The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

The Journal of Immunology, 2007, 179, 6651 -6662
Copyright © 2007 by The American Association of Immunologists, Inc.
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Related articles in The JI
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Shanker, A.
Right arrow Articles by Schmitt-Verhulst, A.-M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Shanker, A.
Right arrow Articles by Schmitt-Verhulst, A.-M.
Right arrowPubmed/NCBI databases
*Substance via MeSH

CD8 T Cell Help for Innate Antitumor Immunity1

Anil Shanker2,3,*,{dagger},{ddagger}, Grégory Verdeil3,*,{dagger},{ddagger}, Michel Buferne*,{dagger},{ddagger}, Else-Marit Inderberg-Suso*,{dagger},{ddagger}, Denis Puthier§, Florence Joly§, Catherine Nguyen§, Lee Leserman*,{dagger},{ddagger},¶, Nathalie Auphan-Anezin*,{dagger},{ddagger} and Anne-Marie Schmitt-Verhulst4,*,{dagger},{ddagger}

* Centre d’Immunologie de Marseille-Luminy, Faculté des Sciences de Luminy, Aix-Marseille Université, Marseille, France; {dagger} Institut National de la Santé et de la Recherche Médicale (INSERM), Unité 631, Marseille, France; {ddagger} Centre National de la Recherche Scientifique (CNRS), Unité Mixte de Recherche 6102, Marseille, France; § INSERM/Equipe de Recherche en Méthodologíes 206, Marseille, France; and CNRS, Groupement de Recherche 2352, Marseille, France


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Innate immunity is considered to initiate adaptive antitumor responses. We demonstrate that monoclonal CD8 T lymphocytes reactive to tumor Ag P1A on P815 mastocytoma cells provide essential "help" to NK cells for rejection of P1A-deficient tumors. RAG-deficient mice have normal NK cells but do not reject either tumor. Reconstitution of these mice with P1A-specific T cells conferred resistance to both P1A-expressing and -deficient tumor cells provided they were present at the same site. Elimination of Ag-negative tumor variants required both activated T and NK cells. Gene expression profiling of NK cells infiltrating P1A-positive tumors in mice with specific CD8 T cells demonstrated an activated effector phenotype. However, CD8 T cell help to NK cells appeared ineffective for P1A-negative variants separated from the P1A-positive tumor. Local tumor Ag-specific T cell-NK cell collaboration results in the elimination of tumor cells whether they express or not the T cell tumor Ag epitope, thus containing the emergence of tumor escape variants before metastasis.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Components of both natural and adaptive immunity contribute to reduce tumor incidence (1). In particular, recent studies showed that RAG-dependent lymphocytes and IFN-dependent signaling synergized in promoting resistance to tumor development (2). It is usually assumed that natural immunity constitutes the first line of defense of the organism and conditions the activation of adaptive immunity (3). A role for NK cells in the induction of CTL responses, which involves the capacity of NK cells to stimulate dendritic cells (DC)5 through an IFN-{gamma}-dependent cascade has recently been documented (4, 5). Furthermore, initial NK cell activation through MHC class Ilow tumor cells was found to induce T cell memory to the parental MHC-sufficient tumor (5, 6).

In antitumor immunity, NK cells have also been proposed to function as a "fail-safe" mechanism after effector T cell selection that leads to the outgrowth of poorly immunogenic tumors (1, 2) or allele-selective HLA loss variants (7). This is illustrated by the capacity of NK cells to lyse MHC class Ilow tumor cells (8) that may have escaped elimination by CD8 CTL effectors (7).

In addition to their activation via disengagement of MHC-recognizing inhibitory receptors (9, 10, 11), NK cells have also recently been shown to express stimulatory receptors, such as NKG2D (12). The NKG2D ligands (Rae-1 and H60 in the mouse, MHC class I-related chain A and B in the human) are poorly expressed on normal cells, but are up-regulated upon stress (13), DNA damage (14), viral infection, and on some tumor cells. Overexpression of the ligands by transfection of tumor cells leads to NK cell-dependent tumor rejection (15), and in some cases to T cell memory for the parental, nontransfected tumor (16). It is not fully understood, however, how NK cells become activated upon encounter with tumor cells expressing ligands for both inhibitory (high MHC class I) and stimulatory receptors.

Reactivity toward tumor Ags poses a particular problem for adaptive immunity, in that many of the identified tumor Ags correspond to self MHC-associated peptide fragments of nonmutated self proteins. Two major Ags of this type have been identified that derive 1) from tissue-specific differentiation Ags or 2) from a class of "cancer-germline" gene products that are encoded on the X chromosome and have restricted expression in gametogenic tissue and tumor cells, akin to the first identified MAGE gene (review in Ref. 17). In DBA/2 mice, the P1A Ag (18), characteristic of such "cancer-germline" genes, has been shown to be the major rejection Ag of mastocytoma P815 (19). Using mice expressing as a transgene the TCR (TCRP1A) from P1A35–43-specific CTL clone P1.5 of DBA/2 origin (20), we established a model where CD8 T cell reactivity was restricted to a single epitope (P815AB) of the natural tumor Ag P1A presented by H-2Ld on the P815 mastocytoma. Remarkably, expression of this monoclonal CD8 T cell population was capable of restricting growth of the P1A-expressing P815 tumor in RAG-deficient mice without outgrowth of P1A-negative tumor variants (21, 22, 23). Analysis of the effectors involved showed that NK cells contributed to the rejection of an MHC class I-expressing variant lacking the P815AB T cell epitope, provided both Ag-expressing and -deficient tumors were established at the same site. The data illustrate how NK cells recruited to a tumor site may become activated in the surrounding of T cells responding to a specific tumor Ag, and contribute to the containment of the emergence of variants escaping adaptive immunity.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
Mice

Mice heterozygous for the H-2Ld/P1A35–43-specific TCR transgene (TCRP1A) were kept on the RAG-1°/°B10.D2 background (TCRP1A RAG-1°/°). Genotyping for the TCRP1A transgene was performed as described previously (24) by PCR or by flow cytometry using anti-V{alpha}8.3 mAb. Mice deficient for the common cytokine receptor {gamma}-chain ({gamma}c°/°), received from J. P. Di Santo (Institut Pasteur, Paris, France) were crossed with RAG-1°/°B10.D2 mice to establish a RAG-1°/°{gamma}c°/°B10.D2 line. All these mice were bred in the animal facility of the Centre d’Immunologie de Marseille-Luminy (CIML; Marseille, France) and used 5–8 wk old. All animal experiments were in accordance with protocols approved by the French and European directives.

Tumor transplantation and monitoring

A total of 106 cells of mastocytoma P815 (H-2d) sublines, namely, P511 expressing P1A (P1A+P511) and P1.204 deficient in P1A but similar in other costimulatory and MHC expression (P1A–P1.204) (25), provided by B. Van den Eynde (Ludwig Institute for Cancer Research, Brussels, Belgium), were transplanted individually or mixed (106 each) s.c. between brachial and inguinal lymph nodes (LN). The tumor growth was monitored every fourth day by measuring the two perpendicular diameters of the solid tumors with vernier calipers. The survival of mice was recorded for at least 3 mo or until the tumor burden reached 400 mm2. Both the tumors showed a well-localized and regularly shaped form. In some experiments, tumor transplantation was performed with tumor cells isolated from the solid tumor mass excised after at least 10 days of in vivo growth in RAG-1°/°, or 40 days of in vivo growth in TCRP1A or T cell-transferred RAG-1°/° mice. Survival analysis was performed using GraphPad Prism (GraphPad Software) and the statistical difference between the survival curves was analyzed by log-rank test. Luciferase expressing P1A+ or P1A– tumor cells were obtained after transfection of P815 or P1.204 cells with vector pEGFPLUC (BD Biosciences/BD Clontech) by Lipofectamine 2000 (Invitrogen Life Technologies), followed by G418 selection. Established clones are referred to as P815-Luc or P1.204-Luc. A similar approach was applied to derive enhanced GFP-expressing P1.204 cells.

Bioluminescence imaging

Mice (RAG-1°/°B10.D2 or RAG-1°/°{gamma}c°/°B10.D2) were shaved on the flanks and were injected s.c. with 106 P815-Luc or 106 P1.204-Luc individually, or with a mixture of 106 P815 plus 106 P1.204-Luc on the same flank. Some mice received an i.v. injection of 2 x 106 TCRP1A CD8 T cells (from LN of TCRP1A RAG-1°/°B10.D2 mice) 5 days earlier. The growth of the luciferase-expressing cells was monitored by bioluminescence imaging. After i.p. luciferin (3 mg/mouse) injection, the mice were anesthetized in a chamber flushed with a mixture of isofluorane (4% in air) and placed in the NightOwl LB981 (Berthold Technologies) under continuous anaesthetization (1.5–3% isofluorane in O2). First, a black and white photographic image was acquired using a 100-ms exposure. Next, 10 min after luciferin injection, the luminescent image was acquired using a 2-min photon integration period with background substraction (pixel binning 8 x 8). Quantification was performed using Berthold Technologies software.

Cell preparation

Cells were prepared from LN according to the standard procedures. Tumor-infiltrating cells (TILs) were prepared from the solid tumor mechanically dispersed into a single-cell suspension and passed over Ficoll-Paque solution (Amersham Biosciences). The interface cells containing viable tumor cells and mononuclear cells were washed and used for analysis of TILs. For microarray analysis, cells from draining LNs (DLN) and TILs were labeled with anti-CD8 and -NK1.1 mAb (BD Pharmingen) and sorted using a FACSVantage (BD Biosciences).

Immunofluorescence staining

H-2Ld tetramers refolded in the presence of the P1A35–43 nonameric peptide LPYLGWLVF (Ld/P1A tetramer) on a scaffold of biotin-streptavidin-PE were used as described (24). Abs used for immunofluorescence staining were obtained from BD Pharmingen. For IFN-{gamma} intracellular staining, cells were restimulated for 4 h with 200 ng/ml ionomycin plus 10 ng/ml PMA in the presence of 10 µg/ml brefeldin A. Cells (0.5–1 x 106) were analyzed on a FACSCalibur cytofluorometer (BD Biosciences). Lamp-1 staining measuring exocytosis by NK cells was performed as described (26). Briefly, 3 x 105 NK TILs were cultured with or without 105 YAC cells for 4 h in medium containing Golgi Stop (BD Biosciences) and anti-Lamp1 (CD107a) mAb coupled to Cy5. After washing, cells were fixed and analyzed by cytofluorometry. Data was analyzed using CellQuest (BD Biosciences) or FlowJo (Tree Star) software.

CFSE staining

The number of T cell divisions was determined by flow cytometry using the intracytoplasmic stable CFSE dye that has been shown to exhibit sequential halving of intracellular fluorescence intensity at each division step (27). Purified TCRP1A CD8 T cells were incubated for 10 min at 37°C with 5 µM CFSE (Molecular Probes).

Adoptive T cell transfer

A total of 2 to 5 x 106 TCRP1A CD8+ T cells labeled with or without CFSE were injected i.v. in RAG-1°/°B10.D2 recipients at the indicated times before or after P1A+P511 or P1A–P1.204 transplantation. CFSE-labeled TCRP1A CD8 T cells were analyzed ex vivo from tumor DLNs or contralateral LNs and spleens on day 1, 2, 3, 6, or 25 after transfer.

NK depletion

TCRP1A or T cell-transferred RAG-1°/° mice were injected i.p. with 200 µg of mAb per injection of anti-NK1.1 (PK136) (28), or an equivalent amount of isotype-matched mouse IgG2a/{kappa} (Ti98anti-TCRBM3.3 or Désiré-1 anti-KB5.C20TCR produced in the laboratory), or nothing on days –1 and +1 of tumor transplantation.

Cytotoxicity assays

Cytolytic activity of TCRP1A CD8 T cells from tumor-bearing mice was assayed ex vivo by incubating FACS-sorted CD8+ cells from TILs with P1A+P511 or P1A–P1.204 tumor target cells labeled with 51Cr (New England Nuclear) for 5.5 h at 37°C. In some experiments, P1A+P511 tumor cells isolated from outgrowing solid tumor mass after at least 10 days of in vivo growth in RAG-1°/° or 40 days of in vivo growth in TCRP1A or T cell-transferred RAG-1°/° mice were used as targets for TCRP1A CD8 T cells stimulated in vitro with peptide P1A35–43 and 51Cr release was assayed after 5.5-h incubation at 37°C performed in the absence or presence of peptide P1A35–43.

Gene expression profiling

cDNA microarray analysis. The nylon microarrays used for this study were produced as described (29) using the same bacterial clones except that 289 of them were replaced by IMAGE clones (Deutsches Ressourcenzentrum für Genomforschung Resource Centre, Berlin, Germany) representing mostly cytokines, their receptors and CD8 T cell-expressed genes. The final clone set includes 8668 bacterial clones representing 7750 nonredundant Unigene clusters. Approximately 10% of the genes included in this clone set are represented by two or more different cDNA clones, providing internal controls to assess the reproducibility of gene expression measurements. RNA isolation and RNA amplification using 50 ng of total RNA as starting material were achieved as described (30). The amount of RNA was measured using RNA 6000 Pico Chip on 2100 Bioanalyzer (Agilent Technologies). Amplifications used a 5-µl reverse transcription (RT) reaction with 50 U of SuperScript II (Invitrogen Life Technologies), 4 µg of T4gp32 (USB) and 50 ng of (dT)-T7 primer (GCATTAGCGGCCGCGAAATTAATACGACTCACTATAGGGAGA(T)21V) in 1x first-strand buffer (Invitrogen Life Technologies) with a 1 h, 42°C incubation. Second-strand synthesis (SSS) was conducted in 20 U of DNA polymerase I, 1 U of Escherichia coli RNase H, and 5 U of E. coli DNA ligase in 1x second-strand buffer (Invitrogen Life Technologies) simply by adding 32.5 µl of an ice-cold SSS premix to the heat inactivated, ice-cold 5 µl of RT reaction and incubating at 15°C for 2 h. The double-stranded (ds) cDNA was polished by adding 5 U of T4 DNA polymerase and incubating for 15 min at 15°C. cDNA was purified by phenol-chloroform extraction followed by chromatography on a BioGel p-6 column (Bio-Rad) and was transcribed in 20 µl using the Ribomax Large Scale T7 KIT (Promega). For the second round of amplification, RT was in 10 µl with 0.5 µg of random hexamers, 16 µg of T4gp32 and incubation for 20 min at 37°C, 20 min at 42°C, 10 min at 50°C, and 10 min at 55°C. For SSS in the second round, 1 U of RNase H was first added to the heat-inactivated 10 µl of RT reaction followed by incubation at 37°C for 30 min, denaturation at 95°C for 2 min and snap cooling on ice. A total of 0.1 µg (dT)-T7 primer2 was then added to the chilled cDNA and the SSS reaction was primed by incubation for 10 min at 42°C followed by snap cooling on ice. Sixty-five microliters of ice-cold SSS premix (same as in round 1 except minus ligase) was then added and incubation and polishing were conducted as in round 1. Complex probes were prepared from 300 ng of amplified RNA as previously described (31) with [{gamma}-33P]dCTP labeling, using random hexamers. After hybridization and washes, arrays were exposed to phosphor imaging plates, which were then scanned with a Fuji BAS 5000 machine (25-µm resolution).

Data processing and analysis. After image acquisition, hybridization signals were quantified using the locally developed Bzscan software (32). Quantile normalization was applied to vector probe data (V) and complex probe data (C), to correct for global intensity and dispersion. Correction by the vector signal was made by calculating a C:V ratio before log transformation (base 2). Genes selected were significantly different with a p value <0.1 between the different conditions. Data were displayed using the TreeView program (Eisen Laboratory, University of California, Berkeley, CA).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
TCRP1A RAG-1°/° mice are resistant to P1A-expressing P511 tumor growth

We set up a model to restrict the analysis of interactions between innate immune components and monoclonal CD8 T cells in antitumor reactivity. When 106 P511 (P1A+) tumor cells were inoculated s.c. in RAG-1-deficient mice, mice succumbed to the tumor (or were sacrificed when the tumor reached a size of 400 mm2) after a median survival of 24 days (Fig. 1A). In TCRP1A RAG-1°/° mice which are RAG-1 deficient but express the TCRP1A transgene, the tumor failed to grow in most animals (81% of the mice) (Fig. 1A). All mice bearing the P1.204 P1A– variant succumbed to tumor after a median of 30–35 days, whether or not they expressed the TCRP1A transgene (Fig. 1A). Therefore, TCRP1A RAG-1°/° mice selectively reject P1A+ tumors.


Figure 1
View larger version (27K):
[in this window]
[in a new window]

  		FIGURE 1. Resistance of TCRP1A RAG-1°/° mice to P1A+P511 tumor growth contrasts with outgrowth of P1A Ag-loss variants in P1A+P511 tumor-bearing mice infused with TCRP1A T cells. A, 106 P1A+P511 or P1A–P1.204 tumor cells were s.c. injected in RAG-1°/° (black squares) and TCRP1A RAG-1°/° (red triangles) mice. Kaplan-Meier survival curves are shown. Survival is scored as days from transplantation until death or a tumor size of 400 mm2 within a period of 90 days. Numbers in parentheses depict median survival in days. Survival was significantly increased in TCRP1A RAG-1°/° mice bearing P1A+P511 as compared with wild-type (WT) RAG-1°/° or to any of the groups of mice bearing P1A–P1.204 (p < 0.001). Curves are representative of seven independent experiments with at least five mice in each group with similar patterns of tumor growth. B, RAG-1°/° mice inoculated s.c. with 106 P1A+P511 cells were i.v. (i.v.) infused (blue squares) or not (black triangles) with 3 x 106 TCRP1A CD8 T cells (see Materials and Methods) 7 days later. Tumor growth was caliper monitored and shown for each individual mouse in one representative experiment. C, RAG-1°/° mice inoculated s.c. with 106 P1A+P815-Luc cells were i.v. infused (+) or not (–) with 2 x 106 TCRP1A CD8 T cells 5 days later (indicated as day 0 (D0)) and tumor growth was monitored by bioluminescence (see Materials and Methods). D, Tumor cells excised from three RAG-1°/° mice infused with TCRP1A CD8 T cells shown in C were used as target cells (without (green lines) or with (red lines) addition of 10–6 M peptide P1A35–43) in a 4-h CTL assay with TCRP1A CD8 T cell effectors (see Materials and Methods) (right panel). The panel on the left side shows the same assay using as targets tumor cells excised from a noninfused RAG-1°/° mouse without addition of peptide during the CTL assay (green line). Data are representative of three independent experiments.

 
Outgrowth of Ag-loss variants in tumor bearing RAG-1°/° mice infused with TCRP1A CD8 T cells

Next, we tested whether TCRP1A CD8 T cells could induce rejection of an established P1A+ tumor. Notably, when 3 x 106 TCRP1A CD8 T cells were transferred to RAG-1°/° hosts 1 wk after tumor inoculation, the P1A+ tumor growth was controlled for 2–3 wk, but resumed thereafter, resulting in tumor burden in most animals (Fig. 1B). In the same setting, P1A–P1.204 cells failed to be rejected by such TCRP1A T cell-reconstituted RAG-1°/° mice (Fig. 1B) and mice succumbed earlier (data not shown). This was also evident when tumor growth was monitored by noninvasive bioluminescence measurement of luciferase-expressing P1A+P815 cells in a similar protocol. The tumor appeared to be completely eradicated by day 12 following infusion of TCRP1A CD8 T cells, with re-emergence of the luciferase signal later on (Fig. 1C). Further, these re-emerging tumors were insensitive to lysis by TCRP1A CD8 CTL effectors unless exogenous peptide P1A35–43 was present in the assay (Fig. 1D). P815-Luc or P511 cells recovered from s.c. injected RAG-1°/° mice were always found to remain sensitive to TCRP1A CD8 CTL, even without addition of exogenous P1A peptide (Fig. 1D and results not shown). These data indicate that the tumor escape variants had lost expression of the P1A35–43 epitope, but had retained H-2Ld MHC expression. Consistently, all outgrowing tumors tested had also lost the capacity to be rejected by a TCRP1A RAG-1°/° host (results not shown, available at www.ciml.univ-mrs.fr/Lab/Schmitt-Verhulst/Docs/JI-07_Shanker.pdf).

These results suggested that Ag-negative tumor variants arose in the presence of immune selection from the Ag-specific TCRP1A T cells. Whether these escape variants arose during or before the infused CD8 T cells became active was not clear. We wanted to know whether the escape of tumor variants correlated with the activation of potential effector cell populations in the tumor-DLN or tumor infiltrate. We thus examined the fate of TCRP1A CD8 T cells upon their transfer in tumor-bearing RAG-1°/° hosts and their activation status in tumor DLN as well as tumor mass.

Tumor Ag-specific migration, activation, and gene expression profile of TCRP1A CD8 T cells transferred in P1A+ tumor-bearing mice

TCRP1A CD8 T cells isolated from TCRP1A RAG-1°/° mice, labeled with CFSE, and transferred to P1A+P511 tumor-bearing RAG-1°/° mice, did not divide in the DLNs by day 1, whereas 44% of the T cells by day 2 and 80% by day 3 had gone through five to seven divisions (Fig. 2A). In comparison, the TCRP1A CD8 T cells in the LNs draining the P1A– tumor were mostly undivided after 2 days, and only a small fraction (24%) went through one or two divisions after 3 days (Fig. 2A). The latter division appeared to be driven by homeostatic proliferation in RAG-1°/° hosts, because it was also observed in the absence of tumor (Fig. 2A). On day 2 posttransfer, the TCRP1A CD8 T cells in the LNs draining the P1A+ tumor expressed the activation marker CD69, and a fraction of the cells had acquired a CD44high and CD62Llow phenotype (Fig. 2B), whereas none of the activation markers was detected for the TCRP1A CD8 T cells in the LN draining the P1A– tumor (Fig. 2B). Upon in vitro restimulation, these CD8 T cells were also capable of producing IFN-{gamma} (Fig. 2B). By day 4, we observed maximal expansion and accumulation of TCRP1A CD8 T cells in the LN draining the P1A+ tumor (Fig. 3A). A significant number of these P1A-specific CD8 T lymphocytes infiltrated the P1A+ tumor mass (Fig. 3, B and C) and had undergone a number of cell divisions with acquisition of an activated phenotype (Fig. 3C, right panel). In contrast, infiltration of TCRP1A CD8 T cells was negligible in the P1A-negative tumor (Fig. 3, B and C). NK cells were also observed in the tumor infiltrates (Fig. 3C, left panel).


Figure 2
View larger version (33K):
[in this window]
[in a new window]

  		FIGURE 2. Tumor Ag-induced proliferation and activation of transferred TCRP1A CD8 T cells. CFSE-labeled (3 x 106) TCRP1A CD8 T cells were transferred i.v. in RAG-1°/° mice inoculated s.c. with 106 P1A+P511 or P1A–P1.204 tumor cells 5 days earlier. Cells from tumor-draining brachial plus inguinal LN were analyzed by cytofluorometry. A, CFSE vs Ld/P1A-tet dot plots gated on CD8 T cells from days 1 to 3 after T cell transfer are shown. B, Surface staining with anti-CD69, anti-CD44, and anti-CD62L and intracellular staining with anti-IFN-{gamma} is shown within Ld/P1A-tet+ CD8 T cells at 48 h post-T cell transfer. The numbers within the histograms depict the percentage of positive cells. One experiment representative of four performed independently with similar results is shown.

 

Figure 3
View larger version (42K):
[in this window]
[in a new window]

  		FIGURE 3. Kinetics of tumor infiltration by NK cells and by TCRP1A CD8 T cells. CFSE-labeled 3 x 106 TCRP1A CD8 T cells were transferred in RAG-1°/°B10.D2 mice inoculated with P1A+P511 or P1A–P1.204 tumor cells s.c. 7 days earlier. LN cells and TILs were isolated at different time points before (day 0) and after T cell transfer. A, Number of TCRP1A CD8 T cells in tumor DLN. B, Ratio of CD8 to NK cells present in TILs calculated from the percentage of each population detected by cytofluorometry in isolated TILs (see Materials and Methods). C, Cytofluorimetric analysis shows CD8 and NK cell populations at days 3 and 4 after T cell transfer among P1A+P511 and P1A–P1.204 TILs. CD25 surface expression and CFSE profile are shown for the CD8 T cells present in P1A+P511 TILs on day 4.

 
To obtain a comprehensive view of the activation program of the infused CD8 T cells, we sorted on day 4 both the TCRP1A CD8 T cells expanding in the tumor DLNs and those infiltrating the P1A+ tumor and profiled their gene expression by microarray analysis. As a control, nonactivated TCRP1A CD8 T cells were collected from LN of P1A– tumor-bearing mice. Two clusters of genes differentially expressed between TCRP1A CD8 T cells present in the LN or in the tumor are presented in Fig. 4A (see also Table I). These include: in Fig. 4Aa, genes down-regulated in CD8 T cells infiltrating the tumor such as Sell (CD62L), consistent with the role of the corresponding surface receptor in the homing from the LN to the peripheral tissue; in Fig. 4Ab, a cluster of transcripts expressed at a much higher level in CD8 TILs as compared with CD8 TCRP1A cells present in the LN. This cluster includes genes encoding molecules involved in cytolytic effector function (GzmB, Prf1, Tnfrsf 4-9-18, Ifng, CCL3–4, Icos). This pattern of gene expression was similar to that observed for optimally in vitro-stimulated CD8 T effector cells (33). Intracellular staining confirmed the higher expression of GzmB (Fig. 4B) in the P1A+P511 as compared with the P1A–P1.204 CD8 TILs. Furthermore, CD8 TCRP1A cells sorted from P1A+P511 TILs were highly cytolytic against Ld/P1A expressing targets (Fig. 4F). These CD8 T cells also expressed higher levels of Tnfrsf9 (4.1BB) (Fig. 4C). We further observed a high level of IFN-{gamma} mRNA in CD8 TILs recovered from the P1A+P511 tumor (Fig. 4Ab) suggesting that the recognition of the cognate Ag in the tumor gave rise to restimulation of IFN-{gamma} secretion. Indeed, we were able to detect IFN-{gamma} and IL-2 secretion upon a brief 4-h in vitro stimulation of the CD8 TILs recovered from the P1A+P511 tumor mass and not from the P1A–P1.204 counterpart (Fig. 4E). Moreover, when measuring the expression level of IFN-{gamma}R1 with the 2E2 mAb (the binding of which is not inhibited by the presence of IFN-{gamma} on its receptor), we observed a higher expression of this chain on the CD8 TCRP1A cells present in the P1A+P511 TILs as compared with the corresponding DLN (Fig. 4D, right panel). When the same measurement was performed with the GR20 mAb, for which the binding to the IFN-{gamma}R1 is blocked in the presence of IFN-{gamma} (34), it appeared that a fraction of the TCRP1A CD8 T cells infiltrating the P1A+P511 tumor had IFN-{gamma} bound to their receptors (Fig. 4D, left panel). Altogether, the kinetic analysis showed that the TCRP1A CD8 T cells became activated in the LN draining the P1A+P511 tumor, and then migrated to the tumor site where they were secondarily stimulated, presumably by the tumor itself.


Figure 4
View larger version (44K):
[in this window]
[in a new window]

  		FIGURE 4. Activation profile of TCRP1A CD8 T cell infiltrating the P1A+ tumor 4 days after transfer. TCRP1A CD8 T cells (3 x 106) were transferred in RAG-1°/°B10.D2 mice inoculated with P1A+P511 or P1A–P1.204 tumor cells s.c. 7 days earlier. LN cells and TILs were isolated at different time points after T cell transfer. A, Gene profiling on purified populations of CD8 T cells (see Materials and Methods) from day 4 P1A+P511 tumor DLN or TILs and from P1A–P1.204 tumor DLN (as too few CD8 T cells were found in the P1A–P1.204 tumor infiltrate). The names of known genes are reported. The asterisks (*) indicate either expressed sequence tags (EST) or Riken cDNA clones or unknown clones (nonreported in Unigene). The results are shown as relative expression levels obtained after normalization and are represented with a color scale indicated at the bottom. The t test values (see Materials and Methods) for comparison between groups are reported in Table I. B and C, Cytofluorimetric analysis of P1A–P1.204 (left side) and P1A+P511 (right side) TILs on day 6 after T cell transfer shows intracellular staining for granzyme B (B) and surface staining for 4-1BB (C, lower graphs). Surface staining for 4-1BB is shown in tumor DLN as a comparison (C, upper graphs). D, Cell surface staining with IFN-{gamma}R1-specific mAb GR20 (left side) or 2E2 (right side) on day 5 after T cell transfer on CD8 T cells in P1A–P1.204 (blue) or P1A+P511 (red) tumor DLN or P1A+P511 TILs (green). Binding of mAb GR20 to the IFN-{gamma}R1 is inhibited when IFN-{gamma} occupies its receptor, whereas binding of mAb 2E2 is not affected by the presence of IFN-{gamma} (see text in Results). E, Cytofluorimetric analysis of P1A–P1.204 (left side) and P1A+P511 (right side) TILs on day 5 after T cell transfer shows intracellular staining for IL-2 and IFN-{gamma} (see Materials and Methods). F, TCRP1A CD8 T cells (3 x 106) were transferred in RAG-1°/°B10.D2 mice inoculated with a mixture P1A+P511 and P1A–P1.204 tumor cells s.c. 7 days earlier. TILs were isolated 4 days after T cell transfer. CD8 TILs were purified by cell sorting and their cytolytic activity was directly assessed (green symbols) against P1.204 targets in the absence (open symbols) or presence (full symbols) of peptide P1A35–43 (10–6 M). As a positive control, in vitro-derived TCRP1A CD8 T effector cells (black curve) were also used. Two independent experiments are shown.

 

View this table:
[in this window]
[in a new window]

  		Table I. Differential gene expression in CD8 T cells recovered from P511 TILs vs P511 DLN

 
P1A-specific CD8 T cells are necessary for the NK cell-dependent rejection of the P1A-negative tumors

We next investigated the potential role of the observed tumor-infiltrating NK cells, and evaluated whether the activation of Ag-specific T cells might lead to a bystander effect with elimination of P1A– as well as P1A+ tumors. In a setting in which P1A+P511 and P1A–P1.204 tumors were transplanted on different sites, respectively, on the right and left flanks of the same mouse, TCRP1A RAG-1°/° mice were capable of rejecting P1A+ tumors, but did not reject P1A– tumors (Fig. 5A). Therefore, the resistance afforded in the TCRP1A RAG-1°/° mice compared with the RAG-1°/° host was selective for the P1A+ tumor (Fig. 5A).


Figure 5
View larger version (25K):
[in this window]
[in a new window]

  		FIGURE 5. Both NK cells and P1A-specific CD8 T cells are required for resistance to P1A-negative tumors present in a mixed inoculum of P1A-positive and -negative tumor cells. s.c. growth of 106 P1A+P511 and P1A–P1.204 tumor cells on the right and the left flanks, respectively, or a mixed sample of both tumor cells (106 each) on the right flank with or without treatments (i.p.) of anti-NK1.1 (PK136: 200 µg mAb per injection) in RAG-1°/°B10.D2 or TCRP1A RAG-1°/°B10.D2 mice was monitored. (A) Kinetics of tumor growth shown for each individual mouse in one representative experiment and (B) Kaplan-Meier tumor-free survival curves from a pool of experiments (total 20 mice), including those shown in A. Survival is scored as days from transplantation until death or a tumor size of 400 mm2 within a period of 3 mo. Numbers in parentheses depict median survival in days. Survival was significantly increased in TCRP1A RAG-1°/°B10.D2 mice transplanted with mixture of P1A+P511 and P1A–P1.204 (16 of 20 showed tumor-free survival) as compared with other test groups (p < 0.0001) (all died of tumor). Curves are representative of three independent experiments with at least five mice in each group with similar patterns of tumor growth. C, Mice (RAG-1°/°B10.D2 or RAG-1°/°{gamma}c°/°B10.D2) were injected s.c. with a mixture of 106 P1A+P511 plus 106 P1A–P1.204-Luc (mix, left flank). All mice had received an i.v. injection of 2 x 106 TCRP1A CD8 T cells (purified from LN of TCRP1A RAG-1°/°B10.D2 mice) 5 days earlier (D-5). The growth of the P1A–P1.204-Luc cells was monitored by bioluminescence imaging (see Materials and Methods). Luminescence is expressed in relative luminescence units (RLU) as a function of time after tumor inoculation. Experimental groups included five and four mice for RAG-1°/° B10.D2 and RAG-1°/°{gamma}c°/° B10.D2, respectively. Full square symbols represent RLU for individual mice. They are all superimposed on the left panel. On the right panel, the bars centered on the full diamonds indicate the mean ± SEM.

 
However, when P511 and P1.204 tumors were transplanted on the same site (mix condition), no tumor growth was detected in 80% of the TCRP1A RAG-1°/° hosts for up to 72 days, whereas all RAG-1°/° mice developed tumors (Fig. 5, A and B). In some instances, apoptosis of cancer cells can cause sufficient release of Ag to sensitize tumor stroma for killing by CTLs (35). We therefore evaluated in the "mix" condition, whether the elimination of the coinjected P1A–P1.204 tumor relied on the transfer of P1A Ag from the P1A+P511 tumor. For this purpose, P1.204-GFP tumor cells were coinjected with P511 cells at a 1:1 ratio on the same site of RAG-1°/° mice. After a week of tumor growth, mice received 3 x 106 TCRP1A CD8 T cells. Four days later, cell suspensions were prepared from the tumors of sacrificed mice and P1.204-GFP cells were isolated by flow sorting and used as targets in a CTL assay against in vitro-activated TCRP1A CD8 T effector cells. Data in Fig. 6A show that the sorted P1.204-GFP tumors were still resistant to TCRP1A CTL killing unless P1A35–43 peptide was exogenously added during the CTL assay. Thus, after their in vivo growth in the vicinity of P1A+P511 tumors under attack by TCRP1A CTLs, the P1A– tumor cells remained insensitive to the CTL effectors. Additionally, cocultivating apoptotic P1A+ cells with P1A–P1.204 cells failed to sensitize the latter for killing by TCRP1A CTL (results not shown). These results exclude Ag loading from the neighboring P1A+ cells to P1A– tumor cells as a mechanism for the bystander elimination of Ag-negative tumors.


Figure 6
View larger version (24K):
[in this window]
[in a new window]

  		FIGURE 6. P1A–P1.204 tumor cells located in the vicinity of P1A+P511 tumor cells remain insensitive to the killing by TCRP1A CD8 effector cells or by NK cells. RAG-1°/°B10.D2 mice were s.c. inoculated with a mixture P1A+P511 and P1A–P1.204-GFP tumor cells (106 each). After 7 days, mice were injected i.v. (+CD8; diamonds) or not (–CD8; triangles) with TCRP1A CD8 T cells (3 x 106). Tumors were recovered 4 days after T cell transfer and P1A–P1.204-GFP cells were purified by cell sorting. In A, sorted P1A–P1.204-GFP cells were incubated in the absence (open symbols) or presence (black symbols) of peptide P1A35–43 (10–6 M) and directly used as targets in a CTL assay against TCRP1A CD8 T effector cells. As a comparison, P1A–P1.204 tumors cultured in vitro (squares) were also included in the test. In B, P1A–P1.204-GFP cells were directly used as targets in a cytolytic assay against poly I:C-activated NK cells. As a comparison, NK-sensitive RMA.S tumors cultured in vitro (open circles) were also included in the test. Data are representative of two independent experiments.

 
We next evaluated whether the mechanism of Ag-independent tumor rejection may involve NK cells, which were found to infiltrate the tumors (Fig. 3C). To this end, a group of TCRP1A RAG-1°/° mice inoculated with the two tumors at the same site (mix) were pretreated for NK cell depletion (mix plus anti-NK1.1), as previously described (36). Clearly, NK cell depletion led to tumor growth in all of the mice (Fig. 5, A and B).

These results suggested that the tumor Ag-specific activation of TCRP1A CD8 T cells possibly led to a local NK cell activation/recruitment that contributed to the elimination of the tumor whether or not it expressed the T cell-recognized Ag. Thus, once activated in proximity of Ag-bearing tumor, the NK cells may have acquired competence to eliminate tumor cells, irrespective of their expression of the T cell Ag epitope.

To further confirm this possibility, RAG-1°/° or RAG-1°/°{gamma}c°/° mice were reconstituted with TCRP1A CD8 T cells and the growth of P1A–P1.204 cells expressing luciferase was followed by bioluminescence when inoculated alone or in combination with P1A+P511 cells (not expressing luciferase). The restriction of the growth of P1A–P1.204 cells present in the mixture with P511 cells was only observed in the RAG-1°/°, but not in the RAG-1°/°{gamma}c°/° mice (Fig. 5C) in line with the suggested requirement for NK cells.

TCRP1A CD8 T cell-dependent activation of NK cells recruited to P1A+ tumors

We next analyzed the activation status of NK cells present in TILs. NK cells could be detected in the tumors by day 7 after their s.c. inoculation, before T cell transfer (day 0: Fig. 3C). Comparison of activation markers on NK cells present in P1A+P511 and P1A–P1.204 TILs revealed that only the former expressed an increased level of surface 4.1BB (Fig. 7B), a marker previously associated with NK cell activation (37). To obtain a comprehensive analysis of the changes induced in tumor-infiltrating NK cells upon recruitment of activated CD8 T cells, NK cells were purified from P511 and P1.204 TILs, as well as from LNs draining the P511 or the P1.204 tumor 4 days after TCRP1A T cell transfer. Gene profiling (see Materials and Methods) identified clusters of genes (Fig. 7A and Table II) with weak up-regulation in P511-DLN, but much stronger expression in P511 TILs (gzmb, gzma, Prf1, Tnfrsf9 or 4.1BB, Klra7, Ifngr, CCR5, Icos). Up-regulated expression of granzyme B in NK cells found in the P1A+P511 TILs vs their P1A–P1.204 counterparts was further confirmed by intracellular staining (Fig. 7C). Moreover, the majority of the NK cells present in the P511 tumor had their IFN-{gamma}R1 chain occupied by IFN-{gamma}, as illustrated by the decreased surface staining measured with the anti-IFN-{gamma}R1 (GR20) mAb (Fig. 7D). No increase in IFN-{gamma} mRNA was observed (Fig. 7A), which correlated with the absence of IFN-{gamma} production as detected by intracellular staining (results not shown). Our results also revealed the coordinated expression of the CCL3 and CCL4 chemokine genes by CD8 TCRP1A cells (Fig. 4Ab) and of the corresponding chemokine receptor CCR5 gene on the NK cells (Fig. 7A). Unlike the critical role played by CCL3 in promoting NK cell recruitment and inflammation in anti-murine CMV responses (38), we found no evidence, using CCL3°/° mice, for such an exclusive role for CCL3 in either NK cell recruitment or activation at the tumor site in the present system (results not shown). Accordingly, we found that TCRP1A RAG-1°/°CCL3°/° mice were not affected in the capacity to reject P1A– tumors in the "bystander" protocol established in Fig. 5 (results not shown).


Figure 7
View larger version (50K):
[in this window]
[in a new window]

  		FIGURE 7. TCRP1A CD8 T cell-dependent activation of NK cells recruited to the P1A+P511, as compared with the P1A–P1.204 tumor. A, Gene profiling on purified populations of NK cells from day 4 P1A+P511 and P1A–P1.204 TILs and DLN. Data are represented as in Fig. 4A with t test values in Table II. B and C, Cytofluorimetric analysis shows NK cells in P1A–P1.204 (left side) and P1A+P511 (right side) TILs at day 4 (B) and 6 (C) after T cell transfer. Cell surface expression of 4.1BB (B) and intracellular staining of granzyme B (C) are shown as a function of NK1.1 surface expression. D, Cell surface staining with IFN-{gamma}R1-specific mAb GR20 or 2E2 (see legend to Fig. 4D) on P1A–P1.204 (blue) or P1A+P511 (red) tumor DLN (left side) or TILs (right side) shows the presence of IFN-{gamma} on the IFN-{gamma}R1 of NK cells in P1A+P511 TILs. E, Exocytosis measured by Lamp-1 staining (see Materials and Methods) on NK1.1+ cells present among P1A+P511 (red) or P1A–P1.204 (blue) TILs in RAG-1°/° B10.D2 mice which had (full bars) or not (empty bars) been transferred with 3 x 106 TCRP1A CD8 T cells 4 days earlier (7 days after the s.c. inoculation of the tumor cells, separately on each flank of the mice). Results are expressed as the percentage of NK cells positive for Lamp-1 staining directly ex vivo (upper graphs) or after incubation with YAC tumor cells for 4 h (lower graphs). The ratios of NK cells positive for Lamp-1 staining from T cell-infused vs noninfused mice are indicated as mean ± SD.

 

View this table:
[in this window]
[in a new window]

  		Table II. Differential gene expression in NK cells recovered from P511 TILs vs P511 DLN

 
Altogether, the gene expression program of NK cells present in TILs is influenced by the local recruitment of activated CD8 T cells and reflects an effector-type (gzmB+, Prf1+, 4.1BBhigh) as well as tissue-migratory (CCR5high) phenotype (Fig. 7A). We thus propose that, within the tumor microenvironment, the expression of the relevant P1A Ag-reactivated CD8 TCRP1A effectors and triggered their secretion of cytokines such as IL-2, IFN-{gamma} (Fig. 4A), and TNF-{alpha} (results not shown) which act both on CD8 TCRP1A and on neighboring NK cells.

The requirement for an Ag-dependent reactivation of the CD8 T cells at the tumor site for their promotion of NK cell activation was further confirmed by experiments comparing NK cells present in P1A–P1.204 and in P1A+P511 TILs from the same mouse. TCRP1A CD8 T cells activated in the P1A+P511 DLN migrated to both P1A+P511 and P1A–P1.204 tumors in such mice, albeit more CD8 T cells accumulated in the P1A+ tumor (ratio of the percentage of CD8/the percentage of NK was 4.9 in P1A+P511 DLN vs 0.37 in P1A–P1.204 DLN) and CD25 surface expression indicates their sustained activation state only in the P1A+ tumor (CD25 mean fluorescence intensity was 81.7 on CD8 T cells present in P1A+P511 TILs vs 5.1 in P1A–P1.204 TILs). Furthermore, only NK cells present in the P1A+ tumor showed an activated (GzmB+) phenotype (data not shown, available at www.ciml.univ-mrs.fr/Lab/Schmitt- Verhulst/Docs/JI-07_Shanker.pdf). Evidence for in situ exocytosis by NK cells, as shown by detection of Lamp-1 ex vivo or after reactivation in the presence of YAC tumor cells, was also selectively increased in P1A-expressing tumors in mice bearing both P1A+P511 and P1A– tumors (Fig. 7E). In aggregate, our results show that tumor Ag-specific CD8 T cells modulate the activation status of tumor-infiltrating NK cells toward a killer effector phenotype.

We further addressed the question of whether the P1A– tumor cell could become sensitive to NK cell killing in a manner depending on the microenvironment in which it is rejected. Thus, P1A–P1.204-GFP cells were sorted from a mixed (P1A+/P1A–) tumor 4 days after infusion of TCRP1A CD8 T cells as for data presented in Fig. 6A and were used as target cells in a 5-h killing assay with poly I:C-induced splenic NK cells (Fig. 6B). Although no killing of the P1.204 cells could be detected in this in vitro assay, we cannot exclude that they may become transiently sensitive to activated NK cells in vivo. This could occur through changes in the expression of cell surface molecules that may be lost during manipulations required to dissociate the tumors and perform the FACS sorting. We also failed to detect any up-regulation of the expression of the NKG2D ligand Rae-1 on those sorted P1.204 cells (results not shown). Defining the exact mechanism by which the P1A– tumor is rejected in vivo will thus require further investigation.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
Arguments for the existence of tumor immune surveillance stem from the increased incidence of spontaneous carcinomas in mice deficient in components of both natural and adaptive immunity (for review, Refs. 39 and 40), as well as from the evidence that tumor Ag-specific immune effectors are found in cancer patients (7, 41). Natural and adaptive immune effectors play complementary roles in which the innate components are considered the vigils that awaken the adaptive components (5, 42) to initiate their program of Ag-driven clonal expansion, differentiation, migration, and establishment of Ag-specific memory. In this study, we describe a situation in which tumor Ag-specific CD8 T cells can activate dormant NK cells present in a tumor. The use of monoclonal CD8 T cells specific for a natural tumor Ag permitted us to address the question of tumor development/escape in the face of a defined effector population. Monoclonal adoptive T cell immunotherapy has previously been shown to result either in restriction of tumor growth (43, 44) or in immune escape by Ag loss or drift (22, 45, 46, 47).

We describe a model in which monoclonal naive TCRP1A CD8 T cells specific for Ag P815AB can lead to marked reduction of s.c. inoculated mastocytoma P815 even if the T cells are transferred into a host bearing an established tumor. Nevertheless, P815 variants having lost the expression of P815AB epitope develop at late time points (Fig. 1), consistent with previous reports (22, 47). The selective long-lasting tumor growth containment observed only in protocols in which monoclonal T cells are present before tumor inoculation could therefore depend either on total elimination of P1A-expressing tumor cells or on a local bystander effect operating on P1A-negative tumors before tumor metastasis, which readily occurs for this tumor (Ref. 47 and results not shown).

The latter possibility is supported by the observation that RAG-1°/° TCRP1A-transgenic mice could reject P1A– tumor cells provided they were present in a mixed inoculum with P1A+ cells (Fig. 5). Rejection of the P1A– tumor in a mixed inoculum was also observed in RAG-1°/° mice infused with TCRP1A CD8 T cells before tumor inoculation (Fig. 5C and data not shown). The bystander effect was thus active on the mixed tumor inoculum before its establishment as a solid tumor. It is therefore distinct from previously described bystander effects mediated by immune attack on tumor stroma (48, 49). Furthermore, no Ag transfer between P1A+ and P1A– tumor cells could be detected on the Ag-negative tumor cells retrieved from mixed inocula (Fig. 6A) and attempts at detecting Ld/P1A complexes on macrophages incubated with apoptotic or necrotic P1A+ tumors were also negative (N. Auphan-Anezin, results not shown).

The observation that pretreatment of the host with depleting anti-NK1.1 mAb prevented the bystander effect suggested that NK cells contribute to the collateral elimination of the P1A– tumor cells. In this model, NK cells represent a prominent population in secondary lymphoid organs. The involvement of NK cells was further suggested by the absence of bystander elimination of P1A– tumors in RAG- and {gamma}c-double deficient mice infused with TCRP1A CD8 T cells before tumor inoculation (Fig. 5C). Additionally, no NKT cells are present in our experimental system and TCRP1A CD8 T cells fail to express the NK1.1 marker (results not shown). The possibility that minor subpopulations of DC cells expressing NK cell markers (50, 51) might be implicated in the antitumor response will require further analysis. The model we have studied uses monoclonal CD8 T cells in RAG-deficient hosts, permitting us to exclude any involvement of CD4 Th or T regulatory populations.

We did not address the question of whether a particular innate immune component (such as NK cells) might be required to activate the TCRP1A CD8 T cells. The fact that RAG-1°/°{gamma}c°/° double deficient mice reconstituted with purified TCRP1A CD8 T cells controlled the initial growth of P511/P815 tumor cells (data not shown; available at www.ciml.univ-mrs.fr/Lab/Schmitt-Verhulst/Docs/JI-07_Shanker.pdf) suggests that, at least in this particular setting, initial CD8 T cell activation was independent of NK cells. It does not rule out, however, a possible role for NK cells or other innate components in the initial activation of CD8 T cells in other settings.

An interesting aspect of this study was the observation that established tumors, whether P1A+ or P1A–, were infiltrated by NK cells which are unable to contain growth of either tumor in RAG-deficient mice. Upon adoptive transfer of monoclonal P1A-reactive CD8 T cells, the infused T cells were first activated in tumor DLN and next migrated to and transiently accumulated in the tumor only if it expressed the P1A Ag. This probably resulted from activation of the CD8 T cells in the LN that were infiltrated by P1A+P815/P511 cells capable of direct CD8 T cell activation (A. Shanker and G. Verdeil, unpublished results). The proximity of the activated CD8 T cells appeared to affect the NK cells present in P1A+ tumors. This was most clearly illustrated by comparing gene expression profiles in NK cell populations present in P1A+ and P1A– tumors (Fig. 7A). Thus, transcripts encoding molecules involved in cytolytic activity (gzmb, gzma, Prf1) increased only in NK cells found in P1A+ TILs. It is not clear whether the increase in expression of these genes in the P1A+ tumor as compared with the DLN occurred locally in the NK population that migrated before CD8 T cell infusion or whether it resulted from the selective migration of the most activated NK cells from the LN. Cotransfer of fluorescently labeled NK cells with P1A-specific CD8 T cells in tumor-bearing mice may help resolve this question.

Icos and Tnfrsf9 (4.1BB) were among activation markers that followed a similar pattern of increased expression from LN to P1A+ tumor-located NK cells. The latter is in agreement with our detection of increased surface expression of the 4.1BB molecule on P1A+ tumor-infiltrating NK cells (Fig. 7B). The increased expression of the CCR5 gene would also contribute to responsiveness to chemokines such as CCL3 or CCL4 that are secreted by Ag-activated CD8 T cells (Ref. 33 , results not shown) and to which NK cells are attracted (52). The combination of IL-2 and CCL3 has been described to boost antineoplastic reactions in a CD8 T and NK cell-dependent fashion in a leukemia/lymphoma vaccine (53).

It is notable that up-regulation of the transcript for the IFN-{gamma}R was also observed in P1A+ tumor-infiltrating NK cells (although IFN-{gamma} transcripts remained low as compared with the level expressed in the CD8 T cells; results not shown). It is thus possible that IFN-{gamma} secreted by Ag-triggered CD8 T cells might contribute to the local activation of NK cells. Indeed, IFN-{gamma} has been shown to up-regulate expression of TNF superfamily members on subsets of NK cells (54). CD8 T cell-secreted IL-2, TNF-{alpha}, or direct T/NK cell contact could also contribute to NK cell activation (55, 56). Activated DCs have been shown to trigger NK cells (57, 58). DC, in turn may mature in response to CD8 T cell-secreted cytokines such as IFN-{gamma} and thereby indirectly activate NK cells. Finally, tumor cell death that results from CD8 T cell cytolytic activity (59) could contribute to NK cell activation, directly or via DC activation, through engagement of NK- or DC-expressed TLR by proteins or DNA released by dead cells (60). So far, we have been unable to pinpoint a unique mediator critical for the local NK cell activation (example, CCL3 or IFN-{gamma}), other than the activation of CD8 T cells themselves, suggesting that it may depend on a complex microenvironment.

NK cells were initially described as natural effectors, in contrast to components of adaptive immunity that required an Ag-induced differentiation and clonal expansion process to become effectors. Nevertheless, it has become increasingly clear that NK cell subpopulations exist that require distinct activation cues (55, 61), not only for their direct effector function, but also for the establishment of transcriptional programs (62). For instance, P815/P511 cells are resistant to lysis by splenic or blood NK cells. They also appear to be resistant to TRAIL-dependent killing (63), but are sensitive to granzyme-dependent killing by hepatic NK cells (64). This superior cytotoxic capacity of hepatic as compared with blood NK cells has been correlated with the differential expression of a set of genes including those encoding the chemokine receptor CCR5, surface receptors such as CD25, NKG2D, CD94, and cytolytic granule components (65). It has been suggested that the apparent difference in the activation states of NK subpopulations may result from their microenvironment. In particular, liver Kupffer cells and sinusoidal endothelial cells secrete chemokines (CCL5, CCL3), ligands for CCR5 that contribute to the cytolytic potential of NK cells (66). Similarly, the microenvironment of the tumor infiltrated by tumor Ag-reactive CD8 T cells provides various cytokines (IL-2, IFN-{gamma}) and chemokines (CCL3, CCL4) capable of activating NK cells. In addition, the complex tumor microenvironment in which CD8 T cells kill Ag-expressing tumor cells and release cytokines may induce a type of "stress" response in neighboring Ag-loss variant tumor cells, leading to their expression of ligands of NK cell receptors such as NKG2D (14). Whether the activated NK cells in turn contribute to further recruitment and activation of other anti-tumor components remains to be evaluated.

CD8 T cells activated in mice bearing P1A+ and P1A– tumors at separate sites fail to induce the rejection of the P1A– tumors although they acquired tissue migratory properties. This suggests that local Ag activation of the CD8 T cells at the site of the tumor may be required for the expression of their NK-activating properties. This is consistent with the activation profile of CD8 T lymphocytes recruited to either P1A+ or P1A– tumors (Fig. 4, B and C).

Altogether, our results illustrate a mechanism by which, in the face of immunoselection/editing, synergy between tumor Ag-specific T cell responses, and natural effectors can contribute to the elimination of tumor cells whether or not they express the T cell tumor Ag epitope. This mechanism appears ineffective, however, once tumor variants lacking a T cell-recognized epitope are separated from the wild-type tumor, as would be the case in metastasis of tumor cells of this type. These findings highlight the importance of designing clinical protocols that promote collaboration between the adaptive and the innate immune effectors for successful therapy of cancer.


   Acknowledgments
 
We thank Benoît Van den Eynde (Ludwig Institute for Cancer Research, Brussels, Belgium) for the gift of P511 and P1.204 tumor cells, Elena Tomasello and Julie Chaix (CIML) for suggestions, Laurent Perrin and Bruno Zeitouni (Institut de Biologie du Dévloppement de Marseille Luminy, Marseille, France) for help with the RNA amplification protocol, and the CIML animal facility personnel for animal care.


   Disclosures
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
References
 
The authors have no financial conflict of interest.


   Footnotes
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 This work was supported by institutional funding from Institut National de la Santé et de la Recherche Médicale and Centre National de la Recherche Scientifique (CNRS), and by grants from Association pour la Recherche sur le Cancer (to A.M.S.V.), Institut National du Cancer (to A.M.S.V. and L.L.), the European Communities (Integrated Project "Cancerimmunotherapy" LSHC-CT-2006-518234 to A.M.S.V.), and the CNRS/Commissariat à l’Energie Atomique Program "Imagerie du Petit Animal" (to L.L. and A.-M.S.-V.). A.S. was supported in part by an International Cancer Research Technology Transfer Fellowship from the International Union Against Cancer. G.V. was the recipient of a Doctoral Fellowship from Ligue Nationale Contre le Cancer. Back

2 Current address: Laboratory of Experimental Immunology, Cancer and Inflammation Program, Science Applications International Corporation-Frederick, National Cancer Institute, Building 560, Room 31-27, Frederick, MD 21702-1201. Back

3 A.S. and G.V. are co-first authors. Back

4 Address correspondence and reprint requests to Dr. Anne-Marie Schmitt-Verhulst, Centre d’Immunologie de Marseille-Luminy, Campus de Luminy, Case 906, 13288 Marseille, Cedex 09, France. E-mail address: verhulst{at}ciml.univ-mrs.fr Back

5 Abbreviations used in this paper: DC, dendritic cell; {gamma}c, common cytokine receptor {gamma}-chain; LN, lymph node; DLN, draining LN; TIL, tumor-infiltrating leukocyte; RT, reverse transcription; SSS; second-strand synthesis; EST, expressed sequence tag. Back

Received for publication June 15, 2007. Accepted for publication August 20, 2007.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 

  1. Dunn, G. P., A. T. Bruce, H. Ikeda, L. J. Old, R. D. Schreiber. 2002. Cancer immunoediting: from immunosurveillance to tumor escape. Nat. Immunol. 3: 991-998. [Medline]
  2. Shankaran, V., H. Ikeda, A. T. Bruce, J. M. White, P. E. Swanson, L. J. Old, R. D. Schreiber. 2001. IFN{gamma} and lymphocytes prevent primary tumour development and shape tumour immunogenicity. Nature 410: 1107-1111. [Medline]
  3. Medzhitov, R., C. A. Janeway, Jr. 2002. Decoding the patterns of self and nonself by the innate immune system. Science 296: 298-300. [Abstract/Free Full Text]
  4. Glas, R., L. Franksson, C. Une, M. L. Eloranta, C. Ohlen, A. Orn, K. Karre. 2000. Recruitment and activation of natural killer (NK) cells in vivo determined by the target cell phenotype: an adaptive component of NK cell-mediated responses. J. Exp. Med. 191: 129-138. [Abstract/Free Full Text]
  5. Mocikat, R., H. Braumuller, A. Gumy, O. Egeter, H. Ziegler, U. Reusch, A. Bubeck, J. Louis, R. Mailhammer, G. Riethmuller, et al 2003. Natural killer cells activated by MHC class Ilow targets prime dendritic cells to induce protective CD8 T cell responses. Immunity 19: 561-569. [Medline]
  6. Kelly, J. M., P. K. Darcy, J. L. Markby, D. I. Godfrey, K. Takeda, H. Yagita, M. J. Smyth. 2002. Induction of tumor-specific T cell memory by NK cell-mediated tumor rejection. Nat. Immunol. 3: 83-90. [Medline]
  7. Coulie, P. G., H. Ikeda, J. F. Baurain, R. Chiari. 1999. Antitumor immunity at work in a melanoma patient. Adv. Cancer Res. 76: 213-242. [Medline]
  8. Glas, R., K. Sturmhofel, G. J. Hammerling, K. Karre, H. G. Ljunggren. 1992. Restoration of a tumorigenic phenotype by β2-microglobulin transfection to EL-4 mutant cells. J. Exp. Med. 175: 843-846. [Abstract/Free Full Text]
  9. Lanier, L. L.. 1998. NK cell receptors. Annu. Rev. Immunol. 16: 359-393. [Medline]
  10. Long, E. O.. 1999. Regulation of immune responses through inhibitory receptors. Annu. Rev. Immunol. 17: 875-904. [Medline]
  11. Moretta, A., C. Bottino, M. Vitale, D. Pende, C. Cantoni, M. C. Mingari, R. Biassoni, L. Moretta. 2001. Activating receptors and coreceptors involved in human natural killer cell-mediated cytolysis. Annu. Rev. Immunol. 19: 197-223. [Medline]
  12. Jamieson, A. M., A. Diefenbach, C. W. McMahon, N. Xiong, J. R. Carlyle, D. H. Raulet. 2002. The role of the NKG2D immunoreceptor in immune cell activation and natural killing. Immunity 17: 19-29. [Medline]
  13. Groh, V., A. Steinle, S. Bauer, T. Spies. 1998. Recognition of stress-induced MHC molecules by intestinal epithelial {gamma}{delta} T cells. Science 279: 1737-1740. [Abstract/Free Full Text]
  14. Gasser, S., S. Orsulic, E. J. Brown, D. H. Raulet. 2005. The DNA damage pathway regulates innate immune system ligands of the NKG2D receptor. Nature 436: 1186-1190. [Medline]
  15. Cerwenka, A., J. L. Baron, L. L. Lanier. 2001. Ectopic expression of retinoic acid early inducible-1 gene (RAE-1) permits natural killer cell-mediated rejection of a MHC class I-bearing tumor in vivo. Proc. Natl. Acad. Sci. USA 98: 11521-11526. [Abstract/Free Full Text]
  16. Diefenbach, A., E. R. Jensen, A. M. Jamieson, D. H. Raulet. 2001. Rae1 and H60 ligands of the NKG2D receptor stimulate tumour immunity. Nature 413: 165-171. [Medline]
  17. Van Der Bruggen, P., Y. Zhang, P. Chaux, V. Stroobant, C. Panichelli, E. S. Schultz, J. Chapiro, B. J. Van Den Eynde, F. Brasseur, T. Boon. 2002. Tumor-specific shared antigenic peptides recognized by human T cells. Immunol. Rev. 188: 51-64. [Medline]
  18. Lethe, B., B. van den Eynde, A. van Pel, G. Corradin, T. Boon. 1992. Mouse tumor rejection antigens P815A and P815B: two epitopes carried by a single peptide. Eur. J. Immunol. 22: 2283-2288. [Medline]
  19. Brandle, D., J. Bilsborough, T. Rulicke, C. Uyttenhove, T. Boon, B. J. Van den Eynde. 1998. The shared tumor-specific antigen encoded by mouse gene P1A is a target not only for cytolytic T lymphocytes but also for tumor rejection. Eur. J. Immunol. 28: 4010-4019. [Medline]
  20. Van den Eynde, B., H. Mazarguil, B. Lethe, F. Laval, J. E. Gairin. 1994. Localization of two cytotoxic T lymphocyte epitopes and three anchoring residues on a single nonameric peptide that binds to H-2Ld and is recognized by cytotoxic T lymphocytes against mouse tumor P815. Eur. J. Immunol. 24: 2740-2745. [Medline]
  21. Uyttenhove, C., J. Maryanski, T. Boon. 1983. Escape of mouse mastocytoma P815 after nearly complete rejection is due to antigen-loss variants rather than immunosuppression. J. Exp. Med. 157: 1040-1052. [Abstract/Free Full Text]
  22. Bai, X. F., J. Liu, O. Li, P. Zheng, Y. Liu. 2003. Antigenic drift as a mechanism for tumor evasion of destruction by cytolytic T lymphocytes. J. Clin. Invest. 111: 1487-1496. [Medline]
  23. Bilsborough, J., A. Van Pel, C. Uyttenhove, T. Boon, B. J. Van den Eynde. 1999. Identification of a second major tumor-specific antigen recognized by CTLs on mouse mastocytoma P815. J. Immunol. 162: 3534-3540. [Abstract/Free Full Text]
  24. Shanker, A., N. Auphan-Anezin, P. Chomez, L. Giraudo, B. Van den Eynde, A. M. Schmitt-Verhulst. 2004. Thymocyte-intrinsic genetic factors influence CD8 T cell lineage commitment and affect selection of a tumor-reactive TCR. J. Immunol. 172: 5069-5077. [Abstract/Free Full Text]
  25. Brichard, V. G., G. Warnier, A. Van Pel, G. Morlighem, S. Lucas, T. Boon. 1995. Individual differences in the orientation of the cytolytic T cell response against mouse tumor P815. Eur. J. Immunol. 25: 664-671. [Medline]
  26. Alter, G., J. M. Malenfant, M. Altfeld. 2004. CD107a as a functional marker for the identification of natural killer cell activity. J. Immunol. Methods 294: 15-22. [Medline]
  27. Lyons, A. B., C. R. Parish. 1994. Determination of lymphocyte division by flow cytometry. J. Immunol. Methods 171: 131-137. [Medline]
  28. Koo, G. C., J. R. Peppard. 1984. Establishment of monoclonal anti-Nk-1.1 antibody. Hybridoma 3: 301-303. [Medline]
  29. Puthier, D., F. Joly, M. Irla, M. Saade, G. Victorero, B. Loriod, C. Nguyen. 2004. A general survey of thymocyte differentiation by transcriptional analysis of knockout mouse models. J. Immunol. 173: 6109-6118. [Abstract/Free Full Text]
  30. Baugh, L. R., A. A. Hill, E. L. Brown, C. P. Hunter. 2001. Quantitative analysis of mRNA amplification by in vitro transcription. Nucleic Acids Res. 29: E29[Medline]
  31. Verdeil, G., D. Puthier, C. Nguyen, A. M. Schmitt-Verhulst, N. Auphan-Anezin. 2002. Gene profiling approach to establish the molecular bases for partial versus full activation of naive CD8 T lymphocytes. Ann. NY Acad. Sci. 975: 68-76. [Medline]
  32. Lopez, F., J. Rougemont, B. Loriod, A. Bourgeois, L. Loi, F. Bertucci, P. Hingamp, R. Houlgatte, S. Granjeaud. 2004. Feature extraction and signal processing for nylon DNA microarrays. BMC Genomics 5: 38[Medline]
  33. Verdeil, G., D. Puthier, C. Nguyen, A. M. Schmitt-Verhulst, N. Auphan-Anezin. 2006. STAT5-mediated signals sustain a TCR-initiated gene expression program toward differentiation of CD8 T cell effectors. J. Immunol. 176: 4834-4842. [Abstract/Free Full Text]
  34. Haring, J. S., G. A. Corbin, J. T. Harty. 2005. Dynamic regulation of IFN-{gamma} signaling in antigen-specific CD8+ T cells responding to infection. J. Immunol. 174: 6791-6802. [Abstract/Free Full Text]
  35. Zhang, B., N. A. Bowerman, J. K. Salama, H. Schmidt, M. T. Spiotto, A. Schietinger, P. Yu, Y. X. Fu, R. R. Weichselbaum, D. A. Rowley, et al 2007. Induced sensitization of tumor stroma leads to eradication of established cancer by T cells. J. Exp. Med. 204: 49-55. [Abstract/Free Full Text]
  36. Wang, M., C. A. Ellison, J. G. Gartner, K. T. HayGlass. 1998. Natural killer cell depletion fails to influence initial CD4 T cell commitment in vivo in exogenous antigen-stimulated cytokine and antibody responses. J. Immunol. 160: 1098-1105. [Abstract/Free Full Text]
  37. Vinay, D. S., B. K. Choi, J. S. Bae, W. Y. Kim, B. M. Gebhardt, B. S. Kwon. 2004. CD137-deficient mice have reduced NK/NKT cell numbers and function, are resistant to lipopolysaccharide-induced shock syndromes, and have lower IL-4 responses. J. Immunol. 173: 4218-4229. [Abstract/Free Full Text]
  38. Salazar-Mather, T. P., J. S. Orange, C. A. Biron. 1998. Early murine cytomegalovirus (MCMV) infection induces liver natural killer (NK) cell inflammation and protection through macrophage inflammatory protein 1{alpha} (MIP-1{alpha})-dependent pathways. J. Exp. Med. 187: 1-14. [Abstract/Free Full Text]
  39. Dunn, G. P., L. J. Old, R. D. Schreiber. 2004. The immunobiology of cancer immunosurveillance and immunoediting. Immunity 21: 137-148. [Medline]
  40. Smyth, M. J., D. I. Godfrey, J. A. Trapani. 2001. A fresh look at tumor immunosurveillance and immunotherapy. Nat. Immunol. 2: 293-299. [Medline]
  41. Zippelius, A., P. Batard, V. Rubio-Godoy, G. Bioley, D. Lienard, F. Lejeune, D. Rimoldi, P. Guillaume, N. Meidenbauer, A. Mackensen, et al 2004. Effector function of human tumor-specific CD8 T cells in melanoma lesions: a state of local functional tolerance. Cancer Res. 64: 2865-2873. [Abstract/Free Full Text]
  42. Tosi, D., R. Valenti, A. Cova, G. Sovena, V. Huber, L. Pilla, F. Arienti, F. Belardelli, G. Parmiani, L. Rivoltini. 2004. Role of cross-talk between IFN-{alpha}-induced monocyte-derived dendritic cells and NK cells in priming CD8+ T cell responses against human tumor antigens. J. Immunol. 172: 5363-5370. [Abstract/Free Full Text]
  43. Hanson, H. L., D. L. Donermeyer, H. Ikeda, J. M. White, V. Shankaran, L. J. Old, H. Shiku, R. D. Schreiber, P. M. Allen. 2000. Eradication of established tumors by CD8+ T cell adoptive immunotherapy. Immunity 13: 265-276. [Medline]
  44. Matsui, K., L. A. O’Mara, P. M. Allen. 2003. Successful elimination of large established tumors and avoidance of antigen-loss variants by aggressive adoptive T cell immunotherapy. Int. Immunol. 15: 797-805. [Abstract/Free Full Text]
  45. Dudley, M. E., D. C. Roopenian. 1996. Loss of a unique tumor antigen by cytotoxic T lymphocyte immunoselection from a 3-methylcholanthrene-induced mouse sarcoma reveals secondary unique and shared antigens. J. Exp. Med. 184: 441-447. [Abstract/Free Full Text]
  46. Lozupone, F., L. Rivoltini, F. Luciani, M. Venditti, L. Lugini, A. Cova, P. Squarcina, G. Parmiani, F. Belardelli, S. Fais. 2003. Adoptive transfer of an anti-MART-1(27–35)-specific CD8+ T cell clone leads to immunoselection of human melanoma antigen-loss variants in SCID mice. Eur. J. Immunol. 33: 556-566. [Medline]
  47. Bai, X. F., J. Q. Liu, P. S. Joshi, L. Wang, L. Yin, J. Labanowska, N. Heerema, P. Zheng, Y. Liu. 2006. Different lineages of P1A-expressing cancer cells use divergent modes of immune evasion for T-cell adoptive therapy. Cancer Res. 66: 8241-8249. [Abstract/Free Full Text]
  48. Blankenstein, T.. 2005. The role of tumor stroma in the interaction between tumor and immune system. Curr. Opin. Immunol. 17: 180-186. [Medline]
  49. Spiotto, M. T., D. A. Rowley, H. Schreiber. 2004. Bystander elimination of antigen loss variants in established tumors. Nat. Med. 10: 294-298. [Medline]
  50. Chan, C. W., E. Crafton, H. N. Fan, J. Flook, K. Yoshimura, M. Skarica, D. Brockstedt, T. W. Dubensky, M. F. Stins, L. L. Lanier, et al 2006. Interferon-producing killer dendritic cells provide a link between innate and adaptive immunity. Nat. Med. 12: 207-213. [Medline]
  51. Taieb, J., N. Chaput, C. Menard, L. Apetoh, E. Ullrich, M. Bonmort, M. Pequignot, N. Casares, M. Terme, C. Flament, et al 2006. A novel dendritic cell subset involved in tumor immunosurveillance. Nat. Med. 12: 214-219. [Medline]
  52. Morris, M. A., K. Ley. 2004. Trafficking of natural killer cells. Curr. Mol. Med. 4: 431-438. [Medline]
  53. Zibert, A., S. Balzer, M. Souquet, T. H. Quang, C. Paris-Scholz, M. Roskrow, D. Dilloo. 2004. CCL3/MIP-1{alpha} is a potent immunostimulator when coexpressed with interleukin-2 or granulocyte-macrophage colony-stimulating factor in a leukemia/lymphoma vaccine. Hum. Gene Ther. 15: 21-34. [Medline]
  54. Smyth, M. J., E. Cretney, K. Takeda, R. H. Wiltrout, L. M. Sedger, N. Kayagaki, H. Yagita, K. Okumura. 2001. Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) contributes to interferon {gamma}-dependent natural killer cell protection from tumor metastasis. J. Exp. Med. 193: 661-670. [Abstract/Free Full Text]
  55. Hayakawa, Y., N. D. Huntington, S. L. Nutt, M. J. Smyth. 2006. Functional subsets of mouse natural killer cells. Immunol. Rev. 214: 47-55. [Medline]
  56. Xu, J., A. K. Chakrabarti, J. L. Tan, L. Ge, A. Gambotto, N. L. Vujanovic. 2007. Essential role of the TNF-TNFR2 cognate interaction in mouse dendritic cell-natural killer cell cross-talk. Blood 109: 3333-3341. [Abstract/Free Full Text]
  57. Fernandez, N. C., A. Lozier, C. Flament, P. Ricciardi-Castagnoli, D. Bellet, M. Suter, M. Perricaudet, T. Tursz, E. Maraskovsky, L. Zitvogel. 1999. Dendritic cells directly trigger NK cell functions: cross-talk relevant in innate anti-tumor immune responses in vivo. Nat. Med. 5: 405-411. [Medline]
  58. Wilson, J. L., J. Charo, A. Martin-Fontecha, P. Dellabona, G. Casorati, B. J. Chambers, R. Kiessling, M. T. Bejarano, H. G. Ljunggren. 1999. NK cell triggering by the human costimulatory molecules CD80 and CD86. J. Immunol. 163: 4207-4212. [Abstract/Free Full Text]
  59. Keefe, D., L. Shi, S. Feske, R. Massol, F. Navarro, T. Kirchhausen, J. Lieberman. 2005. Perforin triggers a plasma membrane-repair response that facilitates CTL induction of apoptosis. Immunity 23: 249-262. [Medline]
  60. Kandil, H., V. Bachy, D. J. Williams, R. Helmi, F. M. Gotch, M. A. Ibrahim. 2005. Regulation of dendritic cell interleukin-12 secretion by tumour cell necrosis. Clin. Exp. Immunol. 140: 54-64. [Medline]
  61. Chiesa, S., M. Mingueneau, N. Fuseri, B. Malissen, D. H. Raulet, M. Malissen, E. Vivier, E. Tomasello. 2006. Multiplicity and plasticity of natural killer cell signaling pathways. Blood 107: 2364-2372. [Abstract/Free Full Text]
  62. Hanna, J., P. Bechtel, Y. Zhai, F. Youssef, K. McLachlan, O. Mandelboim. 2004. Novel insights on human NK cells’ immunological modalities revealed by gene expression profiling. J. Immunol. 173: 6547-6563. [Abstract/Free Full Text]
  63. Kayagaki, N., N. Yamaguchi, M. Nakayama, K. Takeda, H. Akiba, H. Tsutsui, H. Okamura, K. Nakanishi, K. Okumura, H. Yagita. 1999. Expression and function of TNF-related apoptosis-inducing ligand on murine activated NK cells. J. Immunol. 163: 1906-1913. [Abstract/Free Full Text]
  64. Vermijlen, D., D. Luo, C. J. Froelich, J. P. Medema, J. A. Kummer, E. Willems, F. Braet, E. Wisse. 2002. Hepatic natural killer cells exclusively kill splenic/blood natural killer-resistant tumor cells by the perforin/granzyme pathway. J. Leukocyte Biol. 72: 668-676. [Abstract/Free Full Text]
  65. Vermijlen, D., C. Seynaeve, D. Luo, M. Kruhoffer, D. L. Eizirik, T. F. Orntoft, E. Wisse. 2004. High-density oligonucleotide array analysis reveals extensive differences between freshly isolated blood and hepatic natural killer cells. Eur. J. Immunol. 34: 2529-2540. [Medline]
  66. Robertson, M. J.. 2002. Role of chemokines in the biology of natural killer cells. J. Leukocyte Biol. 71: 173-183. [Abstract/Free Full Text]

Related articles in The JI:

IN THIS ISSUE

The JI 2007 179: 6377-6378. [Full Text]  



This article has been cited by other articles:


Home page
 Cancer Res.Home page
A. Shanker, A. D. Brooks, K. M. Jacobsen, J. W. Wine, R. H. Wiltrout, H. Yagita, and T. J. Sayers
Antigen Presented by Tumors In vivo Determines the Nature of CD8+ T-Cell Cytotoxicity
Cancer Res., August 15, 2009; 69(16): 6615 - 6623.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
D. L. Thomas, M. Kim, N. A. Bowerman, S. Narayanan, D. M. Kranz, H. Schreiber, and E. J. Roy
Recurrence of Intracranial Tumors following Adoptive T Cell Therapy Can Be Prevented by Direct and Indirect Killing Aided by High Levels of Tumor Antigen Cross-Presented on Stromal Cells
J. Immunol., August 1, 2009; 183(3): 1828 - 1837.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Related articles in The JI
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Shanker, A.
Right arrow Articles by Schmitt-Verhulst, A.-M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Shanker, A.
Right arrow Articles by Schmitt-Verhulst, A.-M.
Right arrowPubmed/NCBI databases
*Substance via MeSH

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS