Recombinant Core Protein Research Articles

Decorin-Type I Collagen Interaction: PRESENCE OF SEPARATE CORE PROTEIN-BINDING DOMAINS

Interactions between the core protein of the small dermatan sulfate proteoglycan decorin and type I collagen have been considered to influence the kinetics of collagen fibrillogenesis and the diameter of and the distance between the fibrils. A variety of recombinant core protein fragments were expressed in Escherichia coli, extracted from inclusion bodies, and renatured in the presence of bovine serum albumin, which was essential for obtaining functional activity. A recombinant protein lacking the first 14 amino acids of the mature core protein (P15-329) interacted with reconstituted type I collagen fibrils and inhibited collagen fibrillogenesis almost as efficiently as intact decorin purified from fibroblast secretions under non-denaturing conditions. Peptides comprising amino acids 15-183 (P15-183) and 185-329 (P185-329) were able to compete for the binding of wild-type decorin, with P15-183 being more active than P185-329. Several other peptides were much less effective. Binding studies using radioactively labeled peptides P15-183 and P185-329 gave direct evidence for the independent binding of both peptides. Peptides 15-183 and 15-125 had the capability of inhibiting collagen fibrillogenesis, whereas peptide 185-329 was inactive. These data suggest (i) that there are at least two separate binding domains for the interaction between decorin core protein and type I collagen and (ii) that binding is not necessarily correlated with an alteration of collagen fibrillogenesis.

Interaction of Biglycan with Type I Collagen

The small proteoglycan decorin is known to interact with type I collagen fibrils, thereby influencing the kinetics of fibril formation and the distance between adjacent collagen fibrils. The structurally related proteoglycan biglycan has been proposed not to bind to fibrillar collagens. However, when osteosarcoma cells were cultured on reconstituted type I collagen fibrils, both decorin and biglycan were retained by the matrix. Immunogold labeling at the electron microscopic level showed that both proteoglycans were distributed along collagen fibrils not only in osteosarcoma cell-populated collagen lattices but also in human skin. Reconstituted type I collagen fibrils were able to bind in vitro native and N-glycan-free biglycan as well as recombinant biglycan core protein. From Scatchard plots dissociation, constants were obtained that were higher for glycanated biglycan (8.7 x 10(-8) mol/liter) than for glycanated decorin (7 x 10(-10) mol/liter and 3 x 10(-9) mol/liter, respectively). A similar number of binding sites for either proteoglycan was calculated. Recombinant biglycan and decorin were characterized by lower dissociation constants compared with the glycanated forms. Glycanated as well as recombinant decorin competed with glycanated biglycan for collagen binding, suggesting that identical or adjacent binding sites on the fibril are used by both proteoglycans. These data suggest that, because of its trivalency, biglycan could have a special organizing function on the assembly of the extracellular matrix.

High efficiency prokaryotic expression and purification of a portion of the hepatitis C core protein and analysis of the immune response to recombinant protein in BALB/c mice.

Hepatitis C virus (HCV) produces chronic persistent liver infection in 1-2% of the U.S. population and is the leading cause of end stage liver disease in patients presenting for liver transplantation at our center. Efforts to cure persistent HCV infection are frequently unsuccessful, so the development of a HCV vaccine is a high priority. HCV envelope proteins are hypervariable so production of a recombinant surface antigen vaccine such as is available for hepatitis B is not likely to confer widespread, high level protective immunity. As the most highly conserved structural protein in the HCV genome, the core protein is one reasonable target for vaccine production. Presented here are data on the manufacture of recombinant core protein containing partial carboxy terminus deletions in an effort to increase the efficiency of core expression. The maltose binding protein (MBP) and glutathione S-transferase (GST) protein prokaryotic expression systems were used to study two different constructs, expressing the first 140 and 163 amino acids of the core region. Deletion of the 23 amino acids (aa) from aa141-163 led to a marked increase in the efficiency of protein production from < 1 to 3-4 mg/liter for both systems studied. Protein purification was accomplished using affinity chromatography (MBP) or inclusion body isolation (GST) as determined by SDS-PAGE gels and immunotransblot with HCV core protein-specific monoclonal antibody. Finally, the immune response to recombinant protein was assessed in BALB/c mice using a MBP HCV core fusion protein and an ELISA developed using GST HCV core protein as a target. In all mice of this strain, serum anti-HCV core antibody titer increased to 10(-4), two logs above background, following immunization in conjunction with Freund's complete adjuvant. These results represent an encouraging first step toward production of a core protein vaccine. Recombinant core protein is a useful tool to study the immune response to core protein and may be useful to further study the epidemiology and biology of the HCV virus.

Genetic control of the murine humoral response to distinct epitopes of hepatitis C virus core protein.

Recombinant hepatitis C virus (HCV) core protein from aa1-164, designated cp1-10, was used to immunize mice. Antibodies to cp1-10 were produced in all seven strains of congenic mice; none of the strains could be considered low responders relative to the others. The mouse response against individual epitopes of HCV core protein varied from one strain to another: B10.RIII (H-2r) recognized all three peptides aa13-30, aa77-90, aa129-145; B10.D2 (H-2d), B10 (H-2b) and C3H.SW (H-2b) responded to aa13-30, aa77-90; B10.M (H-2f), B10.BR (H-2k) and C3H/Hej (H-2k) reacted with aa13-30 only. Competitive inhibition of binding demonstrated that antibody to the peptide was inhibited by cp1-10 protein and the corresponding peptide only. Recombinant HCV core protein is highly immunogenic and can elicit good antibody response in mice. The aa13-30 is a major epitope of HCV core protein in mice. The humoral response to the distinct epitopes was regulated by the H-2 genes. Further analysis indicated that the I-a locus of H-2 genes determined the antibody response to aa13-30 and 77-90. These results suggest that the variation of antibody responses to HCV in humans may partially contribute to different outcomes of HCV infection.

Effective production of the hepatitis C virus core antigen having high purity in Escherichia coli

The amino-terminal half of putative nucleocapsid (core) protein (amino acids 1–115) of hepatitis C virus (HCV) was directly overproduced in Escherichia coli under the control of the tac promoter. Overproduction of core antigen was achieved by inserting several target genes and by optimizing the culture conditions, whereas a large amount of directly expressed and purified core antigen has not yet been reported. Although the level of expression was comparable to that of the conventional E. coli fused expression system, our recombinant proteins contain only HCV amino acid sequence. Using recombinant E. coli, overproduced large-scale culture system was achieved in jar-fermenter. A highly purified sample of the expressed protein was obtained by ion-exchange and gel permeation column chromatography in the presence of 8 M urea. From a 3.5 1 culture, ~ 440 mg of recombinant core protein was obtained after a two-step purification procedure. An enzyme-linked immunosorbent assay developed using the highly purified antigen satisfactorily diagnosed hepatitis C.

Hepatitis C virus antibody prevalence among human immunodeficiency virus‐1‐infected individuals: Analysis with different test systems

Sera of 383 human immunodeficiency virus (HIV)-1-infected individuals from Frankfurt (Main)/Germany were assayed by two hepatitis C virus (HCV) screening tests (Abbott second generation, Ortho second generation). This population showed a prevalence for reactivity with both tests of 20.8% (80/383). Examination of all reactive sera (91/383) by a supplemental assay (Chiron RIBA 2) gave for 46 sera a positive, for 33 sera an indeterminate, and for 12 sera a negative result. Further analysis focussed on these RIBA 2-indeterminate and -negative samples. Analysis of the sera using an in-house Western blot with three different Escherichia coli-expressed HCV proteins revealed that none of the RIBA 2-negative, but 24 of the 33 RIBA 2-indeterminate sera, including 3 of 4 c33c (NS3)-reactive samples, were reactive with a recombinant core protein. Twenty-one of 22 c22-3 (core) indeterminates stained the core antigen in the in-house Western blot and 3 of them in addition a NS5 moiety. HCV-polymerase chain reaction (PCR) was positive for 14 of the 24 RIBA 2-indeterminate sera, but for none of the RIBA 2-negative or Western blot nonreactive samples. Discrepant results between the two screening tests could not be explained by differences in the antigen compositions (i.e., a NS3-NS4 moiety of 111 amino acids present in the Ortho enzyme-linked immunosorbent assay (ELISA), not present in the Abbott or RIBA 2 assays).

Immunoglobulin M and A antibodies to hepatitis C core antigen in chronic hepatitis C virus infection.

To determine the prevalence and clinical significance of IgM and IgA antibody to hepatitis C virus (HCV) core antigen in chronic HCV infection, sera from 47 patients were tested for immunoglobulin class M (IgM) and immunoglobulin A (IgA) antibody to HCV core antigen by solid-phase enzyme-linked immunoassay using a recombinant core protein (aa1-150). Results were correlated with the clinical, biochemical and histological parameters, serum HCV RNA levels (determined by branched DNA signal amplification assay), and subsequent clinical response to interferon-alpha therapy. IgM anti-HCV core was detected in 11 patients (23.4 percent). There was no correlation between the presence of IgM anti-HCV core and the clinical features (sex, age, mode of acquisition), biochemical parameters (serum ALT, AST, alkaline phosphatase, and albumin level), autoimmune markers [serum globulin levels, anti-nuclear antibody (+ at < 1:80 in 7/47 patients)], serum HCV RNA levels, subsequent response to interferon-alpha therapy, and the histological features. Immunoglobulin A anti-HCV core was not detected in any of the patients. The presence of IgM ant-HCV core in a proportion of patients with chronic HCV infection indicates that the presence of serum IgM anti-HCV core may not be unique to acute HCV infection.

Isolation and epitope characterization of human monoclonal antibodies to hepatitis C virus core antigen.

In this study we describe the establishment of two hybridoma cell lines secreting human monoclonal antibodies to the 22-kD nucleocapsid protein (core, p22) of the hepatitis C virus (HCV). For this purpose we isolated B lymphocytes from an anti-HCV positive blood donor and infected them with Epstein-Barr (EBV). We obtained several lymphoblastoid cell clones secreting antibodies to the recombinant HCV core protein. The B-cell cultures were oligoclonally expanded and two of them were fused with the (mouse:human) heteromyeloma cell line K6H6/B5. The resulting stable hybridomas produce antibodies of the IgG1/kappa (U1/F10) and the IgM/kappa (Ul/F11) isotype reacting specifically with the recombinant core protein p22. To identify the epitopes recognized by these antibodies we synthesized overlapping peptides (13-mer and 6-mer) from the amino terminus of the core amino acid sequence. Antibody reactivity to these peptides was analyzed in an immunoblot assay. Finally, we were able to define a linear epitope recognized by the Ul/F10 antibody on the nucleocapsid protein. The antibody shows specificity to the sequence N-VYLLPR-C, which corresponds to the amino acids 34-39 of the core sequence.

T-cell response to structural and nonstructural hepatitis C virus antigens in persistent and self-limited hepatitis C virus infections

Twenty-nine patients with chronic hepatitis C and 15 asymptomatic hepatitis C virus antibody-positive subjects who clinically recovered from hepatitis C virus infection were studied for their peripheral blood lymphomononuclear cell proliferative response to hepatitis C virus structural and nonstructural antigens (core, envelope, nonstructural 4 and nonstructural 5) expressed in yeast as superoxide dismutase fusion proteins, in an initial attempt to define some of the features of the virus-specific immune response. Hepatitis C virus core was the most immunogenic antigen for human leukocyte antigen class II-restricted T cells in both groups of patients studied, and the proliferative response to it was the most vigorous and the most frequently expressed in comparison with the other antigens tested. The specificity of the results was supported by the lack of response to hepatitis C virus antigens by healthy uninfected controls and confirmed by recognition of recombinant core proteins of different origin (yeast and baculovirus) by polyclonal T-cell lines produced by T-cell stimulation with yeast-derived core. Each of the antigens tested was able to induce significant although variable levels of proliferative response, indicating that all can be immunogenic at the T-cell level. Significant proliferative responses to core, nonstructural 4 and nonstructural 5 antigens were more frequently detected in subjects who were able to eradicate infection than in patients with chronic hepatitis C, although the difference was statistically not significant. No difference was observed between the two groups of patients with respect to the response to the putative envelope antigens.

T-cell response to structural and nonstructural hepatitis C virus antigens in persistent and self-limited hepatitis C virus infections

Twenty-nine patients with chronic hepatitis C and 15 asymptomatic hepatitis C virus antibody–positive subjects who clinically recovered from hepatitis C virus infection were studied for their peripheral blood lymphomononuclear cell proliferative response to hepatitis C virus structural and nonstructural antigens (core, envelope, nonstructural 4 and nonstructural 5) expressed in yeast as superoxide dismutase fusion proteins, in an initial attempt to define some of the features of the virus-specific immune response. Hepatitis C virus core was the most immunogenic antigen for human leukocyte antigen class II–restricted T cells in both groups of patients studied, and the proliferative response to it was the most vigorous and the most frequently expressed in comparison with the other antigens tested. The specificity of the results was supported by the lack of response to hepatitis C virus antigens by healthy uninfected controls and confirmed by recognition of recombinant core proteins of different origin (yeast and baculovirus) by polyclonal T-cell lines produced by T-cell stimulation with yeast-derived core. Each of the antigens tested was able to induce significant although variable levels of proliferative response, indicating that all can be immunogenic at the T-cell level. Significant proliferative responses to core, nonstructural 4 and nonstructural 5 antigens were more frequently detected in subjects who were able to eradicate infection than in patients with chronic hepatitis C, although the difference was statistically not significant. No difference was observed between the two groups of patients with respect to the response to the putative envelope antigens. (Hepatology 1994;19:286–295).

Processing and Subcellular Distribution of the Schwann Cell Lipid-Anchored Heparan Sulfate Proteoglycan and Identification as Glypican

We previously identified a phosphatidylinositol-specific phospholipase c (PI-PLC)-releasable heparan sulfate proteoglycan (HSPG) on the surface of rat Schwann cells (D. J. Carey and R. C. Stahl, J. Cell Biol. 111, 2053-2062, 1990). The present study was carried out to investigate the localization and processing of this proteoglycan. The HSPG was synthesized as a PI-PLC-releasable form that was shed into the culture medium with a T1/2 of 17 h. Degradation of the HSPG was negligible. The HSPG was present on the surface of Schwann cells on small (100-200 nm diameter) cylindrical membrane extensions that resembled filopodia. In neonatal peripheral nerve, brain, heart, and striated muscle the HSPG was found to be localized principally to regions of the cell surface that were in contact with basement membranes. Northern blot analysis with cDNA coding for rat glypican (a previously described human fibroblast HSPG) demonstrated abundant expression of glypican mRNA in Schwann cells. Antibodies made against recombinant rat glypican core protein immunoprecipitated the Schwann cell PI-PLC-releasable HSPG. These data demonstrate that the Schwann cell HSPG is rat glypican and support the hypothesis that this proteoglycan functions in cell-extracellular matrix interactions.

Hepatitis C virus c100 antigen in liver tissue from patients with acute and chronic infection

A pool of murine monoclonal antibodies developed against c100 antigen, a hepatitis C virus–associated protein encoded by the NS3/NS4 virus genome, was used to detect hepatitis C virus in liver biopsy specimens from patients with acute and chronic hepatitis C virus infection. The antigen was present in the cytoplasm of liver cells only. The immunoreactive signal appeared as large, distinct, brilliant fluorescent granules with no clear relationship to cellular structures. No obvious membrane c100 antigen accumulation was observed. Distribution of c100-containing hepatocytes was directly correlated with viral replication in acute hepatitis. All three acute-hepatitis patients were positive for hepatitis C virus RNA (as detected on polymerase chain reaction) in serum and displayed c100 antigen in 50% to 70% of hepatocytes, with a distinct topographical relationship with necrotic areas and inflammatory cell accumulation. Conversely, very low numbers of infected cells and no relationship between tissue c100 antigen expression and sites of liver cell necrosis and inflammation were found in 14 chronic hepatitis C virus infection patients. Furthermore, though all patients had measurable levels of serum hepatitis C virus RNA, only eight (57%) had detectable c100 antigen in liver sections. Indeed, these two distinct immunopathological patterns were inversely related to the development of c100 antibody in serum. Specificity of hepatocellular c100 antigen deposits was established through extensive absorption experiments using structural and nonstructural hepatitis C virus recombinant proteins. However, tissue processing was found to be a crucial step in the demonstration of hepatitis C virus antigen in fresh frozen liver tissue. Furthermore, monoclonal antibodies against recombinant core and NS3-encoded proteins failed to demonstrate immunofluorescent staining, suggesting that the c100 protein is the main antigen demonstrable by this method. (HEPATOLOGY 1993;18:240–245).

Murine HIV-1 p24 specific T lymphocyte activation by different antigen presenting cells: B lymphocytes from immunized mice present core protein to T cells.

Mice were injected with three doses of baculovirus-produced recombinant HIV-1 p24 core protein in alum adjuvant. CD4 positive T lymphocytes from immunized animals proliferated in vitro in the presence of antigen and peritoneal macrophages (Mps) or splenic dendritic cells (DCs) from non-immunized mice as antigen presenting cells (APCs). DCs were approximately three times more efficient than Mps on a cell for cell basis. No synergy was observed between Mps and DCs in this system. B lymphocytes from immunized animals also presented p24 antigen to the specific T cells. Mps did synergize with B cells to enhance the level of T lymphocyte proliferation. This may have implications for the induction of specific immune responses to pathogens after administration of single protein vaccines.

HCV RNA and antibody to HCV core protein in Japanese patients with chronic liver disease.

We evaluated the prevalence of hepatitis C virus (HCV) RNA and antibody (anti-HCVcore) to the putative HCV core protein in Japanese patients with chronic liver disease. Sera were screened by solid-phase enzyme immunoassay with a recombinant HCV core protein and by the reverse transcription-polymerase chain reaction (RT-PCR) test which directly detects the HCV genome. Anti-HCV core was detected with high titers in 95% (69/73) of chronic non-A, non-B hepatitis, and 94% (65/69) of anti-HCVcore-positive patients had the genome. Anti-HCVcore was also found with lower titers in 24% (10/41) of chronic hepatitis B virus carriers, and three of them had the genome. Only one (3%) of the 35 patients negative for anti-HCVcore tested positive to RT-PCR. These findings indicate the overwhelming prevalence of HCV infection in Japanese patients with chronic non-A, non-B hepatitis and a close relationship between the presence of anti-HCVcore and chronic hepatitis C in this population.

Differential expression of cell surface heparan sulfate proteoglycans in human mammary epithelial cells and lung fibroblasts.

Treating the liposome-intercalatable heparan sulfate proteoglycans from human lung fibroblasts and mammary epithelial cells with heparitinase and chondroitinase ABC revealed different core protein patterns in the two cell types. Lung fibroblasts expressed heparan sulfate proteoglycans with core proteins of approximately 35, 48/90 (fibroglycan), 64 (glypican), and 125 kDa and traces of a hybrid proteoglycan which carried both heparan sulfate and chondroitin sulfate chains. The mammary epithelial cells, in contrast, expressed large amounts of a hybrid proteoglycan and heparan sulfate proteoglycans with core proteins of approximately 35 and 64 kDa, but the fibroglycan and 125-kDa cores were not detectable in these cells. Phosphatidylinositol-specific phospholipase C and monoclonal antibody (mAb) S1 identified the 64-kDa core proteins as glypican, whereas mAb 2E9, which also reacted with proteoglycan from mouse mammary epithelial cells, tentatively identified the hybrid proteoglycans as syndecan. The expression of syndecan in lung fibroblasts was confirmed by amplifying syndecan cDNA sequences from fibroblastic mRNA extracts and demonstrating the cross-reactivity of the encoded recombinant core protein with mAb 2E9. Northern blots failed to detect a message for fibroglycan in the mammary epithelial cells and in several other epithelial cell lines tested, while confirming the expression of both glypican and syndecan in these cells. Confluent fibroblasts expressed higher levels of syndecan mRNA than exponentially growing fibroblasts, but these levels remained lower than observed in epithelial cells. These data formally identify one of the cell surface proteoglycans of human lung fibroblasts as syndecan and indicate that the expression of the cell surface proteoglycans varies in different cell types and under different culture conditions.

Characterization of nucleotide sequences from European hepatitis C virus isolates

We characterized the 5' end and parts of the structural genes of European isolates of hepatitis C virus (HCV) and compared them with recently published RNA sequences of American and Japanese HCV isolates. The cDNA, obtained by reverse transcription of viral RNA extracted from different sera, was amplified by nested PCR, cloned and sequenced. Within 239 nucleotides (nt) of the 5' end, we found only three single-nt exchanges compared to two sequences of Japanese origin and one exchange to the prototype HCV sequence (ptHCV) (homology > 99%). The sequence of the core region (534 nt) in two European isolates showed a homology of about 97–98% on the nt level, as compared to ptHCV and one Japanese isolate, and 90% to other Japanese isolates. The amino acid (aa) homology was between 98–99% among all published sequences. A greater discrepancy was found in the European isolates within the 434 nt sequenced from the N-terminus of the putative envelope region, where the nt homology to ptHCV and one Japanese isolate was 90–93% (aa homology 93–95%), and to other Japanese isolates was 72–73% (aa homology 77–78%,), indicating that the European isolates may be more closely related to the ptHCV and one Japanese isolate than to the other Japanese isolates. Amplified genes encoding structural proteins ( core, envelope) were expressed in Escherichia coli. Sera from chronically infected patients reacted strongly with the recombinant core protein, but no specific immunoreactivity occurred with the putative envelope protein.

AIDS recombinant vaccine; recombinant HIV virus core protein cloning in vaccinia virus vector

AIDS recombinant vaccine; recombinant HIV virus core protein cloning in vaccinia virus vector

Detection of antigen—antibody interactions by surface plasmon resonance. Application to Epitope Mapping

Surface plasmon resonance (SPR) detection requires no labeling of antigen or antibodies and allows quantification of two or more interacting molecular species. The automated SPR instrument used here consists of an optical detection unit, an integrated liquid handling unit, and an autosampler. A first molecule is immobilized to the dextran modified surface of the sensor chip. By sequential introduction, the stepwise formation of multimolecular complexes can then be monitored. A two-site binding assay which allows characterization of MoAb epitope specificities is described. A polyclonal rabbit anti-mouse IgG1 (RAMG1) immobilized to the dextran surface is used to capture the first MoAb from unprocessed hybridoma culture supernatants. After introducing the antigen, the ability of a second MoAb to bind to the antigen is tested. The analysis cycle which is fully automated can be performed more than 100 times using the same RAMG1 surface. Since the detection principle allows monitoring of each reactant in the consecutive formation of a multimolecular complex, multi-site binding experiments can be performed. Five MoAbs recognizing different epitopes on an antigen were shown to bind sequentially, forming a hexamolecular complex. MoAbs were further characterized by inhibition analysis using synthetic peptides derived from the primary structure of their antigen. As a model system MoAbs against recombinant HIV-1 core protein p24 were used in all experiments.

Antibodies from patients with rheumatoid arthritis and systemic lupus erythematosus recognize different epitopes of a single heterogeneous nuclear rnp core protein

Antibodies to recombinant heterogeneous nuclear RNP core protein A1 were detected in sera from 27 of 58 patients with rheumatoid arthritis (RA) and from 7 of 31 patients with systemic lupus erythematosus, by immunoblotting and enzyme-linked immunosorbent assay. Protein A1 consists of 2 distinct domains: The N-terminal sequence is identical to a single-stranded DNA binding protein termed UP1, and the C-terminal domain shows a partial homology with keratin. All 7 A1-positive systemic lupus erythematosus sera reacted with UP1, whereas 9 of the 27 A1-positive RA sera did not. In RA, anti-A1 activity was significantly associated with antikeratin antibodies (AKA); these antibodies were present in 23 of 27 A1-positive sera and 10 of 31 A1-negative sera (P less than 0.01). Immunoabsorption with recombinant protein A1 resulted in a significant reduction of AKA titers in 6 of 10 RA sera tested, suggesting that AKA from RA patients may cross-react with the C-terminal portion of the heterogeneous nuclear RNP protein A1.

Multicenter evaluation of a new recombinant enzyme immunoassay for the combined detection of antibody to HIV-1 and HIV-2

A newly developed recombinant antigen-based anti-HIV-1/HIV-2 enzyme immunoassay (Abbott Recombinant HIV-1/HIV-2 EIA) was evaluated against a second generation anti-HIV-1 EIA (Abbott Recombinant HIV-1 EIA). Five thousand and twenty-nine sera from European blood donors and 403 sera from central African blood donors were used in the evaluation, along with four panels and one cohort. The panels included 99 'problem' sera, 733 sera with antibodies to HIV-1 from asymptomatic people and from patients at different disease stages, 25 serial bleeds from five plasmapheresis donors seroconverting for antibodies to HIV-1, and 202 sera with antibodies to HIV-2 collected from healthy and diseased people of European or west African origin. In addition, 734 sera collected from a west African cohort were tested. Using Western blot as the reference standard, the specificity obtained by the recombinant anti-HIV-1 EIA (HIV-i EIA) was 99.90% [99.81-99.99%; 95% confidence limits (95% CL)] with European blood donor sera; 99.50% (98.78-100%) with Central Africa blood donor sera; 92.93% (87.78-98.08%) with 'problem' sera and 99.43% (98.87-100%) with sera from a west African cohort. Using the same samples, the recombinant anti-HIV-1/HIV-2 EIA (HIV-1/HIV-2 EIA) yielded a specificity of 99.84% (99.73-99.95%), 99.50% (98.78-100%), 95.96% (92.00-99.92%) and 98.58% (97.69-99.47%), respectively. All 776 Western blot-confirmed anti-HIV-1 sera were reactive in both EIAs, and the EIA-reactive samples from seroconverting plasma donors were always observed for both assays in the same serial bleed. For HIV-2, the HIV-1 EIA yielded an overall sensitivity of 75.83% (69.93-81.72%) compared with 99.53% (98.58-100%) for HIV-1/HIV-2 EIA. The addition of a recombinant env-protein of HIV-2 to the recombinant env and core proteins of HIV-1 on the solid phase of HIV-1 EIA improved the detection of anti-HIV-2 while preserving the assay's overall specificity and sensitivity for the detection of anti-HIV-1.