Lack of HLA class I antigen expression by cultured melanoma cells FO-1 due to a defect in B2m gene expression

The melanoma cell line FO-1 does not express HLA class I antigens and does not acquire them on the cell surface after incubation with IFN-gamma. Immunochemical studies showed that FO-1 cells synthesize HLA class I heavy chain, but do not synthesize beta 2-microglobulin (beta 2-mu). The latter abnormality is associated with lack of beta 2-mu mRNA which remains undetectable in FO-1 cells incubated with IFN-gamma. The defect was identified as a genetic lesion in the B2m gene, since DNA hybridization analysis detected a deletion of the first exon of the 5'-flanking region, and of a segment of the first intron of the B2m gene. HLA class I antigen expression was reconstituted on melanoma cells FO-1 after transfection with the wild-type mouse B2m gene, thereby confirming the abnormality of the endogenous B2m gene. The defect identified in FO-1 cells is distinct from that underlying the lack of HLA class I antigen expression by lymphoblastoid cells Daudi, but is remarkably similar to that causing lack of H-2 class I antigen expression by mouse lymphoblastoid cells R1 (TL-). These results suggest that genetic recombination in the 5' region of the B2m gene is a recurrent mechanism in B2m gene defects. In addition to contributing to our understanding of molecular abnormalities in HLA class I antigen expression by melanoma cells, FO-1 cells represent a useful model for analyzing the role of HLA class I antigens in the biology of melanoma cells and in their interaction with cells of the immune system.     

Reduction in susceptibility to natural killer cell-mediated lysis of human FO-1 melanoma cells after induction of HLA class I antigen expression by transfection with B2m gene

Induction of HLA class I antigens on cultured melanoma cells FO-1 after transfection with a human or a mouse B2m gene was associated with a statistically significant reduction in their susceptibility to natural killer (NK) cell-mediated lysis. These results indicate that the structural differences between human and mouse beta 2-mu do not abolish the ability of the HLA class I molecular complex to modulate NK cell-mediated lysis of melanoma cells FO-1. The role of HLA class I antigens in the phenomenon is corroborated by the ability of anti-HLA class I MAb to enhance, although to a different extent, the susceptibility of transfected FO-1 cells to NK cell-mediated lysis. Gamma interferon (IFN-gamma) and tumor necrosis factor-alpha (TNF-alpha) significantly reduced the susceptibility to NK cell-mediated lysis of transfected FO-1 cells. Surprisingly, TNF-alpha reduced the extent of lysis more than IFN-gamma, although the latter cytokine enhanced HLA class I antigen expression more than the former one. This finding, in conjunction with a reduction in the susceptibility to NK cell-mediated lysis of untransfected FO-1 cells incubated with IFN-gamma or TNF-alpha, suggests that the two cytokines reduce NK cell-mediated lysis of transfected cells by modulating not only the expression of HLA class I antigens, but also that of other structures. Induction of HLA class I antigens and their modulation with IFN-gamma did not affect the susceptibility to lymphokine-activated killer (LAK) cell-mediated lysis of transfected FO-1 cells. Characterization of the molecular mechanism(s) underlying abnormalities in HLA class I antigen expression by melanoma cells and of the role of these molecules in the interactions of melanoma cells with various types of effector cells may suggest novel immunotherapeutic approaches to melanoma.     

Suppression of class I human histocompatibility leukocyte antigen by c-myc is locus specific

The c-myc oncogene downregulates class I HLA expression in human melanoma. The major class I HLA antigens are encoded by three loci, A, B, and C, and we investigated whether these loci are suppressed equally by c-myc. In three melanoma cell lines with high c-myc expression, we analyzed mRNA, protein, and cell surface expression of the class I HLA antigens. Whereas the HLA-B locus expression was found to be strongly reduced, the HLA-A locus was expressed normally. Analysis of c-myc-transfected clones of two melanoma cell lines confirmed that c-myc preferentially suppresses the class I HLA-B locus. Immunohistochemical analysis of fresh melanoma lesions also showed that in the tumors the HLA-A loci are expressed normally, while on the majority of tumor cells no HLA-B antigen expression was found. This downregulation may have consequences for the recognition of malignant cells by tumor-infiltrating lymphocytes. Our results predict that HLA-B-restricted cytotoxic T cells will be unable to kill high c-myc-expressing melanoma cells.     

Synergistic effect of 5-azacytidine and gamma-interferon or dimethyl sulfoxide on expression of HLA class I antigens by PLC-PRF-5 cells

In order to elucidate the mechanism by which HLA antigens expression is induced or enhanced on the injured or transformed hepatocytes, we have made in vitro studies using human hepatic tumor-derived cell lines as a model system. In the present study, PLC-PRF-5 cells that have the integrated form of hepatitis B virus genome in DNA were treated with 5-azacytidine (5-azaC) in combination with gamma-interferon (IFN-gamma) or dimethyl sulfoxide (DMSO). HLA antigens on the cell surface were quantitated by using a modified cell-ELISA method. As a result, it was demonstrated that DMSO- or IFN-gamma-treatment enhanced expression of HLA class I antigens on the cell surface. In addition, enhanced expression of the antigens on PLC-PRF-5 cells treated with 5-azaC in combination with IFN-gamma or DMSO represented a synergistic effect of these inducers on HLA class I antigens expression although no changes in HLA antigens expression were induced after 5-azaC-treatment alone in short-term experiments. Furthermore, an indirect immunofluorescent analysis of hepatitis B surface antigen on the cells demonstrated increased expression of the antigen after 5-azaC-treatment alone. HLA class II antigens and hepatitis B core antigen were not induced even after those treatments and also not after a long-term experiment. These results might indicate possible modulation of HLA class I and hepatitis B virus antigens expression on the cultured cells by a DNA hypomethylating agent, 5-azaC.     

Butyrate enhances major histocompatibility complex class I, HLA-DR and ICAM-1 antigen expression on differentiated human intestinal epithelial cells

Besides its metabolic role, butyrate, a by-product of colonic fermentation, can modulate colonocyte proliferation and the expression of various molecules. In inflammatory bowel diseases, epithelial cells express HLA class II molecules and may behave as antigen-presenting cells. This study was performed to characterize the effect of butyrate on major histocompatibility complex expression by human colonocytes in comparison with interferon-γ. Five cell lines displaying different differentiation features were analysed for antigen expression by flow cytofluorimetry. All lines expressed class I antigens, whereas only SW 1116 cells express HLA-DR. On these cells, butyrate and interferon-γ strongly enhanced HLA-DR and ICAM-1 expression, whereas a mild increase in class I antigens was noted. Moreover, an increase in class I antigens was observed on two other differentiated cell lines, and it was synergistic with interferon-γ. Butyrate, by its modulation of HLA-DR, ICAM-1 and HLA class I expression, may act on antigen presentation and, thus, influence some inflammatory processes.     

Expression and function of CD8 in a murine T cell hybridoma

In general, the human CD8 molecule is expressed on T cells specific for HLA class I molecules. Studies designed to delineate the function and to define the ligand of the CD8 molecule have been complicated by the fact that the presumptive ligand for CD8 is on the HLA class I molecule, the same molecule encoding the ligand for the antigen-specific T cell receptor. The ability to express genes in cells other than their natural host has produced a new technology with which to approach CD8 functional studies. The insertion of a cDNA clone for CD8 in a defective retroviral vector has allowed the transfer of CD8 by infection with the resulting defective retrovirus. CD8 was then expressed in an HLA class II-specific T cell, thus separating the ligand requirements of the TCR and CD8. By this approach, the human CD8 molecule was expressed in a murine T cell hybridoma specific for human class II antigens. The resulting CD8+ hybridomas demonstrated a 10-fold increase in IL-2 production over the parent cell line when stimulated with JY, a human B lymphoblastoid cell line expressing both class I and II HLA antigens, demonstrating that expression of CD8 increases T cell activation. mAbs directed against the CD8 molecule inhibited the response of CD8+ hybridomas to JY, supporting the conclusion that the CD8 molecule was fractional. The role of CD8 as a receptor for class I MHC antigens was addressed by stimulation with a cell line expressing HLA-DR antigens, but lacking the expression of HLA class I antigens (Daudi). Stimulation of the CD8+ hybridomas by Daudi did not result in increased IL-2 production. The response to Daudi was unaltered by the addition of anti-CD8 mAb, in contrast to the ability of anti-CD8 mAb to block JY stimulation. Furthermore, mAbs directed against the class I antigens present on JY cells were able to block the enhanced response of the CD8+ hybridomas to JY. These data support the hypothesis that HLA class I molecules are the ligands involved in the CD8-dependent enhancement of T cell activation.     

c-erbB-2 and episialin challenge host immune response by HLA class I expression in human non-small-cell lung cancer

The role of major histocompatibility complex expression in cancer prognosis and pathogenesis is contradictory. The aim of the current study was to compare the expression of HLA class I molecules and of oncoproteins that may be sources of peptides presented by HLA class I antigens in non-small-cell lung cancer. For this purpose, the expression of HLA class I antigen and TAP-1 molecule (a transporter in the antigen-processing 1 transport protein) were studied with epidermal growth factor, receptor; c-erbB-2; episialin; wild-type and mutant p53; bcl-2 oncoprotein expression; and angiogenic factor expression (vascular endothelial growth factor and thymidine phosphorylase). The degree of lymphocytic stromal infiltration and of platelet-endothelial cell adhesion molecule-expressing lymphocytes was also studied. A strong association of c-erbB-2 and MUC1 (episialin) expression with HLA class I expression was observed (p = 0.005 and 0.009, respectively). Intense CD31-positive lymphocytic infiltration was also more frequent in HLA class I-positive cases (p = 0.05). Although there was no association of HLA class I expression with survival, loss of the HLA class I expression in MUC1 or c-erbB-2 overexpressing cases conferred a poorer clinical outcome (p = 0.04). Both c-erbB-2 and MUC1 are well-known targets of T-cell-mediated cytotoxicity and cell-cell or cell-matrix adhesion-regulating proteins. The authors provide evidence that the sequence of cell adhesion-disrupting oncoprotein expression, HLA class I induction, and enhanced epitope presentation followed by lymphocytic response is an important pathogenetic three-step sequence of events that define, in part, the clinical outcome in non-small-cell lung cancer.     

Rescue of Daudi cell HLA expression by transfection of the mouse beta 2-microglobulin gene

The Daudi cell line is a B-lymphoblastoid line derived from a Burkitt lymphoma. Daudi cells lack cell surface expression of class I HLA molecules despite the presence of intracellular class I heavy chains. They have a defect in the gene encoding beta 2-microglobulin (beta 2m), resulting in lack of translatable mRNA for this protein. It has been thought that this deficiency is responsible for the lack of cell surface class I expression. However, data have recently been presented demonstrating that at least one mouse class I heavy chain can be expressed on the cell surface in the absence of beta 2m. These results raised the questions of whether the lack of beta 2m is the only defect in Daudi and whether transfer of this single gene could restore surface class I expression. We found that transfection of the mouse beta 2m gene into Daudi indeed rescued cell surface expression of class I HLA molecules, and that these molecules could be recognized both by monomorphic and allospecific mAbs. CTL clones specific for HLA-B17 or a determinant shared by HLA-B17 and HLA-A2 killed the Daudi cells transfected with the beta 2m gene, but not untransfected Daudi or Daudi transfected with vector alone. Mouse beta 2m on the transfected Daudi cells could exchange with human beta 2m when the cells were incubated in human serum. This exchange did not alter the ability of the cells to be killed by the specific CTLs. These results demonstrate that the lack of beta 2m is the sole reason for lack of surface class I molecules in Daudi cells, and that beta 2m is required for cell surface expression of the specific class I heavy chains of Daudi.     

A study of HLA (Bg) on red cells and platelets by immunoblotting with monoclonal antibodies

HLA class I antigens (Bg) on red cells (RBCs) are expressed by some normal donors and by many patients with systemic lupus erythematosus (SLE). To identify the membrane components previously detected by hemagglutination with HLA class I-specific monoclonal antibodies (MoAbs), RBC membrane preparations were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblotted with the HLA class I MoAbs. Two components were obtained that reacted with the MoAbs: a heavy chain of 45 kDa and a light chain termed beta2-microglobulin (beta2-M) of 11 kDa. The effect of chloroquine and acid elution in stripping HLA antigens is shown to be due to the removal of beta2-M, as only that component was detected in eluates from reactive RBCs. Neither antibody elution method affected the heavy chain expression assessed by immunoblotting. It is concluded that HLA class I antigens on RBCs are integral membrane components of the type normally found and wisely distributed on many nucleated cells. Platelets, which have stronger HLA class I antigen expression, were also studied, and their membrane preparations yielded heavy chain and beta2-M molecules; the effect of chloroquine treatment was harder to assess than that of acid elution, owing to the sensitivity with which both components are detected in immunoblotting. In eluates obtained from acid treatment only beta2-M is detected.     

High frequency of melanoma-associated antigen or HLA class I loss does not correlate with survival in primary melanoma

Melanoma-associated antigens are at the center of many immunotherapeutic trials in melanoma. Little is known about the impact of antigen expression on the natural course of disease. We stained 110 cases of primary melanoma with a median follow-up of 13 years (range 10-18 years) for melanoma-associated antigens gp100, MelanA/MART-1, MAGE-3, tyrosinase, and for HLA class I molecules. Of 91 cases evaluated, we found immunoreactivity for gp100, MelanA/MART-1, and tyrosinase in 88%, 80%, and 87% of primary tumors, respectively, for MAGE-3 in 37% and for HLA class I in 86% of primary tumors. Loss, that is, heterogeneous expression within primary tumors, was most pronounced for gp100 (73% of primary tumors) and least for MAGE-3 (27% of primary tumors). MelanA/MART-1 and tyrosinase expression loss was 58% and 59% of primary tumors, respectively. There was a high rate of expression loss for HLA class I (74%). Univariate and multivariate statistical analysis of expression in primary tumors and loss of melanoma antigens as well as HLA class I in individual primary tumors showed no significant correlation to overall survival. Loss of gp100 and loss of tyrosinase expression showed a negative survival trend over homogeneous expression of these antigens, although not reaching statistical significance (P = 0.08 and P = 0.09, respectively). We conclude that loss of melanoma antigen expression as well as HLA class I expression is a frequent observation in primary melanoma. However, no statistically significant correlation between loss of these antigens in individual primary tumors and negative impact on overall survival was found in our cohort.     

Human histocompatibility leukocyte antigen-binding supermotifs predict broadly cross-reactive cytotoxic T lymphocyte responses in patients with acute hepatitis.

The present study was designed to determine if highly conserved hepatitis B virus (HBV)-derived peptides that bind multiple HLA class I alleles with high affinity are recognized as cytotoxic T lymphocyte (CTL) epitopes in acutely infected patients. Peripheral blood mononuclear cells from 67 patients with acute hepatitis B, and 12 patients convalescent from acute hepatitis B, were stimulated with three panels of peptides, each of which bind with high affinity to several class I alleles from the HLA-A2-, HLA-A3-, or HLA-B7-supertypes. In these patients, 8 of the 19 peptides tested were found to represent CTL epitopes recognized by two or more alleles in each supertype. Two sets of nested peptides were recognized in the context of alleles with completely unrelated peptide binding specificities. Finally, promiscuous recognition by the same CTL of a given peptide presented by target cells expressing different A2 subtypes was also commonly observed. In conclusion, several HBV-specific CTL epitopes, recognized by acutely infected or convalescent patients in the context of a wide range of HLA alleles have been identified. These results demonstrate the functional relevance of the supertype grouping of HLA class I molecules in a human viral disease setting. Furthermore, they represent a significant advance in the development of a totally synthetic vaccine to terminate chronic HBV infection and support the feasibility of a systematic approach to development of similar vaccines for prevention and treatment of other chronic viral infections.     

β2-microglobulin gene mutation is not a common mechanism of HLA class I total loss in human tumors

One hundred and sixty-two tumor samples were analyzed for HLA class I expression using immunohistological techniques. HLA class I total loss (phenotype no. I) was detected in 31 cases (19%), comprising 20 colorectal, 3 laryngeal, and 2 bladder carcinomas and 6 melanomas. Twenty-one cases were selected for molecular analysis due to a higher proportion of tumor cells versus stroma cells (75%). We investigated whether beta2-microglobulin mutation was responsible for HLA downregulation. Single-strand conformation polymorphism and sequencing analysis of DNA samples was performed. Alterations were detected only in melanomas M78 (a point mutation in the initiation ATG sequence), M79 (a mutation in codon 31 producing a stop codon), and M34 (a TTCT deletion introducing a termination codon signal). We found no beta2-microglobulin gene mutation in the other 18 samples. Loss of heterozygosity in 15q close to the beta2-microglobulin gene was found in 5 cases. We conclude that HLA class I total loss can frequently occur without beta2-microglobulin gene mutations.     

Potential for transfusion-associated graft-versus-host disease due to apheresis platelets matched for HLA class I antigens

Transfusion-associated graft-versus-host disease (TA-GVHD) has been reported in immunocompetent recipients of nonirradiated cellular blood components from donors who are homozygous for an HLA haplotype shared with the patient. In these cases, donor lymphocytes have no antigens foreign to the recipient, and this similarity in HLA antigens appears important for the development of TA-GVHD. Experience with 65 patients receiving apheresis platelets matched for class I HLA antigens was reviewed to determine the incidence of such a transfusion among HLA- matched, unrelated donor-recipient pairs. In 5 percent of transfusions (31/673), the patient received lymphocytes from a donor exhibiting no antigens foreign to the recipient, but the patient had additional HLA-A or -B antigens not present on donor lymphocytes. Twenty-three percent (n = 15) of patients received at least one such transfusion. In addition, most patients were immunosuppressed as a result of their underlying disease or therapy, which may decrease the degree of antigen matching required to initiate TA-GVHD. Until the pathogenesis of this disease is better understood, it is recommended that the transfusion of an HLA-matched cellular blood component be considered a risk factor for the development of TA-GVHD regardless of the patient's immune status, and that all such blood components be irradiated.     

Intestinal epithelial cells use two distinct pathways for HLA class II antigen processing.

Intestinal epithelial cells express a low level of HLA class II molecules constitutively, with elevated levels seen in the setting of mucosal inflammation including inflammatory bowel disease. The ability of intestinal epithelial cells to act as antigen presenting cells for alphabeta CD4(+) T lymphocytes was examined through a molecular analysis of the HLA class II antigen processing pathway. We have shown that intestinal epithelial cells contain abundant constitutive levels of the cathepsin proteases proven to function in HLA class II mediated antigen presentation. Activation of these cells by gamma-IFN induced the expression of invariant chain and HLA-DM alphabeta, thus facilitating the formation of compact, SDS-stable HLA- DR alphabeta heterodimers. Using HLA-DR-restricted T cells and retroviral mediated gene transfer of HLA-DR alleles into the intestinal epithelial cell lines HT-29 and T84, we demonstrated efficient antigen processing and presentation to CD4(+) T lymphocytes in the presence of the proinflammatory cytokine gamma-IFN. The class II processing pathway and presentation in the presence of gamma-IFN was indistinguishable from that observed with a conventional antigen presenting cell. Antigen processing also occurred in intestinal epithelial cells in the absence of gamma-IFN, and in contrast to that seen after stimulation with gamma-IFN, required high concentration of antigen and was not inhibited by the protease inhibitor leupeptin. These data suggest the use of two distinct pathways of HLA class II antigen processing in enterocytes with differential immunomodulatory properties in the presence or absence of mucosal inflammation.     

Altered MHC class I antigens in tumors

MHC class I antigens are lost or downregulated in invasive tumors compared with autologous normal tissues. This is observed in most of the newly induced experimental tumors analyzed if they are cloned before passaging in vivo. Similarly, this is observed in 40%-90% of human tumors using the available panel of anti-HLA class I monoclonal antibodies. In both systems the tumor populations are heterogeneous for H-2/HLA expression and composed of clones that express different amounts of MHC class I antigens. This heterogeneity may have a profound influence on tumor behavior, considering the role that MHC antigens play in T and natural killer cell-mediated responses. It is possible that the tumor escape mechanisms from T and natural killer cells select variants that express a particular MHC class I altered phenotype. We review the MHC changes detected in different experimental as well as human tumors and demonstrate the relevance of these altered H-2/HLA tumor phenotypes for implementing immunotherapeutic strategies based on T or natural killer cellmediated responses.