Identification of an HLA-A0201-Restricted CTL Epitope ... - Springer Link

P1: GRA pp919-joci-469376

JOCI.cls

August 14, 2003

22:19

C 2003) Journal of Clinical Immunology, Vol. 23, No. 5, September 2003 (°

Identification of an HLA-A0201-Restricted CTL Epitope Generated by a Tumor-Specific Frameshift Mutation in a Coding Microsatellite of the OGT Gene1 EVA RIPBERGER,2 MICHAEL LINNEBACHER,2 YVETTE SCHWITALLE,3 JOHANNES GEBERT,2 and MAGNUS VON KNEBEL DOEBERITZ2,4

in sporadic tumors (3). MSI causes length variations at short repetitive DNA stretches, called microsatellites. If MSI affects coding microsatellites, C-terminally truncated proteins will be generated, often comprising neo-peptide tails. These truncated gene products are believed to have an impaired function and might thereby contribute to MSIdriven carcinogenesis. In line with this hypothesis, several genes known to play important roles in colorectal carcinogenesis as for example transforming growth factor beta receptor type II (TGFbetaRII), Bax and TCF-4 are targets of this mechanism (4–6). Following proteasomal degradation, parts of these novel proteins might be presented as immunogenic neo-peptides at the surface of MSI+ cancer cells in the context of MHC-class I molecules, thereby allowing for recognition of frameshift-derived nonself by effector cells of the immune system. The enhanced lymphocytic infiltration and Crohn’s like reaction observed in MSI+ tumors could be due to the presence of this novel class of tumor-associated antigens (TAA) (7, 8). The high prevalence of activated intraepithelial cytotoxic T lymphocytes and increased neoplastic cell apoptosis in MSI+ colorectal carcinomas (9) similarily suggest the presence of potent tumor rejection antigens in MSI+ cancers. Although the idea that frameshift-derived peptides could be used as targets of antitumor immunity has already been proposed (10), formal proof that such peptides represent true antigens expressed by tumor cells and recognized by human CTL has

Accepted: April 16, 2003

Deficient DNA mismatch repair results in microsatellite instability and might induce shifts of translational reading frames of genes encompassing coding microsatellites. These may be translated in truncated proteins, including neo-peptide tails functioning as tumor rejection antigens, when presented in the context of MHC class I. Recently, others and we identified a frameshift mutation in the coding T(10) microsatellite of the O-linked N acetylglucosamine transferase gene (OGT) occuring in up to 41% of microsatellite unstable colorectal cancers. Here we describe a novel HLA-A0201-restricted cytotoxic T lymphocyte (CTL)-epitope (28 -SLYKFSPFPL; FSP06) derived from this mutant OGT-protein. FSP06-specific CTL-clones killed peptidesensitized target cells and tumor cell lines expressing both HLAA0201 and mutant OGT proteins. This demonstrates that FSP06 is endogenously expressed and represents a CD8+ -T cell epitope. Our data corroborate the concept of frameshift peptides constituting a novel subset of tumor-associated antigens specifically encountered in cancer cells with deficient mismatch repair. KEY WORDS: Microsatellite instability; frameshift peptides; tumor antigens; DNA mismatch repair; T-cell epitopes.

INTRODUCTION

Microsatellite instability (MSI) due to DNA mismatch repair (MMR) deficiency is observed in colorectal adenomas and carcinomas from individuals with hereditary nonpolyposis colorectal cancer (HNPCC) (1, 2) but also occurs

2 Institute

of Molecular Pathology, Department of Pathology, University of Heidelberg, Heidelberg, Germany. 3 Division of Urology, Department of Surgery, University of Heidelberg, Heidelberg, Germany. 4 To whom correspondence should be addressed at Abteilung f¨ ur Molekulare Pathologie, Pathologisches Institut, Universit¨at Heidelberg, Im Neuenheimer Feld 110, D-69120 Heidelberg, Germany; e-mail: [email protected]. or [email protected].

Abbreviations used: cMS, coding microsatellite; FSP, frameshift peptide; MSI, microsatellite instability; MSS, micosatellite stability; MMR, mismatch repair; OGT, O-linked N -acetylglucosamine transferase; OGT(FS), OGT containing a (−1) frameshift mutation; TAA, tumor associated antigen; TGFbetaRII, transforming growth factor beta receptor type II. 1 Eva Ripberger and Michael Linnebacher contributed equally to this work.

415 C 2003 Plenum Publishing Corporation 0271-9142/03/0900-0415/0 °


JOCI.cls

August 14, 2003

22:19

416

only recently been provided (11). To analyze the role of frameshift peptides (FSP) in host immune response against MMR deficient cells, we performed T cell stimulation experiments using a set of HLA-A0201-restricted FSP. In this initial study, we demonstrated the immunogenicity of a frameshift-induced neo-peptide (FSP02), derived from a (−1) shift in a A(10) cMS of the TGFßRII gene (11). These results have been confirmed by subsequent reports of others (12, 13). Since MMR deficiency seems to be relevant for MSI carcinogenesis, investigators have aimed to identify other commonly mutated cMS in MSI+ tumors. As a result of these efforts, numerous candidates have been identified with varying mutation frequencies (14–16). In particular, frameshift mutations affecting the T(10)-microsatellite in exon 5 of the O-linked N -acetylglucosmine transferase (OGT) gene have been observed with a frequency of up to 41% in MSI+ primary colorectal tumors (14, 16). OGT is a unique gene (in human mapped in Xp13) without other obvious family members but very highly conserved form C. elegans to human (17). It modifies many essential regulators of cellular function by addition of OGlcNAc-residues (18). Among those are many proteins which are well known to be involved in carcinogenesis, like P53, c-Myc, or GSK3ß. It has been suggested, that disturbance of O-GlcNAc-modification may interfere with critical cellular control mechanisms, leading to the transformed phenotype (19). Since cMS of different MSI+ tumors display varying mutation patterns, it seems rationale to identify a common set of frameshift mutations which represent putative TAA generated in the vast majority of MSI+ cancer cells. Such a set of multiple frameshift-induced antigenic peptides might prove valuable for therapeutic or even preventive vaccinations against MSI+ cancers. In this study we report the identification of an OGT(FS) antigenic peptide (28 -SLYKFSPFPL; FSP06) presented to CTL by HLAA0201 molecules. This FSP stimulates CTL that recognize tumor cells in an Ag-specific and HLA-restricted fashion. According to these data the OGT(FS)-derived FSP06 represents a novel frameshift Ag potentially useful for MSI+ tumor-specific vaccination approaches. METHODS

Tumor Cell Lines and CD40 Ligand System for the Culture of Human B Cells All tumor cell lines were obtained from the tumor bank of DKFZ or from ATCC and grown in RPMI 1640 medium supplemented with 10% fetal calf serum and antibiotics. The following colorectal carcinoma cell lines were used

RIPBERGER ET AL.

in this study: SW707 (HLA-A0201+ ; OGT(wt); MSS), LoVo (HLA-A0201− ; OGT(FS); MSI+ ), SW480 (HLAA0201+ ; OGT(wt); MSS). In some experiments T2 cells (HLA-A0201+ ; 174 × CEM.T2 hybridoma, TAP1 and TAP2 deficient) and 293 cells were used. Tissue culture media and supplements were purchased from Invitrogen Corporation (Karlsruhe, Germany) unless indicated otherwise. Cytokines were obtained from R&D (Wiesbaden, Germany). The culture of CD40 ligand activated B cells was performed as described (20). B cells from a healthy HLAA0201+ donor were cultured on NIH/3T3-CD154 feeder cells as described (11). Cloning of FSP06-Specific Cytotoxic T Lymphocytes (CTL) Cloning of peptide FSP06 (SLYKFSPFPL)-specific Tcells was performed by limiting dilution of a T-cell bulk culture. Briefly, 0.7 cells/well were plated in V-bottomed 96-well plates. Lethally irradiated (30 Gy), peptide FSP06 (10 µg/mL) loaded CD40 activated B cells of the autologous HLA-A0201+ healthy donor (3 × 105 /well) were added to diluted T cells in a final volume of 200 µL Iscov’s MDM containing 10% human AB-serum, supplements (1:100) and IL-2 (10 IU/mL). This treatment was repeated weekly for restimulation of T cells. After 7 weeks outgrowing clones were screened for peptidespecific cytotoxic activity in a standard chromium release assay (21, 22). Cloning of OGT-cDNAs Total RNA was isolated from tumor cells using RNeasy Mini Kit (Qiagen, Hilden, Germany). Synthesis of cDNA was performed for 1 h at 37◦ C in a final volume of 20 µL including 1 µg total RNA, 0.5 µg oligo dT12−18 , and 200 U M-MLV Reverse Transcriptase. All reagents were obtained from Invitrogen unless stated otherwise. PCR amplification was carried out using 1 µL cDNA of the LoVo (OGT(FS)) and SW480 (OGT(wt)) cell lines in 50 µL total reaction volume, 5 µL 10x reaction-buffer (Platinum Pfx DNA Polymerase Kit), 1 mM MgSO4 , 0.3 mM dNTPs, 0.3 µM of each primer (OGTfor: 50 -tgg ttt gca ctt ggg ttc tg-30 and OGTrev: 50 -tct cct gtg cag tct tta ttt atg c-30 ) and 2.5 U Platinum Pfx DNA polymerase. The following conditions were used: initial denaturation at 94◦ C for 3 min, followed by 35 cycles of denaturation at 94◦ C for 1 min, annealing at 59◦ C for 30 s and primer extension at 68◦ C for 3 min. The final extension step was carried out at 68◦ C for 15 min. The PCR-products (OGT(FS/wt): 2.946/7bp) were cloned

Journal of Clinical Immunology, Vol. 23, No. 5, 2003


JOCI.cls

August 14, 2003

22:19

CTL RECOGNIZES A FRAMESHIFT PEPTIDE OF OGT

into pCR2.1 using the TOPO TA Cloning Kit according to the manufacturer’s protocol. Cloned inserts were confirmed by DNA-sequencing. Subsequently, fragments were subcloned into the eucaryotic expression vectors pUHD10-1 (23) for OGT(FS) and pEGFP-C2 (Invitrogen) for OGT(FS/wt). Stable Transfection of Colorectal Cancer Cell Lines SW707 and LoVo All transfections were performed with Effectene (Quiagen) according to the manufacturers protocol. LoVo cells were cotransfected with pCDNA1-HLA-A0201 (kindly provided by P. van der Bruggen, Ludwig Cancer Institute, Brussels, Belgium) and pSV2neo (BD Clontech, Heidelberg, Germany). Selection was performed using G418 (200 µg/mL). Stable expression of HLA-A0201 in the resulting clones was identified by FACSCalibur (BD, Heidelberg, Germany). One HLAA0201-expressing clone, LoVo-A2, was used for all subsequent studies. SW707 cells were transfected with pUHD10-1-OGT(FS) in cotransfection with pSV2neo and selected with G418 (1100 µg/mL). One OGT(FS) positive clone, SW707-OGT(FS), was identified by cDNAfragment analysis and used in the following experiments. DNA-Fragment Analysis PCR amplification of the thymidine mononucleotide repeat in exon 5 of the OGT coding sequence was carried out in 50 µL total reaction volume including 1 µL cDNA, 5 µL 10x reaction-buffer, 1.5 mM MgCl2, 200 µM dNTPs, 0.15 µM of each primer (OGTs: 5’-tca ctt ttg gct ggt cag ag-30 and OGTa: 50 -ggg agg gaa agg agg taa ag-30 , fluoresceine-labeled) and 0.5 U Taq DNA-polymerase using the following conditions: initial denaturation at 94◦ C for 3 min, followed by 35 cycles of denaturation at 94◦ C for 30 s, annealing at 60◦ C for 45 s and primer extension at 72◦ C for 1.5 min. The final extension step was carried out at 72◦ C for 5 min. For repeat length analysis, 2 µL of diluted PCR-products (1:10) together with 0.5 µL TAMRA 500 length standard (Applied Biosystems, Darmstadt, Germany) and 12.5 µL formamide were loaded on ABI310 genetic analyzer (Applied Biosystems). Results were analyzed using Genescan Analysis Software (Applied Biosystems). Fragment length alterations were confirmed by DNA sequence analysis. Immunofluorescence Analysis For analysis of HLA-A0201-expression, 1-5 × 105 tumor cells were washed in PBS and incubated with primary mAb BB7.2 (hybridoma culture supernatant) for 20 min


417

on ice. After washing twice in PBS, cells were incubated for 20 min with a secondary FITC-conjugated goat antimouse IgG antibody. Cells were washed twice and subsequently resuspended in PBS containing 1% paraformaldehyde. Cells treated without primary antibody were used as negative control. Fluorescence intensity was analyzed with a FACSCalibur flow cytometer (BD). Chromium Release Assay Standard chromium release assays were performed as described (21, 22). Tumor target cells were labeled with 100 µCi [51 Cr]-sodium chromate for 1 h at 37◦ C. For peptide-recognition, T2 cells were incubated overnight at 37◦ C with 5 µg/mL peptide, washed, and subsequently labeled. For each experimental condition, 103 target cells/well were plated in triplicate in V-bottomed 96well plates. Varying numbers of CTL were added to a final volume of 200 µL and incubated for 4 h at 37◦ C. Spontaneous and maximal release was determined in the presence of medium alone or cellular lysis with 1% NP-40, respectively. Supernatants (100 µL/well) were harvested and counted in a gamma-counter (1282 Compugamma, LKBWallac, Stockholm, Sweden). The percentage of specific lysis was calculated as follows: 100% × experimental release – spontaneous release]/[maximal release – spontaneous release]. Western Blot Analysis Two hundred and ninety-three cells were transiently transfected with pEGFP-C2-OGT(FS) or pEGFP-C2OGT(wt) constructs using SuperFect transfection reagent (Qiagen) according to the suppliers instructions. After 48 h cells were harvested and lysed. An equivalent of 5 × 105 cells was loaded per lane on a SDS-PA gel and the proteins were transferred to nitrocellulose membranes after separation. For detection of fusion proteins, anti-GFP monoklonal Ab (Jl-8) (BD Clontech) as primary antibody and rabbit anti-mouse-HRP as secondary antibody were used and ECL-reaction was performed according to the manufacturer’s protocol (Amersham Biosciences).

RESULTS

Generation of Colorectal Cancer Target Cells That Express Both HLA-A0201 and OGT(FS) To investigate the immunogenic potential of a (−1) frameshift mutation of the T(10) tract in exon 5 within the OGT gene (OGT(FS)), we have generated a T-cell bulk culture from PBMC of a normal healthy donor that


JOCI.cls

August 14, 2003

22:19

418

had been stimulated with the OGT(FS)-derived frameshift peptide 28 -SLYKFSPFPL (FSP06) (11). T cells of this bulk culture showed (i) expression of surface markers characteristic for a CTL phenotype (CD8+ , CD16− , CD25+ , CD69+ , and CD45RO+ ), (ii) IFN-g secretion after stimulation with specific peptide, and (iii) specific recognition of peptide-coated target cells. Detailed analysis of the cytotoxic potential of this frameshift peptide-specific T cell bulk culture, however, was hampered by the lack of tumor cell lines concommitantly expressing OGT(FS) and HLA-A0201. To overcome these limitations, we genetically modified two colorectal cancer cell lines, SW707 and LoVo, to express these genes. SW707 cells are MSS, consequently express the wild-type form of OGT, but they express HLAA0201 molecules. LoVo cells are MSI+ , in particular they express only the mutant OGT(FS) gene, however, they are HLA-A0201 negative. First, we generated stable subclones of the HLA-A0201+ cell line SW707 conferring expression of the mutant OGT(FS) cDNA in addition to the endogenous OGT(wt) gene. Clones maintaining stable expression of the OGT(FS) transcripts (SW707-OGT(FS); Fig. 1) were further identified by RT-PCR and subsequent DNA fragment analysis. In addition to SW707 cells, we were interested to use the MMR-deficient colorectal cancer cell line LoVo as target in functional T cell assays, because LoVo cells are homozygous for the (−1)-mutated form of OGT, and the mutant OGT(FS)-gene is transcribed in these cells (16). To achieve stable HLA-A0201-expressing subclones of LoVo, we transfected an HLA-A0201-expression plasmid into LoVo cells and examined HLA-A0201-expression of the resulting clones (LoVo-A2) by flow cytometry (data not shown). A single HLA-A0201-positive clone of LoVoA2 and a single clone of SW707-OGT(FS) (Fig. 1) were used in all subsequent experiments during this study. Frameshift Mutation Leads to the Expression of Truncated OGT Protein The deletion of one thymidine at the T(10) tract in exon 5 of the OGT coding sequence is expected to cause a frameshift mutation resulting in expression of truncated OGT(FS) protein. This mutant protein should still consist of the first 28 amino acid residues of the wild-type OGT protein followed by a short stretch of 19 amino acids encoded by the altered translational reading frame downstream of the frameshift mutation site. Such aberrant and truncated proteins often are recognized and rapidly processed by the cellular protein degradation machinery and thus are difficult to detect. To confirm that the OGT(FS) mutant allele indeed confers expression of such

RIPBERGER ET AL.

a truncated OGT protein, Western blot analysis was performed. For this purpose we chose an indirect approach by marking the truncated OGT(FS) protein with an Nterminal EGFP protein tag thereby facilitating detection of this EGFP-OGT(FS) fusion protein by a tag-specific antibody. Two types of expression vectors were generated consisting of the EGFP protein fused in frame to either wild-type (EGFP-OGT(wt)) or mutant OGT protein sequences (EGFP-OGT(FS)). After transfection of 293 cells with these fusion gene contructs, two bands of the expected sizes of 35 kDa and 137 kDa corresponding to the EGFP-OGT(FS) and EGFP-OGT(wt) fusion proteins, respectively, were detected on Western blots using a monoclonal anti-GFP antibody (Jl-8; Fig. 2). Detection of the 35-kDa fusion protein indicates correct translation of the predicted OGT(FS) neo-sequence and correct termination of translation due to the premature intrinsic stop codon generated by the (−1) frameshift mutation. Degradation of this truncated protein by the cellular proteasome followed by presentation of OGT(FS) peptides in the context of HLA-class I molecules on the cell surface thus is anticipated to allow for recognition by FS-peptide-HLAspecific T cells. Generation and Characterization of FSP06-Specific CTL Clones Therefore we investigated the lytic reactivity of FSP06specific CTL towards the above-mentioned genetically modified colorectal tumor cell lines. We first generated a large panel of FSP06-specific T-cell clones from the FSP06 specific bulk T cell culture because target cell lysis exerted by CTL clones might be obscured by the polyclonal T cell culture. After around seven weekly restimulations with FSP06-loaded, lethally irradiated autologous CD40 activated B cells, 700 outgrowing clones were tested for their lytic activity. Standard chromium release assays were performed using T2 cells coated with either irrelevant P68-peptide or FSP06 as target cells. Clones were considered to be FSP06-specific, if they displayed no reactivity to P68-loaded T2 cells, but were able to lyse FSP06-coated T2. In this cloning experiment, 136 clones mediated 50– 75%; 130 clones 75–95% and 54 clones more than 95% lysis of FSP06-displaying T2 target cells (data not shown). We next examined the cytotoxic activity of the abovementioned CTL clones in more detail. Figure 3 shows the cytotoxic potential of three CTL clones (named clone 3, clone 17, and clone 23), selected by their complete lack of background reactvity to T2 target cells loaded with control peptide P68 (Fig. 3A, open triangles). In contrast, they displayed high FSP06-specific lysis rates and this reactivity could be titrated (Fig. 3A, closed squares).



JOCI.cls

August 14, 2003

22:19


419

Fig. 1. OGT(FS) is expressed in colorectal carcinoma cell lines. Here we show fragment analysis of fluorescinated PCR-products generated from cDNA of colon carcinoma cell lines: microsatellite stable SW707, expressing the wild-type allele of OGT (T(10)); SW707-OGT(FS), an OGT(FS)transfected cell clone expressing the OGT(FS) allele (T(9)) in addition to the wild-type allele and the MSI+ cell line LoVo, expressing only OGT(FS) (T(9)). Fragment length of wild-type and (−1)-FS alleles is indicated in nucleotides.

OGT(FS)-Specific CTL Clones Recognize Endogenously Processed Antigen We next examined whether the identified peptides are also endogeneously processed and presented on MSI+ cells in an HLA-A0201 restricted manner. When analyzed in chromium release assays, the FSP06-specific CTL clones (clones 3, 17, and 23) exhibited significantly higher lytic activity towards SW707-OGT(FS) colorectal tumor cell targets (Fig. 3B, closed squares) as compared to the OGT(wt) control (Fig. 3B, SW707, open triangles). To exclude that this observation was cell-line-specific, we tested the cytolytic capacity of these FSP06-specific CTL clones against an additional MSI+ colorectal carcinoma cell line, LoVo, known to express only the (−1)-mutated form of OGT. Since the parental LoVo cells lacked HLA-A0201 expression, the FSP06-specific CTL clones consequently failed to lyse LoVo target cells (Fig. 3C, open triangles). However, LoVo-A2 cells, genetically modified to stably express HLA-A0201 on their cell surface, were efficiently


killed by the FSP06-specific CTL clones (Fig. 3C, closed squares). CTL mediated lysis of LoVo-A2 and of SW707OGT(FS) target cells was comparatively lower than lysis of peptide-coated T2 cells, but consistently and reproducibly increased with increasing E:T-ratios. These data clearly demonstrate FSP06 and HLA-A0201 specificity of the FSP06 specific CTL clones. Moreover, lytic activity of FSP06-specific CTL towards tumor target cells expressing the corresponding mutated transcript indicates that expression of OGT(FS)-protein followed by proteasomal degradation and presentation of FSP06 in the context of HLA-A0201 must have occurred. DISCUSSION

Many TAAs have been identified over the last decade that are specifically recognized on human tumor cells by CTL. Beside the most common cancer/testis antigens and melanoma differentiation antigens, a growing number of


JOCI.cls

August 14, 2003

22:19

420

RIPBERGER ET AL.

Fig. 2. EGFP-OGT(wt) and EGFP-OGT(FS) fusion protein expression. SDS-PAGE analysis of 293 cells, transiently transfected with expression vectors encoding for EGFP, EGFP-OGT(FS), or EGFP-OGT(wt), respectively. An aquivalent of 5 × 105 cells was loaded per lane. Lane 1: untransfected 293 cells (Panels A and B); lane 2: 293 transfected with pEGFP-C2 (Panels A and B); lane 3: 293 transfected with pEGFP-C2-OGT(FS) (Panel A) or with pEGFP-C2-OGT(wt) (Panel B). The EGFP-protein itselfs has a molecular weight of 27 kDs (Panel A, lane 2). Because of its relatively low-molecular weight, this band cannot be shown in Panel B, displaying a low concentrated gel used for detection of the large EGFP-OGT(wt) fusion protein. EGFP-OGT(FS) shows the expected molecular weight of 35 kDa (Panel A, lane 3). The 137-kDa band in Panel B, lane 3 corresponds to the predicted molecular weight of EGFP-OGT(wt).

antigenic epitopes derived from tumor-specific expressed gene products have been described. Frameshift mutations in cMS as a consequence of defective DNA MMR should result in expression of neopeptides. We (11) and others (12, 13) postulated that these FSPs represent a novel class of TAAs specific for MSI+ cancer cells. In this study we identified a novel HLA-A0201restricted T cell epitope (FSP06) arising from a frameshift mutation in the OGT gene. This particular frameshift mutation occurs in up to 41% of MSI+ colorectal tumors (14, 16) suggesting a direct link between coding region MSI and tumor-specific FSP antigens. T-cell clones isolated from a FSP06-specific T-cell bulk culture not only specifically killed peptide-sensitized target cells, but also lysed human colorectal cancer cell lines genetically modified to express HLA-A0201 and OGT(FS). This T-cellmediated tumor cell lysis provides compelling evidence for FSP06 being expressed and presented at the surface of MSI+ colorectal cancer cells. Since the presentation of FSP06 tended to increase by treatment with IFN-g (data not shown), it is tempting to speculate that this epitope is produced by both constitutive and immunoproteasomes (24), allowing the initiation of an immune response by professional antigen presenting cells. The identification and characterization of such FSPs defining specific T-cell epitopes has major implications for diagnosis and therapy of MSI+ tumors. First, it firmly establishes that FSPs indeed constitute a novel class of tumor-specific antigens specifically marking MSI+ tumor

cells for T-cell-mediated destruction. Second, the high frequency of frameshift mutations in some cMS leads to FSP antigens shared among MSI+ tumors of a particular organ like the colon. However, recent evidence suggests that organ-specific differences in cMS mutation frequency and thus FSP makeups exist among MSI+ tumors which argue for selection of particular mutations and organ-specific differences in MSI+ tumor biology (15) Although, not shown for OGT in the present study, frameshift mutations in other cMS like TGFßRII occur early in tumorigenesis, already detectable in adenomatous lesions (25). Accordingly, therapeutic or even prophylactic FSP vaccination might represent an attractive approach for treatment and prevention of sporadic and hereditary MSI+ tumors. Together with the recently described TGFßRII(FS) frameshift peptide FSP02 (11, 13), and other immunogenic FSPs yet to be identified, FSP06 thus represents a good candidate to be included into a vaccination cocktail against MSI+ tumor cells. A combinatorial approach may substantially increase the possibility of targeting tumor cell, that have lost one or more immunogenic epitopes and such conceivably would be superior to a vaccination strategy based on only a single FSP antigen. In this context it is important to emphasize that our present study on OGT(FS) only focussed on one HLA-A0201-restricted Tcell epitope (FSP06). For a more general approach and to increase the potential therapeutic efficiency of such an MSI+ tumor-specific vaccine other MHC class I restricted epitopes as well as MHC class II epitopes in these



JOCI.cls

August 14, 2003

22:19


421

Fig. 3. Cytotoxic Activitiy of three FSP06-Specific CTL Clones. Colorectal carcinoma target cell lines used in these chromium release assays were treated with IFN-g (250 U/mL) for 24 h, before starting experiments. Panel A shows specific lytic activity against T2 cells coated with FSP06; no background reactivity was seen against irrelevant P68 peptide. Panel B reveals specific response against SW707-OGT(FS) in contrast to the untransfected SW707 controls. Panel C displays lytic activty against the OGT(FS) endogenously expressing cell line LoVo engineered to express HLA-A0201 (LoVo-A2); no significant reactivity was seen against the HLA-A0201-negative parental LoVo cell line.

FSPs need to be investigated. Furthermore, the vaccination against MSI+ tumors seems to be a very promising approach, as autoimmunity following FSP vaccination can virtually be excluded because expression of the corresponding frameshift mutant proteins is confined to neoplastic cells. The novel T-cell epitope described in this study originates from a frameshift mutation in the T(10) coding region microsatellite of the OGT gene. This mutation will abrogate normal protein function but the resulting truncated OGT(FS) protein still might be able to interact with other proteins. This hypothesis is supported by our preliminary observation that transiently transfected OGT(FS) can interfere with ß-Catenin/TCF/LEFmediated transciptional activation as determined by reporter gene assays (E.R. and M.L., unpublished results) thereby potentially contributing to tumorigenesis. Inter-


estingly, OGT mediated O-GlcNAcetylation has been suggested to be involved in carcinogenesis since proteins encoded by oncogenes (c-myc) and tumor suppressor genes (p53) have been reported to be modified by O-phosphate or O-GlcNAc, either synergistically or reciprocally at residues critical for normal function of these trancription factors. However, final proof that imbalanced OGlcNAcylation leads to carcinogenesis is still missing. Although, the results described in this study have been obtained with T cells from a healthy donor and with genetically modified colorectal cancer cell lines, this very simple approach represents a first but important step in characterizing T-cell responses against and immunogenicity of FSP expressing MSI+ tumor cells. Our results further confirm that the MSI+ phenotype leads to the accumulation and subsequent presentation of immunogenic FSPs derived from truncated proteins. Remarkably, MSI+


JOCI.cls

August 14, 2003

22:19

422

colorectal carcinomas have been reported to show increased infiltration of activated CTL, and a high percentage of apoptotic tumor cells (9). Real-time PCR data further support this finding by demonstrating upregulation of MHC class II in MSI+ tumor but not in matched normal epithelial tissue (26), indicating local proinflammatory cytokine production. Moreover, expression of MHC-class II molecules appears to be a favorable prognostic marker in colon cancer (27). Consequently, MSI+ colorectal tumor patients show better survival rates than patients carrying MSS tumors (28). Altogether, these findings strongly argue in favor of a locally ongoing cellular antitumor immune response, most likely driven by MHC-dependent FSP presentation. However, our findings and predictions have to be verified with T cells from sporadic as well as hereditary MSI+ tumor patients. Moreover, future studies also should address the presence of MHC-class II restricted T helper epitopes in OGT(FS)-protein and the frequency of OGT(FS)specific CTL and T helper responses in patients with OGT(FS)-expressing MSI+ tumors. Another aspect of interaction of the patients immune system with MSI+ tumors, the humoral immune response, has to our knowledge not been analyzed so far. Detection of antibodies specific for FSPs may also have remarkable consequences on diagnosis and therapy of MSI+ tumors. ACKNOWLEDGMENT

This work was partially supported by the Deutsche Krebshilfe. This work is dedicated to Professor Harald zur Hausen on the occasion of his retirement as head of the German Cancer Research Center in Heidelberg. REFERENCES 1. Kinzler KW, Vogelstein B: Lessons from hereditary colorectal cancer. Cell 87:159–170, 1996 2. Aaltonen LA, Peltomaki P, Leach FS, Sistonen P, Pylkkanen, L, Mecklin JP, Jarvinen H, Powell SM, Jen J, Hamilton SR, Petersen GM, Kinzler KW, Vogelstein B, De la Chapelle A: Clues to the pathogenesis of familial colorectal cancer. Science 260:812–816, 1993 3. Boland CR, Thibodeau SN, Hamilton SR, Sidransky D, Eshleman JR, Burt RW, Meltzer SJ, Rodriguez-Bigas MA, Fodde R, Ranzani GN, Srivastava SA: National Cancer Institute Workshop on microsatellite instability for cancer detection and familial predisposition: Development of international criteria for the determination of microsatellite instability in colorectal cancer. Cancer Res 58:5248– 5257, 1998 4. Wang J, Sun L, Myeroff L, Wang X, Gentry LE, Yang J, Liang J, Zborowska E, Markowitz S, Willson JK: Demonstration that mutation of the type II transforming growth factor beta receptor inactivates its tumor suppressor activity in replication error-positive colon carcinoma cells. J. Biol. Chem. 270:22044–22049, 1995

RIPBERGER ET AL.

5. Shimizu Y, Ikeda S, Fujimori M, Kodama S, Nakahara M, Okajima M, Asahara T: Frequent alterations in the Wnt signaling pathway in colorectal cancer with microsatellite instability. Genes Chromosomes. Cancer, 33:73–81, 2002 6. Rampino N, Yamamoto H, Ionov Y, Li Y, Sawai H, Reed JC, Perucho M: Somatic frameshift mutations in the BAX gene in colon cancers of the microsatellite mutator phenotype. Science 275:967– 969, 1997 7. Lynch HT, Smyrk T: An update on Lynch syndrome. Curr. Opin. Oncol. 10:349–356, 1998 8. Smyrk TC, Watson P, Kaul K, Lynch HT: Tumor-infiltrating lymphocytes are a marker for microsatellite instability in colorectal carcinoma. Cancer 91:2417–2422, 2001 9. Dolcetti R, Viel A, Doglioni C, Russo A, Guidoboni M, Capozzi E, Vecchiato N, Macri E, Fornasarig M, Boiocchi M: High prevalence of activated intraepithelial cytotoxic T lymphocytes and increased neoplastic cell apoptosis in colorectal carcinomas with microsatellite instability. Am. J. Pathol. 154:1805–1813, 1999. 10. Townsend A., Ohlen C, Rogers M, Edwards J, Mukherjee S, Bastin J: Source of unique tumour antigens [Letter]. Nature 371:662, 1994 11. Linnebacher M, Gebert J, Rudy W, Woerner S, Yuan YP, Bork P, von Knebel Doeberitz M: Frameshift peptide-derived T-cell epitopes: A source of novel tumor-specific antigens. Int. J Cancer 93:6–11, 2001 12. Saeterdal I, Bjorheim J, Lislerud K, Gjertsen MK, Bukholm IK, Olsen OC, Nesland JM, Eriksen JA, Moller M, Lindblom A, Gaudernack G: Frameshift-mutation-derived peptides as tumorspecific antigens in inherited and spontaneous colorectal cancer. Proc. Natl. Acad. Sci. U.S.A 98:13255–13260, 2001 13. Saeterdal I, Gjertsen MK, Straten P, Eriksen JA, Gaudernack G: A TGF betaRII frameshift-mutation-derived CTL epitope recognised by HLA-A2-restricted CD8+ T cells. Cancer Immunol Immunother. 50:469–476, 2001 14. Kim NG, Rhee H, Li LS, Kim H, Lee JS, Kim JH, Kim NK, Kim H: Identification of MARCKS, FLJ11383 and TAF1B as putative novel target genes in colorectal carcinomas with microsatellite instability. Oncogene 21:5081–5087, 2002 15. Duval A, Reperant M, Compoint A, Seruca R, Ranzani GN, Iacopetta B, Hamelin R: Target gene mutation profile differs between gastrointestinal and endometrial tumors with mismatch repair deficiency. Cancer Res. 62:1609–1612, 2002 16. Woerner SM, Gebert J, Yuan YP, Sutter C, Ridder R, Bork P, von Knebel Doeberitz M: Systematic identification of genes with coding microsatellites mutated in DNA mismatch repair-deficient cancer cells. Int. J Cancer 93:12–19, 2001 17. Lubas WA, Frank DW, Krause M, Hanover JA: O-Linked GlcNAc transferase is a conserved nucleocytoplasmic protein containing tetratricopeptide repeats. J. Biol. Chem. 272:9316–9324, 1997 18. Wells L, Vosseller K, Hart GW: Glycosylation of nucleocytoplasmic proteins: Signal transduction and O-GlcNAc. Science 291:2376– 2378, 2001 19. Comer FI, Hart GW: O-Glycosylation of nuclear and cytosolic proteins. Dynamic interplay between O-GlcNAc and O-phosphate. J. Biol. Chem. 275:29179–29182, 2000 20. Schultze JL, Michalak S, Seamon MJ, Dranoff G, Jung K, Daley J, Delgado JC, Gribben JG, Nadler LM: CD40-activated human B cells: An alternative source of highly efficient antigen presenting cells to generate autologous antigen-specific T cells for adoptive immunotherapy. J. Clin. Invest 100:2757–2765, 1997 21. Brossart P, Bevan MJ: Selective activation of Fas/Fas ligandmediated cytotoxicity by a self peptide. J. Exp. Med. 183:2449– 2458, 1996



JOCI.cls

August 14, 2003

22:19


22. Houbiers JG, Nijman HW, van der Burg SH., Drijfhout JW, Kenemans P, van de Velde CJ, Brand A, Momburg F, Kast WM, Melief CJ: In vitro induction of human cytotoxic T lymphocyte responses against peptides of mutant and wild-type p53. Eur. J. Immunol. 23:2072–2077, 1993 23. Deuschle U, Pepperkok R, Wang FB, Giordano TJ, McAllister WT, Ansorge W, Bujard H: Regulated expression of foreign genes in mammalian cells under the control of coliphage T3 RNA polymerase and lac repressor. Proc. Natl. Acad. Sci. USA 86:5400–5404, 1989 24. Morel S, Levy F, Burlet-Schiltz O, Brasseur F, Probst-Kepper M, Peitrequin AL, Monsarrat B, Van Velthoven R, Cerottini JC, Boon T, Gairin JE, Van den Eynde BJ: Processing of some antigens by the standard proteasome but not by the immunoproteasome results in poor presentation by dendritic cells. Immunity. 12:107–117, 2000


423

25. Akiyama Y, Tsubouchi N, Yuasa Y: Frequent somatic mutations of hMSH3 with reference to microsatellite instability in hereditary nonpolyposis colorectal cancers. Biochem. Biophys. Res. Commun. 236:248–252, 1997 26. Bustin SA, Li SR, Phillips S, Dorudi S: Expression of HLA class II in colorectal cancer: Evidence for enhanced immunogenicity of microsatellite-instability-positive tumours. Tumour. Biol. 22:294– 298, 2001 27. Andersen SN, Rognum TO, Lund E, Meling GI, Hauge S: Strong HLA-DR expression in large bowel carcinomas is associated with good prognosis. Br. J. Cancer 68:80–85, 1993 28. Wright CM, Dent OF, Barker M, Newland RC, Chapuis PH, Bokey EL, Young JP, Leggett BA, Jass JR, Macdonald G: A. Prognostic significance of extensive microsatellite instability in sporadic clinicopathological stage C colorectal cancer. Br. J. Surg. 87:1197–1202, 2000