| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Medicine, University of Wales, College of Medicine, Cardiff, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
production
against HCMV pp65 and immediate early 1 Ags following in vitro
coculture with HCMV-AD169-infected Fbs, regardless of the HLA type of
these Fbs. CD8+ T cell stimulation was inhibited by
pretreatment of DCs with cytochalasin B or brefeldin A, indicating a
phagosome/endosome to cytosol pathway. HCMV-infected Fbs were not
apoptotic as measured by annexin V binding, and induction of apoptosis
of infected Fbs in vitro did not augment CTL induction by DCs,
suggesting a mechanism other than apoptosis in the initiation of
cross-presentation. Furthermore, HCMV-infected Fbs provided a
maturation signal for immature DCs during coculture, as evidenced by
increased CD83 and HLA class II expression. Cross-presentation of HCMV
Ags by host DCs enables these professional APCs to bypass some of the
evasion mechanisms HCMV has developed to avoid T cell recognition. It
may also serve to explain the presence of immediate early 1 Ag-specific
CTLs in the face of pp65-induced inhibition of Ag presentation at the
level of the infected cell. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-herpesvirus, a large DNA virus encoding >227
proteins (1). It is a well-adapted virus, establishing
lifelong asymptomatic persistence. HCMV has coevolved with the human
immune system and developed multiple independent mechanisms to evade
recognition by CD8+ CTLs (2, 3, 4). One
of the main mechanisms of evasion is the direct interference with HLA
class I Ag presentation by viral glycoproteins. There are at least four
HCMV proteins that interfere with the assembly, maturation, and
transport of HLA class I molecules. The immediate early (IE)
gene product gpUS3 causes retention of HLA class I molecules in
the endoplasmic reticulum (ER), and the late Ag gpUS6 inhibits TAP
translocation following peptide binding, whereas gpUS2 and gpUS11
translocate assembled class I molecules back into the cytoplasm where
they are rapidly degraded. Due to this blockade, HCMV-infected human
fibroblasts (Fbs) have significantly reduced levels of HLA class I
expression, impaired Ag presentation, and are resistant to lysis by
CD8+ T cells (5). The US2 Ag also
down-regulates HLA class II expression, leading to a reduced T
cell-stimulatory capacity (6). During HCMV infection,
phosphorylation of the abundantly expressed 72-kDa IE I Ag has
been observed (7). This is mediated by the pp65 tegument
protein of HCMV released from virions into the cytosol. This
phosphorylation interferes with the presentation of IE1-derived Ags via
the HLA class I pathway, and, as a result, pp65- and IE-expressing
cells are not recognized and lysed by IE-specific T cells, although
they can be killed by T cells directed against other viral Ags
(7). Furthermore, HCMV encodes three proteins with
homology to CC chemokine receptors: US27, US28, and UL33
(8), which may deplete chemokines from the environment of
infected cells. The reduced local availability of chemokines can have a
negative effect on the attraction of lymphocytes and activation of
CD8+ T cells (9). HCMV also encodes
a viral homologue of IL-10 (10), which may have
immunosuppressive properties favorable to the virus. However, the most
important evasion strategy of HCMV is probably the infection of bone
marrow hemopoietic progenitor cells early in their development
(11, 12, 13, 14), where the virus remains quiescent with limited
nonlytic gene expression. At this stage, the virus is invisible to the
immune system.
Despite the numerous viral evasion strategies, HCMV-specific effector
or memory CD8+ and CD4+ T
cells are present in the peripheral blood of healthy seropositive
carriers at relatively high frequencies, as demonstrated by limiting
dilution analysis (15, 16, 17, 18), intracellular cytokine
staining (19, 20), and by peptide-HLA tetramer staining
(21). To explain how this robust cellular immune response
develops in the presence of viral interference with the HLA class I and
class II Ag presentation pathways, we examined the possibility that
cross-priming via dendritic cells (DCs) enables Ag presentation of
epitopes that would otherwise be blocked in the virally infected cell.
DCs can acquire Ags from apoptotic tumor cells (22) or
influenza virus-infected apoptotic monocytes (23) and
present these via the MHC class I pathway. However, DNA viruses,
including HCMV, inhibit apoptosis of the infected cells (reviewed in
Ref. 24). HCMV IE1 and IE2 proteins inhibit apoptosis by
modulating the expression of cellular proteins, e.g., NF-
B subunits
(25, 26), whereas the product of the UL37 gene, vMIA
(viral mitochondria-localized inhibitor of apoptosis), has broad
anti-apoptotic effects, inhibiting apoptosis at a point downstream
of caspase 8 activation in a bcl-2-related manner
(27).
In this study, we explored whether immature DCs can acquire viral Ags from HCMV-infected Fbs and whether this process is triggered by apoptosis. Our results indicate that cross-presentation of HCMV Ags occurs by DCs in the absence of apoptotic markers on infected cells. This Ag presentation pathway efficiently bypasses some of the T cell evasion strategies of HCMV and may serve to explain how T cell responses against certain HCMV Ags are generated.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Four healthy HCMV-seropositive and three -seronegative
laboratory volunteers were included in this study. Their HCMV
serostatus was determined by high-sensitivity IgG ELISA (Department of
Medical Microbiology, University Hospital of Wales, Cardiff, U.K.). The
HLA types of the donors and the Fbs used in these experiments were
determined by microlymphocyte cytotoxicity assay (Welsh Blood
Transfusion Service, Cardiff, U.K.) or by PCR using sequence-specific
primers (28) (Table I
). MRC5
cells (human embryonic lung Fbs; European Collection of Cell
Cultures, Salisbury, U.K.) and human skin Fbs prepared from
laboratory volunteers (29) were grown in MEM (Life
Technologies, Grand Island, NY) supplemented with 10% FCS (Life
Technologies), 100 IU/ml penicillin, 100 µg/ml streptomycin, 2 mM
L-glutamine, 25 mM HEPES buffer, and nonessential amino
acids. T2 cells (HLA-A2, TAP-negative B cell-derived cell line used in
peptide-specific killing assays) were grown in RPMI 1640 supplemented
as described above except without nonessential amino acids. B
lymphoblastoid cells (BLCL) were prepared according to the standard
method by infecting PBMC with EBV containing B95.8 cell supernatant.
All cells and virus stocks were negative for contamination with
mycoplasma as determined by immunofluorescent staining with a mAb
(CCM-2; ICN Pharmaceuticals, Costa Mesa, CA).
|
HCMV strain AD169 is a laboratory strain adapted to grow in Fbs
(30). RCMV288 is a mutant strain based on strain AD169
that has a copy of enhanced green fluorescent protein (EGFP) inserted
in one copy of the HCMV long repeat under the control of the HCMV
2.7 early promoter. This insertion does not incapacitate the virus
and provides a convenient reporter system for monitoring infection
(31). HCMV strain TB40/E (kindly provided by Dr. C.
Sinzger, University of Tübingen, Tübingen, Germany)
is an endothelial cell (EC)-adapted HCMV strain that maintained its
ability to infect ECs and other cell types including DCs
(32). HCMV strains were propagated in MRC5 cells. The
supernatant of the infected Fbs was collected by centrifugation when
cells exhibited gross cytopathic effect. The filtered supernatant,
stored at -70°C in small aliquots, was used for the experiments. The
virus titers were determined from the tissue
culture-ID50 values on MRC5 cells.
Recombinant adenovirus (rAd)31 is a replication-deficient rAd that
contains the HCMV major immediate early gene (IE1) under the control of
its own promoter (33), whereas rAd60 is an equivalent
(control) adenovirus without a transgene insertion. The adenoviruses
were propagated in complementary 293 cells (33), and small
aliquots of the virus stocks were kept at -70°C.
The HLA-A2-restricted peptide (493-ARNLVPMVATVQGQN-507) (18) of the HCMV pp65 tegument protein was synthesized by the Peptide and Protein Facility (University of Wales, College of Medicine, Cardiff, U.K.). The HLA-B7-restricted peptide (417-TPRVTGGGAM-426) (19) was synthesized by Immune Systems, Paignton, U.K.). Both peptides were >80% pure by HPLC analysis.
In vitro culture of human DCs
PBMC were obtained from peripheral blood by standard Ficoll-Histopaque (Life Technologies) separation. The adherent cells were isolated in serum-free RPMI 1640 at 15 x 106 cells/well in 6-well trays following a 1-h incubation at 37°C and were differentiated into DCs according to the standard method (34). Briefly, adherent cells were grown in RPMI 1640 containing 10% FCS and supplements in the presence of 50 ng/ml human rGM-CSF (Leucomax; Novartis Pharmaceuticals, East Hanover, NJ) and 500 U/ml human rIL-4 (BD PharMingen, San Diego, CA). Fresh culture medium was added on day 3. After 6 days, nonadherent and loosely adherent cells were collected by vigorous pipetting. More than 90% of the cells were of DC phenotype (CD1a+, HLA-DR+, CD14-, CD80+), after gating on size and side scatter by FACS to exclude lymphocytes.
Antibodies
The following Abs and fluorescent reagents were used in this
study: HCMV/IE1-specific mAb 1D6-6, HCMV p52 early Ag-specific
FITC-conjugated mAb (CCH2; Dako, Carpinteria, CA) (35),
HCMV pp65-specific Ab (BioDesign, Carmel, NY), HCMV glycoprotein B
(gB)-specific human mAb (ITC52) (36), mouse
IgG1-negative control (Serotec, Oxford, U.K.), CD3-PE (BD Biosciences,
Mountain View, CA), PKH-26 (Sigma, St. Louis, MO), CD8-CyChrome, anti
IFN--FITC, HLA class II-CyChrome, annexin V-FITC Apoptosis Detection
Kit I (all four from BD PharMingen), CD56-PE (Bioclone, Marrickville,
Australia), CD83-FITC (Immunotech, Luminy, France), and goat
anti-mouse IgG F(ab')2-FITC (Sigma).
Immunofluorescence, FACS analysis
Detection of HCMV Ags. Fibroblasts grown on 13-mm diameter coverslips were fixed with 4% paraformaldehyde, permeabilized with 0.3% Triton X-100 (both for 20 min at room temperature), and blocked with 1 µg/ml goat or mouse IgG for 20 min at 37°C. Ab labeling was conducted either with the IE1- or the pp65-specific Ab followed by goat anti-mouse IgG-FITC (for 45 min at 37°C each step) or with the FITC-conjugated HCMV p52-specific CCH2 Ab (45 min at 37°C). Infection of DCs was determined by FACS after fixing the cells with 4% paraformaldehyde, permeabilizing with 0.025% Triton X-100 (both for 20 min at room temperature), and staining with the Abs as for Fbs.
Early apoptosis detection. HCMV-infected and uninfected MRC5 cells were labeled using the annexin V-FITC Apoptosis Detection Kit I according to the manufacturer’s instructions. Briefly, 5 µl annexin V-FITC and 10 µl propidium iodide (PI) were added to 105 cells in 100 µl labeling buffer. The cells were incubated for 15 min at 37°C in the dark and analyzed on a FACSCalibur (BD Biosciences) using CellQuest 3.1 software.
Intracellular cytokine staining.
A total of 1–2 x 106 T cells/ml were mixed
with DCs and/or Fbs, as indicated in the figure legends, at a 10:1
ratio. one hour later, 1 µl/ml GolgiPlug (BD PharMingen) was added
for 5 h. The cells were washed with PBS after the total incubation
time of 6 h. The samples were first labeled for 45 min at 4°C
with CD8-CyChrome and in some experiments with CD3-PE followed by
fixing with 4% paraformaldehyde for 20 min at room temperature. The
cells were then permeabilized with 0.3% saponin for 20 min at room
temperature and nonspecific Ab binding was blocked with 1 µg/ml mouse
Ig for 20 min in the presence of 0.3% saponin at 37°C.
Anti-IFN--FITC Ab was added in the presence of 0.3% saponin for 45
min at 37°C. Labeled samples were analyzed within 1 h on a
FACSCalibur as described above.
In vitro stimulation of memory CTL responses, CTL assay
The nonadherent fraction of PBMC obtained after 1 h adherence was collected and cryopreserved for 6 days. We refer to this fraction as PBMC throughout this paper. T cell stimulation was conducted in RPMI 1640 supplemented with 10% human AB serum (Sigma), 100 IU/ml penicillin, 100 µg/ml streptomycin, 2 mM L-glutamine, and 25 mM HEPES buffer. DCs were preincubated with Fbs for 1–3 h at a 1:1 ratio (unless otherwise stated) before adding PBMC. Fbs were infected with HCMV strain AD169 (multiplicity of infection (moi) = 5 PFU/cell) for 48 h before adding them to DCs or to PBMC. Fbs in some of the early experiments were irradiated with 3000 rad; however, a comparative experiment revealed that nonirradiated Fbs were equally efficient in the cross-presentation experiments (data not shown). Thus, for easier interpretation of the results, the irradiation step was omitted. For CTL stimulation in bulk cultures, DCs or Fbs were added at a 1:10 ratio to PBMC at 2 x 106 PBMC/ml. For stimulation in 96-well trays, 103 DCs and/or HCMV-infected Fbs were added to each well of PBMC as indicated in the figure legends. Cells were collected and counted from DC/Fb/PBMC bulk cultures 6–7 days later, or cells from 96-well trays were split and added to different targets 9 days later. T cell cytotoxicity was measured in a standard 51Cr-release assay using pp65493–507 peptide-pulsed T2 cells or pp65417–426-pulsed autologous BLCL. To determine IE1-specific killing, Fbs were infected with rAd31 or rAd60 (moi = 50 PFU/cell) for 48 h, then labeled with 5 MBq 51Cr/106 cells for 1 h. A total of 1 x 103 Fbs or 3 x 103 T2 cells or BLCL were added per well into 96-well U-bottom trays. Varying numbers of effector cells in triplicates were incubated with the target cells in 200 µl final volume/well for 4 h (T2 and BLCL) or for 6 h (Fbs), respectively. A total of 25 µl supernatant from each well was harvested onto glass-fiber mats (Wallac, Milton-Keynes, U.K.) and measured on a beta plate counter (Wallac). Spontaneous lysis was usually 8–12% for T2 cells and 10–30% for Fbs. The specific lysis from each well was calculated as [(experimental release - spontaneous release)/(maximum release - spontaneous release)] x 100, where spontaneous and maximum release means target cells incubated with medium only or with 5% Triton X-100, respectively.
Inhibition of Ag presentation
DCs were treated with 5 µg/ml cytochalasin B (Calbiochem, La Jolla, CA) or with 1 µl/ml brefeldin A in the form of GolgiPlug (BD PharMingen) at 105-106 cells/ml for 30 min at 37°C. Following the treatments, DCs were washed three times and incubated overnight with infected Fbs (1:1 ratio) before adding PBMC at a 10:1 PBMC:DC ratio. Intracellular cytokine production by T cells or CTL activity was measured as described above.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Infection of DCs with the laboratory strain AD169 did not result
in the expression of detectable levels of IE Ag, neither at the
routinely used 5 moi (Fig. 1A)
nor at higher rates of infection (data not shown), as measured 6, 16,
and 44 h after infection. However, when DCs were infected under
the same conditions with strain TB40/E (moi = 5 PFU/cell), an
EC-adapted strain of HCMV, immediate early viral Ag was detected in
3.2% of DCs (Fig. 1B) as early as 6 h after infection,
with the rate of Ag-positive cells rising to 53.6% by 44 h after
infection. HCMV pp65 was undetectable early (6 h) after infection with
both strains. Expression of pp65 was first detectable at 44 h
after infection only in DCs infected with strain TB40/E, but not in
those infected with strain AD169. Furthermore, DCs, after infection
with strain AD169 (moi = 5 PFU/cell), did not stimulate
pp65-specific memory CTLs in vitro from the PBMC of seropositive donors
(Fig. 2).
|
|
To study whether DCs, while themselves remaining uninfected with
HCMV strain AD169 (Fig. 1
A), can stimulate HCMV-specific T
cell responses in the presence of cells that are productively infected
with HCMV strain AD169, DCs from seropositive donors were cocultured
with HCMV-infected Fbs and with autologous PBMC. Because the CTL
response in healthy seropositive individuals was shown to be dominated
by pp65-specific T cells (18), we measured CTL activity
against T2 target cells pulsed with the HLA-A*0201-restricted
pp65493–507 peptide. DCs cocultured with
HCMV-infected Fbs stimulated autologous, HLA-A*0201-restricted,
pp65-specific cytotoxic T cells (Fig. 2). DCs preincubated either with
HLA-A*0201-matched or with mismatched (allogeneic) Fbs generated
HLA-A*0201-restricted, pp65 peptide-specific CTL responses at similar
levels. These CTL responses were higher than that following incubation
with HCMV-infected HLA-A*0201 Fbs alone. HCMV-infected
allogeneic Fbs or DCs alone (Fig. 2) did not induce
HLA-A*0201-restricted, pp65-specific CTL responses. In identical
experiments in which PBMC were obtained from two seronegative
HLA-A*0201-positive donors (donors 6 and 7; Table I), HCMV-specific T
cell cytotoxicity was not induced (data not shown). These results
indicate that DCs are able to acquire viral Ags from infected Fbs and
present them to CD8+ cytotoxic T cells and
exclude the possibility that HCMV present in DCs and reactivated by in
vitro manipulation could induce these CTL responses.
DCs stimulate CD8+ CTLs in the presence of low numbers of HCMV strain AD169-infected Fbs
DCs are more powerful than B cells or mononuclear cells in
stimulating proliferative or cytotoxic T cell responses both in terms
of maximum response and amount of Ag required (34). To
confirm this with respect to HCMV-infected Fbs, they were added in
decreasing numbers to a constant number of DCs (2 x
105) and PBMC (2 x
106). DCs were able to stimulate significant
levels of HLA class I-restricted, HCMV pp65 peptide-specific T cells in
these cultures in the presence of as few as 2000 Fbs (Fig. 3). HCMV-infected Fbs alone were not
capable of stimulating T cells at ratios lower than 200,000 Fb/culture.
(Fig. 3, 
). These results indicate that T cell stimulation by DCs is
not simply due to high numbers of infected Fbs present in the cultures
and support a direct Ag-presenting role for DCs in these
experiments.
|
To ascertain that the observed cross-presentation was not unique
to the pp65 tegument protein, which can derive from the virion or from
dense bodies (viral particles without DNA) without de novo protein
synthesis, the CTL response to the 72-kDa IE1 Ag was also
studied. IE1 is expressed abundantly in Fbs from 4 h following
infection with HCMV, and it is a recognized T cell Ag (15, 20, 37, 38, 39). However, the presence of such CTLs in seropositive
individuals has not been fully explained, as pp65 inhibits IE1
presentation at the level of the infected cell (7).
IE1-specific CTLs were generated in bulk culture (Fig. 4, A and B) or in
limiting dilution culture (Fig. 4C) from PBMC of
seropositive donors. The results were similar to those obtained with
pp65: only a low level of IE1-specific CTLs were induced by
HCMV-infected Fbs alone (11% at 50:1 ratio; Fig. 4A),
whereas the CTL activity was 42% following coculture of DCs with HCMV
strain AD169-infected Fbs (Fig. 4A). When PBMC were cultured
in limiting dilution, so that T cell responses at clonal level could be
examined (Fig. 4C), fewer wells containing HLA class
I-restricted IE1-specific T cells were generated at all cell numbers
studied (1–4 x 104 PBMC/well) when
infected Fbs alone were added, compared with cultures which also
contained DCs. The higher levels of specific lysis observed from the
individual wells following stimulation of T cells with DCs plus
infected Fbs (Fig. 4C, •) compared with those with
infected Fbs alone (Fig. 4C, 
) may reflect enhanced cell
proliferation, resulting in larger clone sizes per well. These results
indicate that DCs probably acquire and process several viral Ags from
infected cells for Ag presentation.
|
The enhanced pp65- and IE1-specific CD8+ CTL
induction by DC in the cytotoxicity experiments (
Figs. 2–4) was also
tested by measuring IFN- production by T cells following stimulation
of PBMC from seropositive donors with DCs and/or partially infected
HLA-matched Fbs (Fig. 5). This approach
allowed us to quantitate the total number of T cells stimulated by DCs
and/or infected Fbs. Whereas HCMV AD169-infected Fbs stimulated 0.48%
of the CD8+ cells to produce IFN-, DCs plus
infected Fbs stimulated 1.38% of the CD8+ T
cells. The frequency was only 0.06% when uninfected Fbs were added to
DCs (Fig. 5A). Similar results were obtained with
three-color staining of lymphocytes with CD3-PE and CD8-CyChrome Abs
together before intracellular staining with IFN--FITC Ab (data not
shown). It allowed positive gating on CD3+ T
cells by FACS and ensured that IFN--producing NK cells were not
included in the calculations. The frequency of
CD8+ T cells found by the cross-presentation
experiments described here is in agreement with those found by others
using peptide stimulation or tetramer staining of HCMV-specific T
cells. Thus, cross-presentation is a very efficient Ag-presentation
pathway to stimulate virus-specific CD8+ T cell
recall responses in vitro. CD8- cells were also
stimulated by DC plus infected Fbs (1.49 vs 0.26% control; Fig. 5B). This was probably due to the presence of soluble
antigenic material in the cultures, which may have been presented via
the exogenous Ag presentation pathway to CD4+ Th1
cells.
|
Although we have shown that cell-free HCMV strain AD169 does not
infect DCs (Fig. 1A), the potential of cell-to-cell spread
of the virus during T cell stimulation in the mixed cultures also had
to be addressed. For this, we used an HCMV recombinant based on the
AD169 strain that encodes EGFP under the control of an early promoter.
EGFP provides a strong fluorescent signal that is easy to detect and
avoids the potential of nonspecific-Ab binding. Immature DCs were
cultured on their own (Fig. 6, column A) or were mixed with equal numbers of MRC5 Fbs
infected with RCMV288 (moi = 5 PFU/cell) 48 h earlier (Fig. 6, column C). Infected Fbs were also grown on their own as
controls (Fig. 6, column B). The cells from individual wells
were harvested at different times after the initiation of cultures
(from 1 h to 5 days, as indicated) and labeled with a
CyChrome-conjugated anti-HLA class II Ab. FACS analysis was used to
detect double-positive cells that express both class II molecules and
the GFP at high levels. The small region on the dot plot figures
indicates double-positive cells. The results were similar to those we
obtained with cell-free AD169 virus (Fig. 1A). Only a small
proportion of cells expressed both markers (0.2–0.4%). Because
double-positive cells were detected at 0.25% following only 1 h
incubation on ice in the presence of the class II Ab (Fig. 6, first row), we suggest that some of the double-positive
events observed in the mixed cultures were due to cell adhesion between
two cells with different markers and not to infection of DCs with
RCMV288. A slight shift of the HLA class II-positive cells on the FL1
axis is also detectable with time in the mixed cultures (Fig. 6, column C; days 4 and 5), and it might be due to phagocytosed
material from infected Fbs. Taken together, this coculture experiment
of DCs with RCMV288-infected Fbs indicates that if cell-to-cell spread
of the virus occurs in the mixed cultures, its level remains below
0.4%.
|
We used 0.4% (double-positive cells; Fig. 6) as the highest
possible rate of infection of DCs with AD169 HCMV in coculture
experiments. To determine the functional significance of this low level
of DC infection during T cell stimulation, we used the EC-adapted
strain of HCMV (TB40/E) to infect DCs. The level of infection with
strain TB40/E was measured by FACS following intracellular staining
with an Ab to HCMV early Ag as described earlier for strain AD169 (Fig. 1). Strain TB40/E-infected DCs were added at a gradually decreasing
proportion (Fig. 7), ranging from 19.7
(20) to 0.2%, to uninfected DCs, ensuring a constant number of DCs in
each culture. Strain AD169-infected Fbs alone or mixed with DCs were
also included in this experiment for comparison. Uninfected DCs alone
served as negative control. HLA-B7-restricted HCMV
pp65417–426-specific CTL activity was measured
from the cultures 6 days later. Strain TB40/E-infected DCs were
relatively weak stimulators of CTL even when 20% of the DCs were
infected. CTL stimulation at 0.4% infection, the possible DC infection
rate in the previous coculture experiment (Fig. 6), resulted only in a
very low level of CTL activity (Fig. 7; 2.4%). This extends our
previous observation on the lack of ability of HCMV strain AD169 to
infect DCs in a way or at a level that would enable these DCs to
stimulate virus-specific CTLs. In additional experiments, similar
results were obtained using IFN- production by
CD8+ T cells as a measure of the induced T cell
response (data not shown).
|
DCs have been shown to acquire and cross-present viral Ag to
CD8+ CTLs from influenza-infected apoptotic
cells, but not from necrotic cells, or from infected cells treated with
an apoptosis inhibitor (23). To determine the role of
apoptosis of HCMV-infected Fbs in CD8+ T cell
stimulation by DCs, infected Fbs were labeled with FITC-linked annexin
V (binds to the early apoptotic marker phosphatidylserine) and PI.
Early apoptotic cells are PI-/annexin
V+, late ones are
PI+/annexin V+, and
necrotic cells are PI+/annexin
V-. HCMV did not induce apoptosis of Fbs at
48 h after infection (Fig. 8A,
d) or even at day 7 after infection when cytopathic effects of the
virus were obvious (data not shown). Apoptosis was similarly low
following irradiation of Fbs with 3000 rads (data not shown). Apoptosis
was induced in uninfected Fbs either by deprivation of serum and
glutamine overnight (68% annexin V+) (Fig.
8A, b) or by treating the cells with 50 µg/ml Con A
(40) (41%; Fig. 8A, c).
HCMV-infected Fbs were less sensitive to deprivation than uninfected
ones (19%; Fig. 8A, e) but were as equally sensitive to Con
A treatment as uninfected Fbs (Fig. 8A, f; 53 vs 41%).
However, DCs did not stimulate CD8+ CTLs more
efficiently in the presence of HCMV-infected apoptotic Fbs than in the
presence of nonapoptotic ones (Fig. 8B). This finding
suggests that DCs are able to acquire viral Ags for induction of
specific CTLs from HCMV-infected nonapoptotic Fbs.
|
To gain some insight into the mechanism of the Ag presentation
described in this paper, DCs were pretreated with cytochalasin B, which
partially inhibits soluble Ag presentation by disrupting actin
microfilaments, and with brefeldin A, which blocks secretory protein
traffic and class I Ag presentation (41). Pretreatment of
DCs with cytochalasin B resulted in a 50% inhibition (Fig. 9A), and that with brefeldin A
resulted in a 70% inhibition (Fig. 9B) of
CD8+ T cell stimulation as measured by detecting
CTL activity (Fig. 9A) or the proportion of
IFN--producing CD8+ cells (Fig. 9B). The effect of both inhibitors is reversible; thus,
their removal and incubation of DCs in the absence of the inhibitors
with HCMV-infected Fbs may be responsible for the partial recovery of
DC function. However, the level of inhibition indicates the importance
of protein transport from the endoplasmic reticulum in DCs during
cross-presentation of CMV Ags. An additional experiment showed that
cytochalasin B did not inhibit infection of DC with the EC-adapted HCMV
strain TB40/E, either with free virus or in coculture experiments with
strain TB40/E-infected Fbs (data not shown), confirming that the
cytochalasin B-mediated inhibition was not directed against infectious
virus uptake and trafficking to the nucleus within DCs.
|
Optimal cross-presentation of tumor Ags by DCs to
CD8+ T cells requires two steps: 1) uptake of
tumor Ag by immature DCs, and 2) a maturation signal for DCs
(42). To demonstrate whether HCMV-infected Fbs can provide
a maturation signal for DCs, the expression of CD83 on DCs following 2
days coculture with infected Fbs was examined (Fig. 10). DCs were cocultured with
uninfected Fbs (Fig. 10a), with HCMV-infected Fbs (Fig. 10b), with autologous PBMC (Fig. 10c), or with
infected Fbs and T cells together (Fig. 10d). HCMV-infected
Fbs induced maturation of DCs (54.3% expressed CD83 vs 17.2% when
uninfected Fbs were added) irrespective of the presence of T cells
(45.7%). Up-regulation of HLA class II molecules followed a similar
pattern (data not shown). Thus HCMV-infected Fbs provide both the Ag
for cross-presentation and the signal for DC maturation, which together
result in a successful stimulation of CD8+ T
cells.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The conventional view of Ag presentation to T cells is that exogenous
Ags are presented via MHC class II and endogenous Ags via MHC class I
molecules is simplistic and leaves questions unanswered about priming T
cell responses to viruses that do not infect professional APCs. The
cross-presentation pathway is obligatory in vivo for the initiation of
CTL responses to viruses that infect only nonhemopoietic cells
(45). Its role should be emphasized and studied more
thoroughly also in the stimulation of CTL responses to those viruses
that infect bone marrow-derived APCs. Cross-presentation has been
described when high concentrations of exogenous soluble or particulate
Ags (46), stress protein-associated Ags (47),
or Ags derived from apoptotic virus-infected (23, 48) or
apoptotic tumor cells (42) gained access to the class I
Ag-presenting pathway. We describe here that cross-presentation of HCMV
Ags by DCs occurs in the absence of early or late apoptotic markers on
infected Fbs, and in vitro induction of apoptosis of HCMV-infected Fbs
does not facilitate the process of Ag presentation to
CD8+ T cells by DCs. The mechanism resulting in
CD8+ T cell activation following the interaction
between DCs and HCMV-infected Fbs is being studied in our laboratory.
Because HCMV encodes molecules that block or delay apoptosis to enable
virus replication (27), it is likely that cross-priming of
specific CTLs by DCs is mediated by nonapoptotic signals expressed by
cells infected with HCMV. The possible candidates are 1) cytopathic
effects caused by the virus, especially nonapoptotic cell lysis, which
would allow release of antigenic material from the infected cells; 2)
cell surface molecules that are up-regulated by HCMV on Fbs. CD54
(ICAM-1) and CD58 (LFA-3) (49) may play a role in
cell-cell fusion and exchange of intracellular viral particles. This
has been shown between polymorphonuclear cells and ECs infected with
clinical isolates but not with laboratory strains (50);
and 3) decreased level of surface MHC class I molecules and increased
production of certain cytokines and chemokines, or a combination of
these factors. A recent study on cross-presentation of HCMV pp65 Ag by
DC following coculture with apoptotic (TNF-
-treated) strain
AD169-infected Fbs concluded that apoptosis was the main factor in
cross-presentation of the pp65 Ag (51). However, the
possibility of an alternative mechanism cannot be excluded, given the
high level (52%) of secondary necrosis of Fbs in the coculture
experiments.
Immature DCs are extremely well equipped to capture Ag in tissues. Ag uptake and the inflammatory environment provide signals that mobilize DCs to migrate to secondary lymphoid organs where they stimulate Ag-specific naive T cells. We observed that coculture of HCMV-infected Fbs with immature DCs causes maturation of DCs as measured by the up-regulation of CD83 and HLA class II molecules. Viral infection of DCs can result in the inhibition of maturation, Ag presentation, or migratory capacity of DCs (52, 53, 54). Whether HCMV infection would interfere with the maturation and function of DCs remains to be seen by using virus strains that infect DCs efficiently. Decrease in the viable cell number of DCs 6–8 days following infection with an EC-adapted strain of HCMV has been observed by others (32) and by us, as well as down-regulation of HLA class I molecules on infected DCs (M. Moutaftsi, L. K. Borysiewicz, and Z. Tobi, manuscript in preparation). These preliminary observations also point toward the potential importance of cross-presentation not only in cases in which professional APCs cannot be infected with the virus, but also when the viral infection results in an impairment of the Ag-presenting function. We suggest that infection of DCs with wild-type HCMV in vivo and uptake of viral material in situ from infected tissue can occur simultaneously and may depend on the given environment. Acquisition of noninfectious viral material by DCs and cross-presentation of viral Ags to lymph node T cells would increase the host’s chances to develop a strong CTL response 1) without dissemination of infectious virus by DCs, and 2) without being affected by the inhibitory effects of HCMV.
The pp65 protein inhibits the presentation of IE1-derived Ags in target cells (7), but its effect on the stimulation of IE1-specific T cells has not been studied. Indeed, several groups have reported significant levels of T cells specific for IE1 Ags in persistently infected asymptomatic individuals (15, 37) without explaining how it happens in the face of inhibition by pp65. We have shown here that DCs successfully induce IE1-specific CTLs via the cross-presentation pathway, although we did not address the question of whether the weak T cell stimulation by HCMV-infected Fbs alone was due to inhibition by pp65. This still remains an important question.
It is likely that the nonadherent fraction of PBMC used in our experiments contained some DC precursors. These remaining DCs may, in fact, have been inducing the IE1- and pp65-specific CTL responses when HCMV-infected Fbs alone were added to PBMC, given the poor T cell stimulatory capacity of infected Fbs expressing very low levels of HLA class I molecules. The mechanism involved could also have been cross-priming, although at a much reduced level due to the lower number of DCs present. An alternative explanation for the higher stimulatory capacity of DCs could be that they stimulated a different or an additional subpopulation of memory T cells than HCMV-infected Fbs alone. The recently described subpopulations of memory T cells (central and effector memory T cells) (55) with different activation requirements and recirculation pathways would be likely candidates.
In conclusion, the data reported here show that DCs present HCMV Ags via HLA class I molecules after acquiring viral Ag from HCMV-infected Fbs. Our results also show that HCMV-infected Fbs provide viral Ags to DCs for cross-presentation to stimulate class I-restricted CTLs in the absence of apoptosis of infected cells. DCs also express maturation markers following the encounter with infected Fbs. These findings have potential implications for generating HCMV-specific T cells for adoptive immunotherapy.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Zsuzsanna Tabi, Department of Medicine, University of Wales, College of Medicine, Tenovus Building, Heath Park, Cardiff CF14 4XX, U.K.
3 Abbreviations used in this paper: HCMV, human CMV; BLCL, B lymphoblastoid cell line; DC, dendritic cell; EC, endothelial cell; EGFP, enhanced green fluorescent protein; Fb, fibroblast; IE, immediate early; moi, multiplicity of infection; PI, propidium iodide; rAd, recombinant adenovirus; gB, glycoprotein B.
Received for publication May 1, 2000. Accepted for publication February 22, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B activity by transactivating the NF-B p105/p50 and p65 promoters. J. Virol. 69:5391.[Abstract]
. J. Exp. Med. 179:1109.v5 and CD36, and cross-present antigens to cytotoxic T lymphocytes. J. Exp. Med. 188:1359.This article has been cited by other articles:
|
|
 |
|
  M. M. Weck, S. Appel, D. Werth, C. Sinzger, A. Bringmann, F. Grunebach, and P. Brossart hDectin-1 is involved in uptake and cross-presentation of cellular antigens Blood, April 15, 2008; 111(8): 4264 - 4272. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. M. Weck, F. Grunebach, D. Werth, C. Sinzger, A. Bringmann, and P. Brossart TLR ligands differentially affect uptake and presentation of cellular antigens Blood, May 1, 2007; 109(9): 3890 - 3894. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. N. Posnett, M. E. Engelhorn, and A. N. Houghton Antiviral T cell responses: phalanx or multipronged attack? J. Exp. Med., June 20, 2005; 201(12): 1881 - 1884. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Bosnjak, M. Miranda-Saksena, D. M. Koelle, R. A. Boadle, C. A. Jones, and A. L. Cunningham Herpes Simplex Virus Infection of Human Dendritic Cells Induces Apoptosis and Allows Cross-Presentation via Uninfected Dendritic Cells J. Immunol., February 15, 2005; 174(4): 2220 - 2227. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. J. Manley, L. Luy, T. Jones, M. Boeckh, H. Mutimer, and S. R. Riddell Immune evasion proteins of human cytomegalovirus do not prevent a diverse CD8+ cytotoxic T-cell response in natural infection Blood, August 15, 2004; 104(4): 1075 - 1082. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. N. Fleeton, N. Contractor, F. Leon, J. D. Wetzel, T. S. Dermody, and B. L. Kelsall Peyer's Patch Dendritic Cells Process Viral Antigen from Apoptotic Epithelial Cells in the Intestine of Reovirus-infected Mice J. Exp. Med., July 19, 2004; 200(2): 235 - 245. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Li Pira, L. Bottone, F. Ivaldi, R. Pelizzoli, F. Del Galdo, L. Lozzi, L. Bracci, A. Loregian, G. Palu, R. De Palma, et al. Identification of new Th peptides from the cytomegalovirus protein pp65 to design a peptide library for generation of CD4 T cell lines for cellular immunoreconstitution Int. Immunol., May 1, 2004; 16(5): 635 - 642. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Moutaftsi, P. Brennan, S. A. Spector, and Z. Tabi Impaired Lymphoid Chemokine-Mediated Migration due to a Block on the Chemokine Receptor Switch in Human Cytomegalovirus-Infected Dendritic Cells J. Virol., March 15, 2004; 78(6): 3046 - 3054. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Khan, A. Zimmermann, M. Basler, M. Groettrup, and H. Hengel A Cytomegalovirus Inhibitor of Gamma Interferon Signaling Controls Immunoproteasome Induction J. Virol., February 15, 2004; 78(4): 1831 - 1842. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Fonteneau, D. G. Kavanagh, M. Lirvall, C. Sanders, T. L. Cover, N. Bhardwaj, and M. Larsson Characterization of the MHC class I cross-presentation pathway for cell-associated antigens by human dendritic cells Blood, December 15, 2003; 102(13): 4448 - 4455. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Hertel, V. G. Lacaille, H. Strobl, E. D. Mellins, and E. S. Mocarski Susceptibility of Immature and Mature Langerhans Cell-Type Dendritic Cells to Infection and Immunomodulation by Human Cytomegalovirus J. Virol., July 1, 2003; 77(13): 7563 - 7574. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. Karrer, S. Sierro, M. Wagner, A. Oxenius, H. Hengel, U. H. Koszinowski, R. E. Phillips, and P. Klenerman Memory Inflation: Continuous Accumulation of Antiviral CD8+ T Cells Over Time J. Immunol., February 15, 2003; 170(4): 2022 - 2029. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Koelle and L. Corey Recent Progress in Herpes Simplex Virus Immunobiology and Vaccine Research Clin. Microbiol. Rev., January 1, 2003; 16(1): 96 - 113. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Kondo, M. S. Topp, H.-P. Kiem, Y. Obata, Y. Morishima, K. Kuzushima, M. Tanimoto, M. Harada, T. Takahashi, and Y. Akatsuka Efficient Generation of Antigen-Specific Cytotoxic T Cells Using Retrovirally Transduced CD40-Activated B Cells J. Immunol., August 15, 2002; 169(4): 2164 - 2171. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Moutaftsi, A. M. Mehl, L. K. Borysiewicz, and Z. Tabi Human cytomegalovirus inhibits maturation and impairs function of monocyte-derived dendritic cells Blood, April 15, 2002; 99(8): 2913 - 2921. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X.-Q. Zhao, X.-L. Huang, P. Gupta, L. Borowski, Z. Fan, S. C. Watkins, E. K. Thomas, and C. R. Rinaldo Jr. Induction of Anti-Human Immunodeficiency Virus Type 1 (HIV-1) CD8+ and CD4+ T-Cell Reactivity by Dendritic Cells Loaded with HIV-1 X4-Infected Apoptotic Cells J. Virol., February 22, 2002; 76(6): 3007 - 3014. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Arrode, C. Boccaccio, J.-P. Abastado, and C. Davrinche Cross-Presentation of Human Cytomegalovirus pp65 (UL83) to CD8+ T Cells Is Regulated by Virus-Induced, Soluble-Mediator-Dependent Maturation of Dendritic Cells J. Virol., January 1, 2002; 76(1): 142 - 150. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Kammerer, D. Stober, P. Riedl, C. Oehninger, R. Schirmbeck, and J. Reimann Noncovalent Association with Stress Protein Facilitates Cross-Priming of CD8+ T Cells to Tumor Cell Antigens by Dendritic Cells J. Immunol., January 1, 2002; 168(1): 108 - 117. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) expression in DCs infected
with HCMV strain AD169 (A) or with TB40/E
(B) (moi = 5 PFU/cell for each) for 6, 16, or
44 h. Cells were fixed and permeabilized before staining with HCMV
Ag-specific Abs or mouse IgG1 negative control (
)
followed by goat anti mouse IgG-FITC for FACS analysis as described in
Materials and Methods. Only DCs infected with strain
TB40/E show significant levels of viral Ags, with pp65 being detectable
only at 44 h after infection. The results are representative of
two experiments.
) or mismatched (Fb 3; Table I
) induced a lower
level of CTL activity, whereas mismatched infected Fbs alone (
) did not induce HCMV-specific CTL
responses.
) or
together with autologous DCs (
) or with DCs plus uninfected MRC5 (
) or with strain AD169-infected Fbs alone (
) are shown
following stimulation as indicated under the columns. DCs
with nonapoptotic infected Fbs were the most efficient stimulators of
CTL. The results are expressed as lytic units (see explanation in Fig. 3
