(CANCER RESEARCH57. 2943-2950. July 15. 19971

ABSTRACT

CTLs specific for high-risk human papillomaviruses (HPVs) have recently been found in the peripheral blood of cervical cancer patients.Although cell-mediated immunity is thought to be important in the controlof HPV infection, the functional relevance and site of activation of HPV.specificCTLs are unclear. We identifiedHLA@A*0201@restrictedHPV-16E7 peptide-specific CTLS in the peripheral blood (four of five patients),draining lymph nodes (three of four patients) and tumors (one of threepatients) of cervical cancer patients. In four of four cancer patients, thefrequency of CTLs specific for a recombinant vaccinia virus expressingHPV-16 and -18 E61E7gene products was found to be higher in tumorsand lymph nodes compared with that of peripheral blood. HPV-specificCTLS were not identified In any of seven healthy controls, but primaryresponses could be generated by peptide-pulsed dendritic cells (four offour controls). In a non@HLA@A*O201subject with invasive carcinoma,other HLA alleles also presented HPV antigens. This is the first demonstration that virus-specific CTLs infiltrate the virus-associated tumor,where they may play an important role in restricting disease progression.

INTRODUCTION

Viruses, including HPV,3 EBV, hepatitis B virus, and human T-cellleukemia/lymphoma virus type I, are associated with approximately20% of all human cancers worldwide. Epidemiologicab studies haveestablished a strong association between HPV and cervical cancer,independent of other risk factors (1); HPV DNA is found in 93% ofall cervical cancers (2), when it is often integrated into the hostgenome (3). Cell-mediated immunity may be particularly important inHPV-associated malignancy, because there is an increased incidenceof associated genital cancer in immunosuppressed patients (4). MHCclass I-restricted CTLs are important in the destruction of virusinfected cells in acute and persistent infections (5), but HPV-specificCTLs are difficult to study, possibly because of the inability topropagate the virus in vitro, with a consequent reliance on a numberof expression systems for HPV proteins. HPV-specific CTLs recognizing HPV E6 and E7 proteins have only recently been demonstratedin peripheral blood of cervical cancer patients in our laboratory (6)and by others (7), as well as in healthy donors (8—10).However,although E6 and E7 proteins of high-risk HPVs are constitutivebyexpressed in cervical cancer cells (1 1), such HPV-specific CTLs havenot been found in ilLs.

Although there are over 20 oncogenic HPV genotypes, HPV-16 is

Received 1/17/97; accepted 5/15/97.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 with18 U.S.C. Section 1734 solely to indicate this fact.

t Supported by the Medical Research Council, United Kingdom. E. M-l. E. is a

Medical Research Council Clinical Training Fellow, and S. M. is a Royal SocietyUniversity Research Fellow.

2 To whom requests for reprints should be addressed. Phone: 44-1222-745256; Fax:

44-1222-745003.3 The abbreviations used are: HPV, human papillomavirus; PBL, peripheral blood

lymphocyte; LNL, lymph node lymphocyte; i'lL. tumor-infiltrating lymphocyte; CIL,cervix-infiltrating lymphocyte; DC, dendritic cell; CIN, cervical intraepithelial neoplasia;IL, interleukin; PHA. phytohemagglutinin; PBMC. peripheral blood mononuclear cell;LDA, limiting dilution analysis; B-LCL, B-Iymphoblastoid cell line; NK, natural killer;LAK, lymphokine-activated killer, ECACC, European Collection of Animal Cell Cultures.

the most prevalent and is present in up to 70% of cervical cancersregardless of geographical origin (2). HLA@A*020l@restricted peptideCTL epitopes derived from HPV-16 E6 and E7 have been identifiedby binding assays (12). These peptides have been used to induce CTLresponses in vitro from healthy donors and in vivo in HLA@A*020ltransgenic mice that were also capable of bysing the cervical carcinoma cell line CaSki (10). In this study, we used two of theseHLA@A*O2Ol@restrictedHPV-l6 E7 peptides to map the distributionof HPV-specific CTLs at different anatomical sites (peripheral blood,draining lymph nodes, and tumor tissue) in patients with cervicalcancer. LDA of PBLs, LNLs, and TILs- was used to identify CTLslysing target cells infected with a recombinant vaccinia virus encoding

modified forms of HPV-l6 and -18 E61E7 protein sequences (TAHPV; Refs. 6 and 13). This is the first examination of HPV-specificCTL responses at sites of disease in cervical cancer patients.

PATIENTS AND METHODS

Patients. The study was approved by the South Glamorgan Local ResearchEthics Committee, and cervical cancer patients were recruited before surgery.

Normal controls were either laboratory staff or patients having a hysterectomy

for nonmalignant reasons. Clinical details are shown in Table I.PBL Extraction and HLA Typing. PBLs were separatedfrom hepa

rinized blood samples by centrifugation on a Histopaque-l077 (Sigma Chemicab Co., St. Louis, MO) density gradient. Screening of PBLs for HLA@A*020l

expression with the monoclonal antibody MA2.l (provided by Prof. A. McMichael, Institute of Molecular Medicine, John Radcliffe Hospital, Oxford,United Kingdom) was performed before HLA class I typing at the National

Blood Transfusion Service (Cardiff, United Kingdom) by microlymphocytecytotoxicity.

H.15VTyping. HPV type-specific (HPV-l6, - I8, and -31) PCR of cervicaltissue was performed by Dr L. Ho (University College London Medical

School, London, United Kingdom; Ref. 14).Cell Lines. The C1R.A2 cell line (provided by Prof. A. McMichael)

expresses a transfected genomic clone of HLA@A*020l ( I5). It was grown inRPMI 1640 with 10% FCS and 400 @sg/mlG4l8 (Life Technologies, lnc.,Gaithersburg, MD). EBV-transformed B-LCLs were established by conven

tional methods using the EBV-producing marmoset cell line B95.8 (ECACC

8501 1419). CaSki is a HLA-A5020l cervical carcinoma cell line expressing

HPV-l6 E6 and El proteins (Ref. 16; ECACC 87020501). The K562 humanerythroleukemic cell line (ECACC 89121407) is a NK cell-sensitive target that

was used to “cold-target compete― with NK activity in cytotoxicity assays (7).

All these lines were grown in RPMI 1640 supplemented with 10% FCS (LifeTechnologies, Inc.). Human T-cell lines were cultured in RPMI 1640 supplemented with 10% pooled human AB serum (National Blood TransfusionService), 100 lU/mI penicillin, 100 @g/mlstreptomycin, 2 m@iL-glutamine, 1mM sodium pyruvate, and 25 msi HEPES buffer (Life Technologies, Inc.),referred to as complete medium. This medium was further supplemented with2 @.tg/mlfungizone (Squibb, Princeton, NJ) for culturing TILs/CILs.

Viruses and Synthetic Peptides. TA-HPV is a recombinant vaccinia virusexpressing HPV-l6 and -18 E6/E7 gene products (provided by Cantab Pharmaceuticals, Cambridge, United Kingdom; Refs. 6 and 13). Wyeth strain is theparental vaccine strain of vaccinia used in the construction of TA-HEW(13).HPV-16 E711-20(YMLDLQPE'VF)and Plasmodium falciparum cp36 (YLKTIQNSL) peptides were synthesized by the Peptide and Protein Facility,University of Wales College of Medicine. HPV- 16 E786_93 (TLGIVCPI) and

the influenza matrix Ml58@ (GILGFVFTL) peptides were synthesized byImmune Systems (Bristol, United Kingdom). All peptides were >80% pure by

2943

Infiltration of Cervical Cancer Tissue with Human Papillomavirus-specific

Cytotoxic T-Lymphocytes'

Eve Mei-ling Evans,2 Stephen Man, Alan S. Evans, and Leszek K. BorysiewiczDepartment of Medicine (E. M-l. E., S. M., L K. B.], University of Wales College of Medicine. and Department of Obstetrics and Gynaecology (A. S. El, University Hospital ofWales NHS Trust, Cardiff CF4 4XN, Wales, United Kingdom

Research. on November 4, 2018. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

Table I Details ofpaiients andcontrolsSubjectAge

(yrs)HLA Class Istage/diagnosisHistologyTreatmentHPV typing―Samplesobtained

(PBLJLNLT1'IIJCIL)P142A2

B44 B27lBWell-differentiated squamous carcinomaRadicalhysterectomy—‘PBLJLNLP2

P335 45A2A31 B44 B60

A2@IIB IIBModeratelydifferentiated squamous carcinoma

Moderately differentiated squamous carcinomaObturator node metsRadiotherapy

Radical hysterectomy—( 16PBLPBLJLNIJFILP452A2@IIAModerately

differentiated adenocarcinomaRadicalhysterectomy16PBLJLNLII'ILPS28A24A25 B8 B44lBModerately differentiated squamous carcinomaRadicalhysterectomy16PBIJLNIJI'ILP626Al

A2 B8 B27lIBPoorly differentiated squamous carcinomaRadical hysterectomy &radiotherapy31PB[JLNIJT'ILVascular

invasionLymph nodemetastasesI72A2

A I I B44 B60CIN IIIExtensive CIN Il—IllLoopexcisionNegPBUCILCI49AlA2 B27 B57Labdonor———PBLC230A2A24 B8 B55Labdonor———PBLC332A2A25BI8BSILabdonor———PBLC446A2

Al I B44 B63Labdonor---PBLCS38A2A3BI8B55Labdonor———PBLC668A2

A25 B7 B44FibroidsNo endocervicaldiseaseHysterectomyNegPBIJLNIJCILC743A2A3 B7 BI4Dysfunctional uterine

bleedingNoendocervical diseaseHysterectomyNegPBLJLNLICIL

HPV-SPECIFLCCFLs IN CERVICALCANCER

a FIGO, FédérationIntemationale des Gynaecologistes et Obstetristes.

( Fresh tumor specimen not available.

‘IPBL stained with monoclonal antibody MA2.I.

high-performance liquid chromatography analysis. In cytotoxicity assays for

peptide-specific CTLs, the Ml,5@ peptide was used as the irrelevant peptideon control targets for HPV peptide-stimulated CTL lines, and the P.falciparumcp36 peptide was used as the irrelevant peptide on control targets forMl58@-stimu1ated CTL lines.

Extraction of LNLs and TILs. Lymph node and tumor samples wereprocessed within 30 mm of surgery. Lymph nodes were dissected free of

surrounding tissue and teased open. LNLs were pelleted and washed severaltimes with PBS before counting. Tumor samples were washed, and a smallfragment was snap-frozen in liquid nitrogen for HPV PCR analysis. Theremainder was minced into 1-mm2 fragments, placed in fungizone-containing

complete medium in 6-well plates, and incubated at 37°C.The medium washarvested at 6, 18, 24, and 36 h, and released cells were pelleted, washed, and

purified by Histopaque-b077 density centrifugation before culture.Flow Cytometric Analysis and Monoclonal Antibodies. Flow cytometric

analysis of cell surface molecules was performed on a FACScan 440 flow

cytometer (Becton Dickinson, Mountain View, CA) with anti-CD3(Leu-4) andanti-CD8(Leu-2a) monoclonal antibodies.

in Vitro Induction of Memory CTh Responses. A modified version ofthe method previously described by Resting er a!. (7) was used. Briefly,4 X 106PBMCs or LNLs (and up to 2 X l0@ilLs) were resuspended in I mlof complete medium and incubated with 10 @.tgofpeptide in 24-well plates. Onday 3, wells were topped up with 1 ml of complete medium containingrecombinant human IL-2 (final concentration, 10 IU/ml; Boehringer Mannheim, Indianapolis, IN). On day 7 and weekly thereafter, cells were restimulated as follows: (a) irradiated autologous PBMCs (3000 rads; 4 X 106cells/ml) were incubated for 2 h at 37°Cin 1 ml of complete medium with 10

@g/mlpeptide and 3 gg/ml j32-microglobulin (Sigma); (b) after removal ofnonadherent cells by washing, 106responder T cells were added in 1 ml ofcomplete medium; and (c) wells were topped up with 1 ml of completemedium containing IL-2 (final concentration, 10 units/mI) 2 days later. Cytotoxicity assays were usually performed on days 14 and 21.

In Vitro Induction of Primary CTL Responses. DCs were preparedby amodification of the method of Bakker et aL (17). Briefly, PBMCs wereresuspended at 4 X 106cells/mI in RPMI 1640 with 5% FCS and distributedin 6-well plates at 3 mI/well. After a 2-h incubation at 37°C,nonadherent cellswere removed, and 3 ml of medium containing 800 units/ml granubocytemacrophage colony-stimulating factor (Boehringer Mannheim) and 500units/mb IL-4 (Genzyme Diagnostics, Cambridge, MA) were added to thewells. On day 6, the adherent cells were harvested by trypsinization andresuspended at 5 X 106 cells/mI in serum-free medium containing 40 @g/ml

peptide and 3 @sg/ml @2-microglobuIin. After incubation at 30°C for 4 h, the

DCs were irradiated (3000 rads). Two X 10' peptide-boadedDCs and 2 X 106responder cells/well were then cocultured in a 24-well plate in 2 ml ofcomplete medium with 5 ng/ml IL-i (Genzyme). The cells were restimubated

on day 10 and weekly thereafter as follows: (a) after irradiation (3000 rads),4 X 106 autologous PBMCs were incubated for 2 h with 20 @sg/mlpeptide and

2 @Wml132-microglobulin in I ml ofmedium; (b) after washing, 106 respondercells were added to each well in 1 ml of complete medium; and (c) the nextday, I ml of complete medium was added, containing IL-2 and IL-7 at finalconcentrations of 10 units/ml and 5 ng/ml, respectively. Cytotoxicity assayswere performed after a minimum of three rounds of restimulation.

Cytotoxicity Assays. Target cells (CIR.A2 cells, CaSki cells, or B-LCLs)were loaded with 51Cr(Amersham International, Little Chalfont, United Kingdom) for 1 h. Virus-infected target cells were infected overnight beforelabeling (multiplicity of infection, 10:1). Peptide-pulsed targets were incubatedwith peptide (10 .tg/ml) for 45 mm after “Crlabeling. Various numbers of

effector cells in 50 pAof medium were preincubated with 50 @lof unlabeledK562 cells (I h at 37°C,in 20-fold excess of the “hot―target cells; Ref. 7).Two X l0@51Cr-labeled target cells were then added to triplicate wells of96-well plates in a final volume of 200 @tl.After 4 h at 37°C,20 @lofsupernatant were harvested onto glass-fiber mats (Wallac, Milton Keynes,United Kingdom) and measured on a (3-plate counter (Wallac). Target andK562 cells incubated with 5% Triton X-lOO(Merck, Poole, United Kingdom)or medium alone were used to determine maximum and spontaneous releaseson each plate. Spontaneous release was usually less than 10% in most assaysand never exceeded 20%. The percentage of specific lysis of each well wascalculated by (experimental release —spontaneous release)/(maximal release —spontaneous release) X 100.

Limiting Dilution Cultures for TA-HPV-specific CTLS. Replicate salcrocultures (100 @b;60—600replicates/cell input number) of freshly isolatedPBLs, LNLs, and TILt were set up in 96-well round-bottomed plates, withdilutions ranging from 100—50,000 cells/webb. Complete medium (100 @d)containing 5 X 10―irradiated allogeneic PBMCs was added to each well,together with IL-2 and PHA (Sigma) in final concentrations of 25 units/mband1 @g/ml,respectively. On day 7, the wells were fed with fresh medium andIL-2, and on day 14, the cells were restimubated with 5 X 10―irradiatedallogeneic PBMCs, 25 units/ml IL-2, and 1 @.tg/ml PHA. The plates were

assayed by split-well analysis on day 21. Two 30-pi aliquots were tested

against C1R.A2 target cells infected with TA-HPV and the parental Wyethstrain. Wells with lysis greater than 10% above that of the control target wereconsidered positive (18). Limiting dilution plots for virus-specific killing wereproduced by plotting the proportion of negative wells against the initial

responder cell number/well on a semilogarithmic plot. From the single-hitPoisson model (a straight line relationship demonstrated by the C2 value forgoodness of fit), the frequency of HPV-specific CTLs was estimated from theinitial responder cell number, at which 37% of the wells were negative forcytotoxicity (19). Statistical analysis was performed using GLIM 3.77 (Royal

Statistical Society, London, United Kingdom).

2944

Research. on November 4, 2018. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

HPv-5PECWICCTLsIN CERVICALCANCER

RESULTS

HPV-16 E7 Peptide-specific CTLs Are Present in PeripheralBlood of Patients with Invasive Cervical Cancer. We have previously shown that vaccination of patients with cervical cancer with avaccinia recombinant expressing E61E7 from both HPV-l6 and -18(TA-HPV) induced CTLs in PBLs (6). To determine whether HPVspecific memory CTLs were present without vaccination, two HLAA*O2O1@restricted HPV-16 E7 peptides were used in an in vitro

restimulation protocol (7) to screen peripheral blood of HLA@A*020lcervical cancer patients and healthy controls (Table 1). In four of fivecancer patients, peptide-specific CTLs to HPV-l6 E71 1-20 were de

tected in cytotoxicity assays on day 14 (Fig. la). In one of these fourpositive patients, specific CTLs against HPV-16 E786.93 were alsoobserved (Fig. lb). CaSki, a HLA@A*O2Ol cervical carcinoma cellline expressing HPV-l6 E6 and E7 proteins (16), and CIR.A2 cells

infected with TA-HPV were used as targets to assess recognition ofendogenously processed antigen. All four peptide-specific lines recognized TA-HPV-infected targets at lower levels of specific lysis thanpeptide-pulsed targets. The HPV-l6 E786...93-specific CTL line fromsubject P2 (Fig. lb) was the only peptide-specific line (one of four) tokill CaSki cells (Fig. 2). To determine whether these specific CTLresponses were maintained in vivo, PBMCs from two patients (Pb and

Fig. 1. HLA-A50201 -restricted peptide-specific C'I'Ls in PBLS from patients and controlsafter a 14-day culture with (a) HPV-l6 E71 t2o'(b) HPV-16 E756.,3, (c) influenza A Ml,5@, and(‘0P.fakiparumcp36.Resultsshowtherepresentative percentage of specific lysis at an E:Tratio of 100: 1 of peptide-pulsed (U) CIR.A2 targets in a chromium release assay, in the presenceof K562 cells. Other targets included Cl R.A2 cellsloaded with an irrelevant HLA-A50201-restrictedpeptide (c), no peptide (D)@infected with therecombinant vaccinia virus TA-HPV () or theparentalWyethvaccinia(U).Allassayswereperformed at multiple E:T ratios.

dCl)U)>%

0

011)aU)

2945

P1 P2 P3 P4 P6 Pcinl Cl C2 C3 C4 C5 C6 C7

Subject

b.SQ 75'U)

‘ 600

@45.C.)a 30;@ :@@ . @t@-1,.@—.@3-I,IP,l-l@@ ,t@-i•Id-1,@fl,U@f1@I@ii,

P1 P2 P3 P4 P6 Pcinl Cl C2 C3 C4 C5 C6 C7

Subject

C

@ I . H@@@@ @1@ @T@@@

@30@@ ri P1@ @. I i

@g 15@@@ I L1 i@ L@@ ‘ —

0@@ L@@

P1 P2 P3 P4 P6 Pcinl Cl C2 C3 C4 C5 C6 C7

Subject

90@

75.

60

45

@@ p21-1 @i,@-@ @1@1-i,@,i-i!“-@.@-i.@- ,@_1 @@7i@1@i@1@

P1 P2 P3 P4 P6 Pcinl Cl C2 C3 C4 C5 C6 C7

Subject

Research. on November 4, 2018. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

Cl)

:14030C1C2@C5

Subjectb.@

E1@40@:ILR

0 20

HPv-sPECIFICCTI@sIN CERVICALCANCER

LNLs in patients with cervical cancer, we examined the tumor forHPV-specific CTL activity in TILs. HPV-l6 El1 1-20peptide-specificCTLs were detected in TILs in one of three cancer patients (P4). TheHPV peptide-specific line from P4 did not recognize CaSki cells butkilled TA-HPV-infected targets at 13—15%bysis above those ofWyeth-infected targets (Fig. 5a). Due to the limited amount of tumor

tissue available, phenotypic analysis of fresh ilLs was only performed in one patient with invasive cancer. This showed the cells tobe 53% CD3 positive and 21% CD8@ (data not shown); functionalstudies could not be performed in this patient due to insufficientnumbers of IlLs. No HPV peptide-specific CTh activity was detectedin lymphocytes from a cervical biopsy specimen of a patient with CIN

III (@cIN') or from normal cervical tissue (Cl). Again, Ml58@-specific responses were induced from TILs/CILs of all subjects(Fig. Sc).

HPV-specific CTLs Are Enriched in TILs and LNLs Comparedto PBLs in Patients with Cervical Cancer. If HPV antigen presentation occurs not just in the tumor but also in draining lymph nodes,HPV-specific CTLs not requiring antigen-specific secondary in vitrorestimulation may be present at both sites. Furthermore, such CTLsmight aggregate at these sites preferentially and be present at higher

a

60

50

40

30

20

10

040 60 80 100

E:T ratio

Fig. 2. HPV-l6 E7,6,3 peptide-specific C'l'L responses from PBLs of patient 2.Results show the percentage of specific lysis. at various E:T ratios, of peptide-pulsed (•)CIR.A2 targets in a day 21 chromium release assay, in the presence of 1(562cells. Othertargets included CIR.A2 cells pulsed with an irrelevant HLA@A*020I@restrictedpeptide(0). nopeptide(0), infectedwithTA-HPV(A)ortheparentalWyethstrainvaccinia(is),and CaSki cells (). SD for all assays was <5% specific lysis.

P2) were reexamined 6 months after treatment, and similar HPVspecific CTLs were again obtained in both cases (data not shown).

After conventional HPV peptide secondary in vitro restimulation,no HPV responses were detected in PBLs from the HPV-negative CINIII patient 1), or in those from seven normal controls. Thissuggested that these HPV-specific CTLs were memory CTLs ratherthan in vitro peptide-induced primary CTLs. This was supported bythe induction of CTLs to the influenza A matrix peptide Ml58@ inall subjects (Fig. Ic), but no CTLs to the P. falciparum cp36 peptide(20) were induced, to which neither patients nor controls were previously exposed (Fig. Id).

In Vitro Induction of HPV-16 Peptide-specific Responses inNormal Donors by DCs. HPV-16 E7 peptide-specific responseswere detected in cervical cancer patients but not in normal subjects,which suggested that the culture conditions used were only able todetect memory CTLs. This was further investigated using DCs derived from PBMCs cultured in granubocyte macrophage cobony-stimulating factor and IL-4 as antigen-presenting cells (17). Under theseconditions, HPV-l6 E71 1-20 and E78693 peptide-specific responseswere generated from PBLs of all four control subjects tested (Cl, C2,C4, and CS); these CTL lines also recognized C1R.A2 target cellsinfected with TA-HPV (Figs. 3, a and b). Again, the suggestion thatpulsed DC restimulation induced primary in vitro responses wassupported by the generation of P. falciparum cp36 peptide-specific

Cii responses in all four subjects (Fig. 3c), whereas no such CTLswere detected in the absence of DCs.

HPV Peptide-specific CTLs Are Present in Draining LymphNodes of Patients with Invasive Cervical Cancer. Having established a protocol for restimubating HPV-l6 E7 peptide-specific CTLsfrom the PBLs of cervical cancer patients, we addressed the questionof whether CTLs were present in tumor-draining lymph nodes. Peptide-specific CTLs to HPV-16@ )20 could be demonstrated in

LNLs from three of four cancer patients (P1, P3, and P6; Fig. 4a).Two of these three positive patients (P3 and P6) had lymph nodemetastases. All three positive lines killed TA-HPV-infected targets atbevels of specific bysis 15—20%above those of parental Wyeth strainvaccinia-infected targets. No HPV-l6 peptide-specific responses wereseen in lymphocytes from pelvic lymph nodes in two patients (C6 andC7) having a hysterectomy for nonmalignant disease. CTLs recognizing the influenza Ml58_@ peptide were generated from LNLs of allpatients (Fig. 4c).

HPV Peptide-specific CTLs Infiltrate Tumor Sites. Havingdemonstrated HPV-l6 E7 peptide-specific CTLs in both PBMCs and

UI

U)>1

C.)

C.)UI0.U)

Subject

C40

:4t@Lj@30

Cl C2 C4 CSSubject

Fig. 3. Primary HLA@A*020l@restricted peptide-specific CTLs are generated in normalcontrols after in vitro stimulation by DCs pulsed with (a) HPV-16 E71 -20' (b) HPV-l6E7as,3, and (c) P. falciparum cp36. Results show the representative percentage ofspecific lysis on day 35, at one E:T ratio (100: 1) to peptide-pulsed (R) CIR.A2 targets ina chromium release assay, in the presence of K562 cells. Other targets included CIR.A2cells loaded with an irrelevant HLA-A5020l-restricted peptide (c), no peptide (0),infected with the recombinant vaccinia virus TA-HPV (@) or the parental Wyeth vaccinia(UI).

2946

Research. on November 4, 2018. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

MHC class I restriction. Therefore, the ability of HLA alleles otherthan HLA@A*O2O1to present HPV antigens was investigated in LNLsand TILs of PS (HJ.A@A*24O1, @A*25O1,@B*O8O1,and @B*44O1).Theautologous B-LCL and two albogeneic B-LCLs matched at two HLAalleles, either HL4.A*2401 and @B*44O1or HLA@A*25O1 and@B*O8O1,wereinfectedwitheitherTA-HPVor theparentalWyeth

strain vaccinia, and used as targets in a day 21 cytotoxicity assay withLNLs or ilLs. In both cases, the percentage of positive wells, inwhich lysis was more than 10% above the lysis of control wells (18),was highest for the autobogous B-LCL and for the allogeneic B-LCLmatched at HLA@A*250l and HLA@B*080b, indicating that HLAalleles other than HLA@A*020l can present HPV-l6/l 8 E61E7-denved antigens (Table 3).

DISCUSSION

We and others have previously identified HPV-specific CTLs inperipheral blood of patients with cervical cancer and CIN (6, 7, 22).The present studies were performed to determine whether such CTLslocalized to anatomical sites of disease and whether their in vivogeneration was dependent on the presence of tumor tissue. Initially westudied PBLs, LNLs, and TILs/CILs from HLA@A*O20b patients

HPv-SPECIFIC CTLs IN CERVICAL CANCER

IL @9

40@

P1 P3 P4 P6 C6 C7

P1 P3 P4 P6 C6 C7

Subject

60'U)

Co>@

.2 40C.)UI

@. 20

@e P1P3P4P6C6C7Subject

C60

U)

U)

; 40C.)UIa.U)

@g

Subject

Fig. 4. HLA@A*O2Ol@restrictedpeptide-specific CTL responses of LNLs from patientsand controls after a 14-dayculture with (a) HPV-l6 E7112@,,(b) HPV-16 Elas93, and (c)influenza A Ml58@. Results show the representative percentage of specific lysis at oneE:T ratio (100:1) of peptide-pulsed (W C1R.A2 targets in a chromium release assay, inthe presence of K562 cells. Other targets included CIR.A2 cells loaded with an irrelevantHLA-A5020l-restricted peptide (UI), no peptide (0), infected with the recombinantvaccinia virus TA-HPV (@) or the parental Wyeth vaccinia (s).

frequencies relative to PBLs. We addressed this by culturing fresh

PBLs, LNLs, and TILs in limiting dilution. No specific antigen wasused, but IL-2 and PHA were added to allow polycbonal expansion ofall normal resting lymphocytes (21). The cytotoxicity of individualculture wells was subsequently tested by split-well analysis onC1R.A2 target cells infected with TA-HPV or the parental Wyethstrain. Under these conditions, HPV-specific CTLs were found inPBLs, LNLs, and ilLs of three cancer patients and in the LNLs andTILs of one cancer patient. No CTLs were detected in the HPVnegative patient with CIN III (PCIN') or in a normal control (C7;Table 2). Single-hit kinetics enabling further comparative analysis bymaximum likelihood analysis (19) were confirmed for HPV-specific

CTLs in LNLs from three of the patients and for TILs from twopatients. In these subjects, HPV-specific C'FL frequencies were high

est in ilLs and lowest in PBLs; CTL frequencies were also higher inthe two patients with lymph node metastases (P3 and P6). The highestHPV-specific frequencies in ilLs were found in P3, where there issignificant enrichment of HPV-specific CTLs in TILs over LNLs andin LNLs over PBLs (Table 2).

HLA Class I Alleles Other than HL4@A*O2O1 Present HPV-16E6/E7-derived Epitopes. The LDA approach used in this studyallowed polyclonal expansion of all HPV-specific CTLs regardless of

aU)U)

.‘ 30@

0

.@ 20

P3 P4 P6 Pcinl C6 C7

Subject

b40.

(I)U)

‘ 30

C.).@- 20

0)a; 0@

P3 P4 P6 Pcinl C6 C7

Subject

C

U) 40@!@°@ P-i@ I@20c@@ H

U) 10@@@ ,@@@ @,:,@@@ 1;

@g 0@j@ IL:@@@

P3 P4 P6 Pcjnl C6 C7Subject

Fig. 5. HLA-A50201-restricted peptide-specific CTLs in TILs/CILs from patients andcontrols after a 21-day culture with (a) HPV-16 @,(b) HPV-l6 E7as@ and (c)influenza A Ml55@. Results show representative percentage of specific lysis at one E:Tratio (15:1) of peptide-pulsed ($) CIR.A2 targets in a chromium release assay, in thepresence of K562 cells. Other targets included CIR.A2 cells loaded with an irrelevantHLA-A50201-restricted peptide (UI), no peptide (0), infected with the recombinantvaccinia virus TA-HPV (U) or the parental Wyeth vaccinia (s).

2947

Research. on November 4, 2018. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

Subject(stage)PBLLNLTIL/CILP3

(IIB)P4 (IB)PS (IB)P6 (HB)P@151 (CIN III)C70.8

(0.6-1.3)0.5 (0.3-1.2)0.9 (0.8—1.2)

<0.01<0.01<0.012.5

(2—2.8)―0.9 (0.8—1.2)0.9 (0.4—2)―2.33 (1.9—2.8)―

—

<0.0117.7

(l3.6@23)a2.8 (2.3—6.8)―4 (3.1—5)9.7 (6.8—13.5)

<0.01<0.01a

Results fulfilledP < 0.05.single-hit

kineticsby the method of x2goodness of fit with

Target B-LCLNo.

of positivewellsLNLTILHLA

A24 A25 B8 B4413/180(7.2%)45/420(10.7%)HLAA2 A24 B7B441/180(0.5%)8/540(1.5%)HLAA2 A25 88Cw712/180(6.7%)37/420(8.8%)HLAA2 Al I B7 B5l0/180(0%)0/180(0%)

HPV-SPECIFIC CTLs IN CERVICAL CANCER

occurred in patients with a HPV-associated tumor. Generating naiveresponses in vitro usually requires longer term culture (30) or specialized antigen-presenting cells ( 10, 17, 3 1). In this study, both HPVand P. falciparum-specific responses were only elicited in normalsubjects when peptide-pubsed DCs were used, making it unlikely thatthe HPV-specific CTLs were memory responses with a very lowprecursor frequency. HPV-l6 infection is widely prevalent even in

women with normal cervical cytology (2), and it has been suggestedthat infection with high-risk HPV types can occur in the perinatalperiod (32). This suggests that localized HPV infection of normalgenital or mucosal epithelial cells fails to elicit a CTL responseagainst HPV E6 and E7 proteins. This may be because epithelial cellsfail to generate CTL responses effectively, because they do notpresent antigenic peptides with appropriate costimubatory molecules(e.g., B7; Ref. 33). However, the observation that HPV-specific CUresponses are detected in cancer patients indicates that this may be aCU response akin to those generated against other tumor antigens. Inthese circumstances, priming of CTLs can use bone marrow-derivedantigen-presenting cells in lymphoid tissue (34), which would bypass

any defect in presentation by the epithebium, or alternatively suggeststhat the transformed cervical epithelium is more efficient at processing and presenting E6- and E7-derived peptides.

The role of HPV-specific CTLs in vivo is unknown, but theirdetection and relative accumulation at sites of disease suggest anability to localize as an antitumor response. We used two humanHLA@A*020l@restricted HPV-l6 E7 peptides (12) to restimulate

memory CTLs. C1R.A2 cells either pulsed with peptide or infectedwith a recombinant vaccinia virus expressing E6 and E7 gene products of HPV-16 and -18 (TA-HPV) were killed, the latter at lowerlevels of specific lysis. However, most peptide-specific lines did notrecognize the HLA@A*O20l + HPV-b6 E61E7-expressing CaSki cervical carcinoma line. This suggests that the high-affinity HPV peptides are not processed as efficiently or in sufficient quantity fromnaturally transformed cervical carcinoma cells. Unfortunately, in thisstudy, no autobogous tumor cell lines were available as targets, whichwould have enabled us to assess whether the natural in vivo target cellcould process and present these peptide epitopes. This is of particularimportance if peptide vaccination is to be used for immunotherapy.

Peptide vaccination can induce CTLs that reject HPV-16-inducedtumors in experimental models of disease (35). In man, vaccinationwith peptide-pulsed autologous DCs may induce specific CTLs. DCsexpress high bevels of HLA class I and II and accessory/costimulatorymolecules (CD54, CD58, CD4O, CD8O, and CD86), secrete cytokines,and migrate into bymphoid organs where optimal priming of T cellscan occur (36). Protective and therapeutic antitumor immunity hasbeen generated in animals by this method (37, 38), and this effect ismediated by CTLs (39, 40). In this study, we have shown that thisapproach can induce, even as a primary response, peptide-specific

Table 3 HPV antigens are presented by HL4 alleles other than HLA.A50201LNLs and ilLs from patient 5 (HLA A24 A25 B8 B44) were cultured with allogeneic

irradiated PBMCs, low-dose IL-2, and PHA and assayed for cytotoxicity on day 21.Targets used for split-well analysis were autologous B-LCL, two allogeneic B-LCLs, eachmatched for two HLA alleles (shown in bold), and one allogeneic B-LCL, mismatched forall HLA class I alleles. Target cells were infected with TA-HPV or Wyeth strain vaccinia,and a positive result was regarded as a well showing a percentage of specific lysis >10%above that of control wells. The percentage of positive wells at an input number of 100cells/well is shown for LNLs and TILs.

Table 2 HPV-specific CTLc are concentrated at sites of disease

PBLS. LNLs, and TIL/CILs were cultured in limiting dilution in the presence of PHA(I @g/ml)andIL-2 (25 lU/mI).On day21,cytotoxicityassayswereperformedbysplit-well analysis using CIR.A2 cells (or the autologous B-LCL, for PS) infected withTA-HPV or parental Wyeth strain vaccinia. A positive result was taken as a percentage ofspecific lysis > 10% above that of the control wells. The HPV-specific CFL frequencyshown (per 106 lymphocytes) was estimated from the initial responder cell number atwhich 37% of the wells were negative. Values in brackets represent 95% confidencelimits.

undergoing hysterectomy, using peptides derived from HPV-16 E7 asa specific marker of the HPV CTL response in vivo. HPV peptidespecific CTLs were found at all these sites and, under these in vitroconditions, in cancer patients but not in normal subjects. Furthermore,

this peptide-specific response was probably secondary, because influenza M@ _66 but not P. falciparum cp36 CTLs were detected in thesame samples. In contrast, peptide-pubsed DCs generated primary invitro CTLs to HPV peptides, as well as to P. falciparum cp36, innormal volunteers.

To determine whether there was aggregation of CTLs at sites ofdisease, we used LDA to establish the frequency of HPV-specificCTLs after polyclonal activation of T cells. These studies confirmedthe presence of these CTLs in PBLs, LNLs, and TILs as well assuggesting that they were retained in higher relative numbers at the

sites of disease. The use of polycbonal activation has also enabled

detection of HPV-specific CTL responses restricted by MHC class Ifor which peptides have not been identified.

HPV is only one of several viruses associated with human malignancies. EBV, a ubiquitous y-herpes virus, is associated with Burkiu'slymphoma, Hodgkin's disease, nasopharyngeal carcinoma, and B-celllymphomas in immunosuppressed patients. During normal EBV persistence/latency, MHC class I-restricted CTLs play a major robe in

maintaining the virus/host equilibrium (23). In fact, the developmentof EBV@ B-cell bymphomas in immunosuppressed subjects is associated with loss of EBV-specific CTLs, and their direct replacementby adoptive transfer is therapeutically effective (24). These transferredCTLs not only restored cellular responses against EBV but persistedin vivo for up to 18 months (25). Whether this therapeutic approach isapplicable to other EBV-associated tumors such as Hodgkin's diseaseis an important question. EBV is detected in approximately 40% of

Hodgkin's disease, and EBV-specific CTL responses in TILs (fromclinically affected lymph nodes) and PBLs have been investigated

(26). In contrast to our findings with HPV, MHC class I-restrictedkilling of autologous targets by TILs in Hodgkin's disease was foundonly in EBV tumors, whereas TILs from EBV@ tumors were eithernoncytotoxic or exhibited LAK activity. Comparison of TILs andPBLs in one patient showed EBV-specific cytotoxicity only in thePBL population. These differences highlight the important role offactors that may suppress CTLs locally (e.g., tumor-derived cytokines) and the fact that such effects may vary both with tumor typeand the nature of the virus-induced transformation. The latter effectmay determine whether virus gene products expressed in the tumorcells are efficiently processed, such as the failure of EBNA-l to beefficiently processed (27), or interfere with MHC class I presentationin the transformed cell (28, 29).

Peptide-specific CTLs detected in this study were probably in vitroreactivated secondary responses, implying that in vivo priming has

2948

Research. on November 4, 2018. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

HPV-SPECIFIC CTLs IN CERVICAL CANCER

CTLs that recognize processed E6 and E7 peptides (after TA-HPVinfection). Potential binding epitopes from HPV-16 E6 and E7 proteins have been predicted for five human HLA alleles (HLA@A*O1O1,@A*O2O1,@A*O3O1,@A*11O1,and @A*24O1;Ref. 12), although only

HLA@A*020l@restricted CTL responses have been investigated indetail (7, 22). However, we do not know which epitope may bedominant in an individual. We have found HPV-specific CTL restricHon by alleles other than HLA.A*0201 in this study, and analysis ofthe fine specificity of CTLs induced by adenoviruses encoding HPV16/18 E61E7 from patients4 (6) suggests that multiple peptides wouldbe required to minimize viral escape and provide adequate coverageof the outbred population. CTLs recognizing subdominant epitopesmay in fact be effective in tumor rejection (35), whereas a recentreport that peptide vaccination may result in accelerated tumor outgrowth due to induction of T-cell tolerance (41), and the potentialinduction of escape mutants/peptide antagonism, stresses the need toproceed with caution.

Lymphocyte infiltration is seen in cervical cancer (42) and othersolid tumors (43), yet there is little evidence that it controls tumorgrowth and spread. Directly isolated TILs are poorly cytotoxic unlessstimulated with IL-2 (44). Higher concentrations of IL-2 (>6000

units/mI) favor nonspecific activation of NK and LAK cells (45).Previous studies of CD8@ ilLs from cervical cancers found predom

inantly non-MHC-restricted NK- or LAK-mediated killing (46, 47).Some clones killed autobogous tumor cells specifically, but theirantigen specificity or MHC restriction was never determined (46). Westudied the E61E7 specificity of polyclonally activated TILs/CILsusing low concentrations of IL-2 (25 units/mb) to promote outgrowthofT cells. The precise in vivo relationship between tumor cells and theHPV-specific CTLs detected in ilLs in this study cannot be estab

lished; in particular, expansion of circulating PBLs cannot be cxcluded. However, quantitative analysis by LDA suggests that theincreased number of HPV E6/E7-specific CTLs relative to theirfrequency in PBMCs may be due to a selective accumulation of these

effector cells. The presence of these CTLs in ilLs and LNLs does notnecessarily reflect an effective antitumor response. Local factors inthe tumor may prevent destruction of tumor cells, e.g., MHC class Idown-regulation (29), induction of T-cell tolerance by tumor cells thatfail to deliver costimulatory signals (48), or secretion of solubleinhibitory factors such as transforming growth factor f3by tumor cells(49). However, the ability of CTLs to localize to sites of disease doessuggest the possible use of adoptive immunotherapy with CD8@c:TLs for cervical cancer patients with nonresectable disease. We havemaintained HPV-specific CTL clones established from bulk culturesof these cancer patients, and they retain their specificity in vitro for upto 8 months.5

Some HPV types are closely rebated with respect to E6 and E7primary structure, suggesting that cross-reactive CTLs could be gencrated. Interestingly, PCR typing revealed the tumor present in P6 tobe HPV-31 positive. HPV-16 is the most closely related high-riskgenotype to HPV-31; the HPV-16 E71 1-20 amino acid sequence(YMLDLQPETF) differs from that of HPV-31 E71 1-20 (YVLDLQ

PEAT) by only two amino acids. Both these substitutions should still

allow HLA@A*020l binding and may permit CTLs to cross-reactbetween these two HPV strains. We are currently investigating this

possibility.

In summary, we have shown that HPV-specific CTLs are restimubated from PBLs of cervical cancer patients but not from those of

4 M. Nimako, A. N. Fiander, G. W. G. Wilkinson, L. K. Borysiewicz, and S. Man.

HPV-specific C1'Ls in patients with cervical intraepithelial neoplasia grade 111,submittedfor publication.

5 E. Evans, S. Man, and L. K. Borysiewicz, manuscript in preparation.

normal controls and that these CTLs are enriched in TILs and LNLscompared to PBLs. The presence of this response only in patients withdisease associated with HPV-transformed cells suggests that theseantigens may be presented or processed in a different manner fromthat which would normally occur with restricted infection in epithehum. Their role in vivo remains unknown, but their aggregation at

sites of disease suggests that enhancement of these responses in vivoor by adoptive transfer may be a realistic future therapeutic option.

ACKNOWLEDGMENTS

We thank S. Ismail for histopathobogyreports, L. Ho for HPV PCR typing,and M. Nimako and L. Faulkner for TA-HPV and Wyeth stocks.

REFERENCES

I . Schiffman, M. H. Recent progress in defining the epidemiology of human papillomavirusinfectionandcervicalneoplasia.J. Nail.CancerInst.,84:394—397,1992.

2. Bosch, F. X., Manos, M. M., Muñoz,N., Sherman, M., Jansen, A. M., Peto, J.,Schiffman, M. H., Moreno, V., Kurman, R., and Shah, K. V. Prevalence of humanpapilloma virus in cervical cancer: a worldwide perspective. J. NatI. Cancer Inst., 87:796—802, 1995.

3. Cullen, A. P., Reid, R., Campion, M., and Ldrincz, A. T. Analysis of the physical stateof different human papilloma virus DNAs in intraepithelial and invasive cervicalneoplasm. J. Virol., 65: 606—612,1991.

4. Halpert, R., Fruchter, R. 0., Sedlis, A., Butt, K., Boyce, J. G.. and Sillman, F. H.Human papilloma virus and lower genital neoplasia in renal transplant patients.Obstet. Gynecol., 68: 251—258,1986.

5. Borysiewicz, L. K., and Sissons, J. G. Cytotoxic T cells and human herpes virusinfections. Curr. Top. Microbiol. Immunol., 189: 123—150,1994.

6. Borysiewicz, L. K., Fiander, A., Nimako, M., Man, S., Wilkinson, 0. W. G.,Westmoreland, D., Evans, A. S., Adams, M., Stacey, S. N., Boursnell. M. E. G.,Rutherford, E., Hickling, J. K., and Inglis, S. C. A recombinant vaccinia virusencoding human papilloma virus types 16 and 18, E6 and E7 proteins as immunotherapy for cervical cancer. Lancet, 347: 1523—1527,1996.

7. Ressing, M. E., van Driel, W. J., Celis, E., Sette, A., Brandt, R. M. P., Hartman, M.,Anholts, J. D. H., Schreuder, G. M. T., ter Harmsel, W. B., Fleuren, 0. J.. Trimbos,B. J., Kast, W. M., and Melief, C. J. M. Occasional memory cytotoxic T-cellresponses of patients with human papilloma virus type 16-positive cervical lesionsagainst a human leukocyte antigen.A*020l@restricted El-encoded epitope. CancerRes., 56: 582—588,1996.

8. Altmann, A., Jochmus-Kudielka, I., Frank, R., Gausepol, H., Moebius, U., Gissmann,L., and Meuer, S. C. Definition of immunogenic determinants of the human papillomavirus type 16 nucleoprotein E7. Eur. I. Cancer, 28: 326—333, 1992.

9. Tarpey, I., Stacey, S., Hickling, J., Birley, H. D. L., Renton, A., Mclndoe, A., andDavies, D. H. Human cytotoxic T lymphocytes stimulated by endogenously processedhuman papilloma virus type I I E7 recognize a peptide containing a HLA-A2 (A020l)motif. Immunology, 81: 222—227,1994.

10. Ressing, M. E., Sette, A., Brandt, R. M. P., Ruppert, J., Wentworth, P. A., Hartman,M., Oseroff, C., Grey, H. M., Melief, C. J. M., and Kast, W. M. Human CTL epitopesencoded by human papilloma virus type 16 E6 and E7 identified through in viva andin vitro immunogenicity studies of HLA-A50201-binding peptides. J. Immunol.. 154:5934—5943, 1995.

I 1. Dtlrst, M., Gissmann, L., Ikenberg, H., and zur Hausen, H. A papilloma virus DNAfrom a cervical carcinoma and its prevalence in cancer biopsy samples from differentgeographicregions.Proc.Nail.Acad.Sci.USA, 80:3812—3815,1983.

12. Kast, W. M., Brandt, R. M. P., Sidney, J., Drifjhout, J-W., Kubo, R. T., Grey, H. M.,Melief, C. J. M., and Sette, A. Role of HLA-A motifs in identification of potentialCTL epitopes in human papilloma virus type 16 E6 and E7 proteins. J. Immunol.,152: 3904—3912,1994.

13. Boursnell, M. E. G., Rutherford, E., and Hickling, J. K. Construction and charactensation of a recombinant vaccinia virus expressing human papilloma virus proteins forimmunotherapy of cervical cancer. Vaccine, 14: 1485-1494, 1996.

14. Cuzick, J., Terry, G., Ho, L., Hollingworth, T., and Anderson, M. Type-specifichuman papilloma virus DNA in abnormal smears as a predictor of high-grade cervicalintraepithelial neoplasia. Br. J. Cancer, 69: 167—171,1994.

15. Storkus, W. J., Howell, D. N., Salter, R. D., Dawson, J. R., and Cresswell, P. NKsusceptibility varies inversely with target cell class I HLA antigen expression.J. Immunol., 138: 1657—1659,1987.

16. Smotkin, D., and Wettstein, F. Transcription of human papilloma virus type 16earlygenes in a cervical cancer and a cancer-derived cell line and identification of the E7protein. Proc. Nail. Acad. Sci. USA, 83: 4680—4684, 1986.

17. Bakker, A. B. H., Marband, G., de Boer, A. J., Huijbens, R. J. F., Danen, E. H. 3.,Adema, G. J., and Figdor, C. G. Generation of antimelanoma cytotoxic T lymphocytes from healthy donors after presentation of melanoma-associated antigen-derivedepitopes by dendritic cells in vitro. Cancer Res., 55: 5330—5334,1995.

18. Wills, M. R., Carmichael, A. 3., Mynard, K., Jin, S., Weekes, M. P., Plachter, B., andSissons, J. G. P. The human cytotoxic T-lymphocyte (CTL) response to cytomegalovirus is dominated by structural protein pp65: frequency, specificity, and T-cellreceptor usage of pp65-specific CTL. J. Virol., 70: 7569—7579, 1996.

19. Borysiewicz, L. K., Graham, S., Mason, P. D., and Sissons, 3. G. P. Human

2949

Research. on November 4, 2018. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

HPV-5PECIFIC crt@s IN CERVICAL CANCER

cytomegalovirus-specific cytotoxic T cells: their precursor frequency and stage specificity. Eur. J. Immunol.. 18: 269—275. 1988.

20. Plebanski. M., Allsopp. C. E. M., Aidoo, M., Reybum, H.. and Hill. A. V. S.Induction of peptide-specific primary cytotoxic T-Iymphocyte responses from humanperipheral blood. Eur. 3. Immunol., 25: 1783—1787,1994.

21 . Moretta, A.. Pantaleo, G., Moretta, L.. Cerottini, i-C., and Mingari. M. C. Directdemonstration of the clonogenic potential of every human peripheral blood T cell. 3.Exp. Med., 157: 743—754.1983.

22. Evans, C., Bauer, S., Grubert, T., Brucker, C., Baur, S., Heeg, K., Wagner. H.. andLipford. G. B. HLA-A2-restricted peripheral blood cytolytic T-Iymphocyte responseto HPV type 16 proteins E6 and E7 from patients with neoplastic cervical lesions.Cancer Immunol. Immunother.. 42: 151—160,1996.

23. Rickinson, A. B., and Kieff, E. Epstein-Barr virus. In: B. W. Fields, D. M. Knipe, andP. M. Howley (eds.). Virology, Vol. 2, pp. 2397—2447.Philadelphia: LippincottRaven, 1996.

24. Rooney. C. M., Smith, C. A., Ng. C. Y. C.. Loftin, S., Li, C., Krance, R. A., andBrenner. M. K. Use of gene-modified virus-specific T lymphocytes to controlEpstein-Barr virus-related lymphoproliferation. Lancet, 345: 9—13,1995.

25. Heslop. H. E., Ng, C. Y. C., Li, C.. Smith, C. A., Loftin. S. K., Krance, R. A., Brenner,M. K., and Rooney, C. M. Long-term restoration of immunity against Epstein-Barrvirus infection by adoptive transfer of gene-modified virus-specific T lymphocytes.Nat. Med., 2: 551—555,1996.

26. Fnsan. T., Sjoberg, J., Dolcetti, R., Boiocchi, M., Dc Re, V., Carbone, A., Brautbar,C., Battat, S., Biberfeld. P., Eckman, M., Ost, A., Christensson, C., SundstrOm,C.,Bjtirkholm, M., Pisa, P., and Masucci, M. G. Local suppression of Epstein-Barr virus(EBV)-specific cytotoxicity in biopsies of EBV-positive Hodgkin's disease. Blood,86: 1493—1501, 1995.

27. Levitskaya, 3., Coram, M.. Levitsky. V., Imreh, S., Steigerwald-Mullen. P. M., Klein,G.. Kurilla, M. G., and Masucci, M. G. Inhibition ofantigen processing by the intemalrepeat region of the Epstein-Barr virus nuclear antigen- I. Nature (Lond.). 375:685—688, 1995.

28. Rowe, M., Khanna. R.. iacob, C. A., Argaet. V.. Kelly. A., Powis, S., Belich, M.,Croom-Carter, D., Lee, S., Burrows, S. R., Trowsdale, 3., Moss, D. J., and Rickinson,A. B. Restoration of endogenous antigen processing in Burkitt's lymphoma cells byEpstein-Barrvirus latentmembraneprotein-I: coordinatedup-regulationof peptidetransporters and HLA-class I antigen expression. Eur. 3. Immunol., 25: 1374—1384,1995.

29. Cromme, F. V., Meijer. C. 3. L. M.. Snijders. P. J. F., Uyterlinde, A., Kenemans, P.,Helmerhorst, T. H., Stem, P. L., van den Brule, A. 3. C., and Walboomers, 3. M. M.Analysis of MHC class I and II expression in relation to presence of HPV genotypesin premalignant and malignant cervical lesions. Br. 3. Cancer, 67: 1372—1380,1993.

30. Cemy, A.. Fowler, P., Brothers, M. A., Houghton. M., Schlicht, H. J., and Chisari,F. V. Induction in vitro of a primary human antiviral cytotoxic T-cell response. Eur.3. Immunol.,25: 627—630,1995.

3 1. Macatonia, S. E., Patterson, S.. and Knight. S. C. Primary proliferative and cytotoxicT-cell responses to HIV induced in vitro by human dendritic cells. Immunology, 74:399—406. 1991.

32. Kaye. i. N., Starkey. W. G., Kell, B., Biswas, C., Raju. K. S., Best, J. M.. and Cason,J. Human papilloma virus type 16 in infants: use of DNA sequence analyses todetermine the source of infection. 3. Gen. Virol., 77: 1139—1143, 1996.

33. Nickoloff. B. 3., and Turka, L. A. Immunological functions of nonprofessionalantigen-presenting cells: new insights from studies of T-cell interactions with keratinocytes. Immunol. Today, 15: 464—469, 1994.

34. Huang, A. Y. C., Golumbek, P., Ahmadzadeh, M., Jaifree, E., Pardoll, D., andLevistsky, H. Role of bone marrow-derived cells in presenting MHC class I-restrictedtumor antigens. Science (Washington DC), 264: 961—965,1994.

35. Feltkamp, M. C. W., Smits, H. L., Vierboom, M. P. M., Minnaar, R. P., de Jongh,B. M., Drijthout, 3. W., ter Schegget, 3., Melief, C. 3. M., and Kast, W. M.Vaccination with cytotoxic T lymphocyte epitope-containing peptide protects againsta tumor induced by human papilloma virus type 16-transformed cells. Eur. 3. Immunol., 23: 2242—2249, 1993.

36. Austyn, 3. M. New insights into the mobilization and phagocytic activity of dendriticcells. 3. Exp. Med., 183: 1287—1292,1996.

37. Mayordomo, 3. I., Zorina, T., Storkus, W. 3., Zitvogel, L., Celluzzi, C., Falo, L. D.,Melief, C. 3., Ildstad, S. T., Kast, W. M., Delco, A. B., and Lotze, M. T. Bonemarrow-derived dendritic cells pulsed with synthetic tumour peptides elicit protectiveand therapeutic antitumour immunity. Nat. Med., 1: 1297—1302,1995.

38. Zitvogel, L., Mayordomo, 3. 1., Tjandrawan, T., DeLco, A. B., Clarke, M. R., Lotze,M. T.. and Storkus, W. 3. Therapy of murine tumors with tumor peptide-pulseddendritic cells: dependence on T cells, B7 costimulation, and T helper cell 1-associated cytokines. 3. Exp. Med., 183: 87—97,1996.

39. Ossevoort, M. A., Feltkamp, M. C. W., van Veen, K. J. H., Melief, C. 3. M., and Kast,W. 3. Dendntic cells as carriers for a cytotoxic T-lymphocyte epitope-based peptidevaccine in protection against a human papilloma virus type 16-induced tumor.3. Immunother..18: 86—94,1995.

40. Celluzzi, C. M., Mayordomo, 3. I., Storkus, W. J., Lotze, M. T., and Falo, L. D.Peptide-pulsed dendritic cells induce antigen-specific CTL-mediated protective tumorimmunity. 3. Exp. Med., 183: 283—287,1996.

41. Toes, R. E. M., Blom, R. 3. 3., Offringa, R., Kast, W. M.. and Melief, C. 3. M.Enhanced tumor outgrowth after peptide vaccination: functional deletion of tumorspecific C'l'L induced by peptide vaccination can lead to the inability to reject tumors.3. Immunol.,156: 3911—3918,1996.

42. Ferguson, A., Moore, M., and Fox, H. Expression of MHC products and leucocytedifferentiation antigens in gynaecological neoplasms: an immunohistological analysisof the tumour cells and infiltrating leucocytes. Br. J. Cancer, 52: 551—563,1985.

43. Whiteside, T. L., Jost, L. M., and Herberman, R. B. Tumor-infiltrating lymphocytes:potential and limitations to their use for cancer therapy. Cnt. Rev. Oncol. Hematol.,12: 25—47,1992.

44. Itoh, K., Tilden, A. B., and Balch, C. M. Interleukin 2 activation of cytotoxicT-lymphocytes infiltrating into human metastatic melanomas. Cancer Res., 46: 301 1—3017, 1986.

45. Yannelli, J. R., Hyatt, C., McConnell, S., Hines, K., Jacknin, L., Parker, L., Sanders,M.. and Rosenberg, S. A. Growth of tumor-infiltrating lymphocytes from human solidcancers: summary of a 5-year experience. Int. 3. Cancer, 65: 413—421, 1996.

46. Ghosh, A. K.. and Moore, M. Tumour-infiltrating lymphocytes and cervical cancer.Eur. J. Cancer, 28A: 1910—1916,1992.

47. Hilders, C. G. J. M., Ras, L., van Eenderburg, D. H., Nooyen, Y., and Fleuren, G. J.Isolation and characterization of tumor-infiltrating lymphocytes from cervical cancer.Int. i. Cancer, 57: 805—813, 1994.

48. Bain. C., Merrouche, Y., Puisieux, I., Duc, A., Colombo, M. P., and Favrot. M. B7.lgene transduction of human renal cell carcinoma cell lines restores the proliferativeresponse and cytotoxic function of allogeneic T cells. Int. 3. Cancer, 67: 769—776,1996.

49. Sulitzeanu, D. Immunosuppressive factors in human cancer. Adv. Cancer Res., 60:247—267,1993.

2950

Research. on November 4, 2018. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

1997;57:2943-2950. Cancer Res   Eve Mei-ling Evans, Stephen Man, Alan S. Evans, et al.   Papillomavirus-specific Cytotoxic T-LymphocytesInfiltration of Cervical Cancer Tissue with Human

  Updated version

  http://cancerres.aacrjournals.org/content/57/14/2943

Access the most recent version of this article at:

   

   

   

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

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

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

.http://cancerres.aacrjournals.org/content/57/14/2943To request permission to re-use all or part of this article, use this link

Research. on November 4, 2018. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from