Somatically Mutated Regions of Immunoglobulin on Human B-Cell

[CANCER RESEARCH 61, 5145–5152, July 1, 2001] Somatically Mutated Regions of Immunoglobulin on Human B-Cell Lymphomas Code for Peptides That Bind to Autologous Major Histocompatibility Complex Class I, Providing a Potential Target for Cytotoxic T Cells1

Clair S. Gricks, Eira Rawlings, Letizia Foroni, J. Alejandro Madrigal, and Peter L. Amlot2 Departments of Immunology [C. S. G., E. R., P. L. A.] and Haematology [L. F., J. A. M.], Royal Free and University College Medical School, London NW3 2PF, United Kingdom

ABSTRACT (10). Alternatively, immunization with idiotypic proteins or DNA vaccines without adjuvant produced low levels of idiotypic antibody Lymphoma-derived immunoglobulin idiotype (Id) is a well-character- and poor protective immunity in mice and humans (10, 11). Addition ized, tumor-specific antigen on B-cell malignancies. Immunotherapy using lymphoma immunoglobulin can lead to clinical responses mostly associ- of an adjuvant, for example tetanus toxin fragment C, to DNA ated with anti-Id antibody. We cloned the Id from B-cell lymphomas, encoding the idiotype induced higher levels of idiotypic antibodies in sequenced them, and used bioinformatics to select autologous MHC class mice (12), and this approach is currently being tested in humans. I binding peptides from somatically mutated regions of the lymphoma Id. Although the vast majority of studies have used anti-idiotype antibody Peptides from patients who were HLA-A1, HLA-A2, HLA-A3, or HLA- as an indicator of immune response, more recent work has focused on A11 positive were analyzed in the T2 stabilization assay and a competitive the generation of specific cellular immunity. Immunization of follic- peptide-binding assay. By both methods, approximately half of the pep- ular lymphoma patients with idiotypic protein plus granulocyte/ tides analyzed, regardless of HLA type, bound with intermediate or high macrophage-colony stimulating factor, induced secretion of tumor affinity. Peptide binding affinity was similar to viral peptide sequences known to provide targets for cytotoxic T cells. Further investigation of necrosis factor from postvaccine PBMCs, shown to be mediated by ϩ ϩ lymphocyte responses to stimulation by autologous Id peptides versus Id CD4 and CD8 T cells (13). Specific lysis of autologous tumor cells peptides from other patients revealed that three of five patients in com- was also observed and was blocked by antibodies against MHC class plete remission or with low volume, stable disease responded to self- I, indicating the role of CD8ϩ cytotoxic T cells. Immunization of peptides by IFN-␥ secretion greater than that seen with non-self peptides, lymphoma patients with idiotype protein-pulsed autologous dendritic whereas none of five patients with progressive disease responded to their cells induced cellular proliferative responses rather than humoral own lymphoma Id. We have shown that mutated regions of lymphoma Id immunity (14). A large body of evidence favors cell-mediated rather contain MHC class I binding peptides that are potential targets for cytotoxic T cells. Immunotherapy using the tumor-specific mutated re- than antibody control of tumors, and this is supported by studies in gions from lymphoma Id avoids the need to break innate tolerance toward melanoma (15, 16) and virally induced tumors (17). the germ-line protein sequences present on normal and malignant B cells. The increased knowledge of epitopes within tumor-associated proteins recognized by CTL and bioinformatics enabling prediction of peptide binding to HLA class I have provided tools with which to INTRODUCTION search for the existence of CTLs and immunogenic peptides within Immunotherapy of human cancers requires suitably immunogenic, tumors from patients with malignancy (16, 18, 19). This approach has tumor-specific or -associated antigens. Surface immunoglobulin pro- also been used to monitor CTL responses in patients infected with vides a tumor-specific target in patients with B-cell malignancies viruses such as cytomegalovirus and EBV (20, 21). A recent study by because of its unique method of synthesis (1). The random insertion Trojan et al. (22) described HLA-A2-restricted peptide epitopes from of nucleotides between V(D)J gene segments during antigen-indepen- the tumor-derived immunoglobulin variable region from patients dent rearrangement of immunoglobulin genes followed by somatic mainly with CLL. Some of the peptides were shared among patients, hypermutation, occurring in germinal centers as a result of antigen indicating their unmutated, germ-line nature. The authors described stimulation, leads to immunoglobulin sequences not coded in the the in vitro generation of donor and some patient CTLs against the germ-line DNA (2, 3). The unique variable region of rearranged peptides and reported low-level killing of CLL tumor cells. immunoglobulin is known as its Id.3 Follicular lymphomas and my- The aims of our study were to define peptide sequences from elomas have a high rate of somatic hypermutation (4–7), which somatically mutated regions in lymphoma immunoglobulin Ids capa- increases the probability that their Ids will be immunogenic. ble of binding to the patient’s own HLA class I molecules and Immunization with idiotypic protein and adjuvant induces anti- hypothetically with a greater immunogenicity, because the mutated idiotypic antibodies in both mice (8) and humans in the absence of regions should not be regarded as self. The mutant cell line, T2, clinically detectable lymphoma (9), indicating the immunogenic po- provided a means of testing peptide binding to the HLA-A2 allele by tential of Ids. The production of anti-idiotypic antibodies has been the ability to up-regulate and stabilize cell surface expression of shown to mediate protection against tumor challenge in mouse models HLA-A2 peptide complexes (23). In addition, peptides of lymphoma Ids from patients with HLA-A3 were tested on a T2 cell line trans- Received 1/21/01; accepted 5/1/01. fected with HLA A3 (24). Selected peptides restricted by A1, A2, A3, The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with and A11 were also tested for binding in a competitive inhibition assay 18 U.S.C. Section 1734 solely to indicate this fact. using either LCLs derived from the patients themselves or LCLs 1 Supported in part by the Leukaemia Research Fund (to E. R.) and by the Marc Fisher homozygous for the HLA type of interest. Finally, stimulation of Trust (to C. S. G.). 2 To whom requests for reprints should be addressed, at Department of Immunology, patient lymphocytes by their own idiotypic peptides was examined in Royal Free and University College Medical School, Rowland Hill Street, London NW3 ELISPOT assays. The demonstration of frequent self-HLA binding 2PF. Phone: 44-20-7472-6266; Fax: 44-20-7431-3416; E-mail: [email protected]. 3 The abbreviations used are: Id, idiotype; PBMC, peripheral blood mononuclear cell; lymphoma idiotypic peptides and lymphocytes from lymphoma pa- CLL, chronic lymphocytic leukemia; LCL, lymphoblastoid cell line; ELISPOT, enzyme- tients in long-term remission capable of reacting to idiotypic peptides linked immunospot; RT-PCR, reverse transcription-PCR; MFI, mean fluorescence inten- ␥ sity; FR, framework region; CDR, complementarity-determining region; DLBL, diffuse by IFN- secretion suggests that mutated idiotypic peptides provide a large B-cell lymphoma; FCL, follicle center lymphoma. target for control of lymphomas by cell-mediated immune responses. 5145

MATERIALS AND METHODS A2, A3, or A11 (cell lines JY, HOM-1, VAVY, and FB) or the patient’s own B-LCL were washed twice in PBS and left to rest as a pellet on ice for 5 min. Patients. Lymphoma patients were classified according to the REAL clas- The cells were then acid stripped for 90 s using ice-cold Na2HPO4 buffer, pH sification (25). Lymphoid tissue was analyzed from 17 patients with diffuse 3.1 for A1, pH 3.2 for A2, and pH 2.9 for A3 and A11. The cells were buffered large cell lymphoma, 16 with follicle center lymphoma, 6 with B-cell small with ice-cold Iscove’s modified Dulbecco’s medium (ICN Flow, Basingstoke, lymphocytic lymphoma, 2 with mantle cell lymphoma, 2 with marginal zone Hampshire, United Kingdom), and then washed and resuspended at 1 ϫ 106/ml lymphoma, 1 with plasma cell myeloma, and 1 with lymphoplasmacytoid ␮ ␤ in Iscove’s modified Dulbecco’s medium plus 7.5 g/ml 2-microglobulin lymphoma. All patients were referred to our institution and treated according (Cymbus Biotechnology Ltd.). Twenty-five ␮l of fluorescent-labeled reference to United Kingdom standard lymphoma protocol. peptide [HLA-A1, YLEPACFITCAKY; HLA-A2, FLPSDCFITCFPSV; HLA- Genomic DNA Isolation. Diagnostic biopsies from 45 patients with B-cell A3/11, KVFPCFITCALINK (Department of Immunohematology and Blood malignant lymphomas were processed for DNA extraction using Proteinase K Bank, Leiden University Medical Center)] at a final concentration of 133 nM and a Puregene kit (Flowgen, Lichfield, Staffordshire, United Kingdom). were incubated with 25 ␮l of Id competitor peptide (MWG-Biotech) at final Patients were assessed for IgH clonality by Southern blotting and hybridization concentrations of 0.5–33.3 ␮M in a 96-well, U-bottomed plate with 100 ␮l to a JH probe, as described previously (26, 27). Patients with clonal JH pattern (1 ϫ 105) of cells for 24 h at 4°C (42). The percentage inhibition demonstrated were processed for amplification of the rearranged IgH genes as detailed by immunoglobulin peptides competing with the FITC-labeled reference pep- below. tide was calculated as described (42). PCR. PCR was performed using the manufacturer’s instruction (Promega, ELISPOT Assay for IFN-␥. PBMCs (1 ϫ 105) were plated in triplicate in Southampton, United Kingdom) in a 30-␮l volume containing 1 ␮g of genomic round-bottomed, 96-well plates and stimulated for 18 h with 50 ␮g/ml peptide

DNA and 100 pM of each primer. The forward primers were the family-specific at 37°C in 5% CO2. Culture medium was replaced following the 18-h stimu- VH1–6 leader primers (28, 29), and the reverse primer was a JH consensus lation, and cells were mixed, aliquoted onto the ELISPOT capture plate (30). Restriction sites were added to facilitate cloning. Conditions of amplifi- (Immunodiagnostic Systems Ltd., Bolden, Tyne and Wear, United Kingdom), cation were as follows: 1 cycle for 5 min at 94°C, followed by 30 cycles of 1 and incubated for an additional 20 h. IFN-␥ spots were detected according to min denaturation at 94°C, 1 min annealing at 66°C, and 2 min extension at the manufacturer’s instructions. 72°C, with a final cycle of 10 min extension at 72°C. The amplified PCR products were purified using a Hybaid recovery kit (Hybaid Ltd., Teddington, Middlesex, United Kingdom), cloned into Bluescript KSϩ vector (Promega) RESULTS and DNA, obtained using Qiagen miniprep kit (Qiagen, Crawley, West Sussex, Sequence Analysis. Southern blotting analysis showed the pres- United Kingdom), and sequenced using an ABI 377 Sequencer. ence of one rearranged band in 15 of 45 patients, indicating a mono- RT-PCR. RNA was prepared using the guanidine isothiocyanate method (31) and reverse transcribed as described previously (32), and IgH rearrange- clonal population with a monoallelic rearrangement. In 24 of 45 ments were amplified using antisense primers from IgM, IgG, and IgA con- patients, two rearranged bands of equal intensity were observed (re- stant regions (33) and VH leader primers (as described above). All immuno- ferred to as RR in Table 1), indicating a monoclonal population with globulin sequences were submitted to the IMGT/DNAPLOT website (34) for a biallelic rearrangement, which is commonly associated with an DNA and protein comparison to germ-line sequences. unproductive rearrangement on one allele. In 3 of 45 patients, more Peptides. VDJ protein sequences were searched for nonamer or decamer than two rearranged bands or two rearranged bands of unequal inten- peptides that conformed to the binding motif of the patients’ HLA type by sity (referred to as R and r in Table 1) were identified, which is most submitting the protein sequence to online databases, which predict binding likely explained by the presence of more than one clonal population, peptides and score them accordingly (19, 18). We were also influenced by referred to as oligoclonality, a frequent finding in B-cell acute leuke- published literature regarding anchor and secondary residues (35–38). The mias (43) but infrequent in lymphomas (44). In the remaining 3 majority of peptides selected covered an area of somatic hypermutation or patients, Southern blotting analysis was not carried out, but evidence spanned a VD or DJ junction. Peptides were synthesized to 95% purity (MWG-Biotech UK Ltd., Peartree Bridge, Milton Keynes, United Kingdom). of a clonal B-cell population was provided by immunophenotyping They were supplied lyophilized and then redissolved in DMSO and stored at (data not shown). Ϫ70°C at a concentration of 25 mM. Thirty-six clones were identified from 36 of 45 patients. Amplifi- T2 Assay. T2 cells were cultured in RPMI1640 ϩ 10% FCS and grown in cation by RT-PCR was more efficient (10 of 11 cases) than by

5% CO2 at 37°C. Peptides from HLA-A2-positive patients were assayed for DNA-PCR (26 of 45 cases). Sequence analysis revealed that 7 of 10 their ability to bind to and stabilize A2 molecules on T2 cells (39). Cells JH gene segments contained a high mutation rate affecting the JH (1 ϫ 105) in a volume of 100 ␮l of RPMI 1640 (serum free) were aliquoted primer annealing sequence, with the majority of mutations being into 96-well, U-bottomed plates and incubated with peptide at a final concen- clustered in the 3Ј end of the primer site, which is critical for efficient tration of 100 ␮M plus 5 nM ␤ microglobulin (Cymbus Biotechnology Ltd., 2 amplification. Patient immunoglobulin sequences were submitted to Chandlers Ford, Hampshire, United Kingdom) for 18 h at 37°C in 5% CO . 2 IMGT/DNAPLOT (34) for alignment to germ-line genes. All se- The level of stabilized HLA-A2 on the surface of the T2 cells was detected using the pan HLA class I monoclonal antibody W6/32 (European Collection quences were analyzed for productive rearrangements, conserved of Animal Cell Cultures (ECACC), Porton Down, Salisbury) that recognizes amino acid residues, and correct RNA splicing sites. stabilized HLA-A2 complexes (40). This was detected using a goat antimouse, With the exception of 3 cases, somatic mutations within the VH FITC-conjugated IgG (Cambridge Biosciences, Cambridge, United Kingdom) sequence resulted in between 1 and 37 amino acid replacements. The as the second layer. Samples were fixed in 1% paraformaldehyde in PBS prior highest replacement frequency was seen in DLBL and FCL (Table 1). to analysis and analyzed on a Becton Dickinson FACScan. The fluorescence Peptide Searches. Immunoglobulin sequences from those patients ratio ϭ MFI of the test peptide/MFI of DMSO control. who were HLA-A1, HLA-A2, HLA-A3, or HLA-A11 positive were The T2 cell line transfected with HLA-A3 (24) allowed analysis of peptides searched for nonameric and/or decameric peptides that carried the ϫ 5 ␮ from HLA-A3-positive patients. Cells (1 10 ) in a volume of 100 lof appropriate binding motif (18, 19, 36–38, 45–49). The peptides were ␮ RPMI 1640 (serum free) were incubated with 100 M peptide for 18 h at 26°C analyzed for anchor, preferred, and deleterious residues. Forty-two in 5% CO . Binding of peptides was detected as described above, except that 2 peptides restricted by HLA-A1, HLA-A2, HLA-A3, or HLA-A11 a monoclonal anti-HLA-A3 antibody, GAP-A3, was used (ECACC). EBV Transformation. PBMCs were isolated from patients’ blood using were selected, 39 of which either contained at least one amino acid Lymphoprep (Nycomed Pharma AS, Oslo, Norway). EBV transformed B change attributable to somatic hypermutation or covered a VD or DJ lymphoblastoid lines (B-LCL) were created as described previously (41). junction, making them tumor specific (Tables 2 and 3). The remaining Competition Assay. The competition assay was carried out as described by 3 peptides (MP1, WF1, and GW3) were germ line in nature. Tables 2 van der Burg et al. (42). In brief, LCLs that were homozygous for either A1, and 3 show that 67% of peptides with the potential to bind to A1, A2, 5146

Table 1 Details of patients, lymphoma, and molecular data The positive control for the HLA-A3 assay was also an influenza Southern VH gene Amino acid derived peptide, shown previously to bind well to A3 (45). Various Patient MLa blotb VH genec mutationsd changese conditions were tested when optimizing this assay. Although it was WF B-SLL RR 1–69 0 0 found to work under the conditions used for T2-A2, in contrast to AS B-SLL RG 3–30.5 18 14 T2-A2 the MFI ratios increased considerably when incubated at 26°C JT B-SLL RRG 3–30 28 20 SS B-SLL RRG 3–33 12 10 BJ B-SLL RRG 3–66 3 1 GW B-SLL RRG 1–69 0 0 Table 2 Analysis of potential A2 binding peptides DC DLBL RRG 4–30.2 68 37 Rammensee Parker D’Amaro DD DLBL RR 3–30 23 12 Peptidea Sequenceb Positionc scored scoree scoref Bindingg FAN DLBL RG 3–23 40 19 GK DLBL RRR 5–51 47 31 A2 Flu GILGFVFTL 30 551 54 ϩ GS DLBL RRG 3–23 41 19 EBV1 GLCTLVAML 28 49 61 ϩ MEN DLBL RG 1–2 29 17 JT1 TMYLQMNDL FR3 20 114 55 Ϫ MP DLBL RRrr 4–59 19 13 JT2 VGFRFTLST FR1/CDR1 12 1 0 Ϫ JC DLBL RG 4–59 11 7 JT3 DLTPEDTAL FR3 21 2 51 Ϫ EZ DLBL RRG 4–59 71 34 JT4 GRGTLVTVS FR4 6 0 1 Ϫ MH DLBL RG 4–59 32 19 RH1 GEDTAVYYC FR3 6 0.5 26 Ϫ JPF DLBL RR 3–23 24 11 RH2 KGLEWLSNI FR2/CDR2 17 8 33 Ϫ JB DLBL RG 3–48 3 3 RH3 ILYLQMNSL FR3 26 267 57 Ϫ/ϩ JAL DLBL ND 3–7 24 12 JD1 GDGLSVWGQ CDR3/FR4 6 0 2 Ϫ MA DLBL ND 5–a 60 34 JD2 VWGQGTLVT FR4 10 0 19 Ϫ DT FCL RRG 3–23 59 23 JD3 SVWGQGTLV FR4 19 51 54 ϩ JD FCL RG 3–7 20 8 SJ1 YLYLQMNSL FR3 26 723 62 ϩ JR FCL RRG 3–7 23 14 SJ2 QLMESGGGL FR1 24 42 62 Ϫ MM FCL RRG 4–34 64 31 MP1 QLQESGPGL FR1 24 88 67 Ϫ RH FCL RRG 3–11 47 20 DT1 YVMSWVRQV CDR1/FR2 21 48 54 ϩ SJ FCL RrG 3–9 37 16 DT2 RLMESGGGL FR1 24 42 62 Ϫ PR FCL RR 1–2 6 4 DT3 TMVLQMNRL FR3 19 35 51 Ϫ SP FCL RRG 3–11 27 12 DT4 ALPGGSLRL FR1 27 21 58 Ϫ JF FCL RG 3–23 40 21 FA1 TLFLQMNNL FR3 22 182 57 Ϫ TL FCL RG 3–53 55 26 FA2 FLQMNNLRA FR3 16 23 50 Ϫ/ϩ MB LPL R 1–69 1 1 FA3 NLRAEDTAI FR3 19 0.8 53 Ϫ HB MCL RRG 1–8 0 0 GS1 KLRDGDTAI FR3 23 11 60 Ϫ/ϩ RB MCL RR 1–8 2 1 SS1 TLFLQMNSL FR3 24 181 60 ϩ JL MZBL RR 3–7 30 14 WF1 QLVQSGAEV FR1 24 70 61 Ϫ/ϩ DB MZBL RRG 3–30.3 6 5 ME1 FTDYYIHWV CDR1/FR2 20 22 52 ϩ UD PCM RRG 3–30.5 7 5 ME2 HLVDSGADV FR1 23 10 67 ϩ SLQMNSLRV ϩ a Malignant lymphoma by the REAL classification (25): B-SLL, B-cell small lympho- SP1 FR3 23 70 58 NLIYYADSV ϩ cytic lymphoma; LPL, lymphoplasmacytoid lymphoma/immunocytoma; MCL, mantle SP2 CDR2/FR3 24 34 59 SVGStYMTWV ϩ cell lymphoma; MZBL, marginal zone B-cell lymphoma; PCM, plasma cell myeloma. TL1 CDR1/FR2 16 77 54 RLSCVVSGL ϩ b Southern blot: R, dominant rearrangement; G, germ-line; r, minor rearrangement; TL2 FR1/CDR1 25 49 60 ND, not done. a Patient’s initials and number of the peptide: A2 Flu (38), EBV 1 (21). c Germ-line VH gene used in the rearrangement (34). b Anchor residues are in bold-face, and altered amino acids attributable to somatic d Number of mutations in the VH gene when compared with germ-line by IMGT/ hypermutation are underlined. DNAPLOT (34). c Position of peptide within variable region. e Number of amino acid changes attributable to VH gene mutations. d See Rammensee et al. (18). e See Parker et al.. (19). f See D’Amaro et al. (36). g ϩ, positive; Ϫ, negative. Peptides shown to have Ϫ/ϩ binding bound in the T2 assay A3, and A11 were located in the FR, and only 33% were in overlap- but not in the competition assay. ping FR and CDR regions. T2 Assay. Peptides with the binding motif for HLA-A2 and Table 3 Analysis of potential A1, A3, and A11 binding peptides HLA-A3 were tested in the T2 stabilization assay. All peptides were assayed in triplicate on at least three separate occasions. Figs. 1 and 2 Peptide Rammensee Parker Peptidea sequenceb HLA type Positionc scored scoree Bindingf show the shift in mean fluorescence ratios of peptide binding to A3 Flu ILRGSVAHK A3 30 90 ϩ T2-A2 or T2-A3 cells. The influenza matrix peptide is known to bind AS2 KIYYAESVK A3 CDRs/FR3 32 30 ϩ with high affinity to HLA-A2 (38) and gave a MFI ratio of 2.31 in our AS3 ALTPYDGGK A3 FR2/CDR2 27 27 ϩ experiments. Peptides with a MFI ratio Ͼ1.5 defines high affinity MA1 SLRISCEGY A3 FR1 21 4 Ϫ MA2 ITWVrQVPGK A3 FR2 12 15 ϩ binding to HLA-A2. Altogether, 45% of the peptides selected for EBV3 IVTDFSVIK A11 2 ϩ HLA-A2 binding motifs demonstrated binding in the T2 assay, of TL4 WVSIiYGDGK A11 FR2/CDR2 2 ϩ TL5 FYLQMNNLK A11 FR3 0.6 Ϫ/ϩ which 24% (7 of 29) were high affinity and 21% (6 of 29) bound with JR1 AVYFCAAGK A11 FR3/CDR3 4 ϩ ␤ intermediate affinity (MFI ratio of 1.25–1.5). Addition of 5 nM 2- JD5 RFNIsRDNAK A11 FR3 0.6 Ϫ microglobulin to the cells in serum-free medium when incubated with A1 Flu VSDGGPNLY A1 35 37.5 ϩ JF3 LRDDdTALYY A1 FR3 31 7.5 Ϫ peptide increased the sensitivity and reproducibility of the T2-A2 MM1 VSPNGDTHY A1 CDR2/FR3 22 0 Ϫ/ϩ assay. We also found that because of the hydrophobic motif preferred MM2 ETLSPTCTY A1 FR1 22 1.25 ϩ by HLA-A2, many of the peptides were insoluble in water and had to GW2 HLVPaAMLNY A1 CDR3/FR4 26 25 Ϫ GW3 RSEDTAVYY A1 FR3 29 135 ϩ be dissolved in DMSO. We observed precipitation of peptides when a Patient’s initials and number of the peptide: A3 Flu (45), A1 Flu (45), A11 EBV (21). incubated with cells in RPMI 1640, which may be a limitation in this b Anchor residues are in boldface and altered amino acids attributable to somatic assay because insolubility may impair binding. hypermutation are underlined. c Bioinformatic database scores for peptides that were proven exper- Position of peptide within variable region. d See Rammensee et al. (18). Predictions for HLA-A11 are not available in this imentally to bind in the T2-A2 assay had greater median scores than database. non-binding peptides (14.5 versus 49, Parker score, P ϭ 0.046; 52 e See Parker et al. (19). f ϭ ϩ, positive; Ϫ, negative. Peptides MM1 and TL5 showed no binding on unrelated versus 59, D’Amaro score, P 0.039; 19 versus 23, Rammensee LCLs but low affinity binding to patients’ own cell lines; therefore, binding is shown as score, P ϭ 0.027 using the Mann-Whitney U test). Ϫ/ϩ. 5147

Fig. 3. Competition assay for binding of Id peptides. Id peptides were analysed for binding to unrelated LCLs homozygous for the HLA type of interest. The percentage inhibition of the FL-labeled peptide by Id peptides was calculated as described (42). A ;SJ1; ⅜, SP2 ,ء ;Fig. 1. Binding of Id peptides to HLA-A2 on T2 cells. The binding of Id peptides to representative example of binding peptides is shown (ƒ, MA2; Œ, ME1 the T2-A2 cell line was demonstrated by an increase in MFI. The results are shown as the f, A2 Flu). fold increase in MFI above the DMSO control, with dotted lines to denote intermediate- affinity (1.25) and high-affinity (1.5) thresholds.

Fig. 4. IC50 data for HLA-A2 and HLA-A3 Id peptides binding to unrelated LCLs. The IC (inhibitory concentration at which 50% of the labeled peptide was inhibited) of Fig. 2. Binding of Id peptides to HLA-A3 on T2 cells. The fold increase in MFI above 50 HLA-A2 (f) and A3 (Ⅺ) Id peptides was calculated from the percentage inhibition of the the DMSO control for Id peptides is shown, with dotted lines to denote intermediate- labeled peptide demonstrated in the competition assay. affinity (1.5) and high-affinity (5.0) thresholds.

␤ in serum-free medium without the addition of 2-microglobulin. Seventy-five % (3 of 4) of immunoglobulin peptides were found to increase the MFI ratio, and two of them bound with equivalent affinity to the positive influenza control. Competition Assay. Fig. 3 provides a representative example of inhibition of the FluFITC peptide by Id peptides in the competition

assay. The IC50s (concentration of Id peptide at which the FL-labeled peptide is inhibited by 50%) for A2 and A3 peptides were calculated and are shown in Fig. 4. Sixteen peptides that bound in the T2 assay (A2 and A3) were tested in the competition assay. Eight of 16 (50%) Ͻ ␮ had IC50 5 M, comparable with known CTL epitopes; 1 of 16 (6%) ␮ had IC50s between 5 and 15 M and would be considered a candidate Ͼ ␮ CTL epitope (42); 3 of 16 (19%) competed with IC50 15 M, and 4 of 16 (25%; RH3, FA2, GS1, and WF1) did not compete at all. We

also selected A1- and A11-restricted Id peptides based on the same Fig. 5. IC50 data for binding of Id peptides to unrelated or patients’ LCLs. Id peptides selection procedure as for A2 and A3 peptides and tested them in the were tested in the competition assay for binding to an unrelated (Ⅺ) cell line expressing f competition assay (data not shown). Similar results were gained with the HLA type of interest or the patients’ own cell line ( ). Peptides MM1 and TL5 did not show binding to the unrelated LCLs but showed low-affinity binding (IC50, 20.01 and 2 of 5 (40%) A1 peptides and 2 of 4 (50%) A11 peptides competing 27.35, respectively) to the patient’s own cell line (data not shown above). 5148

Table 4 IFN-␥ ELISPOT and clinical details of patients tested Clinical details ELISPOT response to self-Id peptides

Time in years sinced

Patienta Disease statusb Histology and stagec Diagnosis Treatment Last treatmente Peptidef HLAf IFN-␥g JD CR FCL 3 2 CHOP JD1 A2 0 JD2 A2 5 JD3 A2 0 SJ CR FCL 12 9 Chlor SJ1 A2 6 SP CR FCL 10 9 Chlor SP1 A2 0 SP2 A2 0 MM SD FCL, IIIA 3 0.5 CD20 MM1 A1 6 MM2 A1 1 JR SD FCL, IA 18 2 Chlor JR1 A11 0 JT PD B-SLL, IIIA 12 2 Cyclo JT1 A2 0 JT2 A2 1 JT3 A2 0 JT4 A2 0 TL PD FCL, IA 11 2 BMT TL1 A2 1 TL2 A2 2 DT PD FCL, IIIA 2 1 Chlor DT1 A2 1 DT2 A2 2 DT3 A2 3 DT4 A2 0 GS PD DLBL, IIA 3 1 CHOP GS1 A2 0 GW PD B-SLL, IVA 4 0.5 CD20 GW1 A1 0 GW2 A1 0 a PBMCs from 10 lymphoma patients were stimulated with their self-Id peptides. b CR, complete remission (disease stage, 0); SD, stable disease; PD, progressive disease. c Lymphoma histology and stage of disease at the time of the assay. d The time in years, at the time of the assay, since diagnosis and last treatment. e Details of last treatment (64). Chlor, chlorambucil; CD20, Rituximab; Cyclo, cyclophosphamide; BMT, bone marrow transplant. f Patients’ self-Id peptide is shown with the HLA restriction. g The number of IFN-␥-producing cells/105 PBMCs is shown in response to stimulation with self-Id peptides.

Ͻ ␮ Ͼ ␥ with IC50 5 M. The IC50s of peptides with high ( 1.5) T2 binding tween 37 and 82% of the peptides. Numbers of IFN- -producing cells affinities were significantly lower than those with low (Ͻ1.5) binding ranged from 5 to 65 per 105 PBMCs, with some of the Id peptides affinity (P ϭ 0.02 by Mann-Whitney U test), indicating agreement producing higher responses than those seen with viral peptides. How- between these two assays. Because of the highly polymorphic nature ever, the HLA-A2-negative donors did not respond to the A2 peptides, of the HLA genes, the patient’s own HLA proteins may have slight supporting the HLA-restricted nature of the responses. differences to those HLA proteins presented by the unrelated homozygous LCLs (JY, HOM-1, VAVY, and FB). To investigate this, we DISCUSSION created EBV-transformed B-cell lines from patients’ peripheral blood and used these in the competition assay. Nine peptides (HLA-A1, PCR amplification of the clonal Id was found to have a high failure HLA-A2, HLA-A3, and HLA-A11) were tested on both unrelated and rate when using DNA as the template, also reported in the literature patients’ own cell lines. IC50 data were similar for both LCLs and are (50). A high percentage of DLBL and FCL samples (41 and 63%, shown in Fig. 5. Peptides MM1 and TL5 did not show binding to the respectively) could not be amplified from genomic DNA compared ϭ unrelated LCLs but showed low affinity binding (IC50 20.01 and 27.35, respectively) to the patient’s own cell line (not shown in Fig. 5). Table 5 IFN-␥ ELISPOT responses to viral and non-self-Id peptides ␥ ELISPOT Assay for IFN- . Tables 4 and 5 show the number of ELISPOT response to ELISPOT response IFN-␥-producing cells/105 PBMCs in response to different peptides. non-self Id peptidesc to viral peptidesd

PBMCs were stimulated with peptides derived from the patient’s own No. of positive Range of lymphoma Id, HLA-matched Id peptides derived from other patients’ Patienta HLAb peptides/total tested response EBV Flu lymphomas, and influenza or EBV peptides. PBMCs from 3 patients JD A2 1/9 2 0 0 (JD, SJ, and MM) responded to stimulation with their own lymphoma SJ A2 0/1 8 0 SP A2 2/8 19–25 64 16 Id peptides (JD2, SJ1, and MM1) above any response toward non-self MM A1 1/3 5 14 Id peptides. Both JD and SJ have been in complete remission from JR A11 0/3 0 follicle center lymphoma for 3 and 10 years, respectively. Patient MM JT A2 3/14 1–5 4 8 TL A2 5/7 1–12 40 0 currently has stable follicle center lymphoma with a few small neck DT A2 7/7 2–9 10 3 nodes (Յ1 cm). PBMCs from the other 7 patients did not respond to GS A2 0/1 4 0 self-Id above responses made to non-self-Id peptides, and this in- GW A1 0/3 1 cluded 1 in complete remission of 10 years duration (SP), 1 with a PBMCs from the 10 lymphoma patients described in Table 4 were also stimulated with viral and non-self-Id peptides concurrently (same PBMC sample). stable disease (3-cm neck node), and 5 with progressive disease at the b HLA restriction of peptides used to stimulate patients’ PBMCs. time of testing. Clinical details of the patients tested in this assay are c A variety of non-self, HLA-matched Id peptides were used to stimulate 105 PBMCs shown in Table 4. from lymphoma patients. The number of peptides that induced a response above the background DMSO is shown divided by the total number of peptides tested. The range of The ELISPOT was also carried out on PBMCs derived from HLA- IFN-␥-producing cells above background DMSO in response to non-self peptides is A2-positive and HLA-A2-negative healthy donors using the HLA-A2 shown. d The number of IFN-␥-producing cells/105 PBMCs in response to viral peptides Id and viral peptides (data not shown). The results showed that derived from EBV and influenza type A are shown above the response gained with DMSO PBMCs from all 5 HLA-A2-positive donors made responses to be- alone. 5149

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 2001 American Association for Cancer Research. LYMPHOMA IDIOTYPIC PEPTIDE BINDING TO MHC CLASS I with the failure rate of only 25% for all other lymphomas taken of somatic hypermutation or covered a junctional region (Tables 2 and together. Sequence analysis revealed that the DLBL and FCL VH gene 3) and were therefore tumor specific. We were interested to establish segments contained high load somatic hypermutation, in agreement whether the numbers of binding peptides differ as a result of mutation. with others (7, 51). The hypermutation seen in follicular lymphomas The germ-line counterparts of the variable genes used in the immu- is consistent with their origin in the germinal center, the natural site at noglobulins studied were scanned for potential binding peptides using which somatic hypermutation takes place. DLBLs are also assumed the scoring systems and databases described earlier. Similar numbers to arise from the germinal center of the lymph node. Analysis of of HLA-binding motifs were found in both mutated and germ-line sequences amplified by RT-PCR using 3Ј constant region primers immunoglobulin variable segments. In some cases, peptide binding confirmed that the somatic hypermutation extended into the JH gene motifs were lost because of a mutation that introduced an amino acid segment and would have caused primer misannealing, accounting for into a position affecting binding adversely, and in other cases existing the high failure rate of DNA-PCR on DLBL and FCL samples. peptides had their scores increased when a mutation inserted a favor- Peptides were chosen on the basis of scores predicted by two online able amino acid. databases and published literature. Peptides without anchor residues at Finally, three patients had blood lymphocytes that made low mag- positions 2 and 9 had low scores and did not bind in the T2 assay. The nitude responses to their lymphoma Id peptides by secretion of IFN-␥, use of different peptide scoring algorithms, which are based on natural suggesting previous sensitization to these determinants. Patients SJ, ligands (18, 19, 36) or the dissociation kinetics of HLA complexes JD, and MM made responses to peptides SJ1, JD2, and MM1, respec- (19), provides independent evaluation of peptide binding. The data- tively. Peptides SJ1 and MM1 showed binding in the peptide binding base predictions for any one peptide do not always agree, and even assays; however, peptide JD2 showed no binding on either unrelated when they do correlate, they do not guarantee binding for in vitro or autologous LCLs. HLA class I molecules are known to present assays. Prediction of peptide binding is complex because it depends peptides of fewer than 9 amino acids (56, 57); therefore this peptide on both the presence of anchor residues and the influence of second- may have been further processed before presentation. Two of three of ary residues at other sites (52), and it is unlikely that peptide binding these patients were in complete remission from their disease, and the databases are able to predict all potential epitopes. However, detection third had detectable but low-volume stable disease. This is consistent of all binding peptides from the Id sequences would require analysis with murine models of immunotherapy where immunization was of a panel of overlapping peptides, which would be very labor successful in mice free of lymphoma or with minimal disease but not intensive. Peptide prediction databases reduce the time taken to ana- where the lymphoma was of large volume or progressing (58). All lyze protein sequences for the presence of potential epitopes, and the patients in this study with progressive lymphoma made no or very low accuracy of prediction will increase as our knowledge of binding responses to their lymphoma Ids, whereas they were often capable of peptides expands. The results from this study have shown that high reacting to viral or unrelated lymphoma Id peptides, indicating their binding scores with databases using different peptide binding algo- immunocompetence in this respect. Patients’ PBMCs were also stim- rithms increased the accuracy of prediction. Using only a single MHC ulated with a range of non-self peptides, some of which showed class I allele, HLA-A2, about half of the selected peptides were binding in the peptide assays and some of which did not. Six patients capable of binding, and a quarter of them had binding affinities in the made responses to non-self-Id peptides, including 12 of 19 responses range known to make good viral CTL epitopes (42). The same is true to peptides that did not bind in the peptide binding assays. The reasons for other class I alleles, A1, A3, or A11, suggesting that an effective for this are unclear but may be attributable to binding to other HLA cytotoxic target is achievable from almost any lymphoma Id sequence. class I molecules expressed by the patients’ cells or further processing The range of MFI ratios obtained using T2-A3 cells were of a much of the peptides. Reactivity with non-self, HLA-matched Id peptides higher magnitude than those gained using T2-A2 cells. HLA-A2 may indicate cross-reactive T-cell responses to similar epitopes en- molecules on the surface of T2 cells are reported to be at least partially countered previously by the immune system, as illustrated by the occupied with signal peptides derived from normal cellular proteins, recognition of a myeloma-specific peptide by CTLs specific for an such as IP-30 and calreticulin (53–55). The resulting high background influenza epitope (59). These responses are similar to the generation fluorescence accounts for the differences in MFI ratios. of CTLs from HLA-matched donors against lymphoma immunoglob- There was a statistical correlation between peptide binding in the ulin peptides, described recently (22). It could be argued that the

T2 assay and IC50 data from the competition assay, although there autologous responses made to self-Id peptides are from cross-reacting were some discrepancies in the results. For example, ME2 showed a T cells, but if these T cells could be amplified by immunotherapy, higher IC50 in the competition assay but proved to be a high affinity they may provide useful responses in vivo. The absence of an immune binder in the T2 assay. This may be because of minor differences in response to self-Id peptides in those patients with progressive disease the HLA-A2 molecules presented by the LCL and T2 cell lines. The is likely to indicate the induction of anergy in responding T cells, as T2 assay relies on stabilization and up-regulation of HLA on the cell shown in metastatic melanoma (16). surface, and although this does not yield quantitative data, it provides In a recent study by Trojan et al. (22), HLA-A2-restricted, unmu- an indication of the ability of a peptide to bind. The competition assay tated germ-line peptides from the immunoglobulin V regions of allows the semiquantitative determination of peptide binding, the patients with B-cell malignancies could bind in the T2 assay, also kinetics of which are comparable with that of soluble HLA class I demonstrated by peptide WF1 in this study. They showed that a small molecules (42). Another advantage of the competition assay is that the proportion of these peptides could stimulate the expansion of CTLs patients’ own EBV-transformed B cells can be used instead of a cell from healthy donors and a few patients after repeated rounds of in line. The HLA locus is highly polymorphic, and although peptides vitro stimulation. Many of the peptides were shared among patients, from a patient may bind to the HLA on a cell line, single amino acid indicating their germ-line nature. Unmutated peptide epitopes of this differences in their own HLA molecules may prevent binding. We nature will be present not only in the lymphoma immunoglobulin but demonstrated similar binding characteristics for all of the peptides also within the V regions of normal B cells, making it a tumor- tested with HLA class I homozygous lines and the patients own associated antigen rather than tumor-specific antigen. Immunothera- EBV-transformed cell line (Fig. 5). peutic targeting of these epitopes has the potential to destroy normal Virtually all (39 of 42) lymphoma Id peptides analyzed by in vitro B cells but to do so would need to break tolerance to self-antigens in HLA-binding assays in this study carried amino acid changes because vivo. Id peptides covering mutated regions are tumor specific and 5150

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 2001 American Association for Cancer Research. LYMPHOMA IDIOTYPIC PEPTIDE BINDING TO MHC CLASS I more likely to be immunogenic because there is no need to break to fragment C of tetanus toxin induce protective immunity against lymphoma and self-tolerance. Support for this position was demonstrated recently myeloma. Nat. Med., 4: 1281–1286, 1998. 13. Bendandi, M., Gocke, C. D., Kobrin, C. B., Benko, F. A., Sternas, L. A., Pennington, where unmutated immunoglobulin in CLL had a more aggressive R., Watson, T. M., Reynolds, C. W., Gause, B. L., Duffey, P. L., Jaffe, E. S., course than those in which the immunoglobulin had undergone so- Creekmore, S. P., Longo, D. L., and Kwak, L. W. Complete molecular remissions induced by patient-specific vaccination plus granulocyte-monocyte colony-stimulat- matic mutation (60). ing factor against lymphoma. Nat. Med., 5: 1171–1177, 1999. On the basis of murine studies, lymphoma immunotherapy for over 14. Hsu, F. J., Benike, C., Fagnoni, F., Liles, T. M., Czerwinski, D., Taidi, B., Engleman, a decade has concentrated on antibody-mediated immunity (9, 61), E. G., and Levy, R. Vaccination of patients with B-cell lymphoma using autologous antigen-pulsed dendritic cells. Nat. Med., 2: 52–58, 1996. and attempts at peptide therapy were ineffective in mice (62). How- 15. Yee, C., Savage, P. A., Lee, P. P., Davies, M. M., and Greenberg, P. D. Isolation of ever, the experiments performed at that time were relatively crude, high affinity melanoma-reactive CTL from heterogeneous populations using peptide- and the peptide binding motifs for murine MHC were unknown. MHC tetramers. J. Immunol., 162: 2227–2234, 1999. 16. Lee, P. P., Yee, C., Savage, P. A., Fong, L., Brockstedt, D., Weber, J. S., Johnson, D., Recently, more promising immunotherapy of lymphoma and other Swetter, S., Thompson, J., Greenberg, P. D., Roederer, M., and Davis, M. M. malignancies has been achieved using dendritic cells to immunize Characterization of circulating T cells specific for tumour-associated antigens in patients (14). The efficacy of dendritic cells to prime CTLs may be the melanoma patients. Nat. Med., 5: 677–685, 1999. 17. Chapman, A. L., and Rickinson, A. B. Epstein-Barr virus in Hodgkin’s disease. Ann. means by which immunotherapy has improved in this disease. The Oncol., 9 (Suppl. 5): S5–S16, 1998. identification of CTL epitopes in melanoma and graft versus host 18. Rammensee, H. G., Bachmann, J., Emmerich, N. N., Bachor, O. A., and Stevanovic, disease has enabled MHC:peptide tetramer staining of specific CTL S. SYFPEITHI: database for MHC ligands and peptide motifs. Immunogenetics, 50: 213–219, 1999. populations during disease progression (16, 63), providing invaluable 19. Parker, K. C., Bednarek, M. A., and Coligan, J. E. Scheme for ranking potential data on the immune response. Our current research is involved with HLA-A2 binding peptides based on independent binding of individual peptide side- synthesizing lymphoma Id peptide tetramers to determine whether we chains. J. Immunol., 152: 163–175, 1994. 20. Kern, F., Surel, I., Brock, C., Freistedt, B., Radtke, H., Sceffold, A., Blasczyk, R., can detect specific CTLs in peripheral blood or lymph node biopsies Reinke, P., Schneider-Mergener, J., Radbruch, A., Walden, P., and Volk, H. D. T-cell of these patients to substantiate the ELISPOT data. epitope mapping by flow cytometry. Nat. Med., 4: 975–978, 1998. We have shown that the somatically mutated regions of lymphoma 21. Tan, L. C., Gudgeon, N., Annels, N. E., Hansasuata, P., O’Callaghan, C. A., Rowland-Jones, S., McMichael, A. J., Rickinson, A. B., and Callan, M. F. C. A Id from patients with non-Hodgkin’s lymphoma have the potential to re-evaluation of the frequency of CD8ϩ T cells for EBV in healthy virus carriers. be immunogenic, based on the presence of peptides with the ability to J. Immunol., 162: 1827–1835, 1999. 22. Trojan, A., Schultze, J. L., Witzens, M., Vonderheide, R. H., Ladetto, M., Donovan, bind to the patients’ own HLA type and that patients in complete J. W., and Gribben, J. G. Immunoglobulin framework-derived peptides function as remission from their disease have memory T cells capable of being cytotoxic T-cell epitopes commonly expressed in B-cell malignancies. Nat. Med., 6: stimulated by their lymphoma Id peptides. These are important steps 667–672, 2000. 23. Spies, T., and Demars, R. Restored expression of major histocompatibility class I in defining immunotherapeutic targets for lymphoma. molecules by gene transfer of a putative peptide transporter. Nature (Lond.), 351: 323–324, 1991. 24. Smith, K. D., and Lutz, C. T. Peptide dependent expression of HLA-B27 on antigen ACKNOWLEDGMENTS processing deficient T2 cells. J. Immunol., 156: 3755–3764, 1996. 25. Harris, N. L., Jaffe, E. S., Stein, H., Banks, P. M., Chan, J. K. C., Cleary, M. L., We thank Dr. Tony Dodi for advice regarding peptide binding assays and Delsol, G., De Wolf-Peeters, C., Falini, B., Gatter, K. C., Grogan, T. M., Piris, M. A., Ralfkiaer, E., and Warnke, R. A. A revised European-American classification of the homozygous B-LCLs; Dr. John O’Shea for HLA typing lymphoma patients lymphoid neoplasms: a proposal from the international lymphoma study group. (both at the Anthony Nolan Bone Marrow Institute); and Dr. Catherine Blood, 84: 1361–1392, 1994. McIntyre at the Cancer Medicine Research Unit, St James University Hospital, 26. Southern, E. M. Detection of specific sequences among DNA fragments separated by for the kind gift of the T2-A3 cells. gel electrophoresis. J. Mol. Biol., 98: 503–517, 1975. 27. Foroni, L., Catovsky, D., Rabbitts, T. H., and Luzzatto, L. DNA rearrangements of immunoglobulin genes correlate with phenotypic markers in B-cell malignancies. Mol. Biol. Med., 2: 63–79, 1984. REFERENCES 28. Thompsett, A. R., Ellison, D. W., Stevenson, F. K., and Zhu, D. V(H) gene sequences from primary central nervous system lymphomas indicate derivation from highly 1. Tonegawa, S. Somatic generation of antibody diversity. Nature (Lond.), 302: 575– mutated germinal center B cells with ongoing mutational activity. Blood, 94: 1738– 581, 1983. 1746, 1999. 2. Wilson, P. C., de Bouteiller, O., Liu, Y. J., Potter, K., Banchereau, J., Capra, J. D., and 29. Sahota, S. S., Leo, R., Hamblin, T. J., and Stevenson, F. K. Ig VH mutational patterns Pascual, V. Somatic hypermutation introduces insertions and deletions into immuno- indicate different tumour cell status in human myeloma and monoclonal gammopathy globulin V genes. J. Exp. Med., 187: 59–70, 1998. 3. Wiesendanger, M., Scharff, M. D., and Endelmann, W. Somatic hypermutation, of undetermined significance. Blood, 87: 746–755, 1996. transcription, and DNA mismatch repair. Cell, 94: 415–418, 1998. 30. Campbell, M. J., Zelenetz, A. D., Levy, S., and Levy, R. Use of family specific leader 4. Zhu, D., Hawkins, R. E., Hamblin, T. J., and Stevenson, F. K. Clonal history of a region primers for PCR amplification of the human heavy chain variable region gene human follicular lymphoma as revealed in the immunoglobulin variable region genes. repertoire. Mol. Immunol., 29: 193–203, 1992. Br. J. Haematol., 86: 505–512, 1994. 31. Chomczynski, P., and Sacchi, N. Single step method of RNA isolation by acid 5. Bakkus, M. H. C., Heirman, C., Van Reit, I., Van Camp, B., and Thielemans, K. guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem., 162: 156– Evidence that multiple myeloma Ig heavy chain VDJ genes contain somatic mutations 159, 1987. but show no intraclonal variation. Blood, 80: 2326–2335, 1992. 32. Devaraj, P. E., Foroni, L., Kitra-Roussos, V., and Secker-Walker, L. M. Detection of 6. Kosmas, C., Stamatopoulos, K., Stavroyianni, N., Belessi, C., Viniou, N., and BCR-ABL and E2A-PBX1 fusion genes by RT-PCR in acute lymphoblastic leukaemia Yataganas, X. Molecular analysis of immunoglobulin genes in multiple myeloma. with failed or normal cytogenetics. Br. J. Haematol., 89: 349–355, 1995. Leuk. Lymphoma, 33: 253–265, 1999. 33. Doenecke, A., Winnacker, E. L., and Hallek, M. Rapid amplification of cDNA ends 7. Ottensmeier, C. H., Thompsett, A. R., Zhu, D., Wilkins, B. S., Sweetenham, J. W., (RACE) improves the PCR-based isolation of immunoglobulin variable genes from and Stevenson, F. K. Analysis of VH genes in follicular and diffuse lymphoma shows murine and human lymphoma cells and cell lines. Leukemia (Baltimore), 11: 1787– ongoing somatic mutation and multiple isotype transcripts in early disease with 1792, 1997. changes during disease progression. Blood, 91: 4292–4299, 1998. 34. Lefranc, M. P., Giudicelli, V., Ginestoux, C., Bodmer, J., Muller, W., Bontrop, R., 8. George, A. J. T., and Stevenson, F. K. Prospect for the treatment of B cell tumours Lemaitre, M., Malik, A., Barbie, V., and Chaume, D. IMGT, the international using idiotypic vaccination. Int. Rev. Immunol., 4: 271–310, 1989. ImMunoGeneTics database. Nucleic Acids Res., 27: 209–212, 1999. 9. Hsu, F. J., Caspar, C. B., Czerwinski, D., Kwak, L. W., Liles, T. M., Syrengelas, A., 35. Rammensee, H. G., Friede, T., and Stevanoviic, S. MHC ligands and peptide motifs: Taidi-Laskowski, B., and Levy, R. Tumour specific idiotype vaccines in the treatment first listing. Immunogenetics, 41: 178–228, 1995. of patients with B-cell lymphoma-long term results of a clinical trial. Blood, 89: 36. D’Amaro, J., Houbiers, J. G. A., Drijfhout, J. W., Brandt, R. M. P., Schipper, R., 3129–3135, 1997. Bavinck, J. N., Melief, C. J. M., and Kast, W. M. A computer program for predicting 10. George, A. J., Folard, S. G., Hamblin, T. J., and Stevenson, F. K. Idiotypic vaccina- possible cytotoxic T lymphocyte epitopes based on HLA class I peptide-binding tion as a treatment for a B cell lymphoma. J. Immunol., 141: 2168–2174, 1988. motifs. Hum. Immunol., 43: 13–18, 1995. 11. Stevenson, F. K., Zhu, D., King, C. A., Ashworth, L. J., Kumar, S., and Hawkins, 37. Forero, L., Zwirner, N. W., Fink, C. W., Fernandez-Vina, M. A., and Stastny, P. R. E. Idiotypic DNA vaccines against B-cell lymphoma. Immunol. Rev., 145: Juvenile arthritis, HLA-A2, and binding of DEK oncogene-peptides. Hum. Immunol., 211–228, 1994. 59: 443–450, 1998. 12. King, C. A., Spellerberg, M. B., Zhu, D., Rice, J., Sahota, S. S., Thompsett, A. R., 38. Parker, K. C., Carreno, B. M., Sestak, L., Utz, U., Biddison, W. E., and Coligan, J. E. Hamblin, T. J., Radl, J., and Stevenson, F. DNA vaccines with single-chain Fv fused Peptide binding to HLA-A2 and HLA-B27 isolated from Escherichia coli. Reconsti- 5151

tution of HLA-A2 and HLA-B27 heavy chain/␤ 2-microglobulin complex requires 51. Kuppers, R., Rajewsky, K., and Hansmann, M. L. Diffuse large cell lymphomas are specific peptides. J. Biol. Chem., 267: 5451–5459, 1992. derived from mature B cells carrying V region genes with a high load of somatic 39. Cerundolo, V., Alexander, J., Anderson, K., Lamb, C., Cresswell, P., McMichael, A., mutation and evidence of selection for antibody expression. Eur. J. Immunol., 27: Gotch, F., and Townsend, A. Presentation of viral antigen controlled by a gene in the 1398–1405, 1997. major histocompatibility complex. Nature (Lond.), 345: 449–452, 1990. 52. Ruppert, J., Sidney, J., Celis, E., Kubo, R. T., Grey, H. M., and Sette, A. Prominent 40. Barnstable, C. J., Bodmer, W. F., Brown, G., Galfre, G., Milstein, C., Williams, A. F., role of secondary anchor residues in peptide binding to HLA A2.1 molecules. Cell, and Ziegler, A. Production of monoclonal antibodies to group A erythrocytes. HLA 74: 929–933, 1993. and other human cell surface antigens–new tools for genetic analysis. Cell, 14: 9–20, 53. Hunt, D. F., Henderson, R. A., Shabanowitz, J., Sakaguchi, K., Michel, H., Sevilir, N., 1978. Cox, A. L., Apella, E., and Engelhard, V. H. Characterization of peptides bound to the 41. Haque, T., Amlot, P. L., Helling, N., Thomas, J. A., Sweny, P., Rolles, K., Burroughs, class I MHC molecule HLA-A2.1 by mass spectrometry. Science (Wash. DC), 255: A. K., Prentice, H. G., and Crawford, D. H. Reconstitution of EBV-specific T cell 1261–1263, 1992. immunity in solid organ transplant recipients. J. Immunol., 160: 6204–6209, 1998. 54. Henderson, R. A., Michel, H., Sakaguchi, K., Shabanowitz, J., Appella, E., Hunt, 42. van der Burg, S. H., Ras, E., Drijfhout, J. W., Benckhuijsen, W. E., Bremers, A. J. A., D. F., and Engelhard, V. H. HLA-A2.1-associated peptides from a mutant cell line: Melief, C. J. M., and Kast, W. M. An HLA class I peptide-binding assay based on a second pathway of antigen presentation. Science (Wash. DC), 255: 1264–1266, competition for binding to class I molecules on intact human B cells. Identification of 1992. conserved HIV-1 polymerase peptides binding to HLA-A*0301. Hum. Immunol., 44: 55. Wei, M. L., and Cresswell, P. HLA-A2 molecules in an antigen-processing mutant 189–198, 1995. cell contain signal sequence-derived peptides. Nature (Lond.), 356: 443–446, 1992. 43. van Dongen, J. J., Breit, T. M., Adriaansen, H. J., Beishuizen, A., and Hooijkaas, H. 56. Castelli, C., Tarsini, P., Rini, F., Rivoltini, L., Ravagnani, F., Gallino, F., Belli, F., and Detection of minimal residual disease in acute leukemia by immunological marker Parmiani, G. Novel HLA-Cw8-restricted T cell epitopes derived from tyrosine-related analysis and polymerase chain reaction. Leukemia (Baltimore), 6: 47–59, 1993. protein-2 and gp100 melanoma antigens. J. Immunol., 162: 1739–1748, 1999. 57. Hill, A. V., Elvin, J., Willis, A. C., Aidoo, M., Allsopp, C. E., Gotch, F. M., Gao, 44. Fend, F., Quintanilla-Martinez, L., Kumar, S., Beaty, M. W., Blum, L., Sorbara, L., X. M., Takiguchi, M., Greenwood, B. M., Townsend, A. R., McMichael, A. J., and Jaffe, E. S., and Raffeld, M. Composite low grade B-cell lymphomas with two Whittle, M. C. Molecular analysis of the association of HLA-B53 and resistance to immunophenotypically distinct cell populations are true biclonal lymphomas. Am. J. severe malaria. Nature (Lond.), 360: 434–439, 1992. Pathol., 6: 1857–1866, 1999. 58. Kaminski, M. S., Kitamura, K., Maloney, D. G., and Levy, R. Idiotypic vaccination 45. Dibrino, M., Tsuchida, T., Turner, R. V., Parker, K. C., Coligan, J. E., and Biddison, against a murine B cell lymphoma. Inhibition of tumour immunity by free idiotypic W. E. HLA-A1 and HLA-A3 T cell epitopes derived from influenza virus proteins protein. J. Immunol., 138: 1289–1296, 1987. predicted from peptide binding motifs. J. Immunol., 151: 5930–5935, 1993. 59. Cao, W., Myers-Powell, B. A., and Braciale, T. J. Recognition of an immunoglobulin 46. Dibrino, M., Parker, K. C., Shiloach, J., Turner, R. V., Tsuchida, T., Garfield, M., VH epitope by an influenza virus-specific class I major histocompatibility complex- Biddison, W. E., and Coligan, J. E. Endogenous peptides with distinct amino acid restricted cytolytic T lymphocytes. J. Exp. Med., 179: 195–202, 1994. anchor residue motifs bind to HLA-A1 and HLA-B8. J. Immunol., 152: 620–631, 60. Hamblin, T. J., Davies, Z., Gardiner, A., Oscier, D. G., and Stevenson, F. K. 1994. Unmutated Ig VH Genes are associated with a more aggressive form of chronic 47. Gaugler, B., Van den Eynde, B., van der Bruggen, P., Romero, P., Gaforio, J. J., De lymphocytic leukemia. Blood, 94: 1848–1854, 1999. Plaen, E., Lethe, B., Brasseur, F., and Boon, T. Human gene MAGE-3 codes for an 61. Spellerberg, M. B., Zhu, D., Thompsett, A., King, C. A., Hamblin, T. J., and antigen recognized on a melanoma by autologous cytolytic T lymphocytes. J. Exp. Stevenson, F. K. DNA vaccines against lymphoma: promotion of anti-idiotypic Med., 179: 921–930, 1994. responses induced by single chain Fv genes by fusion to tetanus toxin fragment C. J. 48. DiBrino, M., Parker, K. C., Shiloach, J., Knierman, M., Lukszo, J., Turner, R. V., Immunol., 159: 1885–1892, 1997. Biddison, W. E., and Coligan, J. E. Endogenous peptides bound to HLA-A3 possess 62. Thielemans, K., Rothbard, J. B., Levy, S., and Levy, R. Syngeneic antiidiotypic a specific combination of anchor residues that permit identification of potential immune responses to a B cell lymphoma. Comparison between heavy chain hyper- antigenic peptides. Proc. Natl. Acad. Sci. USA, 90: 1508–1512, 1993. variable region peptides and intact Ig as immunogens. J. Exp. Med., 162: 19–34, 49. Chujoh, Y., Sobao, Y., Miwa, K., Kaneko, Y., and Takiguchi, M. The role of anchor 1985. residues in the binding of peptides to HLA-A*1101 molecules. Tissue Antigens, 52: 63. Mutis, T., Gillespie, G., Schrama, E., Falkenburg, J. H. F., Moss, P., and Goulmy, E. 501–509, 1998. Tetrameric HLA class I-minor histocompatibility antigen peptide complexes demon- 50. Aubin, J., Davi, F., Nguyen-Salomen, F., Leboeuf, D., Debert, C., Taher, M., Valensi, strate minor histocompatibility antigen-specific cytotoxic T lymphocytes in patients F., Canioni, D., Brousse, N., Varet, B., Flandrin, G., and MacIntyre, E. A. Description with graft-versus-host disease. Nat. Med., 5: 839–842, 1999. of a novel FR1 IgH PCR strategy and its comparison with three other strategies for 64. Lister, J. A., and Armitage, J. O. Non-Hodgkin’s lymphomas. In: M. D. Abeloff, J. O. the detection of clonality in B cell malignancies. Leukemia (Baltimore), 9: 471–479, Armitage, A. S. Lichter, and J. E. Niderhuber (eds.), Clinical Oncology, Ed. 2, p. 1995. 2658. New York: Churchill Livingstone, 2000.

5152

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 2001 American Association for Cancer Research. Somatically Mutated Regions of Immunoglobulin on Human B-Cell Lymphomas Code for Peptides That Bind to Autologous Major Histocompatibility Complex Class I, Providing a Potential Target for Cytotoxic T Cells

Clair S. Gricks, Eira Rawlings, Letizia Foroni, et al.

Cancer Res 2001;61:5145-5152.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/61/13/5145

Cited articles This article cites 60 articles, 26 of which you can access for free at: http://cancerres.aacrjournals.org/content/61/13/5145.full#ref-list-1

Citing articles This article has been cited by 5 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/61/13/5145.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/61/13/5145. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.