Analysis of serotonin transporter in human platelets by immunoblotting using site-specific antibodies.

We have produced a panel of site-specific antibodies recognizing different regions of the human serotonin transporter (SERT). This panel included: 1) monoclonal antibodies 23C5 (mAbs 23C5) to the C-terminal region (amino acid residues 597-630); 2) polyclonal antibodies (pAbs) to the N-terminal region (amino acid residues 69-83); 3) pAbs to the region (amino acid residues 86-100) in the beginning of the first transmembrane domain (TMD). The antibodies were produced using recombinant proteins and synthetic peptides (containing certain sequences of SERT) as antigens. These antibodies were purified by affinity chromatography, conjugated to horseradish peroxidase (HRP), and used for immunoblotting analysis of SERT in extracts of human platelets. Sodium dodecyl sulfate extracts were prepared under conditions preventing non-specific proteolytic degradation of the proteins. In platelet extracts, all antibodies were able to detect the 67 kD protein, apparently corresponding to full-length SERT molecule (its theoretical mass is about 70 kD). These antibodies also detected several polypeptides of smaller size (56, 37, 35, 32, 22, and 14 kD), apparently corresponding to N-terminal, C-terminal, and non-terminal SERT fragments. Specificity of immunostaining was confirmed by preincubation of HRP-labeled anti-SERT antibodies with excess of corresponding antigen, which resulted in disappearance of protein band staining. It is suggested that SERT undergoes a programmed proteolytic cleavage (processing) resulting in formation of several SERT-derived polypeptides of smaller size. It is possible that one of the cleaved SERT species is required for serotonin transport activity. Possible sites for specific proteolysis may be located in the region near TMD1 and in the intracellular loop between TMD4 and TMD5.

Analysis of bispecific monoclonal antibody binding to immobilized antigens using an optical biosensor.

The interaction between two different monoclonal antibodies (Mabs) and their corresponding bispecific antibodies (Babs) with immobilized antigens was investigated using an optical biosensor (IAsys). The analyzed panel of affinity-purified antibodies included two parental Mabs (one of which was specific to human IgG (hIgG), and another one to horseradish peroxidase (HRP)), as well as Babs derived thereof (anti-hIgG/HRP). Babs resulting from the fusion of parental hybridomas bear two antigen-binding sites toward two different antigens and thus may interact with immobilized antigen through only one antigen-binding site (monovalently). Using an IAsys biosensor this study shows that the bivalent binding of Mabs predominates over the monovalent binding with immobilized HRP, whereas anti-hIgG parental Mabs were bound monovalently to the immobilized hIgG. The observed equilibrium association constant (K(ass)) values obtained in our last work [1] by solid-phase radioimmunoassay are consistent with those constants obtained by IAsys. The K(ass) of anti-HRP Mabs was about 50 times higher than that of anti-HRP shoulder of Babs. The dissociation rate constant (k(diss)) for anti-HRP shoulder of Babs was 21 times higher than k(diss) for anti-HRP Mabs. The comparison of the kinetic parameters for bivalent anti-HRP Mabs and Babs derived from anti-Mb/HRP and anti-hIgG/HRP, allowed to calculate that 95% of bound anti-HRP Mabs are bivalently linked with immobilized HRP, whereas only 5% of bound anti-HRP Mabs are monovalently linked. In general, the data obtained indicate that Babs bearing an enzyme-binding site may not be efficiently used instead of traditional antibody-enzyme conjugates in the case of binding of bivalent Mabs.

The comparison of the ability of monoclonal antibodies directed to different proteins (human IgG, human myoglobin and HRP) and bispecific antibodies derived thereof to bind antigens immobilized on a surface of a solid phase.

BACKGROUND: Bindings of mouse monoclonal antibodies (mAbs) and affinity purified bispecific antibodies (bAbs), derived thereof, to antigens adsorbed on immunoplates have been compared, using ELISA and RIA methods. METHODS: The analysed panel of antibodies included mAbs specific to human myoglobin (Mb), human IgG (hIgG) and horseradish peroxidase (HRP) and biologically produced bAbs with double specificity to Mb and HRP, and to hIgG and HRP. RESULTS: The degree of difference between different mAbs and corresponding bAbs varied markedly from antibody to antibody, depending on whether the parental mAbs could bind immobilized antigens bivalently. The observed equilibrium binding constant (K(obs)) for anti-HRP mAbs was 21-38 times higher that of anti-HRP site of bAbs (anti-hIgG/HRP or anti-Mb/HRP, respectively), due to bivalent binding of mAbs. Anti-Mb mAbs also bound bivalently with immobilized Mb. On the contrary, anti-hIgG mAbs bound monovalently with immobilized hIgG in the same conditions. The avidity of anti-Mb/HRP bAbs increased, if both antigens were simultaneously adsorbed on a solid phase. CONCLUSIONS: The obtained data indicate that the use of bAbs in heterogeneous immunoassays instead of traditional mAb-enzyme conjugates hardly can provide the significant gain in assay performance if parental mAbs bind bivalently.

Synergistic effects in antigen capture ELISA using three monoclonal antibodies directed at different epitopes of the same antigen.

Using a panel of monoclonal antibodies (mAb) against human myoglobin (Mb), we have shown that the sensitivity of antigen-capture enzyme-linked immunosorbent assay (ELISA) may be significantly increased by the simultaneous immobilization on a solid phase of two co-operating capture mAbs. This method ("a three-site ELISA") uses three mAbs at different epitopes of the same antigen (two capture/one tracer), unlike the traditional two-site assay, using one capture and one tracer mAbs. We established two-site and three-site ELISA assays for Mb, by varying capture and tracer mAbs. Three-site assays showed 4-6 fold increase in sensitivity, if compared with two-site assays. The model for the effect has been suggested, according to which in three-site ELISA the high-affinity cyclic configurations may be formed by an antigen, two-capture mAbs and the surface of solid phase.

Triple-site antigen capture ELISA for human myoglobin can be more effective than double-site assay.

Using a panel of monoclonal antibodies against human myoglobin (Mb), we have shown that the sensitivity of antigen-capture ELISA can be significantly increased by simultaneous immobilization of two cooperating capture monoclonal antibodies on a solid phase. This method ("triple-site ELISA") uses three monoclonal antibodies to different epitopes of the same antigen (two capture/one tracer) unlike the traditional double-site assay using one capture and one tracer monoclonal antibody. We developed double- and triple-site ELISA for Mb by varying the capture and tracer monoclonal antibodies. Triple-site assays showed 4-6-fold increase in sensitivity compared to the double-site assays. A model for this effect is suggested; according to the model, in triple-site ELISA, high-affinity cyclic configurations can be formed by an antigen, two capture monoclonal antibodies, and the surface of the solid phase.

Single-step sandwich immunoassay of myoglobin with bifunctional monoclonal antibodies.

Two protocols for sandwich antigen-capture ELISA of human myoglobin were compared. In the first (routine) variant, 14D6 monoclonal antibodies conjugated to horseradish peroxidase were used as the secondary antibodies. Bifunctional antibodies specific for myoglobin/peroxidase were used as the secondary antibodies in the second variant. The myoglobin-binding site of the bifunctional antibodies was similar to that of the 14D6 antibodies, and the second antigen-binding site of the bifunctional antibodies was bound to horseradish peroxidase. When comparing standard calibration curves, the effective concentration of the bifunctional antibodies and that of antibodies conjugated to horseradish peroxidase were made equal. It is shown that the use of bispecific antibodies as the secondary antibodies does not improve the quality of the parameters tested, i.e., the sensitivity of the assay does not increase and the slope of the calibration curve remains constant.

Quantitative analysis of the products of IgG chain recombination in hybrid hybridomas based on affinity chromatography and radioimmunoassay.

On the model of a hybrid hybridoma (quadroma) to alpha-endorphin (END) and horseradish peroxidase (HRP), we have elaborated a general approach to analyse H and L chain interactions in hybrid hybridomas and to evaluate their efficiency as producers of bispecific antibodies (bAbs). This strategy is based on quantitative analysis of quadroma produced Abs by affinity chromatography and radioimmunoassay. First, Abs produced by quadroma cells in culture media (IgG pools from three quadroma clones) were fractioned with respect to specificity. Second, Ab concentrations in each fraction (bispecific, anti-END, anti-HRP and inactive) were measured by specific radioimmunoassays, using rabbit antiserum against mouse IgG and 125I-labelled affinity purified quadroma Abs. Then the experimentally obtained Ab distributions were compared with the predicted Ab distributions for different models of IgG chain recombination in quadroma cells (random H/L pairing, preferential homologous H/L association). As follows from these models, in a random H/L recombination the yield of bAbs in quadroma produced IgG cannot exceed 12.5%, and the ratio of bAbs and inactive Abs cannot exceed 0.5. In the analysed clones the yield of bAbs amounted to about 30% of total IgG, and the ratio of bAbs and inactive Abs was about 5-8, giving strong evidence for preferential homologous H/L association in these cells. The ratio of anti-HRP and anti-END Abs was about 10:1, suggesting unequal production of parental IgG chains in quadroma cells. The result of quantitative analysis of quadroma IgG was further supported by two-dimensional gel analysis of affinity-purified fractions of quadroma IgG and of two parental mAbs.

Study of antigen-binding properties of bispecific antibodies.

Antigen-binding properties of bispecfic antibodies (bAbs) produced by mouse hybrid hybridomas were studied. One of the bAbs held binding sites for two different antigens with relatively high molecular mass: human IgG (M(r) approximately 160,000) and horseradish peroxidase (HRP, M(r) approximately 40,000). Another bAbs showed specificity to antigens differing in molecular mass by more than an order of magnitude: peptide alpha-endorphin (END, M(r) approximately 1600) and HRP (M(r) approximately 40,000). The studied antibodies also contained different immunoglobulin chains. Both heavy chains of the anti-IgG/HRP bAbs molecule were of mouse subclass IgG1. Anti-END/HRP bAbs was formed by a combination of heavy chains which belong to two subclass of IgG: IgG2a and IgG1. bAbs were purified from ascitic fluid by a two-step affinity chromatography on columns with Sepharose-4B conjugated with the corresponding antigen. Radioimmune and immunoenzyme assays were used to analyze antigen-antibody binding and equilibrium constants of association (Ka) for each parental antibody and bAbs were determined by the Scatchard method. No significant changes in the affinity of bAbs antigen-binding sites were observed as compared to the corresponding parental antibodies. It was also shown that bAbs interaction with an excess of one of the antigens did not affect binding of the other antigen to the second bAbs site. Two-dimensional gel electrophoresis was used to analyze the composition of bAbs light and heavy chains specific to END/HRP. This analysis corroborated that bAbs molecules contained light and heavy chains from both parental hybridomas. Hence, it was demonstrated that the hybridoma fusion method can provide bispecific IgG molecules fully preserving antigen binding properties of the parental antibodies.