Cytotoxic T-Lymphocyte Response to Autologous Human Squamous Cell Cancer of the Lung: Epitope Reconstitution with Peptides Extracted from HLA-Aw68'

[CANCER RESEARCH 54, 2731â€”2737,May 15, 1994) Cytotoxic T-Lymphocyte Response to Autologous Human Squamous Cell Cancer of the Lung: Epitope Reconstitution with Peptides Extracted from HLA-Aw68'

Craig L Slingluff, Jr.,2 Andrea L Cox, John M. Stover, Jr., Marcia M. Moore, Donald F. Hunt, and Victor H. Engeihard

Departments ofSurgery (C. L S., J. M. S., M. M. MI, Chemistrj [A. L. C., D. F. H.J, and Mkrobiology [V. H. E.J, University of Virginia, Charlottesville, Virginia 22908

ABSTRACT associated peptides, MHC-unrestricted tumor-specific CTLs have also been described (ii, 12): the peptide backbone of a mucin Cytotoxic T-lymphocytes (Cfls) specific for autologous human squa molecule appears to be the target for some CTLs specific for mom cell cancer of the lung were generated by stimulation of peripheral blood lymphocytes with autologous tumor cells in vitro. The Cl@Lline was carcinomas of the pancreas and of the breast. It is believed that >97% @1J3+,CD8@,CD16andproducedtumornecrosisfactor-a,y-in identification of the peptide epitopes for tumor-specific CTLs will terferon, and granulocyte-macrophage colony-stimulating factor after impact our understanding of the host:tumor relationship and may stimulation with autologous tumor. The CTLs lysed autologous tumor but permit the rational development of novel immunotherapeutic strat failed to recognize autologous or histocompatability leukocyte antigen egies to treat patients with cancer. Although there is evidence of an matched lymphoid cells, K562, or allogeneic tumor cells of several histo immune response to lung cancer, little is known about the target logical types. Antibody-blocking studies suggested that the CTLs re antigens for lung cancer-specific CTLs. cognized one or more antigens presented by the class I major histocom patibility complex molecule Aw68. To Characterize these antigens f@irtber, In situ evidence for a cellular immune response to lung cancers has histocompatablilty leukocyte antigen Aw68 molecules were extracted from been suggested by electron microscopy of human lung cancers, in the squamous cell cancer of the lung tumor line by immunoaffinity cluding squamous cell cancers (13), in which tumor-infiltrating bym chromatography, and the associated peptides were eluted In acid and phocytes exhibited morphological evidence of activation and neigh separated by reversed-phase high-performance liquid chromatography. boring tumor cells were damaged or destroyed. Laboratory Reconstitution ofthe CTL epitope was evaluated by adding these peptides investigation of the cellular immune response to lung cancers has been to autologous Epstein-Barr vinis-transformed B-cells. Two peaks of re constituting activity were observed, suggesting that these Cl'Ls recognize ongoing for at least 2 decades. Much of the early work focused on atleast two Aw68-associated peptides. This study confirms the existence of proliferative and cytotoxic responses of fresh lymphocytes, which a CTL response against autologous human squamous cell cancer of the suggested some reactivity against autobogous lung cancers (14). Gen lung and suggests that this CFL response is directed against peptide eration and characterization of tumor-specific CFL lines was de epitopes presented by the class I matjor histocompatibifity complex mole scribed in 1982 by Vose and Bonnard (15, 16); several of the CTL eWes. It is anticipated that this approach will permit identification of lines were specific for autologous SCCL. Each lysed autobogous peptide epitopes for lung cancer-specific CTLs tumor, failed to lyse K562 or autobogous normal cells, and, with rare exceptions, failed to lyse allogeneic tumors. The nature ofthe epitopes INTRODUCJION for these crLs was not elucidated, and the role of MHC molecules Tumor-specific CTLs@have been generated in vitro from patients was not specifically addressed. Other investigators studied 7-day with many different solid tumors. Among these, the Cli response IL-2-stimulated mixed lymphocyte tumor cultures from patients with to melanoma has been described in the greatest detail. In the lung cancers of unspecified histological type (17). Even at this early majority of cases, the effector cells are CD3@, CD8@, CD4, time, some of the resultant lines lysed autobogous tumor better than CD16, and T-cell receptor-a/p (1â€”4),and the recognition of ailogeneic tumor, and antibody-blocking studies suggested a role for target cells is restricted by class I MHC molecules (5, 6). The CD8@ cells in autobogous tumor lysis. Kurnick et aL, Kradin et al., specific MHC molecules involved have been identified in many and others have generated tumor-infiltrating lymphocyte cultures cases, the best-defined of which is HLA-A2.1 (6â€”9).The HLA from adenocarcinomas of the lung (18) and have used such cultures A2. 1 molecule presents at least six shared peptides that function as for adoptive immunotherapy (19) with modest responses in a majority epitopes for melanoma-specific CTL (9). Several Al-associated of patients. Autobogous tumor lysis was demonstrated in a minority of CTL epitopes exist, and one HLA-A1-associated melanoma pep these cases, tumor specificity was not demonstrated, and the nature of tide has been sequenced (10). Despite the expectation that CTL the CTL epitopes was not evaluated. responses to other tumors would also be directed against MHC In summary, there is substantial evidence of a cellular immune response to human lung cancers in general and to SCCLS in particular, Received 10/1/93; accepted 3/14/94. Thecostsof publicationofthisarticleweredefrayedinpartby thepaymentofpage and there is some evidence that the immune response may have charges. This article must therefore be hereby marked advertisement in accordance with clinical significance and therapeutic potential. The development of 18 U.S.C. Section 1734 solely to indicate this fact. immunotherapy for SCCL will depend on a more detailed understand 1 This work is supported in part by the Cancer Center Grant NIH P30CA44579 (to C. L S. and V. H. E.) at the University of Virginia, by American Cancer Society Grant ing of the host:tumor response and on the identification of specific 1N149H (to C. L S.), by NIH Grants A120963 (to V. H. B.) and GM37537 (to D. F. H.), epitopes for SCCL-specific CTLs. Initially, it is necessary to define andby theDiabetesCenter(NIHP30DK38942)attheUniversityofVirginia. 2 To whom requests for reprints should be addressed, at Department of Surgery, Box the nature of the CFL epitopes for SCCL-specific CTLs. Subse 3111-MR4,UniversityofVirginiaHealthSciencesCenter,Charlottesville,VA22908. quently, identification of shared antigens and specific characterization 3 The abbreviations used are: Cli, cytotoxic T-lymphocyte; MHC, major histocom patibility complex; HLA, histocompatabifity leukocyte antigen; SCCL, squamous cell of those antigens may permit the rational development of novel carcinoma of the lung; IL-2 interleukin 2; rIL-2, recombinant interleukin 2; MEM, immunotherapy. The goal of the present report is to address the initial minimal essential media; F@S,fetal calf serum; pen/strept, 100 Units/mIpenicillin and 100 @g/mlstreptomycin;EBV, Epstein-Barr virus; PHA, phytohemagglutinin antigen; issue of the nature of the CTL epitopes. We will describe restriction TLR, tumor to lymphocyte ratio; TNF, tumor necrosis factor-a; â€˜y-IFN,y-interferon; of SCCL-specific CTLs by class I MHC molecules,will define a GM-tSF, granuincyte-macrophage colony-stimulating factor, EUSA, enzyme-linked specific restricting antigen, and will document the existence of at beast immunosorbent assay; HPLC, high-performance liquid chromatography; VBT2-EBV, EBV-transformed B-cells; E:T, effectortarget ratio. two MHC-associated peptides as epitopes for these CTLs. 273i

Table 1 Human cell lines used in this study: HL4 types linederivedThelungcancersVBT2andSk-Mes-1aresquamouscellcarcinomas.5k-Lu-iandthebreast,colon,andovariancancersareadenocarcinomas.VBT2-EBVisan EBV-B-cell lymphocytes.Thefrom VBT2 PBL and autologous to VBT2 tumor. HLA typing of VBT2, VMM1, and VMM5 was performed by microcytotoxicity assay (Gentrak) on autologous othercellexpression of HLA-Aw68 on tumor cells was confirmed by specific antibody staining of the tumor cells with the monoclonal antibody CRI1â€”351(26). HLA types of the lineshavebeenreported(6,9,41â€”43).Cell 21,22, -DOVBT2line (ref) Cell type FLLA-A -B -C HLA-DR 4CALU-1 Lung cancer, squamous 34, 68 35 â€”Sk-Mes-l(41) Lungcancer,epidermoid 10,11 15,35 â€” â€” â€”5k-Lu-i(41) Lungcancer,squamous 3, 30 7, 27 â€” â€” â€”DM6(41) Lung cancer, adenocarcinoma 24, 32 27, 41 â€” â€” (7?)â€•DM13(6, 9) Melanoma 2.1 12,13,or 35 1,2 6, 10, NDDM14(6, 9) Melanoma 2.1, 31 13, 18 @.4J)C ND (6) Melanoma 11,28 5, 8 2, 4 â€” â€” DM936bSkMel24 (6, 9) Melanoma 2.1, 33 8, 49 ND 2, 4, â€”HT144(9, 41) Melanoma 1, 2.1 12, 14 â€” â€” â€”1ff144(41) Melanoma 1, 24 13, 15 3 4, 7 â€”VMM1A2-03 Melanoma 1, 2.1, 24 13, 15 3 4, 7 â€”VMM5(9) Melanoma 3, 26 51, w4, w6 ND â€” 7VMM1(9) Melanoma 2.1 39 ND 7, 11, 52, 53 2, â€”MDAMB4681 Melanoma 30, 34 18 â€” â€” â€”CCL228(41)(41) Breast cancer 23, 30 27, 35 2, 4 â€” â€”143b Coloncancer 2.1 8,17 â€” â€” â€”GM126(42) Osteosarcoma 2.1 â€” â€” â€” â€”K562 (42) Fibroblasts 2.1 â€” â€” â€” â€”C1R-Aw68 Erythroleukemia â€” â€” â€” â€” (w3)JY (21, 22) EBV-B 68 â€” 4 5 (w12), 52 w7 â€”Herluff(44) EBV-B 2.1,2.1 7, 7 â€” 4, 6 (9) EBV-B 2.1, 2.1 12, 35 â€” â€” â€” a none reportedor not evaluated. Th@ DR antigensare listed becausecross-reactivitypreventedascertainingwith certaintywhich two were correct. C ND, none detected, either by microcytotoxicity assay or by staining with monoclonal antibodies.

MATERIALS AND METHODS 6 monthsaftertumorresection,while he was receivingradiationtherapyand approximately 8 months before he died. PBLS were isolated by Ficoll gradient Cell Lines and HLA Typing. The VBT2 tumor line was obtained from a centrifugation and were used for HLA typing and for the generation of CTLs, metastatic lesion of SCCL in a 45-year-old man. A portion of the surgical PHA blasts, and an EBV-transformed B cell line, VBT2-EBV. The method of specimen was minced, resulting in a single-cell suspension and several fag C'FL generation is detailed below. P1-IAblasts were generated by stimulation ments. Erythrocytes were lysed by the addition of RBC-lysing media (Sigma of PBLSwith PHA (GIBCO, lot 11K5670)diluted 1:100 in RPM! 1640 + 10% Chemical Co., St Louis, MO.). Tumor cells (1 X 10@)and tumor fragments FCS + glutamine + pen/strept (RPMI) for 48 h. EBV-transformed B cells were plated in 12 ml Eagle's MEM (GIBCO, Grand Island, NY) supplemented were generated by incubating PBLS in 1 ml of EBV-containing supernatant with 10% FCS (GIBCO or Whittaker, Walkersville, MD) and pen/strept @ (Gibco) in a flask (Costar Corp, Cambridge, MA). A tumor line was (from B-958 cells) for 1 h at 37Â°Candthen culturing initially in RPMI:PHA, generated and has been in continuous culture for >18 months, now maintained 1:100. The line is maintained long-term in RPM!. in MEM + 5% FfS and pen-strept. Generation and Characterization of Lung Cancer-specific CTLs by All other cell lines were also of human origin. Lung cancer cell lines Stimulation with Autologous Squamous Cell Cancer ofthe Lung. Detailed CALU-1, Sk-Mes-1, and Sk-Lu-1, melanoma cell lines HT144 and Sk-Mel-24, methods of C@L generation have previously been reported (8, 9, 18, 19). Two osteosarcoma 143b, fibroblasts GM126, colon cancer CCL-228, and breast separate batches of VBT2 CTLs were generated. In both cases, PBLS were cancer MDA-MB-468 were obtained from the American Type Culture Col stimulated in vitro with irradiated (100 Gy) cultured autobogoustumor (VBT2) lection. HT144.A2-03 is a melanoma (HT144) transfected with the HLA-A2.1 at a TLR of 1:5 in RPM! supplemented with 20 units/mI rIL-2 (Cetus). The gene (9). Melanomalines DM6, DM13, DM14, and DM93 were the gift of CF1@swererestimulated with irradiated (100 Gy) autologous tumor (VBT2) Hiffiard F. Seigler and Timothy L Darrow (6). VMM1 and VMM5 are every 2 weeks at a TLR of 1:3â€”1:5andwere assayed for cytotoxicity beginning melanoma cell lines established from metastatic melanoma resected from 39 days after initiation of culture. T-cells were maintained in 24-well plates patients at the University of Virginia. JY is an EBV-transformed B-lympho (Linbro, McLean, VA) at a cell density of 2â€”4X 106 cells/well. blastoid line. K562 is a natural killer cell-sensitive human erythroleukemia Phenotyping of Lymphocytes T-cells were evaluated by flow cytometry line. 12 is a human T-cell/B-cell fusion with an antigen-processing defect (20), after staining with fluorescinated or phycoerythrein-conjugated antibodies to and C1R-Aw68 is an HLA-A,B-negative human lymphoid line transfected CD3, CD4, CD8, CD16, CD2S (GenTrak Inc., Plymouth Meeting, PA. and with the Aw68 gene (21, 22), both provided by Peter Cresswell. HLA typing Olympus Corp., Lake Success, NY), and anti-HLA-DR (Becton-Dickinson, was performed on lymphocytes by microcytotoxicity assay (Gentrak). Cell San Jose, CA, and Olympus). lines used in this study have been repeatedly screened for Mycoplasma (Gen Cytotoxicity. Cell-mediated killing was determined by 4-h chromium re Probe, San Diego, CA), and they have been repeatedly negative. The cell lines lease assays, using methods previously described (9). and their known HLA types are listed in Table 1. Immunohistochemlstry. The VBT2 tumor cell line was plated onto glass Measurement of Cytokine Release by CfL VBT2 CfLs (1-2 X 106 slides and then fixed and stained by the immunoperoxidase method, using the cells) were cultured with or without irradiated (100 Gy) VBT2 tumor cells at Vectastain Elite ABC reagent kit (Vector Laboratories, Burlingame, CA). The 1:3 TLR, in 2 ml RPM! + 20 units/mbrIL-2 in 24-well plates (Linbro), using specificities evaluated were epithelial membrane antigen (Dako Corp., Carpin a modification of a protocol reported by Horn et aL (23). Supernatant (1.5â€”2 teria, CA; clone E9); cytokeratin cocktail of AE1/AE3 (Boehringer-Mann ml) was collected after 6â€”8h and then after 24â€”30h and assayed neat. heim, CBA3O4), CAMS.2 (Becton-Dickinson, M0910), 902 (Enzo, 8KDA2), Production of TNF-a, GM-c@SF,and @y-IFNwasmeasured using ELISA kits 903 (Enzo, 7EFB1), and MAK-6+Triton (GJ1001A); and vimentin (Biogenex (Genzyme, Cambridge, MA). Laboratories, San Ramon, CA; clone V9). CTL aonlng. On day 73 of culture, VBT2 CTLs (batch 2) were plated at Preparation ofAutologous Peripheral Blood Lymphocytes, PHA Blasts, limiting dilution at 4, 10, and 100 cells/well in round-bottom 96-well plates and EBV-transformed B-Cells. Peripheral blood was obtained from the (Linbro) with 2 x 10@irradiated VBT2 tumor cells and 4 X 10â€•irradiated patient, VBT2, by venipuncture, after written informed consent, approximately allogeneic EBV-B-cells (JY) per well in RPM! + 20 units/mI rIL-2. Clones 2732

were expanded by restimulation with irradiated autobogous tumor cells and . @ irradiated allogeneic or autobogous EBV-B-cells. .@, I,. .g Extraction of HLA-Aw68-associated Peptides. Peptides bound to Aw68 S. molecules were acid eluted and isolated by centrifuge filtration using a mod â€¢â€˜. O1 q â€¢.*@ @ ification of a protocol previously described (9). VBT2 tumor cells were a..,. â€¢ @â€˜I If cultured in 10-chamber cell factories (Nunc, Thousand Oaks, CA) in MEM + @ 5% FcS and pen/strept. The cells were harvested with 0.03% EDTA + 0.05% 41%@@@ . i@t â€¢;Oc@. @ trypsin (Trypsin-EDTA from GIBCO), washed 3 times in cold phosphate S P@ â€¢ ,, buffered saline, solubiized in 20 ml, per i09 cells, of 1% Nonidet P-40-0.25% sodium deoxycholate-174 p@g/mlphenybmethylsulfonylfluoride-S p@g/mlap $ â€˜@:@#.â€¢.â€¢4 rotonin-lO @Lg/[email protected]/mbpepstatinA-33 @.tg/mbiodoacetamide . ,. Ã¸@lISl@,@@O%S â€¢@ I S 0.2% sodium azide-0.03 pg/mb EDTA at 4Â°Cfor1 h, and then centrifuged for .@ Sb ...S@ â€¢a 1 h at 90,000 X g at 4Â°C.Thepellets were discarded, and the supernatant was I m @â€¢I. â€˜4 passed through a 0.22-gm filter and then passed slowly over two columns in series. The first column contained protein A Sepharose (Sigma) alone as a preclear, and the second column contained CR11â€”351(monocbonal antibody specific for HLA-A2 and Aw68), covalently bound to Sepharose beads. The columns were separately washed and eluted with 10% (1.7 N) acetic acid, pH 2.1. Aw68 molecules and peptides were dissociated by boiling 5 mm, and peptides were separated from masses of >5000 Da by centrifugation through an Ultrafree-CL filter (5000 NMWL, Millipore). Yields were estimated from the quantitation of @-2-microgbobulinandHLA-Aw68 heavy chain obtained using SDS-PAGE. HPLC Fractionation of Peptide Extracts. The peptide extracts were fractionated by reversed-phase HPLC on an Applied Biosystems model 130A separation system. Peptide extracts were concentrated from 5 ml to 100 p1 by vacuum centrifugation, injected onto a Brownlee narrow-bore C18Aquapore column (2.1 mm X 3 cm, 300 A, 7 pin), and eluted with a 40-min gradient of 0â€”60%(v/v) acetonitrile-0.085% trifluoroacetic acid in 0.1% trifluoroacetic acid. Fractions were collected at 1-mm intervals. Epitope Reconstitution. Soluble peptide fractions were partially dehy drated by vacuum centrifugation, reconstituted in assay media (RPM!, 10% Fc@S,antibiotics),and incubated for 2â€”3hwith 2 X 10@51Cr-labeledautolo Fig. 1. Morphology and surface antigen expression by squamous cell cancer of the gous VBT2-EBV in 150 pi assay media/well in 96-well plates. Effector cells lung VBT2. VBT2 tumor cells were plated onto glass slides, to which they were adherent, were added in 100 p.1 assay medium to give an effector:targetratio of fixed, and then stained by the immunoperoxidase method. A, an irrelevant primary antibody used as a negative control. B, panel of anti-cytokeratin antibodies. Magnification, 10:1â€”40:1and were incubated 4 h at 37Â°C.The remainder of the assay was x 250. performed as in standard chromium release assays previously described (9). Wells containing peptide and target cells but no C1'Ls were used as controls to rule out direct toxicity of the peptide fractions themselves. The first two HPLC (anti-A2, anti-Aw68). Similarly, expression of class II MHC mole fractions were usually acidic, and their pH was corrected with 1 MNaOH prior to addingthem to the targetcells. cubes was demonstrated on the tumor cells by flow cytometry (Fig. 2). Monoclonal Antibody BlOcking. 51Cr-labeled autologous tumor cells The level of expression of Aw68 by VBT2 tumor cells is approxi were preincubatedfor 1 h with dilutions of affinity-purifiedmonocbonal mateby 40% as great as the expression of A2.1 on a human lympho antibodies prior to the addition of effector cells (CTLs). Target lysis with and blastoid cell line, JY, based on mean fluorescence values of 108 and without antibody was evaluated in a 4-h assay. The monoclonal antibodies 282, respectively, using CR11â€”351antibody, which binds both A2.1 used were affinity-purified and included W6/32, specific for a monomorphic and Aw68 (data not shown). determinant on all human class I MHC molecules (24); L243, specific for a Autologous Tumor-specific CTL Two batches of VBT2 CTLs determinant on all human DR molecules (25), CR11-351, specific for HLA were generated from PBLS repeatedly stimulated with autologous Aw68 and HLA-A2 (26); and BB7.2, specific for HLA-A2 (27). cultured tumor cells. The resulting cell lines were maintained more than 50 days and more than 150 days, respectively. They bysed RESULTS autobogous VBT2 tumor cells but failed to lyse 20 other target cells, including autobogous PHA blasts, autobogous EBV-transformed B- Characterization of Autologous Tumor. The VBT2 tumor line, cells, and three ablogeneic lung cancer lines (Fig. 3). In subsequent which was generated from a metastatic human squamous cell experiments, an HLA-A34+ melanoma (VMM11) and an HLA cancer of the lung, has been maintained in continuous culture for B35+ lymphobbastoid cell line (Herluft) were not lysed by VBT2 >2 years. Morphologically, the cells are epithelioid and pleiomor CTL, while autobogous tumor was effectively lysed (data not shown). phic (Fig. 1). The doubling time is 1.5 days. To differentiate them Batch 1 CTLs were a mixture of CD8@ and CD4@ CTLs. Batch 2 from fibroblasts, immunohistochemical evaluation of the cultured CTLs are virtually all CD8@ by fluorescence-activated cell sorter tumor cells was performed: cytokeratin was strongly expressed by (99% CD3@, 99% CD8@, 0% CD4@, 0% CD16@). essentially all cells (Fig. 1), and epithelial membrane antigen was Cytokine Release. The CTLs (batch 2, day 131) were assayed for expressed by a subset of cells (data not shown). These findings are cytokine release after stimulation with autobogous tumor. By ELISA consistent with an epithelial lineage. Moderate expression of vi assay, the unstimulated CFLs released no detectable TNF-a or GM mentin was also observed. CSF, but both were released by stimulated CFLs (Table 2). â€˜y-IFNwas HLA Typing of Lymphocytes and Tumor. The patient's lym constitutively produced, but its production increased after stimubation phocytes express HLA-Aw68, A34, B35, Bw6, and C4 as determined with autobogous tumor (Table 2). by microcytotoxicity assay. The presence of HLA-Aw68 expression Characterization of CTL Clones. Cloning was performed by on the VBT2 tumor cells was confirmed by staining with CR11-351 limiting dilution of Cli. batch 2, day 73, when they were 93% CD8@. 2733

I fractions were added to the autobogous EBV-transformed B-cell line and evaluated in a 4-h chromium release assay. crL epitopes were reconstituted by two distinct HPLC fractions, cRll.351 as manifested by lysis of the target cells incubated with those

@ â€˜@__I

K562 VBT2 I CCL-228 Jy. 12 = GM126@ C, 143b@ @?.2 L243 .. HT144' a VMM1 @. DM14 I-. 5kM&24 HT144.A2-03 VMM5@ DM13 DM93 A, 1.' 1@ir IS' DM6 -10 10 20 30 40 50 60 70 80 1 % Specific Cr-SI Release

CONTROL B. C.

60 35

50. /1 @@ : â€¢@@//â€œ / @@ -t@@__@_ie$ Ir IS'

Fig. 2. Expression of HLA-Aw68 by VBT2 tumor cells. VBT2 tumor cells were @ stained with monoclonal antibodies and then fluorescein isothiocyan.tte-labeled sheep 20@ anti-mouse antibody and were then evaluated by flow cytometry. Primary antibodies includeW6/32,CR11.35!,BB7.2,and 1243. Negativecontrolrepresentsstainingby I1@ secondary antibody alone. Positive results are seen for W6/32, CR11.35!, and 1243. .-@ ,@ @ .@ (I A @, -10 , From CTLs plated at 4 cells/well, T-cell colonies arose in 38% of the 2.5 5 10 20 1.25 2.5 5 10 wells. Using the Poisson distribution, we estimated that approximately Effector:target Ratio Effector:target Ratio 70% are derived from an individual starting cell. Maximal Cli Fig. 3. Specific lysis of autologous tumor by VBT2 CFL. Cytotoxicity of VBT2 CFL proliferation during the cloning process was 106-fold. A small sample was evaluated by 4-h chromium release assays. A, batch 1 CTLs assayed on day 39, at an of each well (10 pA)was assayed for lysis of K562 and of autobogous E:T of 20:1. B, batch 2 Cfl.s assayed on day 142 against autologous tumor (R), VBT2 tumor. E:T ratios were calculated from the concentration of autologous PHA-blasts (0), and lymphoid line C1R-Aw68 (ti). C, batch 2 CFLs assayed onday72againstautologoustumor(U),autologousEBV-transformedB-cells(X),K562 CTL clones in each well assayed. All of 23 CTL clones studied were (+),andthreeallogeneicculturedlungcancersCALU-1(0),5k-Lu-i(â€”),andSk-Mes-1 lytic for autobogous tumor and failed to byse K562. Representative (a). CALU-!sharesHLA-B35withVBT2. results for five clones are detailed in Table 3; among them was clone CS, which was lytic for autologous tumor but not for K562. Its Table 2 Cytokineproductionwithautologous by VB72 CTLs in response to stimulation phenotype was 99.8% CD8@ and 0% CD4@ (data not shown). tumorSupernatants @ Restriction by HLA-Aw68. Autobogous tumor lysis was inhibited withirradiated (1.5â€”2ml) were harvested from 106 8 or 30 h after stimulation wereassayedautologous tumor cells, at a tumor:lymphocyte ratio of 1:3, and samples by anti-class I MHC antibody, W6/32, and by antibody to Aw68, forTNF-afor cytokine concentration by ELISA assay. The lower limits of detection CR11â€”351but not by antibody to anti-class II MHC, L243 (Fig. 4), @y-IFN,andGM-@SFwere 5, 100, and 4 pgJml,respectively.Cytokine

consistent with restriction by Aw68. Similar results were found with production (pg/mi) by: several C1@Lclones (data not shown). Interval after Extraction of MHC-associated Peptides and Reconstitution of CTLsTNF-aCytokine stimulation (h) Unstimulated CFLs Stimulated CTL Epitopes. In order to evaluate further the role of Aw68 as a 605GM-@SF 8 <5 47y-IFN 30 <4 restricting molecule for VBT2 CTLs and to estimate the number of 790Table30 456 peptide epitopes presented by Aw68, peptides were extracted specif ically from immunoaffinity-purified Aw68 molecules of VBT2 clones:5 3 CytotoxicityofVB72CTh tumor cells and evaluated for their ability to reconstitute epitopes representative clones and summaryclones% information on all 23 on VBT2-EBV. HLA-Aw68 molecules were extracted from 3 X VBT2erL lysis of i0@ cells by detergent solubilization and immunoaffinity chroma K562G!0clone E:T tumor % lysisof tography. Based on sodium dodecyl subfate-polyacrylamide gel 1B2 1 17 electrophoresis, the average estimated yield of HLA-Aw68 mole 2CS 2 29 cubes was 20â€”SOp@g/i09cells (data not shown). The MHC-asso â€”8Fl! 6 48 4E9 10 41 ciated peptides were then eluted and separated by centrifuge fib 2Mean 16 54 tration as described above. The resulting peptide mixture was fractionated by reversed-phase HPLC, and portions of the resulting for 23 clones 6.5:1 40.4 0.1 2734

______CTLs against melanoma (1, 4). Specificity has been defined by the .I@ ______lysis of autobogous tumor and the failure to lyse autobogous lymphoid cells, K562, or albogeneic tumors of multiple tissue types. CTLs specific for SCCL have previously been generated, but the basis of I L243 their recognition of the autologous tumor was not elucidated (15, 16). I ______ThefocusofthepresentreportisthecharacterizationoftheCTL @ CR1I.351 response to SCCL at a molecular bevel. @ The CTLs were >90% CD8 and the clones that were evaluated were 100% CD8@; therefore, it is reasonable to conclude that the 687.2 :i cytotoxic function of these CFLs is mediated by CD8@ cells. Antigen recognition by CD8@ CTL is usually restricted by class I MHC -20 0 20 40 60 60 100 molecules, and the antibody-blocking studies confirm that these %Inhibition SCCL-specific CTLs are restricted by class I molecules, with marked Fig. 4. MHC restriction of tumor recognition by CFLs: antibody-blocking studies. inhibition by antibody to a monomorphic determinant on class I MHC Chromium release assays were performed, after preincubating VBT2 tumor cells with (W6/32) but not by antibody to a monomorphic determinant on class selected monoclonal antibodies, as described in â€œMaterialsandMethods.â€•Blackcolumns, II MHC (L243@ Most of the class I MHC alleles ex ressed on this @ results with VBT2 CT'Lused on day 106 at an E:T of 10:1, where autologous tumor lysis . â€˜ â€˜. .. I' withoutantibodieswas27%.Whitecolumns,resultswithCFLsassayedonday 142at an tumor line are not ones for which specific monocbonal antibodies are E:Tof 24:1,whereautologoustumorlysiswithoutantibodywas52%.BB7.2wasusedin available; however, affinity-purified CR1 1â€”351,specific for Aw68 this assay only. Antibody concentrations were 50 and 25 g@g/m1,respectively. ... and A2.1, markedly inhibited lysis of autologous tumor, strongly implicating the Aw68 molecule as a specific restricting molecule for these CTLs. Antibody blocking of autobogous tumor lysis by W6/32 A B C exceedsblockingbyCR11â€”351(Fig.4).Explanationsforthisdis crepancy include (a) the possibility that other class I molecules on this 35 tumor may also present antigen to these autologous CFLs or (b) the 35 @ @L differences in affinity or binding sites of the two antibodies may render one more effective than the other at inhibiting Cli lysis. The 25 25 presence of other MHC alleles that present epitopes to these CFLs has not been excluded, but the focus of the present report is on the

15 15 observed restriction by Aw68. The nature of the CFL epitope for viral, albo-, and xenogeneic 5 5 specific CTLs is a short peptide bound to a cleft on the MHC

- - ______molecule (28â€”31). On melanomas and on ovarian cancers, several

@ -5 5 10 15 distinct epitopes appear to exist (7, 9, 32), but only in melanoma 5 10 15 20 5 10 15 20 HPLCfraction,nimb@ has extraction of MHC-associated peptides and reconstitution of epitopes provided direct evidence of class I MHC-associated pep Fig. 5. Reconstitution of epitopes with MHC-associated peptides derived from autol ogous tumor. Peptides eluted from Aw68 molecules (A and B) or all class I MHC tide epitopes (9). In the present report, similar data are now molecules(C)werefractionatedbyreversed-phaseHPLCandcollectedin 1-minfrac available for SCCL. It has been possible to demonstrate the pres tions. These were added to autologous EBV-B-cells in the presence(â€¢)orthe absence (0) ofVBT2 CFL. Three experiments are illustrated. Background lysis ofthe EBV-B-cells by ence of at least two peptide epitopes for these CTLs, which can be VBT2 CTLs, without peptide, is represented by a horizontal dashed line. There is no lysis reconstituted by adding peptides extracted from the SCCL line to by peptide-only, except by fraction 2 in B. Two peaks of reconstituting activity are autologous EBV-transformed B-cells. observed in each experiment (>3 SD over peptide-only controls). The shift in the active fractions between A and Bâ€”Cis a result of using a new HPLC column for the latter The Aw68 molecule can present viral peptides to CTL (30, 33). experiments.Thequantitiesofpeptideusedforeachexperimentrangedfrom5 X 10' to Although not one of the most common MHC molecules, Aw68 is 1.5 x i09 cell equivalents. E:T ratios were 20:1 in A and B and 40:1 in C. Lysis of expressed by 10% of North American Caucasians and by 17% of autologous tumor by CTL alone was 20, 50, and 30% and in A, B, and C, respectively. North American Blacks (34). Its crystal structure is known, and the peptide-binding motif has been determined to include a valine at fractions (Fig. 5, A and B). In control wells with peptide extracts position 2 and a positively charged residue at the carboxyl terminus but no CTLs, lysis was not observed. The peak at fractions 8â€”10 (josition 9) (35). It may be possible, by tandem mass spectrometry was usually more prominent than the peak at fractions 2â€”4,but (36, 37), to determine the amino acid sequence of the peptide both were reproducibly observed. These data are consistent with epitope(s) for Aw68-restricted CFLs, as has been done for A2.1- the presence of at least two HLA-Aw68-associated peptides that restricted xeno-specific murine CTLs (31) and for A2.1-restricted serves as epitopes for these autobogous-specific CTLs. Similar melanoma-specific CTLs (38). results were obtained by reconstituting epitopes with peptides The potential value of identifying peptide epitopes for lung cancer eluted from all class I MHC molecules, using w6/32 antibody: two specific CTLs will be enhanced if this or other epitopes are shared by peaks were observed in identical fractions (Fig. 5C). There is no albogeneic tumors expressing the same MHC antigen(s). Although the convincing evidence that these CTLs recognize epitopes associated presence of shared MHC-associated Cli epitopes has been docu with molecules other than Aw68. mented for melanoma, sarcoma, and ovarian cancer-derived CFLs (6,7,39,40),theyhavenotyetbeendemonstratedforsquamouscell DISCUSSION cancers of the lung or for lung cancers in general. VBT2 CTLs failed to lyse CALU-1, an albogeneic epidermoid cancer of the lung that CFL have been generated in vitro from PBLS of a patient with expresses B35, which is also expressed by VBT2; but allogeneic metastatic SCCL by stimulation with autobogous tumor in the pres Aw68@ SCCLS have not been available for testing. Transfection of ence of rIL-2. They were specifically cytotoxic for autobogous tumor albogeneic tumors with genes encoding the HLA-Aw68.1 and HLA by day 39, consistent with the time frame described for generating Aw68.2 molecules will permit a more methodical evaluation of the 2735

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1994 American Association for Cancer Research. T-CELL RESPONSE TO LUNG CANCER presence of shared Aw68-associated peptide epitopes for VBT2 in human lung carcinoma. Immunol. Invest., 18: 1095â€”1105,1989. 14. Vose, B. M., Vanky, F., Fopp, M., and Klein, E. Restricted autologous lymphocyto CrLs. This work is in progress. toxicity in lung neoplasia. Br. J. Cancer, 38: 375â€”381,1978. There are several implications of identifying peptide epitopes for 15. Vose, B. M., and Bonnard, G. D. Human tumour antigens defined by cytotoxicity and lung-cancer-specific CTLs restricted by Aw68 and, eventually, proliferative responses of cultured lymphoid cells. Nature (Lond.), 296: 359â€”361, 1982. CTLs restricted by more common MHC molecules such as A2.1, 16. Vose,B.M.,andBonnard,G.D.Specificcytotoxicityagainstautologoustumourand Al, and A3. These peptides, or proteins containing them, hold proliferative responses of human lymphocytes grown in interleukin 2. lot. J. Cancer, promise as a basis for a tumor vaccine. Identification of these 29: 33â€”39,1982. 17. Kiinura, H., Yamaguchi, Y., and Fujisawa, T. Cytotoxicity of autologous and allo epitopes may permit creation of the epitopes in transformed B-cell geneic lymphocytes against cultured human lung cancer cells: optimal conditions for lines, which could be used for stimulation of CTLs in vitro or in the production of cytotoxic lymphocytes. Gaas, 75: 1006â€”1016,1984. vivo and which may become a useful reagent for distinguishing the 18. Kurnick, J. T., Kradin, R. L, Blumberg, R., Schneeberger, E. E., and Boyle, L A. Functional characterization of T lymphocytes propagated from human lung carcino immunostimulatory effects of tumor from the putative immuno mas. can. ImmunoL immunopathol., 38: 367â€”380,1986. suppressive or toleragenic effects of the same tumor cells. More 19. Kradin, R. L, Boyle, L A., Preffer, F. I., Callahan, R. J., Barlal-Kovach, M., Strauss, than 100,000 deaths per year (41) may be avoided if novel curative H. W., Dubineti,S., and Kurnick,J. T. Tumor-derivedinterleukin-2-dependent lymphocytes in adoptive immunotherapy of lung cancer. Cancer. ImmunoL Immu therapy would become available for this disease. It remains an nother., 24: 76â€”85,1987. enigma that patients such as the one from whom this tumor and 20. Salter, R. D., Howell, D. N., and Cresswell, P. Genes regulating H1..Aclass I antigen CFLs were derived can have the ability to mount an immune expression in T-B lymphoblast hybrids. Immunogenetics, 21: 235, 1985. 21. Zemmour, J., Little, A-M., Schendel, D. J., and Parham, P. The HLA-A, B response in vitro to the tumor which eventually kills them. By â€œnegativeâ€•mutantcell line C1R expresses a novel HLA-B35 allele, which also understanding the cellular immune response to this cancer at a has a point mutation in the translation initiation codon. J. Immunol., 148: 1941â€”1948,1992. molecular bevel, we hope that therapeutic manipulation of the 22. Storkus, W. J., Salter, R. D., Alexander, J., Ward, F. E., Ruiz, R. E., Cresswell, host-tumor relationship may eventually be possible. P., and Dawson, J. R. Class I-induced resistance to natural killing: identification of nonpermissive residues in HLA-A2. Proc. Natl. Acad. Sci. USA, 88: 5989â€”5992,1991. ACKNOWLEDGMENTS 23. Hom, S. S., Schwartzentruber, D. J., Rosenberg S. A., and TOpalian, S. L Specific release of cytokines by lymphocytes infiltrating human melanomas in response to We thank Bert Sunden for his assistance in phenotypingCTL for cell shared melanoma antigens. J. Immunother., 13: 18â€”30,1993. 24. Barnstable, C. J., Bodmer, W. F., Brown, G., Galfre, G., Milstein, C., Williams, A. F., surface antigens, Bonnie Baird for immunohistochemical assays, Ye Chen for and Zeigler, A. Production of monoclonal antibodies to group A erythrocytes, lILA providing CR11-351-coupled Sepharose beads for use in extraction of Aw68 and other human cell surface antigensâ€”newtools for genetic analysis. Cell, 14: molecules from VBT2 cells, and Dr. John Jane for providing part of a surgical 9, 1978. specimen for establishing the VBT2 tumor cell line. 25. Lampson, L A., and Levy, R. Two populations of Ia-like molecules on a human B cell line. J. Immunol., 125: 293â€”299,1980. 26. Russo, C., Ng, A-!(, Pellegrino, M. A., and Ferrone, S. The monoclonal antibody REFERENCES CR11351 discriminates HLA-A2 variants identified by T-cells. immunogenetics, 18: 23, 1983. 1. Slingluff, C. L, Jr., Darrow, T. L, and Seigler, H. F. Melanoma-specific cytotoxic T 27. Parham, P., and Brodsky, F. M. Partial purification and some properties of BB7.2â€”a cells generated from peripheral blood lymphocytes. Implications of a renewable cytotoxic monoclonal antibody with specificity for HLA-A2 and a variant of source of precursors for adoptive cellular immunotherapy. Ann. Surg., 210: 194â€”202, HLA-A28.Hum.Immunol.,3:277,1981. 1989. 28. Udaka, K@,Tsomides,T. J., and Eisen, H. N. A naturally occurring peptide recognized 2. Yamada, T., Holmes, E. C., and Golub, S. H. Differential regulation by interleukin-4 by alloreactive CD8+ cytotoxic T lymphocytes in association with a class I MHC and interferon-gamma of an autologous melanoma-specific cytotoxic T-cell clone and protein.Cell,69: 989â€”998,1992. the tumor-infiltrating lymphocytes from which it was established. Cancer Commun., 29. Tsomides, T. J., Walker, B. D., and Eisen, H. N. An optimal viral peptide recognized 2: 113â€”121,1990. by CD8+ T cells binds very tightly to the restricting class I major histocompatibility 3. Mukherji, B., Guha, A., Chakraborty, N. G., Sivanandham, M., Nashed, A. L, Sporn, complex protein on intact cells but not to the purified class I protein. Proc. Nat!. Acad. J. R., and Ergin, M. T. Clonal analysis of cytotoxic and regulatory T cell responses Sci. USA, 88: 11276â€”11280,1991. against human melanoma. J. Exp. Med., 169: 1961â€”1976,1989. 30. Silver, M. L, Guo, H-C., Strominger, J. L, and Wiley, D. C. Atomic structure of a 4. Slingluff,C.L, Jr.,Darrow,T.,Vervaert,C.,Quinn-Allen,M.A.,andSeigler,H.F. human MHC molecule presenting an influenza virus peptide. Nature (Lond.), 360: HumancytotoxicTcellsspecificforautologousmelanomacells:successfulgener 367â€”369,1992. ation from lymph node cells in seven consecutive cases. J. Nat!. Cancer Inst., 80: 31. Henderson, R. A., Cox, A. L, Sakaguchi, IC, Appella, E., Shabanowitz, J., Hunt, 1016â€”1026,1988. D.F.,andEngelhard,V.H.Directidentificationofanendogenouspeptiderecognized 5. Parmiani, G., Fossati, G., Taramelli, D., Anichini, A., Balsari, A., Gambacorti by multiple HLA-A2.1 specific cytotoxic T-cells. Proc. Nail. Aced. Sci. USA, Passerini, C., Sciorelli, G., and Cascinelli, N. Autologous cellular immune response 90: 10275, 1993. to primary and metastatic human melanomas and its regulation by DR antigens 32. loannides, C. G., Freedman, R. S., Platsoucas, C. D., Rashed, S., and Kim, Y. P. expressed on tumor cells. Cancer Metastasis Rev., 4: 7â€”26,1985. Cytotoxic T cell clones isolated from ovarian tumor-infiltrating lymphocytes recog 6. Darrow,T.L.,Slingluff,C.L.,Jr.,andSeigler,H.F.TheroleofHLAclassI antigens nize multiple antigenic epitopes on autologous tumor cells. J. Immunol., 146: in recognition of melanoma cells by tumor-specific cytotoxic T lymphocytes. Evi 1700â€”1707, 1991. dence for shared tumor antigens. J. Immunol., 142: 3329â€”3335, 1989. 33. Missale, G., Redeker, A., Person, J., Fowler, P., Guilbot, S., Schlicht, H. J., Ferrari, 7. wolfel, T., Kiehmann, E., Muller, C., Schutt, K. H., Meyer zum Buschenfelde, K. H., C., and Chisari,F. V. HLA-A31-andHLA-Aw68-restrictedcytotoxicI cell re andKnuth,A.Lysisof humanmelanomacellsbyautologouscytolyticTcellclones. sponses to a single hepatitis B virus nucleocapsid epitope during acute viral hepatitis. Identification of human histocompatibility leukocyte antigen A2 as a restriction J. Exp. Med., 177: 751â€”762,1993. element for three different antigens. J. Exp. Med., 170: 797â€”810,1989. 34. Tiwari, J. L, and Terasaki, P. I. The lILA Complex. In: J. L Tiwari and P. I. Terasaki 8. Hom, S. S., Topalian, S. L, Simonis, T., Mancini, M., and Rosenberg, S. A. Common (eds.), lILA and Disease Associations, pp. 10. New York: Springer-Verlag, expression of melanoma tumor-associated antigens recognized by human tumor 1985. infiltrating lymphocytes: analysis by human lymphocyte antigen restriction. J. 35. Guo, H. C., Jardetzky, T. S., Garrett, T. P., Lane, W. S., Strominger, J. L, and Wiley, Immunother., 10: 153â€”164,199!. D.C. Differentlengthpeptidesbindto HLA-Aw68similarlyattheirendsbutbulge 9. Slingluff, C. L, Jr., Cox, A. L, Henderson, R. A., Hunt, D. F., and Engelhard, V. H. out in the middle. Nature (Lond.), 360: 364â€”366,1992. Recognition of human melanoma cells by HLA-A2.1-restricted C1'L is mediated by 36. Hunt, D. F., Henderson, R. A., Shabanowitz, J., Sakaguchi, K.. Michel, H., Sevilhr,N., at least six shared peptide epitopes. J. Immunol., 150: 2955â€”2963,1993. Cox, A. L, Appella, E., and Engeihard, V. H. Characterization of peptides bound to 10. Strausser,J.L, Mazumder,A.,Grimm,E. A., Lotze,M.T., and Rosenberg,S.A. the class I MHC molecule HLA-A2.1 by mass spectrometry. Science (Washington Lysis of human solid tumors by autologous cells sensitized in vitro to alloantigens. DC),255: 1261â€”1263,1992. J. Immunol.,127:266â€”271,1981. 37. Hunt, D. F., Shabanowitz, J., Moseley, M. A., McCormack, A. L., Michel, H., 11. Jerome, K. R., Barnd, D. L., Bendt, K. M., Boyer, C. M., Taylor-Papadimitriou, J., Martino, P. A., Tomer, K B., and Jorgenson, J. W. Protein and peptide sequence McKenzie,I.F., Bast,R. C., Jr., andFinn,0. J. CytotoxicT-lymphocytesderived analysisbytandemmassspectrometryincombinationwitheithercapillaryelectro from patients with breast adenocarcinoma recognize an epitope present on the protein phoresis or microcapiiary HPLC. In: H. Jomvall, J-O. Hoog, and A-M. Gustavsson core of a mucin molecule preferentially expressed by malignant cells. Cancer Res., (eds.), Methods in Protein Sequence Analysis, pp. 257-266. Basel, Switzerland: 51: 2908â€”2916, 1991. Birkhauser Verlag, 1991. 12. Barnd, D. L, Kerr, L A., Metzgar, R. S., and Finn, 0. 3. Human tumor specific 38. Cox, A. L, Skipper, J., Chen, Y., Henderson, R., Darrow, T. L, Shabanowitz, J., cytotoxic T-ceil lines generated from tumor draining lymph node infiltrate. Trans Engeihard, V. H., Hunt, D. F., and Slingluff, C. L, Jr. Identification of a peptide plant. Proc., 20: 339, 1988. recognized by five melanoma-specific human cytotoxic T-cell lines. Science (Wash 13. Wei, Y. 0., and Hang, Z. B., In situ observation of lymphocyte-tumor cell interaction ington DC), in press, 1994. 2736

39. Slovin, S. F., Lackman, R. D., Ferrone, S., Kiely, P. E., and Mastrangelo, M. J. 42. Hay, R., Caputo, J., Chen, T. R., Macy, M., McClintock, P., and Reid, Y. American Cellular immune response to human sarcomas: cytotoxic T cell clones reactive with Type Culture Collection Catalogue of Cell Lines and Hybridomas. Rockville, MD: autologoussarcomas.I.Development,phenotypes,andspecificity.J.Immunol.,137: American Type Culture Collection, 1992. 3042â€”3048,1986. 43. Bernhard, E. J., La, A. 1., Barbosa, J. A., Lacy, E., and Engelhard, V. H. Cytotoxic 40. loannides, C. G., Fisk, B., Poliack, M. S., Frazier, M. L, Wharton, J. T., and T lymphocytes from HLA-A2 transgenic mice specific for HLA-A2 expressed on Freedman, R. S. Cytotoxic T-ceii clones isolated from ovarian tumor-infiltrating human cells J. Exp. Med., 168: 1157, 1988. lymphocytes recognize common determinants on non-ovarian tumour clones. Scand. 44. Engelhard, V. H., and Benjamin, C. Isolation and characterization of monoclonal J. Immunol.,37:413â€”424,1993. mouse cytotoxic T lymphocytes with specificity for lILA-A, -B or -DR alloantigens. 41. Boring, C. C., Squires, T. S., and Tong, T. Cancer Statistics, 1993. CA-Cancer J. J. Immunol., 129: 2621, 1982. Clin.,43:7â€”26,1993.

2737

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1994 American Association for Cancer Research. Cytotoxic T-Lymphocyte Response to Autologous Human Squamous Cell Cancer of the Lung: Epitope Reconstitution with Peptides Extracted from HLA-Aw68

Craig L. Slingluff, Jr., Andrea L. Cox, John M. Stover, Jr., et al.

Cancer Res 1994;54:2731-2737.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/54/10/2731

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/54/10/2731. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.