| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Institute of Medical Microbiology and Hygiene, Department of Immunology, University of Freiburg, Freiburg, Germany;
Institute of Experimental Immunology, University Hospital, Zurich, Switzerland; and
Cell Genix Technologie Transfer GmbH, Freiburg, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
20%) of CD11b+CD11c-
macrophage-phenotype cells that were able to cross-present the gp33
epitope to T cells. Tumor-infiltrating CD8 T cells exhibited a highly
activated phenotype but lacked effector cell function. Strikingly, a
significant portion of tumor-infiltrating lymphocytes expressed TCRs
specific for gp33 but bound MHC tetramers only after cell purification
and a 24-h resting period in vitro. The phenomenon of
"tetramer-negative T cells" was not restricted to
tumor-infiltrating lymphocytes from MCA102gp33 tumors, but
was also observed when Ag-specific T cells derived from an environment
with high Ag load were analyzed ex vivo. Thus, using a novel tumor
model, allowing us to trace tumor-specific T cells at the single cell
level in vivo, we demonstrate that the tumor microenvironment is able
to alter the functional activity of T cells infiltrating the tumor
mass. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
have been shown to suppress the induction of
T cell responses in vitro (8, 9, 10). To further analyze these processes, we have established a novel murine tumor model using MCA102 fibrosarcoma cells transfected with a minigene encoding residues 33–41 (gp33 epitope) of the glycoprotein from lymphocytic choriomeningitis virus (LCMV).3 The gp33 epitope represents a strong CTL epitope of LCMV (11) and was selected as a model Ag because it allows tracing of gp33/tumor-specific CD8 T cells in normal mice using MHC class I tetramers complexed with the gp33 peptide (12, 13). The gp33 epitope has been used as a tumor-associated model Ag before and, depending on the tumor cell type and Ag expression levels, T cell priming (14, 15), spontaneous tumor regression (16), ignorance (17, 18, 19), or tolerance induction (20) has been observed. The MCA102gp33 tumor model described here differs strongly from these previously described models because progressive tumor cell growth is seen despite induction of systemic gp33/tumor-specific T cell immunity. Thus, the model exhibits striking similarities to the situation in cancer patients where high numbers of tumor-specific T cells were frequently detected, which nonetheless failed to contain tumor progression (21, 22, 23).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
C57BL/6 (B6) mice were obtained from our breeding colony and from Harlan Winkelmann (Borchen, Germany). P14 TCR-transgenic (tg) mice (24) (line 327) specific for amino acids 33–41 (gp33 epitope) of the LCMV glycoprotein in association with H-2Db and H8-tg mice (25) ubiquitously expressing the LCMV gp33 epitope as a transgene have been previously described. Female or male mice were used at 8–16 wk of age. Mice were bred and kept in a conventional animal house facility.
Tumors
MCA102 tumor cells (26) were kindly provided by S. Rosenberg (National Institutes of Health, Bethesda, MD) and MCA102gp33 cells were derived from parental tumor cells by gene transfection using the LCMV gp33 minigene as described previously (15). The resulting transfectants were cloned and screened for gp33 expression by CTL assays. The experiments reported in this study were performed with MCA102gp33 clone 3. Cells were cultured in DMEM high glucose supplemented with 10% FCS, glutamine, streptomycin, and penicillin (all from Life Technologies, Gaithersburg, MD). MCA102gp33 cells were kept in culture under G418 (600 µg/ml; Life Technologies) selection. Mice were injected with the indicated number of MCA102gp33 tumor cells in 100 µl PBS s.c. into the flank and tumor size was calculated as the product of bisecting tumor diameters. Mice bearing a tumor with a diameter >15 mm were killed according to animal care regulations.
Peptides and virus
LCMV gp33 (KAVYNFATM), nucleoprotein (NP)396 (FQPQNGQFI), and adenovirus E1A234–243 (SGPSNTPPEI) peptides were purchased from Neosystem (Strasbourg, France). The original cysteine at the anchor position 41 in the LCMV gp33 peptide was replaced by methionine. The LCMV-WE isolate used in this study was originally obtained from R. Zinkernagel (University Hospital, Zurich, Switzerland). Mice were infected i.v. with 200 pfu and viral titers were determined in a virus plaque assay as described (27).
In vivo analysis of cell rejection
Spleen cells from H8-tg mice were incubated at 5 x 106 cells/ml in ice-cold PBS containing 0.5 mM CFSE (Molecular Probes, Eugene, OR) for 10 min at 37°C. The cells were washed once in PBS containing 1% FCS and 108 cells were injected i.v. into recipient mice. The percentage of CFSE+ donor cells among recipient PBL was determined by flow cytometry.
Immunohistochemistry
Tumor sections (5–7 µm) were cut on a cryostat microtome, air dried, fixed in acetone, and blocked with TBS containing 5% mouse serum. Anti-CD8-biotin, anti-CD11b-biotin, and anti-CD11c-biotin (all from BD PharMingen, San Diego, CA) were used as primary mAb followed by avidin-conjugated alkaline phosphatase (StreptAB Complex/AP; DAKO, Hamburg, Germany) and developed with the alkaline phosphatase substrate kit I (Vector Laboratories, Burlingame, CA). Sections were counterstained with Mayer’s hemalum.
Cell isolation
For isolation of tumor-infiltrating cells, tumors were cut into small pieces and digested in PBS containing 0.1% collagenase (Sigma, Taufleirchen, Germany), 0.01% hyaluronidase (Sigma), and 0.002% DNase I (Sigma) for 2 h at 37°C. Undigested material was allowed to settle and released cells were recovered and washed. CD8 T cells were isolated using anti-CD8-conjugated beads (MACS; Miltenyi Biotec, Bergisch-Gladbach, Germany) and MS+ columns. For isolation of macrophages, anti-CD11b-conjugated beads (MACS; Miltenyi Biotec) were used.
T cell cultures and cytotoxicity assay
To test APC function, spleen cells (2 x 106) from P14 TCR-tg or LCMV-immune B6 mice were stimulated with the indicated numbers of CD11b+ cells (purity >95% CD11b+, <0.1% CD11c+), 2 x 105 irradiated (10,000 rad) MCA102gp33 tumor cells, or 2 x 106 B6 spleen cells loaded with 10-6 M gp33 peptide (1 h at 37°C). The cultures were performed in 24-well plates in 1 ml IMDM supplemented with 10% FCS, penicillin/streptomycin, and 0.001 M 2-ME. To induce a CTL response in vitro, 2 x 106 purified tumor-infiltrating lymphocytes (TILs) or 2 x 106 lymphocytes from PBL, spleen, or draining lymph nodes from tumor-bearing mice were stimulated for 5 days in vitro with 2 x 106 gp33 peptide-loaded B6 spleen cells in 24-well plates in 1 ml tissue culture medium. P14 TCR-tg spleen cells (2 x 106) were stimulated similarly for 3 days. Cytolytic activity was tested in 51Cr release assays using EL-4 target cells loaded with gp33 peptide (10-6 M for 2 h at 37°C) or with the control Db-binding adenovirus peptide aa 234–243. To test for Ag-specific cell proliferation, 2 x 106 purified TILs or 2 x 106 spleen cells from P14 TCR-tg mice were labeled with CFSE and stimulated with 2 x 106 gp33 peptide-loaded B6 spleen cells for 3 days in 24-well plates in 1 ml tissue culture medium.
Flow cytometry
Lymphocytes were resuspended in PBS containing 2% FCS and 0.1%
NaN3 at a concentration of
106–107 cells/ml, followed
by incubation at 4°C for 20 min with 100 µl of properly diluted
mAb. For PBL staining 10 U/ml heparin was added to the staining buffer.
The following mAb were used: CD8 (clone 53-6.7), CD11b (clone M1/70),
CD11c (clone HL3), CD25 (clone 7D4), CD44 (clone IM7), CD45 (clone
30-F11), CD62L (clone MEL-14), CD69 (clone H1.2F3), CD90.1 (clone
OX-7), TCR V
2 (clone B20.1), TCR- (clone H57-597),
I-Ab (clone AF6-120.1). All mAb were purchased
from BD PharMingen. H-2Db MHC class I tetramers
complexed with streptavidin-PE and containing the LCMV gp33 or NP396
peptides were prepared as described (12, 13). The modified
Db and 2 plasmids were a kind gift of Dr. J.
Altman (Emory University, Atlanta, GA). Tetramer staining was performed
at 4°C for 30 min followed by FITC-conjugated anti-CD8 mAb for a
further 20 min. TILs or spleen cells were stained with tetramers either
directly after cell purification or after preculture
(106 cells/ml) in tissue culture medium (IMDM
plus 10% FCS) for 24 h at 37°C. Addition of IL-2 to the culture
medium did not influence tetramer staining results. For direct ex vivo
functional analysis of PBL, RBCs were removed by nonfixating PharM Lyse
buffer (BD PharMingen) according to the instructions of the
manufacturer. For intracellular IFN-
staining, purified TILs or PBL
(105) were stimulated for 5 h in 96-well
flat-bottom microtiter plates with gp33 peptide-loaded bone
marrow-derived dendritic cells (28)
(105) in the presence of 1 µg/ml brefeldin A
(Golgistop; BD PharMingen). Cells were then surface stained with
anti-CD8-FITC, washed, permeabilized, and restained with
PE-conjugated rat anti-mouse IFN- mAb (clone XMG1.2; BD
PharMingen). Cells were analyzed on a FACSort flow cytometer (BD
Biosciences, Mountain View, CA) using CellQuest software. Before
analysis of PBL, RBCs were lysed using FACS-Lysing Solution (BD
PharMingen).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To examine the immunogenicity of the gp33 CTL epitope as a
tumor-associated Ag, parental and gp33-transfected MCA102 fibrosarcoma
cells (106) were transplanted s.c. into unprimed
B6 mice. As shown in Fig. 1
A,
injection of both types of tumor cells resulted in a similar,
progressive tumor growth. To exclude the possibility that
MCA102gp33 tumor development was due to the
selection of gp33 Ag loss variants, tumor cells were isolated and
tested in 51Cr release assays.
MCA102gp33 cells isolated from tumors
were lysed by gp33-specific CTL as efficiently as
MCA102gp33 cells kept in in vitro culture,
indicating that the gp33 epitope was still expressed in
MCA102gp33 tumors growing in B6 mice (Fig. 1B). In LCMV-immune B6 mice possessing gp33-specific
memory T cells, MCA102gp33 but not parental
MCA102 tumor cells were rejected (Fig. 1C). LCMV infection
of MCA102gp33 tumor-bearing B6 mice 10 days after
tumor cell injection led to a significant decrease in tumor size after
day 15 (Fig. 1D). However, tumor regression was transient
and after day 25 progressive tumor growth was observed in all mice. All
tumors isolated after this time point represented gp33 Ag loss variants
(data not shown). Thus, the gp33 epitope on MCA102 tumor cells serves
as a tumor rejection Ag in gp33-primed but not unprimed B6
mice.
|
MHC class I tetramers containing the gp33 epitope
(Db-gp33) were used to examine whether
inoculation of B6 mice with MCA102gp33 tumor
cells could induce gp33-specific T cells. Strikingly, a high number
(0.5–8% of CD8) of gp33-tetramer+ cells was
detected in PBL of MCA102gp33 tumor-bearing mice
(Fig. 2A). To investigate
whether this response represented gp33-specific T cell immunity, three
functional assays were used.
|
B). To demonstrate that the observed antiviral protection
was due to gp33-specific T cells, the experiment was repeated using
H8-tg mice (25) that ubiquitously express the gp33 epitope
as a transgene. These mice have a normal T cell repertoire, but due to
central tolerance induction they lack gp33-specific T cells. Fig. 2B, column 3, shows that injection of
MCA102gp33 tumor cells did not reduce viral
titers in H8-tg mice, indicating that the observed antiviral protection
in MCA102gp33 tumor-bearing B6 mice was due to
the gp33-specific T cell response. Efficient antiviral protection was
also obtained when irradiated MCA102gp33 tumor
cells were used for immunization (Fig. 2B, columns 4 and 5),
whereas injection of nonviable tumor cells (freeze/thawed) failed to
induce a gp33-specific response.
Second, the ability of MCA102gp33 tumor-bearing
mice to reject gp33 Ag-expressing spleen cells was examined. For these
experiments, spleen cells from H8-tg mice, labeled with the vital dye
CFSE, were transferred into tumor-bearing mice 2 wk after tumor cell
injection, and the kinetics of the transferred donor cells was
followed. As shown in Fig. 2C, shortly after adoptive
transfer, CFSE+ H8 spleen cells represented
3–5% of host PBL. Afterward, the number of
CFSE+ cells decreased over the course of 3 days
in mice bearing MCA102gp33 tumors, whereas they
persisted in mice bearing parental tumors.
Third, we compared gp33-specific CTL activity of in vitro stimulated
lymphocytes from PBL, spleen, and draining lymph nodes of
MCA102gp33 tumor-bearing mice. In mice bearing
day 10 tumors, gp33-specific CTL activity was predominantly observed in
cultures of PBL (Fig. 2D, upper panel). At a
later stage (day 20), strong gp33-specific CTL activity was also
detected in cultures of spleen and draining lymph nodes (Fig. 2D, lower panel). gp33-specific CTL activity was
not detected in T cell cultures from normal mice or mice bearing
parental MCA102 tumors (data not shown). Taken together, these data
demonstrate that growth of MCA102gp33 tumors
induced a strong gp33-specific CTL response capable of mediating
antiviral protection and specific rejection of gp33-expressing spleen
cells in vivo.
Resistance of MCA102gp33 tumor-bearing mice to a second tumor challenge
The unrestricted growth of MCA102gp33 tumors
in the presence of a potent gp33-specific T cell response was
unexpected. A kinetic disparity between tumor growth and induction of
functional gp33-specific immunity may have occurred, and the tumor has
already grown beyond the critical mass that can still be eliminated by
T cells (29, 30). To provide T cells with a kinetic
advantage, B6 mice bearing 5-day (group 1) or 10-day (group 2)
MCA102gp33 tumors on the left flank were
challenged with 106
MCA102gp33 tumor cells in the right flank. As
shown in Fig. 3A, both groups
of tumor-bearing mice were resistant to a second tumor challenge.
Concomitant tumor immunity was gp33 Ag-specific because mice bearing
MCA102gp33 tumors were not protected against a
second tumor challenge with parental MCA102 tumor cells (Fig. 3B). These results demonstrate that gp33-specific T cells
induced by the first tumor cell injection could mediate rejection of
newly transplanted tumor cells. Thus, the kinetic balance between tumor
growth and induction of gp33-specific immunity plays an important role
in this model.
|
MCA102gp33 tumors that grew for 2–3 wk in
B6 mice were infiltrated with CD4 and CD8 T cells (2–5%), a high
number (20%) of CD11b+, and a few (<1%)
CD11c+ cells (Fig. 4A). The
CD11b+ cells expressed CD45 and MHC class II
molecules, but not the dendritic cell marker CD11c, and thus
represented most likely tumor-infiltrating macrophages (Fig. 4B). To test whether these cells could present the gp33
epitope to T cells, CD11b+ cells were isolated
from the tumor mass (purity >95% CD11b+,
<0.1% CD11c+) and were used as APC for in vitro
stimulation of responder spleen cells from P14 TCR-tg mice. Irradiated
MCA102gp33 tumor cells were included in the assay
to judge the direct priming capacity of the tumor cells themselves. As
shown in Fig. 4C, top panel,
CD11b+ cells isolated from the tumor mass induced
gp33-specific CTLs as efficiently as gp33 peptide-loaded B6 spleen
cells, whereas MCA102gp33 tumor cells failed to
elicit a CTL response. The same results were obtained when spleen cells
from LCMV-immune B6 mice were used as responder cells (Fig. 4C, bottom panel). Thus,
CD11b+ cells isolated from the tumor mass were
able to cross-present the gp33 epitope to CD8 T cells.
|
Next, the Ag specificity of TILs from
MCA102gp33 tumors was determined 2–3 wk after
tumor cell injection. Purified CD8 TILs exhibited a highly activated
phenotype with up-regulated CD25, CD44, and CD69 and down-regulated
CD62L expression (Fig. 5A).
Tetramer staining further revealed that 2–3% of CD8 T cells isolated
ex vivo from MCA102gp33 tumors appeared to be
specific for gp33 Ag. However, the number of
gp33-tetramer+ cells increased strikingly when
the staining was performed with purified TILs that had been precultured
for 24 h in the absence of "contaminating" cells from the
tumor mass. Under these conditions, a significant portion of the
isolated TILs could be stained with gp33-tetramers, whereas control
staining with NP396-tetramers or staining with gp33-tetramers on TILs
from parental MCA102 tumors was negative, with or without preculture
(Fig. 5B, columns 2 and3). Kinetic analysis showed
that the ability of gp33-tetramers to stain purified TILs from
MCA102gp33 tumors gradually increased with the
duration of in vitro culture (Fig. 5C). However, the
intensity of gp33-tetramer staining of TILs was lower than the
intensity of PBL from tumor-bearing mice (Fig. 2A). The
reduced ability to stain freshly isolated TILs with tetramers was not
due to TCR down-regulation, because the cells expressed TCRs at normal
levels (Fig. 5A). Thus, a significant portion of the T cells
infiltrating MCA102gp33 tumors were specific for
the gp33 epitope. However, TILs bound gp33-tetramers only after 24
h preculture in the absence of gp33-expressing/presenting cells.
|
The altered gp33-tetramer binding of TILs could be due to specific
factors present in the MCA102gp33 tumor mass or,
alternatively, could represent a more general phenomenon where T cells
analyzed ex vivo from an Ag-expressing environment cannot efficiently
be stained with tetramers. To address this issue, T cells from P14
TCR-tg mice specific for gp33 were adoptively transferred into H8-tg
mice that express the gp33 epitope on all MHC class
I+ cells (25). As a control, P14 T
cells were also transferred into B6 mice. Immediately after cell
transfer, both types of recipient mice were infected with LCMV to
induce expansion of the transferred donor cells. P14 T cells
(Thy1.1+) were traced in the recipient mice
(Thy1.2+) using Thy1.1-specific mAb. In B6 mice,
P14 T cells could be stained with both mAb specific for the tg TCR
V2 chain and with gp33-tetramers (Fig. 6A, left panels).
In striking contrast, a large fraction of P14 T cells isolated from
H8-tg mice could not be stained with gp33-tetramers, despite positive
staining with TCR V2-specific mAb (Fig. 6A, right
panels). Similar to TILs from MCA102gp33
tumors, the ability of gp33-tetramers to stain P14 T cells from H8-tg
mice increased significantly after overnight in vitro culture at 37°C
but not at 4°C (Fig. 6B). Tetramer staining was specific
because P14 T cells were not stained with NP396-tetramers.
|
The paradox of progressive tumor growth despite the high
percentage of gp33-tetramer+ cells within the CD8
subset led us to examine CTL activity, IFN- secretion, and
proliferative capacity of purified TILs from
MCA102gp33 tumors. As shown in Fig. 7, PBL from
MCA102gp33 tumor-bearing mice directly used in 5-
or 18-h 51Cr release assays efficiently lysed
gp33 peptide-loaded target cells. Ex vivo CTL activity of spleen cells
was also detected in 18-h 51Cr release assays
(30% specific lysis at a CD8 to target ratio of 60:1). However, the
activity was considerably lower than that of PBL. CTL activity of TILs
was <15% in 5-h assays, even after the purified cells had been
precultured for 24 h. Under these conditions, most of the cells
could be stained with gp33-tetramers (Fig. 5B). In 18-h
assays (Fig. 7, right panels), gp33-specific CTL activity of
TILs was detectable; however, this activity was 10-fold lower on a
CD8 cell to target ratio when compared with PBL. IFN- secretion of
TILs after gp33 Ag stimulation was determined by intracellular cytokine
staining. As shown in Fig. 8A,
TILs isolated from MCA102gp33 tumors were unable
to produce notable amounts of IFN- after gp33 Ag stimulation when
assayed ex vivo or after a 24-h preculture. In contrast, most of
gp33-tetramer+ cells in PBL of
MCA102gp33 tumor-bearing mice produced high
levels of IFN- after gp33 Ag stimulation (Fig. 8B).
Finally, the proliferative capacity of TILs after gp33 Ag stimulation
was tested. TILs purified from MCA102gp33 tumors
were labeled with the vital dye CFSE and were stimulated for 4 days in
vitro using gp33 peptide-loaded spleen cells. Under these conditions
P14 T cells proliferated vigorously, as indicated by a decrease in
their CFSE content. In contrast, TILs failed to divide at all (Fig. 8C), even when the culture medium was supplemented with 40
U/ml IL-2 (data not shown). Furthermore, gp33 stimulation of TILs for 5
days in vitro failed to induce gp33-specific CTL activity, whereas
spleen cells from the same tumor-bearing mice specifically lysed target
cells when stimulated under the same conditions (Fig. 8D).
In short, these data demonstrate that TILs in
MCA102gp33 tumor-bearing mice were functionally
impaired.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The frequency of gp33-specific T cells in tumor-bearing mice was
determined by using MHC class I tetramers containing the gp33 epitope.
The analysis revealed a high proportion (0.5–8% of CD8) of
gp33-tetramer+ cells in PBL. In naive B6 mice,
the precursor frequency of cells specific for the gp33 epitope has been
estimated to be in the range of
1:104/105 in CD8 T cells
(13). Thus, gp33-specific T cells expanded
103-fold within 2–3 wk after tumor cell
challenge. Clonal expansion of these cells was accompanied by
acquisition of effector cell function because
MCA102gp33 tumor-bearing mice were protected
against LCMV infection and were able to reject gp33-expressing spleen
cells in vivo. In this respect, the MCA102gp33
model exhibits similarities to the allogeneic
H-2K216 tumor model in which mice bearing
K216-positive tumors were able to reject
K216-positive skin transplants (31).
However, it differs from most other models with defined
tumor-associated Ags showing T cell priming (14, 15),
tumor regression (16), T cell ignorance
(17, 18, 19), or tolerance induction (2, 20, 32).
In a murine colon carcinoma model, peripheral T cells from late stage
tumor-bearing mice expressed Ag receptors that contained low amounts of
CD3- and completely lacked CD3-
(33). The model
described here differs because peripheral T cells in
MCA102gp33 tumor-bearing mice were functional and
immune dysfunction was restricted to T cells inside the tumor mass.
Interestingly, in patients with colorectal carcinomas CD3- levels
were low in TILs but only slightly reduced in PBL
(34).
The potent induction of a gp33-specific T cell response in tumor-bearing mice raises the question of the mechanism of gp33 Ag presentation. MCA102gp33 tumors were infiltrated with a high number of CD11b+ macrophage phenotype cells. This could be due to soluble factors produced by the tumor cells, specific cell-cell interactions, and/or tumor cell necrosis. Cell debris and possibly also heat shock protein-associated peptides (35) could be taken up by these cells and subsequently cross-presented to CD8 T cells, either inside the tumor mass or after migration to lymphoid organs. Our experiments further showed that tumor-infiltrating CD11b+ cells, but not MCA102gp33 tumor cells themselves, were able to activate gp33-specific T cells in vitro. However, it is doubtful that the systemic gp33-specific T cell immunity observed in tumor-bearing mice was due to Ag presentation by macrophages inside the tumor mass. CD11c+ cells, most likely dendritic cells, were also detected in MCA102gp33 tumors, albeit at a lower frequency. In addition, metastasizing MCA102gp33 cells were found by PCR in local lymph nodes and spleens of tumor-bearing mice (data not shown). Thus, it is more likely that gp33-specific T cell immunity was induced in lymphoid organs of tumor-bearing mice.
In contrast to unprimed B6 mice, LCMV-immune mice or mice already
bearing a tumor on the opposite flank rejected
MCA102gp33 tumor cells. How can these findings be
explained? It has been demonstrated that tumors grow more effectively
when transplanted as solid fragments compared with cell suspensions,
even though the latter contain more cells (5, 18). Thus, T
cells are likely to be more effective against isolated tumor cells
compared with tumor cell elimination from a solid tumor mass. In
unprimed B6 mice, gp33-tetramer+ T cells were
below the detection limit on day 10 after tumor cell injection (Fig. 2A). This time period would allow the transplanted tumor
cells to develop a solid tumor mass that is more difficult to control.
In mice with an increased frequency of gp33-specific T cells, this
kinetic balance is shifted toward protective antitumor immunity.
LCMV infection of B6 mice bearing day-10
MCA102gp33 tumors induced transient tumor
regression, followed by outgrowth of gp33 loss variants (Fig. 1D). Due to the low level of viral replication in these mice
(Fig. 2B), the number of
gp33-tetramer+ T cells in PBL was 10-fold reduced
when compared with LCMV-infected control mice (data not shown).
Nevertheless, the tumor regression observed indicates that
gp33-specific CTL induced by LCMV infection were able to eliminate
gp33-expressing tumor cells, even at a later stage of tumor
development. It is therefore possible that additional factors triggered
by the viral infection, such as induced CD4 T cells and activated APC,
help to improve and sustain gp33-specific T cell activity.
The peculiar gp33-tetramer binding behavior of TILs from tumor-bearing mice deserves discussion. Most of the TILs could only be stained with gp33-tetramers after cell purification and a resting period. The impaired tetramer staining of TILs was not due to Ag-induced TCR down-regulation, because ex vivo isolated cells expressed TCRs at normal levels. In contrast, gp33-tetramer+ T cells in PBL could be immediately stained by conventional methods. These circulating T cells probably had less direct contact with gp33-expressing/presenting cells than T cells from the tumor mass. Our results further revealed that "tetramer-negative T cells" were not restricted to TILs in MCA102gp33 tumors, but were also found among gp33-specific P14 T cells isolated from gp33-expressing H8-tg mice. Thus, our data suggest that the inability to stain Ag-specific T cells with tetramers represents a phenomenon that is observed when T cells derived from an environment with high Ag load are analyzed ex vivo.
Failure to stain Ag-specific CD8 T cells with MHC tetramers is not an unprecedented observation. Tetramer-negative Ag-specific CD8 T cells have been described in newborn mice infected with oncogenic polyoma virus (36) and in chronic hepatitis B virus patients (37). Discrepancies between cytokine release or cytolytic activity and tetramer binding have also been found in melanoma patients (38) and have been reported for cultured T cells (39). Furthermore, triggering T cell clones under certain in vitro stimulation conditions has been demonstrated to lead to a loss of tetramer labeling (40). Moreover, tetramer binding has been shown to be an active cellular process requiring cytoskeletal rearrangements (41), and efficient binding also depends on the integrity of lipid rafts on the surface of T cells (42). Thus, interaction of tetrameric MHC molecules with TCRs appears to be influenced by the activation state of the T cell and follows different rules when compared with binding of Abs to cell surface molecules.
The lack of effector cell function of tumor-specific T cells inside the
MCA102gp33 tumor provides an explanation for the
tumor outgrowth despite potent systemic gp33-specific T cell immunity.
It is possible that soluble factors such as TGF-
(8, 9, 10) produced by the tumor cells themselves or by
infiltrating cells inhibited the response of the TILs at the tumor
site. However, supernatant of MCA102gp33 tumor
cells contained only low levels of TGF-1 or -2 (20 pg/ml, data
not shown). Alternatively, permanent stimulation of gp33-specific T
cells by gp33-expressing tumor cells and by gp33 cross-presenting
macrophages in the absence of costimulatory signals may lead to the
observed loss of effector cell function of the TILs. In this respect,
it is noteworthy that neither MCA102gp33 tumor
cells nor tumor-infiltrating macrophages express costimulatory
molecules such as CD80, CD86, or ICAM-1 (data not shown). The activated
phenotype and the lack of effector function of TILs in
MCA102gp33 tumors are reminiscent of
virus-specific "Sisyphean" CD8 T cells found in persistently
infected hosts (43). Similarly, P14 T cells isolated from
LCMV-infected H8-tg mice ubiquitously expressing gp33 have been shown
to exhibit an activated phenotype and to be anergic (25).
As demonstrated in this study, these T cells also display an altered
gp33-tetramer staining behavior similar to TILs from
MCA102gp33 tumors (Fig. 6). Thus, we favor the
view that gp33-specific T cells activated in lymphoid organs of
MCA102gp33 tumor-bearing mice "exhaust"
inside the tumor mass due to the high Ag load present mainly on
nonprofessional APC.
In conclusion, the present study provides a striking in vivo example of how a tumor can grow progressively despite potent systemic tumor-specific T cell immunity. The model helps us to understand failures in tumor regression in patients despite successful induction of tumor-specific CTLs in the periphery. Our study emphasizes the importance of understanding the local mechanisms that prevent destruction of an established tumor by T cells.
|
Acknowledgments |
|---|
2 microglobulin
constructs. We also thank Oliver Schweier for generation of MHC
tetramers, Marlies Rawiel for expert technical assistance, and Theresa
Treuer, Rainer Bronner, and Thomas Imhof for animal husbandry. |
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Hanspeter Pircher, Institute of Medical Microbiology and Hygiene, Department of Immunology, Hermann-Herder-Strasse 11, University of Freiburg, D-79104 Freiburg, Germany. E-mail address: pircher{at}UKL.uni-freiburg.de
3 Abbreviations used in this paper: LCMV, lymphocytic choriomeningitis virus; tg, transgenic; TIL, tumor-infiltrating lymphocyte; NP, nucleoprotein.
Received for publication June 14, 2002. Accepted for publication September 24, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2-microglobulin in metastatic melanomas from five patients receiving immunotherapy. J. Natl. Cancer Inst. 88:100.. EMBO J. 6:1633.[Medline]
in growth, motility, angiogenesis, and immune escape. Microsc. Res. Tech. 52:401.[Medline]
in immune-evasion of cancer. Microsc. Res. Tech. 52:387.[Medline]
in CD8 T cell-mediated elimination of 3LL-A9 Lewis lung carcinoma cells in vivo. J. Immunol. 164:3645. secretion and binding of A2/peptide multimers to TCR reveals interclonal dissociation of CTL effector function from TCR-peptide/MHC complexes half-life. Proc. Natl. Acad. Sci. USA 98:10302.This article has been cited by other articles:
|
|
 |
|
  Y. Zhang, Q. Liu, M. Zhang, Y. Yu, X. Liu, and X. Cao Fas Signal Promotes Lung Cancer Growth by Recruiting Myeloid-Derived Suppressor Cells via Cancer Cell-Derived PGE2 J. Immunol., March 15, 2009; 182(6): 3801 - 3808. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Hargadon, C. C. Brinkman, S. L. Sheasley-O'Neill, L. A. Nichols, T. N. J. Bullock, and V. H. Engelhard Incomplete Differentiation of Antigen-Specific CD8 T Cells in Tumor-Draining Lymph Nodes J. Immunol., November 1, 2006; 177(9): 6081 - 6090. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Ryan and T. D. Schell Accumulation of CD8+ T Cells in Advanced-Stage Tumors and Delay of Disease Progression following Secondary Immunization against an Immunorecessive Epitope J. Immunol., July 1, 2006; 177(1): 255 - 267. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Kao, M. A. Daniels, and S. C. Jameson Loss of CD8 and TCR binding to Class I MHC ligands following T cell activation Int. Immunol., December 1, 2005; 17(12): 1607 - 1617. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Curtsinger, D. C. Lins, C. M. Johnson, and M. F. Mescher Signal 3 Tolerant CD8 T Cells Degranulate in Response to Antigen but Lack Granzyme B to Mediate Cytolysis J. Immunol., October 1, 2005; 175(7): 4392 - 4399. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. R. Drake III, R. M. Ream, C. W. Lawrence, and T. J. Braciale Transient Loss of MHC Class I Tetramer Binding after CD8+ T Cell Activation Reflects Altered T Cell Effector Function J. Immunol., August 1, 2005; 175(3): 1507 - 1515. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Avogadri, C. Martinoli, L. Petrovska, C. Chiodoni, P. Transidico, V. Bronte, R. Longhi, M. P. Colombo, G. Dougan, and M. Rescigno Cancer Immunotherapy Based on Killing of Salmonella-Infected Tumor Cells Cancer Res., May 1, 2005; 65(9): 3920 - 3927. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Cham and T. F. Gajewski Glucose Availability Regulates IFN-{gamma} Production and p70S6 Kinase Activation in CD8+ Effector T Cells J. Immunol., April 15, 2005; 174(8): 4670 - 4677. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Salio, M. J. Palmowski, A. Atzberger, I. F. Hermans, and V. Cerundolo CpG-matured Murine Plasmacytoid Dendritic Cells Are Capable of In Vivo Priming of Functional CD8 T Cell Responses to Endogenous but Not Exogenous Antigens J. Exp. Med., February 17, 2004; 199(4): 567 - 579. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Jackaman, C. S. Bundell, B. F. Kinnear, A. M. Smith, P. Filion, D. van Hagen, B. W. S. Robinson, and D. J. Nelson IL-2 Intratumoral Immunotherapy Enhances CD8+ T Cells That Mediate Destruction of Tumor Cells and Tumor-Associated Vasculature: A Novel Mechanism for IL-2 J. Immunol., November 15, 2003; 171(10): 5051 - 5063. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. H. Maris, J. D. Miller, J. D. Altman, and J. Jacob A Transgenic Mouse Model Genetically Tags All Activated CD8 T Cells J. Immunol., September 1, 2003; 171(5): 2393 - 2401. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
