1808
Anal. Chem. 1988, 58, 1898-1900
are metabolites of ABPC in human urine (20). Recently, it has been reported that the piperazine-2,5-dioine (3) is excreted in rat (21) and human (22) urine after dose of ABPC is given. Figure 4 shows the structures of these metabolites, and Figure 5 shows a chromatogram of ABPC, 1,2, and 3 detected by the proposed method. The limits of detection of 1,2, and 3 are as low as 0.5 pg/mL with a 50-pL injection. These results reveal that the method is able to detect not only a penicillin but also a metabolite having a thiazolidine ring.
1
ACKNOWLEDGMENT We are grateful to T. Uno and H. Yasuda, Mukogawa Women's University, for their interest and support. W S t W NO. 1,32746-94-4;2,71774-55-5; 3,94595-14-9; PCG, 61-33-6; ABPC, 69-53-4; ACPC, 3485-14-1; CBPC, 4697-36-3; TIPC, 34787-01-4; penicillin, 1406-05-9; sodium hypochlorite, 7681-52-9.
2
LITERATURE CITED
4. Stnrctue of (5RW)genidlkic acid (l), the (5S,6R)-epimer (2) and piperazlne-2,5dlone (3). 1
4
1
h3 b
(1) Lam, S.; aushka, E. J . L i s . Chromatcar. 1978, 1 , 33-41. (2) Rogers, M. E.; Adlard, M. W.; Saunderc G.; Hoit, G. J . Llq. Chromat00r. 1983. 6 . 2019-2031. (3) MGazakl, K.; Ohtanl, K.; Sunada, K.; Arita, T. J . Chromatogr. 1983, 276, 478-482. (4) Rogers, M. E.; Adlard, M. W.; Saunders, G.; Holt, G. J . Chromatogr. 1984. 297, 385-391. (5) Haginaka, J.; Wakai, J. Amlyst(London) 1985, 110, 1185-1188. (6) Haglnaka. J.; Wakal, J. Analyst (London) 1985, 110, 1277-1281. (7) Lw, T. L.; D'arconte, L; Brooks, M. A. J . Pharm. Sci. 1979, 6 8 , 454-458. (8) Westerlund, 0.; Cerlqvlst, J.; Theodorsen, A. Acta Pharm. Sue. 1979, 16, 187-214. (9) Cadqvist, J.; Westerlund. D. J . Chromatogr. 1979, 164, 373-381. (IO) Rogers, M. E.; Adlard, M. W.; Saunders, G.; Holt, G. J . Chromatogr. 1088, 257, 91-100. (11) Buchberger, W.; Wlnsauer, K.; Nachtmann. F. Fresenius' 2. Anal. Chem. 1988, 315, 525-527. (12) Haaalnaka, J.: Wakai, J. Anal. Chem. 1985. 5 7 . 1568-1571. (13) Kok, W. Th.; Halvax, J. J.; Voogt, W. H.; Brinkman, U. A. Th.; Frei, R. W. Anal. Chem. 1085.. 5 7 ,~ . 2580-2583. (14) Carlqvist, J.; Westerlund, D. J . Chromatogr. 1985, 344, 285-296. (15) Haglnaka, J.; Wakal, J.. submitted for publication in J . Pharm. Phar~
macol .
0
10
20
30
Tlmo (mln) Flgure 5. Separation of ABPC (4), 1 (3), 2 (2), and 3 (1). Each concentratkn was 25 pg/rnL. Injection volume was 20 &. Detection was petfumed at 270 nm and at a SenSitMty of 0.016 AUFS. Fleadon coil length was 1 m. For other conditions, see the Experimental Section.
cisions (relative standard deviation) were 1.0-4.5% and 0.9-3.3%. Detection of ABPC and Its Metabolites. It was known that (5R,GR)-penicilloic acid (1)and the (5S,6R)-epimer (2)
(16) Uno, T.; Masada. M.; Yamaoka, K.; Nakagawa, T. Chem. Pharm. BuU. 1981, 2 9 , 1957-1968. (17) Bundgaard, H.; Larsen, C. I n t . J . Pharm. 1979, 3 , 1-11. (18) Bohbn, P.; Meibt, M. Anal. Blochem. 1970, 9 4 , 313-321. (19) Ishida, Y.; Fujita, T.; Asai, K. J . Chromatogr. 1981, 204, 143-148. (20) Bird, A. E.;Cutmore, E. A.; Jennlngs, K . R.; Marshall, A. C. J . Pharm. PhErrnz3COl. 1983, 35, 138-143. (21) Everett, J. R.; Jennlngs, K. R.; Woodnutt, J.; Buckingham, M. J. J . Chem. Soc., Chem. Commun. 1984, 894-895. (22) Haginaka. J.; Wakai, J. J . Pharm. Pharmacol. 1988, 38, 225-226.
RECEIVED for review January 2, 1986.
Resubmitted March
14, 1986. Accepted April 1, 1986.
Spectruphatometric System tor Klnetic Absorbance Measurements in Two-Phase Enzyme Immunoassays Vadiraja V. Murthy,* Lawrence Freundlich, and Arthur Karmen Department of Laboratory Medicine, Albert Einstein College of Medicine of Yeshiva University, 1300 Morris Park Avenue, The Bronx, New York 10461 Many microimmuncmsay methods currently in favor in the clinical laboratory involve measurement of the activity of an *Address all correspondence to this author at Hospital of the Albert Einstein College of Medicine, Room 3-33, 1600 Tenbroeck Ave., The Bronx, NY 10461. 0003-2700/86/035&1898$01.50/0
enzyme that is immobilized to the surface of a plastic or glass bead by an antibody-antigen reaction (1,2). These methods are preferable to radioimmunoassays for their relative ease of operation, freedom from exposure to radioactivity, and cost-effectiveness. In a typical enzyme immunoassay, a "sandwich" consisting 0 1986 American Chemical Society
ANALYTICAL CHEMISTRY, VOL. 58, NO. 8, JULY 1986 1899
Table I. Stirring in the Presence of a Plastic Bead Has No Effect on the Absorbance of Potassium Dichromate Solutions" sample blank dichromate
mean absorbance f SD (coeff variation) with stirring no stirring 0.003 f 0.001 0.065 f 0.001 (0.8%)
0.003 f 0.001 0.066 f 0.001 (0.7%)
12.5 /AM 0.163 f 0.001 (0.5%) 15.0p M 0.186 f 0.001 (0.3%) 25.0 pM 0.356 f 0.001 (0.2%) 60.0/AM 0.786 i 0.001 (0.1%)
0.163 f 0.001 (0.4%) 0.185 f 0.001 (0.3%) 0.356 f 0.001 (0.2%) 0.784 f 0.001 (0.1%)
5.0 p M
'Shown are means of 10 successive readings recorded at 15-sintervals.
Figure 1. (a) Schematic view of the modified dual cell holder. Position of the bead and the spinning vane lnslde the cuvette (dotted line) Is indicated. (b) Isometric views of the cell compartment.
of the bead/analyte/antibody-reagent enzyme is formed. The reagent enzyme activity still attached to the bead, after the unreacted antibody is washed away, can be related to the quantity of analyte present. The reagent enzyme activity is measured by incubating the bead with a suitable substrate solution usually for a period of time and taking an end point absorbance reading of the chromophore in a spectrophotometer. A more rapid "kinetic" enzyme method, desirable for "stat" assays, would be interfered with by the presence of the bead in the light path of the spectrophotometer. In the work described here we designed and tested a cuvette holding system for a conventional ultraviolet/visible spectrophotometer that permits continuous monitoring of the activity of an enzyme bound to a plastic bead. The floor of the cuvette housing of the temperaturecontrolled, dual-cell holder (part 585753) for the Beckman Model 35 spectrophotometer (Beckman Instruments, Fullerton, CA) was milled away so that the bottom portion of the cuvette containing the bead was below the light path of the photometer (Figure 1). The cuvette contents were stirred continuously by a Teflon-coated magnetic micro stirring bar, 7 mm long and 2 mm diameter (catalog no. 14511-67,Fisher Scientific Co., Springfield, NJ),kept in motion by a cobalt disk magnet, 1.9 cm 0.d. and 6 mm thick (catalog no. D30.962), mounted eccentrically on the shaft of a 1.5-V,25-mA, 'mini-motor" (catalog no. D40.872) both from Edmond Scientific Co., Barrington, NJ. To facilitate positioning and easy removal of the cuvette, the top portion of the rear wall of the cuvette compartment was also milled down 6 mm. The heating element, fixed to the bottom of the compartment, was relocated slightly to permit the cuvette to pass through the housing. The external disk magnet was fastened to the motor shaft by gluing it to a disk cam 1.9 cm in diameter, and 6 mm thick, machined from a 1.9-cm-diameter mild steel rod, which was then pressed on the shaft. The stirring motor was clamped to the base of the housing with the help of an aluminum plate, 9 mm thick, in which a semicircle that fit the motor casing closely was machined, and which was bolted to the base by two no. 8 bolts, 3.8 cm long. A 1.5-V battery external to the spectrophotometer housing provided the power for the motor. The assay mixture consisted of 2.5 mL of 4 mM p-nitrophenyl phosphate in 0.14 M phosphate buffer, pH 9.1. The reaction was initiated by carefully inserting the stirring bar and the bead containing the enzyme with a pair of forceps. An equal volume of the substrate solution was placed in the reference cuvette. A continuous record of the absorbance at 405 nm of a series of dilute solutions of potassium dichromate instead of the usual assay mixture employed in the creatine kinase-MB isoenzyme (CK-MB) assay showed less than 3 milliabsorbance units of noise whether the solutions were stirred or not. There were no differences in absorbances due to stirring alone as
24
E
t
Ln
0
18
t c, [I
12
ui m < . 06 2
4 6 T I M E
10
Figure 2. Alkaline phosphatase activity on the bead as a function of CKMB concentration and time of incubation. The enzyme activity was measured by following the formation of p-nitrophenol at 405 nm: (m) 14.3 ng/mL CK-ME; (0)28.5 ng/mL, (A)57 ng/mL.
shown in Table I. To test the proportionality of alkaline phosphatase activity to the quantity of CK-MB bound to the bead, graded quantities of CK-MB standard solutions (14.2,28.5, and 57 ng/mL) were incubated with beads and the enzyme-labeled second antibody according to the Hybritech protocol; the beads were washed and assayed for alkaline phosphatase. The rates oE production of p-nitrophenol were constant during the first 10 min of incubation and were proportional to the concentration of CK-MB added (Figure 2). Stirring was necessary to ensure uniform distribution of the p-nitrophenol. Without stirring the apparent rate of production was significantly lower. If stirring was started midway in the reaction the absorbance could be observed to have reached the same level as that of the stirred solution, indicating that the diffusion of the product away from the bead, not the access of the substrate to the enzyme on the bead, was the rate-limiting step.
DISCUSSION Constant agitation of the bead and solution ensures uniformity of concentration without detectably increasing the electrical noise of the measurement. Measurement of the bound enzyme, kinetically, with the system described offers two significant advantages over the end point method usually employed. The enzyme assay can be performed more rapidly, particularly when only a few samples are to be analyzed. In addition, measurement of the rate of reaction, when the rate is constant, affords greater certainty that what is being measured is enzyme activity rather than a trace of colored material in the sample adsorbed to the bead. This increased sensitivity is particularly helpful in distinguishing small quantities of analyte from background or confirming the
Anal. Chem. 1986, 58, 1900-1901
WOO
Dresence of small auantitiea. in such assavs as those for serum gonadotropin or bK-MB in which th;! presence of small quantities is clinically significant (3, 4). Registry No. Creatine kinase, 9001-15-4.
(2) Procedure swDlied with Tandem4 CKMB Immunoenzvmetric assay kit, Hybriiech,. inc., Sen Diego, CA 92121. (3) Bock, J.; Furgiuele, J.; Wenz, B. Clin. Chim. Acta. 1985 747, 241. (4) Sheehan, M.; Haythorn, P. Clin. Chem. (Whston-Salem, N . C . ) 1985, 37, 160.
LITERATURE C I T E D (1) Myrtle, J.; Shimizu, I.; Varga, M.; Kotler, H.; Bartholomew, R. Clin. Chem. (Winston-Salem, N . C . ) 1983, 29, 1232.
RECEIV~, for review January 9,1986. Accepted March 3,1986.
Interface for Computer-Based Alternating Current Voltage Control Dean A. Bass, David K. Eaton,’ and James A. Holcombe*
Department of Chemistry, The University of Texas at Austin, Austin, Texas 78712 The control of ac power is important in a number of applications, and the computer represenb a convenient means of monitoring and controlling delivery. As one example, furnaces used for electrothermal atomization (ETA) in atomic absorption analysis require frequent feedback and power adjustments if linear heating rates of 1000-2000 K/s are to be realized. The circuit described in this paper is a versatile computer interface for triac control of ac voltage. Temperature control in ETA is often accomplished by use of direct feedback from an optical transducer (1-6) or other temperature-sensitive devices (7). The simplest case involves heating with constant applied ac voltage to the desired temperature, followed by off-on control to maintain the final temperature (1-3). This is one means of achieving rapid heating with a limited maximum temperature. Feedback throughout the heating cycle can produce any desired heating profile and can provide good control over the temperature region in which the transducer is sensitive. The primary device for ac voltage control is a triac or a pair of antiparallel silicon controlled rectifiers. The triac acta as a bidirectional switch, which is turned on by a pulse a t the gate terminal. As long as the gate is not activated, the device goes out of conduction when the main current goes to zero (i.e., zero crossing). Ac control of 60-Hz line frequency is obtained by sending a 120-Hz pulse train to the triac. By phase shifting the pulse train relative to the zero crossing, the output power can be varied. The circuit described in this paper can provide a specific amount of power during each half-cycle for precise voltage control. It is controlled with an Apple II+, although any computer with an 8-bit parallel output port may be used. Some advantages can be gained if a 6522 versatile interface adapter (VIA) is available (8). EXPERIMENTAL S E C T I O N Circuit Description. The circuit shown in Figure 1is divided into three parta: the zero crossing circuit (I),the counter circuit (11), and the triac circuit (III). Section I produces a 120-Hz pulse train synchronizedto the zero crossing. The counter circuit phase shifts this pulse train using digital counter circuitry, and the triac circuit actually provides the power to the load. A full-wave rectified ac signal from transformer T1 (Figure 2a) modulates switching transistor Q-1 at 120 He (Figure 2b). This zero crossing pulse train is used in conjunction with the 50-kQ variable resistor and IC-1 to phase shift the falling edge of the output (Figure 2c) to coincide with the zero crossing point of the voltage driving the load. The falling edge of the zero crossing pulse triggers the monostable multivibrator (IC-2a) to produce a 1-ps pulse shown in Figure 2d. The rising edge of this pulse strobes the data latch, IC-3, which loads an 8-bit number from the computer’s parallel Present address: Eaton Instruments, Inc., Horseshoebay, TX.
Table I. Electrical Components for Figure 1 abbreviation
component
F1
fuse (1/2 A)
F2 T1 T,
Q-1,2 IC-1
fuse (30 A) transformer 2:l step-up pulse transformer (no. 1122003, Sprague) ECG 128 SN74121
abbreviation IC-2a,b IC-3 IC-4,5 IC-6 Triac
component SN74123 14LS374 SN74193 NE555 ECG 5697 (500 V, 40 A) (may vary depending on power requirements)
output port. The falling edge of the same pulse allows the number on IC-3 to be loaded onto IC-4 and IC-5, which comprise the 8-bit counter. IC-4 immediatelybegins incrementing at the clock (IC-6) frequency. When all 8 bits in IC-4 and IC-5 are one (Le., the next clock pulse sets the carry and sends a pulse to IC-2b. This multivibrator produces a 20-ps pulse (Figure 2e), which is displaced from zero crossing by the time required to increment the counter from the input value to 25510. The output of IC-2b is then used to trigger the triac. Switching transistor 6 - 2 is used for current amplification of the signal sent to the triac gate. The pulse transformer, T,, serves as an isolation device for the rest of the circuit. When the triac receives the pulse at its gate it goes into conduction for the remainder of the half-cycle. To obtain the heating rates used in ETA, it is necessary to use a high-wattage, step-down transformer (e.g., 9 V/250 A), which is not shown in Figure 1, but would be connected at the ac out in section 111, Figure 1. It should be noted that problems can arise in this type of application when driving an inductive load with a triac near full power (9). Power Resolution. This circuit has 256 different power settings. The incremental resolution on the power delivered to the load is limited by the size of the counter and can be set by adjusting the clock frequency (IC-6). To maximize resolution, the clock frequency should be set to count at 31 kHz (i.e., 256 clock cycles per 1/120 s for a 60-Hz line frequency). At peak voltage, where the power resolution is most limited, the change in power is about 0.8% from one count to the next. For increased power resolution a larger counter could be used. For example, if a 12-bit counter is used with a 492-kHz clock, 4096 power settings would be obtained with a resolution of better than 0.1 % at peak voltage. Circuit Adjustment for Zero and Full Power. To adjust zero and full power to correspond to counts of zero and 255, respectively, the triac output is monitored by using an oscilloscope at ac out (Figure 1). The zero crossing pulse can be adjusted by using the 50-kQvariable resistor on IC-1 so that 255 on the counter
0003-2700/86/0358-1900$01.50/00 1986 American Chemical Society