Practical quantitative and kinetic applications of bio-layer interferometry for toxicokinetic analysis of a monoclonal antibody therapeutic

Bio-layer interferometry (BLI) is a label-free technology that can be used for kinetic characterization of proteins. Although other label-free platforms have been used for quantitation purposes (most notably surface plasmon resonance), little work has been done using BLI. Here we present rationale and strategies for the development and analytical qualification of a BLI assay for the quantitation of a humanized antibody therapeutic in cynomolgus monkey plasma. Results of the qualification were compared to those of a validated ELISA used to quantitate the same therapeutic. Selectivity, matrix effect, and precision and accuracy were similar between the two methods. Target interference was more pronounced in the BLI assay compared to the ELISA. The main difference between the two assays was in the dynamic range (0.1-10 μg/mL for ELISA vs. 0.4-50 μg/mL for BLI). The monkey plasma BLI assay was applied to rat plasma for the comparison of study samples generated in the same matrix by ELISA. A direct quantitation comparison of sample results for the two methods shows a high degree of agreement (r(2)=0.979, slope=1.017). However, an evaluation of low concentration samples showed a bias of over-recovery in the BLI compared to the ELISA. In addition to utilizing the quantitative capabilities of the platform, we evaluated the utility of using the kinetic properties of the quantitative assay to detect anti-drug antibodies (ADA) and illustrated the potential for ADA to cause either over recovery (non-neutralizing ADA) or under recovery (neutralizing ADA) of a biotherapeutic using the BLI assay.     

Multiplexed serum measurement of IgG, IgA, IgM, and IgE antibody responses to therapeutic biologicals

Analytical methods characterizing the immunogenicity of therapeutic proteins are useful for monitoring, characterizing and predicting reactions to biopharmaceuticals. A multiplexed assay capable of isotyping the specific IgG, IgA, IgM and IgE (IgGAME) antibody responses against a biotherapeutic was demonstrated in a hyper-immunized cynomolgus monkey, over a 15-month period. The quantitative range of the antibody measurements was determined to be 15 ng/ml to 50 ng/ml in 10% serum. By the use of any biotinylated or fluorescently tagged therapeutic as a detector, this multiplexed isotyping assay can be broadly applied to human and non-human primate IgG, IgA, IgM and IgE immunogenicity studies.     

Preclinical efficacy and safety of pascolizumab (SB 240683): a humanized anti-interleukin-4 antibody with therapeutic potential in asthma

The type 2 helper T cell (T(H)2) cytokine interleukin (IL)-4 is thought to play a central role in the early stages of asthma. In an effort to develop an antibody treatment for asthma that neutralizes the effects of IL-4, a murine monoclonal antibody, 3B9, was generated with specificity for human IL-4. In vitro studies demonstrated that 3B9 inhibited IL-4-dependent events including IL-5 synthesis, (T(H)2) cell activation and up-regulation of immunoglobulin E expression. 3B9 was then humanized (pascolizumab, SB 240683) to reduce immunogenicity in humans. SB 240683 demonstrated species specificity for both monkey and human IL-4 with no reactivity to mouse, rat, cow, goat or horse IL-4. Pascolizumab inhibited the response of human and monkey T cells to monkey IL-4 and effectively neutralized IL-4 bioactivity when tested against several IL-4-responsive human cell lines. Affinity studies demonstrated rapid IL-4 binding by pascolizumab with a slow dissociation rate. In vivo pharmacokinetic and chronic safety testing in cynomolgus monkeys demonstrated that pascolizumab was well tolerated, and no adverse clinical responses occurred after up to 9 months of treatment. Three monkeys developed an anti-idiotypic response that resulted in rapid pascolizumab clearance. However, in the chronic dosing study the antibody response was transient and not associated with clinical events. In conclusion, pascolizumab is a humanized anti-IL-4 monoclonal antibody that can inhibit upstream and downstream events associated with asthma, including (T(H)2) cell activation and immunoglobulin E production. Clinical trials are under way to test the clinical efficacy of pascolizumab for asthma.     

Quantitative determination of humanized monoclonal antibody rhuMAb2H7 in cynomolgus monkey serum using a Generic Immunoglobulin Pharmacokinetic (GRIP) assay

Preclinical pharmacokinetic (PK) assays are important to help evaluate the safety and efficacy of a potential biotherapeutic before clinical studies. The assay typically requires a biotherapeutic-specific reagent to minimize matrix effects especially when the host species are non-human primates such as cynomolgus monkeys and the biotherapeutic is a humanized monoclonal antibody (MAb). Recombinant humanized mAb 2H7 (rhuMAb2H7) binds to the extracellular domain of CD20 that is expressed on B cells and results in B cell depletion. It is currently being evaluated for its therapeutic potential in rheumatoid arthritis (RA) in clinical studies. During the early development of rhuMAb2H7, a cynomolgus monkey PK assay was needed to help assess the pharmacokinetic parameters of rhuMAb2H7 in a pilot cynomolgus monkey study. However, development of a cynomolgus monkey PK assay was challenging due to lack of rhuMAb2H7-specific reagents. Here we describe an alternative method for detection of rhuMAb2H7 in cynomolgus monkey serum using polyclonal antibodies against human IgGs. This assay quantifies rhuMAb2H7 in 10% cynomolgus monkey serum with high sensitivity, accuracy, and precision. This assay successfully supported the rhuMAb2H7 development, and has the potential to be used to quantify other humanized MAb biotherapeutics in serum from a variety of non-human species.     

Strategies for detection,measurement and characterization of unwanted antibodies induced by therapeutic biologicals

An important aspect of evaluating the safety of therapeutic biologicals is the assessment of the unwanted immunogenicity of such biologicals in recipients. Properly planned immunogenicity studies with appropriately devised strategies are critical if valid conclusions concerning the unwanted immunogenicity are to be derived. Such studies need to be conducted using carefully selected and validated procedures. Several techniques are available for detection and measurement of immunogenicity including immunoassays, radioimmunoprecipitation assays (RIPAs), surface plasmon resonance (SPR) and bioassays. A combination of methods for characterization of the induced antibodies is usually necessary for a detailed understanding of the type(s) of antibodies generated against a therapeutic product. This review considers the benefits and limitations of the various techniques available for antibody detection and outlines a strategy for the assessment of unwanted immunogenicity of therapeutic products.     

An enzyme-linked immunosorbent assay for enumeration of Pneumocystis carinii in vitro and in vivo.

An enzyme-linked immunosorbent assay (ELISA) to quantitate Pneumocystis carinii organisms from culture supernatant and rat lung has been developed. A polyclonal antibody specific to P. carinii was produced in Sprague-Dawley rats by allowing P. carinii-infected animals to recover from infection. This antibody reacted strongly to P. carinii proteins of 50 to 55 kDa and weakly to those of 33 and 116 kDa. The ELISA used this convalescent-phase antibody to quantitate the number of P. carinii organisms in lung homogenates of infected rats and supernatants from infected tissue cultures which were used to screen drugs for P. carinii. The results of the ELISA were compared with those of direct microscopic counting of organisms, and the two methods were highly correlated (r > 0.9). Thus, the ELISA can be used as an alternative method for the quantitation of P. carinii organisms, and it is superior to the conventional microscopic method because it is easier to perform and less labor-intensive.     

Detection of Lassa virus antigens and Lassa virus-specific immunoglobulins G and M by enzyme-linked immunosorbent assay.

Rapid diagnosis of Lassa fever is desirable for the timely therapeutic intervention and implementation of strict quarantine procedures both in West Africa field hospitals where the disease is endemic and at international crossroads. An enzyme-linked immunosorbent assay (ELISA) to measure Lassa virus antigens in viremic sera was developed in which experimentally infected monkeys were used as a model for the human disease. In this test, Lassa virus antigens in test sera were captured in wells of microtiter plates by monkey anti-Lassa virus immunoglobulin. Guinea pig anti-Lassa virus immunoglobulin was then added, and binding of specific immunoglobulin was quantitated by the addition of rabbit anti-guinea pig immunoglobulin followed by alkaline phosphatase-labeled anti-rabbit immunoglobulin. This test detected viremia titers as low as 2.1 log10 PFU/ml in experimentally infected monkey sera, a titer often exceeded in patients with Lassa fever. Inactivation of infectious virus by beta-propiolactone or gamma-irradiation did not diminish reactivity. Antigen-ELISA concentrations increased with infectivity for the first 10 days after infection but then declined while infectivity titers remained high, suggesting that the presence of humoral antibody in viremic sera diminishes the sensitivity of the antigen ELISA. Lassa virus-specific immunoglobulin M (IgM) titers measured in an IgM capture ELISA were detectable within 10 days of infection and peaked after 36 days but remained detectable for 1.5 years. The Lassa virus-specific IgG ELISA response was slightly delayed, peaking on day 73 but declining only slightly thereafter. These studies in a realistic primate model suggest that the antigen detection ELISA or the IgM capture ELISA described, in which beta-propiolactone-inactivated sera are used, should be useful for the rapid diagnosis of human Lassa fever.     

Species-dependent serum interference in a sandwich ELISA for Apo2L/TRAIL

To support pre-clinical studies of Apo2L/TRAIL in rodents and non-human primates, a sandwich ELISA was developed using two mouse monoclonal anti-Apo2L/TRAIL antibodies. Mouse, rat, cynomolgus monkey, and chimpanzee serum at concentrations of > or =1% were found to interfere with accurate quantitation of Apo2L/TRAIL. Moreover, the characteristics of the serum interference for each species were different. In order to resolve the observed serum effect, studies were performed in which salts, detergents, and blocking proteins were added to the sample diluent, and optimized sample diluents that eliminated serum interference were developed for mouse, cynomolgus monkey, and chimpanzee serum. These buffers consisted of a base assay diluent (PBS/0.5% BSA/0.05% Tween-20/10 ppm ProClin 300) supplemented with: NaCl (mouse serum); NaCl, EDTA, CHAPS, bovine gamma globulin (BGG), and human IgG (cynomolgus monkey serum); and NaCl and EDTA (chimpanzee serum). Full characterization studies were performed for the "buffer" ELISA run in base assay diluent (intended for non-serum samples) as well as the assays optimized for mouse serum and cynomolgus monkey serum. Precision, accuracy, linearity, and specificity were found to be satisfactory. With the availability of a rabbit polyclonal antibody against Apo2L/TRAIL, a new pAb/mAb ELISA was developed. This assay was not only more sensitive by > or =6-fold, but it was also much less subject to serum interference.     

丙型肝炎病毒多表位抗原重组体免疫原性分析

目的：研究丙型肝炎病毒(HCV)多表位抗原重组体在大肠杆菌中的非融合表达及其免疫原性分析。方法：利用DNA重组技术将构建的HCV多表位抗原重组体克隆入原核表达载体pBV220并在BL-21中表达，非融合表达蛋白经SDS-PAGE检测，排阻层析纯化，利用ELISA和免疫印迹分析其抗原反应性。免疫昆明鼠和猕猴，并检测其抗体和特异性CTL(Cyto-toxic T lymphocyte)，在抗体转阴后再次用HCV病人血清攻击以检测其免疫应答能力。结果：HCV多表位重组体在pBV220／BL-21中表达量占菌体总蛋白量的15％。Western-blot和ELISA分析表明HCV多表位抗原能与HCV病人血清特异性结合，抗体水平和特异性CTL检测结果显示该抗原肽能诱导小鼠和猕猴产生较好的抗体水平和特异性CTL效应。用病人血清再次攻击后受试动物的抗体迅速阳转。结论：表达的多表位重组体具有特异的抗原反应性和免疫原性。     

Pneumocystis carinii antigen detection in rat serum and lung lavage.

We developed a modified double-antibody sandwich enzyme-linked immunosorbent assay (ELISA) that detected relatively low concentrations of known Pneumocystis carinii antigen added to buffer or rat sera. Artificial immunization-derived polyclonal rabbit anti-P. carinii antibody was used on the solid phase to capture the antigen. Infection-derived (after P. carinii pneumonia) polyclonal rat anti-P. carinii antibody or a mixture of five murine monoclonal antibodies was used as the antigen detector antibody. Rabbit anti-rat immunoglobulin G antibody or goat anti-mouse immunoglobulin G antibody conjugated to alkaline phosphatase was used as the final antibody. After standardization and optimization of the various reactants in this ELISA system, approximately 53 ng of known P. carinii antigen per ml suspended in phosphate-buffered saline-Tween 20 buffer or 210 ng of antigen per ml suspended in normal rat serum diluted 1:4 could be detected. In addition, an indirect ELISA for P. carinii antibody measurement was developed, using as the antigen a soluble supernatant from a sonicated preparation of Percoll-purified whole cysts and trophozoites to coat the solid phase. Limited studies with sera from a small number of caesarian-obtained, barrier-sustained rats from Charles River Breeding Laboratories, Inc., and the National Institutes of Health and sera from normal and heavily infected rats indicated that the caesarian-obtained, barrier-sustained rats had negligible levels of antibody. The normal and heavily infected rats had variable antibody titers. A significantly high level of P. carinii antigenemia was detected in only 2 (11%) of 18 heavily infected rats. Extensive studies of the P. carinii pneumonia rat model with the ELISA did not reveal significant serum P. carinii antigenemia during the acute stage of infection. However, soluble P. carinii antigen was detected by the ELISA and Western blot assays in the supernatant of lavage fluid after centrifugation to sediment intact organisms. As expected, P. carinii antigens were detected by these assays in the lavage pellet recovered after centrifugation. In conclusion, the antigen assay used in this study detected P. carinii antigen in lung lavage but failed to detect P. carinii antigen in rat serum during the acute phase of infection.     

Enzyme-linked immunosorbent assay for mumps IgM antibody: comparison of IgM capture and indirect IgM assay

We have established two enzyme-linked immunosorbent assays (ELISAs) for detection of mumps IgM antibody, i.e., indirect IgM ELISA and IgM capture ELISA, for serodiagnosis of recent mumps infection. In the latter method, peroxidase-conjugated monoclonal antibody to mumps virus was employed. Both methods detected mumps antibody of IgM class only in serum fractions separated by centrifugation through a sucrose density gradient. Optical density values given by both ELISAs were correlated for most sera examined. Indirect IgM ELISA, however, gave a false positive reaction for sera containing both rheumatoid factor and mumps IgG antibody, while giving a false negative reaction for sera containing high titers of mumps IgG antibody. This technique was, therefore, less reliable than IgM capture ELISA. IgM antibody detectable by IgM capture ELISA was present in all patients with mumps by the fifth day of illness and persisted for up to 3 mth in most and up to 5 mth in same cases.     

An ELISA assay for heme oxygenase (HO-1)

A double antibody capture ELISA for the HO-1 protein has been developed to separately quantitate HO-1 protein. The use of 2.5% NP40 detergent greatly assists in freeing HO-1 protein from membranes and/or other cellular entities and increased the amount of HO-1 protein found in rat liver whole homogenates as well as the nuclear, mitochondrial and microsomal fractions. Use of the detergent NP40 did not substantially change HO-1 protein standard curves. The ELISA assay for HO-1 has been shown to be reproducible over (i) a 4-day trial period as well as (ii) almost 1 year of general laboratory use. Excellent specificity for the HO-1 isoform is shown by the failure of either the human HO-2 protein or HO-2 peptide (at concentrations as high as 1000 ng/ml) to generate any signal above background. At least a 300-fold greater signal comes from HO-1 protein as compared to the HO-2 protein. The EC(50) is about 200 ng/ml for HO-1, and the minimum detectable level of the HO-1 protein is about 1 ng/ml. The ELISA assay for the HO-1 protein requires a total of 6 h to complete. Of the total cellular HO-1 protein, 20, 19, 9 and 3% appeared in the nuclear, microsomal, mitochondrial and high speed supernatant fractions, respectively. As expected, the highest concentration of HO-1 protein per total protein in a subcellular fraction was found in the microsomes. For many research projects utilizing this ELISA assay for HO-1 protein concentration, use of the whole homogenate will be an excellent choice, rather than use of the postmitochondrial or microsomal fractions. Much higher HO-1 protein levels were found in tissues of rats rather than mice. This may be because the capture antibody and secondary antibody were both raised against the rat and not the mouse forms of the HO-1 protein. In rats the HO-1 concentrations were 1067, 364, 194, 31, 28 19, 5 and 2 ng/g tissue in whole homogenates from testes, brain, liver, lung, spleen, kidney, small intestines and urinary bladder, respectively. The ELISA assay for HO-1 described here will be useful for HO-1 research studies in tissues and cell cultures of rats and mice. This ELISA for HO-1 may also work with human tissues and cells.     

Immunogenicity issues in drug development

Immunogenicity is an important factor that manufacturers must consider as they develop new protein therapeutics. It is important to understand the immunogenicity of new proteins both at the preclinical phase and in the clinical phase of development. This paper provides an overview of the issues that manufacturers should consider including some of the potential reasons that some proteins induce an immune response, a discussion regarding current methodology used to understand immunogenicity, and some examples of marketed protein therapeutics with immunogenicity issues. Given the increasing scrutiny from regulatory agencies around the way immunogenicity is assessed by manufacturers, the strategy of detecting and characterizing antibodies that are formed against protein therapeutics is becoming an important topic. Screening assays are typically performed first on all serum samples collected in the course of a trial to detect the presence of antibodies that can bind to the protein therapeutic. There are several platforms in use: radioimmune precipitation assays (RIP), enzyme linked immunosorbent assays (ELISA), electrochemiluminescent assays (ECL), and biosensor-based assays. Each has its advantages and disadvantages, and needs to be evaluated to identify the optimal platform for a specific therapeutic protein. Once antibodies are identified, a confirmatory assay is performed to verify and characterize the antibodies. A biological assay should be used next to test if these antibodies are capable of neutralizing the biological effect of the drug. Any sample that is positive for neutralizing antibodies, indicates that the antibody is probably having an impact on the patient's ability to derive full benefit from the therapeutic protein, and may be critical for patient safety.     

Variable quantitation of Haemophilus influenzae type b anticapsular antibody by radioantigen binding assay.

The measurement of antibody to Haemophilus influenzae type b capsular polysaccharide is important in the study of natural immunity and in the immunogenicity evaluation of H. influenzae type b vaccines. Several radioantigen binding assays (RABA) have been developed to measure H. influenzae type b anticapsular antibody, but recent immunogenicity data obtained with structurally similar vaccines suggest major differences in antibody quantitation in different laboratories. To evaluate interlaboratory variability in the measurement of anticapsular antibody levels, we blindly evaluated a sample of 40 pre- and postimmunization sera by eight RABAs in different laboratories. Evaluation of RABA methods revealed differences in polysaccharide antigens, radiolabeling methods, concentration and volume of antigen and antibody, and other assay methods. The reported results of assays varied significantly between laboratories (up to sixfold differences in geometric means), in part because of differences in assay sensitivity and different proportions of samples having undetectable levels of antibody (0 to 65% of specimens with undetectable levels). After standardizing the limit of sensitivity for all assays (0.125 microgram/ml), the results of all combinations of paired analyses of RABA assays correlated well (r = 0.88 to 0.99) but the geometric mean levels still varied as much as twofold. For individual sera, the differences between paired assays often were substantial (P less than or equal to 0.0001, paired t test), with some results varying as much as 64-fold. Differences were greatest for lower levels of antibody. There was good comparability and interlaboratory reproducibility of some assays but not of others. Intrinsic or extrinsic labeling of the antigen was not a major determinant of comparability. In most instances, the current variation in the quantitation of antibody levels by these assays precludes interassay comparisons. A standardized measurement of antibody needs to be developed to adequately compare results between different H. influenzae type b immunogenicity studies.     

Collaborative study for the evaluation of enzyme-linked immunosorbent assays used to measure human antibodies to Bordetella pertussis antigens.

Acellular pertussis vaccines are being evaluated in multiple clinical studies, and human immunogenicity data will likely be pivotal in the appraisal of vaccine responses between populations and the responses to different vaccine combinations. Antibody response to pertussis antigens is also used in the diagnosis of pertussis. An international study was designed to assess the comparability of data generated in different laboratories by enzyme-linked immunosorbent assays (ELISAs). Thirty-three participating laboratories were asked to quantitate specific antibody to pertussis toxin (PT), filamentous hemagglutinin (FHA), pertactin (PRN), or fimbrial proteins (FIM) in 21 samples. Samples were to be assayed in triplicate in five independent assays by each ELISA routinely performed in the laboratory to assess intra-assay, interassay, and population variability. The mean sample values were used to compare quantitative results among the laboratories. Thirteen of the 32 laboratories which submitted evaluable data for an assay to measure antibodies to PT, 12 of 30 laboratories with assays for FHA, 10 of 17 laboratories with assays for PRN, and 6 of 13 laboratories with assays for FIM maintained a coefficient of variation below 20% for 75% of the samples tested. Assays that measure antibodies to FIM appear to be less precise than the other assays. Precision varied among laboratories that used similar methods. The relative values of intra- and interassay variabilities were not consistent for a given assay within a laboratory, indicating that the sources of these variability components may be unrelated. Precision and agreement appeared less reliable for samples with low antibody levels. Ranking and regression analyses suggest that some laboratories generated comparable quantitative results, although direct comparison or combination of results from different laboratories remains difficult to support. Calibration to the U.S. Reference Pertussis Antisera appears to have been successful at standardizing the results in some laboratories. Statistical analyses are affected by assay precision and are not necessarily reliable sole predictors of biologically relevant differences in quantitative results. If results from different laboratories must be compared, appropriate studies of precision and quantitative agreement should be conducted to support the specific comparisons.     

IDOL,inducible degrader of low-density lipoprotein receptor,serves as a potential therapeutic target for dyslipidemia

Low-density lipoprotein cholesterol (LDL-C) is the hall marker for the atherosclerotic cardiovascular disease (ASCVD). It has been shown that over 70% of circulating LDL-C is metabolized through binding and activation of hepatic LDL receptor (LDLR). Genetic LDLR mutations cause hypercholesterolemia in the patients. Therefore, elevation of LDLR levels is beneficial for the treatment of dyslipidemia. LDLR expression is regulated by the SREBP2/PCSK9 pathways. Targeting SREBP2/PCSK9 pathways by statins and human monoclonal PCSK9 antibody has been shown to reduce the progression of ASVCD. Recent studies identified that inducible degrader of LDLR (IDOL) is a novel regulator of LDLR. IDOL is an E3-ubiquitin ligase regulated via liver X receptors (LXRs) binding to the upstream of translation start site of IDOL. IDOL modulates LDLR distribution through ubiquitination and degradation of LDLR in lysosomes. Genome-wide association studies (GWAS) have revealed that the nonsynonymous substitution rs9370867 of IDOL probably contributes to the variability of circulating LDL levels. Recently studies also demonstrated that IDOL influences PCSK9 expression in a LDLR/SREBP2-dependent manner. Based upon these novel findings, we hypothesize that IDOL and PCSK9 would have a synergistic effect on LDLR distribution. Specifically, loss of IDOL increases LDLR distribution in the hepatic cell, and subsequently reduces serum LDL-C levels in dyslipidemic patients. IDOL might be a potential therapeutic target for the treatment of ASCVD.     

Comparison of two rapid diagnostic assays for detection of immunoglobulin M antibodies to dengue virus

Two easy-to-use commercial diagnostic assays, a dipstick enzyme-linked immunosorbent assay (ELISA) (Integrated Diagnostics, Baltimore, Md.) and an immunochromatographic card assay (PanBio, Brisbane, Australia) were evaluated for detection of immunoglobulin M (IgM) antibody to dengue virus with an in-house IgM antibody capture microplate ELISA as a reference assay. The dipstick ELISA was based on the indirect-ELISA format using dengue 2 virus as the only antigen and enzyme-labeled goat anti-human IgM antibody as the detector. The total assay time was 75 min. The immunochromatographic card assay was based on the antibody capture format and separately measured both anti-dengue virus IgM and IgG in the same test. Colloidal-gold-labeled anti-dengue virus monoclonal antibody bound with dengue virus 1 to 4 antigen cocktail was the detector, and anti-human IgM and IgG were the capture antibodies. The total assay time was <10 min. Sera from 164 individuals classified as either anti-dengue virus IgM positive (94) or anti-dengue virus IgM negative (70) in the reference microplate ELISA with a dengue virus 1 to 4 antigen cocktail were tested in the two commercial assays. The dipstick ELISA missed 7 of 94 positive samples, for a sensitivity of 92.6%, while the immunochromatographic card assay missed two positive samples, for a sensitivity of 97.9%. Of the 70 negative samples, four were false positive by the dipstick ELISA and two were false positive in the immunochromatographic card assay, resulting in specificities of 94.3 and 97.1%, respectively. Both commercial assays provide sensitive and specific detection of anti-dengue virus IgM antibody and could prove useful in settings where the microplate ELISA is impractical.     

Sensitive ELISA for interleukin-6. Detection of IL-6 in biological fluids: synovial fluids and sera

A monoclonal antibody and an affinity purified polyclonal antibody, both raised against recombinant human IL-6, have been employed in an ELISA procedure to quantitate human IL-6. Both antibodies were very potent in neutralizing the biological activity of recombinant as well as natural human IL-6. The monoclonal antibody was used as the capture antibody whilst the polyclonal antibody, in biotinylated form, was used as the detecting antibody in combination with a streptavidin horseradish peroxidase conjugate and a signal amplification system. The detection limit for natural as well as recombinant IL-6 was 1 pg/ml. A good correlation was found between the ELISA and the B9 biological assay when IL-6 was measured in crude culture supernatants, in synovial fluids of rheumatoid arthritis patients and in the sera of patients with diverse diseases. Immunoprecipitation of IL-6, produced by different cell types, such as monocytes, endothelial cells and smooth muscle cells or derived from biological fluids, such as the serum of a patient with septic shock or the synovial fluid of a rheumatoid arthritis patient, revealed in every case only molecules in the molecular weight range of 21,000-26,000.