Differential expression of murine CD81 highlighted by new anti-mouse CD81 monoclonal antibodies

We describe the use of a soluble CD81-Fc fusion protein to screen for novel monoclonal antibody (MAb) reactive with the extracellular loops of murine CD81 (TAPA-1). Two such MAbs, Eat1 and Eat2 (for Extracellular Anti-TAPA1), were used to assess the expression and function of CD81 on murine lymphocytes. Although CD81 is expressed uniformly on all human lymphocytes, murine CD81 was found to be expressed at much higher levels on resting B cells than on resting T cells. This was particularly evident when staining with the new MAbs, Eat1 and Eat2. The molecule is also functionally active on B cells, as Eat1 and Eat2 induce homotypic adhesion of B lymphocytes. Stimulated B cells undergo early apoptotic events in the presence of Eat2, as shown by binding of Annexin V-fluorescein isothiocyanate (FITC). Polyclonal activation of murine T cells also induces higher level CD81 expression, and many immortalized murine T-cell lines express high levels of the protein. In contrast to human CD81, which is expressed equally on all thymocytes, murine CD81 is induced during thymic development, being expressed at high levels on CD4+CD8+ thymocytes, in contrast to other subsets of thymocytes. Finally, murine dendritic cells, splenic macrophages, and non-killer (NK) cells all express high levels of CD81. We conclude that CD81 is differentially expressed in the murine immune system, and is involved in regulating the adhesion and activation of murine B cells.     

T cell costimulation by the hepatitis C virus envelope protein E2 binding to CD81 is mediated by Lck

Binding of the hepatitis C virus (HCV) envelope protein E2 to CD81 provides a costimulatory signal for human T cells. This phenomenon may play a role in liver damage and autoimmune manifestations associated with HCV infection. Here we show that cross-linking of CD81 by HCV E2 induced a calcium flux in T cells that depends on Lck since it was blocked by PP1 and absent in Lck-deficient Jurkat T cells. In wild-type Jurkat cells, Lck was activated by CD81 cross-linking, and CD81, like Lck, was found in lipid rafts. Indeed, the integrity of the raft compartment was required for the induction of a calcium flux by E2, since methyl-beta-cyclodextrin abolished this response. A requirement for TCR/CD3 expression was indicated by the absence of a calcium flux following E2 stimulation of TCR/CD3-deficient Jurkat cells. CD81 cross-linking increased and prolonged the anti-CD3-induced tyrosine phosphorylation of TCR1 and of other proteins, indicating that the CD81-mediated signal converges with the TCR/CD3 signaling cascade at its most upstream step. In conclusion, we propose that the costimulatory effects of HCV E2 on T cells depend on CD81 cross-linking that activates Lck through raft aggregation and thus leads to enhanced TCR signaling.     

Isolation and characterization of human monoclonal antibodies against hepatitis C virus envelope glycoproteins

The isolation and characterization of human monoclonal antibodies (humAbs) against the hepatitis C Virus (HCV) glycoproteins E1 and E2 are described. B-cells from blood donors with anti-HCV were transformed with Epstein-Barr virus. The supernatants of the resulting lymphoblastoid clones were screened by ELISA with an extract of cells infected with a recombinant vaccinia virus RMPA95 expressing the envelope proteins E1 and E2 of an HCV genotype 1a virus (H strain). Positive clones were fused to the heteromyeloma cell line K6H6/B5. Fifteen heterohybridoma cell lines have been established. The specificity of the isolated humAbs was determined both by ELISA and Western blot assays. Several recombinant extracts expressing either the E1 or E2 protein or truncated forms were used in an attempt to map the epitopes on the viral glycoproteins. Some of the humAbs were used successfully for immunofluorescence investigation of transfected cells. Seven specific anti-E2 humAbs, which react with the envelope protein 2 of genotype 1a and 1b isolates, were characterized. J. Med. Virol. 55:28–34, 1998. © 1998 Wiley-Liss, Inc.     

Predominance of antibodies to hepatitis C virus envelope proteins in various disease statuses of hepatitis C

The antibody profile to various proteins of hepatitis C virus (HCV) was studied in 113 patients positive for HCV RNA in various disease statuses of hepatitis C (HC). A single peptide (E2/NS1, aa 413-436 of HCV polyprotein) chosen from a conserved region at the C-terminus of the hypervariable region (HVR) HVR1 of HCV was found to be sufficient for reliable diagnosis of the infection, even in the acute phase. Six hundred and one suspected HC cases and 200 voluntary blood donors were tested by this peptide. The sensitivity of detection of HCV antibodies by this peptide did not increase with addition of peptides from other HCV proteins. Our results clearly demonstrate that antibodies to HCV envelope proteins occur in a higher percentage of the infected population than those to other proteins. This emphasizes the necessity of using representative sequences from HCV envelope proteins in diagnostic immunoassays of this viral infection.     

Subclass distribution of human myeloma proteins as determined with monoclonal antibodies

Subclasses of 176 IgG and 62 IgA myeloma proteins were determined by indirect ELISA with monoclonal antibodies, as well as by an immunoblotting technique (for monoclonal IgG) and by immunoelectrophoresis against the lectin jacalin (for IgA). The subclass distribution of monoclonal IgG did not reflect mean normal serum levels of the correspondent isotypes, with an over-representation of IgG1 and IgG4 and an under-representation of IgG2 and IgG3 in myeloma. Similarly, IgA2 myeloma were clearly under-represented.     

Small molecule inhibition of hepatitis C virus E2 binding to CD81

The hepatitis C virus (HCV) is a causal agent of chronic liver infection, cirrhosis, and hepatocellular carcinoma infecting more than 170 million people. CD81 is a receptor for HCV envelope glycoprotein E2. Although the binding of HCV-E2 with CD81 is well documented the role of this interaction in the viral life cycle remains unclear. Host specificity and mutagenesis studies suggest that the helix D region of CD81 mediates binding to HCV-E2. Structural analysis of CD81 has enabled the synthesis of small molecules designed to mimic the space and hydrophobic features of the solvent-exposed face on helix D. Utilizing a novel bis-imidazole scaffold a series of over 100 compounds has been synthesized. Seven related, imidazole-based compounds were identified that inhibit binding of HCV-E2 to CD81. The inhibitory compounds have no short-term effect on cellular expression of CD81 or other tetraspanins, do not disrupt CD81 associations with other cell surface proteins, and bind reversibly to HCV-E2. These results provide an important proof of concept that CD81-based mimics can disrupt binding of HCV-E2 to CD81.     

A comparative immunogenicity study of HIV-1 virus-like particles bearing various forms of envelope proteins, particles bearing no envelope and soluble monomeric gp120

To assess the potential of native Envelope glycoprotein (Env) trimers as neutralizing antibody vaccines, we immunized guinea pigs with three types of VLPs and soluble gp120. Particles included "SOS-VLPs" (bearing disulfide-shackled functional trimers), "UNC-VLPs" (bearing uncleaved nonfunctional Env) and "naked VLPs" (bearing no Env). The SOS-VLPs were found to have a density of about 27 native trimers per particle, approximately twice that of live inactivated HIV-1 preparations. As immunogens, UNC- and SOS-VLP rapidly elicited anti-gp120 antibodies focused on the V3 loop and the gp120 coreceptor binding site. Reactivity to the gp41 immunodominant domain was absent in SOS-VLP sera, presumably because gp120-gp41 association is stabilized, effectively covering this epitope. Gp120-immune sera reacted with the receptor binding sites of gp120 and were less focused on the V3 loop. Some Env-VLP sera neutralized primary isolates at modest titers. The measurement of neutralization was found to be affected by the cell lines used. Depending on the assay particulars, non-Env specific antibodies in VLP sera could enhance infection, or nonspecifically neutralize. However, a neutralization assay using TZM-BL cells was essentially clear of these effects. We also describe a native trimer binding assay to confirm neutralization activity in a manner that completely eliminates nonspecific effects. Overall, our data suggests that Env-VLP sera were primarily focused on nonfunctional forms of Env on VLP surfaces, possibly gp120/gp41 monomers and not the trimers. Therefore, to make progress toward a more effective VLP-based vaccine, we will need to find ways to refocus the attention of B cells on native trimers.     

Recombinant hepatitis C virus-like particles expressed by baculovirus: utility in cell-binding and antibody detection assays

Hepatitis C virus (HCV) is difficult to study due to the lack of an efficient cell culture system or small animal model. As a result, HCV-cell interactions are not well-defined. In addition, several studies have identified a subset of patients in whom HCV RNA is present, but HCV antibody is not detected. We produced recombinant baculoviruses that expressed HCV structural proteins (core, E1 and E2, nt 342-2651) or control proteins. The HCV structural protein precursor was processed into immunoreactive proteins of appropriate size, and sucrose density sedimentation and electron microscopy of infected cell lysates demonstrated particle formation. To evaluate HCV antigenicity, particularly in patients who tested negative for HCV antibody in commercial HCV immunoassays but had persistent viremia, we evaluated the virus-like particles (VLPs) in solid-phase immunoassays. VLPs reacted with sera from HCV antibody positive subjects in these solid phase immunoassays, but not with control sera. Plasma samples from 19% (5/26) of HCV antibody negative subjects who were persistently HCV RNA positive also reacted with the HCV VLPs. When incubated with MOLT-4 cells at 4 degrees C, HCV VLPs demonstrated cell binding, and behaved similar to plasma-derived HCV preparations in a flow cytometry-based cell binding assay. These data suggest that recombinant HCV VLPs may allow identification of HCV antibody in patients, including some patients with persistent viremia and who are seronegative with current assays. In addition, HCV VLPs seem useful for evaluating HCV-cell interactions.     

Structural analysis of monoclonal anti-arsonate antibodies: idiotypic specificities are determined by the heavy chain

Hybrid immunoglobulin molecules were constructed from the isolated heavy and light chains of monoclonal anti-arsonate antibodies which differed in their expression of a cross-reactive idiotype. The reconstructed molecules were tested for public and private idiotypic determinants associated with the A-strain response to the hapten p-azophenylarsonate and it was found that both an appropriate heavy chain and an appropriate light chain were required for expression of idiotypic determinants. While appropriate heavy chains could be derived only from idiotype-positive antibodies, appropriate light chains could be derived not only from idiotype positive antibodies, but also from certain idiotype-negative molecules. Similarly, recombinant (hybrid) molecules constructed from two idiotype-positive immunoglobulins differing quantitatively in the degree of idiotypic character, expressed the character at a level which correlated with that of the heavy chain donor. Thus, while the serological expression of idiotypic determinants occurs only in the context of the appropriate VHVL combination, these data demonstrate that the determinants comprising the major cross-reactive idiotype of the anti-arsonate response of A/J mice have their bases in heavy chain structures.     

Binding of the hepatitis C virus envelope protein E2 to CD81 provides a co-stimulatory signal for human T cells

Chronic hepatitis C virus (HCV) infection frequently develops into liver disease and is accompanied by extra-hepatic autoimmune manifestations. The tetraspanin CD81 is a putative HCV receptor as it binds the E2 envelope glycoprotein of HCV and bona fide HCV particles. Here we show that HCV E2 binding to CD81 on human cells in vitro lowers the threshold for IL-2 receptor alpha expression and IL-2 production, resulting in strongly increased T cell proliferation. HCV E2-induced co-stimulation also enhances the production of IFN-gamma and IL-4 and causes increased TCR down-regulation. This suggests that binding of HCV particles to CD81 on T cells in vivo may lead to activation by otherwise suboptimal stimuli. Therefore, co-stimulation of autoreactive T cells by HCV may contribute to liver damage and autoimmune phenomena observed in HCV infection.     

Increased heterogeneity of serum thyroglobulin in thyroid cancer patients as determined by monoclonal antibodies

Summary We investigated the immunological heterogeneity of plasma Tg in thyroid cancer patients using monoclonal antibodies in an immunoradiometric assay and a conventional RIA system with a polyclonal rabbit antibody. The results were compared with measurements of plasma Tg in patients with nonmalignant disease. We can demonstrate an increased immunological heterogeneity in tumor patients compared with patients with non-malignant thyroid diseases. In one case the Tg value measured by a monoclonal test system exceeded the value obtained by a polyclonal RIA system in the same sample by a factor of 25. It has to be further investigated whether this increase in heterogeneity is of diagnostic value in the follow-up of thyroid cancer patients.Abbreviations IRMA 11, IRMA 13 Immunoradiometric assay using mabTg 11 or mabTg 13 - mabTg monoclonal antithyroglobulin antibodies - Tg Thyroglobulin     

CD81 nucleotide mutation in hepatocellular carcinoma and lack of CD81 polymorphism in patients at stages of hepatitis C virus infection

Mechanisms determining the chronicity or the pattern of clinical course of hepatitis C virus (HCV) infections have not been clarified. Recently, CD81 was reported to bind the E2 protein of HCV and was suggested to function as a cellular receptor for HCV. Accordingly, the hypothesis was examined that CD81 polymorphism, if it exists, might correlate with certain clinical courses of HCV infection. CD81 cDNA sequences were determined from peripheral blood mononuclear cells (PBMCs). Twenty-four Japanese subjects were enrolled initially as follows: patients with chronic hepatitis C without cirrhosis (n = 3), patients with cirrhosis (n = 3), patients with cirrhosis complicated by hepatocellular carcinoma (HCC) (n = 3), patients with persistent HCV viremia without ALT elevation (n = 3), those with positive anti-HCV antibodies without evidence of HCV viremia (n = 3), and healthy volunteers (n = 9). In all PBMCs samples analyzed, no polymorphism was found in the CD81 cDNA sequence. The sequence was different, however, from the one reported previously at three nucleotide positions: a transversion to thymine instead of cytosine at nt 1130, a deletion at nt 1206, and a guanine insertion at nt 71. Subsequently, CD81 cDNA sequences from PBMCs and HCC tissue were compared among the other 6 patients with chronic hepatitis C bearing HCC. A comparative study of the CD81 sequences from HCC and PBMCs revealed that various nucleotide mutations existed only in the HCC samples in 3 out of 6 patients. Several mutations in the 3' non-coding region of CD81 cDNA were observed exclusively in HCC tissue suggesting its possible role in hepatocarcinogenesis. Because of the absence of polymorphisms, however, CD81 is unlikely to affect the progression of chronic hepatitis C in terms of chronicity, hepatitis activity, or disease stage.     

Mapping of staphylococcal enterotoxin A functional binding sites and presentation by monoclonal antibodies and fusion proteins

Staphylococal enterotoxins (SE) bind with high affinity to major histocompatibility complex (MHC) class II proteins and stimulate large number of T cells via the Vbeta region of the T-cell receptor (TCR). To map the epitopes of SE type A (SEA) involved in MHC binding and cell proliferation, 20 specific anti-SEA monoclonal antibodies (MAbs) and two large glutathione S-transferase fusion proteins corresponding to the amino and carboxy termini, respectively, of SEA were used. The functionality of these antibodies was tested, by MHC binding inhibition, interleukin-2 production, and T-cell proliferation assays. Moreover, I studied the ability of the MAbs to present SEA in vitro to human and murine cells and their reactivity with the two fusion proteins. This study showed that all of the MAbs have a defined effect on one or both immunological properties of SEA and were able to present SEA to human and murine cells. However, one MAb (4H8) recognized SEA but without any interference with its biological activities. When the MAbs were tested to react with the two fusion proteins representing the SEA molecule, all of the MAbs were negative except for two. These results confirmed the presence of two functionally different binding sites of SEA with MHC class II molecules and the importance of the disulfide loop for the mitogenic activity of SEA. I further demonstrated that MAbs can present SEA to immune cells independent of the site recognized by the antibody and that the integrity of the SEA molecule is very important for its functions.     

Morphology of hepatitis C and hepatitis B virus particles as detected by immunogold electron microscopy

We performed indirect immunogold electron microscopy (EM) for immunological identification and characterization of hepatitis C virus (HCV). To clarify the morphology of HCV, an indirect immunogold EM of two plasma samples from patients with high HCV RNA titers was carried out using antibodies specific for the putative HCV envelope protein (E) 1. Spherical virus particles 55–65 nm in diameter with delicate spike projections were detected in the 1.14–1.16 g/ml fractions after sucrose density gradient centrifugation. Polyclonal and monoclonal antibodies to the putative HCV E1 specifically recognized these particles. In addition, immunogold EM of the samples was also performed to uncover the morphology of HCV core particles. Spherical particles 33–40 nm in diameter (average, 37 nm) were detected in the 1.22- to 1.25-g/ml fractions by conventional EM after sucrose density gradient centrifugation. Immunogold EM using rabbit polyclonal antibody (RR8) specific for the putative HCV core protein and colloidal gold-labeled goat antirabbit IgG showed binding of the gold particles with RR8. Some of the HCV core particles showed icosahedric morphology. Optical rotation technique showed that the HCV core particles exhibit sixfold symmetry and that the length of the regular hexagon side is approximately 20 nm, suggesting that they have an icosahedric structure. Further, the detection limit of the indirect immunogold EM was evaluated in 11 plasma samples from chronic hepatitis B patients with different degrees of hepatitis B virus (HBV) DNA titers using antihepatitis B surface antigen antibody. The study showed that the detection limit of virus using this method is 10⁷ virions/ml.     

Controlling complement resistance in cancer by using human monoclonal antibodies that neutralize complement-regulatory proteins CD55 and CD59

Expression of the complement-regulatory proteins (CRP) CD46, CD55 and CD59 represents a strategy used by tumor cells to evade complement-dependent cell cytoxicity stimulated by monoclonal antibodies. We have isolated two single-chain variable fragments (scFv) to CD55 and CD59 from a human phage-display library and from these scFv we have produced two miniantibodies (MB), MB-55 (against CD55) and MB-59 (against CD59), containing the human hinge-CH2-CH3 domains of IgG1. The specificity of the two MB for the corresponding CRP was assessed by ELISA using purified CD46, CD55 and CD59. MB-55 and MB-59 neutralized the inhibitory activity of CD55 and CD59, respectively, restoring the complement-mediated lysis of sheep and guinea pig erythrocytes that was otherwise inhibited by the two CRP. FACS analysis revealed binding of MB-55 and MB-59 to the lymphoma cell line Karpas 422. The two MB induced a two-fold increase in the complement-dependent killing of these cells stimulated by Rituximab, a chimeric anti-CD20 monoclonal antibody. Transfection of HEK293T cells with vectors encoding MB-55 or MB-59 markedly reduced the expression of CD55 and CD59. We conclude that the human antibodies MB-55 and MB-59 may represent a therapeutic tool to increase the complement-dependent killing activity of Rituximab in the treatment of non-Hodgkin's lymphoma.     

Presentation of the hydrophilic domains of hepatitis C Viral E2 envelope glycoprotein on hepatitis B surface antigen particles

Subviral particles of hepatitis B virus have been used to present foreign epitopes. We attempted to present the hydrophilic domains of E2 envelope protein of hepatitis C virus (HCV) as a fusion protein with hepatitis B virus surface antigen (HBsAg). The five hydrophilic domains of HCV E2 antigen were inserted into HBsAg such that the inserted hydrophilic domains were presented on the outer surface of HBV subviral particles. In addition, a fusion encoding the hypervariable region (HVR) of E2 antigen was also made. Cell lysate and culture medium were analyzed for the synthesis and secretion of the fusion proteins by immunoprecipitation with polyclonal anti-HBsAg antibody using recombinant vaccinia virus system. The results showed that the fusion proteins containing these six E2 domains were made in the cell, but only two out of six fusion proteins were secreted into culture medium. Further, cesium chloride density gradient analysis and electron microscopy revealed that these fusions were secreted into culture media as particles. It will be of interest to test immunogenicity of the HBsAg fusion particles containing the HCV E2 domains in animal model. © 1996 Wiley-Liss, Inc.     

Immunosuppression of canine renal allograft recipients by CD4 and CD8 monoclonal antibodies

Abstract: A state of tolerance to MHC mismatched allografts can be generated in rodents by treatment with CD4 and CD8 monoclonal antibodies (mAb). In order to transpose this type of therapy to large animals and ultimately to the clinic, a suitable model is required. To this end we have generated a series of mAb to the canine CD4, CD8, and Thy-1 antigens and have tested their ability to prevent rejection of renal allografts. Donor-recipient pairs were selected from a colony of mongrel dogs in which untreated rejection of two haplotype-mismatched kidneys occurred by day 7 (defined as a serum creatinine > 300 μmol/l). Therapy with either the CD4 or the CD8 mAb, using no other immunosuppression, did not prolong graft survival. Depletion of T cells by a Thy-1 mAb prior to surgery only extended graft survival to day 9. However, treating with combinations of mAb up to day 10 (CD4 plus Thy-1; CD4 plus CD8; or CD4 plus CD8 plus Thy-1) prolonged renal allograft function up to 25 days. Combination of the triple mAb therapy with a sub-therapeutic immunosuppressive drug regimen (cyclosporin A plus azathioprine that alone gave a median survival of 15 days) favored survival to a median of 38 days. This protocol also inhibited the antiglobulin response that had curtailed the effects of mAb treatment, opening the way to more extended, and potentially tolerizing, mAb plus drug regimens.     

The role of higher-order protein structure in supporting binding by heteroclitic monoclonal antibodies: the monoclonal antibody KIM185 to CD18 also binds C4-binding protein

Heteroclitic monoclonal antibodies are characterized by the ability to bind multiple epitopes with little or no similarity. Such antibodies have been reported earlier, but insight into to the molecular basis of this propensity is limited. Here we report that the KIM185 antibody to human CD18 reacts with the plasma protein C4b-binding protein (C4BP). This was revealed during affinity purification procedures where human serum was incubated with surfaces coated with monoclonal antibodies to CD18. Other monoclonal antibodies to CD18 (KIM127 and TS1/18) showed no such interaction with C4BP. We constructed a sandwich-type time-resolved immunofluorometric assay using KIM185 both as capture and developing antibody. By use of proteolytic fragments of KIM185 and recombinant deletion mutants of C4BP the interaction sites were mapped to the variable region of KIM185 and the oligomerization domain of C4BP, respectively. C4BP is a large oligomeric plasma protein that binds activated complement factor C4b and other endogenous ligands as well as microorganisms. By use of the recent crystallographic data on the structure of CD11c/CD18 and prediction of the secondary structure of the C4BP oligomerization domain, we show that epitopes bound by KIM185 in these proteins are unlikely to share any major structural similarity. However, both antigens may form oligomers that would enable avid binding by the antibody. Our report points to the astonishing ability of heteroclitic antibodies to accommodate the binding of multiple proteins with no or little structural similarity within the confined space of the variable regions.