, Twine ME, Myers M, Amerson E, Kepler JA, Taylor GF. Theophylline radioimmuunassay: synthesis of antigen and characterization of antiserum. Res Commun Chem Pathol Pharmacol 1976;13:497-505. 29 Jasiewicz ML, Sc.hownberg DR, Mueller GC. Selective retrieval of biotin-labeled cells using immobilized avidin. Exp Cell Res 1976:100:213-217.
Clinica Chimica Acta 218 (1993) 59-71
A rate turbidimetric immunoassay for theophylline using biotin-avidin system Chan Oh*, Julie Kim, Betty Kearns, Anthony Cheng, Tom Dobashi Beckman Instruments. Brea. California. USA
(Received9 November 1992; ~evisionreceived22 March 1993; accepted 22 March 1993)
Abstract
We describe a biotin-avidin based rate turbidimetric homogeneous immunoassay for the determination of serum theophylline with a measuring range of 0-40 #g/ml. The assay features an avidin-biotin-labeled theophylline conjugate and a monoclonal antibody. The rate of change of turbidity caused by the antibody-conjugate complexing was monitored on the Beckman Synchron CX® 4 System at 340 nm and 37°C. Theophylline in a sample inhibited the complexation, and the extent of inhibition allowed the quantitation of theophylline in the sample. The within-run coefficient of variation (CV) was < 3,7% and the between-run CV was 5,2-9.2% depending on 1he theophylline concentration. Linear regression analysis showed a good correlation with an established fluorescence polarization immunoas~ay (r = 0.9866, n = 94). Key words: Theophylline; Turbidimetric immunoassay; Biotin.avidin
I. Introduction
Theophylline, 1,3.dimethylxanthine, is a potent smooth-muscle relaxant especially affecting bronchial smooth muscle. It is used extensively for the prevention and treatment of asthma, and for the treatment ofapnea and bradycardia in infants [!,2]. Monitoring of serum levels is important for the safe and effective use of theophylline because it has a narrow therapeutic range, and patients vary considerably in their * Corresponding author. 0009.8981/93/$06.00 © 1993 ElsevierScience Publishers B.V. All rights reserved. SSDI 0009.8981(93)05558-V
60
C Oh et al. / Clin. Chim. Acta 218 (1993) $9-71
ability to clear the drug. Serum concentrations of 5-15 ~g/ml theophylline reportedly control apneic spells in neonates without causing apparent side effects [3]. In most patients, theophylline serum concentrations of 10-20 ~g/ml effectively suppress chronic asthmatic and other bronchospastic symptoms [3]. Above 20 ~gtml, toxic effects such as nausea, headache, vomiting, diarrhea, irritability, and insomnia have been reported [4]. Severe toxicity is characterized by seizures and cardiac arrhythmias. The elimination kinetics of the drug are influenced by age, smoking, and liver function. Methods available for theophylline quantitation are gas chromatography [5,6], high performance liquid chromatographic techniques [7-9], and a variety of isotopic [10] and non-isotopic immunoassays. Non-isotopic immunoassays include the homogeneous enzyme immunoassay [ 11-15], nephelometric immunoassay [! 5-19], fluorescence polarization immunoassay (FPIA) [13,15,20-22], homogeneous substrate-labeled fluorescent immunoassay [23], apoenzyme reactivation immuno. assay system (ARIS) [24], particle enhanced turbidimetric inhibition assay [20,25], ligand displacement assay [26], and latex particle counting [27]. The biotin-avidin rate turbidimetric method described in this report represents a new homogeneous immunoturbidimetric method for the measurement of theophylline in serum. In addition to being specific, fast and reliable, the method allows the use of neat sample and small sample volume (3 ~!). Fig. ! illustrates the principle of the biotin-avidin assay.
2. Biotin-avidin based rate turbidimetric immunoassay The use of biotin-avidin for nephelometric inhibition immunoassay for small molecules was previously described by Oh and Sternberg [16]. The authors used biotinylated hapten and avidin in place of the conventional pseudo-antigen or developer antigen, such as polyhaptenylated apoferritin, Since each avidin molecule can bind up to four biotin molecules, the developer antigen thus takes the ~'orm, avidin-(biotin-hapten)4. When this tetrahaptenylated avidin conjugate (avidin conjugate) was mixed with an antibody specific to the hapten, immunoprecipitin reaction occurred at high rate. The fast immunoprecipitin reaction and the high scatter intensity from the immunocomplex formation appeared to be associated with the positive charge of the abundant lysine termini of avidin. This charge assisted immunoprecipitin reaction is specific and characterizes the unique feature of the biotin-avidin methodology. Under similar condition, streptavidin or charge neutralized avidin failed to produce the immunoprccipitin reaction with high rate. When the spacer length between biotin and avidin was about 27 A, steric hindrance between the Fab portion of the antibody and avidin permitted only two antibodies, one at each end of the avidin molecule, to bind to the avidin conjugate [16]. The binding led to the formation of the linear polymer, (antibody..haptenbiotin..avidin..biotin.hapten)n, as shown in Fig, !. In an assay medium containing polyethylene glycol, the linear polymer resulted in light scattering, and the rate of change of light scattering could be monitored on a nephelometer or turbidimeter. The hapten, when present in the sample, competed with the hapten-biotin of the avidin conjugate for the antibody and led to reduction of rate signal. Thus, the rate
C Oh et aL/Clin. Chim. Acta 218 (1993) 59-71
61
Without Theophylline in Sample: Formation of Complexes Resulting in Turbidity Chango
With Theophylline in Sample: Inhibition of Formation of Complexes Resulting in Oecrease in Turbidity Change
L •
Biotin
Drug
e.....~ Biotin-Drug
~
Avidin
Antibody Conjugate
Fig. I. Principle of biotin-avidin immunoturbidimetric assay.
of nephelometric or turbi~Jimetricresponse was inversely proportional to the concentration of the hapten in the specimen. The biotin-avidin assay presented in this report for the quantitation of theophylline is based on the turbidimetric rate measurement. Unlike nephelometry, turbidimetric measurement allows the use of neat sample without interference from excess light scattering. A theophylline monoclonal antibody solution (6-75 #1) was
62
C. Oh et al. / Ciin. Chim. Acta 218 (1993) 59-71
introduced into a reaction cell which contained the assay buffer and was maintained at 37°C. A steady optical absorbance level was observed until the neat sample, ranging from 3 to 25/zl, was added when a higher and steady optical absorbance baseline was reached. Subsequent addition of the avidin..biotin-theophyilineconjugate to the above mixture initiated the formation of the immunocomplex. The rate of turbidity change caused by the complex was then monitored spectrophotometricallyat 340 nm using a fixed read window of 8-64 s with reading to begin from 0-16 s after the trigger reagent was introduced. Theophyiline in the sample inhibited the complex formation, and a dose response curve was constructed for the quantitation of theophylline. 3. Materials and methods
3.1. Chemicals N-Hydroxysuccinimide (NHS), !,1 '-carbonyldiimidazole (CDI), 1,5-pentanediamine, 4-aminobutyric acid, dimethylaminocinnamaldehyde spray reagent were purchased from Aldrich (Milwaukee, USA). Ninhydrin was obtained from Calbiochem (San Diego, CA, USA). Biotin was available from Schweizerhall (Piscataway, N J, USA). Avidin was obtained from Boehringer (Mannheim, FRG). Theophylline for calibrator preparations was purchased from USP (Md, USA). 3.2. Buffers Buffer A. Phosphate-buffered saline (PBS) solution containing 0,01 mol/I sodium phosphate, 0.15 mol/I NaCI and 0.1% sodium azide, pH 7.0. Buffer B. Citrate.buffered saline (CBS) solution containing 0.025 mmol/I sodium citrate, 5 mmol/I disodium EDTA, 0.15 mmol/I NaCl and 0. I% sodium azide, pH 6.0. Buffer C, Phosphate buffer containing 0. I mmol/I sodium phosphate, 0.15 mol/I NaCI, 0.1% sodium azide and 2,6% polyethylene glycol, pH 7.0. 3.3. Serum specimens Serum specimens for correlation study were provided by Damon Laboratories (Rancho Cucamonga, CA, USA). 3.4. Theophylline calibrators Theophylline calibrators were prepared in a human serum pool, which was determined to be theophylline-free based on results from a commercially available fluorescence polarization kit. A stock solution containing 4 mg/ml of theophylline in 50 mmol/i NaOH was prepared. The stock solution was then diluted in the serum matrix to give theophyllinecalibrators consisting of six levels at 0, 2.5, 5, 10, 20 and 40 ~g/ml. 3.5. Antibody reagent Theophylline monoclonal antibody was produced by Beckman Instruments (Irvine, CA, USA). The antibody was diluted in phosphate buffered saline (buffer A) for assays.
63
C Oh et at./Clin. Chim. Acta 218 t1993) 59-71
O
H3C
O ,
N
0
H-N~N-H
0 H
H
O
CH5
Biotin-theophylline derivotive Fig, 2. Chemical structure of biotin-theophylline derivative
3.6. Biotin-theophyiline derivative The structure of the biotin-theophyilinederivative used in this study is shown in Fig. 2, and was synthesized by coupling theophylline-8-butyramidopentylamineto biotinamidobutyric acid as described below. Theophylline-8-butyramidopentylamine. Theophylline-8-butyric acid (14.1 ram), synthesized according to the method of Cook [28], was activated with CDI (16.5 ram) and NHS (16.7 ram) in 180 ml of dimethylformamide (DMF) at 70-75°C for 30 rain and at room temperature for 16 h. A solution of 1,5.diaminopentane (38.1 nun) in 100 ml DMF was introduced and the resulting mixture was stirred at room temperature for an additional 15 h. A white solid was formed and filtered. After the filtrate was evaporated, the resulting oil was chromatographed on a silica gel column. Elution with methanol/chloroform/ammonium hydroxide (25:71:4, by vol.) gave theophylline-g-butyramidopentylamine.The product was pure based on TLC using the same solvent system, and positive to ninhydrin spray for the presence of amino group (yield 19°/o, m.p. 125-126°C). Biotinamidobutyric acid. A solution of 4.aminobutyric acid (22 mm) in 75 ml of 0.1 molll NaHCO~ was added to 100 ml of DMF containing biotin-N-hydroxysuccinimide (20 ram), prepared as described by Jasiewicz [29]. After overnight stirring, the solid formed was filtered, washed with ethyl acetate, and dried to give 3.61 g of product (yield 50c%,m.p. 219-220°C). The product was pure as judged by TLC using dichloro.methane/methanol/acetic acid (67:32:1, by vol.) as the eluent, and positive to dimethylaminocinnamaldehydespray for the presence of biotin. Biotin-theophyiline derivative (Fig. 2). A solution of biotinamidobutyric acid (2.57 ram), CDi (2.83 ram) and NHS (2.~;', ram) in 50 ml of DMF was stirred at 70-75°C for 30 rain and at room temperature for 15 h. Theophyiline-8-butyramidopentylamine (2.57 ram) dissolved in 17 mi of DMF was then added, and stirring continued for an additional 15 h. The precipitate formed was filtered, washed with ethyl acetate, and recrystallized twice from hot methanol to afford 1.26 g of product (yield 74%). TLC revealed one spot using 25% methanol in chloroform as the cluent. The spot was positive to dimethylaminocinnamaldehyde spray indicating the presence of biotin. Avidin-biotin-theophylline conjugate. The avidin-biotin-theophyllineco:ljugate was prepared by mixing a solution of avidin (I I I rag) in 9 ml of 0. ! rnmol~!phosphate buffer (pH 7.4) and a solution of the biotin-theophylline derivative (5 rag) in 1 mi
64
C Oh el aL /Clin. C'him. Acto 218 (1993) 59-71
0"
/
.
30 °
0 O O/
i
o
O
y : 1.011X - 0 4 r = 0.~7
10
N~94
v
0
10
2O
30
FPIA (Thoo, pglrnl)
FiB, 3, Correlation of serum theophylline values obtained by biotin.uvidin immunoturbidimetric ussay and FPIA methods.
of methanol/dimethylsulfoxide (2: I, v/v), After I h of reaction followed by dialysis with the citrate buttered saline (CBS, buffer B), the conjugate concentrate was diluted with CBS for immunoassays, 3.7. FPIA reagents and assays A commercially available FPIA theophylline assay kit was employed for the serum sample correlation study, The reagents were handled and the assays run according to procedure supplied by the manufacturer, 3.8. Assay protocol on Beckman Synchron CX ® 4 System Reagents and buffers for the biotin-avidin based theophylline assay were introduced and mixed automatically by the Beckman Synchron CX ® 4 System employing
40
C. Oh el al. / Clin, Chim, Acta 218 (1993) 59-71
65
Ablorbance change/rain 0.3 --I-
0.2S
- - O - - - - - - OAY 0 2 --"¢~--
0.2
DAY 3 DAY 5
~ D A Y ?
0.15
0.1
0.06
/
0
I
I
5
10
15
20
25
30
35
40
Theophylllne (/~g/mi) Fig. 4. Stability of avidin.,biotin-theophylline conjugate at 4S°C.
the User Defined Reagents (UDR) mode. The assay protocol and addition sequence were as follows: (a) introduction of reaction buffer (buffer C, 230 ~l) and antibody (45 ~d), (b) addition of neat sample (3/d), i c) addition of avidin..biotin-theophylline conjugate (43/~l) 16 s after step (b) and (d) measurement of rate of change of turbidimetric response at 340 nm for 36 s at 37°C immediately after step (c).
C Oh et al./Clin. Chim. Acta 218 (1993) 59-71
66
Table ! Precisionevaluation Sample pool
Mean (t~g/ml)
S.D.
C.V.
(pg/ml)
(%)
4.7 15.5 34.9
0.173 0.521 1.012
3.7 3.4 2.9
A
4.6
B
15.4
C
35.2
0.422 0.801 2.091
9.2 5.2 5.9
Within-run precision ( n = 20)
A B C Between.run precision {n = 20)
4. Results 4. I. T h e o p h y l l i n e d o s e - r e s p o n s e c u r v e
A typical calibration curve for the biotin-avidin based theophylline assay on the Beckman Synchron CX ® 4 System is displayed in Fig. 4 as represented by the Day 0 curve. The assay covered a measuring range o f 0 - 4 0 ~g/ml, corresponding to a turbidimetric rate change of 0.286-0.069 absorbance units/rain at 340 nm, respectively.
4.2. Precision Within- and between-run precision of the assay were evaluated using three levels of commercial controls which contained mean theophylline concentrations of 4.6, Table 2 Cross.reactivity
Sample' containing theophylline and compound studied Compound
Theophylline observed (v ~g/ml)
% cross,reactivity (100 * 0 ' - 9.2)/x)
10 50 100 100 100 20 200 200
9.2 11.7 13.8 11.9 10.6 13.I 9.7 9.4
--
13
I 1,4
16.9
100
I !.0
1.8
Cone,
(x/~B/ml) b Caffeine Theobromin¢ 3-Methyixanthine 1,3-Dimethyluric acid I-Methyluric acid Uric acid
8-Chlorotheophylline Dyphyllinc
5.0 4.6 2.7 1.4 19.5 0.3 O.I
a Sampleswere prepared by mixingequal volumes of 20 Isg/mlof theophylline in a normal human serum pool and twice the listed concentration of the compound in the same matrix. b Sample containing theophylline only.
67
C Oh et al./Clin. Chim, Acta 218 (1993) 59-71
i 5.4 and 35.2 pg/ml of theophylline (Table !). The within-run precision (CV < 4%) was determined by running 20 replicates of each sample within the same run. Ten separate runs of the three controls gave the results for the between-run CV, with each control assayed in duplicate per run. The between-run CVs were 9.2%, 5.2% and 5.9%, respectively. 4.3. Specificity
The cross-reactivity of the theophylline metabolites and structurally related compounds was studied (Table 2). The cross-reactivity o f a compound was assessed by running as known in the theophylline assay a sample containing both the compound at or greater than its pharmacological concentration and theophyiline at 9.2 ;tg/ml. Cross-reactivity of a compound is defined as the percent increase of the theophylline concentration divided by the concentration of the compound in the sample containing 9.2 pg/ml o f theophylline. As can be seen from Table 2, 1,3dimethyluric acid and 8-chlorotheophylline exhibited some interference in the theophylline assay. The cross reaction observed with 1,3-dimethyluric acid poses no problem for serum theophylline measurements because methylated uric acids are not present in the serum of patients receiving theophylline therapy. Drugs such as dimenhydrinate, which contain 8-chlorotheophylline, could increase the apparent
concentration of theophylline. No significant cross-reactivity was observed for caffeine and 3-methylxanthine, the latter being one of the major metabolites of theophylline.
Table 3 Analytical recovery Serum sample
Theophylline supplemented
Theophylline target
Theophyiline observed
pool
(p8)
(/~g/mi)
(pg/ml)
A
0 4.9 9,7 19,4 29,I 38.8 0 4.9 9.7 29,I 38,8
C
0
% recovery
9,2 17,6 25,3 32.3
0 4,9 9,3 18,7 26.8 34.7
-103,6 100,6 106,0 105.9 107.3
-9.5 13.9 22.I 29.6
4.9 9. I 13.8 23.2 30.6
95.7 . 99.3 105,0 103.5
4,7
0
--
4.9
9.7 29. I
14,4 18.7 26.6
15.2 20.0 28.2
9.9
--
105,6 107.1 105.9
38,8
33.9
35.3
104.1
68
C Oh et aL ~C/in. Chim, Acta 218 (1993) 59-71
4.4. Correlation with FPIA A group of 94 serum samples covering a range of approx. 1-30/~g/ml was assayed by FPIA and by the avidin-biodn rate turbidimetric method on the Beckman Synchron CX ® 4 System. A good correlation between the two methods was observed (Fig. 3). The linear regression analysis resulted in the following correlation parameters: Synchron = 1.0106 * TDX - 0.38, r = 0.9866.
4.5. Analytical recovery Three serum samples containing 0, 4.9 and 9.9 /~g/ml of theophylline were supplemented with known amounts of theophylline, and then reassayed. Good agreement between expected and recovered theophylline concentrations were obtained (Table 3).
4.6. Sample dilution linearity Three serum samples containing 22.4, 26.5 and 28.1/~g/ml of theophyiline were serially diluted to 1:6, 1:8 and 1:8, respectively, with normal human serum, and the resulting samples were assayed for dilutional linearity. The observed theophylline concentrations were 100 4- 10% of the expected concentrations in the range of 3-30 t~g/ml, and were considered by us as acceptable.
4.7. Interference study Interference to triglycerides, hemoglobin and bilirubin were tested by assaying three serum pools which contained mean theophylline concentrations of 8.78, 19.3 and 33. I ~8/ml, and were supplemented with different levels of the interferant of interest. No significant interference was detected: triglycerides (500 mg/dl), hemoglobin (500 mg/dl) and bilirubin (30 mg/dl).
4.8. Stability of avidin.biotin.theophylline conjugate The avidin-biotin-theophylline conjugate was heat stressed at 45°C for up to 7 days. No significant change in both the immunoreactivity (94% retention) and dose-response curve was observed as compared to the conjugate kept at 4°C (Fig. 4). Thus, the conjugate is very stable to heat stressing.
S. Discussion We have described a biotin-avidin rate immunoturbidimetric procedure for the quantitation of theophylline on the Beckman Synchron CX ® 4 System and its performance characteristics. The biotin-avidin based assay is homogeneous and involves the mixing of reagent components and sample in a reaction buffer at 37°C. A fast immunoprecipitin reaction was obtained, which permitted the use of sl~ort and fixed read window for the measurement of the turbidity change at 340 nm. Measurement could begin immediately after the trigger reagent was introduced and completed in approx. 36 s for each sample covering the entire assay range. Unlike nephelometry rate turbidimetry allows the use of neat serum samples without interference from background light scattering. Because larger sample size can be
C Oh et al./Clin. Chim. Acta 218 (1993) 39-71
69
utilized, rate turbidimetry may enhance the detection of those analytes with low serum concentration. With the Beckman Synchron CX®, as much as 25 ~1 sample size can be used. We selected the rate measurement over the end-point method to eliminate the sample to sample variation in background turbidity. The biotin-avidin theophylline assay protocol described in this report employed the sequential saturation method as the sample was first allowed to incubate with the antibody until the change in background optical absorbance was stabilized and before the conjugate was introduced. The sequential saturation method not only increased the assay sensitivity as expected, but also ensured that only the specific rate was measured. Since the avidin..biotin-theophyilineconjugate exhibited substantially less turbidity than the antibody component and gave no non-specific rate in the reaction buffer, conjugate triggering was preferred. The biotin-theophyllinederivative (Fig. 2) selected for this study has a spacer arm of 21 atoms equivalent to approx. 27 .~ long between theophyllineand biotin moiety [16]. This spacing led to the formation of essentially linear repeating units o£ the form [16]: [(antibody)(theophyiline-biotin)(avidin)(biotin-theophylline)],. The linear polymer formation is ideal for inhibition assays, since in an inhibition assay, the binding of a single hapten molecule to an antibody binding site effectively terminates the polymer chain. This leads to an improved sensitivity and a greater economy of antibody than can be obtained with the conventional polyhaptenylated pseudo-antigens. The avidin..biotin-theophyllineconjugate can be readily prepared for large scale manufacturing. The conjugation procedure is easy to perform and reproducible. The procedure involves merely the mixing of avidin and the biotinylated drug derivative followed by the removal of the excess organic components by dialysis. Each lot of the avidin conjugates prepared contained stoichiometric quantity of hapten molecules per molecule of avidin, and a simple ultraviolet spectrum allowed the quantitation ¢f the conjugate. This feature obviates the cumbersome cross-titering of the conjugate with a given batch of antibody. The conjugate is very stable to heat stressing. After seven days at 45°C, the dose response curve remained basically unchanged and 94% of immunoreactivity was retained, As for the biotintheophylline derivative (Fig. 2), large quantity can be prepared from the readily available commercial materials. The biotin-avidin theophylline assay performed well in all the areas described in this report. Reproducible results were obtained over the assay range of 0-40/~g/ml. The within-run CV was < 3.7%, and the between-run CV values ranged from 5.2% (> 15 ~g/ml) to 9.2% (< 5/~g/ml). Correlation with a well established FPIA gave a correlation coefficient of 0.9866, a slope of 1.01 and an intercept of -0.4. Dilution of samples through the assay range gave the expected linear response to dilution. Interference study indicated negligible effects at concentration of up to 500 mg/dl for triglycerides, 500 mg/dl for hemoglobin and 30 rag/d! for bilirubin. Cross-reactivity to caffeine was insignificant. In summary, the biotin-avidin method represents the use of a new class of conjugate for immunoprecipitin reaction, and the resulting turbidity change can easily be measured spectrophotometrically. The method is simple, rapid, reliable, and applicable to other drug assays.
70
C Oh et al./Clin. Chim. Acta 218 fl993) 59-71
6. References i 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 '~.0 21 22 23 24
Uaury R, Shapiro DL, Smith B, Warshaw JB. Treatment ofsevere apnoea irt prematures with orally administered theophylline. Pediatrics 1975:55:595-598. Weinberger M, Riegelman S. Rational use of theophylline for bronchodilatation. N Engl J Med 1974;291:151-152. Hendeles L, Weinberger MM. Theophyiline therapeutic use and serum concentration monitoring. In: Taylor WJ, Finn AI, eds, Individualizing drug therapy: Practical applications of drug monitoring. New York: Gross Townsend Franck Inc. 1981;1:31-36. Aranda JV et al. Pharmacologic considerations in the therapy of neonatal apnea. Ther Pediatr Clin Am 1981;28:113-133. Shah VP, Riegelman S. GLC determination of theophylline in biological fluids, J Pharmacol Sci 1974;63:1283-1285. $chwemer HA. Ludden TM. Wallace JE, Determination of theophylline in plasma by electron capture gas chromatography, Anal Chem 1976:48:1875-1878. Thompson RD. Nagasawa HT, Jenne JW. Determination of theophylline and its metabolites in human urine and serum by high-pressure liquid chromatography, J Lab Clin Med 1974;84:584-593. Midha KK, Sved S, Hossie RD. McGiiveray IJ, High performance liquid chromatography and mass spectrometric identification of dimethylxanthine metabolites of caffeine in human plasma. Biomed Mass Spectrom 1977;4:i72-177. Pickard CE. Stewart AD, Hartley R, Lucock MD. A rapid HPLC method for monitoring plasma levels of caffeine and theophylline using solid phase extraction columns, Ann Clin Biochem 1986;23:440-446. NeeseAL. Soyka LF, Development of a radioimmunoassay for theophyllinc. Clin Phamacol Ther 1977;21:633-641, OellerichM. Sybrecht GW, Haeckel R. Monitoring of serum theophylline ~oncentrations by a fully mechanized enzyme immunoassay (EMIT). J Clin Chem Clin Biochem 1979:17:299,-302. Chang J, Gotcher S, Gushaw JB. Homogeneous enzyme immunoassay for theophylline in serum and plasma, Clin Chem 1982:28:361-367, Vasiliades J, Halsted T, giteley "L Cox RS, Determination of theophylline, phenytoin, phenobarbital, gentamicin, and tobramycin by fluorescence polarization immunoassay and enzyme immunoassay, J Clin Lab Autom 1984:4:315~320. Singh P, Hu MW, Gushaw JB, UIIman EF, Speoific antibodies to theophylline for use in a homogeneous enzyme immunoassay, J Immunoassay 1980;1:309-322, Yatscoff RW, Hayter J, Competitive nephelometric inhibition immunoassay of theophylline with use of monoclonal reagents: comparison with fluorescence polarization and EMIT procedures. Clin Chem 1983;29:1857-1858. Oh CS. Sternberg J. Nephelometri¢ inhibition immultoas~y for small molecules, In: Ngo TT, ed. Nonisotopi¢ immunoassay. New York: Plenum Press, 1988:457-476, Sternberg, J. Monitoring the precipitin reaction t~sing rate neph¢1ometry. Am Clin Prod Rev 1984:24-26. Deaton C. Perry C. Decker M. A nephelometric inhibition immunoassay system for therapeutic monitoring of anticonvulsants, aminoglyeosides and theopl,ylline. Clin Chem 1982:28:161I, Cambiaso CL. Riccomi HA, Masson PL, Heremans JF. Automated nephelometri¢ immunoassay !1. its application to the determination of hapten, J Immune Methods 1974:5:293-302, Lawrence JL, Elser RC. Comparison of particle.enhanced turbidimetric inhibition immunoassay and fluorescence polarization immunoassay using monoclo,:al antibodies for theophylline. Ther Drug Monitoring 1986;8:228-231. Jolley ME, Stoupe SD, Schwenzer KS, et al. Fluorescence polarization immunoassay, III. An automated system for therapeutic drug determination, Clin Chem 1981;27:i575-1579, Kraws LM, Sweet RV, Fluorescent immunoassay of theophylline, Clin Chem 1981;27:1085. Li TM, iknovi¢ JL, Buckler TT, Bard JF, Homogeneous substrate-labelled fluorescent immunoassay for theophylline in serum, Clin Chem 1981:27:22-26, Rupchock P, Sommer R, Greenquist A, Tyhack R, Walter B, Zipp A, A Dry-reagent strips used for detgrmination of theophyUine with a centrifugal analyser, Clin Chem 1985;31:737-740,
C Oh et al./Clin. Chim. Acta 218 (1993) 59-71
71
25 OoteUi G, Mitchell M. Particle-enhanced turbidimetric inhibition immunoassay for theophylline with a centrifugal analyser. Clin Chem 1985:31:1085-1086. 26 Hinds JA, Pincombe CF, Kanowski RK, Sanderson JC, Duffy P. Ligand displacement immunoassay: A novel enzyme immunoassay demonstrated for measuring theophylline in serum. Clin Chem 1984;30:1174-1 |78. 27 Poncelet SM, Limet JN, Noel JP, Kayaert MC, Galanti L, Collet-Cassart D. Immunoassay of theophylline by latex particle counting. J Immunoassay 1990;I 1:77-88. 28 Cook CE, Twine ME, Myers M, Amerson E, Kepler JA, Taylor GF. Theophylline radioimmuunassay: synthesis of antigen and characterization of antiserum. Res Commun Chem Pathol Pharmacol 1976;13:497-505. 29 Jasiewicz ML, Sc.hownberg DR, Mueller GC. Selective retrieval of biotin-labeled cells using immobilized avidin. Exp Cell Res 1976:100:213-217.