Dec 23, 1991 - 500 CLINICAL CHEMISTRY, Vol. 39, No. 3, 1993 ... An automated method based on the principles of flow- ..... conventional manual counterpart.
5. Holt DW, Tucker GT, Jackson PR, Storey GCA. Amiodarone pharmacokinetics. Am Heart J 1983;106:840-7. 6. Talajic M, DeRoode MB, Nattel S. Comparative electrophysiologic effects of intravenous amiodarone and desethylamiodarone in dogs; evidence for clinically relevant activity of the metabolite. Circulation 1987;75:265-71. 7. Mostow ND, Noon DL, Myers CM, Rakita L, Blumer JL. Determination of amiodarone and its N-deethylated metabolite in serumby high performance liquid chromatography. J Chromatogr 1983;227:229-37. 8. RessK, Liebich HM, Kramer B, Ickrath 0, Risler T, Seipel L. Determination of amiodarone and its metabolite N-desethylamiodarone in serum by high performance liquid chromatographycomparison of difference extraction procedures. J Chromatogr 1987;417:465-70. 9. Pollak PT, Carruthers SG, Freeman DJ. Simplified liquidchromatographic assay of amiodarone and desethylamiodarone after solid-phase extraction. Clin Chem 1986;32:890-3. 10. Verbesselt R, andreamaga TB, Schepper PJD. High-performance liquid chromatographic determination of 12 antiarrhyth-
mic drugs in plasma using solid-phase column extraction. Ther Drug Momt 1991;13:157.-65. 11. Bliss M, Mayersohn M, Nolan P. High-performance liquid chromatographic analysis of amiodarone and desethylaniiodarone in serum. J Chromatogr 1986;381:179-84. 12. Harkey MR. Bonded phase extraction in analytical toxicology. In: Deutsch DO, ed. Analytical aspects of drug testing. New York: John Wiley & Sons, 1989:59-85. 13. Glick MB, Ryder KW, Jackson SA. Graphical comparisons of interferences
in clinical
chemistry
instrumentation.
Clin
Chem
1986;32:470-5. 14. Lalloz M, Byfleld P, Himsworth R. Binding of serum amiodarone by serum proteins and the effects of drugs, hormones and other interacting ligands. J Phsrm Pharmacol 1984;36:366-72. 15. Weir SJ, Ueda CT. Rapid liquid chromatographic assay forthe determination of amiodarone and its N-deethyl metabolite in plasma, urine, and bile. J Pharm Sci 1985;74:460-5. 16. Siebers RWL, Chen CT, Ferguson RI, Maling TJB. Effect of blood sample tubes on amiodarone and desethylainiodarone concentrations. Ther Drug Monit 1988;10:349-51.
CLIN.CHEM.39/3, 500-503 (1993)
Automated Enzymatic Determination of Sodium in Serum R. Quiles,’
J. M. Fern#{225}ndez-Romero,2 E. Fern#{225}ndez,’ M. D. Luque de Castro,2
An automated method based on the principles of flowinjection analysis is proposed for the enzymatic determination of sodium ion in serum. The method relies on the activation of 3-galactosidase by the analyte. The features of the proposed method include linear range between 1 and 1700 mol/L, a sampling rate of 50 samples/h, a sample volume of 50 pL, and the absence of interferences from species usually present in serum. The results obtained were consistent with those provided by widely used methods such as those based on flame spectrometry and direct potentiometry with ion-selective electrodes.
and M. Vaic#{225}rcel2
on the concentration of sodium ion in the medium, according to the reaction
ONPG
Na
o-nitrophenol
+ galactose
$.galacto8idaae
The method proposed here is based on the above reaction and the flow-injection (Fl) technique (5). It requires small sample volumes and provides a high sampling rate and good precision with low reagent
consumption. We used it with three Fl manifolds: stopped-flow (6), open-closed (7, 8), and conventional. MaterIals and Methods
Indexing
Terms:
flow-Injection anaysis
ctosldase
Research into new analytical methods for determining ions in serum has focused chiefly on optical methods. Methods have been based on colorimetry (1), photometry with dry reagents (2), optical sensors (3), and enzymatic reactions (4). Enzymatic methods typically require use of an enzyme whose activity depends on the
ion being determined. The enzymatic photometric determination of sodium entails measuring the absorbance at 405 nm of the o-mtrophenol produced by hydrolysis of o-nitrophenylgalactoside (ONPG), which is catalyzed by 13-galactosidase (EC 3.2.1.23). 13-Galactosidaseactivity depends ‘Department of Biochemistry, Hospital Virgen de la Salud, 45005 Toledo, Spain. 2Department of Analytical Chemistry, Faculty of Sciences, University of C#{243}rdoba, 14004 C#{243}rdoba, Spain. Received December 23,1991; accepted August 24, 1992. 500 CLINICAL CHEMISTRY, Vol. 39, No. 3, 1993
Apparatus
We used Philips PU 8700 spectrophotometer (IJnicam Ltd., Cambridge CB1 2PX, UK) furnished with a Hellma 178.12QS (Hellma Hispania S.A., E 08011 Barcelona, Spain) flow cell (18 pL inner volume), a Julabo-5 recirculating thermostat, and a Minipuls-2 (Gilson Medical Electronics S.A., BP 45-95400 Villier Ii Bel, France) four-channel peristaltic pump with a rate selector, Rheodyne (Cotati, CA) 5041 injection valves of variable volume, and Teflon tubing [0.5 mm (i.d.)]. A standard personal computer with an in-house-built active interface was used to synchronize halting of the flow with injection in the stopped-flow mode. Whenever necessary, the pH was measured with a Radiometer Copenhagen 5 (Radiometer A/S, Copenhagen, Denmark) and a Beck-
‘Nonstandard abbreviations: ONPG, o-nitrophenyl--D-ga1actoside; EGTA, ethylene bis(oxyethylenenitrilo)tetraacetic acid; Fl, flow injection; and IMER, immobilized enzyme reactor.
man 072 (Beckman Instruments, 92634) pH meter.
Inc., Fullerton,
CA
A)
The concentrations of sodium ions in serum obtained by the Fl method were compared with those provided by an Instrumentation Laboratory (Lexington, MA) 243 flame photometer with an automatic dilution system and an Ektachem 700 XR automated analyzer (Kodak, Rochester, NY) based on direct potentiometry with an ion-selective electrode, without sample dilution.
il
Reagents /3-Galactosidase,
grade VIII from Escherichia coli (Sigma, St. Louis, MO; cat. no. G-5635); ONPG (Sigma; cat. no. N-1127); MgCl2 6H20 (Merck, Darmstadt, Germany; cat. no. 5833); Titriplex VI EEGTA, ethylene bis(oxyethylenenitrilo)tetraacetic acid] (Merck; cat. no. 8435); Tris (Merck; cat. no. 8382); NaC1 (Merck; cat. no. 6404); DL-dithiothreitol (Sigma; cat. no. D-0632); and MgSO4 7H20 (Merck; cat. no. 5886) were used. Reagent 1 was an aqueous solution containing 12.10 g of Tris, 125.50 g of EGTA, and 1.85 g of MgSO4 . 7H20 per liter with pH adjusted with 6 molfL HC1. Reagent 2 was a 4 mmol/L solution of DL-dithiothreitol in reagent 1. Reagent 3 was a 4 mmolJL solution of ONPG in reagent 1. All solutions were prepared in doubly distilled water of high purity obtained from a Millipore (Milford, MA) Milli-Ro system.
Samples samples from a routine clinical laboratory 1000-fold (100 L of sample diluted to 100 mL) with reagent 1 and injected into the Fl system in Serum
were diluted triplicate.
Standards and Controls
A stock aqueous solution containing 1 g/L of NaC1 in reagent 1 was prepared. From this, more dilute solutions were prepared as required. A Lab-trol Chemistry Control from Baxter Healthcare Co. was used to calibrate the flame spectrometer. Kodatrol I and II (Kodak) were used to calibrate the Ektachem automated analyzer. Immobilization
of /3-Galactosidase
The biocatalyst was immobilized on controlled-pore glass (CPG; 120-200 mesh; Electronucleonics, Fairfield, MA) by the procedure described by Masoom and Townshend (9). The immobilized enzyme reactors (IMERs), were constructed from different lengths of glass tubing (1.0 mm i.d.).
Results and DIscussIon Fl Manifolds Figure 1 depicts the three Fl configurations used. Manifold A (a conventional Fl system) includes an additional merging point after the IMER, where a basic solution is inserted to intensifr the color of the monitored product without causing deterioration of the enzymatic reactor. Halting the flow while the injected plug is in the IMER allows increased contact time between the
C)
F2
Fig. 1. Fl manifoldsused for the enzymaticphotometricdeterminationofsodiumion in serum (A) Conventional Fl manifold; (B) stopped-flow manifold; (C) open-closed manifold.S, sample;R, reagent; P. peristaltic pump; IV, Injection valve; IMER, immobilizedenzymereactor;T, thermostat; L, reactor;0, detector; W, waste; MC, microcomputer; SV, selecting valve; t, time
the substrates, and the analyte in manifold B. When the selecting valve in manifold C is switched, the sample plug is trapped in the closed circuit, where it is recirculated and passed through the IMER and the detector; thus each peak appears at a higher absorbance than the previous peak. The variables affecting the analytical signal obtained with each manifold were classified by the chemical, physical, and hydrodynamic characteristics of each configuration used (Table 1). The difference between the optimal pH for the development of the enzymatic reaction (pH 7.2) and for maximum intensity of the colored reaction product (pH 10-11) resulted in our having to biocatalyst,
Table 1. OptImum Value for Range of VarIables Studied Variable
Range
n-FIA
sf-FIA
oc-FIA
Physical Temp., #{176}C
25-45 #{176}C 37
37
37
Chemical
[ONPG], mmol/L [Mg(ll)], mmol/L [EGTA], mmoL’L pH of Al [NaOHI, mol/L FIA Flowrate, mLimin Injectedvolume,L IMER length,cm ReactorL1length,cm Delay time, S Stop time, s Switchingtime, s
0.1-8.0
4
2
2
0-20 0-1 5-9 0.5-6
7.5
7.5
7.5
0.4 7.2
0.4 8.7
0.4 8.7
0.5-1.1 50-500 0.5-2
50-200
4
0.87 50 0.5 100
35-50 15-120 30-105
0.64
50 1.0 50 42 60
0.87 100
1.0 150
60
n, conventional; sf, stopped-flow;oc, open-closed.
CLINICAL CHEMISTRY, Vol. 39, No. 3, 1993 501
Table 2. Percentage of ActivIty RemaIning Storage of IMER
after OI9S
20 h Solution 100 mmol/LTns HCI, pH 8.7 + 0.4 mmol/L EGTA + 7.5 mmol/LMg(II) 87 100 mmol/LTns HCI, pH 7.3 + 0.4 mmol/L 97 EGTA + 7.5 mmol/L Mg2 100 mmol/LTris HCI, pH 7.3 + 1.7 mmol/L 83 (NH4)2S04 + 10.0 mmol/LMg2 100 mmoVL phosphate buffer, pH 7.0 + glycerol (500 mI/L) 67
5 days
70 ga 72
The same activity percentage was found after storage for 1
4CC.
month at
pH (8.7) for the stopped-flow and methods because it was impossible to change the pH between the enzymatic reaction and detection (10). A confluence point of the main channel of the conventional Fl manifold with a 4 mol/L NaOH solution after the IMER allowed each step to be run at its optimum pH. A temperature of 37#{176}C for the IMER was selected for further experiments because the biocatalyst activity choose a compromise
open-closed
quate stability of the immobilized enzyme, and this activity was retained even after storage for 1 month at 4#{176}C (Table 2). A 5% loss of activity was observed after 40 serum samples were injected. This was overcome by adding 8 mmol/L DL-dithiothreitol (reagent 2) to the buffer solution, because thiol provides effective protection of the active sites of the biocatalyst. Calibration Curves and Sampling Frequency The features of the calibration curve for each configuration obtained by injecting standard solutions of 1-1700 mo1/L NaCl and the sampling frequency achieved with each manifold are listed in Table 3. The enhanced sensitivity of the conventional FT method compared with the other two methods is because in the former both the reaction and detection steps are developed at their optimum pHs.
Selection of Fl Manifold for Determining Sodium in Actual Samples Because the conventional
FT manifold
surpasses the
other two in simplicity, sampling frequency, regression coefficient, linear range, detection limit, and sensitivity, and because it requires smaller sample volumes and a was lost through denaturation at temperatures >40 #{176}C.shorter IMER, it was selected for determining sodium ion in actual samples. There was no appreciable loss of IMER activity over a working day with standard solutions; this was checked Analytical Performance of the Conventional Fl Manifold by injecting the same standard solution at the beginCalibration. A serum pool of samples from a clinical ning and end of the day. The IMER showed serious deterioration on storage, which necessitated the assay of laboratory was prepared. The samples were diluted as required to cover a sodium concentration range between different storage solutions. Solution 2 provided adeTable Msthod rt-FIA
Unsar rang., pmoVL 17-1700
sf-FIA
170-1700 170-1700 170-1700
oc-FIA a b
3.
Detection lImit, pmoIIL
12 12 100
90
Features of Three ConfiguratIons Equatlonb A = 6.86 ± 0.08.102 + 5.06 ± 0.05 iO [Nal A 0.11 ± 0.04 + 4.73 ± O.04104[Nai = 0.41 ± 0.03 + 1.33 ± 0.05. 104.[Na] #{163}4 = 0.62 ± 0.02 + 8.10 ± 0.04 iO[Na]
r
n
Sampling frsqu.ncy,h’
0.992 7 0.994 5 0.965 6 0.968 6
50 50 20 10
Abbreviations as in Table 1. A, absorbanceunIts;#{163}4, absorbanceincrement In a preset interval for the at methodand between consecutive peaksforthe oc method. (NaJ Numberof points In the linearregion.
=
moVL.
-J
-J
S.-
0
0
E
E E
E V
0
E
E -
V
-
o
o
E
(I)
E V 0
Cl)
LL
-
100
110
ItO
130
140
ISO
leO
LSodiumL ISE method
170
ISO
100
110
120
130
mmol/L
160
leO
[Sodiumi,
170
ISO
mmol/L
FP method
FIg. 2. Correlationof the proposed Fl methodwithconventionalmethods basedon direct potentiometry (ISE) (A)
502 CLINICAL CHEMISTRY, Vol. 39, No. 3. 1993
140
andflame spectrometry (B)
65 and 250 jzmol/L. The equation of the transient signal (Fl recording) obtained by injecting the sample into the manifold in triplicate was A=-0.107 + 2.50 iO [Na] (r = 0.9982), where A is in milliabsorbance units and the concentration of sodium in the serum samples is in moI/L. Withinand between-run reproducibility. A serum pool from the same laboratory was used for these studies after the concentration of sodium ion (137.7 mmol/L) was determined by direct potentiometry (Ektachem analyzer). The within-run CV was 3.3% (mean Na 136.3 mmol/L, SD 4.6, n = 25) and the between-run CV was 1.42% (mean Na 136.8 mmol/L, SD 1.86, n = 10). Each sample was injected in triplicate, and a separate stan-
dardization was performed Comparison
for each run.
of the proposed
method
with conventional
The proposed method (Fl) was compared with flame spectrometry (FP) and direct potentiometry (ISE). The results are shown in the two-dimensional diagrams in Figure 2. The correlation equations, coefficients, and deviations (a) of the slopes, intercepts, and estimates are as follows: FT = 1.O6ISE - 8 (r = 0.9716, n = 30, a8 = 0.05, a1 = 6, a, = 1.18) and FT = 1.O9FP - 12 (r = 0.9394, n = 30, a8 = 0.07, a1 = 10, a = 1.71). Thus an acceptable correlation exists between the proposed method and the comparison methods. Study of interfere nts. The potential disturbances caused by uni- and divalent cations usually present in serum were investigated by keeping constant the concentration of sodium ion in the samples (170 mol/L). methods.
No interferences were detected at concentrations 250 anol/L for K, and Ca2 and 20 mmol/L for Li. The catalytic effect of Mg was avoided by saturating the medium with this cation, because its influence was constant for concentrations 7.5 mmol/L in reagent 1. We concluded that the proposed method is selective for determining sodium ion in serum samples. The proposed Fl method is a valid alternative to
determining sodium ion concentration by flame spectrometry and with ion-selective electrodes, with which FT shows an excellent correlation. Fl surpasses its conventional manual counterpart (4) because the low concentration range of analyte in the proposed method makes it unnecessary to decrease its concentration as free species because of a possible catalytic effect. Thus, reagent consumption is decreased dramatically with the FT technique. Also, less biocatalyst is required because it is immobilized on a suitable support. The high dilution of the sample (1:1000) avoids interferences even though it involves additional sample manipulation. We thank Direcci#{244}n Generalde Investigaci#{243}n Cientifica y T#{233}cnica, for financial support (grant no. PB9O-0925).
References 1. Kumar A, Chapoteau
E, Czech BP, Gebaver CR, Chimenti MZ, Rainiondo D. Chromogenic ionophore-base for spectrophotometric assay of sodium and potassium in serum plasma. Clin Chem 1988;34:1709-12. 2. Walter B. Dry reagent chemistries. Anal Chem 1983;55:498A514A. 3. Seiler K, Wang K, Barrer E, Morf WE, Rustermols B, Spichiger VE, Simon V. Characterization of sodium selective optrode membranes based on neutral ionophores and assay of sodium in plasma. Clin Chem 1991;36:1350-.5. 4. Berry MN, Mazzachi ED, Pejakovic M, Peare J. Enzymatic determination of sodium in serum. Clin Chem 1988;34:2295-308. 5. Valc#{225}rcel M, Luque de Castro MD. Flow injection analysis: principles and applications. Chichester, UK: Ellis Horwood, 1987. 6. Luque de Castro MD, Valcdrcel M. Kinetic-based determinations in continuous flow analysis. J Autom Chem 1986;8:186-92. 7. Rios A, Luque de Castro MD, Valcdrcel M. Multidetection in unsegmented flow systems with a single detector. Anal Chem 1985;57:1803-9. 8. RIos A, Luque de Castro MD, Valc#{225}rcel M. Simultaneous determination by iterative spectrophotometric detection in a closed flow system. Anal Chim Acta 1986;179:463-8. 9. Masoom M, Townshend A. Determinition of glucose in blood by flow injection analysis and an immobilized glucose oxidase column. Anal Chim Acta 1984;166:111-8. 10. Boyer PD. The enzymes, 3rd ed. Vol. VII. New York: Academic Press, 1972.
CLIN.CHEM.39/3, 503-508 (1993)
Rate Nephelometric Assay of Serum Lipoprotein(a) Philippe
Gillery,’
Patricia
Arthuis,’
Chantal
Cuperlier,’
This nephelometric assay of serum lipoprotein(a) [Lp(a)] is characterized by the use of a specific antibody to generate a high rate of light-scatter formation and the
and Raymond Circaud2
elimination of nonspecific reactions from serum samples
by diluting samples in phosphate-buffered saline containing polymer enhancer polyethylene glycol (PEG), 40 g/L, and detergent before the assay. We reacted 100 p.L of sixfold-diluted serum in 500 L of buffer containing PEG
‘Laboratory of Biochemistry, Centre Hospitalier Universitaire de Reims, H#{244}pital Robert Debr#{233}, Rue Alexis Carrel, F51092 Reims, France. 2Beckman Instruments France, Chemin des Bourdons, Gagny, France. Received January 6, 1992; accepted October 7, 1992.
with 42 L of pure polyclonalrabbit antiserum (Dakopatts) directed against human Lp(a) and monitored the reaction by rate nephelometry with the Array Protein System nephelometer (Beckman). The standard curve for the reaction was linear in the Lp(a) range 10-1280 mg/L; antigen excess occurred between 1300 and 1400 mg/I. Calibration was performed with serial dilutions of a stanCLINICAL CHEMISTRY, Vol. 39, No. 3, 1993
503
