| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Clinical Trials |
University of Pittsburgh Cancer Institute [T. S., G. D., W. G., M. T. L., T. L. W.] and Departments of Pathology [G. D., T. L. W.], Surgery [M. T. L.], and Molecular Biology and Biochemistry [M. T. L.], University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15213
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The proportion of terminal deoxynucleotidyl transferase-mediated dUTP nick end
labeling-positive PBMCs undergoing spontaneous apoptosis was found to
be comparable with that induced by CH-11 antibody or TNF-.
Three-color flow cytometry revealed that CD3+
Fas+ T cells were especially sensitive to apoptosis and
were preprogrammed in vivo to die. Apoptosis occurred in
all subsets of PBMCs but was significantly higher
(P = 0.01) in the CD3+ CD8+
T-cell subset in patients relative to controls. In two patients
with melanoma, who responded clinically to dendritic cell-based peptide
vaccines, the proportion of apoptotic T cells was decreased by half
after therapy. In patients who were treated previously with
vaccination-based therapies, levels of T-cell apoptosis were lower than
in the other melanoma patients. The observed accelerated death of T
cells, which are activated and susceptible to apoptosis in patients
with melanoma, may contribute to a rapid turnover of immune cells,
resulting in a decreased immunocompetence. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
chain as well as other signaling molecules in the peripheral blood T
lymphocytes of patients with melanoma (3
, 4)
. The presence
of chain defects in T cells has been correlated with defective
signaling in T cells upon TcR cross-linking and to other functional
abnormalities (4, 5, 6)
. Although both functional
abnormalities and decreased expression of the chain are most
pronounced in TILs, the finding of similar defects in circulating T
cells of patients with cancer suggested that the influence of tumor on
immune cells extends far beyond its microenvironment (5
, 7
, 8)
. The mechanisms that are responsible for functional
deficiencies in immune cells of patients with cancer are unknown, but
it is surmised that the presence of the tumor has immunosuppressive
effects (9)
.
In examining human tumor biopsies, including melanoma, for the
presence of apoptotic cells, we observed recently that DNA
fragmentation was evident not in tumor cells but in TILs, which were
TUNEL positive in situ (5
, 10
, 11)
. Only rare
TUNEL-positive lymphocytes were present in control biopsy specimens
obtained from non-tumor sites in patients with cancer (5)
.
In addition, when fresh tumor cells or tumor cell lines were
coincubated with activated normal lymphocytes, evidence of caspase
activation and DNA fragmentation in the lymphocytes was obtained,
leading to the conclusion that normal lymphocytes, coming in contact
with human tumors, are induced to die (9
, 11
, 12)
.
Clearly, only a proportion, not all, lymphocytes are induced to
apoptose at the tumor site, and preliminary experiments suggest that
low or absent expression of the chain in TILs is a manifestation of
early apoptosis (10
, 13
, 14)
. In addition to TILs,
circulating lymphocytes in patients with cancer appear to
simultaneously undergo DNA fragmentation upon incubation in culture
medium for several hours, as we reported recently for patients
with HNC (15)
.
To determine whether spontaneous ex vivo apoptosis of circulating PBMCs is a general phenomenon in cancer and to explore its nature, we studied a cohort of patients with advanced melanoma. The expectation was that ex vivo apo-ptosis would be more extensive in patients with advanced disease involving multiple organs than in those with localized or inactive disease (NED). Because some of the melanoma patients were receiving anti-melanoma peptide-based vaccines at the time of this study, their PBMCs were obtained before and after therapy to evaluate the effects of vaccination on the proportion of circulating lymphocytes undergoing spontaneous apoptosis. Overall, the results indicated that Fas+ T cells were particularly sensitive to spontaneous apoptosis in the peripheral circulation of patients with melanoma. Two distinct groups of patients were identified: those with higher levels and those with lower levels of spontaneously apoptotic CD3+ T cells. All of the patients treated previously with vaccination-based therapies were in the lower subgroup. Thus, decreased spontaneous apoptosis of PBMCs after the administration of biotherapies might be a measure of immunological recovery in response to therapy.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
lists patient
characteristics, including their sex and age, the clinical status of
each at the time of a blood donation for this study, the disease stage,
and history of previous or current therapies.
|
PBMCs.
Venous blood was obtained from patients and controls (20 ml) in the
morning and collected in heparinized tubes. Blood samples were
hand-carried to the laboratory and immediately processed by
Ficoll-Hypaque gradient centrifugation. PBMCs were recovered from the
gradient interface, washed in DPBS, counted in trypan blue, and
promptly used for experiments.
To determine the proportion of PBMCs with evidence of DNA
fragmentation, TUNEL, Annexin V binding, or caspase-3 activation assays
(see below) were performed on aliquots (from 1 x
106 to 2 x 106
cells/assay) of the patients’ or control cells immediately after their
isolation (i.e., at time 0). The remaining PBMCs were
divided into aliquots (3.5 x 106 cells),
and each aliquot was incubated for 24 h at 37°C under the
following conditions: (a) medium alone; (b)
medium plus CH-11 antibody (agonistic anti-Fas antibody; UBI
Biotechnology, Lake Placid, NY) at a concentration of 400 ng/ml;
(c) medium plus recombinant TNF- (Knoll Pharmaceuticals,
Whippany, NJ) used at the concentration of 50 ng/ml. After incubation,
cells were harvested and immediately tested in TUNEL (or in some cases
also in Annexin V binding or caspase-3 activation) assays for evidence
of apoptosis. The proportions of PBMCs that underwent spontaneous
apoptosis (medium alone) or induced apoptosis (medium plus CH-11
antibody or medium plus TNF-) during the 24-h incubation period were
determined by flow cytometry. The medium used for all experiments was
RPMI 1640 supplemented with 1 mM
L-glutamine, 10% (v/v) of FCS, 100 µ/ml
penicillin, and 100 µg/ml streptomycin. These reagents were purchased
from Life Technologies, Inc. (Grand Island, NY).
Staining for Flow Cytometry.
PBMCs (3.5 x 106) were washed in DPBS (Life
Technologies), divided into 5 x 105 cell
aliquots, and individually incubated in the presence of the following
PE-labeled monoclonal antibodies: anti-CD3, anti-CD4, anti-CD8,
anti-CD14, anti-CD16, anti-CD56, anti-CD19, and IgG1 or IgG2b isotype
controls (all from Becton Dickinson, San Jose, CA) for 45 min on ice.
All antibodies were pretitered on normal PBMCs to determine their
optimal dilutions. After incubation, the PBMCs were washed twice in
DPBS containing 0.1% BSA and 0.1% NaN3 and
fixed with 2% (w/v) paraformaldehyde in DPBS for 30 min at room
temperature prior to flow cytometry.
To determine the percentage of PBMCs expressing Fas or TNFR1, staining was performed with ZB4 anti-Fas monoclonal antibodies purchased from Upstate Biotechnology (Lake Placid, NY) and anti-TNFR1 monoclonal antibodies purchased from R&D systems (Minneapolis, MN). FITC-conjugated goat antimouse IgG used as a negative control was obtained from Caltag (South San Francisco, CA).
TUNEL Assay.
The TUNEL assay was performed to study DNA fragmentation in
freshly harvested or ex vivo incubated cells. PBMCs were
washed after fixation with paraformaldehyde and permeabilized with
0.1% (w/v) sodium citrate in phosphate saline buffer containing
0.1% Triton X-100 for 3 min on ice. After washing, cells were
incubated with FITC-conjugated dUTP in the presence of terminal
deoxynucleotidyl transferase enzyme solution for 1 h at 37°C,
using reagents purchased from Boehringer Mannheim Corp. (Indianapolis,
IN). After incubation, the cells were washed, and 10,000 events were
acquired and analyzed by two-color analysis, using FACScan (Becton
Dickinson, San Jose, CA). Controls included cells incubated without the
enzyme in labeling buffer.
Annexin V Binding Assay.
T cells were initially stained with labeled antibodies to surface
markers as described above. Surface staining was performed on ice. The
cells were resuspended in 100 µl of Annexin V binding buffer without
NaN3 (PharMingen) and incubated with 1 µl of
Annexin V-FITC (PharMingen) for 20 min at room temperature. Next, a
400-µl aliquot of Annexin V binding buffer was added to each tube,
and the flow cytometry analysis was performed within 60 min, gating on
lymphocytes by FSC/SSC.
Caspase-3 Activation Assay.
A fluorescence-based assay, using Phi Phi Lux as a substrate, was
performed to measure caspase-3 activity in PBMCs, which were freshly
harvested from normal donors and patients with melanoma. The assay was
performed immediately after a phlebotomy and PBMC separation, using a
kit purchased from OncoImmunin, Inc. (Gaithersburg, MD) according to
the instructions provided by the manufacturer. In some cases, the assay
was also performed after ex vivo incubation of PBMCs. The
flow cytometry results were analyzed after 10,000 events were acquired.
The normal levels of caspase-3 activity were determined using PBMCs of
17 normal donors. As a positive control, PBMCs incubated with CH-11
antibody (400 ng/ml) for 4 h were used.
Flow Cytometry.
Two-color flow cytometry analysis was performed on a FACScan
equipped with a single 488-nm argon ion laser. To reliably detect small
subpopulations of cells, at least 10,000 events were acquired for each
sample. The following settings were used: 540 V and 640 V on
photomultiplier tubes for FL1 (FITC-TUNEL) and FL2 (PE), respectively.
FITC and PE fluorescence were measured through 530/30 and 585/42
bandpass filters. Compensation for FL1-FL2 and FL2-FL1 was 0.7%.
Threshold was set on 30 for forward scatter.
The percentages of apoptotic cells were calculated by scoring TUNEL+, Annexin V+, or caspase-3+ cells in two-color gated subpopulations: CD3+ apoptosis+, CD3- apo-ptosis+ in comparison with CD3+ apoptosis- and CD3- apoptosis-. All of the gated mononuclear cell subpopulations were visualized on forward angle scatter/side angle scatter dot-plots. To include all apoptotic cells and avoid debris with a high SSC signal, the gate was set to include a wide boundary of mononuclear cells ("open gate"), because apo-ptotic cells accumulated mainly in the lower FSC/SSC channels (16) . The strategy of double gating eliminated a majority of debris, which usually overlaps with apoptotic cells. The level of debris within the gated populations did not exceed 10%, and it was not significantly higher in samples with high than low (control) apoptotic values. MFI was determined for all gated cell populations.
Cytospin Preparation.
PBMCs obtained from patients and normal control samples were doubly
stained for TUNEL and CD3 surface marker, as described above. Stained
cells were centrifuged onto glass slides, using a cytocentrifuge
(Cytospin 2; Shandon) and directly mounted with fluorescent mounting
medium (Dako). The slides were examined independently by two observers,
using a fluorescent microscope (Microphot-FX; Nikon), and photographs
of representative fields were taken, using appropriate filters.
Statistical Analyses.
The Wilcoxon test was used to evaluate differences between melanoma
patients and normal controls. All within-patient or within-control
differences between assay results at time zero and each of the three
24 h assays (spontaneous, as well as induced by CH-11 antibodies
or TNF-) were tested using the signed rank test. P =
0.01 was chosen as being significant because of the large number of
statistical tests performed and the need to guard against
false-positive results.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, only a very low proportion of
PBMCs was found to be TUNEL+ at time 0 in
patients (mean ± SD, 3.2 ± 1.8%) or in controls (mean ± SD, 1.9 ± 1.1%), with a significant difference
(P = 0.0264) between the patient and control values.
When PBMCs were incubated in medium alone at 37°C for 24 h, the
proportion of cells that were TUNEL+ increased in
both control (10 ± 4.7%) and patients’ (17.2 ± 8.2%)
PBMCs. Consistently, the percentage of spontaneously apoptotic
(TUNEL+) PBMCs was significantly greater
(P = 0.0027) in patients’ than normal PBMCs (Table 2
;
Fig. 1A).
The fraction of
TUNEL+ PBMCs, i.e., PBMCs with
evidence of spontaneous apoptosis after 24-h incubation ex
vivo ranged from 1 to 33% in patients with melanoma and from 4 to
20% in normal controls (Table 2
; Fig. 1A).
|
|
was determined
by flow cytometry, it became apparent that relative to the total
percentage of TUNEL+ PBMCs, the percentage of
apoptotic T cells was consistently lower (Table 2)
. Spontaneously
apoptotic CD3+ T cells generally accounted for
less than half of the total PBMCs, with evidence of DNA fragmentation
in most of the patients with melanoma, an indication that cells other
than CD3+ were dying as well (Table 2)
. When
Annexin V binding or caspase-3 activation were measured by flow
cytometry in CD3+ T cells at time 0, consistently
the patients had increased percentages of apoptotic T cells relative to
control values, as illustrated in Fig. 2
.
The results indicated that in the peripheral circulation, more
CD3+ T cells were undergoing apoptosis in
patients with melanoma than in normal donors. Similarly, a
significantly higher proportion of CD3+ T cells
was TUNEL+ in patients than controls after
24 h incubation in medium alone (Fig. 1B;
P = 0.0001).
|
and Fig. 1
, it is important to note that the data obtained for
age-matched controls were not significantly different from those
obtained for younger laboratory controls. Also, the statistical
analysis performed to establish the correlation between the patients’
age and the proportion of TUNEL+ PBMCs after
24 h incubation indicated that spontaneous apoptosis in PBMCs was
not dependent on age in patients with melanoma (data not shown).
Induced Apoptosis in PBMCs of Patients with Melanoma.
To investigate the mechanisms responsible for the observed
greater sensitivity to apoptosis of the PBMCs obtained from
patients than those from controls, we used agonistic anti-Fas antibody,
CH-11, and TNF- to induce apoptosis ex vivo in these
PBMCs. After 24 h of incubation in medium supplemented with
anti-Fas CH-11 antibody or TNF-, the patients’ PBMCs contained a
significantly higher proportion (P = 0.0001 for both
conditions) of TUNEL+ cells than similarly
treated normal control cells (Fig. 1)
. This finding implied that the
patients’ cells were enriched in Fas+ and
TNFR1+ cells, and that these activated PBMCs were
predestined to apoptose.
Preferential Apoptosis of Fas+ and
TNFR1+ PBMCs.
To examine the possibility that more circulating PBMCs expressed
Fas or TNFR1 in patients than controls, we measured by flow cytometry
the proportion and/or MFI of Fas+ and
TNFR1+ cells in PBMCs of a subset of patients
with melanoma (n = 10) and normal controls
(n = 10). The TNFR1 was expressed mainly on monocytes,
with mean values of 20% in patients versus 9% in controls.
The MFI for TNFR1 on monocytes was considerably higher in patients than
in controls (Fig. 3A).
The
representative flow cytometry data shown in Fig. 3A
are
consistent with the increased expression of TNFR1 in PBMCs of patients
with melanoma relative to PBMCs of normal controls. Similarly, Fas
expression was seen on a significantly greater proportion of PBMCs in
melanoma patients (Fig. 3B).
Overall, 68 ± 8%
(mean ± SD) of PBMCs in patients were found to be
Fas+ compared with 44 ± 15% in controls
(P < 0.05; n = 10).
|
The data obtained using
PBMCs of a representative patient with melanoma indicated that at time
0, 11% of Fas+ CD3+ T
cells were apoptotic (i.e., showed caspase-3 activation) as
opposed to only 4% of Fas-
CD3+ T cells (Fig. 4A).
Similarly,
using the Annexin V binding assay, 19% of Fas+
CD3+ T cells were apoptotic, in contrast to only
11% of Fas+ CD3+ T cells
at time 0 (Fig. 4A).
When the patients T cells were
incubated in medium for 24 h prior to the caspase activation or
Annexin V binding assays, 21 and 25% of T cells, respectively, were
apoptotic among Fas+ T cells versus
13% (for both assays) among Fas-
CD3+ T cells (Fig. 4A).
The shift to
the right in Annexin V+ T cells at 24 h is
accentuated by the appearance of a small, distinct subset of intensely
Annexin V+ T cells (Fig. 4B),
which
seem to be in the process of losing CD3. Fig. 4B
illustrates
the shift in the MFI of Annexin V+
CD3+ T cells that occurs between time 0 and
24 h of incubation. On the basis of these results, it can be
surmised that in patients with melanoma, who have a significantly
higher proportion of Fas+ T cells than controls,
more spontaneous apoptosis occurs. Indeed, with or without incubation
of these cells in medium alone, apoptosis was significantly higher in a
Fas+ fraction of T cells in patients
(P = 0.0027) than in normal controls. In aggregate, our
data point to the Fas-mediated mechanisms of T-cell elimination in
patients with melanoma. Preliminary evidence suggests that activated
mononuclear cells expressing Fas or TNFR1 are the population which is
sensitive to ex vivo apoptosis, because a significant linear
relationship (P = 0.01) exists between the percentage
of Fas+ CD3+ T cells and
CD3+ T cells with elevated caspase-3 activity in
patients with melanoma (data not shown).
|
, it became apparent that
two distinct patient populations could be identified. One patient group
had a lower percentage of apoptotic T cells, not exceeding the normal
mean + 3 SD, whereas the other group had a high percentage of apoptotic
T cells (i.e., more than normal mean + 3 SD; Fig. 5
). The
mean percentage + 3 SD T cells induced to apoptose by CH-11 antibody
was calculated to be 17.9%. There were no readily identifiable
clinical differences between these two groups of patients. However, in
the group with lower levels of apoptosis, 9 of 11 patients were treated
previously with one or more of the following biotherapies:
peptide-based vaccines (n = 7), a tumor-based vaccine
then a genetically modified vaccine and gene therapy (n = 1); or a genetically modified vaccine (n = 1). Only
one patient of nine was concurrently treated with a vaccine-based
therapy in the group of patients with high levels of T-cell apoptosis.
|
, we used an open gate to evaluate both
apoptotic and nonapoptotic lymphocytes. Of the two distinct T-cell
subpopulations: CD3+
TUNEL- (black) and
CD3+ TUNEL+
(gray), the latter was increased in the patient’s PBMCs
relative to normal control. It is important to note that the proportion
of CD3+ TUNEL- T cells
(black) was decreased in this patient’s PBMCs. When the
proportion of CD3+ TUNEL-
T-cell subpopulation in PBMCs of patients versus controls
was compared (Table 3)
, a significant
decrease (P < 0.0001) in nonapoptotic T cells was
noted in the patients with melanoma. At the same time, the proportion
of CD3+ TUNEL+ cells was
not significantly larger in the patients than in controls (Table 3)
.
Because flow cytometry indicated that a shift from
TUNEL- to TUNEL+ T cells
occurs during the incubation period, we suspected that a loss in CD3
expression on the cell surface of apoptotic T cells could explain these
results. When the proportion of CD3-
TUNEL+ cells was determined, we observed an
increase in this cell population in patients as compared with controls
(Table 3)
, which was of borderline significance (P =
0.0374). The data are consistent with a shift toward lower expression
of CD3 in T cells of the patients versus controls (Fig. 6, B and D).
This shift was more apparent in the
patient than control samples, and it indicated that a loss of the
CD3-
chain from the cell surface occurred during 24-h incubation of
patients’ lymphocytes in medium, which might reflect spontaneous
apoptosis.
|
|
Flow Cytometry Analysis of CD3- TUNEL+
Cells.
We observed consistently that DNA fragmentation ex vivo
occurred not only in CD3+ lymphocytes but also in
PBMCs that were CD3-. Therefore, two-color flow
cytometry was used to discriminate between these two populations and to
identify the CD3- subsets that became
TUNEL+ during the period of 24 h incubation
of PBMCs in medium alone. The representative flow cytometry data shown
in Fig. 7
indicate that: (a)
spontaneous apoptosis was considerably greater at 24 h relative to
0 time in PBMCs of the patient versus the accompanying
control; and (b) CD3+ T cells were the
main lymphocyte subsets undergoing spontaneous apoptosis in the PBMCs
obtained from the patients. The CD3-
CD56+ natural killer cells,
CD19+ B cells, and CD14+
monocytes were lesser contributors to spontaneous apoptosis (Fig. 7)
.
|
, the
representative results for two patients and the accompanying control
show that: (a) for all of the PBMC subsets, the proportion
of TUNEL+ cells is higher in patients than the
control; and (b) CD4+ and
CD8+ T lymphocyte subsets constitute the largest
proportions of spontaneously apoptotic PBMCs in both patients with
melanoma. The cumulative data presented in Table 4
suggest that although all of the PBMC
subsets were more sensitive to ex vivo apoptosis in the
patients than in controls, only the value for the
CD8+ T-cell subset (P = 0.01) was
significantly different in patients versus controls. These
data suggest that CD8+ immune effector cells are
particularly sensitive to spontaneous ex vivo apoptosis in
patients with melanoma. Incidentally, after incubation of PBMCs in
the presence of TNF-, apoptosis was induced in a significantly
higher proportion of the CD14+ cells (monocytes)
in patients with melanoma than in controls (P =
0.0083). The data are consistent with the higher level of TNFR1
expression on monocytes in patients than in normal controls, as
illustrated in Fig. 3A.
|
|
. All four were enrolled in the Phase I
clinical trial of a DC-based vaccine with a mixture of melanoma
peptides pulsed on autologous DCs. Patient 8 was a complete responder
to the vaccine. Patient 13 completed three courses of the vaccine and
had a favorable (partial) clinical response. Patients 9 and 12 had
disease progression. The data on spontaneous apoptosis of the PBMCs
obtained in the course of the protocol or follow-up of these patients
are summarized in Table 5
.
|
Clinical Correlations.
The patients’ sex, age, disease stage, or disease activity (NED
versus active disease) were examined for association with
apoptosis. None of the clinical characteristics were significantly
correlated with spontaneous or induced apoptosis in PBMCs or T
lymphocytes (Fig. 9)
. However, as shown
in Fig. 5
, the subgroup of nine melanoma patients treated previously
with vaccination-based therapies had lower levels of spontaneous
apoptosis compared with those determined in the 11 other patients who
were not treated with similar therapies.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To explain the high levels of spontaneous apoptosis in PBMCs of
patients with melanoma, we hypothesized that more cells expressed Fas
or TNFR1 in these patients relative to controls. The hypothesis was
based on the observation made initially that the high level of
spontaneous ex vivo apoptosis in PBMCs of patients with
melanoma was not different from that induced by CH-11 antibody or
TNF-. This finding was consistent with the conclusion that increased
expression by the patients’ PBMCs of Fas or TNFR1, relative to lower
expression of these receptors by normal PBMCs, was responsible for
greater sensitivity of these patients’ cells to apoptosis. Indeed,
flow cytometry confirmed increased expression of the two receptors on
the surface of PBMCs in patients with melanoma, as compared with normal
controls. Furthermore, using patients’ PBMCs, it was possible to
demonstrate that Fas+ CD3+
T cells preferentially undergo apoptosis, as evidenced by increased
endogenous caspase-3 activity in these cells. Also, a direct and
significant (P = 0.01) correlation between the
percentage of Fas+ T cells and T cells with
increased caspase-3 activity was established for patients with melanoma
in preliminary experiments. Although the reason for this increased
expression of TNF family of "death receptors" on mononuclear cells
of patients with melanoma remains unknown, it is reasonable to
speculate that it is related to the presence of tumor. Tumor-induced
activation of immune cells, resulting in up-regulated expression of the
"death receptors " on these cells, could lead to their apoptosis.
We have reported previously that a high proportion of TILs were
TUNEL+ in human tumors (5
, 9
, 10)
.
Because TILs consist mainly of activated T cells (19)
, it
was reasonable to conclude that these T cells were induced to apoptose
in the tumor microenvironment (9
, 10) . It now appears that
the effect of tumor on immune cells can extend beyond the tumor site,
and that systemic effects might result in a high rate of apoptosis in
chronically activated PBMCs of patients with cancer.
Ex vivo apoptosis of PBMCs observed in patients with melanoma included not only T lymphocytes but all subsets of mononuclear cells, although only CD8+ T cells were found to be significantly depleted. This finding is consistent with preliminary in situ data, which show that more CD8+ than CD4+ TILs undergo apoptosis in human tumor biopsies.4 We have particularly focused on T lymphocytes because of the previous data reported from our and other laboratories that functional impairments, detectable in TILs as well as circulating T lymphocytes, were associated with the presence of tumor (reviewed in Refs. 9 and 19 ). In our hands, coincubation of activated T cells with tumor cells induced a variety of functional impairments, including decreased or absent expression of TcR-associated signaling molecules, and led to activation of caspases and to apoptosis in these T cells (10 , 13 , 20) . More recently, tumor-specific T cells captured and isolated from PBMCs of patients with melanoma by peptide-specific tetramers were found to be selectively nonfunctional (21) . The explanation for these findings may be that activated tumor-specific T lymphocytes become susceptible to activation-induced cell death or tumor-induced cell death or both. The distinction between these two mechanisms of T-cell death was recently discussed by Chappell and Restifo (22) . In principle, however, the two mechanisms cannot be distinguished in a tumor setting. Human tumors are able to induce activation of tumor-specific T cells via the MHC-restricted TcR-dependent recognition, which culminates in activation-induced cell death (23) . On the other hand, human tumors that express Fas ligand (13 , 24, 25, 26) could directly induce death of activated Fas+ immune cells via the Fas/Fas ligand pathway (9 , 13) . Non-MHC-restricted immune cells in the tumor microenvironment are activated through interactions not involving TcR and likely to up-regulate expression of Fas, TNFR1, or other death receptors. Tumor-induced cell death might be a consequence of such activation, especially when a tumor expresses Fas ligand (13 , 24 , 25) or when TNF is present in the tumor microenvironment. Although controversy exists about expression of Fas ligand on human melanomas (27) , it is still possible that other mechanisms, not involving the death receptors, could directly induce apoptosis in immune cells (28) .
The high levels of ex vivo apoptosis seen in PBMCs of patients with melanoma could be adequately explained by the mechanism, which involves a rapid turnover of circulating mononuclear cells through their death, removal from the circulation of preapoptotic cells, and increased sequestration and utilization of precursors from the bone marrow stores. In practical terms, this means that PBMCs are dying and are replaced at a much higher rate in patients with cancer than in controls. A similar mechanism has been invoked for patients with HIV infections (29 , 30) . The rationale for this mechanism is based on our observations that: (a) implicate effector cell subsets of circulating PBMCs in tumor-induced apoptotic death; (b) demonstrate that more PBMCs in patients with cancer than normal volunteers express death receptors, which implies that more PBMCs are activated and susceptible to apoptosis; (c) show that spontaneous or induced apoptosis in the CD8+ subset of PBMCs of many patients with cancer is significantly greater than that in healthy controls; and (d) indicate that Fas+ (activated) T cells are predestined in vivo to die. The future challenge will be to determine how to control this rapid turnover and to increase survival of antitumor effector cells in patients with cancer.
In this study, only 55% of patients with advanced melanoma showed highly elevated levels of ex vivo apoptosis in T lymphocytes. Using the conventional clinical criteria (disease activity, stage, and organ involvement), this group of patients could not be distinguished from the group of patients with lower levels of spontaneous apoptosis. Furthermore, the statistical analysis of patients classified into groups based on sex, age, disease stage, or activity did not indicate any significant differences in the level of spontaneous or induced apoptosis of PBMCs or T cells. Thus, it remains unclear why T lymphocytes in some patients with melanoma have a greater predisposition to apoptosis than in others, and further studies are necessary to determine the potential prognostic importance of this observation. However, an important clue is provided by the finding that 9 of 11 patients with lower levels of T-cell apoptosis received prior vaccination-based therapies. Also, an increase in the proportion of circulating CD3+ T cells accompanied by a substantial decrease in the proportion of spontaneously apoptotic PBMCs occurred in a clinical responder to vaccination therapy with a DC-based melanoma peptide vaccine. These preliminary results suggest that serial assessments of ex vivo spontaneous apoptosis might correlate with or predict a favorable response to biological therapy. Because few, if any, reliable immunological in vitro correlates of a clinical response to therapy exist in cancer, our preliminary data are intriguing and provide a strong rationale for further serial studies of spontaneous apoptosis in melanoma patients treated with biotherapies.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported in part by NIH Grant P01 DE 12321 (to
T. L. W.). 
2 To whom requests for reprints should be
addressed, at University of Pittsburgh Cancer Institute, W1041
Biomedical Science Tower, 211 Lothrop Street, Pittsburgh, PA
15213-2582. Phone: (412) 624-0096; Fax: (412) 624-0264; E-mail: whitesidetl{at}msx.upmc.edu
3 The abbreviations used are: TcR, T-cell
receptor; TIL, tumor-infiltrating lymphocyte; TUNEL, terminal
deoxynucleotidyl transferase-mediated dUTP nick end labeling; HNC, head
and neck cancer; PBMC, peripheral blood mononuclear cell; NED, no
evidence of disease; DPBS, Dulbecco’s PBS; TNF, tumor necrosis factor;
TNFR, TNF receptor; PE, phycoerythrin; FSC/SSC, forward angle
scatter/side angle scatter; MFI, mean fluorescence intensity; DC,
dendritic cell.
4 T. E. Reichert and T. L. Whiteside,
unpublished data.
Received 10/26/99; revised 1/16/00; accepted 1/18/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
chain in T cells at the tumor site correlates with poor survival in patients with oral carcinoma. Cancer Res., 58: 5344-5347, 1998. chain. Cancer Res., 59: 1422-1427, 1999.This article has been cited by other articles:
|
|
 |
|
  T. Das, G. Sa, E. Paszkiewicz-Kozik, C. Hilston, L. Molto, P. Rayman, D. Kudo, K. Biswas, R. M. Bukowski, J. H. Finke, et al. Renal Cell Carcinoma Tumors Induce T Cell Apoptosis through Receptor-Dependent and Receptor-Independent Pathways J. Immunol., April 1, 2008; 180(7): 4687 - 4696. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Bhattacharyya, D. Mandal, B. Saha, G. S. Sen, T. Das, and G. Sa Curcumin Prevents Tumor-induced T Cell Apoptosis through Stat-5a-mediated Bcl-2 Induction J. Biol. Chem., June 1, 2007; 282(22): 15954 - 15964. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. B. Thoren, A. I. Romero, and K. Hellstrand Oxygen Radicals Induce Poly(ADP-Ribose) Polymerase-Dependent Cell Death in Cytotoxic Lymphocytes. J. Immunol., June 15, 2006; 176(12): 7301 - 7307. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-W. Kim, R. L. Ferris, and T. L. Whiteside Chemokine C Receptor 7 Expression and Protection of Circulating CD8+ T Lymphocytes from Apoptosis Clin. Cancer Res., November 1, 2005; 11(21): 7901 - 7910. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Takahashi, M. G. V. Hanson, H. R. Norell, A. M. Havelka, K. Kono, K.-J. Malmberg, and R. V. R. Kiessling Preferential Cell Death of CD8+ Effector Memory (CCR7-CD45RA-) T Cells by Hydrogen Peroxide-Induced Oxidative Stress J. Immunol., May 15, 2005; 174(10): 6080 - 6087. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Wysocka, B. M. Benoit, S. Newton, L. Azzoni, L. J. Montaner, and A. H. Rook Enhancement of the host immune responses in cutaneous T-cell lymphoma by CpG oligodeoxynucleotides and IL-15 Blood, December 15, 2004; 104(13): 4142 - 4149. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Rayman, A. K. Wesa, A. L. Richmond, T. Das, K. Biswas, G. Raval, W. J. Storkus, C. Tannenbaum, A. Novick, R. Bukowski, et al. Effect of Renal Cell Carcinomas on the Development of Type 1 T-Cell Responses Clin. Cancer Res., September 15, 2004; 10(18): 6360S - 6366S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-W. Kim, T. Tsukishiro, J. T. Johnson, and T. L. Whiteside Expression of Pro- and Antiapoptotic Proteins in Circulating CD8+ T Cells of Patients with Squamous Cell Carcinoma of the Head and Neck Clin. Cancer Res., August 1, 2004; 10(15): 5101 - 5110. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Boissonnas, C. Combadiere, E. Lavergne, M. Maho, C. Blanc, P. Debre, and B. Combadiere Antigen Distribution Drives Programmed Antitumor CD8 Cell Migration and Determines Its Efficiency J. Immunol., July 1, 2004; 173(1): 222 - 229. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
