Modeling of immunosensors under nonequilibrium conditions. II. Experimental determination of performance characteristics.

In an attempt to optimize immunosensors operating with an immobilized antibody as binding protein and an analyte-enzyme conjugate as signal generator that is significantly larger in molecular size than the analyte, in a previous communication (Part I) (S.-H. Paek and W. Schramm (1991) Anal. Biochem. 196) we developed mathematical models for the prediction of performance characteristics. These models are compared in this contribution with experimentally obtained results. As an example, a monoclonal antibody to the steroid hormone progesterone has been used as binding protein, an 125I-progesterone derivative, and a progesterone-horseradish peroxidase derivative as tracers for signal generation. A minimum of parameters needs to be experimentally determined to calculate the performance: the amount of immobilized antibody, the diffusion coefficient of antigens, the thickness of the penetration layer, and the on- and off-rates for binding of the antigen to the antibody. We have described simple methods to obtain these data for the labeled antigen and for the unlabeled analyte that does not provide a signal per se. Kinetic binding curves for antigen-antibody complex formation obtained with the mathematical models correlated well with experimentally obtained results for antigens of different sizes. Although equilibrium of the antigen-antibody complex for the enzyme-labeled analyte conjugate requires about 4 h in the absence of free analyte, dose-response curves can be obtained after 5 min and the relative position of these curves does not change significantly after 30 min. Using a total volume of 200 microliters for the analytical procedure in microtiter wells, agitation as a means to accelerate convective diffusion during an incubation period of 30 min is not necessary with the analyte-enzyme conjugate. However, immunosensors using large analyte-enzyme conjugates as signal generators for the detection of small analytes require strict control of the incubation time if operated within short periods of time (less than 30 min).     

Homogeneous enzyme immunoassay for lipoic acid based on the pyruvate dehydrogenase complex: a model for an assay using a conjugate with one ligand per subunit.

A homogeneous enzyme immunoassay for lipoic acid was developed by using an enzyme-ligand conjugate containing only one ligand per enzyme subunit. Theoretical studies have shown that the traditional use of multisubstituted enzyme-ligand conjugates has limited the detection limits and sensitivity obtainable with these assays. The use of conjugates with a smaller number of ligands should allow for improved assays. The pyruvate dehydrogenase complex was chosen for this study because each polypeptide chain of dihydrolipoyl transacetylase contains one lipoic acid as a covalently attached prosthetic group. Thus, the naturally occurring enzyme can be considered as an enzyme-lipoic acid conjugate. Anti-lipoic acid antibodies were developed in New Zealand White rabbits to be used as the analyte-specific binders. Association and binder dilution curves were prepared in order to optimize the reagent concentrations and the analytical conditions. Unexpected inhibition by free lipoic acid resulted in a biphasic dose-response curve with a detection limit of 5 x 10(-6) M lipoic acid. This technique has several advantages over previous electrochemical and chromatographic techniques for lipoic acid determination.     

Test of a theory relating to the cross-linking of IgE antibody on the surface of human basophils

Recent mathematical models of bivalent hapten-induced histamine release from basophils predict that under appropriate conditions histamine release is maximum when cross-link formation is maximum, at a hapten concentration equal to 1/(2Ka), where Ka is the average affinity constant of the hapten for a single IgE binding site. To test this prediction we sensitized human basophils with a monoclonal anti-dinitrophenol IgE and generated histamine release dose-response curves with a bivalent hapten, alpha, epsilon-DNP-lysine. The monoclonal IgE has a published affinity constant of 7.1 X 10(7) M-1 for epsilon-DNP-lysine as determined by equilibrium dialysis. From the position of the maximum of the histamine dose-response curves, both in the presence and in the absence of monovalent DNP hapten, we determine that the sensitizing IgE has an intrinsic affinity constant of 6.9 +/- 0.5 X 10(7) M-1 for epsilon-DNP-lysine and 1.2 +/- 0.6 X 10(6) M-1 for alpha-DNP-lysine. The agreement between the two estimates of the epsilon-DNP-lysine affinity constant, one from histamine release experiments involving surface bound IgE and one from binding experiments involving IgE free in solution, 1) is consistent with a central prediction of the theory of cross-linking and 2) indicates that the hapten-binding properties of the IgE are unaffected by its being bound to Fc epsilon receptors on the basophil surface.     

Determination of an antibody-antigen binding constant by enzyme immunoassay and a theory for analysis of competitive binding of two ligands to heterogeneous receptor

A method for determining antigen-antibody binding constants by using enzyme-labeled antigens has been developed. In the measurement, enzyme-labeled and unlabeled antigens (Ag* and Ag) were allowed to compete in binding to the antibody (Ab) under conditions where Ag* much less than Ab much less than Ag. The data were analyzed according to a new theory developed for the analysis of competitive binding of two ligands to a heterogeneous receptor. The theory indicates that the binding degree of a labeled ligand measured at various concentrations of the receptor can be used to prepare a standard curve relating the binding degree of the labeled ligand and the average of the concentrations of the free receptor components which are in binding equilibrium with another unlabeled ligand. For homogeneous receptors, the method gives usual binding constants for the unlabeled ligand, but for heterogeneous receptors, it gives a new type of average binding constant for the unlabeled ligand in which the contribution of each receptor component is amplified in proportion to its affinity against the labeled ligand. This average binding constant was named the "affinity-average binding constant." A rabbit anti-blasticidin S (BLS) antiserum analyzed by the present method using beta-galactosidase-labeled BLS as the labeled ligand was found to be fairly homogeneous with respect to the affinity and to have a binding constant of 1.48 +/- 0.24 (S.D.) X 10(8) M-1 for unlabeled BLS.     

Modification of the insulin receptor by diethyl pyrocarbonate: effect on insulin binding and action

Insulin binding to rat liver plasma membranes is inhibited in a time- and dose-dependent fashion by prior treatment of membranes with the histidine-specific reagent diethyl pyrocarbonate. If all receptors are occupied by unlabeled hormone during diethyl pyrocarbonate treatment, no inhibition of 125I-labeled insulin binding is observed folowing washout of unlabeled hormone and unreacted reagent. Scatchard analysis of the binding inhibtion due to diethyl pyrocarbonate reveals a loss in receptor number rather than a change in receptor affinity for hormone. Fat cells treated with diethyl pyrocarbonate exhibit a rightward shift in the dose-response relationship for insulin-stimulated glucose oxidation consistent with a loss in receptor number due to the reagent. The pH profile for inhibition of insulin binding by diethyl pyrocarbonate and the partial reversibility of this inhibition by hydroxylamine are consistent with modification of a histidine residue. These results suggest that a histidine residue at or near the receptor binding site is required for formation of the biologically relevant insulin - receptor complex.     

High affinity ScFvs from a single rabbit immunized with multiple haptens

We report the generation of single-chain Fv (scFv) fragments with high affinities against four different hapten molecules from a single immunised rabbit. The rabbit was immunised with a mixture of protein conjugates of four different haptens, namely the herbicide mecoprop and derivatives of the herbicides atrazine, simazine, and isoproturon. An scFv phage display library was constructed, and several scFvs with high affinity against each hapten were isolated. For each hapten, a single binder was selected by k(off) ranking and used for affinity determination. The affinities were in sub-nanomolar range and the lowest K(d) value obtained was 6.75 x 10(-10) M. An unusual feature of one of the anti-isoproturon scFvs was its ability to retain binding activity at pH1.7. The utility and potential of using a single animal and immunisation with multiple antigens for the production of multiple, specific, high affinity scFvs by phage display is discussed.     

Functionalization of covalent DNA-streptavidin conjugates by means of biotinylated modulator components.

Covalent DNA-streptavidin conjugates are versatile biomolecular coupling reagents, since they have binding capacity for both a complementary nucleic acid and four molecules of biotin. The DNA-streptavidin hybrid molecules have been investigated for their capabilities to bind two different types of biotinylated components. Thus, (i) a functional biomolecule, e.g., a single-stranded DNA fragment or an enzyme and (ii) low-molecular weight biotin derivatives ("modulators") were coupled stepwise with the hybrid molecules. Modulators were D-biotin, aminobiotin, and biotin-fluorescein conjugate as well as a lysine-rich 10mer peptide, containing a biotin and a fluorescein substituent. These modulators were chosen to affected the hybridization properties of the DNA-streptavidin conjugates. As investigated by surface-plasmon resonance and microplate solid-phase hybridization measurements, D-biotin, biotin-fluorescein, and aminobiotin decreased the efficiency of hybridization with complementary, surface-bound oligonucleotides to a varying extent. The basic peptide increased the conjugate's hybridization efficiency. Moreover, it was demonstrated in two examples how modulators can be utilized as additional functional domains of streptavidin-based conjugates. First, fluorescein-containing modulators were used as hapten groups, allowing a sensitive detection by means of specific antibodies directed against the modulator. Second, the biotinylated peptide was used as a carrier molecule to attach multiple fluorogenic lanthanide-chelate groups to the streptavidin conjugate, enabling its sensitive detection by time-resolved fluorometry. The applicability of this kind of bioconjugation strategy to generate sensor-probes for gene detection assays was demonstrated.     

Dual analyte detection using tandem flash luminescence

A heterogeneous, dual analyte-binding assay which makes use of the flash luminescence from both aequorin and an acridinium-9-carboxamide label is presented. The signal generating species were triggered both differentially and sequentially using Ca(2+) followed by basic peroxide. Both signals were resolved readily using a single photomultiplier tube without the need for multiwavelength detection. To demonstrate the tandem luminescence concept in a model assay system, dose-response curves for two analytes, biotinylated BSA and myoglobin, were generated using a competitive binding format. Because of the relatively short assay time and the well-resolved signals, this format will be useful in the development of dual analyte high-throughput assays.     

A heterologous enzyme immunoassay of prostaglandin E2 using a stable enzyme-labeled hapten mimic

A sensitive heterologous enzyme immunoassay for prostaglandin E2 was developed using 9-deoxy-9-methylene-prostaglandin F2 alpha as a stable prostaglandin E2 mimic. beta-Galactosidase was conjugated to the hapten mimic. Anti-prostaglandin E2 IgG was bound to a polystyrene tube. The enzyme-labeled hapten mimic mixed with unlabeled prostaglandin E2 was allowed to react in a competitive manner with the immobilized antibody. Then, the beta-galactosidase specifically bound to the antibody was assayed fluorometrically, and the enzyme activity was correlated with the amount of unlabeled prostaglandin E2. According to the calibration curve thus obtained, prostaglandin E2 could be determined in a range of 1.2-430 fmol. Prostaglandin E2 was extracted from human urine by the use of an octadecylsilyl silica column. The crude extract contained a substance(s) which disturbed the enzyme immunoassay and gave an apparently high content of prostaglandin E2. The interfering substance was separated from prostaglandin E2 by reverse-phase high-performance liquid chromatography. The purified urinary extract was examined by the enzyme immunoassay for prostaglandin E2, and the validity of the results was confirmed by gas chromatography-selected ion monitoring.     

Organomercury bioconjugate synthesis and characterization by matrix-assisted laser desorption ionization mass spectrometry

Synthesis of an organomercury hapten and conjugation of the hapten to proteins and peptides is described. Starting with allylamine, synthesis of the organomercury hapten was completed in five steps using readily available and inexpensive reagents. The key transformation in the synthesis, intramolecular oxymercuration, was achieved in good yield and under mild conditions. Hapten conjugation was afforded via disuccinimide active ester coupling chemistry, and the resulting conjugates were analyzed by matrix-assisted laser desorption ionization mass spectrometry (MALDI-MS). To exploit the accurate mass measuring capabilities of MALDI-MS, the conjugates were digested with trypsin prior to analysis. The masses of the peptides resulting from tryptic digestion of the organomercury conjugates were accurately measured, and five hapten attachments were identified in the mass range of 1000-2200 m/z. The organomercury bioconjugate synthesized in this study was designed to contain a stable carbon-metal bond, constituting an underutilized approach for preparing protein-metal complexes and may result in mAbs consisting of unique recognition capabilities.     

Affinity capillary electrophoresis: important application areas and some recent developments

Affinity capillary electrophoresis (ACE) is a broad term referring to the separation by capillary electrophoresis of substances that participate in specific or non-specific affinity interactions during electrophoresis. The interacting molecules can be found free in solution or can be immobilized to a solid support. Every ACE mode has advantages and disadvantages. Each can be used for a wide variety of applications. This paper focuses on applications that include purification and concentration of analytes present in diluted solutions or complex matrices, quantitation of analytes based on calibration curves, and estimation of binding constants from direct and derived binding curves based on quantitation of analytes or on analyte migration shifts. A more recent chemicoaffinity strategy in capillary electrophoresis/capillary electrochromatography (CE/CEC) termed molecular imprinting (`plastic antibodies') is discussed as well. Although most ACE studies are aimed at characterizing small-molecular mass analytes such as drugs, hormones, and peptides, some efforts have been pursued to characterize larger biopolymers including proteins, such as immunoglobulins. Examples of affinity interactions that have been studied are antigen–antibody, hapten–antibody, lectin–sugar, drug–protein, and enzyme–substrate complexes using ultraviolet, laser-induced fluorescence, and mass spectrometer detectors. This paper also addresses the critical issue of background electrolyte selection and quantitation of analytes. Specific examples of bioaffinity applications are presented, and the future of ACE in the biomedical field is discussed.     

Quantitative protein profiling using antibody arrays

Traditional approaches to microarrays rely on direct binding assays where the extent of hybridisation and the signal detected are a measure of the analyte concentration in the experimental sample. This approach, directly imported from the nucleic acid field, may fail if applied to antibody-antigen interactions due to the shortage of characterised antibodies, the significant heterogeneity of antibody affinities, their dependence on the extent of protein modification during labelling and the inherent antibody cross-reactivity. These problems can potentially limit the multiplexing capabilities of protein affinity assays and in many cases rule out quantitative protein profiling using antibody microarrays. A number of approaches aimed at achieving quantitative protein profiling in a multiplex format have been reported recently. Of those reported, the three most promising routes include signal amplification, multicolour detection and competitive displacement approaches to multiplex affinity assays. One in particular, competitive displacement, also overcomes the problems associated with quantitation of affinity interactions and provides the most generic approach to highly parallel affinity assays, including antibody arrays.     

Measuring antibody affinity and performing immunoassay at the single molecule level

Fluorescence correlation spectroscopy (FCS) enables direct observation of the translational diffusion of single fluorescent molecules in solution. When fluorescent hapten binds to antibody, analysis of FCS data yields the fractional amounts of free and bound hapten, allowing determination of the equilibrium binding constant. Equilibrium dissociation constants of anti-digoxin antibodies and corresponding fluorescein-labeled digoxigenin obtained by FCS and fluorescence polarization measurements are identical. It is also possible to follow a competitive displacement of the tracer from the antibody by unlabeled hapten using FCS in an immunoassay format. The fluorescence polarization immunoassay for vancomycin detection was used to test the FCS approach. Fitting of the FCS data for the molar fractions of free and bound fluorescein-labeled vancomycin yielded a calibration curve which could serve for determination of the vancomycin concentration in biological samples.     

Recognition of HIV-TAR RNA using neomycin–benzimidazole conjugates

Synthesis of a novel class of compounds and their biophysical studies with TAR-RNA are presented. The synthesis of these compounds was achieved by conjugating neomycin, an aminoglycoside, with benzimidazoles modeled from a B-DNA minor groove binder, Hoechst 33258. The neomycin–benzimidazole conjugates have varying linkers that connect the benzimidazole and neomycin units. The linkers of varying length (5–23 atoms) in these conjugates contain one to three triazole units. The UV thermal denaturation experiments showed that the conjugates resulted in greater stabilization of the TAR-RNA than either neomycin or benzimidazole used in the synthesis of conjugates. These results were corroborated by the FID displacement and tat-TAR inhibition assays. The binding of ligands to the TAR-RNA is affected by the length and composition of the linker. Our results show that increasing the number of triazole groups and the linker length in these compounds have diminishing effect on the binding to TAR-RNA. Compounds that have shorter linker length and fewer triazole units in the linker displayed increased affinity towards the TAR RNA.     

Effect of diethyl pyrocarbonate modification on the calcium binding mechanism of the sarcoplasmic reticulum ATPase

Diethyl pyrocarbonate was used to modify histidyl residues on the sarcoplasmic reticulum ATPase. Difference spectra of the N-carbethoxyhistidyl derivative indicated that most all the histidyl residues on the enzyme had been modified. These residues could be divided into two populations on the basis of their reaction rate with the reagent. It could then be shown that enzyme inhibition followed modification of the slower reacting population. Reversal with hydroxylamine verified that the loss of activity was due specifically to histidyl modification. Using [32P]ATP as a substrate it was further determined that the modified ATPase could form a phosphoenzyme intermediate, but that the hydrolysis of this intermediate was inhibited. Size exclusion chromatography was used to obtain equilibrium binding curves for high affinity Ca2+ sites on the enzyme. With the normal ATPase a cooperative binding curve for two Ca2+ with a Hill coefficient of 1.8 was observed. With the modified ATPase binding to two independent sites was observed; however, the dissociation constants remained the same as in the cooperative mechanism (K1 = 14 microM; K2 = 0.5 microM). That is, modification had eliminated cooperativity without changing the site specific binding affinities. E-P formation was then shown to follow binding to the higher affinity of the two sites. This would be the second site to bind Ca2+ in a sequential, cooperative mechanism. A model is suggested in which the binding of Ca2+ to an initial site allows for the binding of a second Ca2+ to an occluded site, this second site being responsible for enzyme activation. Modification apparently allows the binding properties of both sites to be observed independently.     

Deaminating enzyme labels for potentiometric enzyme immunoassay

Adenosine deaminase, asparaginase, and urease are examined as possible enzyme labels for immunoassays using potentiometric detection with the ammonia gas-sensing membrane electrode. Considerations of binding ability, retained activity, and stability reveal asparaginase to be the most effective enzyme label for immunoassay purposes. The utility of the potentiometric approach with this enzyme label is demonstrated via model hapten assays for dinitrophenyl groups and for cortisol.     

Gonadal receptors. I. Evidence for irreversibility in the binding of human chorionic gonadotropin and human luteinizing hormone.

The binding of human chorionic gonadotropin and human luteinizing hormone to particulate receptors of rat testes has generally been assumed to follow an equilibrium model similar to that proposed for many enzyme systems. Our work shows that equilibrium dissociation constant (Kd) and number of hormone binding sites (Bmax) are highly sensitive to changes in hormone and/or receptor concentration and to treatment received by tissue or receptor preparation prior to the assay. The results of binding assays obtained using receptor preparation pretreated with hormone (labeled as well as unlabeled) indicated that the binding reaction between hormone and receptor was irreversible and that pretreatment of the tissue with hormone greatly alters the number of high affinity gonadotropin binding sites in the testicular homogenate. Data from studies involving increasing receptor concentrations revealed that increasing the mass of particulate receptors in the binding assays leads to higher Kd as well as Bmax values. These findings are incompatible with a binding model based upon occupancy of receptor sites and the state of equilibrium implied. The incompatibilities are analyzed and an alternate model advanced (Bhalla, V.K., Trowbridge, C.G., Chen, C.J.H., Lindeman, J.G. and Rojas, F.J. (1979) Biochim. Biophys. Acta 584, 436--453).     

Solid phase enzyme-linked competitive binding assay for riboflavin

A new solid-phase enzyme-linked assay for riboflavin (vitamin B2) is described. The assay is based on the competition between analyte vitamin molecules and a glucose-6-phosphate dehydrogenase-3-carboxymethylriboflavin conjugate for a limited number of riboflavin-binding protein sites immobilized on Sepharose particles. Significant improvements in conjugate catalytic activity and thus detectability are achieved by optimizing the reaction conditions used to covalently link 3-carboxymethylriboflavin to the enzyme. Optimization experiments include studying the effects of reaction pH and organic solvent composition. Final assay detection limits and the sensitivity of the dose-response curves are dependent on the ratio of conjugate to binding protein sites utilized in an equilibrium assay protocol. Selectivity of the method correlates well with that predicted based on the known association constants of riboflavin-binding protein with flavin analogs. The assay is shown to offer adequate detection limits and selectivity for direct measurement of riboflavin in urine, infant formula, and vitamin capsules.     

A high-affinity anti-salbutamol monoclonal antibody: key to a robust lateral-flow immunochromatographic assay

Among the components that make up a lateral-flow immunochromatographic assay (ICA), antibody is the key. In this paper, salbutamol (SAL) as a model analyte was meticulously designed to prepare immunogen and coating antigen in distinctly different ways. Four hybridoma cell lines were prepared and identified. Among them, C9 had highest affinity, best dose-response behavior, lowest limit of detection, and highest specificity and was chosen to be labeled with colloidal gold as the detector reagent and applied on the conjugate pad. Goat anti-mouse antibody and SAL-BSA conjugate were sprayed on a nitrocellulose membrane as test line and control line, respectively. Under the optimized conditions, the ICA strip was constructed based on a binding inhibition format. Color intensity on the test line was visually distinguishable from that of the negative sample within 5 min, with the visual detection limit of 1 ngml(-1) in phosphate-buffered saline. Cross-reactions with other β-agonists were not found (<1%). The results from ICA were in a good agreement with those obtained by enzyme-linked immunosorbent assay. The developed ICA has potential as a useful on-site screening tool for SAL in swine urine.     

Characterization of canine renal receptors for the parathyroid hormone-like protein associated with humoral hypercalcemia of malignancy

Parathyroid hormone-like proteins (PTHLP) display actions in the kidney which are similar to those of parathyroid hormone (PTH). We compared the binding properties of PTHLP and PTH in canine renal cortical membranes to determine if they interacted with the same or different receptors. Radioiodination to high specific activity (greater than 400 microCi/micrograms) of [Nle8,18,Tyr34]human PTH-(1-34)amide and [Tyr36]PTHLP-(1-36)amide was performed using the lactoperoxidase method. Complete enzymatic digestion of both radioligands demonstrated that the peptides were monoiodinated. Both radioligands retained full biological activity in the renal adenylate cyclase assay, and neither was significantly degraded during incubation with highly purified canine renal membranes under binding assays conditions. Specific binding reached equilibrium by 20 min at 20 degrees C. Competition binding studies using unlabeled [Nle8,18,Tyr34]human PTH-(1-34)amide, [Tyr36] PTHLP-(1-36)amide, and bovine PTH-(1-34) with either radioligand revealed similar binding affinities for all three peptides. Biologically inactive PTHLP fragments did not show significant displacement. In contrast to its similar binding affinity, [Tyr36]PTHLP-(1-36)amide was 6-15-fold less potent than bovine PTH-(1-34) in the renal adenylate cyclase assay, suggesting less efficient receptor-effector coupling. Photoaffinity cross-linking using either radioligand in canine renal membrane labeled indistinguishable 70,000-dalton proteins. In the presence of multiple protease inhibitors, binding to an 85-kDa component was observed. Labeling of both receptor forms was specifically abolished by an excess of either cold peptide and dose-response curves using affinity cross-linked membranes corroborated the apparent binding affinities determined by conventional radioligand binding assays. We conclude that PTHLP-(1-36) and amino-terminal PTH analogues bind to indistinguishable receptors in canine renal cortical membranes, but display differential coupling to post-receptor events.