186,14-18
ANALYTICALBIOCHEMISTRY
(1990)
Application of a Fusion Protein, Metapyrocatechase/ Protein A, to an Enzyme Immunoassay Eiry Kobatake,
Yuji Nishimori,
Yoshihito
Ikariyama,
Masuo Aizawa,
and Seishi Kate*
Department of Bioengineering, Tokyo Institute of Technology, 2-12-1, Ookayama, Meguro-ku, Tokyo 152, Japan, and *Genetic Engineering Section, Sagami Chemical Research Center, Nishi-ohnuma, Kanagawa 229, Japan
Received
November
1,1989
A fusion protein of metapyrocatechase and protein A was genetically produced for demonstration of effective conjugation of an enzyme with a binding protein employed in enzyme immunoassay. Plasmid pMPRA3, constructed by inserting the protein A gene into a plasmid pMK12 vector derived directly from the structural gene of metapyrocatechase, was expressed in Escherichia coli. The resulting fusion protein was shown to have promising properties for use in enzyme immunoassays due to the specific binding of the protein A moiety to the Fc portion of immunoglobulin G and to the high amplification of enzyme. Bovine serum albumin, a model antigen, was successfully determined in the concentration range from 1 X 10e3 to 1 X lo-’ g/ml. 0 1990 Academic
Press,
Inc.
Enzyme immunoassay has become an indispensable analytical method for the determination of trace substances especially in clinical laboratories due to the outstanding amplification effect of enzyme (1,2). In this technique, enzyme-labeled antigen and enzyme-labeled antibody are required for competitive and sandwich immunoassays, respectively. Enzyme labeling has been performed by covalent binding between an enzyme and an antibody/antigen (3,4). However, covalent preparation of a conjugate with a bifunctional agent generally results in the production of a complicated conjugate conglomerate; i.e., enzyme labeling hardly generates a conjugate of a 1:l binding ratio, which occasionally invites steric hindrance in an antigen-antibody reaction, particularly in the binding between a hapten and an antibody (5). Despite the random covalent coupling and the resulting steric hindrance of the enzyme in immunoreaction, enzyme immunoassay has become a prevailing sen-
sitive analytical technique due to the high catalytic activity of the enzyme to be employed (6,7). The advantage is most important in the enzyme immunosensor, since the demonstration of enzyme activity at a localized area on the sensor is required for the sensitive, rapid determination of an antigen (8,9). In the enzyme immunosensors, we used a Clark-type oxygen electrode as the amperometric transducer whose surface is covered with an antigen (or antibody)-bound membrane. Therefore, the oxygen-consuming or oxygen-producing enzyme was used as the labeling enzyme (8,9). In recent years, much attention has focused on genetically fused proteins because of the easy stoichiometric control of the fused protein. This technique is especially important for the mass production of an enzyme-antigen/binding protein conjugate with lesshindrance, since site-specific conjugation seems to be carried out by recombinant DNA techniques (10,ll). As staphylococcal protein A binds to the Fc region of the immunoglobulin of several mammalian species, the binding protein of the enzyme-bound form has been employed as a useful immunochemical reagent in enzyme immunoassay (12-14). Metapyrocatechase (catechol 2,3-dioxygenase, EC 1.13.11.2) is an oxygen-consuming enzyme, which catalyzes the oxygenation of catechol to cu-hydroxymuconic t-semialdehyde (1516). The enzyme seemsto be advantageous in immunoassay when it is used as a labeling enzyme, because the yellowish product is easily detected by a conventional spectrophotometer. In addition, the oxygen-consuming enzyme seemsto be applicable as an enzyme label in an amperometric-type immunosensor by taking advantage of the high catalytic activity of the enzyme in oxygen consumption (17). In this paper, an approach to the genetic production of metapyrocatechase-protein A and the resulting fun-
14 All
Copyright 0 1990 rights of reproduction
0003-2697/90 $3.00 by Academic Press, Inc. in any form reserved.
APPLICATION
damental properties are described enzyme immunoassay. MATERIALS
AND
OF
A FUSION
PROTEIN
from the viewpoint
of
METHODS
Materials. The IgG for BSA’ was purchased from Organ0 Teknika Corp. (West Chester, PA) and bovine serum albumin (fraction V, BSA) was purchased from Sigma (St. Louis, MO). IgG-bound Sepharose was obtained from Pharmacia (Uppsala, Sweden). All other chemicals and reagents were from Sigma and Kanto Chemicals (Tokyo, Japan). Plastic plates for solid-phase immunoassay were purchased from Becton-Dickinson & Co. (Lincoln Park, NJ). Plasmids pUC19, pKK223-3, and pRIT5 were purchased from Pharmacia. The plasmid pSLMK1, encoding the metapyrocatechase gene, was described in a previous paper (18). Plasmid construction. The metapyrocatechase fusion vector, pMK12, was constructed as follows. A tat promoter, a terminator, and an upstream region of the p-lactamase gene were prepared from pKK223-3. On the other hand, a downstream region of the p-lactamase gene was obtained from pUC19. An SD sequence and the structural gene of metapyrocatechase were prepared from pSLMK1. The metapyrocatechase/protein A gene fusion vector, pMPRA3, was obtained by inserting the fragment of the protein A gene, isolated from the plasmid pRIT5, into a SmaI site of pMK12. Fusion protein preparation. The preparation of the metapyrocatechase/protein A hybrid was carried out as follows. The transformed Escherichia coli cells were cultured in LB broth in the presence of ampicillin (50 pg/ ml) at 37°C with shaking for 6 h and harvested by centrifugation at 4000g for 10 min at 4°C. The cells were washed and resuspended in Tris-acetate buffer (50 mM, pH 7.5) containing 10% acetone and finally lysed by treating with lysozyme (100 pg/ml) for 30 min. The lysate was centrifuged at 5000g for 30 min to remove cell debris, and the supernatant was used for further analytical purpose. Assessment of specificity of the fusion protein. Nonspecific adsorption of the bifunctional protein on the solid-phase matrix was assessedwith and without an agent for blocking surface active points. Before the adsorption of the fusion protein, the surface of the wells was blocked with a protein such as skim milk, bovine serum albumin, or gelatin. Metapyrocatechase activity retained on the solid phase was determined after the titer wells were rinsed. The specific adsorption of the fusion protein to the primary antibody was evaluated with 1 Abbreviations used: BSA, bovine serum albumin; SD, Shine-Dalgarno; SDS-PAGE, sodium dodecyl phosphate-polyacrylamide gel electrophoresis; ELISA, enzyme-linked immunoadsorbent assay.
TO
AN
ENZYME
IMMUNOASSAY
15
and without a blocking agent. IgG-adsorbed titer wells were coated with one of the blocking proteins and then reacted with the fusion protein. The results were compared with those obtained for the protein-noncoated IgG-adsorbed wells by determining the pyrocatechase activity of the hybrid protein. Immunoassay procedure. The typical assay procedure was as follows: Anti-BSA antibody was dissolved in TBS buffer (20 mM Tris, 500 mM NaCl, pH 7.5) to a final concentration of 1 mg/ml, and an aliquot (500 ~1) was added to each well of a 24-well microtiter plate. The antibody was adsorbed overnight at 4°C. The solution in each well was discarded and the plates were washed three times with TBS. After washing, each well was incubated with 600 ~1 of 1% skim milk in TBS buffer at 30°C for 1 h to block any remaining free binding sites. The wells were then washed three times with TBS and a series of BSA solutions of different concentrations were added to undergo immunoreaction with the primary antibody at 30°C for 2 h. After washing, 500 ~1 of the secondary anti-BSA antibody was reacted for 2 h at 30°C. The wells were then washed three times with TBS and the crude extract (500 ~1) containing the fusion protein, diluted 20 times with KPi buffer (50 mM potassium phosphate, pH 7.5), was added to each well. After washing, the bound fusion protein was determined by adding 500 ~1 of 0.3 mM catechol solution. After 10 min of incubation at room temperature, the reaction was terminated by the addition of 50 ~1 of inhibitor solution (0.1 M of l,lO-phenanthroline) to each well. The absorbance change of each well at 375 nm was measured by a Jusco uv spectrophotometer (Tokyo, Japan). RESULTS
Plasmid construction. Figure 1 shows a schematic drawing of the plasmid, pMPRA3, containing the fused gene coding for metapyrocatechase and protein A. The plasmid was designed for the C-terminal of metapyrocatechase to be fused to the N-terminal of protein A in the resulting protein. The binding ratio of these two proteins was designed to be 1:l. Characterization of the fusion protein. As this plasmid possessesa tat promoter and an SD sequence of metapyrocatechase, the E. coli transformed by pMPRA3 was able to express large amounts of the fusion protein. Densitometric analysis of the crude extract from these transformed cells revealed that the extract contained over 70% of the target fusion protein; therefore, the crude extract was employed for the following experimentation without further purification. From 1 liter of culture, approximately 400 mg of the fusion protein was obtained. The protein hybrid was stored at -80°C and a required amount was thawed when needed. Degradation and decreases in enzymatic and binding activities were
16
FIG. 1. gene-fused coding for restriction Dalgarno cated. The
KOBATAKE
Schematic drawing of the metapyrocatechase/protein A plasmid, pMPRA3. Boxes show the positions of the genes metapyrocatechase (MPC) and p-lactamase (AMP’). Some sites, replication origin (Ori), tat promoter (tat), Shinesequence (SD), and terminator (rmBT,T2) are also indiarrows indicate the orientation of genes.
scarcely observed after 1 year of storage at -80°C. The fundamental characterization of the fusion protein was performed as follows. Human IgG covalently coupled to Sepharose was incubated with the fusion protein for 1 h and washed thoroughly with TBS buffer. When the catechol solution (0.3 mM) was added, the Sepharose gel turned a yellowish color, which indicates that the fusion protein not only possesses enzymatic activity but also is capable of binding to IgG. The molecular weight of the fusion protein was ca. 65 kDa, which was confirmed by SDS-PAGE analysis. This result indicates that the bifunctional protein forms a 1:l fusion ratio between metapyrocatechase and protein A. The estimated molecular weight of 65,000 was in good agreement with the fact that the fusion protein is combined with metapyrocatechase with a molecular weight is 35,000 and protein A with molecular weight 30,000. In addition, it was confirmed by Western blotting analysis that the fusion protein possesses binding properties to rabbit IgG. From these results, the feasibility of using the bifunctional protein in enzyme immunoassay was clearly demonstrated. The K,,, value of the fusion protein for catechol was 1.3 X 10m5, and the molecular activity of the hybrid protein was ca. 1.6 X lo4 mol/min at an optimum pH of 6.5. One of Assessment of specificity of the fusion protein. the main factors for increasing the sensitivity of the sandwich assay is to keep the nonspecific adsorption of the labeled antibody to the solid-phase matrix as low as possible. Generally, when the molecular weight of protein is larger, the nonspecific adsorption is more significant (19). Since chemical conjugation of the enzyme with the antibody easily forms a complicated conglomerate of large molecular weight, the conglomerate often
ET
AL.
suffers from nonspecific adsorption on the solid-phase matrix. However, by taking advantage of the gene fusion technique, the molecular weight of the compound can be strictly controlled through formation of a fusion protein with a binding ratio of 1:l. In order to assess the nonspecific adsorption of the fusion protein to the solid-phase matrix, the fusion protein was directly added to the wells. Before the adsorption of the fusion protein, the surfaces of the wells were coated for 1 h with 1% skim milk, 1% BSA, and 3% gelatin. The effect of surface coating was compared with the effect of noncoating. After a thorough washing, the plastic well with the fusion protein was incubated for 2 h, and then 500 ~1 of catechol solution (0.3 mM) was added. Metapyrocatechase activity was scarcely observed in each well when the well surface was coated with one of the above proteins. The result obtained is shown in Table 1. Among the protein coatings, milk showed the most remarkable effect in inhibiting nonspecific adsorption. Also, on the noncoated immunoplate, the fusion protein showed little adsorption. To evaluate the adsorption of the fusion protein to the primary antibody, the fusion protein was added to the primary antibody-bound well. The nonprotein-coated well on which surface IgG was adsorbed beforehand was employed for the assessment of protein coating. The surface of each well was then coated with either 1% skim milk or 3% gelatin. After incubation of the fusion protein for 2 h, catechol solution (0.3 mM) was added. A little adsorption of the fusion protein on the well surface occurred when the well was not coated with protein. However, when the well surface was coated with milk or gelatin, nonspecific adsorption was almost negligible. This result shows that the protein A moiety of the fusion protein does not bind to the Fc region of the primary antibody adsorbed on the solid-phase matrix, as long as the surface is coated with skim milk or gelatin. Almost all of the hydrophobic Fc portions of IgG molecules are expected to bind to the plastic well; therefore, the fusion TABLE Effect
of Protein
Protein
coating
None Milk (1%) BSA (1%) Gelatin (3%)
1
Coating on Nonspecific of the Fusion Protein
Adsorption
Relative absorbance at 375 nm (W) 100 60 70 80
Note. Plastic wells were coated with blocking protein to investigate the nonspecific adsorption of the hybrid protein. Nonspecifically adsorbed metapyrocatechase-protein A was determined by its enzymatic activity.
APPLICATION
OF
A FUSION
PROTEIN
TO
AN
ENZYME
17
IMMUNOASSAY
100 -
01
0, 01
0. 1
0
1
30
Time
Concentration of primaryantibody(mdmli FIG. 2. Effect of the concentration of the primary anti-BSA antibody on solid-phase immunoassay. The concentration of the primary antibody adsorbed to the solid phase was varied. After the incubation with BSA (10 wg/ml), the secondary antibody, and the fusion protein for 2 h at 3O”C, respectively and successively, 300 pM catechol solution was added. The cu-hydroxymucoic c-semialdehyde produced in each well was measured at 375 nm after 10 min of enzymatic reaction.
protein does not seem to bind to IgG molecules. However, when the IgG-coated plastic surface was not coated with milk or gelatin, the fusion protein bound to the well surface probably due to the interaction between the Fc portion and protein A. This result indicates that some of the IgG molecules adsorb to the well surface with their Fab portions. Thus, the Fc region recognized by protein A seems to be effectively blocked by the protein coating. Recently, as a substitute material for BSA or gelatin, skim milk has been used as a blocking agent (20). Also in this case, milk coating was shown to be a very effective way of preventing nonspecific protein adsorption. Optimization ofprimary IgG concentration. The concentration of the primary anti-BSA antibody was varied for the determination of the optimum conditions for solid-phase immunoassay. The concentration of BSA selected was 10 pg/ml, and all the other procedures were performed as described under Materials and Methods. As shown in Fig. 2, when the concentration of the primary antibody was 1 mg/ml, the enzymatic activity was drastically increased. The optimal reaction time for the binding reaction between the secondary antibody and the protein A hybrid was investigated. After incubation for a given time, the solution of the fusion protein in each well was discarded, and 0.3 mM catechol solution was then added. Each enzyme reaction was terminated 10 min after the binding of the reaction, and the change in absorbance was measured at 375 nm. The change in absorbance was normalized and plotted to the reaction time in Fig. 3, where it can be seen that the binding of
60
90
120
(min)
FIG. 3. Time dependence of the binding reaction between the secondary antibody and the fusion protein. After the incubation of the fusion protein for 30, 60, and 120 min, the solution in each well was discarded, and 300 pM catechol solution was added. The change in absorbance was measured at 375 nm after enzymatic reaction for 10 min. BSA concentrations were 0 (U), 10m4 (w), lo-’ (O), and 1 (0) mg/ml.
the hybrid protein increased remarkably at 2 h of incubation. It is considered that the efficiency of the binding of protein A with IgG depends on the concentrations of the proteins concerned. In a highly sensitive immunoas-
0 $
BSA
concentration
(g/ml)
FIG. 4. Standard curve for BSA. Anti-BSA antibody-adsorbed wells were incubated with various concentrations of BSA at 30°C for 2 h and then washed with TBS buffer, following which the secondary antiBSA antibody was incubated at 30°C for 2 h. After washing, the fusion protein that reacted with the secondary antibody underwent binding at 30°C for 2 h. After thorough washing, the activity of metapyrocatechase bound to the solid-phase surface via protein A-IgG binding was determined in the presence of 300 pM catechol. The change in absorbance at 375 nm was normalized in such a way that absorbance of 1.0 = 100%. The details are described under Materials and Methods.
18
KOBATAKE
say, a long incubation time is necessary for a high magnitude of amplification, especially when an antibody concentration is very low. Despite the very small amounts of fusion protein on the secondary antibody, only 10 min were required for the colorization reaction of this enzyme. Due to the high catalytic activity of metapyrocatechase, the colorization took less time than the typical labeling of enzymes such as /3-galactosidase and alkaline phosphatase. In addition, this enzyme has a rather small K,,, value (1.3 X 10A5) compared with those of other typical enzymes for enzyme immunoassay. ELBA plate assay. Finally, sandwich immunoassay of BSA, a model antigen, was performed under the optimized conditions as clarified above. A series of BSA solutions was incubated in the primary antibody-adsorbed wells. After an incubation with the secondary antibody, the binding reaction between the secondary antibody and the metapyrocatechase/protein A fusion protein was carried out. The enzymatic activity bound on the titer well was determined in the presence of 300 PM catechol for 10 min at room temperature. Normalized absorbance at 375 nm was plotted against the logarithmic BSA concentration. The standard curve for BSA is shown in Fig. 4. In the concentration range 10e3 to 10e7 g/ml, BSA was determined by taking the bifunctional fusion protein. The detection limit of this assay was 0.7 pmol of BSA. DISCUSSION
Metapyrocatechase/protein A fusion protein is very useful for the detection of trace amounts of antigen in a very simple and rapid procedure. The approach of gene fusion described here offers a promising alternative to chemical coupling, especially to enzyme labeling with a binding protein. We have also successfully demonstrated that genetic conjugation is very useful for stoichiometric controlling of the binding ratio of the two proteins. In the preparation of the enzyme-labeled binding protein by chemical conjugation, complicated conglomerate formation is unavoidable. On the other hand, the binding ratio can be strictly controlled in genetic conjugation. The nonspecific adsorption of the fusion protein on the solid-phase surface was very little. In addition, the fusion protein adsorption was repressed to a negligible level by blocking with a protein, especially with skim milk. The hybrid protein showed specific binding to the
ET
AL.
Fc moiety of IgG, which was confirmed by the solidphase immunometric assay and Western blotting. With protein A as a binding protein, the fusion protein can be used successfully as a universal immunological reagent in an enzyme immunoassay. In addition, as metapyrocatechase consumes oxygen, the fusion protein can be applied to an amplifying element in oxygen electrode-based immunosensors, since the protein is expected to cause a drastic change in oxygen concentration at a localized area of an oxygen electrode. REFERENCES 1. Engvall, E. (1980) in Methods in Enzymology (Van Vunakis, H., and Langone, J. J., Eds.), Vol. 70, pp. 419-439, Academic Press, San Diego, CA. 2. Van Weemen, B. K., and Schuurs, A. H. W. M. (1971) FEBS Z&t. 15,232-236. 3. Feldman, G., Druet, P., Bignon, Immunoenzymatic Techniques, E. (1973) J. Biochem. 4. Ishikawa,
J., and Avrameas, North-Holland, 73,1319-1321.
S. (Eds.) Amsterdam.
(1976)
5. Ikariyama, Y., Furuki, M., and Aizawa, M. (1985) Anal. Chem. 496-500. 6. Ishikawa, E., Kawai, T., and Miyai, K. (Eds.) (1981) Enzyme munoassay, Igaku-shoin, Tokyo. 7. Kaplan, L. A., and Pesce, A. J. (Eds.) tives to Radioimmunoassay, Dekker,
(1981) Nonisotopic New York.
57, Im-
Alterna-
8. Aizawa, M., Morioka, A., Matsuoka, H., Suzuki, S., Nagamura, Y., Shinohara, R., and Ishiguro, I. (1976) J. Solid-Phase Biochem. 1, 319-328. M., Morioka, A., Suzuki, S., and Nagamura, Y. (1979) 9. Aizawa, Anal. Biochem. 94,22-28. 10. Uhlen, M., Nilsson, B., Guss, M., Gatenbeck, L. (1983) Gene 23,369-378. A., Mecklenburg, M., Meussdoerffer, 11. Peterhans, K. (1987) Anal. Biochem. 163,470-475. 12. Goding, 13. Kessler, 14. O’Keefe,
J. W. (1978) J. Zmmunol. S. W. (1976) J. Zmmunol. E., and Bennett,
V. (1980)
15. Nozaki, M., Ono, K., Nakazawa, (1968) J. Biol. Chem. 243,2682-2690.
Methods
S., and Philipson, F., and Mosbach, 20,241-253.
117,1482-1490. J. Biol.
Chem.
T., Katori,
16. Nozaki, M. (1970) in Methods in Enzymology White Tabor, C., Eds.), Vol. 17A, pp. 522-525, San Diego, CA. 17. Aizawa, M., Morioka, A., and Suzuki, S. (1980)
255,561-568.
S., and Hayashi, (Tabor, Academic Anal.
0.
H., and Press,
Chim.
Acta
115,61-67. 18. Ohmori, M., Narushima, H., Miki, T., Numao, N., and Kondo, K. (1988) Agric. Biol. Chem. 52,2823-2830. 19. Hamaguchi, Y., Yoshitake, S., Ishikawa, E., Endo, Y., and Ohtaki, S. (1979) J. Biochem. 85,1289-1300. 20. Johnson, D. A., Gautsch, J. W., Sportsman, J. R., and Elder, J. H. (1984) Gene Anal. Tech. 1,3-S.
ANALYTICALBIOCHEMISTRY
(1990)
Application of a Fusion Protein, Metapyrocatechase/ Protein A, to an Enzyme Immunoassay Eiry Kobatake,
Yuji Nishimori,
Yoshihito
Ikariyama,
Masuo Aizawa,
and Seishi Kate*
Department of Bioengineering, Tokyo Institute of Technology, 2-12-1, Ookayama, Meguro-ku, Tokyo 152, Japan, and *Genetic Engineering Section, Sagami Chemical Research Center, Nishi-ohnuma, Kanagawa 229, Japan
Received
November
1,1989
A fusion protein of metapyrocatechase and protein A was genetically produced for demonstration of effective conjugation of an enzyme with a binding protein employed in enzyme immunoassay. Plasmid pMPRA3, constructed by inserting the protein A gene into a plasmid pMK12 vector derived directly from the structural gene of metapyrocatechase, was expressed in Escherichia coli. The resulting fusion protein was shown to have promising properties for use in enzyme immunoassays due to the specific binding of the protein A moiety to the Fc portion of immunoglobulin G and to the high amplification of enzyme. Bovine serum albumin, a model antigen, was successfully determined in the concentration range from 1 X 10e3 to 1 X lo-’ g/ml. 0 1990 Academic
Press,
Inc.
Enzyme immunoassay has become an indispensable analytical method for the determination of trace substances especially in clinical laboratories due to the outstanding amplification effect of enzyme (1,2). In this technique, enzyme-labeled antigen and enzyme-labeled antibody are required for competitive and sandwich immunoassays, respectively. Enzyme labeling has been performed by covalent binding between an enzyme and an antibody/antigen (3,4). However, covalent preparation of a conjugate with a bifunctional agent generally results in the production of a complicated conjugate conglomerate; i.e., enzyme labeling hardly generates a conjugate of a 1:l binding ratio, which occasionally invites steric hindrance in an antigen-antibody reaction, particularly in the binding between a hapten and an antibody (5). Despite the random covalent coupling and the resulting steric hindrance of the enzyme in immunoreaction, enzyme immunoassay has become a prevailing sen-
sitive analytical technique due to the high catalytic activity of the enzyme to be employed (6,7). The advantage is most important in the enzyme immunosensor, since the demonstration of enzyme activity at a localized area on the sensor is required for the sensitive, rapid determination of an antigen (8,9). In the enzyme immunosensors, we used a Clark-type oxygen electrode as the amperometric transducer whose surface is covered with an antigen (or antibody)-bound membrane. Therefore, the oxygen-consuming or oxygen-producing enzyme was used as the labeling enzyme (8,9). In recent years, much attention has focused on genetically fused proteins because of the easy stoichiometric control of the fused protein. This technique is especially important for the mass production of an enzyme-antigen/binding protein conjugate with lesshindrance, since site-specific conjugation seems to be carried out by recombinant DNA techniques (10,ll). As staphylococcal protein A binds to the Fc region of the immunoglobulin of several mammalian species, the binding protein of the enzyme-bound form has been employed as a useful immunochemical reagent in enzyme immunoassay (12-14). Metapyrocatechase (catechol 2,3-dioxygenase, EC 1.13.11.2) is an oxygen-consuming enzyme, which catalyzes the oxygenation of catechol to cu-hydroxymuconic t-semialdehyde (1516). The enzyme seemsto be advantageous in immunoassay when it is used as a labeling enzyme, because the yellowish product is easily detected by a conventional spectrophotometer. In addition, the oxygen-consuming enzyme seemsto be applicable as an enzyme label in an amperometric-type immunosensor by taking advantage of the high catalytic activity of the enzyme in oxygen consumption (17). In this paper, an approach to the genetic production of metapyrocatechase-protein A and the resulting fun-
14 All
Copyright 0 1990 rights of reproduction
0003-2697/90 $3.00 by Academic Press, Inc. in any form reserved.
APPLICATION
damental properties are described enzyme immunoassay. MATERIALS
AND
OF
A FUSION
PROTEIN
from the viewpoint
of
METHODS
Materials. The IgG for BSA’ was purchased from Organ0 Teknika Corp. (West Chester, PA) and bovine serum albumin (fraction V, BSA) was purchased from Sigma (St. Louis, MO). IgG-bound Sepharose was obtained from Pharmacia (Uppsala, Sweden). All other chemicals and reagents were from Sigma and Kanto Chemicals (Tokyo, Japan). Plastic plates for solid-phase immunoassay were purchased from Becton-Dickinson & Co. (Lincoln Park, NJ). Plasmids pUC19, pKK223-3, and pRIT5 were purchased from Pharmacia. The plasmid pSLMK1, encoding the metapyrocatechase gene, was described in a previous paper (18). Plasmid construction. The metapyrocatechase fusion vector, pMK12, was constructed as follows. A tat promoter, a terminator, and an upstream region of the p-lactamase gene were prepared from pKK223-3. On the other hand, a downstream region of the p-lactamase gene was obtained from pUC19. An SD sequence and the structural gene of metapyrocatechase were prepared from pSLMK1. The metapyrocatechase/protein A gene fusion vector, pMPRA3, was obtained by inserting the fragment of the protein A gene, isolated from the plasmid pRIT5, into a SmaI site of pMK12. Fusion protein preparation. The preparation of the metapyrocatechase/protein A hybrid was carried out as follows. The transformed Escherichia coli cells were cultured in LB broth in the presence of ampicillin (50 pg/ ml) at 37°C with shaking for 6 h and harvested by centrifugation at 4000g for 10 min at 4°C. The cells were washed and resuspended in Tris-acetate buffer (50 mM, pH 7.5) containing 10% acetone and finally lysed by treating with lysozyme (100 pg/ml) for 30 min. The lysate was centrifuged at 5000g for 30 min to remove cell debris, and the supernatant was used for further analytical purpose. Assessment of specificity of the fusion protein. Nonspecific adsorption of the bifunctional protein on the solid-phase matrix was assessedwith and without an agent for blocking surface active points. Before the adsorption of the fusion protein, the surface of the wells was blocked with a protein such as skim milk, bovine serum albumin, or gelatin. Metapyrocatechase activity retained on the solid phase was determined after the titer wells were rinsed. The specific adsorption of the fusion protein to the primary antibody was evaluated with 1 Abbreviations used: BSA, bovine serum albumin; SD, Shine-Dalgarno; SDS-PAGE, sodium dodecyl phosphate-polyacrylamide gel electrophoresis; ELISA, enzyme-linked immunoadsorbent assay.
TO
AN
ENZYME
IMMUNOASSAY
15
and without a blocking agent. IgG-adsorbed titer wells were coated with one of the blocking proteins and then reacted with the fusion protein. The results were compared with those obtained for the protein-noncoated IgG-adsorbed wells by determining the pyrocatechase activity of the hybrid protein. Immunoassay procedure. The typical assay procedure was as follows: Anti-BSA antibody was dissolved in TBS buffer (20 mM Tris, 500 mM NaCl, pH 7.5) to a final concentration of 1 mg/ml, and an aliquot (500 ~1) was added to each well of a 24-well microtiter plate. The antibody was adsorbed overnight at 4°C. The solution in each well was discarded and the plates were washed three times with TBS. After washing, each well was incubated with 600 ~1 of 1% skim milk in TBS buffer at 30°C for 1 h to block any remaining free binding sites. The wells were then washed three times with TBS and a series of BSA solutions of different concentrations were added to undergo immunoreaction with the primary antibody at 30°C for 2 h. After washing, 500 ~1 of the secondary anti-BSA antibody was reacted for 2 h at 30°C. The wells were then washed three times with TBS and the crude extract (500 ~1) containing the fusion protein, diluted 20 times with KPi buffer (50 mM potassium phosphate, pH 7.5), was added to each well. After washing, the bound fusion protein was determined by adding 500 ~1 of 0.3 mM catechol solution. After 10 min of incubation at room temperature, the reaction was terminated by the addition of 50 ~1 of inhibitor solution (0.1 M of l,lO-phenanthroline) to each well. The absorbance change of each well at 375 nm was measured by a Jusco uv spectrophotometer (Tokyo, Japan). RESULTS
Plasmid construction. Figure 1 shows a schematic drawing of the plasmid, pMPRA3, containing the fused gene coding for metapyrocatechase and protein A. The plasmid was designed for the C-terminal of metapyrocatechase to be fused to the N-terminal of protein A in the resulting protein. The binding ratio of these two proteins was designed to be 1:l. Characterization of the fusion protein. As this plasmid possessesa tat promoter and an SD sequence of metapyrocatechase, the E. coli transformed by pMPRA3 was able to express large amounts of the fusion protein. Densitometric analysis of the crude extract from these transformed cells revealed that the extract contained over 70% of the target fusion protein; therefore, the crude extract was employed for the following experimentation without further purification. From 1 liter of culture, approximately 400 mg of the fusion protein was obtained. The protein hybrid was stored at -80°C and a required amount was thawed when needed. Degradation and decreases in enzymatic and binding activities were
16
FIG. 1. gene-fused coding for restriction Dalgarno cated. The
KOBATAKE
Schematic drawing of the metapyrocatechase/protein A plasmid, pMPRA3. Boxes show the positions of the genes metapyrocatechase (MPC) and p-lactamase (AMP’). Some sites, replication origin (Ori), tat promoter (tat), Shinesequence (SD), and terminator (rmBT,T2) are also indiarrows indicate the orientation of genes.
scarcely observed after 1 year of storage at -80°C. The fundamental characterization of the fusion protein was performed as follows. Human IgG covalently coupled to Sepharose was incubated with the fusion protein for 1 h and washed thoroughly with TBS buffer. When the catechol solution (0.3 mM) was added, the Sepharose gel turned a yellowish color, which indicates that the fusion protein not only possesses enzymatic activity but also is capable of binding to IgG. The molecular weight of the fusion protein was ca. 65 kDa, which was confirmed by SDS-PAGE analysis. This result indicates that the bifunctional protein forms a 1:l fusion ratio between metapyrocatechase and protein A. The estimated molecular weight of 65,000 was in good agreement with the fact that the fusion protein is combined with metapyrocatechase with a molecular weight is 35,000 and protein A with molecular weight 30,000. In addition, it was confirmed by Western blotting analysis that the fusion protein possesses binding properties to rabbit IgG. From these results, the feasibility of using the bifunctional protein in enzyme immunoassay was clearly demonstrated. The K,,, value of the fusion protein for catechol was 1.3 X 10m5, and the molecular activity of the hybrid protein was ca. 1.6 X lo4 mol/min at an optimum pH of 6.5. One of Assessment of specificity of the fusion protein. the main factors for increasing the sensitivity of the sandwich assay is to keep the nonspecific adsorption of the labeled antibody to the solid-phase matrix as low as possible. Generally, when the molecular weight of protein is larger, the nonspecific adsorption is more significant (19). Since chemical conjugation of the enzyme with the antibody easily forms a complicated conglomerate of large molecular weight, the conglomerate often
ET
AL.
suffers from nonspecific adsorption on the solid-phase matrix. However, by taking advantage of the gene fusion technique, the molecular weight of the compound can be strictly controlled through formation of a fusion protein with a binding ratio of 1:l. In order to assess the nonspecific adsorption of the fusion protein to the solid-phase matrix, the fusion protein was directly added to the wells. Before the adsorption of the fusion protein, the surfaces of the wells were coated for 1 h with 1% skim milk, 1% BSA, and 3% gelatin. The effect of surface coating was compared with the effect of noncoating. After a thorough washing, the plastic well with the fusion protein was incubated for 2 h, and then 500 ~1 of catechol solution (0.3 mM) was added. Metapyrocatechase activity was scarcely observed in each well when the well surface was coated with one of the above proteins. The result obtained is shown in Table 1. Among the protein coatings, milk showed the most remarkable effect in inhibiting nonspecific adsorption. Also, on the noncoated immunoplate, the fusion protein showed little adsorption. To evaluate the adsorption of the fusion protein to the primary antibody, the fusion protein was added to the primary antibody-bound well. The nonprotein-coated well on which surface IgG was adsorbed beforehand was employed for the assessment of protein coating. The surface of each well was then coated with either 1% skim milk or 3% gelatin. After incubation of the fusion protein for 2 h, catechol solution (0.3 mM) was added. A little adsorption of the fusion protein on the well surface occurred when the well was not coated with protein. However, when the well surface was coated with milk or gelatin, nonspecific adsorption was almost negligible. This result shows that the protein A moiety of the fusion protein does not bind to the Fc region of the primary antibody adsorbed on the solid-phase matrix, as long as the surface is coated with skim milk or gelatin. Almost all of the hydrophobic Fc portions of IgG molecules are expected to bind to the plastic well; therefore, the fusion TABLE Effect
of Protein
Protein
coating
None Milk (1%) BSA (1%) Gelatin (3%)
1
Coating on Nonspecific of the Fusion Protein
Adsorption
Relative absorbance at 375 nm (W) 100 60 70 80
Note. Plastic wells were coated with blocking protein to investigate the nonspecific adsorption of the hybrid protein. Nonspecifically adsorbed metapyrocatechase-protein A was determined by its enzymatic activity.
APPLICATION
OF
A FUSION
PROTEIN
TO
AN
ENZYME
17
IMMUNOASSAY
100 -
01
0, 01
0. 1
0
1
30
Time
Concentration of primaryantibody(mdmli FIG. 2. Effect of the concentration of the primary anti-BSA antibody on solid-phase immunoassay. The concentration of the primary antibody adsorbed to the solid phase was varied. After the incubation with BSA (10 wg/ml), the secondary antibody, and the fusion protein for 2 h at 3O”C, respectively and successively, 300 pM catechol solution was added. The cu-hydroxymucoic c-semialdehyde produced in each well was measured at 375 nm after 10 min of enzymatic reaction.
protein does not seem to bind to IgG molecules. However, when the IgG-coated plastic surface was not coated with milk or gelatin, the fusion protein bound to the well surface probably due to the interaction between the Fc portion and protein A. This result indicates that some of the IgG molecules adsorb to the well surface with their Fab portions. Thus, the Fc region recognized by protein A seems to be effectively blocked by the protein coating. Recently, as a substitute material for BSA or gelatin, skim milk has been used as a blocking agent (20). Also in this case, milk coating was shown to be a very effective way of preventing nonspecific protein adsorption. Optimization ofprimary IgG concentration. The concentration of the primary anti-BSA antibody was varied for the determination of the optimum conditions for solid-phase immunoassay. The concentration of BSA selected was 10 pg/ml, and all the other procedures were performed as described under Materials and Methods. As shown in Fig. 2, when the concentration of the primary antibody was 1 mg/ml, the enzymatic activity was drastically increased. The optimal reaction time for the binding reaction between the secondary antibody and the protein A hybrid was investigated. After incubation for a given time, the solution of the fusion protein in each well was discarded, and 0.3 mM catechol solution was then added. Each enzyme reaction was terminated 10 min after the binding of the reaction, and the change in absorbance was measured at 375 nm. The change in absorbance was normalized and plotted to the reaction time in Fig. 3, where it can be seen that the binding of
60
90
120
(min)
FIG. 3. Time dependence of the binding reaction between the secondary antibody and the fusion protein. After the incubation of the fusion protein for 30, 60, and 120 min, the solution in each well was discarded, and 300 pM catechol solution was added. The change in absorbance was measured at 375 nm after enzymatic reaction for 10 min. BSA concentrations were 0 (U), 10m4 (w), lo-’ (O), and 1 (0) mg/ml.
the hybrid protein increased remarkably at 2 h of incubation. It is considered that the efficiency of the binding of protein A with IgG depends on the concentrations of the proteins concerned. In a highly sensitive immunoas-
0 $
BSA
concentration
(g/ml)
FIG. 4. Standard curve for BSA. Anti-BSA antibody-adsorbed wells were incubated with various concentrations of BSA at 30°C for 2 h and then washed with TBS buffer, following which the secondary antiBSA antibody was incubated at 30°C for 2 h. After washing, the fusion protein that reacted with the secondary antibody underwent binding at 30°C for 2 h. After thorough washing, the activity of metapyrocatechase bound to the solid-phase surface via protein A-IgG binding was determined in the presence of 300 pM catechol. The change in absorbance at 375 nm was normalized in such a way that absorbance of 1.0 = 100%. The details are described under Materials and Methods.
18
KOBATAKE
say, a long incubation time is necessary for a high magnitude of amplification, especially when an antibody concentration is very low. Despite the very small amounts of fusion protein on the secondary antibody, only 10 min were required for the colorization reaction of this enzyme. Due to the high catalytic activity of metapyrocatechase, the colorization took less time than the typical labeling of enzymes such as /3-galactosidase and alkaline phosphatase. In addition, this enzyme has a rather small K,,, value (1.3 X 10A5) compared with those of other typical enzymes for enzyme immunoassay. ELBA plate assay. Finally, sandwich immunoassay of BSA, a model antigen, was performed under the optimized conditions as clarified above. A series of BSA solutions was incubated in the primary antibody-adsorbed wells. After an incubation with the secondary antibody, the binding reaction between the secondary antibody and the metapyrocatechase/protein A fusion protein was carried out. The enzymatic activity bound on the titer well was determined in the presence of 300 PM catechol for 10 min at room temperature. Normalized absorbance at 375 nm was plotted against the logarithmic BSA concentration. The standard curve for BSA is shown in Fig. 4. In the concentration range 10e3 to 10e7 g/ml, BSA was determined by taking the bifunctional fusion protein. The detection limit of this assay was 0.7 pmol of BSA. DISCUSSION
Metapyrocatechase/protein A fusion protein is very useful for the detection of trace amounts of antigen in a very simple and rapid procedure. The approach of gene fusion described here offers a promising alternative to chemical coupling, especially to enzyme labeling with a binding protein. We have also successfully demonstrated that genetic conjugation is very useful for stoichiometric controlling of the binding ratio of the two proteins. In the preparation of the enzyme-labeled binding protein by chemical conjugation, complicated conglomerate formation is unavoidable. On the other hand, the binding ratio can be strictly controlled in genetic conjugation. The nonspecific adsorption of the fusion protein on the solid-phase surface was very little. In addition, the fusion protein adsorption was repressed to a negligible level by blocking with a protein, especially with skim milk. The hybrid protein showed specific binding to the
ET
AL.
Fc moiety of IgG, which was confirmed by the solidphase immunometric assay and Western blotting. With protein A as a binding protein, the fusion protein can be used successfully as a universal immunological reagent in an enzyme immunoassay. In addition, as metapyrocatechase consumes oxygen, the fusion protein can be applied to an amplifying element in oxygen electrode-based immunosensors, since the protein is expected to cause a drastic change in oxygen concentration at a localized area of an oxygen electrode. REFERENCES 1. Engvall, E. (1980) in Methods in Enzymology (Van Vunakis, H., and Langone, J. J., Eds.), Vol. 70, pp. 419-439, Academic Press, San Diego, CA. 2. Van Weemen, B. K., and Schuurs, A. H. W. M. (1971) FEBS Z&t. 15,232-236. 3. Feldman, G., Druet, P., Bignon, Immunoenzymatic Techniques, E. (1973) J. Biochem. 4. Ishikawa,
J., and Avrameas, North-Holland, 73,1319-1321.
S. (Eds.) Amsterdam.
(1976)
5. Ikariyama, Y., Furuki, M., and Aizawa, M. (1985) Anal. Chem. 496-500. 6. Ishikawa, E., Kawai, T., and Miyai, K. (Eds.) (1981) Enzyme munoassay, Igaku-shoin, Tokyo. 7. Kaplan, L. A., and Pesce, A. J. (Eds.) tives to Radioimmunoassay, Dekker,
(1981) Nonisotopic New York.
57, Im-
Alterna-
8. Aizawa, M., Morioka, A., Matsuoka, H., Suzuki, S., Nagamura, Y., Shinohara, R., and Ishiguro, I. (1976) J. Solid-Phase Biochem. 1, 319-328. M., Morioka, A., Suzuki, S., and Nagamura, Y. (1979) 9. Aizawa, Anal. Biochem. 94,22-28. 10. Uhlen, M., Nilsson, B., Guss, M., Gatenbeck, L. (1983) Gene 23,369-378. A., Mecklenburg, M., Meussdoerffer, 11. Peterhans, K. (1987) Anal. Biochem. 163,470-475. 12. Goding, 13. Kessler, 14. O’Keefe,
J. W. (1978) J. Zmmunol. S. W. (1976) J. Zmmunol. E., and Bennett,
V. (1980)
15. Nozaki, M., Ono, K., Nakazawa, (1968) J. Biol. Chem. 243,2682-2690.
Methods
S., and Philipson, F., and Mosbach, 20,241-253.
117,1482-1490. J. Biol.
Chem.
T., Katori,
16. Nozaki, M. (1970) in Methods in Enzymology White Tabor, C., Eds.), Vol. 17A, pp. 522-525, San Diego, CA. 17. Aizawa, M., Morioka, A., and Suzuki, S. (1980)
255,561-568.
S., and Hayashi, (Tabor, Academic Anal.
0.
H., and Press,
Chim.
Acta
115,61-67. 18. Ohmori, M., Narushima, H., Miki, T., Numao, N., and Kondo, K. (1988) Agric. Biol. Chem. 52,2823-2830. 19. Hamaguchi, Y., Yoshitake, S., Ishikawa, E., Endo, Y., and Ohtaki, S. (1979) J. Biochem. 85,1289-1300. 20. Johnson, D. A., Gautsch, J. W., Sportsman, J. R., and Elder, J. H. (1984) Gene Anal. Tech. 1,3-S.
