- Epi2occurs whenever a relatively slow chemical equilibrium containing the electroactive substance as the major component of the equilibrium precedes a charge transfer (12). At the concentrations of I- and I2 used here, the electroactive 13- is presumably about 90% of the total analytical concentration of iodide if the formation constant for 13- is taken as lo6 (3). Summary of Mechanisms. Cyclic voltammograms of the oxidation of I- in acetonitrile agree with theory of Reaction 2 for the anodic scan. Voltammograms of 13- oxidation agree with theory for the reaction i P / d without a change in Ep
13- e 13.
+ e-
e
2 1 ~ . 312
(5)
for anodic scans. However, the cathodic currents of voltammograms for these two ions are at substantial variance with theory. The reverse current of 13- oxidation is affected by product adsorption and by a slow, irreversible loss of products through reaction. Both the adsorption process and the irreversible reaction cause a cathodic shift in the reverse scan of an oxidation (10, 11) and could account for the deviation from theory seen here. The irreversibility of the I- oxidation cannot be easily con(12) M. S. Shuman, Ph.D. Thesis, Univ. of Wisconsin, Madison, Wis., 1966.
strued. The peak currents of the first and second oxidation waves of a tetrabutylammonium iodide solution (Figure 6) are in the ratio of 2 :1 which agrees with the theoretical current ratio 2.1 :1 for the mechanisms proposed for these two waves (7, 9). Thus, all the product of the first wave is oxidized at the second wave, i.e. 13- is stable during the time the experiment is performed. As is evident in Figure 8, the cathodic scan of the experimental curve does not return to zero current as would be expected for a totally irreversible wave, but reaches a constant, positive value apparently independent of potential. This behavior could be caused by regeneration of reductant through a chemical reaction (11). However, no reaction of I- with other iodine species can be envisioned that would not also affect the forward scan. It is possible that a specific interaction between reductant and adsorbed species inhibits the electrode reaction (13). RECEIVED for review August 11, 1969. Accepted January 30, 1970. This work was supported by funds from the Texas Christian University Research Foundation under Grant No. 6894 and from the Robert A. Welch Foundation under Grant NO.P-270. (13) C. N. Reilley and W. Strumm, in “Progress in Polarography,” P. Zuman and I. M. Kolthoff, Eds., Interscience, New York, N. Y., 1962, Vol I, p 81.
Differential Amperometric Measurement of Monamine Oxidase Activity at Tubular Carbon Electrodes William D. Mason1 and Carter L. Olson The College of Pharmacy, The Ohio State University, Columbus, Ohio 43210
Methodology for the measurement of monamineoxidase activity in flowing streams based on differential amperometric measurement at tubular carbon electrodes is described. The method yields highly reproducible results with good sensitivity and is very insensitive to normal interferences found in biological fluids. The enzyme activity can be measured directly in diluted biological fluids with no prior separation steps required.
All of these measurements except 0 2 uptake, which has a sensitivity disadvantage, involve initial separation procedures in that they are not run on crude biological extracts. Blaedel and Olson reported a continuous amperometric method for the measurement of the glucose oxidase catalyzed system (6). This method has been modified and applied to the monamine oxidase system, and is the subject of this paper. The monamine oxidase reaction is given by the Equation 1.
WORKERSIN SEVERAL health related areas are interested in measuring the activity of the enzyme monamine oxidase in biological fluids. Several methods for analysis have been reported. Included among these are spectrophotometric measurement of substrate disappearance ( 1 , 2 ) , fluorimetric measurement of substrate disappearance (3),measurement of 0 2 uptake (2, d), and radiotracer measurements (5).
R-CHz-NH2
Present address, School of Pharmacy, University of Georgia, Athens, Ga. (1) H. Weissbach, T. E. Smith, J. W. Daly, B. Witkop, and S. Undenfriend, J. Biol. Chem., 235, 1160 (1960). (2) C. W. Tabor, H. Tabor, and S. M. Rosenthal, ibid., 208, 645 (1954). (3) M.Kraml, Biochem. Pharmacol. 14, 1683 (1965). (4) S. Sho et al., Showa Igakukai Lasshi, 21, 932 (1967); Chem. Abstr. 69, 24686d (1967). (5) R. J. Wurtman and J. Axelrod, Biochem. Pharmacol., 12, 1439 (1963). 488
ANALYTICAL CHEMISTRY, VOL. 42, NO. 4,APRIL 1970
+ H2O +
(MAO)
0 2
RCHO
+ NH3 + HzO2
(Slow) (1)
The reaction is followed electrochemically by coupling the
H202enzymatically to another redox couple which can be conveniently measured by means of the peroxidase reaction as illustrated in Equation 2. KaFe(CW6
+ H202+ 2H+
(Peroxidase)
K3Fe(CN)6
+ 2H20 4- 2K+
(fast) (2)
The K3Fe(CN)6formation which is proportional to the rate of the M A 0 catalyzed reaction since Reaction 2 is much faster than Reaction 1 , is measured amperometrically by reduction to K4Fe(CN)6 at a tubular carbon electrode (TCE) through (6) W. J. Blaedel and C. Olson, ANAL.CHEM., 36, 343 (1964).
PUMP
BRIDGE
SALT
PUMP
VA m
Figure 1. Flow stream arrangement TOP
VIEW
10.5"
4-
T
Figure 3. Electrode assembly A-TCE B-SCE
Figure 2. Mixing chamber
electrical contact electrical contact 1.5
which the reaction mixture flows. The differential amperometric measurement is remarkably free from interferences and enzyme activity measurements can be made directly on diluted crude biological fluids with no other separation steps required.
v
EXPERIMENTAL Flow System. A variety of flow stream arrangements were evaluated and the one diagramed in Figure 1 was settled on as being most satisfactory since it permitted convenient thermostating and provided contact of the reaction mixture with only glass and Teflon (Dupont) prior to measurement at the electrodes. The latter fact is important since some M A 0 inhibitors absorb on tygon tubing affecting subsequent experimental results. Solutions are pumped by means of a Harvard Apparatus Model 500-1200 variable speed peristaltic pump. Tygon tubing (formula B44-3, 1/16-inchi.d.) was used in the pump. The flow lines were made from 16gauge Teflon tubing. Smaller tubing was tried, but since the pumps were placed in a downstream pulling position, it was relatively easy for a pressure drop to occur causing bubbles to form in the flow lines with the smaller diameter tubing. Typical tube lengths were 10 feet for the long delay line and 1 foot for the short delay line. The mixing chamber fabricated from a piece of 2-mm. i.d. glass tubing is illustrated in Figure 2. A micro Teflon-coated stirring bar was placed in the chamber and caused to vibrate rapidly with a waterdriven magnetic stirrer. Downstream from the mixing chamber a 2-inch piece of kinked 26-gauge platinum wire was inserted to further ensure mixing. Typical solution flow rates through the electrodes were between 2 and 3 ml per minute. The pump is adjusted so that essentially equal flow rates are attained through each electrode. The delay lines, mixing chamber, and magnetic stirrer are immersed in a water bath in order to thermostat the reaction. Electrodes. Two tubular carbon electrodes (TCE) (0.25 inch long by 0.081 inch i.d.) in Teflon holders constructed as described by Mason and Olson (7) were mounted vertically and connected through a "T" shaped salt bridge to a large saturated calomel reference electrode as shown in Figure 3. These electrodes were found to be very stable and gave reproducible results for several weeks to months if rinsed daily after use with 0.1NHC1 for a few minutes followed by a water wash. (7) W. Mason and Carter L. Olson, ANAL.CHEM., 42,548 (1970).
L
S
-
I
J
RECORDER
Figure 4. Measurement circuit R1-200 Q potentiometer Rz,R8-10,000 Q 10-turn potentiometers TCE-Tubular carbon electrode SCE-Saturated calomel electrode R,-l0,000,
3332, 1000,332, 100 Q resistor (5~0.1%) for current
measurement S-Position for total current measurement &-Position for bridge potential measurement Differential Measurement System. A bridge circuit, which is basically the same as previously reported (6), is used to measure the difference in currents passing through the two TCE's when a fixed voltage is applied to them. The bridge circuit is shown in Figure 4. The bridge voltage is measured directly with a strip chart recorder (Varian Model G-40) which has a potentiometric input impedance at the low millivolt ranges. The recorder was damped by a parallel capacitance of 500 to 1000 pF. Reagents. Monamine oxidase enzyme was obtained from beef plasma. Initial enzyme samples were prepared by the Method of Tabor, Tabor, and Rosenthal (2). It was then found that the plasma could be diluted and studied directly without purification. The standard procedure is to collect the blood at the slaughter house and immediately add ' 1 9 volume of a citrate solution containing 8 grams of citric acid ANALYTICAL CHEMISTRY, VOL. 42, NO. 4, APRIL 1970
489
I (CATHODIC) .?
- )iAMPERES
0.1
// IO0 MV
'(-I
E (VS) SCE
1
P OJt
J.
Figure 5. Residual current Reagent stream-0.2mM K4Fe (CN)B, 3 mg/liter peroxidase and 4mM benzylamine in pH = 7.2 phosphate buffer. Sample stream-Buffer only VI = 2.0ml/min; T = 37 "C
and 26.7 grams of sodium citrate.5.5 H 2 0 per liter. The plasma is collected by centrifugation at 2500 rpm for 20 minutes. The plasma is studied immediately or frozen in 3- to 10-ml glass vials for future study. The frozen M A 0 was found to retain its activity for periods up to four months. Horseradish peroxidase was purchased from the Worthington Biochemical Corp. All other chemicals were reagent grade and used without purification. The enzyme samples are prepared by diluting the plasma in 0.1M phosphate buffer (pH 7.2) which contained 0.1M KC1. The reagent solution contained K4Fe(CN)6, peroxidase and the substrate benzylamine. This solution is deaerated and stored under N2 to prevent oxidation of the K4Fe(CN)6. Procedure. The peristaltic pump is initially adjusted so that the flow rate OT solution through each delay line is as equal as possible. The bridge is balanced and calibrated by pumping a standard K3Fe(CN)6 solution through the reagent stream and diluted plasma in buffer through the sample stream. The standard K3Fe(CN)6 solution normally is prepared by adding K3Fe(CN)6 to the regular reagent solution containing buffer, peroxidase, and K4Fe(CN)6 but less the substrate benzylamine so that the enzyme reaction does not proceed. The usual K3Fe(CN), concentration is 5 X lO-6M. The bridge is balanced by adjusting the values of resistors Rz and R3, while the calibrating KsFe(CN)6 is passing through both electrodes, so that the differential voltage across the bridge becomes zero. The bridge sensitivity can be controlled by the size of R2 and RB. Calibration of current us. concentration is done by alternating the solution pumped through the reagent stream between the solutions containing the desired K3Fe(CN)6concentrations and a blank containing everything but the K3Fe(CN)6. When the calibrating solution reaches the short delay line electrode the current at the electrode will be measured. When the solution reaches the second electrode the current will cancel and read zero across the bridge. When the blank is placed in the reagent stream it will again reach the short delay line electrode first, and during the time interval it takes to reach the second electrode the differential response will be that of the calibrating solution passing through the long delay line electrode. Response linearity is normally checked with calibrating K3Fe(CN)6 concentrations between 1 X and 10 X 10-6M. Since the calibrating solution is mixing with the sample solution, the actual concentrations passing through the electrodes is one half the concentration introduced through the reagent stream. The time difference for solutions to reach the two electrodes is measured and the recorder response is calculated in terms of concentration change per unit time, usually in units of micromolar per 490
ANALYTICAL CHEMISTRY, VOL. 42, NO. 4, APRIL 1970
2 3 4 5 6 PEROXIDASE MGM 1 L. Figure 6. Peroxidase concn in reagent solution Reagent contained: 5 mM benzylamine, 0.2 mM K4Fe (Cn)6 in buffer O.1MKCI Sample: 1 plasma in buffer (pH = 7.2) T = 37 "C, V f = 2.0 ml min ( A t = 2 min, 10 sec)
minute. The recorder response is therefore a direct readout of the reaction rate. RESULTS AND DISCUSSION
The optimum voltage to be applied to the electrodes was determined by measuring a residual current voltage curve. Figure 5 illustrates the total residual current for the two TCE M K$e(CN)s, 3 when a solution containing 0.2 X mg/liter peroxidase and 4 X lO-3M benzylamine in pH 7.2 phosphate buffer at 37 O C was pumped through the reagent stream and buffer was pumped through the sample stream. A voltage of -0.050 V us. SCE which is on the diffusion limited current plateau for the reduction of K4Fe(CN)6 was applied to the electrodes for 'the remainder of the studies reported. Biological samples lowered the residual current slightly; 1 beef plasma lowered the total current 0.03 FA, but the differential base line did not change. Instrument stability is illustrated by the fact that during studies of up to 5 hours of continuous running, the differential base line drift was less than 1 % full scale. When the flow system and electrodes were rinsed daily with 0.1NHC1 after use to prevent the growth of microorganisms in the delay lines, measurements could be made for periods of several weeks with the response of the electrodes remaining essentially constant. In any case, because of the direct readout capabilities of the instrumentation, it is very convenient to calibrate the electrodes with standard solutions. As a precaution against breakage during a prolonged experiment, the three pieces of tygon tubing in the pump head were replaced daily. The volume flow rate, typically 2 to 3 ml/min, measured frequently during each day, remained constant, varying less than 0.5% during the day. This flow rate constancy indicates the accuracy of the time measurements for the reaction rate calculations. Studies to determine optimum concentrations for the reagent solutions were conducted. For all the studies, 1 % beef plasma in buffer was pumped in the sample stream. Dependence of response on peroxidase concentration is shown in Figure 6. In this study, 5 m M benzylamine, 0.2 m M K4Fe(CN)6, and varying peroxidase concentration were pumped through the reagent stream. Figures 7 and 8 are the results of similar studies showing the dependence of response on benzylamine and ferrocyanide concentration, respectively. The reagent stream contained 3 mg/liter peroxidase and 0.2 m M ferricyanide for the benzylamine study
c
, 1 2 K,FE(CN)~
3 x 10,
4 5 MOLAR
DIFFERENTIAL CURRENT Figure 9. Differential current (typical chart) Each inch = 4 min 5 inches = 1 pM/min or 0.5 mV See text for description
Figure 7. KPe(CN)O concn in reagent solution Reagent: 2 mg/liter peroxidase, 5mM benzylamine in pH = 7.2 phosphate buffer and O.1MKCI Sample: 1 plasma in buffer T = 37 "C,V f = 2.0ml/m
Table 11. M A 0 Activity as a Function of Plasma Concentration
t
2.0
zPlasma
Bridge voltage, mV
0.20
0.10
Y
I
I
2 4 6 8 10 12 14 BENZYLAMINE x io3 MOLAR Figure 8. Benzylamine in reagent solution
16
Reagent: 0.2mM K4Fe(CN)6,3 mg/liter peroxidase Sample: 1% plasma, V , = 2.0 ml/min, T = 37 "C Plasma: After 1/6 diluted with citrate solution
Rz =
"C f 0.2 "C
2 z Plasma
0.58
vo
mV pM/min mV pM/min 0.36 0.68 0.34 0.18 25.0 0.45 0.86 0.22 0.42 27.5 0.64 1.26 0.33 0.64 31.5 0.87 1.72 0.44 0.87 35.0 2.00 1.02 1 .oo 0.51 37.0 R1 = R a = 9.0 KO Rec. 1 mV full scale Reagent = 2 X lO-4M K4Fe(CN)6; 4 mg/liter Peroxidase, 4mM Benzylamine pH = 7.20. V , = 2.0 ml/min 4t = 2 min 10 sec vo = d[KaFe(cN)e.l dt
~
Plasma 1.00 0.98 0.97 1.04 1.02 1.02 1.02 1 .00 1.00 0.98 0.83
Reagent: 0.2mM KaFe(CN)0; 4mM benzylamine; 3 mg/liter peroxidase 4t = 2 min, 10 sec V , = 2 ml/min
Table I. Temperature Effects
vo
Vo,pM/min
vo = dlKaFe(CN)d dt
~~
1% Plasma
voic,
p M/min
0.20 0.196 0.39 0.58 0.60 0.29 0.83 0.80 0.416 1.00 0.51 1.02 1.22 1.20 0.61 1.02 1.0 0.51 2.00 2.0 1.00 3.0 1.50 3.00 3.90 4.0 1.95 5.00 6.0 2.50 5.80 10.0 2.90 R 3 = 9.0 K Rec. = 5 mV f.s., T = 37 "C 0.40
~~~
~-
~
~~
and 3 mg/liter peroxidase and 5 m M benzylamine for the ferrocyanide study. All studies were done at 37 "C. Dependence of the reaction rate of temperature is shown in Table I. Solutions containing 1 and 2z beef plasma in buffer were pumped through the sample stream and the reagent stream contained 0.2 m M ferrocyanide, 4 m M benzylamine, and 4 mg/liter peroxidase in buffer. It is clear that for reproducible results the solutions and delay lines must be thermostated. During the temperature study, it was important that the electrode response be calibrated with standard ferncyanide at each temperature.
z
Table I1 contains typical data illustrating the linearity and sensitivity of the instrumentation for measuring monamine oxidase activity. The reagent stream contained 0.2mM K4Fe(CN)6, 4mM benzylamine, and 4 mg/liter peroxidase (in pH 7.2 phosphate buffer) at 37 "C. The sample stream contained beef plasma diluted in the same phosphate buffer. The percentage listed in the Table is a volume per centLe., a 1 plasma sample is prepared by diluting 1 ml of plasma to 100 ml with buffer. Typically, the average deviation for a series of measurements was less than 2x, although often even better results were obtained as illustrated in Table 11. Each number reported in the table is the average for two or three runs whose average deviation is less than 1 Under the experimental conditions listed, linearity of response was good up to about 4 plasma concentration. An example of the experimental data is shown in Figure 9. The pen response at A is the base line, the negative current at B is the current present at the short delay line electrode. C is the differential current between both electrodes where the distance between A and C divided by the delay line time difference gives the reaction rate. D shows blank, clearing the short delay line electrode where only current at the long delay line electrode is measured, and then the current
z.
ANALYTICAL CHEMISTRY, VOL. 42, NO. 4, APRIL 1970
491
finally returns to the base line when blank is passing through both electrodes. If each sample is not separated by a blank, samples could be run easily at a rate of 15 to 20 samples per hour. Assuming a flow rate of about 2 ml/min and 5 minutes per sample using a 1 plasma sample, a measurement could be conveniently made with 0.1 ml of plasma. In conclusion, the method developed for the assay of M A 0 activity was found to have good sensitivity but, more important, to be very free from interferences because the measurement can be made fortuitously in a voltage pocket that exists between the reduction of oxygen and the oxidation of biological amines, where the background noise is small and constant, and because the differential measurement effectively cancels out faradic interferences which do not change during the time interval of passage between the two electrodes, Initial studies were begun using tubular platinum electrodes, but it was the authors' experience that superior results were obtained at the tubular carbon electrodes.
Although the data are the subject of subsequent publications, kinetic constants for the enzyme reaction of the crude preparations are in excellent agreement with data reported on crystallized enzyme. Also, measurements can be made on other crude biological preparations-for example, measurements have been made on diluted rat brain and rat liver homogenates. In addition, extensive studies of M A 0 inhibitors in the crude preparations are being made. In all cases, excellent results, typical of those presented in the present paper, indicate the general utility of the method.
RECEIVED for review September 30, 1969. Accepted January 29, 1970. This study was supported in part by Grant GM15821, from the National Institutes of Health, U.S. Public Health Service, Bethesda, Md. William D. Mason was a fellow of the American Foundation for Pharmaceutical Education.
Ferrocene as a Primary Standard for Oxidation-Red uction Titrations in Acetonitrile Byron Kratochvil and Peter F. Quirk Department of Chemistry, University of Alberta, Edmonton, Alberta, Canada
Ferrocene is a suitable primary standard for solutions of hydrated copper(l1) perchlorate in acetonitrile. Commercial ferrocene can be purified readily by recrystallization and sublimation, and is stable and nonhygroscopic on storage. Concentrations of copper(ll) solutions, determined by weight titrations of ferrocene and aqueous EDTA, agree within 0.2 ppt.
IT HAS RECENTLY been shown that copper(I1) perchlorate can be used as an analytical oxidant in acetonitrile (I, 2). The reduction potential in acetonitrile of the copper(I1)-(I) couple vs. a silver-0.01M silver nitrate reference is 0.801 V (3), 1.27 V greater than that of the copper(I)-(O) couple. Although copper(I1) solutions in acetonitrile are prepared easily and are stable ( I ) , a method for determining their redox titer to a part per thousand or better has not been established. A preliminary investigation showed that copper(I1) perchlorate was too hygroscopic to be used directly. Thiourea, potassium iodide, hydroquinone, and quinhydrone gave drawn out potential breaks and concentration-dependent equivalence points. Other standards used in water-such as arsenic trioxide, sodium oxalate, potassium ferrocyanide, and iron(I1) ethylenediammonium sulfate-are not sufficiently soluble in acetonitrile. Oxalic acid dihydrate, though soluble, is not oxidized by copper(I1). Ferrocene is oxidized readily to the ferricenium ion by copper(I1) in acetonitrile. Its availability and ease of oxidation prompted a study as a primary standard for acetonitrile solutions of hydrated copper(I1) perchlorate. (1) B. Kratochvil, D. A. Zatko, and R. Markuszewski, ANAL. CHEM., 38, 770 (1966).
(2) H. C . Mruthyunjaya and A. R. V. Murthy, Indian J. Chem., 7, 403 (1969). (3) E. Lorah and B. Kratochvil, University of Alberta, unpublished
work, 1969. 492
0
ANALYTICAL CHEMISTRY, VOL. 42, NO. 4, APRIL 1970
EXPERIMENTAL
Reagents. Technical grade acetonitrile (Matheson, Coleman and Bell) was purified according to the method of O'Donnell, Ayers, and Mann (4), except that the decantation following addition of sulfuric acid was replaced by a vacuum distillation below 55 "C. The pure solvent had an absorbance of less than 0.2 at 200 mp. Technical grade heptane (Fisher Scientific Co.) was used as received. Deionized water was used throughout. Crystallization and sublimation were used to purify commercial ferrocene (Arapahoe Chemicals). Recrystallization was done from heptane; the product was dried under vacuum for at least 12 hr. Commercial acetonitrile was not satisfactory for recrystallization. Sublimations were performed under vacuum at 60 to 80 "C; the product was collected on a water-cooled cold finger. The melting point of ferrocene recrystallized twice, then sublimed, was obtained from DTA curves recorded on a DuPont Model 900 differential thermal analyzer; a range of 175.4 to 175.9 "C was found. The solubility of ferrocene in acetonitrile was determined by equilibration under nitrogen with pure solvent at 25 f 0.05 "C; equilibrium was reached within 4 hr. The solutions were analyzed by careful evaporation of the solvent and weighing, and by titration of weighed aliquots with copper(I1) perchlorate in acetonitrile. Anal. by evaporation (av. of 4), 3.41 weight %; by titration (av. of 4), 3.37 weight %. Hydrated copper(I1) perchlorate was prepared by adding a slight excess of 72% perchloric acid (J. T. Baker Chemical Co.) to a suspension of copper(I1) carbonate (Fisher Certified Reagent) in water. The solution was boiled to eliminate carbon dioxide, then cooled, and the resulting crystals were collected and dried at room temperature under vacuum. The copper was determined by EDTA titration with murexide
(4) J. F. O'Donnell, J. T. Ayres, and C . K. Mann, ANAL.CHBM., 37, 1161 (1965).
