Journal of Immunological Methods, 131 (1990) 99-104 Elsevier
99
JIM 05627
Covalently linked peptides for enzyme-linked immunosorbent assay Jan Sonderg/ird-Andersen a, Edgar Lauritzen 2, Klaus Lind 1 and Arne H o l m 3 Research Center for Medical Biotechnology, 1Mycoplasma Laboratory/Neisseria Department, 2 H I V Laboratory/Rubella Department, Statens Seruminstitut, Amager Boulevard 80, DK-2300 Copenhagen S, Denmark, and 3 Chemical Laboratory II, The H. C. Orsted Institute, University of Copenhagen, Denmark (Received 6 October 1989, revised received 24 January 1990, accepted 29 January 1990)
A general method is described, by which synthetic peptides are covalently linked via their carboxyl group to microtiter plates (CovaLink) for enzyme-linked immunosorbent assay (ELISA). Plates were prepared by this method with an angiotensin II peptide and with an HIV-2 peptide and attachment detected by rabbit anti-angiotensin serum and with a positive serum from an HIV-2-infected patient, respectively, using the common ELISA procedure in the last steps. The method is simple to perform, it constitutes an alternative to the common ELISA method, and eliminates the risk of inadvertent loss of peptide during the procedure. The method is highly reproducible and has a high sensitivity. It may be used for either antigen or antibody detection. Key words: Immobilization; ELISA; Synthetic peptide; CovaLink; Angiotensin II; HIV-2
Introduction
Since the introduction of solid-phase peptide synthesis by Merrifield (1963), synthetic peptides have greatly facilitated both the study of antigenantibody interaction (Hodges et al., 1983; Geysen, 1985) and the mapping of antigenic determinants (Goudsmit, 1988), and they have a promising future in the field of synthetic vaccines (Arnon, 1987) and specific diagnostic tests for microorganisms (Gnann, 1987a). However, when working with synthetic peptides a major problem is the test system used for measuring their reactivity with antibodies. It is possible to coat synthetic peptides directly onto a microtiter plate, but it involves testing each
Correspondence to: J. S~nderg~trd-Andersen, Mycoplasma Lab., Neisseria Dept., Bygn. 7, Statens Seruminstitut, Amager Boulevard 80, DK-2300 Copenhagen S, Denmark.
peptide with a panel of buffers with different pH and ionic strength for optimal binding to the solid phase (Geerlings et al., 1988). Still the method does not work equally well for all peptides, and loss of peptides during the ELISA procedure cannot be excluded. Therefore, when synthetic peptides are applied directly to a microtiter plate and reacted with antibodies, it is difficult to verify if a negative result is due to lack of binding between peptide and antibody, or to a lack of primary attachment of the peptide to the surface. Radioactively labelled peptides have been used to test for attachment, but besides the drawbacks of working with radioactive materials, a considerable workload is involved when dealing with hundreds or thousands of peptides as described by some workers (Geysen, 1985). Another approach is to use peptides coupled to a carrier like bovine serum albumin (BSA) which is known to attach to the solid phase (Shirahama et al., 1985). Besides being time consuming, it is
0022-1759/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
100
difficult to control how the peptide is coupled to the carrier. Results may vary according to whether the coupling is to a terminal amino acid, to one or more of the side chains, or whether the coupling method itself has modified functional groups of the peptide (Briand et al., 1985). Synthesis of peptides directly onto the solid support has also been used (Geysen et al., 1984). This has obvious advantages in comparison with other methods, since peptides are coupled to the surface in a controlled manner. By this technique overlapping peptides may be synthesized in a single microtiter plate, and since peptides are covalently coupled to the surface, bound antibodies can be removed, and the peptides retested with a new set of antibodies. However, by this method the identity and purity of the peptides cannot be verified directly. The need for simple and reliable methods to covalently couple purified peptides to a surface in a controlled manner seems obvious. A method where peptides are covalently coupled to nitrocellulose has recently been described (Lauritzen et al., 1990). In the present report we wish to describe a general method by which peptides can be coupled covalently via the carboxyl group to microtiter plates (CovaLink, Nunc, Roskilde, Denmark).
Immunoplates Nunc Immuno Modules, CovaLink NH, cat. no. 478042, were used. These plates have been made from polystyrene plates by covalent attachment of the linker shown in Fig. 1 (5). A total activated surface area corresponding to 100 #1 O
O
N-OH
EDc
3
•
-OCOR
\b
a 1
2
4
CH3
5
Peptides The angiotensin II peptide, with the sequence H - A s p - A r g - V a l - T y r - I l e - H i s - P r o - P h e - O H , was purchased from Sigma, and the HIV-2 peptide with the sequence H-Leu-Asn-Ser-Trp-Gly-CysAla-Phe-Arg-Gln-Val-Cys-OH ( G n a n n et al., 1987b) was from Cambridge Research Biochemicals, London, U.K.
A n tisera Anti-angiotensin II antiserum production in rabbits was performed as described previously by coupling angiotensin to the thyroglobulin carrier (Sofreniev et al., 1978). Antiserum and prevaccination serum from rabbit K-529 was used. Human serum reactive against the HIV-2 peptide was from a confirmed HIV-2 seropositive patient (Poulsen et al., 1989). Normal human serum was from a healthy, anti-HIV-1 and anti-HIV-2 negative person.
Coupling of peptides to CovaLink plates
Materials and methods
RCOOH *
tilling volume, which again corresponds to approximately 1 cm 2, is available. The number of linking groups on the surface has been radioactively estimated to approximately 1014 g r o u p s / c m 2 corresponding to 1.67 x 10 -a° m o l / c m 2 (patent pending).
~
CH3
6
Fig. 1. Covalent coupling of peptide to CovaLink. EDC: 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide.
To 5 /tl of the HIV-2 peptide (5.0 m g / m l in H 2 0 ) were added 5 ~1 of 0.1 M aqueous N-hydroxysuccinimide (NHS) and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide, HC1 (EDC), both from Sigma. To 18/tl of the angiotensin II peptide (1000 # g / m l in H 2 0 ) were added 1 8 / d of N H S + EDC. After 30 min at room temperature the activated peptides were diluted from 20 to 0.001 /~g/ml in icecold 0.1 M carbonate buffer p H 8.6 and 100 /~1 applied per well of the CovaLink plates, which were then incubated for 30 min at 4 o C, adapted from an earlier coupling procedure (Cuatrecasas and Parikh, 1972). To test for the ability of the peptides to bind noncovalently to the plates, distilled water was added to the peptides instead of N H S + EDC in the activation step. As a control for unspecific binding of antibodies to the plates, N H S + EDC was added to distilled water instead of peptide in the activation step.
101
ELISA procedure The plates were washed four times in PBS p H 7.2, postcoated with PBS p H 7.2 with 5% ( w / v ) skimmed milk powder (PBS-M) for 40 min at r o o m temperature on a rocking table, washed four times and incubated rocking at 3 7 ° C for 1.5 h with antisera diluted in PBS-M. H u m a n anti-HIV2 serum and normal h u m a n serum were diluted from 1 / 5 0 to 1/400. Rabbit anti-angiotensin II serum and prevaccination serum were diluted 1 / 1 0 0 to 1/32,000 and 1/100 to 1/2000, respectively. Plates were washed four times and incubated rocking for 45 min at 3 7 ° C with peroxidase labeled anti-antibodies (P214 (anti-human IgG) and P217 (anti-rabbit Ig), both Dakopatts, Denmark) diluted 1/1000 in PBS-M. The plates were washed four times, and colour developed for 10 min with o-phenylenediamine (0.5 m g / m l ) in citrate buffer p H 5.0 with 0.05% H202. Color development was stopped with 150/~1 1 N H2SO 4. The plates were read on an I m m u n o R e a d e r (InterMed N J-2000) at 490 nm. In addition to concentrations of peptide and serum, the following parameters were varied during optimalization of the test: (a) the amount of N H S + E D C in the final solution added to the plates; (b) the duration of peptide activation with N H S + EDC; (c) the
3000]
,
activation and coupling of peptide by N H S and E D C alone; (d) the temperature at which peptide was coupled to the plates; (e) the duration of peptide coupling to the plates; (f) the type of coupling buffer; (g) the addition of Tween 20 to the washing buffer and PBS-M.
Results
Angiotensin H Checkerboard titrations of angiotensin II peptide against rabbit anti-angiotensin II serum showed, that 1 t t g / m l peptide (Fig. 2A) and a 1/1000 dilution of antiserum (Fig. 2B) gave an optimal distinction between covalently coupled peptide, and unspecifically bound peptide. The background from antiserum bound non-covalently to plates without peptide was minimal, and normal rabbit serum showed no reaction at all with the peptide. Each well contains 1014 functional groups or 1.67 × 10-1° mol on approximately 1 cm 2 corresponding to a filling volume of approximately 100 /~1 (see above). With an average molecular weight of the peptide of 1000, 0.167/~g/100/~1 of peptide would be needed to occupy all functional groups
A
2500I
~oo~~
250O
,
i
,
,
r 8
2ooc 1500
i 1000
5OO
10
5
1 0.1 Peptide conc. in j~g/ml
0.01
0.001
1/1-00 1/5-00 1/1000 112000 1/,~000 1/1~)00 1/1~000 1132000 Antibody dilution
Fig. 2. A: rabbit serum diluted 1/1000 and angiotensin II peptide in serial dilutions. B: angiotensin II peptide 1 #g/ml and serial dilutions of rabbit serum. © ©, rabbit anti-angiotensin II serum against covalently coupled peptide. [] O, rabbit anti-angiotensin II serum against non-covalently bound peptide. • •, rabbit anti-angiotensin II serum against immunoplates without peptide. • O, normal rabbit serum against covalently coupled peptide. OD: optical density. Standard deviation is indicated on A (n = 6).
102
in the well. The highest concentration of peptide used was 10/xg/ml, and at this concentration the dose-response curve seemed to flatten (Fig. 2A). Therefore a surplus of peptide was needed to occupy all functional sites in the well, but die non-covalent binding of peptide was high at this concentration. However, at our working dilution of 1/~g peptide/ml, not all functional sites would be occupied. When the duration of peptide activation by NHS + EDC exceeded 30 min, reactivity of the activated peptide dropped. Coupling of peptide to the CovaLink plates was optimal with 30 min at 4°C. If coupling of peptide was performed at room temperature a higher background from peptide binding non-covalently to the plates was observed. When the 0.1 M carbonate buffer pH 8.6 used for coupling of peptide to the plates was substituted with PBS pH 7.2 or carbonate buffer pH 9.6, a drop in the reactivity of the plates was observed. The activating solution (peptide and NHS + EDC) should be diluted preferably 100 times in carbonate buffer. Otherwise the higher concentration of NHS + EDC would result in increased background from unspecific binding of serum to the plates. It was observed, that peptide, that had not been activated with NHS + EDC, could bind noncovalently to the plates. The effect of Tween 20 on the noncovalently bound peptide was investigated. When the microtiter plates were treated with PBS pH 7.2 with 0.05-1% Tween 20 before coupling of peptide, the plates lost their reactivity. The reactivity observed with covalently bound peptide increased when 0.05% Tween 20 was added to the washing buffer and PBS-M, but the reactivity of non-covalently bound peptide increased equally. Tween 20, therefore, did not give a better distinction between covalently and non-covalently bound peptide. If Tween 20 was added only to the washing buffer applied just after coupling of peptide to the plates, the same general increase in reactivity was observed. After coupling of peptide to the plates, they may be washed, dried, and stored in a refrigerator overnight without loss of reactivity. Earlier results with Nunc immunoplates MaxiSorp showed an almost complete desorption of
2500
2ooo~
T
20
10
0
1
(~1
0.01
0001
0.0001
Peptide ¢on¢. in jJg/ml
Fig. 3. Serum from an HIV-2-infected individual diluted 1/100 and serial dilutions of HIV-2 peptide, o o, anti-HIV-2 serum against covalently coupled peptide. [] n, antiHIV-2 serum against non-covalently coupled peptide. • • , normal human serum against covalently coupled peptide. • • , anti-HIV-2 serum against immunoplates without peptide. OD: optical density. Standard deviation is indicated on the figure (n = 6).
angiotensin II and a C terminal extension of angiotensin II the decapeptide called angiotensin I from the plates during the ELISA procedure. These peptides were desorbed after having been coated to the plates overnight using the carbonate buffer pH 9.6. HIV-2
By checkerboard titrations of HIV-2 peptide and the serum from an HIV-2 infected individual, it was shown, that 10-20 /tg peptide/ml and a 1/100 dilution of serum gave the best distinction between covalently coupled and non-covalently bound peptide (Fig. 3). As for angiotensin II the addition of Tween 20 to the buffers increased the overall signal, but did not give a better distinction between covalently and non-covalently bound peptide.
Discussion
In this paper we have described a general method by which peptides containing a carboxylic
103 acid group, terminal or side chain, are coupled covalently to CovaLink microtiter plates. The secondary amino group, to which the peptide is coupled, is placed at the end of a spacer (Fig. 1 (5)). Besides the covalent linkage the spacer includes 11 carbon atoms and 3 O/NH groups and is relatively hydrophilic. If the spacer is assumed to be fully extended into the aqueous surroundings then the N H group is estimated to be approximately 2 nm from the polystyrene surface wall of the well. This, together with the specific manner in which peptides are coupled to the functional group, increases the probability that antibodies specific for the peptides will bind to them. If peptides were absorbed onto the surface in an unspecific way, as is the case with the commonly used methodology, the probability exists, that specific antibodies would be unable to bind because of steric hindrance or inability of the peptides to assume a conformation necessary for binding. With the present method we cannot exclude, of course, that highly lipophilic covalently bound peptides will still have a tendency to bind by hydrophobic forces to the polystyrene surface. We should also point out, that the presence of side chain carboxylic acid groups, due to the amino acids aspartic acid and glutamic acid, will compete with the terminal carboxyl acid group during covalent coupling. Therefore, even though all peptides will be covalently coupled to the CovaLink plates by our method, peptides that have one or more side chain carboxylic acid groups will probably be presented to antibodies in more than one way in the same well, and this should be taken into consideration when examining the results. To achieve optimal coupling of peptide to the CovaLink plates and optimal reaction with antiserum, certain steps in the procedure are important. Coupling of peptide (Fig. 1 (1)) is obtained by using w a t e r soluble 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC (3)) as coupling agent in the presence of N-hydroxysuccinimide (NHS (2)). This procedure has been used previously for activation and coupling of peptide segments, where absence of N-hydroxysuccinimide led to a partially racemized product, and excess of both this reagent and the water soluble carbodiimide lowered the yield (Sheppard, 1980). In the
present investigation we have observed that activation with EDC alone gives rise to a much lower covalent coupling (6) than in the presence of NHS. Clearly, in order to achieve an optimal formation of the activated, intermediate peptide N-hydroxysuccinimide ester (4), the N H S + EDC concentration must not be too low during the activation step. On the other hand it must not be too high in the solution when peptide is coupled to the plates, since we have observed, that this results in an increased background from antibodies binding unspecifically to the plates. Thus, high initial concentrations of peptide, N H S and EDC is used followed by dilution in carbonate buffer prior to peptide coupling to the plate. Furthermore, activation of peptide should not exceed 30 min as reactivity drops with longer incubation time, most likely due to hydrolysis of the activated ester (4). Optimal coupling of peptide was achieved with 30 min incubation. That covalent attachment of the peptide to the plate has taken place is demonstrated by the fact, that more peptide is bound to the plate when covalent conditions are used contrary to non-covalent conditions (Fig. 2A). At high concentrations of peptide there was a considerable non-covalent binding of peptide to the plates. To achieve a good distinction between peptide covalently coupled to the plates, and peptide that bound non-covalently, checkerboard titrations of peptide and antiserum should be used (Figs. 2A and 2B). Also, coupling of peptide to the plates at 4 o C, instead of at room temperature, resulted in a better distinction, as the reactivity of noncovalently bound peptide increased more than the reactivity of covalently bound peptide at room temperature. With patients' sera other than those mentioned above, we observed that with Tween 20 added to all of the buffers certain of these sera gave a very high background by binding unspecifically to the plates (unpublished results). This is in contrast to the usual effect of non-ionic detergents like Tween 20, which will normally reduce the nonspecific binding of antibodies to microtiter plates. Thus, if Tween 20 is added to one or more of the buffers, it is of paramount importance to include tests for background of each serum. The use of the CovaLink plates for testing the reaction between synthetic peptides and antiserum has several advantages. First of all the risk of
104 w a s h i n g o u t the c o v a l e n t l y b o u n d p e p t i d e is a b sent. S e c o n d l y , if the p e p t i d e is a t t a c h e d to a m i c r o t i t e r p l a t e via a p r o t e i n c a r r i e r to test for r e a c t i v i t y w i t h specific a n t i b o d i e s , this carrier, as well as the l i n k e r u s e d for c o u p l i n g , m u s t b e d i f f e r e n t f r o m the o n e s u s e d for i m m u n i z a t i o n , as a n t i b o d i e s will b e p r o d u c e d a g a i n s t b o t h ( B r i a n d et al., 1985; G e e r l i g s et al., 1988). T h u s the n e e d for c o u p h n g to a c a r r i e r for the E L I S A is ehminated with the present methodology. Thirdly, the m e t h o d is h i g h l y r e p r o d u c i b l e a n d has a sensit i v i t y c o m p a r a b l e to p r e s e n t l y u s e d m e t h o d s ( G n a n n et al., 1987a; G e e r l i g s et al., 1988). F i n a l l y , as the p e p t i d e s are c o v a l e n t l y c o u p l e d , it m a y b e p o s s i b l e to elute b o u n d a n t i b o d i e s a n d retest the p e p t i d e s w i t h a n e w set o f a n t i b o d i e s .
References Arnon, R. (Ed.) (1987) Synthetic Vaccines, Vols. I and II. CRC Press, Boca Raton, FL. Briand, J.P., Muller, S. and Van Regenmortel, M.H.V. (1985) Synthetic peptides as antigens: pitfalls of conjugation methods. J. Immunol. Methods 78, 59. Cuatrecasas, P. and Parikh, I. (1972) Adsorbents for affinity chromatography. Use of N-hydroxysuccinimide esters of agarose. Biochemistry 11, 2291. Geerhgs, H.J., Weijer, W.J., Bloemhoff, W., Welling, G.W. and Welling-Wester, S. (1988) The influence of pH and ionic strength on the coating of peptides of Herpes simplex virus type 1 in an enzyme-linked immunosorbent assay. J. Immunol. Methods 106, 239. Geysen, M.H. (1985) Antigen-antibody interactions at the molecular level: adventures in peptide synthesis. Immunol. Today 6, 364. Geysen, M.H., Meloen, R.H. and Barteling, S.J. (1984) Use of
peptide synthesis to probe viral antigens for epitopes to a resolution of a single amino acid. Proc. Natl. Acad. Sci. U.S.A. 81, 3998. Gnann, Jr., J.W., Schwimmbeck, P.L., Nelson, J.A., Truax, A.B. and Oldstone, M.B.A. (1987a) Diagnosis of AIDS by using a 12-amino acid peptide representing an immunodominant epitope of the human immunodeficiency virus. J. Infect. Dis. 156, 261. Gnann, Jr., J.W., McCormick, J.B., Mitchell, S., Nelson, J.A. and Oldstone, M.B.A. (1987b) Synthetic peptide immunoassay distinguishes HIV type 1 and HIV type 2 infections. Science 237, 1346. Goudsmit, J. (1988) Immunodominant B-cell epitopes of the HIV-1 envelope recognized by infected and immunized hosts. AIDS 2 (suppl. 1), $41. Hodges, R.S., Heaton, R.J. and Parker, J.M.R. (1988) Antigen-antibody interaction. Synthetic peptides define linear antigenic determinants recognized by monoclonal antibodies directed to the cytoplasmic carboxyl terminus of rhodopsin. J. Biol. Chem. 263, 11768. Lauritzen, E., Masson, M., Rubin, I. and Holm, A. (1990) Dot immunobindingand immunoblotting of small peptides onto activated nitrocellulose. Submitted. Merrifield, R.B. (1963) Solid phase peptide synthesis I. The synthesis of a tetrapeptide. J. Am. Chem. Soc. 85, 2149. Poulsen, A.G., Kvinesdal, B., Aaby, P., Molb~ek, K., Frederiksen, K., Dias, F. and Lauritzen, E. (1989) Prevalence and mortality from human immunodeficiency virus type 2 in Bissau, West Africa. Lancet, ~pril 15, 827. Sheppard, R.C. (1980) In: E. Gross and J. Meienhofer (Eds.), The Peptides. Analysis, Synthesis, Biology, Vol. 2. Marcel Dekker, New York, p. 463. Shirahama, H. and Suzawa, T. (1985) Adsorption of bovine serum albumin onto styrene/acrylic acid copolymer latex. Colloid Polym. Sci. 263, 141. Sofreniev, M.V., Madler, M., MUller, A.O. and Seriba, P.C. (1978) A method for the consistent production of high quality antisera to small peptide hormones. Frezenius 2. Anal. Chem. 290, 163.
99
JIM 05627
Covalently linked peptides for enzyme-linked immunosorbent assay Jan Sonderg/ird-Andersen a, Edgar Lauritzen 2, Klaus Lind 1 and Arne H o l m 3 Research Center for Medical Biotechnology, 1Mycoplasma Laboratory/Neisseria Department, 2 H I V Laboratory/Rubella Department, Statens Seruminstitut, Amager Boulevard 80, DK-2300 Copenhagen S, Denmark, and 3 Chemical Laboratory II, The H. C. Orsted Institute, University of Copenhagen, Denmark (Received 6 October 1989, revised received 24 January 1990, accepted 29 January 1990)
A general method is described, by which synthetic peptides are covalently linked via their carboxyl group to microtiter plates (CovaLink) for enzyme-linked immunosorbent assay (ELISA). Plates were prepared by this method with an angiotensin II peptide and with an HIV-2 peptide and attachment detected by rabbit anti-angiotensin serum and with a positive serum from an HIV-2-infected patient, respectively, using the common ELISA procedure in the last steps. The method is simple to perform, it constitutes an alternative to the common ELISA method, and eliminates the risk of inadvertent loss of peptide during the procedure. The method is highly reproducible and has a high sensitivity. It may be used for either antigen or antibody detection. Key words: Immobilization; ELISA; Synthetic peptide; CovaLink; Angiotensin II; HIV-2
Introduction
Since the introduction of solid-phase peptide synthesis by Merrifield (1963), synthetic peptides have greatly facilitated both the study of antigenantibody interaction (Hodges et al., 1983; Geysen, 1985) and the mapping of antigenic determinants (Goudsmit, 1988), and they have a promising future in the field of synthetic vaccines (Arnon, 1987) and specific diagnostic tests for microorganisms (Gnann, 1987a). However, when working with synthetic peptides a major problem is the test system used for measuring their reactivity with antibodies. It is possible to coat synthetic peptides directly onto a microtiter plate, but it involves testing each
Correspondence to: J. S~nderg~trd-Andersen, Mycoplasma Lab., Neisseria Dept., Bygn. 7, Statens Seruminstitut, Amager Boulevard 80, DK-2300 Copenhagen S, Denmark.
peptide with a panel of buffers with different pH and ionic strength for optimal binding to the solid phase (Geerlings et al., 1988). Still the method does not work equally well for all peptides, and loss of peptides during the ELISA procedure cannot be excluded. Therefore, when synthetic peptides are applied directly to a microtiter plate and reacted with antibodies, it is difficult to verify if a negative result is due to lack of binding between peptide and antibody, or to a lack of primary attachment of the peptide to the surface. Radioactively labelled peptides have been used to test for attachment, but besides the drawbacks of working with radioactive materials, a considerable workload is involved when dealing with hundreds or thousands of peptides as described by some workers (Geysen, 1985). Another approach is to use peptides coupled to a carrier like bovine serum albumin (BSA) which is known to attach to the solid phase (Shirahama et al., 1985). Besides being time consuming, it is
0022-1759/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
100
difficult to control how the peptide is coupled to the carrier. Results may vary according to whether the coupling is to a terminal amino acid, to one or more of the side chains, or whether the coupling method itself has modified functional groups of the peptide (Briand et al., 1985). Synthesis of peptides directly onto the solid support has also been used (Geysen et al., 1984). This has obvious advantages in comparison with other methods, since peptides are coupled to the surface in a controlled manner. By this technique overlapping peptides may be synthesized in a single microtiter plate, and since peptides are covalently coupled to the surface, bound antibodies can be removed, and the peptides retested with a new set of antibodies. However, by this method the identity and purity of the peptides cannot be verified directly. The need for simple and reliable methods to covalently couple purified peptides to a surface in a controlled manner seems obvious. A method where peptides are covalently coupled to nitrocellulose has recently been described (Lauritzen et al., 1990). In the present report we wish to describe a general method by which peptides can be coupled covalently via the carboxyl group to microtiter plates (CovaLink, Nunc, Roskilde, Denmark).
Immunoplates Nunc Immuno Modules, CovaLink NH, cat. no. 478042, were used. These plates have been made from polystyrene plates by covalent attachment of the linker shown in Fig. 1 (5). A total activated surface area corresponding to 100 #1 O
O
N-OH
EDc
3
•
-OCOR
\b
a 1
2
4
CH3
5
Peptides The angiotensin II peptide, with the sequence H - A s p - A r g - V a l - T y r - I l e - H i s - P r o - P h e - O H , was purchased from Sigma, and the HIV-2 peptide with the sequence H-Leu-Asn-Ser-Trp-Gly-CysAla-Phe-Arg-Gln-Val-Cys-OH ( G n a n n et al., 1987b) was from Cambridge Research Biochemicals, London, U.K.
A n tisera Anti-angiotensin II antiserum production in rabbits was performed as described previously by coupling angiotensin to the thyroglobulin carrier (Sofreniev et al., 1978). Antiserum and prevaccination serum from rabbit K-529 was used. Human serum reactive against the HIV-2 peptide was from a confirmed HIV-2 seropositive patient (Poulsen et al., 1989). Normal human serum was from a healthy, anti-HIV-1 and anti-HIV-2 negative person.
Coupling of peptides to CovaLink plates
Materials and methods
RCOOH *
tilling volume, which again corresponds to approximately 1 cm 2, is available. The number of linking groups on the surface has been radioactively estimated to approximately 1014 g r o u p s / c m 2 corresponding to 1.67 x 10 -a° m o l / c m 2 (patent pending).
~
CH3
6
Fig. 1. Covalent coupling of peptide to CovaLink. EDC: 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide.
To 5 /tl of the HIV-2 peptide (5.0 m g / m l in H 2 0 ) were added 5 ~1 of 0.1 M aqueous N-hydroxysuccinimide (NHS) and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide, HC1 (EDC), both from Sigma. To 18/tl of the angiotensin II peptide (1000 # g / m l in H 2 0 ) were added 1 8 / d of N H S + EDC. After 30 min at room temperature the activated peptides were diluted from 20 to 0.001 /~g/ml in icecold 0.1 M carbonate buffer p H 8.6 and 100 /~1 applied per well of the CovaLink plates, which were then incubated for 30 min at 4 o C, adapted from an earlier coupling procedure (Cuatrecasas and Parikh, 1972). To test for the ability of the peptides to bind noncovalently to the plates, distilled water was added to the peptides instead of N H S + EDC in the activation step. As a control for unspecific binding of antibodies to the plates, N H S + EDC was added to distilled water instead of peptide in the activation step.
101
ELISA procedure The plates were washed four times in PBS p H 7.2, postcoated with PBS p H 7.2 with 5% ( w / v ) skimmed milk powder (PBS-M) for 40 min at r o o m temperature on a rocking table, washed four times and incubated rocking at 3 7 ° C for 1.5 h with antisera diluted in PBS-M. H u m a n anti-HIV2 serum and normal h u m a n serum were diluted from 1 / 5 0 to 1/400. Rabbit anti-angiotensin II serum and prevaccination serum were diluted 1 / 1 0 0 to 1/32,000 and 1/100 to 1/2000, respectively. Plates were washed four times and incubated rocking for 45 min at 3 7 ° C with peroxidase labeled anti-antibodies (P214 (anti-human IgG) and P217 (anti-rabbit Ig), both Dakopatts, Denmark) diluted 1/1000 in PBS-M. The plates were washed four times, and colour developed for 10 min with o-phenylenediamine (0.5 m g / m l ) in citrate buffer p H 5.0 with 0.05% H202. Color development was stopped with 150/~1 1 N H2SO 4. The plates were read on an I m m u n o R e a d e r (InterMed N J-2000) at 490 nm. In addition to concentrations of peptide and serum, the following parameters were varied during optimalization of the test: (a) the amount of N H S + E D C in the final solution added to the plates; (b) the duration of peptide activation with N H S + EDC; (c) the
3000]
,
activation and coupling of peptide by N H S and E D C alone; (d) the temperature at which peptide was coupled to the plates; (e) the duration of peptide coupling to the plates; (f) the type of coupling buffer; (g) the addition of Tween 20 to the washing buffer and PBS-M.
Results
Angiotensin H Checkerboard titrations of angiotensin II peptide against rabbit anti-angiotensin II serum showed, that 1 t t g / m l peptide (Fig. 2A) and a 1/1000 dilution of antiserum (Fig. 2B) gave an optimal distinction between covalently coupled peptide, and unspecifically bound peptide. The background from antiserum bound non-covalently to plates without peptide was minimal, and normal rabbit serum showed no reaction at all with the peptide. Each well contains 1014 functional groups or 1.67 × 10-1° mol on approximately 1 cm 2 corresponding to a filling volume of approximately 100 /~1 (see above). With an average molecular weight of the peptide of 1000, 0.167/~g/100/~1 of peptide would be needed to occupy all functional groups
A
2500I
~oo~~
250O
,
i
,
,
r 8
2ooc 1500
i 1000
5OO
10
5
1 0.1 Peptide conc. in j~g/ml
0.01
0.001
1/1-00 1/5-00 1/1000 112000 1/,~000 1/1~)00 1/1~000 1132000 Antibody dilution
Fig. 2. A: rabbit serum diluted 1/1000 and angiotensin II peptide in serial dilutions. B: angiotensin II peptide 1 #g/ml and serial dilutions of rabbit serum. © ©, rabbit anti-angiotensin II serum against covalently coupled peptide. [] O, rabbit anti-angiotensin II serum against non-covalently bound peptide. • •, rabbit anti-angiotensin II serum against immunoplates without peptide. • O, normal rabbit serum against covalently coupled peptide. OD: optical density. Standard deviation is indicated on A (n = 6).
102
in the well. The highest concentration of peptide used was 10/xg/ml, and at this concentration the dose-response curve seemed to flatten (Fig. 2A). Therefore a surplus of peptide was needed to occupy all functional sites in the well, but die non-covalent binding of peptide was high at this concentration. However, at our working dilution of 1/~g peptide/ml, not all functional sites would be occupied. When the duration of peptide activation by NHS + EDC exceeded 30 min, reactivity of the activated peptide dropped. Coupling of peptide to the CovaLink plates was optimal with 30 min at 4°C. If coupling of peptide was performed at room temperature a higher background from peptide binding non-covalently to the plates was observed. When the 0.1 M carbonate buffer pH 8.6 used for coupling of peptide to the plates was substituted with PBS pH 7.2 or carbonate buffer pH 9.6, a drop in the reactivity of the plates was observed. The activating solution (peptide and NHS + EDC) should be diluted preferably 100 times in carbonate buffer. Otherwise the higher concentration of NHS + EDC would result in increased background from unspecific binding of serum to the plates. It was observed, that peptide, that had not been activated with NHS + EDC, could bind noncovalently to the plates. The effect of Tween 20 on the noncovalently bound peptide was investigated. When the microtiter plates were treated with PBS pH 7.2 with 0.05-1% Tween 20 before coupling of peptide, the plates lost their reactivity. The reactivity observed with covalently bound peptide increased when 0.05% Tween 20 was added to the washing buffer and PBS-M, but the reactivity of non-covalently bound peptide increased equally. Tween 20, therefore, did not give a better distinction between covalently and non-covalently bound peptide. If Tween 20 was added only to the washing buffer applied just after coupling of peptide to the plates, the same general increase in reactivity was observed. After coupling of peptide to the plates, they may be washed, dried, and stored in a refrigerator overnight without loss of reactivity. Earlier results with Nunc immunoplates MaxiSorp showed an almost complete desorption of
2500
2ooo~
T
20
10
0
1
(~1
0.01
0001
0.0001
Peptide ¢on¢. in jJg/ml
Fig. 3. Serum from an HIV-2-infected individual diluted 1/100 and serial dilutions of HIV-2 peptide, o o, anti-HIV-2 serum against covalently coupled peptide. [] n, antiHIV-2 serum against non-covalently coupled peptide. • • , normal human serum against covalently coupled peptide. • • , anti-HIV-2 serum against immunoplates without peptide. OD: optical density. Standard deviation is indicated on the figure (n = 6).
angiotensin II and a C terminal extension of angiotensin II the decapeptide called angiotensin I from the plates during the ELISA procedure. These peptides were desorbed after having been coated to the plates overnight using the carbonate buffer pH 9.6. HIV-2
By checkerboard titrations of HIV-2 peptide and the serum from an HIV-2 infected individual, it was shown, that 10-20 /tg peptide/ml and a 1/100 dilution of serum gave the best distinction between covalently coupled and non-covalently bound peptide (Fig. 3). As for angiotensin II the addition of Tween 20 to the buffers increased the overall signal, but did not give a better distinction between covalently and non-covalently bound peptide.
Discussion
In this paper we have described a general method by which peptides containing a carboxylic
103 acid group, terminal or side chain, are coupled covalently to CovaLink microtiter plates. The secondary amino group, to which the peptide is coupled, is placed at the end of a spacer (Fig. 1 (5)). Besides the covalent linkage the spacer includes 11 carbon atoms and 3 O/NH groups and is relatively hydrophilic. If the spacer is assumed to be fully extended into the aqueous surroundings then the N H group is estimated to be approximately 2 nm from the polystyrene surface wall of the well. This, together with the specific manner in which peptides are coupled to the functional group, increases the probability that antibodies specific for the peptides will bind to them. If peptides were absorbed onto the surface in an unspecific way, as is the case with the commonly used methodology, the probability exists, that specific antibodies would be unable to bind because of steric hindrance or inability of the peptides to assume a conformation necessary for binding. With the present method we cannot exclude, of course, that highly lipophilic covalently bound peptides will still have a tendency to bind by hydrophobic forces to the polystyrene surface. We should also point out, that the presence of side chain carboxylic acid groups, due to the amino acids aspartic acid and glutamic acid, will compete with the terminal carboxyl acid group during covalent coupling. Therefore, even though all peptides will be covalently coupled to the CovaLink plates by our method, peptides that have one or more side chain carboxylic acid groups will probably be presented to antibodies in more than one way in the same well, and this should be taken into consideration when examining the results. To achieve optimal coupling of peptide to the CovaLink plates and optimal reaction with antiserum, certain steps in the procedure are important. Coupling of peptide (Fig. 1 (1)) is obtained by using w a t e r soluble 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC (3)) as coupling agent in the presence of N-hydroxysuccinimide (NHS (2)). This procedure has been used previously for activation and coupling of peptide segments, where absence of N-hydroxysuccinimide led to a partially racemized product, and excess of both this reagent and the water soluble carbodiimide lowered the yield (Sheppard, 1980). In the
present investigation we have observed that activation with EDC alone gives rise to a much lower covalent coupling (6) than in the presence of NHS. Clearly, in order to achieve an optimal formation of the activated, intermediate peptide N-hydroxysuccinimide ester (4), the N H S + EDC concentration must not be too low during the activation step. On the other hand it must not be too high in the solution when peptide is coupled to the plates, since we have observed, that this results in an increased background from antibodies binding unspecifically to the plates. Thus, high initial concentrations of peptide, N H S and EDC is used followed by dilution in carbonate buffer prior to peptide coupling to the plate. Furthermore, activation of peptide should not exceed 30 min as reactivity drops with longer incubation time, most likely due to hydrolysis of the activated ester (4). Optimal coupling of peptide was achieved with 30 min incubation. That covalent attachment of the peptide to the plate has taken place is demonstrated by the fact, that more peptide is bound to the plate when covalent conditions are used contrary to non-covalent conditions (Fig. 2A). At high concentrations of peptide there was a considerable non-covalent binding of peptide to the plates. To achieve a good distinction between peptide covalently coupled to the plates, and peptide that bound non-covalently, checkerboard titrations of peptide and antiserum should be used (Figs. 2A and 2B). Also, coupling of peptide to the plates at 4 o C, instead of at room temperature, resulted in a better distinction, as the reactivity of noncovalently bound peptide increased more than the reactivity of covalently bound peptide at room temperature. With patients' sera other than those mentioned above, we observed that with Tween 20 added to all of the buffers certain of these sera gave a very high background by binding unspecifically to the plates (unpublished results). This is in contrast to the usual effect of non-ionic detergents like Tween 20, which will normally reduce the nonspecific binding of antibodies to microtiter plates. Thus, if Tween 20 is added to one or more of the buffers, it is of paramount importance to include tests for background of each serum. The use of the CovaLink plates for testing the reaction between synthetic peptides and antiserum has several advantages. First of all the risk of
104 w a s h i n g o u t the c o v a l e n t l y b o u n d p e p t i d e is a b sent. S e c o n d l y , if the p e p t i d e is a t t a c h e d to a m i c r o t i t e r p l a t e via a p r o t e i n c a r r i e r to test for r e a c t i v i t y w i t h specific a n t i b o d i e s , this carrier, as well as the l i n k e r u s e d for c o u p l i n g , m u s t b e d i f f e r e n t f r o m the o n e s u s e d for i m m u n i z a t i o n , as a n t i b o d i e s will b e p r o d u c e d a g a i n s t b o t h ( B r i a n d et al., 1985; G e e r l i g s et al., 1988). T h u s the n e e d for c o u p h n g to a c a r r i e r for the E L I S A is ehminated with the present methodology. Thirdly, the m e t h o d is h i g h l y r e p r o d u c i b l e a n d has a sensit i v i t y c o m p a r a b l e to p r e s e n t l y u s e d m e t h o d s ( G n a n n et al., 1987a; G e e r l i g s et al., 1988). F i n a l l y , as the p e p t i d e s are c o v a l e n t l y c o u p l e d , it m a y b e p o s s i b l e to elute b o u n d a n t i b o d i e s a n d retest the p e p t i d e s w i t h a n e w set o f a n t i b o d i e s .
References Arnon, R. (Ed.) (1987) Synthetic Vaccines, Vols. I and II. CRC Press, Boca Raton, FL. Briand, J.P., Muller, S. and Van Regenmortel, M.H.V. (1985) Synthetic peptides as antigens: pitfalls of conjugation methods. J. Immunol. Methods 78, 59. Cuatrecasas, P. and Parikh, I. (1972) Adsorbents for affinity chromatography. Use of N-hydroxysuccinimide esters of agarose. Biochemistry 11, 2291. Geerhgs, H.J., Weijer, W.J., Bloemhoff, W., Welling, G.W. and Welling-Wester, S. (1988) The influence of pH and ionic strength on the coating of peptides of Herpes simplex virus type 1 in an enzyme-linked immunosorbent assay. J. Immunol. Methods 106, 239. Geysen, M.H. (1985) Antigen-antibody interactions at the molecular level: adventures in peptide synthesis. Immunol. Today 6, 364. Geysen, M.H., Meloen, R.H. and Barteling, S.J. (1984) Use of
peptide synthesis to probe viral antigens for epitopes to a resolution of a single amino acid. Proc. Natl. Acad. Sci. U.S.A. 81, 3998. Gnann, Jr., J.W., Schwimmbeck, P.L., Nelson, J.A., Truax, A.B. and Oldstone, M.B.A. (1987a) Diagnosis of AIDS by using a 12-amino acid peptide representing an immunodominant epitope of the human immunodeficiency virus. J. Infect. Dis. 156, 261. Gnann, Jr., J.W., McCormick, J.B., Mitchell, S., Nelson, J.A. and Oldstone, M.B.A. (1987b) Synthetic peptide immunoassay distinguishes HIV type 1 and HIV type 2 infections. Science 237, 1346. Goudsmit, J. (1988) Immunodominant B-cell epitopes of the HIV-1 envelope recognized by infected and immunized hosts. AIDS 2 (suppl. 1), $41. Hodges, R.S., Heaton, R.J. and Parker, J.M.R. (1988) Antigen-antibody interaction. Synthetic peptides define linear antigenic determinants recognized by monoclonal antibodies directed to the cytoplasmic carboxyl terminus of rhodopsin. J. Biol. Chem. 263, 11768. Lauritzen, E., Masson, M., Rubin, I. and Holm, A. (1990) Dot immunobindingand immunoblotting of small peptides onto activated nitrocellulose. Submitted. Merrifield, R.B. (1963) Solid phase peptide synthesis I. The synthesis of a tetrapeptide. J. Am. Chem. Soc. 85, 2149. Poulsen, A.G., Kvinesdal, B., Aaby, P., Molb~ek, K., Frederiksen, K., Dias, F. and Lauritzen, E. (1989) Prevalence and mortality from human immunodeficiency virus type 2 in Bissau, West Africa. Lancet, ~pril 15, 827. Sheppard, R.C. (1980) In: E. Gross and J. Meienhofer (Eds.), The Peptides. Analysis, Synthesis, Biology, Vol. 2. Marcel Dekker, New York, p. 463. Shirahama, H. and Suzawa, T. (1985) Adsorption of bovine serum albumin onto styrene/acrylic acid copolymer latex. Colloid Polym. Sci. 263, 141. Sofreniev, M.V., Madler, M., MUller, A.O. and Seriba, P.C. (1978) A method for the consistent production of high quality antisera to small peptide hormones. Frezenius 2. Anal. Chem. 290, 163.
