Bioconjugate Chem. 2001, 12, 329−331
329
Design of Acridinium-9-carboxamides and Anti-acridinium Antibodies for Chemiluminescent Signal Enhancement Maciej Adamczyk,* Phillip G. Mattingly, Jeffrey A. Moore, You Pan, Kevin Shreder, and Zhiguang Yu Department of Chemistry (9NM), Abbott Diagnostics Division, Abbott Laboratories, Abbott Park, Illinois 60064-6016. Received December 22, 2000
A novel system of signal enhancement is presented in which every labeled antibody is capable of generating a signal. Three chemiluminescent acridinium-9-carboxamide haptens (1, 2, and 3) which incorporated differences in charge and location of the linker were designed and synthesized. Antiacridinium polyclonal antibodies for each hapten were screened using surface plasmon resonance instrumentation to determine specificity for each hapten. Anti-acridinium 2 antibodies were found to be non-cross-reactive to acridinium 1. This property was exploited to design secondary antibody conjugates which would bind to primary antibodies labeled with 2 yet could still be labeled with the structurally similar acridinium 1. Consequently, both layers contributed to the overall chemiluminescent signal. This format is an advance over other signal amplification formats which employ nonsignal-generating, labeled antibodies to construct multilayered systems.
A variety of assays used for biological research, clinical diagnostics, and high throughput screening employ a detection reagent labeled with species capable of generating a colorimetric, fluorescent, or chemiluminescent signal (1). In turn, the signal generated is used to detect an analyte of interest. Frequently, multiple layers of signal generating conjugates must be applied to a captured analyte to achieve signal enhancement. In the area of immunoassays, immunoconjugates can be prepared with increasing label-to-antibody ratios until an optimal signal is achieved (2). While this approach is straightforward, it does have its practical limitations. Conjugates with high label-to-antibody ratios can result in detection reagents with lowered solubility and compromised binding affinity relative to their unconjugated state (3, 4). To overcome such limitations, systems that build multiple layers of detection reagents can be used. In this way, the number of labeled antibodies increases, not the number of labels per antibody (5). A common method makes use of a biotin label (B) in conjunction with an anti-biotin antibody (Figure 1a). Detection is achieved with an anti-biotin antibody labeled with a signal generating species (S). In this scheme, the system achieves an enhanced signal relative to a single labeled antianalyte antibody (Figure 1b). Many analogous systems exist; schemes incorporating labeled avidin are perhaps the best known (6-8). While such systems are effective, they are simultaneously wasteful. In the example detailed in Figure 1a, the biotin “extender” is merely a means to noncovalently bind the signal bearing antibody and itself does not generate a signal. We reasoned that a signal generating extender would increase the overall effectiveness of the system by ensuring the contribution of every labeled antibody. In such a proposition, B, the extender, becomes a signal-generating species, S1, and the signal-generating species on the secondary antibody is designated S2 (Figure 1c). The design of a signal generating extender and its corresponding receptor poses certain challenges. First,
Figure 1. Schematic representation of immunoassay formats.
Figure 2. Acridinium-9-carboxamides used in this study.
S1 and S2 should be capable of generating a signal using an identical method of induction. Otherwise, multiple triggering events may be needed, thereby increasing the complexity of the system. Finally, because an anti-S1 antibody would need to be labeled with S2, cross-reactivity between the secondary label and the secondary antibody must be eliminated. If not, the secondary antibody conjugate would have limited capacity to bind S1 and would be prone to insolubility caused by immunoprecipitation. To overcome these challenges, three chemiluminescent acridinium-9-carboxamide haptens (1a, 2a, and 3a) were designed and synthesized (Figure 2). Acridinium deriva-
10.1021/bc000152j CCC: $20.00 © 2001 American Chemical Society Published on Web 04/19/2001
330 Bioconjugate Chem., Vol. 12, No. 3, 2001
Adamczyk et al.
Figure 3. Determination of antibody cross-reactivity using SPR.
tives have proved themselves to be efficient chemiluminescent labels used in a variety of applications (9). In the case of antibodies specific for acridinium-9-carboxamides, the antibodies would need to discriminate between acridinium labels that, by necessity, must be structurally conserved to retain their chemiluminescent properties. Differences between the structures, such as their overall charge and the location of the linker, were incorporated to generate hapten specific antibodies. Conjugation of 1b, 2b, and 3b to keyhole limpet hemocyanin (KLH) yielded the hapten-carrier conjugates 1c, 2c, and 3c, respectively, which were used to immunize New Zealand White rabbits. Three anti-sera for each immunogen were generated. IgG polyclonal antibodies (pAbs) were isolated from rabbit serum using ammonium sulfate precipitation followed by protein A chromatography. To ascertain specificity, surface plasmon resonance (SPR) was employed to gauge the relative affinities of the polyclonal antibodies for the three different haptens. SPR has previously been shown to be an effective method for characterization of antibody binding phenomena (10). Amine-bearing congeners of the haptens (1d, 2d, and 3d) were covalently attached to a carboxymethylated dextran sensor chip using NHS and EDAC. The resulting sensor chips were then capped using ethanolamine. Solutions of each of the anti-hapten pAbs were injected over the surface of the sensor chips. The resulting saturated responses were then compared (Figure 3). Significantly, of the nine possible combinations of pAbs versus hapten screened, one combination exhibited no observable cross-reactivity: Anti-2 pAbs showed no response for an acridinium 1 derivatized chip. This particular combination showed the greatest divergence with respect to charge and position of the linker. Postconjugation to KLH, 2b, yielded a hapten with a positive charge in contrast to the negative charge of the conjugated form of 1. Furthermore, the linker of 2 emanates from the nitrogen of the sulfonamide. By contrast, the linker of 1 is attached to the tosyl ring. In general, antibody specificity was similar among the set of three rabbits immunized with the same immunogen. This trending would support that the observed specificity was the result of a general immunological response to hapten structure as opposed to the isolated response of a single rabbit. Interestingly, the opposite configuration, anti-1
pAbs in combination with a 2d derivatized sensor chip showed cross-reactivity. Once a requisite pair of hapten and anti-hapten pAbs was found, immunoconjugates were designed to demonstrate the utility of the combination for construction of a signal generating layer in a model system. An anti-LT4 monoclonal antibody (mAb) was used to detect a bovine serum albumin (BSA)-L-T4 conjugate passively absorbed to a polystyrene microtiter plate (11). The antiL-T4 mAb was labeled with the active ester 2b in sodium borate buffer, pH 8.5. The anti-2 pAbs were oxidized using sodium periodate and purified from excess periodate into sodium acetate buffer, pH 5.1. Coupling of the oxidized pAbs with 1e resulted in the formation of a stable, oxime-linked immunoconjugate. Both conjugates were purified using a desalting column equilibrated with PBS in order to remove unconjugated label. The resulting conjugates were characterized using UV-visible spectroscopy and were shown to have label-to-antibody ratios of 3.4 and 4.0, respectively. Prior to or after the purification of the labeled anti-2 pAbs, no precipitation of the antibodies was observed, as might be expected if crosslinking and subsequent immunoprecipitation was to be a problem. To demonstrate the use of 2 as a signal generating extender, the chemiluminescent signals of the labeled anti-L-T4 mAb complexed to labeled or unlabeled anti-2 pAbs were compared. Specifically, the anti-L-T4 mAb conjugate was incubated with a BSA-L-T4 conjugate passively absorbed to a polystyrene plate and unbound antibody then was washed away. This incubation was followed by a second incubation using either labeled or unlabeled anti-2 pAbs (Figure 4a and 4b, respectively). After a second wash, the addition of basic peroxide resulted in a chemiluminescent signal. The observed increase in signal when labeled secondary antibody was used was taken as evidence for the ability of the label 2 to act as a chemiluminescent extender between the primary and secondary antibodies used for detection. To rule out apparent signal enhancement caused by nonspecific binding of the labeled antibodies, two controls were performed. Neither replacement of the BSA-L-T4 conjugate with BSA (Figure 4c) nor the replacement of the anti-L-T4 mAb conjugate with unlabeled mAb (Figure 4d) in the system yielded any significant signal. Both of these latter experiments indicated that the observed
Communications
Bioconjugate Chem., Vol. 12, No. 3, 2001 331 LITERATURE CITED
Figure 4. Normalized chemiluminescence microtiter plate results.
signal enhancement in Figure 4b is due to specific binding as opposed to nonspecific absorption of the immunoconjugates. In conclusion, a novel system of chemiluminescent signal enhancement has been presented in which every antibody layer is capable of generating a signal. A combination of synthetic design and ligand screening was utilized to find the appropriate hapten-antibody pair to enhance signal. Alternative synthetic or enzymatic labels also can be envisioned. Furthermore, the application of this concept to build multilayered systems should provide even higher degrees of signal enhancement. Supporting Information Available: Synthesis of 1be, 2a-d, and 3c,d; preparation and isolation of pAbs; conjugation procedures; experimental procedures for surface plasmon resonance and chemiluminesence analyses. This material is available free of charge via the Internet at http://pubs.acs.org.
(1) Davies, C. (1994) The Immunoassay Handbook (D. Wild, Ed.) pp 3-80, Stockton Press, New York. (2) Adamczyk, M., Fishpaugh, J., Mattingly, P. G., and Shreder, K. (1998) Tracermer signal generators: An arborescent approach to the incorporation of multiple chemiluminescent labels. Bioorg. Med. Chem. Lett. 8, 3595-3598. (3) Baldwin, R. W., Durrant, L., Embleton, M. J., Garnett, M., Pimm, M. V., Robins, R. A., Hardcastle, J. D., Armitage, N., and Ballantyne, K. (1985) Design and Therapeutic Evaluation of Monoclonal Antibody 791T/36-Methotrexate Conjugates, in Monoclonal Antibodies and Cancer Therapy, pp 215-231, Alan R. Liss, Inc., New York. (4) Kanellos, J., Pietersz, G. A., and McKenzie, I. G. C. (1985) Studies of Methotrexate-Monoclonal Antibody Conjugates for Immunotherapy. J. Natl. Cancer Inst. 75, 319-329. (5) Kricka, L. J. (1994) Selected Strategies for Improving Sensitivity and Reliability of Immunoassays. Clin. Chem. 40, 347-357. (6) Cordell, J. L., Falini, B., Erber, W. N., Ghosh, A. K., Abdulaziz, Z., MacDonald, S., Pulford, K. A. F., Stein, H., and Mason, D. Y. (1984) Immunoenzymatic Labeling of Monoclonal Antibodies Using Immune Complexes of Alkaline Phosphatase and Monoclonal Anti-Alkaline Phosphatase (APAAP Complexes). J. Histochem. Cytochem. 32, 219-229. (7) Mack, D., Couzens, S., Reardons, D., and Marcus, R. (1989) Rapid diagnosis of leukemia: a three-stage immunophenotyping technique. Med. Lab. Sci. 46, 316-323. (8) Kuo, H.-H. (1999) Method for amplification of the response signal in a sandwich immunoassay. U.S. Patent 5,876,944. (9) Kricka, L. J. (2000) Application of Bioluminescence and Chemiluminescence in the Biomedical Sciences, in Methods in Enzymology (M. M. Ziegler and T. O. Baldwin, Eds.) pp 333-346, Academic Press, San Diego. (10) Adamczyk, M., Moore, J. A., and Yu, Z. (2000) Application of Surface Plasmon Resonance toward Studies of LowMolecular Weight Antigen-Antibody Binding Interactions. Methods 20, 319-328. (11) Adamczyk, M., Gebler, J. C., Gunasekera, A. H., Mattingly, P. G., and Pan, Y. (1997) Immunoassay Reagents for Thyroid Testing. 2. Binding Properties and Energetic Parameters of a T4 Monoclonal Antibody and Its Fab Fragment with a Library of Thyroxine Analogue Biosensors Using Surface Plasmon Resonance. Bioconjugate Chem. 8, 133-145.
BC000152J
