Biotechnol. Prog. 1992, 8, 268-274
288
Use of Monoclonal Anti-enzyme Antibodies for Analytical Purposes P. Chandrani Gunaratna*and George S. Wilson Department of Chemistry, University of Kansas, Lawrence, Kansas 66045
This review discusses the analytical applications of monoclonal antibodies specific for enzymes. One important, but not well-studied, application of these monoclonal antibodies is their use in immobilizing enzymes on solid supports. This method is based on binding the enzymes to an immobilized antibody through the antigen bindng site of the antibody. Enzymes immobilized this way retain much of their activity. The utility of immobilized enzyme reactors prepared by immobilizing the enzymes through antibodies is demonstrated by using them in the determination of acetylcholine and choline in brain tissue extracts. Currently available methods for immobilizing antibodies and enzymes are reviewed. Other issues discussed in this review include the problems and advantages of immobilized enzyme reactors, especially when used in conjunction with HPLC. In addition, the applications of monoclonal antibodies for the detection and measurement of enzymes and their isoforms are summarized.
Contents Introduction Analytical Applications of Monoclonal Anti-enzyme Antibodies Immobilized Enzyme Reactors Methods of Enzyme Immobilization Immobilization of Antibodies Detection of Acetylcholine Tissue Analysis of Acetylcholine Conclusion Future Applications
268 268 269 270
270 27 1 272 273
273
Introduction Monoclonal antibodies (MoABs) specific for enzymes have been used in numerous analytical applications in recent years. They are becoming very powerful tools in enzyme research and are extensively used in analytical methodology because they offer many advantages over polyclonal antibodies. A major drawback of polyclonals is the variation of the antibody affinity from lot to lot or from animal to animal. Since the affinity constant of the particular antibody used plays an important role in analytical applications, one needs to affinity purify the polyclonal antibody and evaluate the affinity constant with eachnew source of antibody. Unlike the polyclonals, monoclonals have a homogeneousantibody population specific to a particular antigenic epitope. Therefore, once the cell line producing the specific antibody is isolated, then a continuous source of homogeneous analytical reagent is available.
Analytical Applications of Monoclonal Anti-enzyme Antibodies Monoclonal anti-enzyme antibodies can be used for a variety of purposes. The most common application is the use of these antibodies in the immunoaffinity purification of enzymes (Santos et al., 1984;Brent et al., 1990; Hanson
* Address correspondence to this author at the following present address: Department of Pharmaceutical Chemistry, University of Kansas, Lawrence, KS 66045.
and Beavo, 1982). Antibodies with low or moderate affinity constants are desirable for this purpose. MoABs affecting the enzyme activity represent a valuable tool for suggesting structure-function relationships, and hence, they have been used extensively as tools for probing the molecular mechanism of enzyme catalysis (Wilson, 1987; Tipton et al., 1990). If the MoABs to a particular epitope on the enzyme affect the catalytic activity, then it suggests that a particular region is directly or indirectly involved in the enzyme catalysis. Functional roles of different subunits in multisubunit enzymes can be studied in this way (Lewis et al., 1982; Park et al., 1982). Another important application of MoABs is their use for discrimination between multiple molecular forms of enzymes and isozymes. MoABs have been used to identify (Denney et al., 1982a),purify (Hauri et al., 1982; Denney et al., 1982b),and assay (Uemura et al., 1990; Jorgensen et al., 1990) these isozymes. Major clinical applications of MoABs to these isozymes is for the measurement of isozyme concentration for diagnostic purposes. One such application is the assay of creatine kinase MB (CK-MB) isozyme in serum for the diagnosis of myocardial infarction. The popular “two-site” assay for CK-MB uses two monoclonal antibodies, one to the M subunit and the other to the B subunit of CK-MB. New assay procedures using a CK-MB-specific monoclonal antibody have been developed (Vaidya et al., 1986). Other applications of isozyme measurements in the clinical field include the assay of the pancreatic isozyme of a-amylase (Gerber et al., 1987), the prostatic isozyme of acid phosphatase (Lin et al., 1983), and the isozymes of alkaline phosphatase, especially the placental alkaline phosphatase (de Groote et al., 1983). There are many more applications in the above-mentioned areas in the literature. Interested readers are referred to reviews concerning these applications (Vora, 1985; Rosalki, 1989). An important, but not well-documented application of MoABs to enzymes is their use for the immobilization of enzymes on solid supports. Preparation of supports with highly active immobilized enzymes has been described (de Alwis et al., 1987; Lomen et al., 1989). In another application, mouse MoABs have been used to immobilize carboxypeptidaseA on Eupergit C and Sepharose-protein A (Solomon and Coppel, 1987). The immobilized enzyme was considerably more stable than the native enzyme at
8756-7938/92/3008-0268$03,00/00 1992 American Chemical Society and American Institute of Chemical Engineers
269
Biotechnol. Rug..,1992, Vol. 8, No. 4
a) Binding through the catalytic site
b) Binding near the catalytic site Conformational change
P. Chandrani Gunaratna is a Postdoctoral Fellow in the
A c)
Sterically hindered catalytic site
d) Ideal binding
Figure 1. Illustration of different orientations in enzyme coupling to a solid support.
lower pH and retained 70% of its initial activity at pH 4.5, whereas the native enzyme retained only 30 % of its original activity. Relatively few applications have been reported in the literature on this subject, despite the importance and the convenience of using MoABs for enzyme immobilization. This article will demonstrate the use of monoclonal antibodies to obtain highly active immobilized enzymes for the preparation of immobilized enzyme reactors (IMERs), on the basis of our experience with the development of a bioenzyme reactor (Gunaratna and Wilson, 1990). The applicability of this IMER is demonstrated by the detection of acetylcholine (ACh) in brain tissue samples.
Immobilized Enzyme Reactors Recent advances in the chemistry of enzyme coupling allow the preparation of immobilized enzymes, which can be used repeatedly. In addition to being reusable, immobilized enzymes also eliminate the instability problems associated with the soluble enzyme. IMERs have found many useful applications in biotechnological processes such as fermentation, in clinical and biochemical applications, and in analytical assay methods. One of the main advantages of IMERs in analytical assays is their high specificity, particularly for the analytes in biological samples. The desired analyte can be separated from the other components in the matrix by high-performanceliquid chromatography (HPLC),and a post-column IMER can be used to convert the analyte into a product which can be detected by usual analytical methods, spectroscopic or electrochemical. This is often necessary because the concentration of the eluting analyte is too low to be detected directly. IMERs are convenient and allow rapid manipulation of reactions to permit the simultaneous determination of more than one component in a sample (Gorton and Ogren, 1981). However, IMERs are not without limitations. One of the main problems in using IMERs in conjunction with HPLC is the incompatibility of the optimal media for separation and for the enzymatic reaction. Reverse-phase analytical columns are vulnerable to degradation at the alkaline pH sometimes necessary for optimal enzyme activity. Most of the chromatographic separations require mobile phases with pH < 7 and/or contain organic solvents or modifiers such as ion-pairing agents which may inactivate the enzyme column (Svobodnik et al., 1987; Lam et
Pharmaceutical Chemistry Department a t University of Kansas. She received her B.Sc. degree from University of Colombo in Sri Lanka and her Ph.D. in analytical chemistry from the University of Pittsburgh in 1986under the direction of J. F. Coetzee. She then joined the National Institute of Standards and Technology as a research associate in the Inorganic Analytical Division. From 1988 to 1990, she was employed as a Postdoctoral Research Associate in the Chemistry Department a t University of Kansas. Her research interests include development of immunoassay methods for anti-cancer drugs, applications of antibodies in biomedical analysis, and the use of immobilized enzymes in conjunction with HPLC for analysis of biological materials.
George S. Wilson is currently Higuchi Distinguished Professor of Chemistry and Pharmaceutical Chemistry a t the University of Kansas. He received his A.B. degree from Princeton University in 1961 and a Ph.D. from the University of Illinois in 1965. His current research interests include structural effects on the electron transfer properties of proteins, particularly cytochromes, the redox chemistry of thioethers such as methionine, and the properties of immobilized proteins which are used in analytical applications of biological recognition, especially in problems of biomedical and pharmaceutical interest. He serves presently on the Editorial Board of Biosensors and Bioelectronics and is the Chairman of the IUPAC Commission on Electrochemistry.
al., 1988; Boppana et al., 1986). In these cases, either an addition of post-column buffer into the analysis stream by another pump or adjustment of the chromatographic conditions to suit the enzyme column is necessary. Addition of another buffer stream can be a source of baseline instability and may result in a lower S/N ratio. Also, a mixing coil may be needed to assure the proper mixing of the two streams. Adjusting the mobile phase may compromise the separation of the desired component. In either case, the sensitivity of the assay method will be greatly diminished. Another major disadvantage of these IMERs is the loss of enzyme activity during the immobilization step. Commonly used immobilization methods covalently bind an active amino group of the enzyme to an activated functional group of the support. Usually these active amino groups are located at or near the catalytic site. The enzyme can bind to the surface in different orientations as shown in Figure 1. As a result of the covalent attachment, the
Biotechnol. Prog., 1992, Vol. 8, No. 4
270
Figure 2. Schematic representation of various immobilization methods: (a) direct, (b) spacer, (c) avidin-biotin method, (d) antibody method.
conformation of the catalytic site may change and the substrate will no longer recognizethe enzyme. Multipoint attachment of the enzyme to the surface through several functionalities can also cause a substantial loss of activity. In addition, the enzyme can bind to the surface in such a way that the active site is sterically hindered and the substrate cannot access the active site freely. This type of covalent binding affects enzymes which have low intrinsic activity and are very sensitive to their environment. Any changes in the immediate surroundings can cause a significant loss of enzyme activity. Such an example is choline oxidase in our studies. Enzymes with high intrinsic activity or with several subunits (e.g., acetylcholinesterase) can be immobilized with retention of much of their activity.
Methods of Enzyme Immobilization Typical covalent coupling of an enzyme causes it to lose 80-90% of its activity. One way to couple enzymes to retain high activity is to attach them through a spacer arm such as triethylenetetramine tetrahydrochloride (TETA) at a specific distance from the active site. Several spacer arms are commercially available. A long spacer arm could hold the enzyme farther away from the solid support so that the active site is effectively accessible to the substrate. An excellent system for this purpose is the avidin-biotin system. The biotin molecule can easily be activated and coupled to enzymes with complete retention of activity. Avidin, a tetramer containing four identical subunits, can be immobilized on the support by the usual covalent coupling methods. The avidin-biotin linkage is strong and undisturbed by extremes of pH or ionic strength. These coupling methods are illustrated in Figure 2a-c using a carbonyldiimidazole-activated solid support. Immobilization of enzymes can be carried out on support matrices activated with other functional groups such as epoxy tresyl and N-hydroxysuccinimide (Leckband and Langer, 1991; Ulbrich et al., 1991).
Coupling through a spacer arm might not alleviate the activity problem for enzymes which undergo conformational changes due to the covalent coupling. This is where a monoclonal antibody to the enzyme could be very useful because the interaction between the antibody and the enzyme (antigen) is strong, yet mild enough not to cause a noticeable conformational change. Also, since the antibody molecule is a rather large molecule, it separates the enzyme from the surface. Therefore, the enzyme immobilized through antibodies might have less steric hindrance and might possess more accessible active sites. This is particularly true if antibodies are selected which have the property that they bind to an enzyme without a resulting loss of enzymatic activity. Enzyme immobilization through antibodies is a novel technique. In this method, an antibody specific to the desired enzyme is first immobilized on the support and then the enzyme is allowed to bind immunochemically to the antigen binding site of the antibody as shown in Figure 2d. The monoclonalantibodies used for this purpose must have an affinity constant high enough (>lo9 M) for effective binding and must not inhibit the enzyme activity. Also, they should be of the IgG class. Enzyme reactors prepared by immobilization via monoclonal antibodies have the added advantage of easy and quick regeneration. This feature is very valuable especially in cases where the chromatographic conditions could cause the rapid deterioration of the enzyme column as mentioned above. Once the activity of the enzyme is lost after repeated use, the inactive enzymecan be eluted from the reactor by changing the pH of the buffer and new enzyme can be immobilized within minutes. Another significant benefit of enzyme immobilization via monoclonal antibodies is realized in the preparation of multistep enzyme reactors. Multienzyme systems coimmobilized on the same support have been used as model systems to study in vivo mechanisms (Bouin et al., 1976). Such systems can also replace single IMERs used in series in analytical assay methods where several enzyme reactors are needed to carry out a reaction sequence (Koerner and Nieman, 1988). Co-immobilization using antibodies can provide higher sensitivity in the assay system due to the increased local intermediate substrate concentration surrounding the first enzyme in the sequence which is then immediately available for reaction with the next enzyme. In conventional multienzyme systems when the enzymes are simultaneously immobilized by covalent coupling techniques, it is difficult to predict the enzyme activity ratio (Mosbachand Mattiasson, 1976). This problem does not arise when the enzymes are immobilized using antibodies since the enzyme loading can be carefully controlled as described later in this article. The importance of controlling the enzyme activities in multienzyme systems has been indicated (Lomen et al., 1986).
Immobilization of Antibodies The amount of active immobilized enzyme depends on the antigen binding capability of the immobilizedantibody molecule. When an antibody is immobilized, it also can attach to the surface in different orientations (see Figure 2d). Therefore, it is important to immobilizethe antibody in an orientation such that the immobilization does not adversely affect immunological activity of the antibody. In direct immobilization of antibodies, the surface of the support has functional groups which react with the amino groups of the antibody molecule. A considerable amount of antibody activity is lost this way, and the effective activity is only about 10-15 7%. Even though most antibody immobilizationmethods involve random coupling
Biotechnol. Rug., 1992, Vol. 8, No. 4
Oxldized antibody
271
3 ;I
(a)
1
+
+y-
Y-I
H +
H20
P
(b)
to a functional group on the immobilizing matrix. This method has the advantage of not having to immobilize the antibody. In addition, multienzyme immobilization on the same support can be accomplished easily by this method.
Detection of Acetylcholine The applicability of the monoclonal antibodies to prepare IMERs has been demonstrated by using these IMERs to detect acetylcholine in brain tissue samples. IMERa have been used with high-performance liquid chromatography (HPLC) and electrochemical detection to detect acetylcholine (ACh) and choline (Ch) (Potter et al., 1983; Beley et al., 1987). Acetylcholine, the main neurotransmitter in the cholinergic neurotransmitter system, has been found to be involved in various neurological disorders (Davis and Berger, 1979). Increasingly sensitive assay techniques are required to measure the low concentrations of ACh and Ch in brain. In our assay method after the separation of ACh from brain tissue extracts by reverse-phase HPLC, the effluent is allowed to pass through acetylcholinesterase (AChE) and choline oxidase (ChO) immobilized reactors. Hydrogen peroxide produced by the following reaction scheme is then detected amperometrically.
(d)
Figure 3. Schematic representation of orientated coupling of antibodies: (a) through the carbohydrate moiety, (b) immunochemical attachment, (c) Fab’ fragment, (d) protein A.
of amino groups at or near the antigen binding site, it is preferable to couple the antibody molecule through its Fc region by an immunochemical attachment (Mifflin et al., 1989)or through the carbohydrate moiety (Cress and Ngo, 1989) or via the -SH group of an Fab’ fragment (de Alwis and Wilson, 1987). The antibody molecule contains a carbohydrate moiety linked to the Fc region. Under mild conditions, sodium periodate can oxidize the vicinal hydroxyl groups of this carbohydrate to aldehyde groups. The antibody can then be linked to a hydrazine-activated support through these aldehyde groups. The IgG molecule can be reduced to Fab’ after the molecule is cleaved into F(ab’)2 by pepsin digestion. The monovalent Fab’ fragment can then be immobilized through the -SH group a t the hinge region (de Alwis and Wilson, 1987; Prisyazhony et al., 1988). Protein A and protein G are known to bind antibodies of the IgG class through the Fc region (Sisson and Castor, 1990; Janis and Regnier, 19891, and they also can be used for site-directed coupling. The antibody molecule can also be immobilized via another antibody by an immunochemical attachment. For example, mouse immunoglobulin can be immobilized by first coupling polyclonal goat antimouse antibodies to the support. A support prepared this way is considered to be a universal system since the bound antibody can be eluted by washing with a chaotropic agent and another mouse antibody can be immobilized onto the same support. It is also possible to bind two MoABs simultaneously to the same support. These oriented coupling methods are shown in Figure 3. Wimalasena and Wilson (1991)have carried out a more comprehensivestudy on the factors affecting the antigen binding capability of the immobilizedantibodies and antibody fragments. Their study shows the protocols for optimizing immobilized antibody performance. Enzyme immobilizationthrough bifunctional antibodies has also been suggested (Nolan and O’Kennedy, 1990). Half of the bifunctional antibody can be raised to a single epitope on the desired enzyme so that binding does not inhibit the enzyme. The other half can specifically bind
-
AChE
ACh
choline (Ch) + CH,COO-
-
ChO
Ch
betaine
+ 2H,O,
Performance of the IMERs depends on the properties of both of the immobilized enzymes. Since AChE is more durable than ChO, it can be immobilized without a loss of activity. ChO is very sensitive to its environment, and significant losses on immobilization are quite common. The two enzymes used in these experiments, ChO and AChE, have very different properties. ChO is a monomer with an approximate molecular weight of 72 OOO determined by gel filtration (Ohta-Fukuyama et al., 1980)and which contains an FAD group as the active site. AChE exists as a dimer usually containing two or three tetrameric seta of catalytic subunits (MacPhee-Quigley et al., 1986). AChE has a very high turnover number (Zubay, 1986)and has a high specific activity of about 1000 units/mg. ChO, on the other hand, has a very low intrinsic turnover rate and specific activity ( 10units/mg) as compared to AChE. The enzymes were coupled to the support material by using the methods shown in Figure 2 in order to prepare the AChE and ChO IMERs. No inhibition of the enzyme activity was observed when an excess of each antibody was mixed in solution with its respective enzyme antigen. This suggeststhat the antibodies do not inhibit the enzyme activity. Antibodies to both ChO and AChE were immobilized on the support as described above. Before the introduction of the enzymes, the column was saturated with bovine serum albumin to prevent the nonspecific adsorptionof the enzymes. The enzymeswere immobilized by injecting an aliquot of enzyme solution in carrier buffer. The bound and free fractions were evaluated for different injection volumes and different enzyme concentrations at different flow rates to establish an optimum concentration of enzyme and injection volume to get the maximum enzyme loading on the column. In this way, the enzyme loading on the column can carefully be controlled as desired and the specific activity of the enzyme on the column can be known. Performance of these IMERs was then evaluated individually in the FIA mode (HPLC/IMER
-
Biotechnol. Prog., 1992, Vol.
272
a, No. 4
Table I. Comparison of Immobilization Methods method detection response4 limit (M) (nA) direct direct 10-4 spacer 10-7 5.1 direct 10-8 13.6 avidin-biotin avidin-biotin 10-9 72.4 avidin-biotin avidin-biotinb antibody antibody 10-8 10.9 a Current response to 200 pmol of ACh. *Both enzymes are immobilized on the same support. (Reprinted with permission from Anal. Chem. 1990, 52, 402. Copyright 1990 American Chemical Society.)
Table 11. Performance of ChO Supports
immobilization method direct spacer avidin-biotin avidin-biotin both enzymes antibody antibody
loading of ChO (mg/mL of support) 1.2 (5.5)" 1.0 (4.6)" 2.5 (11.5)" 2.5 (11.5)" 0.02 (0.09)" 0.02 (0.09)"
%
uptake of ChO 85
81 91 84 100 100
immobilized enzyme percent activity activityb (units) 0.32 0.015 0.23 6.2 36.3 3.8 32.1 3.1 0.023 0.018
25.6 20.0
Units of enzyme. Percent of activity on the support relative to that of the soluble enzyme. Antibody immobilized through goat anti-mousespecific antibody (Reprinted with permission from Anal. Chem. 1990,62,402. Copyright 1990 American Chemical Society.)
system without the reverse-phase LC column). These results are tabulated in Table I. Direct immobilization of ChO yielded poor results. It was found that the activity of AChE was not significantly affected by the coupling except in the case of the coupling through a spacer arm, where AChE activity was somewhat reduced. The enzymes retained the highest activity when they were coupled through avidin-biotin linkages. The sensitivity was further enhanced by co-immobilizing the biotinylated enzymes to the same avidin-bound support material. This observation is in agreement with the reported literature (Mosbach and Mattiasson, 1976),where it has been shown that co-immobilization of two enzymes on the same support causes enhanced efficiency. They explained this observation as due to the availability of increased local concentration of the product from the first enzyme around the second enzyme in the reaction sequence. This is possible when the enzymes are coimmobilized on the same particle because the enzymes are in close proximity of each other. Therefore, the second enzyme will catalyze its reaction more efficiently and increase the overall reaction rate leading to the enhanced sensitivity in the assay system. When ChO antibody was bound to the support and used to immobilize ChO, it was found that the enzyme binds to the antibody effectively despite the fact that it has an apparently lower binding affinity than the AChE antibody. Our studies on these two IMERs showed that the activity of the ChO column governs the sensitivity and the detection limit of the assay. Therefore, a thorough study has been done to evaluate the ChO activity on the support after immobilization. These results are summarized in Table 11. The data in Table I1 show that when ChO is immobilized through the avidin-biotin linkage, it retains the highest activity. Enzyme activity on the supports which have ChO immobilized through the antibody is considerably higher than with the other methods but lower than with the avidin-biotin method. This can be attributed to the random coupling of the antibody which leads to a low number of
ii
eChEantibody
288
Use of Monoclonal Anti-enzyme Antibodies for Analytical Purposes P. Chandrani Gunaratna*and George S. Wilson Department of Chemistry, University of Kansas, Lawrence, Kansas 66045
This review discusses the analytical applications of monoclonal antibodies specific for enzymes. One important, but not well-studied, application of these monoclonal antibodies is their use in immobilizing enzymes on solid supports. This method is based on binding the enzymes to an immobilized antibody through the antigen bindng site of the antibody. Enzymes immobilized this way retain much of their activity. The utility of immobilized enzyme reactors prepared by immobilizing the enzymes through antibodies is demonstrated by using them in the determination of acetylcholine and choline in brain tissue extracts. Currently available methods for immobilizing antibodies and enzymes are reviewed. Other issues discussed in this review include the problems and advantages of immobilized enzyme reactors, especially when used in conjunction with HPLC. In addition, the applications of monoclonal antibodies for the detection and measurement of enzymes and their isoforms are summarized.
Contents Introduction Analytical Applications of Monoclonal Anti-enzyme Antibodies Immobilized Enzyme Reactors Methods of Enzyme Immobilization Immobilization of Antibodies Detection of Acetylcholine Tissue Analysis of Acetylcholine Conclusion Future Applications
268 268 269 270
270 27 1 272 273
273
Introduction Monoclonal antibodies (MoABs) specific for enzymes have been used in numerous analytical applications in recent years. They are becoming very powerful tools in enzyme research and are extensively used in analytical methodology because they offer many advantages over polyclonal antibodies. A major drawback of polyclonals is the variation of the antibody affinity from lot to lot or from animal to animal. Since the affinity constant of the particular antibody used plays an important role in analytical applications, one needs to affinity purify the polyclonal antibody and evaluate the affinity constant with eachnew source of antibody. Unlike the polyclonals, monoclonals have a homogeneousantibody population specific to a particular antigenic epitope. Therefore, once the cell line producing the specific antibody is isolated, then a continuous source of homogeneous analytical reagent is available.
Analytical Applications of Monoclonal Anti-enzyme Antibodies Monoclonal anti-enzyme antibodies can be used for a variety of purposes. The most common application is the use of these antibodies in the immunoaffinity purification of enzymes (Santos et al., 1984;Brent et al., 1990; Hanson
* Address correspondence to this author at the following present address: Department of Pharmaceutical Chemistry, University of Kansas, Lawrence, KS 66045.
and Beavo, 1982). Antibodies with low or moderate affinity constants are desirable for this purpose. MoABs affecting the enzyme activity represent a valuable tool for suggesting structure-function relationships, and hence, they have been used extensively as tools for probing the molecular mechanism of enzyme catalysis (Wilson, 1987; Tipton et al., 1990). If the MoABs to a particular epitope on the enzyme affect the catalytic activity, then it suggests that a particular region is directly or indirectly involved in the enzyme catalysis. Functional roles of different subunits in multisubunit enzymes can be studied in this way (Lewis et al., 1982; Park et al., 1982). Another important application of MoABs is their use for discrimination between multiple molecular forms of enzymes and isozymes. MoABs have been used to identify (Denney et al., 1982a),purify (Hauri et al., 1982; Denney et al., 1982b),and assay (Uemura et al., 1990; Jorgensen et al., 1990) these isozymes. Major clinical applications of MoABs to these isozymes is for the measurement of isozyme concentration for diagnostic purposes. One such application is the assay of creatine kinase MB (CK-MB) isozyme in serum for the diagnosis of myocardial infarction. The popular “two-site” assay for CK-MB uses two monoclonal antibodies, one to the M subunit and the other to the B subunit of CK-MB. New assay procedures using a CK-MB-specific monoclonal antibody have been developed (Vaidya et al., 1986). Other applications of isozyme measurements in the clinical field include the assay of the pancreatic isozyme of a-amylase (Gerber et al., 1987), the prostatic isozyme of acid phosphatase (Lin et al., 1983), and the isozymes of alkaline phosphatase, especially the placental alkaline phosphatase (de Groote et al., 1983). There are many more applications in the above-mentioned areas in the literature. Interested readers are referred to reviews concerning these applications (Vora, 1985; Rosalki, 1989). An important, but not well-documented application of MoABs to enzymes is their use for the immobilization of enzymes on solid supports. Preparation of supports with highly active immobilized enzymes has been described (de Alwis et al., 1987; Lomen et al., 1989). In another application, mouse MoABs have been used to immobilize carboxypeptidaseA on Eupergit C and Sepharose-protein A (Solomon and Coppel, 1987). The immobilized enzyme was considerably more stable than the native enzyme at
8756-7938/92/3008-0268$03,00/00 1992 American Chemical Society and American Institute of Chemical Engineers
269
Biotechnol. Rug..,1992, Vol. 8, No. 4
a) Binding through the catalytic site
b) Binding near the catalytic site Conformational change
P. Chandrani Gunaratna is a Postdoctoral Fellow in the
A c)
Sterically hindered catalytic site
d) Ideal binding
Figure 1. Illustration of different orientations in enzyme coupling to a solid support.
lower pH and retained 70% of its initial activity at pH 4.5, whereas the native enzyme retained only 30 % of its original activity. Relatively few applications have been reported in the literature on this subject, despite the importance and the convenience of using MoABs for enzyme immobilization. This article will demonstrate the use of monoclonal antibodies to obtain highly active immobilized enzymes for the preparation of immobilized enzyme reactors (IMERs), on the basis of our experience with the development of a bioenzyme reactor (Gunaratna and Wilson, 1990). The applicability of this IMER is demonstrated by the detection of acetylcholine (ACh) in brain tissue samples.
Immobilized Enzyme Reactors Recent advances in the chemistry of enzyme coupling allow the preparation of immobilized enzymes, which can be used repeatedly. In addition to being reusable, immobilized enzymes also eliminate the instability problems associated with the soluble enzyme. IMERs have found many useful applications in biotechnological processes such as fermentation, in clinical and biochemical applications, and in analytical assay methods. One of the main advantages of IMERs in analytical assays is their high specificity, particularly for the analytes in biological samples. The desired analyte can be separated from the other components in the matrix by high-performanceliquid chromatography (HPLC),and a post-column IMER can be used to convert the analyte into a product which can be detected by usual analytical methods, spectroscopic or electrochemical. This is often necessary because the concentration of the eluting analyte is too low to be detected directly. IMERs are convenient and allow rapid manipulation of reactions to permit the simultaneous determination of more than one component in a sample (Gorton and Ogren, 1981). However, IMERs are not without limitations. One of the main problems in using IMERs in conjunction with HPLC is the incompatibility of the optimal media for separation and for the enzymatic reaction. Reverse-phase analytical columns are vulnerable to degradation at the alkaline pH sometimes necessary for optimal enzyme activity. Most of the chromatographic separations require mobile phases with pH < 7 and/or contain organic solvents or modifiers such as ion-pairing agents which may inactivate the enzyme column (Svobodnik et al., 1987; Lam et
Pharmaceutical Chemistry Department a t University of Kansas. She received her B.Sc. degree from University of Colombo in Sri Lanka and her Ph.D. in analytical chemistry from the University of Pittsburgh in 1986under the direction of J. F. Coetzee. She then joined the National Institute of Standards and Technology as a research associate in the Inorganic Analytical Division. From 1988 to 1990, she was employed as a Postdoctoral Research Associate in the Chemistry Department a t University of Kansas. Her research interests include development of immunoassay methods for anti-cancer drugs, applications of antibodies in biomedical analysis, and the use of immobilized enzymes in conjunction with HPLC for analysis of biological materials.
George S. Wilson is currently Higuchi Distinguished Professor of Chemistry and Pharmaceutical Chemistry a t the University of Kansas. He received his A.B. degree from Princeton University in 1961 and a Ph.D. from the University of Illinois in 1965. His current research interests include structural effects on the electron transfer properties of proteins, particularly cytochromes, the redox chemistry of thioethers such as methionine, and the properties of immobilized proteins which are used in analytical applications of biological recognition, especially in problems of biomedical and pharmaceutical interest. He serves presently on the Editorial Board of Biosensors and Bioelectronics and is the Chairman of the IUPAC Commission on Electrochemistry.
al., 1988; Boppana et al., 1986). In these cases, either an addition of post-column buffer into the analysis stream by another pump or adjustment of the chromatographic conditions to suit the enzyme column is necessary. Addition of another buffer stream can be a source of baseline instability and may result in a lower S/N ratio. Also, a mixing coil may be needed to assure the proper mixing of the two streams. Adjusting the mobile phase may compromise the separation of the desired component. In either case, the sensitivity of the assay method will be greatly diminished. Another major disadvantage of these IMERs is the loss of enzyme activity during the immobilization step. Commonly used immobilization methods covalently bind an active amino group of the enzyme to an activated functional group of the support. Usually these active amino groups are located at or near the catalytic site. The enzyme can bind to the surface in different orientations as shown in Figure 1. As a result of the covalent attachment, the
Biotechnol. Prog., 1992, Vol. 8, No. 4
270
Figure 2. Schematic representation of various immobilization methods: (a) direct, (b) spacer, (c) avidin-biotin method, (d) antibody method.
conformation of the catalytic site may change and the substrate will no longer recognizethe enzyme. Multipoint attachment of the enzyme to the surface through several functionalities can also cause a substantial loss of activity. In addition, the enzyme can bind to the surface in such a way that the active site is sterically hindered and the substrate cannot access the active site freely. This type of covalent binding affects enzymes which have low intrinsic activity and are very sensitive to their environment. Any changes in the immediate surroundings can cause a significant loss of enzyme activity. Such an example is choline oxidase in our studies. Enzymes with high intrinsic activity or with several subunits (e.g., acetylcholinesterase) can be immobilized with retention of much of their activity.
Methods of Enzyme Immobilization Typical covalent coupling of an enzyme causes it to lose 80-90% of its activity. One way to couple enzymes to retain high activity is to attach them through a spacer arm such as triethylenetetramine tetrahydrochloride (TETA) at a specific distance from the active site. Several spacer arms are commercially available. A long spacer arm could hold the enzyme farther away from the solid support so that the active site is effectively accessible to the substrate. An excellent system for this purpose is the avidin-biotin system. The biotin molecule can easily be activated and coupled to enzymes with complete retention of activity. Avidin, a tetramer containing four identical subunits, can be immobilized on the support by the usual covalent coupling methods. The avidin-biotin linkage is strong and undisturbed by extremes of pH or ionic strength. These coupling methods are illustrated in Figure 2a-c using a carbonyldiimidazole-activated solid support. Immobilization of enzymes can be carried out on support matrices activated with other functional groups such as epoxy tresyl and N-hydroxysuccinimide (Leckband and Langer, 1991; Ulbrich et al., 1991).
Coupling through a spacer arm might not alleviate the activity problem for enzymes which undergo conformational changes due to the covalent coupling. This is where a monoclonal antibody to the enzyme could be very useful because the interaction between the antibody and the enzyme (antigen) is strong, yet mild enough not to cause a noticeable conformational change. Also, since the antibody molecule is a rather large molecule, it separates the enzyme from the surface. Therefore, the enzyme immobilized through antibodies might have less steric hindrance and might possess more accessible active sites. This is particularly true if antibodies are selected which have the property that they bind to an enzyme without a resulting loss of enzymatic activity. Enzyme immobilization through antibodies is a novel technique. In this method, an antibody specific to the desired enzyme is first immobilized on the support and then the enzyme is allowed to bind immunochemically to the antigen binding site of the antibody as shown in Figure 2d. The monoclonalantibodies used for this purpose must have an affinity constant high enough (>lo9 M) for effective binding and must not inhibit the enzyme activity. Also, they should be of the IgG class. Enzyme reactors prepared by immobilization via monoclonal antibodies have the added advantage of easy and quick regeneration. This feature is very valuable especially in cases where the chromatographic conditions could cause the rapid deterioration of the enzyme column as mentioned above. Once the activity of the enzyme is lost after repeated use, the inactive enzymecan be eluted from the reactor by changing the pH of the buffer and new enzyme can be immobilized within minutes. Another significant benefit of enzyme immobilization via monoclonal antibodies is realized in the preparation of multistep enzyme reactors. Multienzyme systems coimmobilized on the same support have been used as model systems to study in vivo mechanisms (Bouin et al., 1976). Such systems can also replace single IMERs used in series in analytical assay methods where several enzyme reactors are needed to carry out a reaction sequence (Koerner and Nieman, 1988). Co-immobilization using antibodies can provide higher sensitivity in the assay system due to the increased local intermediate substrate concentration surrounding the first enzyme in the sequence which is then immediately available for reaction with the next enzyme. In conventional multienzyme systems when the enzymes are simultaneously immobilized by covalent coupling techniques, it is difficult to predict the enzyme activity ratio (Mosbachand Mattiasson, 1976). This problem does not arise when the enzymes are immobilized using antibodies since the enzyme loading can be carefully controlled as described later in this article. The importance of controlling the enzyme activities in multienzyme systems has been indicated (Lomen et al., 1986).
Immobilization of Antibodies The amount of active immobilized enzyme depends on the antigen binding capability of the immobilizedantibody molecule. When an antibody is immobilized, it also can attach to the surface in different orientations (see Figure 2d). Therefore, it is important to immobilizethe antibody in an orientation such that the immobilization does not adversely affect immunological activity of the antibody. In direct immobilization of antibodies, the surface of the support has functional groups which react with the amino groups of the antibody molecule. A considerable amount of antibody activity is lost this way, and the effective activity is only about 10-15 7%. Even though most antibody immobilizationmethods involve random coupling
Biotechnol. Rug., 1992, Vol. 8, No. 4
Oxldized antibody
271
3 ;I
(a)
1
+
+y-
Y-I
H +
H20
P
(b)
to a functional group on the immobilizing matrix. This method has the advantage of not having to immobilize the antibody. In addition, multienzyme immobilization on the same support can be accomplished easily by this method.
Detection of Acetylcholine The applicability of the monoclonal antibodies to prepare IMERs has been demonstrated by using these IMERs to detect acetylcholine in brain tissue samples. IMERa have been used with high-performance liquid chromatography (HPLC) and electrochemical detection to detect acetylcholine (ACh) and choline (Ch) (Potter et al., 1983; Beley et al., 1987). Acetylcholine, the main neurotransmitter in the cholinergic neurotransmitter system, has been found to be involved in various neurological disorders (Davis and Berger, 1979). Increasingly sensitive assay techniques are required to measure the low concentrations of ACh and Ch in brain. In our assay method after the separation of ACh from brain tissue extracts by reverse-phase HPLC, the effluent is allowed to pass through acetylcholinesterase (AChE) and choline oxidase (ChO) immobilized reactors. Hydrogen peroxide produced by the following reaction scheme is then detected amperometrically.
(d)
Figure 3. Schematic representation of orientated coupling of antibodies: (a) through the carbohydrate moiety, (b) immunochemical attachment, (c) Fab’ fragment, (d) protein A.
of amino groups at or near the antigen binding site, it is preferable to couple the antibody molecule through its Fc region by an immunochemical attachment (Mifflin et al., 1989)or through the carbohydrate moiety (Cress and Ngo, 1989) or via the -SH group of an Fab’ fragment (de Alwis and Wilson, 1987). The antibody molecule contains a carbohydrate moiety linked to the Fc region. Under mild conditions, sodium periodate can oxidize the vicinal hydroxyl groups of this carbohydrate to aldehyde groups. The antibody can then be linked to a hydrazine-activated support through these aldehyde groups. The IgG molecule can be reduced to Fab’ after the molecule is cleaved into F(ab’)2 by pepsin digestion. The monovalent Fab’ fragment can then be immobilized through the -SH group a t the hinge region (de Alwis and Wilson, 1987; Prisyazhony et al., 1988). Protein A and protein G are known to bind antibodies of the IgG class through the Fc region (Sisson and Castor, 1990; Janis and Regnier, 19891, and they also can be used for site-directed coupling. The antibody molecule can also be immobilized via another antibody by an immunochemical attachment. For example, mouse immunoglobulin can be immobilized by first coupling polyclonal goat antimouse antibodies to the support. A support prepared this way is considered to be a universal system since the bound antibody can be eluted by washing with a chaotropic agent and another mouse antibody can be immobilized onto the same support. It is also possible to bind two MoABs simultaneously to the same support. These oriented coupling methods are shown in Figure 3. Wimalasena and Wilson (1991)have carried out a more comprehensivestudy on the factors affecting the antigen binding capability of the immobilizedantibodies and antibody fragments. Their study shows the protocols for optimizing immobilized antibody performance. Enzyme immobilizationthrough bifunctional antibodies has also been suggested (Nolan and O’Kennedy, 1990). Half of the bifunctional antibody can be raised to a single epitope on the desired enzyme so that binding does not inhibit the enzyme. The other half can specifically bind
-
AChE
ACh
choline (Ch) + CH,COO-
-
ChO
Ch
betaine
+ 2H,O,
Performance of the IMERs depends on the properties of both of the immobilized enzymes. Since AChE is more durable than ChO, it can be immobilized without a loss of activity. ChO is very sensitive to its environment, and significant losses on immobilization are quite common. The two enzymes used in these experiments, ChO and AChE, have very different properties. ChO is a monomer with an approximate molecular weight of 72 OOO determined by gel filtration (Ohta-Fukuyama et al., 1980)and which contains an FAD group as the active site. AChE exists as a dimer usually containing two or three tetrameric seta of catalytic subunits (MacPhee-Quigley et al., 1986). AChE has a very high turnover number (Zubay, 1986)and has a high specific activity of about 1000 units/mg. ChO, on the other hand, has a very low intrinsic turnover rate and specific activity ( 10units/mg) as compared to AChE. The enzymes were coupled to the support material by using the methods shown in Figure 2 in order to prepare the AChE and ChO IMERs. No inhibition of the enzyme activity was observed when an excess of each antibody was mixed in solution with its respective enzyme antigen. This suggeststhat the antibodies do not inhibit the enzyme activity. Antibodies to both ChO and AChE were immobilized on the support as described above. Before the introduction of the enzymes, the column was saturated with bovine serum albumin to prevent the nonspecific adsorptionof the enzymes. The enzymeswere immobilized by injecting an aliquot of enzyme solution in carrier buffer. The bound and free fractions were evaluated for different injection volumes and different enzyme concentrations at different flow rates to establish an optimum concentration of enzyme and injection volume to get the maximum enzyme loading on the column. In this way, the enzyme loading on the column can carefully be controlled as desired and the specific activity of the enzyme on the column can be known. Performance of these IMERs was then evaluated individually in the FIA mode (HPLC/IMER
-
Biotechnol. Prog., 1992, Vol.
272
a, No. 4
Table I. Comparison of Immobilization Methods method detection response4 limit (M) (nA) direct direct 10-4 spacer 10-7 5.1 direct 10-8 13.6 avidin-biotin avidin-biotin 10-9 72.4 avidin-biotin avidin-biotinb antibody antibody 10-8 10.9 a Current response to 200 pmol of ACh. *Both enzymes are immobilized on the same support. (Reprinted with permission from Anal. Chem. 1990, 52, 402. Copyright 1990 American Chemical Society.)
Table 11. Performance of ChO Supports
immobilization method direct spacer avidin-biotin avidin-biotin both enzymes antibody antibody
loading of ChO (mg/mL of support) 1.2 (5.5)" 1.0 (4.6)" 2.5 (11.5)" 2.5 (11.5)" 0.02 (0.09)" 0.02 (0.09)"
%
uptake of ChO 85
81 91 84 100 100
immobilized enzyme percent activity activityb (units) 0.32 0.015 0.23 6.2 36.3 3.8 32.1 3.1 0.023 0.018
25.6 20.0
Units of enzyme. Percent of activity on the support relative to that of the soluble enzyme. Antibody immobilized through goat anti-mousespecific antibody (Reprinted with permission from Anal. Chem. 1990,62,402. Copyright 1990 American Chemical Society.)
system without the reverse-phase LC column). These results are tabulated in Table I. Direct immobilization of ChO yielded poor results. It was found that the activity of AChE was not significantly affected by the coupling except in the case of the coupling through a spacer arm, where AChE activity was somewhat reduced. The enzymes retained the highest activity when they were coupled through avidin-biotin linkages. The sensitivity was further enhanced by co-immobilizing the biotinylated enzymes to the same avidin-bound support material. This observation is in agreement with the reported literature (Mosbach and Mattiasson, 1976),where it has been shown that co-immobilization of two enzymes on the same support causes enhanced efficiency. They explained this observation as due to the availability of increased local concentration of the product from the first enzyme around the second enzyme in the reaction sequence. This is possible when the enzymes are coimmobilized on the same particle because the enzymes are in close proximity of each other. Therefore, the second enzyme will catalyze its reaction more efficiently and increase the overall reaction rate leading to the enhanced sensitivity in the assay system. When ChO antibody was bound to the support and used to immobilize ChO, it was found that the enzyme binds to the antibody effectively despite the fact that it has an apparently lower binding affinity than the AChE antibody. Our studies on these two IMERs showed that the activity of the ChO column governs the sensitivity and the detection limit of the assay. Therefore, a thorough study has been done to evaluate the ChO activity on the support after immobilization. These results are summarized in Table 11. The data in Table I1 show that when ChO is immobilized through the avidin-biotin linkage, it retains the highest activity. Enzyme activity on the supports which have ChO immobilized through the antibody is considerably higher than with the other methods but lower than with the avidin-biotin method. This can be attributed to the random coupling of the antibody which leads to a low number of
ii
eChEantibody
