Fluorocarbon-based immobilization method for preparation of enzyme

John G. Stevens , Lawrence H. Bowen , and Katherine M. Whatley. Analytical Chemistry ... David J. Stewart , Duncan R. Purvis , Christopher R. Lowe. Jo...
1 downloads 0 Views 406KB Size
1996

Anal. Chem. 1988, 60,1996-1998

(3) Gaspar, G.; Annino, R.; Vidai-Madjar, C.; Guiochon. G.; Anal. Chem. 1978, 50, 1512-1518. (4) Annino, R.; Leone, J. J. Chromtogr. Sci. 1982, 20, 19-26. (5) Trehy, M. L.; Yost, R . A.; Dorsey, J. G. Anal. Chem. 1988, 58. 14-19. (6) Ewels, B. A.; Sacks, R . D. Anal. Chem. 1985, 57, 2774-2779. (7) Myers, M. N.; Giddings, J. C. Anal. Chem. 1985, 3 7 , 1453-1457. (8) van Es, A.; Janssen, J.; Baiiy, R.; Cramers, C.; Rijks, J. HRC CC,J . High Resolut. Chromatogr . Chromatogr . Commun . 1987, 10 273-279. (9) Tijssen, R.;van den Hoed, N.; van Kreveld, M. E. Anal. Chem. 1987. 59, 1007-1015. (10) Doue, F.; Guiochon, G. Sep. Sci. 1970, 5 , 197-218. (11) Wade, R. L.; Cram, S. P. Anal. Chem. 1972, 4 4 , 131. (12) Hopkins, B. J.; Pretorius, V. J. J. Chromatogr. 1978, 158,465-489. (13) Berg, S.;Jonsson, A. HRC CC,J. High Resolut. Chromatogr. Chromatogr. Commun. 1984, 7 , 887-695. (14) Matuska, P.; Kovai, M.; Seller, W. HRC CC,J. High Resolut. Chromatogr. Chromatogr. Commun. 1988, 9 , 577-583.

(15) Hagman, A.; Jacobsson, S. HRC CC,J . High Resolut. Chromatogr. Chromatogr. Commun. 1985, 8 , 332-336. (16) Craydon, J. W.; Grob, K. J. Chromatogr. 1983, 254,265-269. (17) Arnts. R. R. J. Chromatogr. 1985, 399-405. (18) Yasuoko, T.; Higuchi. H.; Kidokoro, T.; Mitsuzawa. S.;Zimmerman, P. R. Bunseki Kagaku 1984, 33, 528-532. (19) Cox, R. D.; Earp, R. F. Anal. Chem. 1982, 5 4 , 2265. (20) Baiischmiter, K.; Mayer, P.; Class, Th. Fresenius' Z . Anal. Chem. 1986, 323, 334-339 (21) McCienney, W. A.; Pleil, J. D.; Holdren, M. W.; Smith, R. N. Anal. Chem. 1984, 56, 2947-2951.

RECEIVED for review January 12, 1988. Accepted April 28, 1988. We acknowledge the support of the Centers for Disease Control (CDC-NIOSH), Grant 1-R01-OH02303, and of the Dow Chemical Co.

Fluorocarbon-Based I mmobiliration Method for Preparation of Enzyme Electrodes R. K. Kobos,* J. W. Eveleigh, M. L. Stepler, B. J. Haley, a n d S. L. Papa E . I . d u Pont de Nemours and Company, Inc., Medical Products Department, Glasgow Site, Building 100, Wilmington, Delaware 19898 Enzyme electrodes are an important type of electrochemical biosensor that have application in clinical diagnostics, biomedical research, process monitoring, and artificial organs. An enzyme electrode consists of a thin layer of enzyme immobilized on the surface of an electrochemical sensor (1-4). Many methods have been used to immobilize the enzyme, including entrapment using a dialysis membrane or within a polymer gel, adsorption onto the electrode surface or a support membrane, and covalent attachment to the electrode surface or to a support membrane (2-4). Enzyme electrodes have been constructed with a variety of electrochemical sensors. Of these, gas-sensing electrodes for ammonia, carbon dioxide, and oxygen are preferred because of their high selectivity. These gas-sensing electrodes utilize a hydrophobic membrane, typically a fluorocarbon membrane, to separate the internal solution, in which the electrochemical measurement is made, from the sample solution. Because only dissolved gases diffuse through the membrane into the internal solution and are detected, these sensors are not affected by ionic species. However, the preparation of enzyme electrodes using gas sensors is complicated by the difficulty in attaching the enzyme to the fluorocarbon membrane. In most cases, the enzyme is immobilized onto a support membrane, e.g., nylon net ( 5 ) ,pig or cellulose acetate (3,which is physically held intestine (6), over the gas-permeable membrane of the sensor. Alternatively, covalently cross-linked enzyme membranes, prepared with glutaraldehyde and bovine serum albumin, are held on the gas sensor by means of a dialysis membrane (8, 9). The enzyme has also been entrapped in a polyacrylamide or gelatin matrix on the electrode surface (10). These methods are complicated and use an additional membrane, which adversely affects the response time of the enzyme electrode. Another approach is to directly attach the enzyme to the fluorocarbon membrane by adsorption of the native enzyme ( l l ) ,or by chemically binding the enzyme, with glutaraldehyde, to a fluorocarbon membrane that has been etched with a sodium dispersion in naphthalene (12). It has also been reported that enzymes can be directly polymerized onto certain fluorocarbon membranes with glutaraldehyde without etching the membrane (13). Moreover, enzyme electrode membranes have been prepared by first treating a fluorocarbon membrane with a perfluoroalkyl surface active agent

to make it hydrophilic to a prescribed depth, exposing the membrane to the enzyme solution, and cross-linking the enzyme within the hydrophilic region with glutaraldehyde (14). The resulting immobilized enzyme membrane contains a hydrophobic region and, therefore, functions as a gaspermeable membrane. In this paper, we report a novel method for enzyme immobilization, in which the enzyme is chemically modified by perfluoroalkylation of available amino groups, as shown in Figure 1. This modification greatly enhances adsorption onto fluorocarbon surfaces. This method was used to immobilize the enzyme urease (EC 3.5.1.5) onto the gas-permeable membrane of an ammonia sensor to prepare a urea electrode.

EXPERIMENTAL SECTION Apparatus. An Orion Model 95-10 ammonia gas-sensing electrodewas used to construct the urea electrode. Potentiometric measurements were made with a Corning Model 130 research pH meter in conjunction with a Hewlett-Packard Model 7132A strip chart recorder. All measurements were made at room temperature. Reagents. Urease, Type VII, 573 000 International Units (IU)/g, was obtained from Sigma Chemical Co., St. Louis, MO. Urea, electrophoresis purity reagent, was obtained from Bio-Rad Laboratories, Richmond, CA. Tetrahydrofuran (THF) was of HPLC grade; dimethylformamide(DMF)was of analytical reagent grade. (Perfluoroocty1)propanoyl imidazolide was prepared from (perfluoroocty1)propanoic acid (Fluorochem Limited, Glossop, Derbyshire, England) as follows: a 4.9-g aliquot of the acid was dissolved in 15 mL of dry THF and added to a stirred solution of 1.8 g of carbonyldiimidazole (Sigma) in 35 mL of dry THF at room temperature. The reaction mixture was stirred for 30 min, during which time the product began to crystallize. The mixture was cooled in an ice bath and filtered. The crystals were washed with ice-cold THF and dried with a stream of air. The yield of (perfluoroocty1)propanoylimidazolide,which had a melting point of 128 "C, was 68% of theoretical yield. Procedure. Perfluoroalkylated urease was prepared by adding 2.0 mL of a solution of (perfluoroocty1)propanoyl imidazolide, dissolved in either THF or DMF (20 mg/mL), to 10.0 mL of urease solution (2 mg/mL in 0.1 M, pH 8.5 phosphate buffer). The reaction mixture was stirred for 2 h at room temperature, and then applied to a 25 X 2.2 cm gel permeation column of Bio-Gel P-6 (Bio-Rad Laboratories), which had been equilibrated with pH 8.5 phosphate buffer. The perfluoroalkylated urease was

0003-2700/88/0360-1996$01.50/0Q 1988 American Chemical Society

ANALYTICAL CHEMISTRY, VOL. 60,

NO. 18, SEPTEMBER 15, 1988

1997

I

#%$

Fm F F F F F F F F H H

PF-Reagent

I

150 -

o

!.R

F F F F F F F F H H

>

t

t

E

J

100 -

O H H F F F F F F F F

H H F F F F F F F F

#

Perfluoroalkylated

50 .

Enzyme

Flgure 1. Basic principle of fluorocarbon immobilization method; R = imidazol-1-yl.

0

-4.5

eluted with phosphate buffer in the exclusion volume of the column, free from unreacted imidazolide reagent and organic solvent. An Amicon 8050 ultrafiltration system with a 100 000 molecular weight cutoff membrane was used to concentrate the perfluoroalkylated urease solution 4-5-fold. Alternatively, the reaction mixture was dialyzed with phosphate buffer using the Amicon system to remove unreacted reagent and organic solvent. The degree of substitution of urease was determined from the difference in available amino groups present before and after reaction. The number of amino groups was determined with standard procedures using trinitrobenzenesulfonic acid (15,16). The activity of the perfluoroalkylated enzyme, which is completely soluble in aqueous solution, was measured and compared to that of the native enzyme. An ammonia gas-sensing electrode was used to measure the rate of ammonia production from a urea solution (0.1 M in 0.2 M, pH 8.5 Tris buffer), catalyzed by the addition of urease. The change in potential with time in the initial stage of the reaction was used as a measure of enzyme activity (17).

Perfluoroalkylated urease was immobilized onto fluorocarbon membranes (Orion ammonia gas-permeablemembrane 95-10-04, Gelman TF-450, and poly(tetrafluoroethylene), 0.2-wm pore size from W. L. Gore Associates) by using the following procedure. The membrane was first partially wetted by dipping it into an aqueous solution of 10% tert-butyl alcohol and 2% Tween 20 (v/v) for 5-10 s. The excess wetting solution was blotted off, and the membrane was placed into the concentrated perfluoroalkylated urease solution for 5-90 min. The membrane was washed with Tris buffer and then dried under vacuum in a desiccator or under a stream of nitrogen. The immobilized enzyme activity was determined by placing the membrane into a vigorously stirred urea solution and measuring the rate of ammonia production with the ammonia sensor, as described above for the soluble enzyme. The urea electrode was prepared by placing the immobilized urease membrane into the body of the ammonia sensor and assembling the sensor as recommended by the manufacturer. The response of the urea electrode was tested by placing it into 25.0 mL of 0.2 M, pH 8.5 Tris-HC1 buffer (18)containing 1.0 mM ethylenediaminetetraacetic acid (EDTA) and making additions of a stock solution of 0.1 M urea, prepared in the same buffer. The steady-state potential reading was recorded after each addition. The eletrodes were stored in Tris buffer at room temperature.

RESULTS AND DISCUSSION At the conditions used for the perfluoroalkylation reaction, 10-1570 of approximately 317 available amino groups of urease were substituted. There was 10-18% loss of urease activity upon perfluoroalkylation. Greater losses of activity were observed with increasing age of the T H F used as cosolvent in the perfluoroalkylation reaction. This loss of activity is probably due to the formation of peroxides in the THF, which oxidize sulfhydryl groups of the enzyme. Consequently, DMF was the preferred cosolvent. Approximately 0.5 IU of urease activity was immobilized on a 1 3 mm diameter membrane after soaking for 30 min. Somewhat higher activities, up to 0.7 IU, were obtained after

-4.0

-3.5

-3.0

-2.5

-2.0

log [Urea]. M

Flgure 2. Calibration curves for urea electrodes prepared with (0) perfluoroalkylated urease and (0) native urease.

Table I. Urea Electrode Stability" slope, mV/decade electrode 2 electrode 3

day

electrode 1

1 2 3 4 6 7

49.0 49.4 47.3 46.8

51.3 46.1

38.5

39.1

50.4 46.3

44.4 45.2

Electrodes were stored in buffer at room temperature.

a 90-min exposure. In most cases, a 30-min soaking was sufficient to give good electrode response characteristics. With the native enzyme, only 0.1 IU was immobilized, thereby demonstrating the enhanced adsorption of the perfluoroalkylated enyzme. Although no significant differences were observed among the membranes studied, most of the results were obtained with the Orion ammonia membrane. The membranes were only partially wetted before exposure to the enzyme by placing them into a mixed aqueous-organic solution for 5-10 s to decrease the surface tension and, thus, improve contact with the solution. Therefore, the amount of enzyme immobilized is limited by the external surface area of the membrane. If the membrane was completely wetted by longer exposure or by using neat organic solvent, the hydrophobicity was destroyed, and the membrane leaked when used in the ammonia electrode. A typical response curve for a urea electrode prepared with the fluorocarbon immobilization method is shown in Figure 2. The response was linear with the logarithm of the urea concentration from 4 X lo4 to 1X M with a slope of 49-51 mV/decade. This response is comparable to that for other urea electrodes reported in the literature (18). The response curve for a urea electrode prepared in the same manner, but with native urease, is also shown. As can be seen, the response of this electrode was very poor, i.e., a response slope of 16.8 mV/decade. The urea electrode had an analytically useful response for up to 7 days (Table I) when stored in buffer a t room temperature. This stability is not as good as that of other urea electrodes (13,18). Typically, enzyme electrodes are designed to operate under mass-transfer-limited conditions, i.e., high enzyme loading, in which the response is independent of enzyme concentration (3). This high enzyme loading results in longer electrode lifetimes because enzyme activity can be lost without a decrease in response. High enzyme loadings are not possible with the fluorocarbon method because of the limited

Anal. Chem. 1988, 60, 1998-2000

1998

50

0.0

I

1 .o

2.0 Time. min

3.0

4 0

Figure 3. Response times of urea electrode for a concentration change of (a) 1 X to 1 X IO-*and (b) 1 X lo4 to 1 X M.

surface area of the membrane. The decrease in response slope that was observed over a 7-day period is apparently due to a loss of enzyme activity. No urease activity was detected in buffer solution in which immobilized enzyme membranes were soaked for 4 days under refrigeration, indicating that the perfluoroalkylated enzyme does not leach from the membrane. Indeed, subsequent studies with fluorescently labeled proteins indicate that the adsorption of perfluoroalkylated protein on fluorocarbon surfaces is essentially irreversible, even in the presence of organic solvents such as acetonitrile and methanol. The nature of the interaction of the perfluoroalkylated protein with the fluorocarbon support is not certain and is currently being investigated. The adsorption may be a strong hydrophobic interaction, or a more specific “fluorophilic” interaction may be involved. The response time characteristics of the urea electrode are shown in Figure 3. The response time, defined as the time required for the potential to become equal to its steady-state value within 1 mV (19),was 3 min for a urea concentration change from 1 x IOm5to 1 x IO-* M and 2 min for a concento 1 X M. These rapid tration change from 1 x IOm4 response times, which result from the thin enzyme layer and the absence of a support membrane, are about twice as fast as other urea electrodes fabricated with the Orion ammonia sensor (13, 20). The fluorocarbon-based immobilization method described herein provides a simple method for the preparation of enzyme electrodes that utilize gas sensors, e.g., ammonia, oxygen, and carbon dioxide electrodes. The perfluoroalkylated enzyme

is adsorbed directly onto the gas-permeable membrane of these sensors, eliminating the need for a support membrane and resulting in improved response times. The main disadvantage of the method is that the amount of enzyme immobilized is small because of the limited external surface area of the membrane. The low enzyme loading results in shorter electrode lifetimes. However, this is not seen as a major disadvantage because of the ease with which a new immobilized enzyme membrane can be prepared. We are currently studying the use of the fluorocarbon immobilization method to immobilize glucose oxidase on the fluorocarbon membrane of an oxygen electrode to make a glucose sensor. A different chemistry is required to perfluoroalkylate glucose oxidase because of the unavailability of amino groups (21). Consequently, perfluoroalkylation via the carbohydrate residues of glucose oxidase is being explored. Furthermore, the fluorocarbon immobilization method provides a general method for protein immobilization on fluorocarbon surfaces, e.g., particles, films, liquids, and fibers, for application in other types of biosensors, enzyme reactors, affinity chromatography, and immunoassays.

LITERATURE CITED (1) Kobos, R. K. TrAC, Trends Anal. Chem. (Pers. Ed.) 1987, 6, 6-9. (2) Gullbault, G. G. Analytical Uses of Immobillzed Enzymes; Dekker: New York, 1984;Chapter 3. (3) Carr, P. W.; Bowers, L. D. Immobllized Enzymes in Ana/yticai and Clinical Chemistry; Wiley: New York, 1980 Chapter 5. (4) Kobos, R. K. in Ion-Selecthe Electrodes in Ana/ytical Chemistry; Freiser, H.. Ed.; Plenum: New York, 1980;Volume 2,Chapter 1. (5) Mascini, M.; Iannello, M.; Palleschi, G. Anal. Chim. Acta 1985, 746, 135-148. (6) Gullbault, G. G.;Czarnecki, J. P.; Nabi Rahnl, M. A. Anal. Chem. 1985, 57. 2110-2116. (7) Koyama, M.; Sato, Y.; Suzuki, S. Anal. Chim. Acta 1980, 716, 307-314. (8) Tran-Minh, C.; Broun, G. Anal. Chem. 1975,4 7 , 1359-1364. (9) Lubrano, G. J.; Gullbault, G. G. Anal. Chim. Acta 1978,9 7 , 229-236. (10) Weaver, M. R.; Vadgama, P. M. Clin. Chim. Acta 1988, 755, 295-308. (I 1) Fishman, J. H. US. Patent 3 843 443, 1974. (12) Busby, M. G.; Hartwig, D. E. U S . Patent 4317879,1982. (13) Anfalt, T.; Graneli, A.; Jagner, D. Anal. Lett. 1973,6 , 969-975. (14) Hato, M.; Shimura, Y.; Tsuda, K. U.S. Patent 4819897, 1986. (15) Means, G. E.; Feeney, R . E. Chemkal Modlflcetions of Proteins; Holden-Day: San Francisco, 1971;pp 30-33, 121. (16) Okuyama, T.; Satake, K. J. Blochem. (Tokyo) 1960, 47. 454-468. (17) Gullbault, G. G.; Smith, R. K.; Montalvo, J. G. Anal. Cham. 196S94 7 , 600-805. (18) Mascini, M.; Gullbault, G. G. Anal. Chem. 1977,4 9 , 795-798. (19) Pure Appl. Chem. 1976, 48, 127-132. (20) Papastathopoulos, D. S.;Rechnltz, G. A. Anal. Chim. Acta 1975, 7 9 , 17-26. (21) Degani, Y.; Heller, A. J. Phys. Chem. 1987,91. 1285-1269.

RECEIVED for review February 8,1988. Accepted May 23,1988. This work was presented in part at the 1987 Annual Meeting of the Society for Industrial Microbiology, Baltimore, MD.

Display of the Composltlon of Polychlorinated Biphenyls Vladimir Zitko Marine Chemistry Division, Department of Fisheries and Oceans, St. Andrews, New Brunswick, Canada EOG 2x0 Because of environmental and toxicological concerns, POlychlorinated biphenyls (PCBs) are a widely studied group of chemicals consisting of 209 chlorobiphenyls. Most of the studies now include individual chlorobiphenyls and result in long tables of data. This note describes a graphical presentation of such data. The presentation helps to visually classify

samples, to detect patterns in the presence of chlorobiphenyls in the samples, and to relate qualitatively substitution and properties of chlorobiphenyls. The derivation of quantitative relationships may be possible in some instances. The condensation of substitution patterns into single numbers (coordinates) also leads to an abbreviated nomenclature of

0003-2700/88/0360-1998$01.50/0 0 1988 American Chemical Society