Evidence for in Situ Amplification of Cytotoxic T-Lymphocytes with Antitumor
Activity in a Human Regressive MelanomaAndreas Mackensen,1 Laurent Ferradini, Guislaine Carcelain, Frederic Triebel, Florence Faure,
Sophie Viel, and Thierry HercendLaboratoire d'Hemato-lmmuno/ogie. INSERM U333. Instimi Gustave Rnussy, 39 rue Camille Desmmi/ins, F-94805 Villejuif Cedex, France
ABSTRACT
We have derived from lymphocytes infiltrating a human regressivemelanoma lesion a series of T-cell receptor o/ß-dependent, III,A-H14-restricted cytotoxic T-lymphocyte clones reactive against the autologoustumor. Analysis of the T-cell receptor gene expression revealed that all the
clones represented a unique cell expressing a VßO.l/Jßl.l gene segment.T-cell receptor transcripts expressed in the cloned cells were compared to
those present in the uncultured tumor tissue. This analysis demonstratedthat the specific cytotoxic T-lymphocyte clones characterized in vitro was
actually selected and amplified in vivo at the lesion site. These resultsprovide strong evidence that effector T-cells have contributed to tumor
regression.
INTRODUCTION
The contribution of the immune system to host defense againsttumors has been an extensively debated question over the last decades(1,2). In the human, there is still little if any direct evidence to supportthe immunosurveillance concept in particular for solid tumors. Malignant melanoma has represented one of the important models toassess the putative mechanisms of antitumor immunity. Numerousstudies have shown that one can generate TIL2-derived T-cell lines or
clones with major histocompatibility complex-restricted cytotoxic activity against autologous tumor cells in this disease (3-5). The rele
We have analyzed previously a case of melanoma displaying thepathological and clinical criteria of tumor regression. It was shownthat the regressive lesion was infiltrated by T-lymphocytes with aCD3 +, CD4+, or CD8+ and TCRa/ß+/TCRy/g- phenotype (8). The
molecular structure of the TCR ßchains expressed by TILs wascharacterized directly in situ using the bispecific PCR technique. Theresults demonstrated that certain TCR gene segments were markedlyoverexpressed in the tumor compared to autologous peripheral bloodlymphocytes. Extensive sequence analysis of some Vßsubfamilytranscripts revealed that clonal T-cell populations were indeed selected
and amplified at the tumor site (8). In the present study, we havegenerated cell lines and clones derived from the TIL of this regressivemelanoma. All the tumor-reactive CTL clones tested were shown to be
identical, thus deriving from a unique cell. Comparing the sequences
Received 5/3/93; accepted 6/16/93.The costs of publication of this article were defrayed in pan 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.
1Recipient of a fellowship from the Deutsche Forschungsgemeinschaft (Ma 1351/1-1 ).
To whom requests for reprints should be addressed.2 The abbreviations used are: TIL, tumor-infiltrating lymphocyte; TCR. tumor cell-
of the TCR transcripts in the clones and those expressed in the tumortissue, we provide evidence that this specific CTL was amplified insitu. Together, these results strengthen the view that cytotoxic effectorT-cells have actually contributed to the cancer regression.
MATERIALS AND METHODS
Patient. The patient under study was a 65-year-old woman suffering from
a primary malignant melanoma with clinical and histological signs of tumorregression, as described previously (8).
TIL Cultures and Clones. TIL were obtained from tumor fragments mechanically dispersed into single cell suspensions and cultured in 12-well plates
(Becton Dickinson, Lincoln Park, NJ) in medium consisting of RPMI 1640(Biochrom. Berlin. Germany) supplemented with 10% human AB serum,amino acids, pyruvate. antibiotics, 100 units/ml human recombinant IL-2(Roussel-Uclaf, Romainville, France), and 3% TCGF. The culture medium wasreplenished every 3-5 days.
After 3 weeks of expansion, a TIL cell line was generated plating 5 X IO3
TIL/well in 96-well. V-shaped microtiter plates (Nunc, Roskilde, Denmark) inthe presence of irradiated autologous EBV-B cells (60 Gy, 1.5 X IO4cells/well)plus allogeneic PBL (35 Gy, 7 x IO4 cells/well) and in the presence of
IL-2/TCGF.
After 2 weeks of culture, the TIL cell line was further cloned by limitingdilution (0.3 cell/well) in the presence of autologous EBV-B cells, allogeneicPBL, and IL-2/TCGF. Two weeks later, the generated T-cell clones wereexpanded by transferring them to 96-well round-bottomed microtiter plates(Becton-Dickinson).
Tumor Cell Lines. Fresh melanoma cells were obtained from mechanicallydispersed tumor fragments. One portion of the tumor specimen was rapidlyfrozen in liquid nitrogen awaiting RNA extraction. Another portion was frozenin PCS plus I0"7r dimethyl sulfoxide for later use in cytotoxicity assays. Theremaining portion was placed in a 75-cm2 flask in RPMI 1640 supplemented
with 10% PCS, amino acids, pyruvate. and antibiotics. Tumor lines weremaintained in monolayer and passaged with 0.05% trypsin. Early passageswere cryopreserved with PCS plus 10% dimethyl sulfoxide.
Additional tumor cell lines maintained in culture include a human allogeneic melanoma cell line (K562), a nonadherem human myeloid leukemia line,and natural killer target. LAZ388 and LCL are EBV-transformed lymphoblas-toid (EBV-B) cell lines established from a healthy donor (LAZ388) and from
the melanoma patient (LCD. respectively.m Vhs Serological Reagents, and Phenotypic Analysis of TIL Cell Lines
and T-Cell Clones. BMA 031, kindly provided by Dr. R. Kurrle (Behring
Company, Marburg. Germany), reacts with a monomorphic determinant ofTCR-a/0 receptor (9). Anti-TCR-ol mAb (T Cell Science, Cambridge, MA)recognizes a constant epitope of the human TCR-8 chain (10). OKT3, OKT4,
and OKT8 (Ortho Diagnostics Systems, Inc., Westwood, MA) react with theCD3, CD4, and CDS proteins, respectively. W6/32 (11) and 9^t9 (12) mAbs
recognize nonpolymorphic determinants of HLA class I and class II geneproducts, respectively. Ami-NKTa mAb recognizes an infrequent TCR-a/ßclonotypic determinant (13). XC3, 3B8 (14), and anti-LFAla react with theCD2, CD56. and CD1 la antigens, respectively. EP-2 mAb. kindly provided by
Dr. S. Ferrone (Department of Microbiology and Immunology. New YorkMedical College, Valhalla, NY), detects the high molecular weight melanoma-associated antigen expressed on a large percentage of human malignant melanoma (15). Each mAb was used at a saturating concentration predeterminedby titration curves on positive cloned cell lines.
Human HLA-reactive pregnancy sera (anti-locus A. anti-A2. anti-B14, anti-B35. and anti-BI2) were kindly provided by Dr. P. Bierling (Transfusion
IN SITU AMPLIFICATION OF CTL IN REGRESSIVE MELANOMA
Phenotypic analysis of TIL cell lines and T-cell clones was carried out by
indirect immunofluorescence using an Elite flow cytometer (Coulter Electronics, Inc.. Hialeah, FL) as described previously (13).
Cytotoxicity Assays. The cytotoxic activity of TIL cell lines and T-cellclones was measured by a conventional 4-h 5lCr release assay using triplicate
cultures in V-bottomed plates. E:T ratios were 25:1, 5:1, 1:1. 0.2:1, and 0.04:1
on 2000 target cells/well. The percentage of specific cytotoxicity was calculated conventionally; SDs were <5%. Functional effects of the antibodies andantisera, either on effector cells (OKT3, OKT4, OKT8, XC3, BMA031. anti-LFAla, 3B8, NKTa) or on target cells (W6/32, 9-49, EP-2. NKTa, antiseraanti-HLA-A, anti-HLA-A2, anti-HLA-B12, anti-HLA-B14, anti-HLA-B35),were tested by incubating each of them for 2 h at 37°Cbefore the assay at the
predetermined saturating concentration.The percentage of inhibition of lysis was calculated as
% of specific lysis in mAb/antisera-treated wells1 „¿�, ... , . . ¡ x 100%
% of specific lysis in control wells
Proliferation of T-Cell Clones to Autologous Tumor Cells. Proliferativeresponses of the T-cell clone was examined by incubating 5 x IO4 respondercells with 2.5 X IO4 mitomycin C-treated autologous or allogeneic tumor or
EBV-B cells for 2 days in 96-well round-bottomed plates in RPMI 1640
supplemented with 10% AB serum alone or in combination with 5 units/ml ofIL-2. The cells were pulsed with 1 jxCi of ['H]thymidine 16 h before harvest.['HJThymidine incorporation was measured in a Beckman liquid scintillation
counter.Synthesis of the First Strand of cDNA. Total RNA was prepared from the
tumor fragment, autologous PEL, the TIL-derived cell line, and one T-cellclone using a single-step guanidinium isothiocyanate-phenol chloroform extraction method (16). First strand cDNA was synthesized with oligodeoxythy-
midine priming and reverse transcriptase.Molecular Analysis of TCR V Gene Segments Using PCR. The proce
dure used to analyze TCR Va and Vßgene segment expression has beendescribed previously (17). Briefly, these TCR transcript cDNAs were amplifiedin a 30-cycle PCR, using the 29 different Va and 24 Vß5'-oligonucleotide
primers ( Val-w29/Vßl-w24). each of them paired with the corresponding Ca-or Cß-specific3' primer. Positive a and ßcontrols consisted of constant region
amplifications (180 and 190 base pairs, respectively) and negative controlswere performed without cDNA (17). The amplified products were detected bySouthern blot analysis using a Ca or Cßoligonucleotide probe. Autoradio-graphs were scanned by a computer-assisted imaging system (Bio-Profil. Vil-ber-Lourmat, France) providing an absolute value for each autoradiographic
spot. Each V«or Vßspot was expressed as a percentage of the sum of all Vaor Vßsignals detected on the autoradiogram, respectively.
Cloning and Sequencing of Vßl3Transcripts. The technique used forcloning and sequencing of Vßtranscripts has been described previously (8).Briefly, the primer for the cloning of Vßl3sequences were 5'-CACTGCG-GTGTACCCAGGATATGA-3' in the V region and 5'-ACCAGCTCAGCTC-CGCGGGGTCGG-3' in the Cßregion, the latter containing a Sodi restriction
site. The amplifications were performed in 2 rounds of 30 cycles. After ethanolprecipitation, the amplified material was digested with Sarll, separated on a2% agarose gel, and purified by absorption on glass beads (Gene Clean; BIO101, Inc.. La Jolla, CA). The material was ligated into a Sacll-EcoRV pBS-SK* cloning vector (Stratagene, La Jolla, CA) and used to transform XL-1
blue Escherichia coli strains (Stratagene). The white colonies were screened bythe dot blot technique and a Cßoligonucleotide probe. Plasmid DNA wasextracted from positive colonies and sequenced by the dideoxy-chain termi
nation procedure (Sequenase 2.0; United States Biochemicals, Cleveland, OH).
RESULTS
Derivation and Characterization of TIL Cell Lines. TILs,freshly isolated from the previously in situ studied tumor specimen,were expanded with IL-2/TCGF for 3 weeks and then plated at 5000cells/well in the presence of irradiated autologous EBV-B cells plus
allogeneic PBL as feeder cells.This TIL-derived T-cell line increased by 120-fold at 2 weeks in the
presence of IL-2/TCGF and displayed high levels of cytotoxicity
against autologous tumor cells (data not shown). In contrast, no lysis
of allogeneic melanoma, autologous and allogeneic EBV-B cell lines,
or K562 was observed.Immunofluorescence analysis of the TIL T-cell line revealed a
CD3 ', CDS ', CD4-, TCRo/ß+, and TCRyo' phenotype.
The expression of TCR Va and Vßgene segments in the culturedTIL line (day 13) was assessed here and compared to that observedpreviously in situ, where an increased expression of the Vßl3,Vßl4,and Vßl6gene segments has been demonstrated (8). As shown in Fig.IA, marked alterations in the expression of the Va gene segmentswere observed in the TIL-derived cell line. Some of them were absent
or strongly decreased such as Va3, Va6, Va 14, or Va20 gene segments, while others such as Va5 and Va 15 were markedly overex-
pressed. With respect to TCR ßchains, the Vßl3and Vß22genesubfamilies were increased in the TIL cell line (Fig. Iß).Conversely,Vßl6,previously shown to be the most overrepresented Vßgenesegment in situ (8), had strongly decreased during in vitro culture (Fig.Iß).
Note that another TIL-derived T-cell line, where autologous tumor
cells has been used for in vitro stimulation, has been analyzed. Resultsregarding the expression of TCR Va and Vßgene segments weresimilar (data not shown).
Generation of CTL Clones Directed against Autologous Melanoma. Cells from the TIL line were cloned in the presence of irradiated autologous EBV-B cells, allogeneic PBL, and IL-2/TCGF.More than 100 T-cell clones were obtained. Twenty-five were examined by flow cytometry and all were found to display a CD3 ', CD4",CDS *, TCRa/ß', TCRy/S' phenotype. They were subsequently an
alyzed by PCR amplification with a series of previously describedoligonucleotides (17). Of the 25 clones, 23 were found to express aVßl3transcript and two a chain signals involving the Va5 and Va25gene segments. Note that it is now well established that a unique T-cellmay express two distinct TCR Va chain transcripts (18-20).
Sequence of the Vßl3gene segment from the TIL cell line and oneof the T-cell clones, designed 5G, were compared. cDNAs from theTIL-derived cell line and that of the T-cell clone 5G were found to
represent a unique transcript including the Vßl3.1gene segmentrearranged to Vßl.l with a WGGD encoding junctional sequence(Table 1).
It thus became evident that a unique T-cell had been greatly amplified in culture and the representative T-cell clone 5G was selected
for functional studies. It was tested for cytotoxicity against autologousand allogeneic melanoma cells, autologous and allogeneic EBV-B
cells, autologous and allogeneic PHA blasts, and K562 in conventional MCr release assays.
As shown in Fig. 2, 5G cells displayed high levels of cytotoxicity(>60% specific lysis at an E:T ratio of 5:1) against the autologousmelanoma cells but failed to lyse autologous EBV-B cells, autologous
PHA blasts (data not shown), and all the allogeneic targets (melanomacells, EBV-B cells, PHA blasts) (see Fig. 2).
Note that a specific cytotoxic activity was also found against thefresh autologous melanoma cells (data not shown).
Inhibition of Cytolysis with Monoclonal Antibodies. A panel ofmonoclonal antibodies was tested for the inhibition of specific cy-
tolytic activity against the cultured autologous melanoma target cells(see Fig. ÌA).Anti-CD3, anti-TCRa/ß, anti-CDlla, and to a lesserdegree anti-CD2 mAb inhibited CTL activity. Anti-CD8 mAb wasalso strongly inhibitory, while anti-CD4 mAb had no effect. We testedthe ability of either anti-HLA class I mAb (W6/32) or anti-HLA-classII mAb (9-49) to block the cytotoxic activity at the target cell level.
Pretreatment of autologous melanoma cells with mAb W6/32 inhibited the lytic interactions while mAb 9-49 and other control reagents
Fig. 1. Comparative analysis of TCR Va and Vßgene segment expression on the tumor tissue and a TIL-derived cell line. Relative Va 1-29 (Ai and Vßl-24(ß)subfamily usagewas determined by quantitative PCR and expressed as percentage of the sum of the total Va or Vßsignal obtained after scanning, as indicated in "Materials and Methods."
Inhibition of Cytotoxicity by Human Anti-HLA-B14 Sera. HLAtyping of the patients cells showed that they are HLA-A3, -AIO, -B14,-B35, -CW4, -CW8. To characterize more precisely the restriction
element recognized by the CTL clone 5G, we used a series of relevantantisera from pregnant women in blocking experiments. It was foundthat HLA-B14 is likely to be involved in the recognition of the
autologous tumor cells. Indeed, as shown in Fig. 3B, the corresponding antiserum inhibited the CTL activity up to 80% in 4-h s'Cr release
assays at an E:T ratio of 1:1, while those directed against otherspecificities had little if any effect.
Proliferation of the CTL Clone 5G in Response to AutologousMelanoma Cells. Proliferative response of the 5G CTL clone wastested in 24-h coculture in the presence of either autologous or allo-geneic melanoma cells and autologous or allogeneic EBV-transformedB-cells. As shown in Table 2, clone 5G proliferated following stim
ulation by the autologous melanoma cells, whereas other stimulator
Table 1 Analysis of TCR Vßl3chain transcriptsPCR-generated Vßl3TCR ßchain fragments were cloned into Sacil-EcoRV pBS-SK* vector. Double-stranded plasmid DNAs were sequenced directly with the Sequenase kit
(United States Biochemical). The encoded amino acid sequences of the junctional region are shown according to the one-letter code.
IN SITU AMPLIFICATION OF CTL IN REGRESSIVE MELANOMA
% Specific Lysis
0.04 0.2 1 6 25Effector:Target Ratio
Fig. 2. Specificity of CTL clone 5G directed against autologous melanoma cells. Thecytolytic activity was measured in a 4-h 5lCr release assay on autologous and allogeneicmelanoma cells, autologous and allogeneic EBV-B cells, and K562. Values representmeans of triplicates (SD <5%) at E:T ratios of 25:1. 5:1, I:I, 0.2:1, and 0.04:1. D,autologous melanoma; I, allogeneic melanoma; #. autologous EBV-B, A, allogeneicEBV-B; O, K562.
cells had no effect. Addition of low doses of IL-2 (5 units/ml) aug
mented substantially the proliferation of the CTLs in response to theirradiated autologous melanoma cells.
Together, the results demonstrate that the Vßl3+T-cell clone is a
specific CTL recognizing the autologous melanoma in a conventionalHLA class I-restricted, TCR-dependent fashion.
Evidence for in Situ Clonal Expansion of the Vßl3+CTL. Therelevance of the observation that 5G cells are able to kill autologousmelanoma cells was assessed by testing whether this clone was actually present and potentially overrepresented at the tumor site.
Analysis of the Vß13 transcripts in the tumor in situ revealed thatthe cDNA clone, encoding the same WGGD junctional sequence asCTL clone 5G, represented 30% (6 of 20 sequences) of the Vßl3transcripts in the tumor (Table 1). Note that another Vßl3cDNAclone, encoding an FRSAPP junctional sequence, found in 7 of 20sequences (33%) in the tumor (Table 1), was not identified in the cellcultures.
In contrast, Vßl3+TCR ßchain transcripts from autologous PBL
were polyclonal; i.e., none of the 20 sequences studied was repeated(data not shown). Moreover, the predominant cDNA clone encodingthe WGGD amino acid junctional sequence was not found among the20 Vßl3transcripts of the autologous PBL (data not shown).
DISCUSSION
In a previous study, we have shown through direct sequencing ofTCR transcripts that unique T-cells were selected and amplified at the
tumor site in a spontaneously regressing melanoma lesion (8). Herewe have generated T-cell lines derived from lymphocytes infiltrating
this melanoma. Analysis of the TCR gene expression in the culturedcells revealed a profile substantially different from that evidenced insitu. Indeed, multiple Vßgenes present in vivo were no longer detected while others were overexpressed. Further analysis indicatedthat a unique cell expressing a VßB.l/Jßl.l gene segment had become predominant in the culture. Mechanisms underlying such astrong selection are unknown. It may be related to a growth advantagedependent upon cytokine susceptibility because the Vßl3cells overgrew whether or not autologous tumor cells were used for in vitrostimulation. In addition, cells which were previously found to bepredominant in vivo (such as, e.g., Vßl6f T-lymphocytes) may have
reached a final stage of proliferation and differentiation in situ, preventing further in vitro amplification. In any case, our results demonstrate that analysis of tumor responses based on in vitro studies
exclusively may be misleading. In this regard, we have established anoriginal experimental system which allows us to compare in vitro andin situ data.
The Vßl3+ T-lymphocyte clones characterized here display a con
ventional CTL activity against autologous melanoma cells. Indeed,cytolysis was found to be CD3/TCR and CDS dependent, HLA classI restricted, without cross-reactivity to any additional target cells
(Figs. 2 and 3). In addition, the cloned lymphocytes are able toproliferate specifically in the presence of autologous tumor cells(Table 2).
Blocking experiments with antisera to the relevant HLA moleculesprovided evidence that HLA-B14 is the restriction element recognizedby the Vßl3+ CTLs (Fig. 3). Most of the melanoma-specific CTL
clones presently described have been shown to use HLA-A alÃelesfortumor recognition, in particular HLA-AÃŒ/A2,known to be representedin the population with a high frequency (21-25). In contrast, the roleof HLA-B molecules as restriction elements has been suggested only
following immunoselection experiments (26) or after pretreatment oftumor cells with y-interferon (27). Our results demonstrate thatHLA-B alÃelesare indeed used operationally to present antigens in
malignant melanoma. These findings are in line with the view that in
OKT3
OKT4
OKT»
XC3
BMA031
«ntl-LFAl
SM
W6/32»-4«
EP-2
NKTl
20 40 «0% INHIBITION
B
WS/32
•¿�ntlHLA-A
•¿�nilHLA-A1
anil HLA-BW
•¿�ntlHLA-B36
anil HLA-B12
20 40 CO
% INHIBITION•¿�0 100
Fig. 3. Inhibition of the specific cytotoxicity of clone 5G towards autologous melanoma cells by monoclonal antibodies and antisera. In .4, either effector cells (for OKT3,OKT4. OKT8, XC3, BMA031. anti-LFAla, 3B8, NKTa) or target cells (for W6/32. 9-^49,EP-2, NKTa) were preincubated with each of the antibodies for 2 h before the cytotoxicityassay. In B, target cells were preincubated with mAb (W6/32) or with human antisera toHLA-A, -AI. -B 14. B35, -B12 for 2 h prior to the addition of the effectors (clone 5G).Cytotoxicity was determined in a 4-h 5lCr release assay at an E:T ratio of 1:1, and thepercentage of inhibition was calculated as indicated in "Materials and Methods."
IN SITU AMPLIFICATION OF CTL IN REGRESSIVE MELANOMA
Table 2 Pmliferutive responses of CTL clone 5C to autologous and allogeneic iunior and KBV-B cells
NoneMedium
IL-2"Clone2,520±214
2.657 ±401Autologous
tumor2.572
±142"
6.969 ±33812.877 ±1.133Allogeneic
tumor2.633
±2()61.967 ±1122.5 14 ±199Autologous
EBV-B3,958
±5023,642 ±3053.936 ±197Allogeneic
EBV-B3,779
±4023.300 ±2853.717 ±601
" Results are expressed as mean epm ±SD for [ 'H|thymidine incorporation by clone 5G for triplicate cultures with a responder:stimulator ratio of 2:1.'' Five units/ml.
vivo antitumor responses are complex involving different HLA loci forpresentation of either the same or distinct tumor-associated antigenic
peptides.The major contribution of this study is the demonstration that the
specific Vßl3' melanoma-reactive CTL clone, generated in vitro, is
actually selected and amplified at the tumor site even though thisparticular cell may not be the most represented in vivo (i.e., in light ofin vitro selection). The fact that it was characterized here as a specificCTL while cancer regression was unequivocally documented bothclinically and histologically in this patient strongly supports the viewthat the adaptative immune response has represented an active mechanism of defense against the tumor.
Further studies in this experimental system should provide usefulinformation on the nature of the immune response. Attempts to expandand characterize additional T-cell populations, known to be repre
sented in situ, may help to assess the diversity of the antitumorreaction. This approach should also lead to characterize tumor-asso
ciated antigenic peptides such as for, example, those recently described in the MAGE series (28) in a situation of actual clinicalrelevance.
ACKNOWLEDGMENTS
We thank Drs. Marie-Françoise Avril and Pierre Duvillard for providing
patient materials used in this study.
REFERENCES
1. Burnet. F. M. The concept of immunological surveillance. Prog. Exp. Tumor Res., 13:1-27. 1970.
2. Old. L. J. Cancer immunology: the search for specificity. Cancer Res.. 41: 361-375.1981.
3. Itoh. K.. Platsoucas. C. D., and Balch, C. M. Autologous tumor-specific cytotoxic Tlymphocytes in the infiltrate of human metastatic melanomas: activation by interleu-kin 2 and autologous tumor cells, and involvement of the T cell receptor. J. Exp. Med..168: 1419-1441, 1988.
4. Muul. L. M.. Spiess. P. J., Director. E. P., and Rosenberg. S. A. Identification ofspecific cytolytic immune responses against autologous tumor in humans bearingmalignant melanoma. J. Immunol., MX: 989-995. 1987.
5. Topalian. S. L.. Solomon, D.. and Rosenberg. S. A. Tumor-specific cytolysis bylymphocytes infiltrating human melanomas. J. Immunol., 142: 3714—3725, 1929.
6. Knuth. A.. Danowski. B.. Oettgen. H. F.. and Old. L. J. T-cell-mediated cytotoxicityagainst autologous malignant melanoma: analysis with interleukin 2-dependent T-cellcultures. Proc. Nati. Acad. Sci. USA, 81: 3511-3515, 1984.
7. Van den Eynde. B.. Hainaut. P.. Herin. M.. Knuth. A.. Lemoine, C.. Weynants, P., Vander Bruggen, P., Fauchet, R., and Boon, T. Presence on a human melanoma of multipleantigens recognized by autologous CTL. Int. J. Cancer, 44: 634-640. 1989.
8. Ferradini. L.. Mackensen, A., Genevee, C., Bosq. J., Duvillard. P., Avril, M. F., andHercend. T. Analysis of TCR variability in tumor infiltrating lymphocytes from ahuman regressive melanoma: evidence for in situ T-cell clonal expansion. J. Clin.Invest., 91: 1183-1190, 1993.
9. Lanier. L. L.. Ruitenberg. J. J.. Allison. J. P.. and Weiss. A. Biochemical and flowcytometric analysis of CD3 and Ti expression on normal and malignant T cells. A. J.McMichael (ed.l. In: Leukocyte Typing III. pp. 175-178. Oxford. United Kingdom;Oxford University Press, 1987.
10. Band. H.. Hochstenbach, F.. McLean. J.. Hata, S.. Krangel, M. S.. and Brenner. M. B.
Immunochemical proof that a novel rearranging gene encodes the T cell receptor ôsubunit. Science (Washington DC). 238: 682-685, 1987.
11. Brodsky. F. M.. and Parham. P. Monomorphic anti-HLA-A, B, C monoclonal anti
bodies detecting molecular subunits and combinatorial determinants. J. Immunol..128: 129-135, 1982.
12. Todd. R. F., Meuer. S. C.. Romain. P. L., and Schlossman, S. F. A monoclonalantibody that blocks class II histocompatibility related immune interactions. Hum.Immunol., 10: 23^10, 1984.
13. Hercend. T. Reinherz. E. L.. Meuer. S. C., Schlossman, S. F.. and Ritz. J. Phenotypicand functional heterogeneity of human cloned natural killer cell lines. Nature (Lund.),Ml: 158-160. 1983.
14. Hercend. T, Griffin. J. D.. Bensussan. A.. Brennan, A.. Schmidt. R. E., Edson, M. A.,Daley. J. F.. Schlossman. S. F., and Ritz. J. Generation of monoclonal antibodies to ahuman NK clone: characterization of two NK associated antigens. NKH1A (200-220
KD) and NKH2 (60 KD), expressed on subsets of large granular lymphocytes. J. Clin.Invest., 75: 932-943. 1985.
15. Natali. P. G., Imai, K.. Wilson, B. S., Bigotti. A., Cavaliere. R., Pellegrino, M. A., andFerrone. S. Structural properties and tissue distribution of the antigen recognized bythe monoclonal antibody 653.40s to human melanoma cells. J. Nati. Cancer Inst.. 76:591-601. 1981.
16. Belyavsky. A., Vinogradova. T. and Rajewsky. K. PCR-based cDNA library con
struction: general cDNA libraries at the level of a few cells. Nucleic Acids Res. 17:2919-2932, 1989.
17. Genevee, C., Diu, A.. Nierat. J.. Caignard. A., Dietrich, P. Y., Ferradini. L.. Roman-
Roman. S.. Triebel. F.. and Hercend. T. An experimentally validated panel of subfamily-specific oligonucleotide primers (Val-w29/Vßl-w24) for the study of humanT cell receptor variable V gene segment usage by polymerase chain reaction. Eur. J.Immunol., 22: 1261-1269. 1992.
18. Malissen. M., Trucy. J., Letourneur, F.. Rebai, N.. Dunn, D. E., Fitch, F. W., Hood, L.,and Malissen. B. A T cell clone expresses two T cell receptor a genes but uses onett/ßheterodimer for allorecognition and self MHC-restricted antigen recognition.Cell, 55: 49-59, 1988.
19. Furutani, M.. Yanagi, Y.. Fujisawa, I.. Nakayama. T. Kishimoto, H., Kuida, K.,Asano, Y., and Tada, T. Post-transcriptional allelic exclusion of two functionallyrearranged T cell receptor a genes. Int. Immunol., /: 281-288, 1989.
20. Couez, D.. Malissen. M., Búleme, M.. Schmitt-Verhulst, A. M.. and Malissen. B.Each of the two productive T cell receptor a-gene rearrangements found in both the
AIO and BM 3.3 T cell clones give rise to an a chain which can contribute to theconstitution of a surface-expressed o/ßdimer. Int. Immunol., 3: 719-729, 1991.
21. Van der Bruggen, P., Traversai!, C., Chômez, P., Lurquin, C., De Plaen, E.. Van denEynde, B.. Knuth. A., and Boon, T. A gene encoding an antigen recognized bycytolytic T lymphocytes on a human melanoma. Science (Washington, DC), 254:1643-1647, 1991.
22. Darrow. T. L., Slingluff, C. L., and Seigler. H. F. The role of HLA class I antigens inrecognition of melanoma cells by tumor-specific cytotoxic T lymphocytes. Evidencefor shared tumor antigens. I. Immunol., 142: 3329-3335. 1989.
23. Crowley. N. J., Slingluff. C. L.. Darrow, T. L., and Seigler, H. F. Generation of humanautologous melanoma-specific cytotoxic T-cells using HLA-A2-matched allogeneicmelanomas. Cancer Res., SO: 492^*98. 1990.
24. Wolfel. T., Klehmann, E., Muller. C.. Schutt. K. H., and Meyer zum Buschenfelde. K.H. Lysis of human melanoma cells by autologous cytolytic T cell clones: identification of human histocompatibilily leukocyte antigen A2 as a restriction element forthree different antigens. J. Exp. Med., 170: 797-815. 1989.
25. Chen. Q.. and Hersey, P. MHC-restricted responses of CD8+ and CD4+ T-cell clonesfrom regional lymph nodes of melanoma patients. Int. J. Cancer, 51: 218-224, 1992.
26. Crowley, N. J., Darrow. T. L.. Quinn-Allen. M. A., and Seigler, H. F. MHC-restricledrecognition of autologous melanoma by tumor-specific cytotoxic T cells. Evidence forrestriction by a dominant HLA-A alÃele.J. Immunol.. 146: 1692-1699, 1991.
27. Wolfel. T. Klehmann. E., Meyer zum Buschenfelde. K. H., and Knuth, A. Recognition of a human melanoma antigen in association with two different HLA class 1molecules by autologous CTL clones. Proc. Am. Assoc. Cancer Res., 31: 252, 1990.
28. Traversati. C., Van der Bruggen, P., Luescher. I. F.. Lurquin. C.. Chômez. C.. Van Pel.A.. De Plaen. E.. Amar-Costesec. A., and Boon. T. A nonapeptide encoded by humangene MAGE-1 is recognized on HLA-A 1 by cytolytic T lymphocytes directed againsttumor antigen MZ2-E. J. Exp. Med.. 176: 1453-1457. 1992.
![Page 1: Evidence for in Situ Amplification of Cytotoxic T …cancerres.aacrjournals.org/content/canres/53/15/3569...[CANCER RESEARCH 53. 3569-3573, August 1. 1993] Evidence for in Situ Amplification](https://reader030.documents.pub/reader030/viewer/2022040722/5e31abba8887c7279f13b558/html5/thumbnails/1.jpg)
![Page 2: Evidence for in Situ Amplification of Cytotoxic T …cancerres.aacrjournals.org/content/canres/53/15/3569...[CANCER RESEARCH 53. 3569-3573, August 1. 1993] Evidence for in Situ Amplification](https://reader030.documents.pub/reader030/viewer/2022040722/5e31abba8887c7279f13b558/html5/thumbnails/2.jpg)
![Page 3: Evidence for in Situ Amplification of Cytotoxic T …cancerres.aacrjournals.org/content/canres/53/15/3569...[CANCER RESEARCH 53. 3569-3573, August 1. 1993] Evidence for in Situ Amplification](https://reader030.documents.pub/reader030/viewer/2022040722/5e31abba8887c7279f13b558/html5/thumbnails/3.jpg)
![Page 4: Evidence for in Situ Amplification of Cytotoxic T …cancerres.aacrjournals.org/content/canres/53/15/3569...[CANCER RESEARCH 53. 3569-3573, August 1. 1993] Evidence for in Situ Amplification](https://reader030.documents.pub/reader030/viewer/2022040722/5e31abba8887c7279f13b558/html5/thumbnails/4.jpg)
![Page 5: Evidence for in Situ Amplification of Cytotoxic T …cancerres.aacrjournals.org/content/canres/53/15/3569...[CANCER RESEARCH 53. 3569-3573, August 1. 1993] Evidence for in Situ Amplification](https://reader030.documents.pub/reader030/viewer/2022040722/5e31abba8887c7279f13b558/html5/thumbnails/5.jpg)
![Page 6: Evidence for in Situ Amplification of Cytotoxic T …cancerres.aacrjournals.org/content/canres/53/15/3569...[CANCER RESEARCH 53. 3569-3573, August 1. 1993] Evidence for in Situ Amplification](https://reader030.documents.pub/reader030/viewer/2022040722/5e31abba8887c7279f13b558/html5/thumbnails/6.jpg)
