Applications for the new electrochemiluminescent (ECL) and biosensor technologies.

Biosensor and electrochemiluminescent (ECL) assays are replacing enzyme-linked immunosorbent assays (ELISAs) at Schering-Plough as immunoassays of choice to monitor cytokine levels and detect anti-cytokine antibody responses during cytokine therapy. These new assays provide increased sensitivity and a better correlation with biological assays. Biosensor assays using the BIACORE 2000 (BIACORE, Uppsala, Sweden) are being adopted to support preclinical and clinical trials for the detection of antibodies capable of binding to IL-10 and IL-4. Significant advantages when using a biosensor assay are that real-time and label-free detection permit increased throughput and direct detection of binding interactions which enables detection of low affinity antibodies that are not detected by ELISA. The ECL assays using the ORIGEN Analyser (IGEN, Gaithersburg, MD) that we have implemented to replace existing ELISAs for quantification of serum IL-10 and serum interferon alfa levels are more sensitive and less subject to matrix effects. Data obtained during the validation of these assays are described.

A highly sensitive electrochemiluminescence immunoassay for interferon alfa-2b in human serum.

A highly sensitive immunoassay for the quantitative measurement of recombinant interferon alfa-2b (IFN alfa-2b) in serum was developed using the ORIGEN electrochemiluminescence (ECL) detection system. The ECL assay developed uses a ruthenylated mouse monoclonal anti-IFN-alfa antibody and a sheep polyclonal anti-IFN-alfa antibody that has been biotinylated. An immune complex, formed between the antibodies and the IFN in the serum, is captured by streptavidin coated paramagnetic beads. The assay is sensitive and can accurately measure 4 IU/ml in undiluted serum samples. In addition, this assay requires less labor than classical ELISAs and is not significantly affected by individual serum factors. The total assay variance for both intra- and inter-assay precision is < 10%. The assay is specific for the analyte. Mean percent recoveries in pooled and individual donor serum samples range from 84 to 114%. In addition, results of the ECL assay correlate well with a bioassay and were more accurate that an ELISA for the detection of interferon alfa-2b in individual human serum samples. The assay can be modified to measure other forms of the analyte and/or interferon in other matrices.

Microencapsulated antibodies in radioimmunoassay. III. Determination of cortisol.

We describe a method for directly measuring total plasma cortisol by use of a microencapsulated antibody. The antibody microcapsules were incubated with plasma and 125I-labeled cortisol in a competitive reaction at 37 degrees C for 15 min, then separated by centrifugation; the radioactivity of the pellet was counted for 1 min. Analytical recovery of cortisol from three plasma pools was 97.6% (SD 8.5%) and was unaffected by triglyceridemia, hemolysis, or icterus. The standard curve was linear to 500 microgram/L and the sensitivity was 10 microgram/l. Recovery of cortisol added as the hemisuccinate was 102% (SD 10.6%), whereas that of the conjugate cortisol hemisuccinate/bovine serum albumin was 4.3% (SD 3.5%). This confirmed that the microcapsule excludes large molecules. The within-day CV for two control pools was 9.8 and 12.8%, the day-to-day variation 13.4 and 13.8%.

Microencapsulated antibodies in radioimmunoassay. II. Determination of free thyroxine.

We describe a method for directly measuring free thyroxine in human serum. Antibody to thyroxine was encapsulated within a semipermeable nylon membrane that excludes substances of relative molecular mass greater than 20,000. The microencapsulated antibody was then presaturated with [125I]thyroxine. Serum incubated with the microcapsules initiated a displacement reaction between free thyroxine and [125I]thyroxine bound to the antibody. The displaced thyroxine was separated from the bound thyroxine by centrifugation and the concentration of free thyroxine determined from a standard curve prepared by use of known amounts of free thyroxine. The within-day coefficient of variation for two control samples was 8.8 and 6.8%; day-to-day precision for the same material was 12.5 and 13.5%. Lipemia, icterus, or hemoglobin had no adverse effect. Interlaboratory evaluation of the procedure revealed no significant difference in mean values for 20 speciments (p greater than 0.05).

Radioimmunoassay of total urinary estrogens.

We describe a radioimmunoassay for total estrogens in urine. Estrogens are extracted by adsorption on XAD-2 resin, eluted with methanol, and acid-hydrolyzed. Estrogens are then determined in a diluted aliquot of the hydrolysate. Within-run coefficients of variation for estriol values between 6.0 and 100 micrograms/L were less than 5%; between-run CV was 19.6% at an estriol concentration of 16.8 microgram/L. A correlation coefficient of 0.96 was found on comparison of the radioimmunoassay (y) with a standard fluorometric assay (x); the regression equation is y = 2.18 x - 6.70 micrograms/L. Determination on pregnancy and non-pregnancy urine can be done within 4 h.

Microencapsulated antibodies in radioimmunoassay--I. Determination of digoxin.

We describe the application of the microencapsulated-antibody technique to the radioimmunoassay of digoxin in serum. Droplets of emulsified rabbit antibody are microencapsulated in a semipermeable nylon membrane by an interfacial polymerization technique. The antibody microcapsules are incubated with 125I-labeled digoxin and unlabeled digoxin for 15 min at 37 degrees C, then free and bound digoxin are separated by centrifugation. Subtherapeutic, therapeutic, and toxic concentrations of digoxin in sera can be determined, with use of a standard curve prepared by use of known amounts of digoxin. With this technique we obtained an intra-laboratory correlation coefficient of 0.945 for 100 patients' sera and one of 0.940 for interlaboratory results for 21 sera (10 laboratories) when compared to a routine clinical laboratory radioimmunoassay for digoxin. Icterus, lipemia, hemoglobin, or disproteinemia had no effect on the analytical recovery of digoxin. The standard curve was linear to 6 microgram/L; the sensitivity was 0.25 microgram/L.