Detection and quantification of CD8(+) T cells specific for HLA-A*0201-binding melanoma and viral peptides by the IFN-gamma-ELISPOT assay.

Blood lymphocytes from HLA-A*0201-subtyped melanoma patients and healthy controls were screened for the presence of T cells specific for HLA-A*0201-binding melanoma and viral peptide antigens by the enzyme-linked immunoSPOT (ELISPOT) assay. CD8(+) cells were tested for peptide-specific IFN-gamma release immediately after selection as well as after 2 weeks of in vitro stimulation. After in vitro stimulation, CD8(+) T cells specific for influenza were measured in all patients and controls, whereas these T cells could be detected among nonstimulated CD8(+) cells in only 52% of individuals. Similarly, T cells specific for EBV were more frequently measured among in vitro-stimulated than nonstimulated CD8(+) cells. In nonstimulated CD8(+) cells, T cells specific for MART-1/Melan-A, gp100, tyrosinase and CAMEL were present in 4 (33%), 1 (8%), 1 (8%) and 3 (25%) of 12 patients, respectively. Only MART-1/Melan-A-specific CD8(+) T cells were found in 1 (11%) of 9 healthy controls. CD8(+) T cells specific for MAGE-2 were not observed. After in vitro stimulation, CD8(+) T cells specific for MART-1/Melan-A could be demonstrated in 6 (46%) of 13 patients and 2 (20%) of 10 controls. CD8(+) T cells specific for gp100 were detected in 1 patient after in vitro stimulation. No CD8(+) T cells specific for tyrosinase, MAGE-2 or CAMEL could be measured after in vitro stimulation. These data show that the ELISPOT assay allows direct ex vivo detection of CD8(+) T cells specific for viral and melanoma antigens. Furthermore, the data show that the sensitivity of the ELISPOT assay to measure influenza- and EBV-specific CD8(+) T cells can be enhanced by a short in vitro stimulation step, whereas opposing effects on numbers of CD8(+) T cells specific for melanoma antigens have been observed.

Direct, autocrine and paracrine effects of cyclic stretch on growth of myocytes and fibroblasts isolated from neonatal rat ventricles.

Several studies have demonstrated that static stretch of cardiomyocytes induces cardiomyocyte hypertrophy. We investigated the effects of cyclic stretch, a more physiological stimulus, on protein synthesis and DNA synthesis of rat ventricular cardiomyocytes and cardiofibroblasts. Further-more, we investigated whether these effects are caused by autocrine mechanisms. In addition, we studied the paracrine influences of stretched cardiofibroblasts on cardiomyocyte growth. Short-term cyclic stretch (0-24 h) of cardiomyocytes induced a growth response indicative of cardiomyocyte hypertrophy, given the fact that increased rates of protein synthesis and DNA synthesis were accompanied by an elevated release of atrial natriuretic peptide into the culture medium. In cardiofibroblasts, short-term cyclic stretch also induced a growth response as indicated by an increased rate of protein synthesis and DNA synthesis. Furthermore, incubation of stationary cardiofibroblasts with conditioned medium derived from stretched cardiofibroblasts revealed an autocrine effect of stretch as illustrated by an increased rate of protein synthesis in stationary cardiofibroblasts. In analogy, there was an autocrine effect of stretch on stationary cardiomyocytes incubated with conditioned medium derived from stretched cardiomyocytes. Moreover, we observed a paracrine effect of the conditioned medium derived from stretched cardiofibroblasts on stationary cardiomyocytes. Thus, short-term cyclic stretch of cardiomyocytes and cardiofibroblasts induces growth responses that are the result of direct, autocrine, and paracrine effects. These autocrine/paracrine effects of stretch are most probably due to release of factors from stretched cells.

The role of angiotensin II, endothelin-1 and transforming growth factor-beta as autocrine/paracrine mediators of stretch-induced cardiomyocyte hypertrophy.

Cardiac hypertrophy is a compensatory response of myocardial tissue upon increased mechanical load. Of the mechanical factors, stretch is rapidly followed by hypertrophic responses. We tried to elucidate the role of angiotensin II (AII), endothelin-1 (ET-1) and transforming growth factor-beta (TGF-beta) as autocrine/paracrine mediators of stretch-induced cardiomyocyte hypertrophy. We collected conditioned medium (CM) from stretched cardiomyocytes and from other stretched cardiac cells, such as cardiac fibroblasts, endothelial cells and vascular smooth muscle cells (VSMCs). These CMs were administered to stationary cardiomyocytes with or without an AII type 1 (AT1) receptor antagonist (losartan), an ET-1 type A (ET(A)) receptor antagonist (BQ610), or anti-TGF-beta antibodies. By measuring the mRNA levels of the proto-oncogene c-fos and the hypertrophy marker gene atrial natriuretic peptide (ANP), the molecular phenotype of the CM-treated stationary cardiomyocytes was characterized. Our results showed that c-fos and ANP expression in stationary cardiomyocytes was increased by All release from cardiomyocytes that had been stretched for 60 min. Stretched cardiomyocytes, cardiac fibroblasts and endothelial cells released ET-1 which led to increased c-fos and ANP expression in stationary cardiomyocytes. ET-1 released by stretched VSMCs, and TGF-beta released by stretched cardiac fibroblasts and endothelial cells, appeared to be paracrine mediators of ANP expression in stationary cardiomyocytes. These results indicate that AII, ET-1 and TGF-beta (released by cardiac and vascular cell types) act as autocrine/paracrine mediators of stretch-induced cardiomyocyte hypertrophy. Therefore, it is likely that in stretched myocardium the cardiomyocytes, cardiac fibroblasts, endothelial cells and VSMCs take part in intercellular interactions contributing to cardiomyocyte hypertrophy.

HLA-B locus-specific downregulation in human melanoma requires enhancer A as well as a sequence element located downstream of the transcription initiation site.

Human MHC class I molecules are encoded by three different loci (HLA-A, -B, and -C), which are regulated at the transcriptional level through several conserved cis-acting promoter elements. The presence of locus-specific residues throughout the entire promoter region strongly suggests that the various HLA class I loci are differentially regulated. To identify regulatory sequences involved in locus-specific HLA class I gene transcription, a series of truncated HLA-A2 and HLA-B7 promoter-reporter constructs were transfected into melanoma cell lines expressing high and low levels of endogenous HLA-B, but comparable levels of HLA-A. These experiments showed that differential regulation of HLA-B expression in melanoma cell lines is mediated by a previously unidentified co-operative action of enhancer A, located 175 bp upstream of the transcription initiation site (+1), and a specific region of 20 nucleotides located at +13 to +33 bp downstream of the transcription initiation site. Furthermore, we demonstrated binding of transcription factor Yin Yang 1 to the HLA-A +13/+33 bp region, but not to the equivalent HLA-B region. Based on these results, we present a model suggesting that YY1 displaces either activating or repressing transcription factors, thereby making the HLA-A gene resistant to differential regulation.

Rapid effects of stretched myocardial and vascular cells on gene expression of neonatal rat cardiomyocytes with emphasis on autocrine and paracrine mechanisms.

Passive stretch of the heart has a direct effect on cardiomyocytes and other cell types including cardiac fibroblasts, endothelial cells, and vascular smooth muscle cells (VSMCs). Cardiomyocytes are targets for the action of peptide growth factors found in myocardium, suggesting an autocrine or paracrine model of the hypertrophic process. In this study we examined stretch-dependent cellular communication between cardiomyocytes, cardiac fibroblasts, endothelial cells, and VSMCs. Stationary cardiomyocytes were incubated with stretch-conditioned medium (CM0-CM60) derived from stretched (for 0-60 min) cardiomyocytes, cardiac fibroblasts, endothelial cells, and VSMCs. The expression levels of protooncogenes (as c-fos, c-jun, and fra-1) were measured, and as an indication of a hypertrophic response the expression of atrial natriuretic peptide (ANP) was measured. Stationary cardiomyocytes that have been incubated for 30 min with CM from stretched (for 0-60 min) cardiomyocytes, cardiac fibroblasts, endothelial cells, and VSMCs showed distinct gene expression patterns that were time-dependent and cell-type specific. In stationary cardiomyocytes, CM derived from stretched cardiomyocytes caused decreased c-fos and fra-1 expression by 37 and 20%, respectively (CM30), elevated c-jun expression by 20% (CM45-CM60), and increased ANP expression by 106% (CM45). CM derived from stretched cardiac fibroblasts caused increased c-fos expression by 41% (CM60), no significant changes in c-jun expression, and increased fra-1 and ANP expression by 39 and 20%, respectively (CM45). CM derived from stretched VSMCs induced an initial decrease in c-fos expression followed by an increase of 13% (CM45) and induced increased c-jun, fra-1, and ANP expression by 39, 24, and 22%, respectively. CM15-CM60 derived from stretched endothelial cells caused decreased c-fos, c-jun and fra-1 expression by 20, 25, and 25%, respectively, and increased ANP expression by 18%. Our data indicate that gene expression of cardiomyocytes in stretched myocardium is regulated by mediators released by cardiomyocytes, cardiac fibroblasts, endothelial cells, and VSMCs. This observation emphasizes the involvement of nonmyocyte cells in the early stages of cardiomyocyte hypertrophy caused by cardiac stretch.

Retargeting of a T cell line by anti MAGE-3/HLA-A2 alpha beta TCR gene transfer.

BACKGROUND: The T cell receptor (TCR) is an heterodimeric protein on the cell membrane of cytotoxic T cells (CTLs). In CTLs TCRs mediate the recognition of target cells through interaction with specific, MHC class I presented peptides. MATERIALS AND METHODS: As a model system to show proof of principle we chose the Jurkat/MA cell line and the HLA-A2.1 binding MAGE-3 derived peptide 271-279, as target specificity. RESULTS: We show that this cell line can be successfully transduced with the dicistronic retroviral vector (LZRS) containing cDNAs encoding for the complete alpha and beta chains of the selected TCR. Following retroviral transduction, Jurkat/MA cells do express the anti-MAGE-3 TCR on their membrane. The transduced TCR is functional as travoductants are successfully triggered, upon stimulation with T2 cells or MAGE-3+ melanoma cells loaded with the MAGE-3 peptide. CONCLUSION: We conclude that TCR gene transfer is possible and it represents a powerful therapeutic tool for the genetical modification of T calls of patients sullering from cancer.

Rapid gene transcription induced by stretch in cardiac myocytes and fibroblasts and their paracrine influence on stationary myocytes and fibroblasts.

Functional adaptation of cardiac cells in response to haemodynamic load requires dynamic alteration of gene expression. In this study, we examined early changes in gene expression following stretch in myocytes and fibroblasts isolated from neonatal rat hearts. In the first hour of biaxially applied static stretch, the changes in expression of immediate-early genes, such as c-fos, c-jun and fra-1, were quantified. The expression of the atrial natriuretic peptide (ANP) gene in myocytes was measured as an indication of the hypertrophic response. In stretched myocytes, expression of c-fos and ANP increased transiently to 227% and 176% respectively after 30 min stretch, whereas c-jun and fra-1 expression decreased in the 1st hour of stretch. In stretched fibroblasts the expression of c-fos and fra-1 increased transiently to maxima of 145% and 146% respectively after 30 min stretch, whereas c-jun expression did not change significantly. To study the indirect effects of stretch, as an indication of cross-talk between cardiac cells, stationary myocytes and fibroblasts were incubated with stretch-conditioned medium (CM) from stretched (0-60 min) myocytes and fibroblasts. CM from stretched myocytes reduced c-fos and induced c-jun expression in myocytes and fibroblasts, reduced fra-1 expression in myocytes but induced fra-1 expression in fibroblasts. CM from stretched fibroblasts induced c-fos expression and had little effect on c-jun expression in myocytes and fibroblasts, induced the fra-1 expression in myocytes but had little effect on fra-1 expression in fibroblasts. CM from myocytes and CM from fibroblasts induced ANP expression in myocytes to 206% and 120% respectively after 45 min stretch. Static stretch of myocytes and fibroblasts appears to stimulate, within 1 h, secretion of cell type-specific factors that participate in the regulation of proto-oncogene and ANP expression of stationary myocytes and fibroblasts. These early changes in gene transcription suggest that stretch of the myocardium initiates intracellular gene expression as well as cross-talk between the cell types.

The influence of the oncogene N-ras on cell cycle delay in a human melanoma cell line with reduced radioresistance.

In a previous study we found that transfection of a human melanoma cell line with the oncogene N-ras led to increased radiosensitivity as measured by clonogenic assays. Since a shift in radiosensitivity is often correlated with altered G2/M delay, we investigated whether this was also the case in this oncogene containing melanoma cell line (IGRras). A human melanoma cell line, stably transfected with mutated N-ras, and its parental cell line transfected with the neomycin phosphotransferase gene only (IGRneo), were irradiated with 5 Gy and cell cycle distribution was measured at hourly time intervals by DNA staining with propidium iodide. Next, the effect of ionising radiation on the duration of the S-phase was determined by pulse labelling cells with BrdUrd before irradiation. Both cell lines showed a radiation induced G2/M delay, which was most prolonged for the ras transfected cell line. After 5 Gy, the S-phase duration was unaltered, although the shape of the relative movement (RM) curves was slightly different. No G1 delay was observed in either cell line. Ras transfection in a melanoma cell line leads to prolonged G2/M delay after radiotherapy. This prolongation is associated with increased radiosensitivity and not with radioresistance. These data throw doubt on the use of oncogene expression or G2/M delay as predictors of radiosensitivity.

Vaccination of melanoma patients with an allogeneic, genetically modified interleukin 2-producing melanoma cell line.

Thirty-three metastatic melanoma patients were vaccinated according to a phase I-II study with an allogeneic melanoma cell line that was genetically modified by transfection with a plasmid containing the gene encoding human interleukin 2 (IL-2). The cell line expresses the major melanoma-associated antigens and the HLA class I alleles HLA-A1, -A2, -B8, and Cw7. All patients shared one or more HLA class I alleles with this cell line vaccine. Patients were immunized by three vaccinations, each consisting of 60 x 106 irradiated (100 Gy) melanoma cells (secreting 120 ng of IL-2/10(6) cells/24 hr) administered subcutaneously at weekly intervals for 3 consecutive weeks. Side effects of treatment consisted of swelling of locoregional lymph nodes and induration at the site of injection, i.e., a delayed-type hypersensitivity (DTH) reaction. In three patients, vaccination induced inflammatory responses in distant metastases containing necrosis or apoptosis along with T cell infiltration. Apoptosis occurred only in Bcl-2-negative areas, not in Bcl-2-expressing parts of the metastases. Two other patients experienced complete or partial regression of subcutaneous metastases. Seven patients had protracted stabilization (4 to >46 months) of soft tissue metastases, including one patient who developed vitiligo after vaccination. Immune responses to the vaccine could be detected in 67% of the 27 patients measured. Vaccination was shown to induce a variable change in the number of anti-vaccine cytotoxic T lymphocytes (CTLs) in peripheral blood, which did not correlate with response to treatment. However, in two of five patients the frequency of anti-autologous tumor CTLs measured was significantly higher than before vaccination. This study demonstrates the feasibility, safety, and therapeutic potential of vaccination of humans with allogeneic, gene-modified tumor cells, and that frequencies of vaccine-specific CTLs among patient lymphocytes can be determined by using a modified limited dilution analysis (LDA).

Interleukin-2-induced, melanoma-specific T cells recognize CAMEL, an unexpected translation product of LAGE-1.

Melanoma-specific cytotoxic T lymphocytes (CTLs) were induced by in vitro stimulation of peripheral blood mononuclear cells of a melanoma patient with autologous IL-2-producing melanoma 518/IL2.14 cells. CTL clone 1/29 recognized, in addition to autologous melanoma cell lines, a panel of HLA-A*0201-expressing allogeneic melanoma cell lines but was not reactive with normal melanocytes. Here, we report the full molecular characterization of the target structure for CTL 1/29, which was identified by cDNA expression cloning. The recognized antigen was named CAMEL (CTL-recognized antigen on melanoma). The CAMEL cDNA turned out to be derived from the LAGE-1 gene, a recently described tumor antigen that is strongly homologous to NY-ESO-1. CAMEL, however, is not encoded by the putative open reading frame (ORF) of LAGE-1 but by an alternative frame starting from the second ATG of the mRNA. The first 11 amino acids of the CAMEL protein, MLMAQEALAFL, constitute the epitope of CTL 1/29. This epitope is also encoded by a similar alternative ORF in NY-ESO-1. In summary, CTL induction with IL-2-transfected melanoma cells has revealed a new tumor antigen that may serve as a target for immunotherapy.

p53 mutations in human cutaneous melanoma correlate with sun exposure but are not always involved in melanomagenesis.

In melanoma, the relationship between sun exposure and the origin of mutations in either the N-ras oncogene or the p53 tumour-suppressor gene is not as clear as in other types of skin cancer. We have previously shown that mutations in the N-ras gene occur more frequently in melanomas originating from sun-exposed body sites, indicating that these mutations are UV induced. To investigate whether sun exposure also affects p53 in melanoma, we analysed 81 melanoma specimens for mutations in the p53 gene. The mutation frequency is higher than thus far reported: 17 specimens (21%) harbour one or more p53 mutations. Strikingly, 17 out of 22 mutations in p53 are of the C:G to TA or CC:GG to TT:AA transitional type, strongly suggesting an aetiology involving UV exposure. Interestingly, the p53 mutation frequency in metastases was much lower than in primary tumours. In the case of metastases, a role for sun exposure was indicated by the finding that the mutations are present exclusively in skin metastases and not in internal metastases. Together with a relatively frequent occurrence of silent third-base pair mutations in primary melanomas, this indicates that the p53 mutations, at least in these tumours, have not contributed to melanomagenesis and may have originated after establishment of the primary tumour.

Influence of increased c-Myc expression on the growth characteristics of human melanoma.

Overexpression of the proto-oncogene c-myc has been associated with neoplastic transformation in a variety of tumors. For human melanoma high c-myc expression has been found in the vertical growth phase and higher positivity reported in metastases than primary tumors. The principle aim of this study was to determine, whether c-Myc expression influences the metastatic behavior of human melanoma in the absence of lymphocyte-mediated immune phenomena. The growth characteristics and tumor biology of two c-myc transfectants of the human melanoma cell line IGR39D, expressing c-Myc 1.7 and three times over baseline and the respective vector control were analyzed both in vitro and in a severe combined immunodeficient mouse model in vivo. Both c-myc transfectants showed increased growth rates, anchorage independent growth and directed cell movement in culture. Subcutaneously implanted IGR39D melanomas highly overexpressing c-Myc spontaneously formed macroscopic metastases (lymph nodes and lung) in severe combined immunodeficient mice in all cases (n = 7 per group), whereas less prominent c-Myc overexpression caused the development of only lung micrometastases. During the time period leading to terminal disease in animals injected with c-myc transfected human melanoma cells, melanoma development was not seen in vector controls. These findings suggest that constitutive high c-Myc expression in human melanoma results in a more aggressive growth behavior both in vitro and in vivo and favors metastasis in severe combined immunodeficient mice by factors unrelated to immune phenomena such as class I human leukocyte antigen downregulation known to be associated with c-Myc expression.

Role of the interferon-stimulated response element in HLA-B downregulation in human melanoma cell lines.

Tumor cells are thought to escape immune surveillance from T cells by suppressing expression of major histocompatibility complex (MHC) class I molecules at their cell surface. Human MHC class I molecules are encoded by three different loci (HLA-A, -B, and -C). In primary human melanomas as well as melanoma cell lines, HLA class I expression is frequently downregulated in a B locus-specific manner. To study the involvement of promoter elements in HLA-B locus-specific downregulation, a series of reporter constructs containing 5'-flanking sequences of the HLA-A2 and -B7 genes were transfected into melanoma cell lines expressing high and low levels of HLA-B antigens. It is shown that enhancer A, which is generally believed to be a potent enhancer in HLA class I gene transcription, only weakly activates transcription in melanoma cell lines. In contrast, the interferon-stimulated response element (ISRE), known to induce MHC class I expression in response to IFNs, as well as a region comprising site alpha/enhancer B significantly stimulate constitutive transcription of HLA class I genes. Although none of the promoter elements tested could be demonstrated to mediate HLA-B locus-specific downregulation, high and low HLA-B melanoma cell lines do differ in ISRE activity as well as in ISRE-binding nuclear factors. The finding that high and low HLA-B melanoma cell lines contain different transcription factors binding to elements not actively involved in the process of HLA-B locus abrogation suggests that these cell lines originate from distinct types of melanocyte precursor cells expressing a different set of transcription factors.

HLA-B down-regulation in human melanoma is mediated by sequences located downstream of the transcription-initiation site.

Major histocompatibility complex (MHC, HLA in humans) class I molecules play an important role in cellular immunology by presenting viral, tumor-associated or minor histocompatibility antigen-derived peptides to T cells. Tumor cells frequently fail to express one or more of the different MHC class I loci (HLA-A, -B and -C), thereby avoiding elimination by T cells. In primary human melanomas as well as melanoma cell lines, HLA class I expression is frequently down-regulated in a B locus-specific manner. The HLA class I promoter contains a number of cis-regulatory elements located upstream of the transcription-initiation site, among them enhancer A and an interferon-stimulated response element. In the present study, we show that novel sequences located 13 to 33 bp downstream of the transcription-initiation site mediate HLA-B locus-specific down-regulation in human melanoma cell lines. Furthermore, involvement of the +13 to +33-bp region in HLA-B locus-specific down-regulation in vivo is supported by in vitro experiments showing locus-specific binding of protein complexes to the +13 to +33-bp region.

Repression of the minimal HLA-B promoter by c-myc and p53 occurs through independent mechanisms.

Major Histocompatibility Complex (MHC, HLA in humans) class I antigens play an important role in cellular immunology by presenting antigens to T cells. Downregulation of MHC class I expression is thought to be a mechanism by which tumor cells escape from T cell-mediated lysis. In primary human melanomas and melanoma cell lines, HLA-B expression is frequently downmodulated, correlating with elevated expression of the c-myc oncogene. Transfection experiments have shown that c-myc induces HLA-B downregulation through a -68 to +13 base pairs (bp) core promoter fragment, containing CCAAT and TATA-like (TCTA) boxes. Since (i) c-myc has been reported to activate the human p53 promoter and (ii) p53 is capable of repressing a large array of basal promoters, we investigated whether c-myc-induced HLA-B abrogation is mediated by p53. In this article, it is shown that the HLA-B core promoter is indeed repressed by wild-type p53, making p53 a candidate for mediating c-myc-induced HLA-B downregulation. However, transfection of c-myc into p53-null cell lines still resulted in suppression of the basal HLA-B promoter, demonstrating that c-myc and p53 repress the minimal HLA-B promoter through independent mechanisms.

Definition of unique and shared T-cell defined tumor antigens in human renal cell carcinoma.

From peripheral blood mononuclear cells of a patient with renal cell carcinoma (RCC), we isolated several T-cell clones, which efficiently lyse the autologous RCC cell line (LE-8915-RCC), but not the autologous Epstein Barr virus-transformed lymphoblastoid cell line. Most of the cytotoxic T lymphocyte (CTL) clones recognize HLA-A1-positive allogeneic RCC cell lines, indicating that HLA-A1 is the restricting element for these T cells. One CTL clone exclusively recognizes the autologous tumor cells. The HLA-A1-restricted CTL clones can be divided further into two subsets of T-cell clones, one blocked by an HLA-A1-specific monoclonal antibody, the other not. The reactivity of HLA-A1-restricted T-cell clone 6/135 was studied in greater detail. This T-cell clone also recognizes a number of melanoma cell lines, indicating that expression of the antigen seen by this CTL clone is not restricted to RCC. Strikingly, the antigen is not exclusively expressed by tumor cell lines, because primary cultures of proximal tubulus epithelium cells, adult mesangial cells, and normal breast epithelium cells are also lysed. These results corroborate the notion that renal carcinoma cells are immunogenic by virtue of a broadly distributed antigenic structure that may serve as a target for cytotoxic T cells and may be a potential candidate for tumor vaccine development.

Expression of MAGE, gp100 and tyrosinase genes in uveal melanoma cell lines.

In order to determine the possible use of uveal melanoma cell lines as stimulators in immunotherapy, we evaluated the expression of the human genes for MAGE-1, -2 and -3, gp100 and tyrosinase in uveal melanoma cell lines. mRNA expression of the MAGE-1, -2 and -3, gp100 and tyrosinase genes and the HLA class I specificity were determined in five primary and three metastatic uveal melanoma cell lines. Expression of the examined genes was heterogeneous in the primary and metastatic cell lines. The cell lines OCM-1 and OMM-1 expressed MAGE-1, -2 and -3, whereas EOM-3, MEL202, 92-1 and OMM-3 were negative for these antigens. gp100 was expressed in all cell lines, and tyrosinase in all but three (EOM-29, OMM-2 and OMM-3). Except for EOM-3, the HLA-A type of all the cell lines could be determined by complement-dependent microlymphocytotoxicity assay. Since at least two melanoma-associated antigens can be found in uveal melanoma cell lines, as well as the HLA class I molecules, these cell lines may be applicable as immunogens for specific immunotherapy against metastatic uveal melanoma.

Identification of HLA-A*0201-restricted CTL epitopes encoded by the tumor-specific MAGE-2 gene product.

MAGE-2 is expressed in many tumors, including melanoma, laryngeal tumors, lung tumors and sarcomas, but not in healthy tissue, with the exception of testis. Thus, MAGE-2-derived peptides that bind to HLA class I molecules and elicit cytotoxic T lymphocyte (CTL) responses could be of significant therapeutic importance. In this study, we show that several MAGE-2-derived peptides bind with high affinity to HLA-A*0201. Three of them form complexes with HLA-A*0201 that are stable at 37 degrees C and are immunogenic in HLA-A*0201Kb transgenic mice. Moreover, CTLs against 2 of them (M2 112-120, and M2 157-166) specifically recognize cells that express both the MAGE-2 protein and HLA-A*0201Kb. These 2 peptides are processed and presented in the context of HLA-A*0201. Therefore, these peptides are candidate components in peptide-based vaccines for the treatment and prevention of several types of MAGE-2-expressing cancers.

Transfection of IL-2 augments CTL response to human melanoma cells in vitro: immunological characterization of a melanoma vaccine.

We have transfected human melanoma cell line 518A2 with the cDNA encoding interleukin-2 (IL-2) or granulocyte-macrophage colony-stimulating factor (GM-CSF), and compared cytokine-producing clones for their ability to induce melanoma-specific cytotoxic T lymphocytes (CTL) from autologous peripheral blood mononuclear cells (PBMC) in vitro. The parental cell line expressed HLA-A1, HLA-A2, ICAM-1, LFA-3, in addition to the common CTL antigens MAGE-1, MAGE-3, tyrosinase, gp100, and Melan-A/MART-1. Stimulation of autologous PBMC responders with the IL-2-transfected clone 518/IL2.14 specifically induced CTL lines reactive with all cell lines derived from the autologous patient. Strikingly, GM-CSF-transfected 518A2 cells did not induce anti-tumor CTL reactivity. CTL induction against 518/IL2.14 was independent of HLA class II expression or CD4 help. The parental cell line 518A2 gained immunogenic properties when high concentrations of IL-2 were supplied exogenously, indicating that IL-2 produced and present at high levels locally by itself enhanced immunogenicity. From the autologous CTL line reactive with 518/IL2.14, clones were generated against an as yet unknown antigen, which was present in all autologous melanoma cell lines as well as in 7 of 15 HLA-A2+ melanoma cell lines tested, but not in melanocytes. These results will be discussed with respect to the possibility of using IL-2-transfected melanoma cells as a vaccine for treatment of patients with melanoma.