Boris Kubuschok

Sustained High Frequencies of Specific CD4 T Cells Restricted to a Single Persistent Virus

ABSTRACT Replication of cytomegalovirus (CMV) is largely controlled by the cellular arm of the immune response. In this study the CMV-specific CD4 T-cell response was characterized in a cohort of apparently healthy individuals. In 11% of all individuals, extremely high frequencies, between 10 and 40%, were found. High-level frequencies of CMV-specific CD4 T cells persisted over several months and were not the result of an acute infection. Specific T cells were oligoclonal and were phenotypically and functionally characterized as mature effector cells, with both cytokine-secreting and proliferative potential. These high-level frequencies do not seem to compromise the immune response towards heterologous infections, and no signs of immunopathology were observed. Whereas a large temporary expansion of virus-specific T cells is well known to occur during acute infection, we now show that extremely high frequencies of virus-specific T cells may continuously exist in chronic CMV infection without overtly compromising the remaining protective immunity.

Prospective study on the expression of cancer testis genes and antibody responses in 100 consecutive patients with primary breast cancer.

To determine the expression of cancer testis (CT) genes and antibody responses in a nonselected population of patients with primary breast cancer, we investigated the composite expression of 11 CT genes by RT‐PCR in fresh biopsies of 100 consecutive cases of primary breast carcinoma and by immunohistology in selected RT‐PCR‐positive cases. Antibody responses against 7 CT antigens were analyzed using recombinant antigen expression on yeast surface. In 98 evaluable cases, SCP‐1 and SSX‐4 were expressed most frequently (both 65%), followed by HOM‐TES‐85/CT‐8 (47%), GAGE (26%), SSX‐1 (20%), NY‐ESO‐1 (13%), MAGE‐3 (11%), SSX‐2 (8%), CT‐10 (7%), MAGE‐4 (4%) and CT‐7 (1%). One CT gene was expressed by 90% of the cases; 79% expressed ≥2, 48% ≥3, 29% ≥4, 12% ≥5, 6% ≥6, 3% ≥7, 2% ≥8 and one case coexpressed 9 antigens. Of 100 serum samples screened for CT antigen‐specific antibodies, antibodies against NY‐ESO‐1 were detected in 4 patients, against SCP‐1 in 6 patients and against SSX‐2 in 1 patient, while no antibodies were detected against MAGE‐3, CT‐7 and CT‐10. Expression of CT genes or antibody responses was not correlated with clinical parameters (menopausal status, tumor size, nodal involvement, grading, histology and estrogen receptor status) or the demonstration of CT gene expression at the protein level, by immunohistology. Our results show that breast carcinomas are among the tumors with the most frequent expression of CT antigens, rendering many patients potential candidates for vaccine trials.

EXPRESSION OF CANCER TESTIS ANTIGENS IN PANCREATIC CARCINOMA CELL LINES, PANCREATIC ADENOCARCINOMA AND CHRONIC PANCREATITIS

In order to define antigens that might be suitable as vaccines for pancreatic carcinoma, we investigated the composite expression of 10 cancer testis (CT) antigens (SCP‐1, NY‐ESO‐1, SSX‐1, SSX‐2, SSX‐4, GAGE, MAGE‐3, MAGE‐4, CT‐7 and CT‐8) by Reverse Transcriptase‐PCR (RT‐PCR) in fresh biopsies of human pancreatic adenocarcinoma, chronic pancreatitis and pancreatic carcinoma cell lines. While all CT genes were frequently expressed in cell lines derived from pancreatic cancer, no expression of MAGE‐3, SSX‐1, SSX‐2, NY‐ESO‐1 and CT‐7 was detected in fresh tumor biopsies, and MAGE‐4 (1/52), SSX‐4 (1/39) and CT‐8 (2/41) were only rarely expressed. In contrast, HOM‐TES‐14/SCP‐1 was expressed in 48% (29/61) and GAGE in 21% (13/61) of cases, respectively. One CT gene was expressed by 59% (75% in male, 46% in female patients; p = 0.05) and 2 or more CT genes by 15% of the samples. SCP‐1 protein expression correlated well with mRNA expression. While SCP‐1 and GAGE were absent in normal pancreas, they were found in 2/8 (SCP‐1) and 1/8 (GAGE) samples of chronic pancreatitis, respectively, supporting the concept of chronic pancreatitis as a premalignant condition. SCP‐1 and GAGE represent promising candidates for vaccine development in pancreatic carcinoma. Whether SCP‐1 and GAGE expression identify cases of chronic pancreatitis with a high risk of malignant transformation remains to be shown.

Serological identification of breast cancer‐related antigens from a Saccharomyces cerevisiae surface display library

A yeast cell surface display technology was used for the isolation and characterization of tumor antigens recognised by autologous or allogeneic breast cancer serum. More than 100 clones recognized by patient serum were isolated using high‐through‐put fluorescence activated cell sorting. Combined serological and sequence analysis confirmed that a number of proteins known to be overexpressed in breast cancer tissue could be detected. A recently identified small breast epithelial mucin almost exclusively expressed in mammary gland tissue was isolated as a mutated protein variant. Subsequent serological analysis using the yeast expression system for the wild‐type and mutant form showed a strong recognition by patient sera, whereas no significant recognition was observed for the respective prokaryotically expressed proteins. The small breast epithelial mucin is present to a large extent in a membrane bound format and might be used for tumor targeting strategies.

Naturally occurring T-cell response against mutated p21 ras oncoprotein in pancreatic cancer.

Mutated p21 ras proteins (muRas) are present in approximately 90% of pancreatic adenocarcinomas and express mutants which can function as cancer-specific antigens. To evaluate the frequency and magnitude of the natural T-cell response against muRas in 19 HLA-A2-positive patients with muRas-positive pancreatic carcinomas, antigen-experienced T lymphocytes in fresh peripheral blood mononuclear cells were shown by IFN-gamma enzyme-linked immunospot using muRas peptides (5-21) that encompass both HLA class I (HLA-A2)- and class II-restricted (HLA-DRB1) epitopes. Six of 19 patients (32%) were found to have a specific T-cell response against individual mutation-specific ras(5-21) but not against other ras mutations or wild-type ras. In contrast, none of 19 healthy subjects had T cells specifically secreting IFN-gamma (P = 0.004). The T-cell response consisted of both CD8(+) and CD4(+) T cells but was dominated by CD8 T cells in three of four patients. MuRas(5-14) and muRas(6-14) were shown to specifically induce CD8(+) T-cell mediated cytotoxicity against HLA-A2-positive, muRas-bearing pancreatic carcinoma cells. The T-cell response was not correlated with prognostic or clinical variables such as tumor-node-metastasis status, stage, or survival. In conclusion, a natural T-cell response against muRas proteins that could be exploited for immunostimulatory therapeutic approaches has been shown in a significant proportion of patients with pancreatic cancer.

Identification of an HLA‐DR‐restricted peptide epitope with a promiscuous binding pattern derived from the cancer testis antigen HOM‐MEL‐40/SSX2

The SSX2 gene encodes the tumor‐specific antigen HOM‐MEL‐40/SSX2 expressed in a broad spectrum of tumors of different origin, against which humoral and CD8+ T‐cell‐mediated MHC‐I‐restricted responses have been demonstrated. Searching for promiscuous MHC‐II‐restricted peptides that might be suitable as a CD4+ stimulating vaccine for many patients, we used the SYFPEITHI algorithm and identified a HOM‐MEL‐40/SSX2‐derived pentadecamer epitope (p45–59) that induced specific CD4+ T‐cell responses restricted by the HLA‐DRB1 subtypes *0701, *1101 and *1302 that have a cumulative prevalence of ∼ 25% in the Caucasian population. The CD4+‐mediated response against p45–59 and its DR restriction was demonstrated by inhibition with anti‐CD4 and HLA‐DR antibodies, respectively, and by blocking experiments using HLA‐specific antibodies. The natural processing and presentation of p45–59 was demonstrated by recognition of the SSX2+ melanoma cell line Me 275 as well as autologous and allogeneic dendritic cells pulsed with whole‐protein SSX2 by T cells with specificity for p45–59. p45–59 was able to induce responses in 3/6 breast cancer patients and 1/5 healthy controls. No correlation was found between CD4+ T‐cell responses against p45–59 reactivity and anti‐SSX2 antibody titers in the serum of patients, suggesting that CD4+ and B‐cell responses are regulated independently. p45–59 holds promise as a broadly applicable peptide vaccine for patients with SSX2‐positive neoplasms.

Identification of an antigenic peptide derived from the cancer-testis antigen NY-ESO-1 binding to a broad range of HLA-DR subtypes.

NY-ESO-1 is a SEREX-defined cancer-testis antigen of which several MHC I, but only few MHC II–restricted epitopes have been identified. Searching for highly promiscuous MHC II–restricted peptides that might be suitable as a CD4+ stimulating vaccine for many patients, we used the SYFPEITHI algorithm and identified an NY-ESO-1–derived pentadecamer epitope (p134–148) that induced specific CD4+ T-cell responses restricted to the HLA-DRB1 subtypes *0101, *0301, *0401, and *0701 that have a cumulative prevalence of 40% in the Caucasian population. The DR restriction of the p134–148 pentadecamer was demonstrated by inhibition with an HLA-DR antibody and a functional peptide displacement titration assay with the pan-DR-binding T-helper epitope PADRE as the competitor. The natural processing and presentation of this epitope was demonstrated by recognition of an NY-ESO-1+ melanoma cell line by T cells with specificity for p134–148. The pentadecamer p134–148 was able to induce CD4+ responses in 4/38 cancer patients tested. However, no strict correlation was found between CD4+ T-cell responses against p134–148 reactivity and anti-NY-ESO-1 antibody titers in the serum of patients, suggesting that CD4+ and B-cell responses are regulated independently. In conclusion, p134–148 holds promise as a broadly applicable peptide vaccine for patients with NY-ESO-1–positive neoplasms.

Analysis of the B cell repertoire against autoantigens in patients with giant cell arteritis and polymyalgia rheumatica

The analysis of the antibody repertoire of patients with giant cell arteritis (GCA) and polymyalgia rheumatica (PMR) might identify target antigens of the autoimmune response with potential relevance to our understanding of the pathogenesis of the disease and to the development of serodiagnostic tests. To detect such antigens, we screened a cDNA library derived from normal human testis for antigens reacting with IgG antibodies in the 1 : 250 diluted sera of three patients with untreated GCA using SEREX, the serological identification of antigens by recombinant cDNA expression cloning. Of 100 000 clones screened with each serum, six, 28 and six clones, respectively, were positive, representing a total of 33 different antigens. Most of the antigens reacted only with the serum used for identification and/or at a similar frequency with normal control sera. However, lamin C and the nuclear antigen of 14 kD reacted specifically with 32% of GCA/PMR, but with none of the control sera, while human cytokeratin 15, mitochondrial cytochrome oxidase subunit II, and a new gene product were detected preferentially, but not exclusively by sera from GCA/PMR patients. We conclude that patients with GCA/PMR develop antibodies against a broad spectrum of human autoantigens. Antibodies against human lamin C, the nuclear autoantigen of 14 kD as well as human cytokeratin 15, mitochondrial cytochrome oxidase subunit II and the product of a new gene should be investigated further to determine their value as tools for the diagnosis and/or the definition of clinical subgroups of patients with GCA/PMR.

Serological immune response to cancer testis antigens in patients with pancreatic cancer

Serological screening approaches have allowed for the identification of a large number of potentially relevant tumor antigens in cancer patients. Within this group, cancer testis antigens represent promising targets for cancer immunotherapy, since they are widely expressed in a variety of human cancer entities. In pancreatic cancer, however, there are only few data available about the expression pattern and serological response to cancer testis antigens and other serological‐defined tumor antigens. Therefore, we investigated the IgG antibody response against 11 cancer testis antigens (SCP‐1, GAGE, LAGE‐1a,‐1b, CT‐7, NY‐ESO‐1, SSX‐1‐5) recombinantly expressed on yeast surface (RAYS) in patients with pancreatic cancer (n = 96), chronic pancreatitis (n = 18) and healthy donors (n = 48). We found in 14% of all patients antibody responses to SCP‐1, but not to other cancer testis antigens (GAGE, LAGE‐1a,‐1b, CT‐7, NY‐ESO‐1, SSX‐1‐5). Antibody response correlated with the expression of SCP‐1 in the primary tumor of the respective patient as shown by RT‐PCR, immunohistochemistry and Western blot. In contrast, no serological response to cancer testis antigens was observed in healthy donors. The humoral immune response against SCP‐1 was associated with the size of tumor, but not with other clinico‐pathological parameters such as histology, stage, presence of lymph node metastases, grading, age, gender or gemcitabine treatment. In conclusion, antibody response to cancer testis antigen SCP‐1 is found in a proportion of pancreatic carcinoma patients. These results indicate that identification of additional tumor antigens by serological screening of tumor cDNA expression libraries by RAYS is a promising goal in pancreatic cancer.

Use of Spontaneous Epstein-Barr Virus-Lymphoblastoid Cell Lines Genetically Modified to Express Tumor Antigen as Cancer Vaccines: Mutated p21 ras Oncogene in Pancreatic Carcinoma as a Model

Spontaneous Epstein-Barr virus (EBV)-transformed lymphoblastoid cell lines (SP-LCLs) can be easily obtained from latently EBV-infected cancer patients and used as a source of antigen-presenting cells (APCs) for immunotherapy. Using point-mutated (codon 12) p21(ras) (muRas) as a model tumor antigen, we evaluated the practicability of using genetically modified SP-LCLs as cancer vaccines for patients with pancreatic cancer expressing mutated Ras (muRas). The repeated stimulation of peripheral blood mononuclear cells (PBMCs) from patients with muRas-LCLs elicited a strong, muRas-specific T cell response. A significant cytotoxic activity against EBV virus proteins or components of the expression vector was not observed. The T cells were able to recognize naturally presented muRas, as shown by their cytotoxicity against muRas (Gly-12 to Val-12 or Asp-12)-expressing tumor cells. The T cell response was mainly MHC class I restricted, and peptides containing amino acids 5 to 14 of muRas-Val-12 and muRas-Asp-12 were identified as immunogenic peptides for HLA-A2. In contrast to the situation in patients with putatively muRas-primed T cells, muRas-LCLs were not able to prime naive T lymphocytes from healthy controls. Vaccination of a pancreatic cancer patient with muRas-LCL induced muRas-specific T cells in PBMCs after 4 weeks. We conclude that genetically modified muRas-LCLs can efficiently present tumor antigens to the immune system and induce antigen-specific cytotoxic T cell responses in vitro and in vivo.