Antigen processing and presentation by EBV-carrying cell lines: cell-phenotype dependence and influence of the EBV-encoded LMP1.

Burkitt's-lymphoma (BL) lines which have maintained in vitro the tumor-cell phenotype (group-I BLs) are poor antigen-presenting cells (APC), in spite of a relatively high surface expression of MHC class II. In order to investigate the mechanism of this deficiency, we have compared group-I BL lines, their sub-lines which have progressed in vitro towards an LCL-like phenotype (group-III BLs), and EBV-transformed lymphoblastoid cell lines (LCLs), for their ability to bind and process tetanus toxoid (TT). The uptake and internalization of 125I-labelled TT was equivalent in the 3 cell types. Only LCLs and group-III BL lines were able to process the TT, as shown by the identification of discrete proteolytic products after separation of whole-cell extracts in tricine-SDS-polyacrylamide gels, and by the recovery of TCA-soluble radioactivity in the culture supernatant. Processing of TT was induced by expression of the EBV-encoded membrane protein LMP 1 in transfected group-1 BLs. The present findings suggest that the inability of group-1 BLs to act as APC is due to their failure to process exogenous antigens. This function appears to be related to phenotypic properties that can be modulated by the expression of LMP1.

LAK cells release a novel form of directly acting TGF beta, tightly bound to a high molecular weight carrier(s).

We showed in previous work that LAK cell supernatants contain a large molecular weight factor with toxic activity for A375 melanoma and other cell lines. The factor, Fr1, was identified tentatively as TGF beta-related, since its activity was abolished by anti-TGF beta serum. This relatedness is further confirmed in the present work, which demonstrates that, like TGF beta, Fr1 stimulates the release and deposition of fibronectin and induces morphological changes indistinguishable from those induced by TGF beta. The TGF beta derived from LAK cells, although associated with a large carrier molecule, is directly acting and does not dissociate from its carrier following gel filtration in acetic acid. Its carrier is different from alpha-2 macroglobulin. SDS-PAGE and immunoblotting showed that FR1 contains TGF beta complexed with large molecules (150 and > 200 kDa), which dissociate in reduced gels to molecules of 60-67 kDa. We interpret these data as showing that TGF beta secreted by LAK cells is, presumably, covalently linked to monomeric carrier molecules of approximately 60 kDa, which, in turn, are S-S bonded to form multimeric molecules of 150 kDa and > 200 kDa.

Cloning and characterization of the latent membrane protein (LMP) of a specific Epstein-Barr virus variant derived from the nasopharyngeal carcinoma in the Taiwanese population.

A DNA fragment containing Epstein-Barr virus (EBV) terminal fragment sequence was obtained from a genomic library of nasopharyngeal carcinoma (NPC). One of the clones (clone 1510) contained the gene encoding latent membrane protein (LMP). Sequence analysis revealed that this gene had 95% homology with the LMP sequence of the B95-8 strain. Among the sequence variations, there was a change from G to T at nucleotide position 169,426, resulting in the loss of an XhoI site in exon 1 of the LMP gene. A pair of primers bracketing the XhoI site were designed to synthesize the EBV DNA fragment from nucleotides 169,081-169,577 by using the polymerase chain reaction (PCR) method. The PCR products were then subject to XhoI digestion and to DNA sequencing analysis. This restriction enzyme site polymorphism along with the sequence variations were also observed in 50 biopsy tissues as well as in the throat washings of 6 out of 20 healthy individuals that we examined, indicating that the EBV strain predominantly existing in these biopsy tissues was different from strains of B95-8, Jijoye or nude mouse passaged cells (C15) with an African origin, but closely resembled other nude mouse passaged CAO cells which were originally derived from China. Balb/c 3T3 cells carrying this NPC-LMP gene showed a transformed cell morphology and were tumorigenic in nude mice. The relationship between this unique type of EBV and NPC has yet to be established.

Degenerative changes in the A375 melanoma line induced by transforming growth factor beta 1.

A375 human melanoma cell cultures grown in the presence of TGF beta contained greatly reduced cell numbers and exhibited drastic alterations in cell morphology compared to the control cultures. Preincubation of the cells with the cytokine for only 18 h was sufficient to induce these changes irreversibly. Examination of TGF beta-treated cells in the electron microscope revealed large numbers of lipid-filled vacuoles in the cytoplasm, greatly contracted nuclei and some loss of the otherwise abundant microvilli. Thus TGF beta may have a direct toxic effect on the A375 melanoma cells.

A high molecular weight cytotoxic lymphokine in supernatants of lymphokine-activated killer cells cross-reacts with transforming growth factor beta.

Supernatants of lymphokine-activated killer (LAK) cells were highly cytotoxic for melanoma A375 cells. A high-molecular-weight fraction was isolated from such supernatants by gel filtration on an S-300 Sephacryl column (Fraction 1; Fr1). The cytotoxic activity in Fr1 was heat- and acid-resistant and was completely abolished by a rabbit antibody against TGF-beta. We conclude that Fr1 contains TGF-beta or a cross-reactive molecule, associated with a high-molecular-weight carrier.

Cross-reactivity between the EBNA-1 p107 peptide, collagen, and keratin: implications for the pathogenesis of rheumatoid arthritis.

An unusually heavy load of Epstein-Barr virus (EBV) infection and autoimmunity to collagen are believed to be contributing factors to the pathogenesis of rheumatoid arthritis (RA). The present report presents data showing that p107, the major epitope of the EBV-encoded EBNA-1 antigen, cross-reacts with denatured collagen (DC) and keratin (K), suggesting a new likely link among RA, EBV-1, and these autoantigens. A radioimmunoassay using antigen-coated microtiter plates was used to demonstrate antibodies in sera of patients with RA and sera of healthy donors against p107, DC, and K. Specificity of the antibodies was ascertained by inhibition tests with the homologous antigens. Cross-reactivity among anti-p107, anti-DC, and anti-K antibodies was assayed by the ability of a given antigen to block the binding of nonpurified or affinity-purified antibodies to plates coated with another antigen. Most of the sera contained antibodies to all three antigens, but only anti-DC antibodies were present in higher titers in RA sera. Preincubation of sera with p107 appreciably reduced their binding to plates coated with DC or K. On the other hand, preincubation with DC (in solution or bound to Sepharose) did not result in consistent reduction of anti-p107 titers. Tests with affinity-purified antibodies revealed the existence of two antibodies populations, one of which reacted preferentially with p107, the other with DC. The cross-reactivity of the anti-p107 antibodies with DC and K suggests that such antibodies, produced by RA patients following persistent stimulation with EBV, might react in vivo with collagen (and keratin) exposed in previously damaged areas and thus reinforce the disease process.

Culture supernatants of lymphokine-activated killer (LAK) cells contain a high-molecular-weight cytotoxic lymphokine.

Supernatants of human lymphokine-activated killer (LAK) cells grown in vitro were tested for cytotoxic activity against several mouse and human neoplastic cell lines. All LAK preparations tested (14/14) exhibited cytotoxic activity (40-90% killing of the target cells). Sephacryl S-300 Gel filtration experiments indicated that the biological activity of the LAK supernatant is associated with molecular moieties ranging from 800 kDa or more, to less than 10 kDa. The finding of strong cytotoxic activity in LAK supernatants against several tumor lines points to the possibility of employing soluble products of these cells, rather than the living cells themselves, for therapeutic purposes.

Antibodies in human sera against the Epstein-Barr virus encoded latent membrane protein (LMP).

Antibodies reactive with the Epstein-Barr (EBV)-encoded latent membrane protein, LMP, were detected in human sera. Membrane fractions of the EBV-carrying Raji cells and a fusion protein (LMFP), that represents the carboxy part of LMP, were used as antigens. These were assayed by the reduction of the leukocyte migration inhibition (LMI) reaction. Reactivity with LMFP was detected in 10 of 14 sera from patients with Burkitt's lymphoma (BL) and in 3 of 23 sera from healthy EBV-seropositive individuals. The antibody levels were higher in the BL sera. Since the tumor cells do not express LMP, this may be due to the high virus load in the patients.

Proteins in normal and malignant cells, cross-reacting with the latent membrane protein encoded by Epstein-Barr virus.

Human B lymphocytes transformed by infection with the Epstein-Barr virus (EBV) express a new membrane protein of 63 kDa (latent membrane protein, LMP) encoded by the virus. The function of this protein in the virus-cell interaction is not known. In this work we have identified in EBV- human and mouse cell molecules which cross-react with LMP. Two types of reagents were employed: (a) antibodies against LMP-derived synthetic peptides, affinity purified from antisera against a fusion protein containing the carboxy half of the LMP molecule and (b) antisera prepared by immunizing rabbits directly with the peptide conjugates. Cross-reactions were determined by radioimmunoblotting experiments. At least six molecules (Mr = 110, 85, 63, 53, 45 and 23 kDa), present in a variety of human cells (peripheral blood lymphocytes, B cell lines and epithelial cell lines) were found to cross-react with the LMP-derived peptides. Cross-reacting proteins were also identified in normal mouse tissues. The specificity of the cross-reacting antibodies was confirmed by inhibition experiments with the corresponding peptide. Furthermore, antibodies eluted from individual bands were shown to bind to the same band when reacted with new blots of the same extracts. Our data suggest that normal cells contain a family of highly conserved proteins cross-reacting with the LMP molecule. If, indeed, these proteins share common functions, their study may lead the way to unraveling the function of LMP.

Down-regulation of the EBV-encoded membrane protein (LMP) in Burkitt lymphomas.

A radioimmunoassay (RIA) has been developed and used to determine the expression of LMP-a membrane protein encoded by the LT3 region of the Epstein-Barr virus (EBV) genome-in cell lines of various origins. The RIA was highly sensitive, specific and reproducible. All EBV-negative cell lines were LMP-negative and 18 of 21 EBV-carrying cells were LMP-positive. LMP concentrations varied widely, ranging approximately from less than 4 ng up to 650 ng/mg protein. In several instances comparisons were made between lymphoblastoid (LCLs) and Burkitt lymphoma (BL) cell lines (EBV-positive or EBV-converted sublines of originally EBV-negative BL) originating from the same patient. In all such cases LMP and LMP-specific mRNA levels were higher in the LCLs. Most of the LMP was found in the cytosol fraction, yet this fraction was negative in immunoblotting tests. However, antiserum preincubated with the cytosol lost its ability to react in immunoblotting with membrane LMP, indicating that the 2 LMP forms (membrane and cytosol) are completely cross-reactive.

Stable transfection of a human lymphoma line by sub-genomic fragments of Epstein-Barr virus DNA to measure humoral and cellular immunity to the corresponding proteins.

An Epstein-Barr virus (EBV)-negative human lymphoid B-cell line, DG75, was stably transfected with recombinant selection vectors that carry a subfragment of the BamHI WYH region (nucleotides 44664 to 50628), the BamHI K fragment, or a subfragment of the EcoRI D region (nucleotides 166614 to 170149) of B95-8 EBV DNA. These fragments contain the coding exons for the EBV-determined nuclear antigens EBNA2 and EBNA1, and the membrane antigen LMP, respectively. Antigen expression of the cells was detected by immunofluorescence. EBNA2 was expressed in 80-100% of the transfected cells, in contrast to EBNA1 which was expressed in only 25%, and LMP in only about 5% of the cells. Humoral antibody responses were measured by immunofluorescence and compared to cellular immunity as determined by the leukocyte migration inhibition (LMI) technique. Extracts from transfected cell lines expressing EBNA1, EBNA2 or LMP elicited an LMI response with cells from healthy EBV-seropositive individuals whereas the extract from the parental DG75 cell line did not. The results demonstrate the value of stably transfected cell lines expressing a defined EBV antigen for the monospecific analysis of host responses to the EBV-encoded antigen complex in growth-transformed cells.

A radioimmunoassay for the diagnosis of malaria.

A newly developed radioimmunoassay for the diagnosis of malaria has been tested in South Africa. The radioimmunoassay is an antibody binding-inhibition assay, based on a monoclonal antibody (D5) cross-reacting with Plasmodium berghei and P. residual binding activity was tested on antigen-coated microtiter plates. A sample was considered positive if it inhibited binding of the antibodies to an extent exceeding that of the microscopically negative blood samples. Blood was collected on 3 separate occasions from a total of 530 individuals living in a malaria-endemic area and was examined by radioimmunoassay and microscopy. Group 1, consisting of 194 samples, yielded 12 samples positive by microscopy and 10 of these (83%) were also positive by radioimmunoassay. One sample in this group was "positive" in the radioimmunoassay but negative on microscopy (false positive). In the 320 samples of group 2, 13 were positive by microscopy and 6 (46%) by radioimmunoassay. Group 3, which included 16 samples preselected as positive by microscopic examination and 16 controls, was examined after 4 weeks storage at -20 degrees C. Twelve samples (75%) were positive by radioimmunoassay. Tests carried out to determine the effect of blood storage on the activity of the antigen indicated that activity was preserved with little loss over a 3-month period.

(Auto)antibodies in human breast cancer sera against antigens associated with breast cancer cells, detected by immunoblotting.

Sera of patients with breast cancer (as well as control normal sera and sera of patients with ovarian cancer or melanoma) were screened for the presence of antibodies against antigens expressed by the MDA breast cancer cell line. The techniques employed were radioimmunoassay with radioiodinated protein A and immunodotting with peroxidase-conjugated anti-human immunoglobulin antibodies. Sera reacting strongly by immunodotting were subsequently tested against antigens of the MDA and T47D cell lines in immunoblotting experiments. Both the breast cancer and the control sera yielded highly complex band patterns, which varied from serum to serum. The cancer sera differed from the normal sera, however, as they produced in most cases one or several bands that were distinctly stronger than the others. One of the strong bands, in fact a doublet of approximately 50 kilodaltons (kd), was produced preferentially (although not exclusively) when breast cancer sera were reacted with T47D cell membrane antigens. Absorption of selected sera with normal tissue or MDA antigens abolished or greatly reduced the intensity of some of the bands. It is concluded that, with the possible exception of the 50-kd band, most (probably all) of the bands seen in immunoblots resulted from the binding of autoantibodies to normal antigens expressed by the breast cancer cell lines. The main difference between cancer and normal sera would seem to be an increased content of autoantibodies in cancer, the specificity of these autoantibodies varying, however, from serum to serum.

Epstein-Barr virus (EBV) antigen-specific leukocyte migration inhibition (LMI) in infectious mononucleosis (IM). I. Kinetics and response to a membrane protein on EBV-transformed cells.

Cell-mediated immune response of mononucleosis (IM) patients to Epstein-Barr virus (EBV)-determined antigens was measured by the leukocyte migration inhibition (LMI) assay. Patients in the acute phase of the disease failed to respond to partially purified nuclear antigen, EBNA, or to cell extracts that contained EBNA as the predominant EBV antigen. They showed a strong specific response to cell extracts enriched in early antigen (EA) and virus capsid antigen (VCA). The LMI response to EBNA appeared in convalescence in parallel with EBNA-specific antibodies, slightly later in children than in adults. Membrane fractions of EBV-carrying, virus nonproducer Raji cells induced an EBV-specific LMI at approximately the same time. A bacterial fusion protein containing the hydrophilic part of the virus-coded membrane antigen of latently EBV-infected cells also induced an EBV-specific response that parallelled the LMI reaction elicited by the Raji membrane fraction. This is in line with our previous finding (D. Sulitzeanu et al., J. Virol. 58, 230, 1986) that this fusion protein shares an epitope with Raji cell membranes.

Leukocyte migration inhibition demonstrates a human T-cell response to a membrane protein expressed in latent Epstein-Barr virus infection.

Leukocyte migration inhibition tests show that lymphocytes of Epstein-Barr virus-seropositive individuals recognize a Raji cell membrane antigen and a membrane protein encoded by Epstein-Barr virus in latently infected cells. Antiserum against the latter blocks the leukocyte migration inhibition triggered by both preparations, suggesting that the two antigens are associated with the same protein complex.

Serial circulating immune complex values and development of metastatic disease in breast cancer and malignant melanoma patients.

A polyethylene glycol precipitation technique was used to determine the levels of circulating immune complexes (CIC) in breast cancer and melanoma patients. All patients in the study had undergone surgery and were free of distant metastatic disease. CIC were measured at two to four time intervals, of 3 to 6 months each, over an average follow-up period of 13.5 months (range 7-20 months). In both groups of patients, metastatic disease developed with a higher frequency in patients who had undetectable CIC levels throughout the follow-up period or had become negative at the time metastases were discovered.

Detection of Plasmodium falciparum using a radioimmunoassay based on a crossreacting, monoclonal anti-P. berghei antibody-P. berghei antigen system.

An antibody binding-inhibition test is described, which allows the detection of P. falciparum in red blood cells (RBC) infected in vitro, using a crossreacting, monoclonal anti-P. berghei antibody and P. berghei coated microtiter plates. Experiments carried out to determine the coating efficiency of various P. berghei and P. falciparum derived antigen preparations showed that intact, saponin freed P. berghei parasites and sonicated, RBC parasitized with P. falciparum had the highest binding activity. Binding of the monoclonal antibody to the antigen coated plates was effectively inhibited by preincubation with sonicated, P. falciparum infected RBC. The minimal degree of infection detectable was about 0.008% parasitemia (400 parasitized RBC/microliters blood). The sensitivity of detection was not appreciably affected by the source of the coating antigen. We conclude that the difficulty and expense involved in the use of P. falciparum based immunodiagnostic tests for large scale screening for malaria can be obviated by making use of P. berghei based assays.