close

Вход

Забыли?

вход по аккаунту

?

884

код для вставкиСкачать
Int. J. Cancer: 68,251-258 (1996)
Q 1996 Wiley-Liss, Inc.
Publication of the International U n m Against Cancer
Publication de I’Union Internationale Contre 1e Cancer
DEFECTIVE PRESENTATION OF MHC CLASS I-RESTRICTED CYTOTOXIC
T-CELL EPITOPES IN BURKITT’S LYMPHOMA CELLS
Teresa FRlSAN1*3,Qian-Jin Z W G ” , Jelena LEVITSKAYA’,
Michael CORAM~,Michael G. KURILLA~
and Maria G. MASUCCI~.~
lMicrobiologyand Tumor Biology Center, Karolinska Institute, S-I71 77 Stockholm, Sweden; =Departmentof Pathology,
University of Erginia Health Science Center, Charlottesville, Erginia, USA.
Defects of antigen processinglpresentation have been suggested to play a role in the escape of Burktt’s lymphoma (BL)
from cytotoxic T lymphocyte (CTL-ediated
rejection. Impaired presentation of an immunodominant H I A A l l - r e stricted epitope from the resident or recombinant vaccinia
virus-expressed Epstein-Barr virus nuclear antigen (EBNA)4
was demonstrated in the EBV-positive E95B-BU8 and its
EBV-negative parental B U 8 cell lines. We have investigated
whether this was due to (i) impaired production of the antigenic
peptide, (ii)poor peptide translocation into the ER lumen or (iii)
inefficient maturation and transport of the MHC-peptide complexes at the cell surface. The defect was not overcome by
cytosolic expression of a pre-formed epitope, suggesting that
presentation of EBNA4 is not limited by inefficient production of
the antigenic peptide. B U 8 expressed 5- to 10-fold lower levels
ofthe transporter associated with antigen presentation(TAP) I
and TAP2 proteins and behaved poorly in a stre o
mediated peptide translocation assay, whereas
showed higher TAP expressionand virtually normal transporter
function. Up-regulation of HIA A l I and reconstitution of TAP
activity by treatment with 1FN-y did not restore presentationof
the resident EBNA4 in E95B-BU8 and did not enhance presentation of the vaccinia virus-expressed intact protein or preformed epitope. Efficient maturation of class I molecules to
Endo H-resistant species was demonstrated in pulse-chase
experiments. Taken together, our findings identify a previously
uncharacterized defect of antigen presentationwhich appears
to affect events occurring after proteasomal degradation but
before TAP-dependent peptide transport and MHC class I
assemb and maturation.
g5r:gi
o 1996
8iley-Liss,
Inc.
The capacity of cytotoxic T lymphocytes (CTLs) to recognize
and destroy tumor cells is well established, and there are now
numerous examples of therapeutic efficacy following adoptive
transfer of CTLs to cancer patients. The majority of tumorspecific CTLs express the CD8 surface glycoprotein and
clonally distributed T-cell receptors (TCRs) that recognize
8-10 amino acid long peptides derived from proteolytic degradation of cytosolic and nuclear proteins. These are presented
at the cell surface in association with MHC class I heavy chain:
p2-microglobulin(p2m) heterodimers (Townsend et al., 1986).
Recent evidence suggests that processing of antigens for class
I-restricted presentation requires both ubiquitination and
subsequent degradation by the proteasome. Two proteasome
subunits, the low-molecular-mass protease (lmp)2 and lmp7,
are encoded in the MHC class I1 region. Lack of these subunits
in mutant cell lines correlates with altered peptidase activities
of the proteasome (Gaczynska et al., 1993). In addition, the
products of 2 genes of the ABC transporter family, the
transporter associated with antigen presentation (TAP)l and
TAP2, are required for efficient presentation of most class
I-restricted CTL epitopes. Functional TAPl /TAP2 heterodimers translocate cytosolic peptides to the site of class 1
assembly in the lumen of the endoplasmic reticulum (ER) in
an ATP-dependent fashion. They also bind to pre-formed class
I heavy chain:Pzm heterodimers, which associate loosely in the
absence of peptides, and facilitate assembly, probably by
increasing the effective peptide concentration or by acting as
molecular chaperones in promoting class I folding (reviewed in
Townsend and Trowsdale, 1993). It has been suggested that
the heat shock proteins (HSPs) gp96 and hsp70 are accessory
to antigen presentation by MHC class I molecules (Srivastava
el al., 1994). In particular, the cytosolic hsp70 is believed to
associate with antigenic peptides derived from proteolytic
degradation of cellular proteins, while gp96 may act as a
general peptide-acceptor in the ER and facilitate the transfer
of peptides to class I molecules.
Despite the potential of CTLs to eradicate neoplasms, it is
obvious from the progressive growth of most tumors that
malignant cells escape CTL surveillance in vivo. The EpsteinBarr virus (EBV)-associated Burkitt’s lymphoma (BL) provides one of the best examples of tumour that evades CTL
recognition (reviewed in Rickinson et al., 1992). Several
consistent features of the BL cell phenotype are likely to
contribute to CTL resistance. Thus, EBV-carrying BL lines
that have retained the phenotypic characteristics of the original tumor do not express the highly immunogenic viral proteins
that are regularly detected in EBV-transformed immunoblasts
or several accessory molecules that are required to strengthen
the interaction of CTLs with their targets. In addition, most BL
lines express relatively low levels of HLA class I and show
selective loss of certain class I alleles. Recent evidence
suggests that phenotypically representative BL lines display
impaired antigen-presenting function which correlates with
transcriptional down-regulation of TAP genes. Partial reconstitution of antigen presentation was achieved by transfection of
a minigene encoding a viral epitope linked to an ER localization sequence (Khanna et al., 1994) or by expression of the
EBV-encoded latent membrane protein (LMP)l in parallel
with up-regulation of TAPl (Rowe et al., 1995). However, the
relative inefficiency of these procedures and the failure to
reconstitute antigen presentation in some BL lines suggest that
additional defects may be involved.
In the present study we have analyzed various steps of
antigen processing/presentation in one EBV-negative BL line
derived from an HLA All-positive individual and its in vitro
EBV-converted subline that expresses the full spectrum of
latent viral genes. We demonstrate that low levels of presentation of an immunodominant HLA A1 1-restricted epitope
from the EBV nuclear antigen (EBNA)4 could be achieved
only by over-expression of the intact protein in cells infected
with a recombinant vaccinia virus. The defect was not overcome by expression of the pre-formed epitope or by treatment
with IFNy, which resulted in up-regulation of HLA A l l and
full reconstitution of TAP expression and function. Efficient
maturation of class I heavy chain to Endo H-resistant species
was demonstrated in pulse-chase experiments indicating that
peptide loading and transport of class I-peptide complexes
occur normally.
3The authors contributed equally to this work and are therefore
placed in alphabetical order.
jTo whom correspondence and re rint requests should be sent, at
MTC,Karolinska Institute, Box 28g S-171 77 Stockholm, Sweden.
Fax: 46 8 330498.
Received: April 25, 1996 and in revised form July 5,1996.
252
FRISAN ETAL
MATERIAL AND METHODS
Cell lines
The BL28 cell line and the EBV-transformed lymphoblastoid cell line (LCL) IARC 139 were established from the
tumor and by in vitro EBV infection of normal B lymphocytes
from a patient bearing an EBV-negative BL (HLA A l , A l l ,
B8, B16). The E95B-BL28 cell line was obtained by infection
of BL28 with the spent supernatant from the virus producer
B95.8 cell line. The DHsp LCL was established by spontaneous outgrowth from lymphocytes of a normal EBV-seropositive donor (HLA A l l B31.2, B51) cultured in the presence of
0.1 pg/ml cyclosporin A. This LCL carries a mutated EBV
strain and is not lysed by CTLs specific for the prototype
EBNA4 416-424 epitope (IVTDFSVIK, E441M24) (de CamposLima et al., 1994). The TAP-deficient mutant T2 and the
parental T1 cell lines (Salter and Cresswell, 1986) were used as
controls in the analysis of TAP protein expression and peptide
transport assays. All lines were cultured in RPMI 1640
(GIBCO, Grand Island, NY), supplemented with 2 mM
glutamine, 100 U/ml penicillin, 100 U/ml streptomycin and
10% heat-inactivated FCS. Where indicated, the cell lines
were cultured for 48 hr in medium containing 500 IU/ml
interferon (IFN)-y before the assays.
Recombinant vaccinia viruses
The Vacc-EBNA4 recombinant that expresses the EBNA4
gene of the prototype B95.8 EBV strain has been described
previously (Murray et al., 1990). A recombinant virus expressing the E4416-424
epitope preceded by a methionine codon was
constructed by cloning the minigene into the Sma I site of the
pSCl1 transfer vector. Briefly, the oligonucleotides 5 ' GAATCATGATAGTAACTGACTTTAGTGTAATCAAGTAGAC-3' and 5'-GTCTACTTGATTACACTAAAGTCAGTTACTATCATGATTC-3' were annealed by
incubating 10 pM of each oligo in a 20 p,l reaction with 10 mM
TRIS-HCl, 50 mM NaCI, 10 mM MgC12 and 1 mM DTT at pH
7.9. The reaction was heated to 97°C for 5 min, allowed to cool
slowly until <5O"C and incubated at 37°C for an additional 5
min. The annealed oligos were diluted 5-fold, and 2 pmoles
were ligated to the pSCll vector. Colonies with inserts were
isolated and subjected to DNA sequencing by the Exo Pfu
cycling sequencing strategy (Stratagene, La Jolla, CA) according to the instructions of the manufacturer. A plasmid of
correct orientation relative to the P7.5 promoter was selected
and vaccinia virus recombinants were produced according to
standard protocols. After plaque purification and amplification under selection, viral stocks were amplified in BSC-40
cells, harvested and titrated as previously described (Murray et
al., 1990). Two P-galactosidase-positiveplaques were isolated
due to inability to assess expression of the minigene by
Western blotting or immunofluorescence techniques.
Synthetic peptides
Synthetic peptides were synthesized by the Merrifield solidphase method. Peptides were dissolved in DMSO at the
concentration of lo-* M and further diluted in PBS to obtain
the indicated concentrations before the assays. A synthetic
peptide library with fixed amino acids to allow iodination (Y),
glycosylation (N and T) and variable COOH-terminus was
constructed from the reference TYNRTRALI peptide by
incorporation of different amino acids in a single coupling
cycle, as described (Heemels et al., 1993).
Detection of EBVproteins and TAP components
by immunoblotting
Expression of viral antigens was assayed in uninfected and
Vacc-EBNA4-infected cells using a previously characterized
human serum (HR) containing high antibody titers to all
EBNAs(VCA 1:640, EBNAl 1:80, EBNA-2A 1~160,EBNA-2B
1:80, EBNA-6 1:320) and the LMP1-specific mouse monoclo-
nal antibody (MAb) S-12 (Mann et al., 1985). Total extracts
from lo6 cells were separated in 8% polyacrylamide-SDS gels
and blotted onto nitrocellulose filters. Blots were probed with
1:20 dilution of the HR serum or 1:10,000 dilution of the S-12
MAb. TAPl and TAP2 were detected using rabbit anti-sera
specific for C-terminal peptides at 1:1,000 dilution (Kellyet al.,
1992). Blots were then reacted with horseradish peroxidaselabeled anti-human, anti-mouse or anti-rabbit antibodies,
respectively, and immunocomplexes were visualized by enhanced chemiluminescence (ECL) according to the instructions of the manufacturer (Amersham, Aylesbury, UK). For
quantitative analysis, the relative expression of TAPl and
TAP2 molecules was assessed by densitometric scan of the
ECL-developed immunoblots.
Surface class I antigen expression
HLA class I expression was assessed by indirect immunofluorescence using the MAbs W6/32 (anti-HLA A, B, C monomorphic determinant) and AUF5.13 (anti-HLA A3, A1 1). Fluorescein isothiocyanate-conjugated (FITC) rabbit anti-mouse IgGs
(Dakopatts, Glostrup, Denmark) were used as second antibody. Aliquots of 106 cells were incubated for 30 min at 4°C
with 50 pl of the first antibody. After 2 washings, the cells were
resuspended in 50 p1 of the FITC-conjugated secondary
antibody and incubated for 30 min at 4°C. All samples were
washed twice and fluorescence intensity was measured with a
FACSort analyzer (Becton Dickinson, Mountain View, CA).
MHC class I maturation
Pulse-chase experiments were performed as previously described (Ljunggren et al., 1990). Briefly, 6 x lo7 BL28 and 3 x
lo7IARC 139 cells were incubated in methionine-free medium
for 30 rnin at 37°C and then pulse-labeled by adding 300 pCi
[3sS]methionine (Amersham) for 10 min. Warmed medium
containing excess of cold methionine was added and cells were
then incubated at 37°C for the indicated times before harvesting. AIiquots of 5 x lo6 LCL cells and 1 x lo7 BL cells were
lysed for 20 rnin at 0°C in a buffer containing 140 mM NaCI, 1
mM EDTA, 1% Triton X-100, 1 mM PMSF and 10 mM
TRIS-HCl (pH 7.4). Nuclei were pelletted by centrifugation at
15,00Og, and supernatants were incubated with protein ASepharose previously loaded with the MAb W6/32. After
extensive washing, class I molecules were eluted from the
beads by boiling in 0.1 M sodium acetate, 3 mM EDTA and
0.25% SDS (pH 6). One-half of each sample was digested
overnight at 37°C with 2 mU endo-p-N-acetylglucosaminidase
H (Endo H; Boehringer, Mannheim, Germany) and the other
half was mock-digested. Samples were then analyzed by
SDS-PAGE using a 12% polyacrylamide gel. Gels were fixed,
incubated with Amplify (Amersham), dried and exposed to
Kodak XAR-5 film.
Peptide translocation assays
Streptolysin-0-mediated transport assays were performed
as described by Momburg et al. (1994) with minor modifications. Briefly, the peptide library was 1251-labeled by chloramine T-catalyzed iodination to a specific activity of 10 Ci/
mmol (Heemels et al., 1993). Cells ( lo6) were permeabilized
with 2 IU/ml streptolysin-0 (Wellcome, Dartford, UK) for 10
min at 37°C. The iodinated peptide library (approx. 66 ng) was
added and the incubation continued for 10 min in the presence
or absence of 10 mM ATP (Sigma, St. Louis, MO). Cells were
then transferred to ice, lysed in a buffer containing 1% NP40,
150 m M NaCI, 5 mM MgClz and 50 mM TRIS-HCI (pH 7.5)
and nuclei were removed by centrifugation at 15,000 g for 10
min. Translocated peptides that had been glycosylated in the
ER were recovered by adsorption to concanavalin A-Sepharose beads (Pharmacia, Uppsala, Sweden). Beads were washed
5 times in lysis buffer, and the associated radioactivity was
measured in a gamma-counter (LKB, Uppsala, Sweden). Time
kinetics experiments were performed by incubating permeabi-
DEFECTS OF ANTIGEN PROCESSING IN BL CELLS
253
lized cells in the presence of labeled peptides with or without
10 mM ATP for the indicated times before harvesting.
Generation of CTL clones and cytotoxicity tests
HLA A1 1-restricted EBV-specific CTLs were obtained by
stimulation of lymphocytes from EBV-seropositive donors
with the autologous B95.8 virus-transformed LCL as previously described (Torsteinsdottir et al., 1986). After 2 or 3
consecutive restimulations, the culture was expanded in complete medium supplemented with 10 U/ml of human recombinant interleukin (1L)-2 and 30% (vol/vol) culture supernatant
from the gibbon lymphoma cell line MLA 144. Single-cell
cloning was done by limiting dilution in 96-well plates in 200 kl
of IL-Zcontaining medium and 105 irradiated (3,000 rads)
allogeneic PHA-pulsed blood lymphocytes as feeder (Torsteinsdottir et al., 1986). Growing cultures were transferred into
48-well plates and fed twice a week by replacing half of the
medium. The EBV specificity and class I restriction of the
clones was investigated by testing their cytotoxic activity
against a panel of EBV-positive and -negative targets, including the autologous LCL and PHA-activated blasts, allogeneic
LCLs sharing single class I alleles and class I-mismatched
LCLs. Targets were labeled with Nas1Cr04(0.1 @/lo6 cells)
for 1 hr at 37”C, and cytotoxic activity was assayed in a standard
5 hr SICr-releaseassay (Torsteinsdottir er a/., 1986). Cytotoxicity tests were routinely run at 5:1,2.5:1 and 1.2:l effector:target
ratios in triplicate. For peptide sensitization, one aliquot of the
targets was pulsed with lo-* M of the synthetic peptide during
labeling and extensively washed before the assay.
RESULTS
BL cells fail topresent endogenous viral antigens
to EBV-specijic CTLs
The EBV-negative BL28 cell line and its in vitro B95.8
virus-converted subline, E95B-BL28, were tested for their
ability to present the E441M2.1 epitope derived from EBNA4.
The E95B-BL28 cell line was not sensitive to lysis by the
E441M424-~pe~ifi~
CTLclone BK289, though it expressed EBNA4
at levels comparable to a regular B95.8 virus-transformed LCL
(Fig. 1). Both cell lines were lysed after pre-incubation with
lo-* M of the relevant synthetic peptide, confirming that
sufficient amounts of the restricting class I allele were expressed at the cell surface (Fig. 2u). The BL cells were also
lysed after infection with a recombinant vaccinia virus encoding the EBNA4 protein from the prototype B95.8 EBV strain
(VaccEBNA4), but the level of killing was, in repeated
experiments, lower than that achieved in the control DHsp
LCL. This was investigated by comparing CTL sensitization
over a wide range of multiplicity of infection (m.0.i.). To avoid
biases of interpretation due to the variable expression of class I
antigens and adhesion molecules, the results are expressed as
the ratio between the lysis observed after Vacc-EBNA4
infection and the maximal lysis achieved by peptide sensitization. Maximal lysis was induced by infection of DHsp at an
m.0.i. of 1, whereas only 20-60% of the peptide-induced lysis
was achieved in the BL cells even at a 10-fold higher m.0.i.
(Fig. 26). Similar levels of EBNA4 were detected in all
Vacc-EBNA4-infected cells (Fig. 1). Thus, processing and/or
presentation of the endogenously expressed protein appear to
be deficient in BL cells.
Defective antigen presentation is not overcome by cytosolic
expression of a pre-formed E4411wz4epitope
To assess whether the defect could be explained by impaired
production of the relevant antigenic peptide, a pre-formed
E441-24 peptide was expressed in the cytosol by infection with
a vaccinia recombinant encoding the epitope preceded by a
methionine codon (Vacc-MetIVT). Infection of DHsp for 4 hr
at a multiplicity of 1 was sufficient to induce epitope-specific
mw
205
A
A B C D E F
-
140
-
83
-
45
-
E4
E6
E3
E2
El
B
A B C E
mw
83
LMPl
45
-
Flcuw 1-Expression of EBV antigens in uninfected and
Vacc-EBNA4-infected cells. Lane A, B95.8 virus-transformed
LCL; B, Vacc-EBNA4-infectedDHsp; C, BL28; D, Vacc-EBNA4infected BL28; E, E95B-BL28; F, Vacc-EBNA4-infected E95BBL28. Total cell extracts from 106 cells were fractionated by
SDS-PAGE, transferred to nitrocellulose filters and probed with
the EBV antibody-positive polyclonal human serum HR (a) or
with the LMP1-s ecific mouse MAb S-12 (b). The position of the
m.w. markers anfthe EBNA4 and LMPl proteins are indicated.
lysis at levels comparable to those achieved by sensitization
with the synthetic peptide, and 50% of the maximal lysis was
obtained at an m.0.i. as low as 0.1 (Fig. 3). In contrast,
infection of BL28 and E95B-BL28 at a multiplicity of 1
induced levels of killing that did not exceed 20% and 50% of
the maximal lysis, respectively. Higher m.0.i. could not be
evaluated since, under those conditions of infection, the
epitope is efficiently presented also in HLA A1 1-transfected
T2 cells (not shown). The failure to restore antigen presentation by expression of a pre-formed epitope suggests the
involvement of functions down-stream of the cytosolic production of the antigenic peptide.
TAP expression and peptide transport are deficient in BL celk hut
can he restored by IFN- y treatment
A functional TAP1/TAP2 heterodimer is required for
translocation of peptides from the cytosol into the ER lumen
in an ATP-dependent fashion. TAP expression was therefore
investigated by probing Western blots of total cell extracts with
previously characterized TAP1- and TAPZspecific polyclonal
FRISAN ETAL.
254
50
BL28
-
E95B-BL28
+ DHSP
'"1
DHSp
-0-
.-aa 40 -
-
peptide
Vacc-EBNA4
loo
1
+
0
+
.
-
+
-
-
+
-
-
+
-
-
+
-
-
n
"\
B
I
1:l
1:l
031
0.1:l
Vacc-Met IVT m.0.i.
DHsp
EL28
E95B-EL28
3:l
BL28
+ E95B-BL28
FIGURE 3 - Presentation of Vacc-MetIVT. Aliquots of 1.5 X lo6
cells were either infected for 4 hr with the Vacc-MetIVT recombinant at the indicated m.0.i. or pre-incubated with the synthetic
E441u24pe tide and then tested for sensitivity to lysis by the CTL
clone BK2h. The percent specific lysis was assayed at a 2 5 1
effector:target ratio, and the relative lysis was calculated as
described in Figure 2. The mean k SE of 3 experiments is shown in
the figure.
0.3:l
Vacc-EBNA4 m.o.1.
FIGURE2 - Lysis of uninfected and Vacc-EBNA4-infected cells
by the HLA All-restricted E 4 4 1 M 2 4 - ~ p e CTL
~ i f i ~clone BK289.
(a) Aliquots of 1.5 x 106 cells were either infected for 4 hr with the
Vacc-EBNA4 recombinant virus at an m.0.i. of 10 or incubated for
1 hr at 37°C in the presence of
M of the synthetic E4416-424
peptide and extensively washed before labeling. The percent
specific lysis at a 2 5 1 effector:target ratio in 1 representative
experiment out of 6 is shown in the figure. The percent specific
lysis of cells infected with a control Vacc-TK- recombinant was
always lower than 1% (data not shown). (b) Aliquots of 1.5 x lo6
cells were either infected for 4 hr with Vacc-EBNA4 at the
indicated m.0.i. or pre-incubated with the synthetic E441~24
yptide as described above. The percent relative lysis was calcuated as the ratio between the specific lysis of Vacc-EBNA4infected cells and the maximal lysis obtained after pre-incubation
with the synthetic peptide. One representative experiment out of 3
is shown in the figure.
rabbit sera. Lower levels (5-to 10-fold) of both proteins were
detected in BL28 compared with DHsp (Fig. 4u, b, lanes B, C),
while intermediate levels of expression were demonstrated in
E95B-BL28 (Fig. 4a, b, lane E). Treatment of the BL cell lines
for 48 hr in medium containing 500 IU/ml IFN-y resulted in
up-regulation of T A P l and TAP2 at levels comparable to
those observed in the control LCL (Fig. 4u, b, lanes B, D, F).
To assess whether the lower expression of TAPS correlated
with impaired peptide transport, the BL lines were tested in a
streptolysin-0-mediated transport assay in parallel with a
B95.8 virus-transformed LCL derived from the normal B
lymphocytes of the same donor. A 12sI-labeled peptide library,
containing 2,304 peptides with variable COOH-terminus
(Heernelset al., 1993), was used to quantitate E R translocation
by virtue of N-linked glycosylation. The TAPl /TAPZdeficient
T2 and the T1 parental lines were included as negative and
positive controls, and basal levels of peptide uptake were
determined by incubating the streptolysin-0-pcrmcabilizcd
cells in the absence of exogenous ATP. The results of 5
experiments are summarized in Table I. BL28 showed significantly reduced levels of ATP-dependent E R uptake compared
with the corresponding LCL (ATP+/ATP- ratio 1.8 vs. 7.1),
while intermediate levels were observed in the E95B-BL28
A
A B C D E F
IFN-Y
04
- - - + - +
- TAP1
OD ratio
04
r o ( D 0 Q O m
h e m - *
m
m - e
6
-TAP2
FIGURE 4 - Expression of the TAP heterodimer. Lysates from
2 x 105 untreated or IFN-y-treated cells were separated by
SDS-PAGE using an 8% resolving gel and transferred to the
nitrocellulose filter. Blots were robed with rabbit sera specific for
human TAPl (a) or TAP2 (b). &ne A, T2;B, DHsp; C, BL28; D,
IFN-y-treated BL28; E, E95B-BL28; F, IFN-y-treated E95BBL28. The positions of the m.w. markers are indicated. The
intensity of the specific bands was measured by densitometry, and
the corresponding OD is shown at the bottom of each blot. One
representative experiment out of 3 is shown in the figure.
convertant. The extent of the defect and the effect of IFN-y
treatment were investigated in time kinetics experiments.
ATP-dependent translocation increased with time in the
IARC 139 cell line, with a maximum after 15-20 min (Fig. 5).
DEFECTS O F ANTIGEN PROCESSING IN BL CELLS
45
AUF 5.13
IARC 139
/“
35]
255
* untreated - ATP
untreated + ATP
IFN-y-treated ATP+ IFN-y-treated ATP’
-8-
BL28
-A-
W
E
::wc
4 7
35
BL28
J’ 1
E95B-BL28
Eg5B-BL7
IARC 139
5
0
0
5
10
15 20
250
5
10
15 20
25
log [ Fluorescence intensity 3
Time (min)
F~GURE
5 - Time kinetics of peptide translocation in untreated
and IFN-y-treated cells. Aliquots of 1.5 x lo6cells were permeabilized with streptolysin-0. After addition of the 12SI-peptidelibrary,
cells were incubated at 37°C for the indicated times either alone
(ATP-) or in the presence of 10 mM ATP (ATP+). The glycosylated peptides were recovered by absorption to Con-A Sepharose and quantitated in a gamma-counter. One representative
e eriment out of 3 performed with each pair of untreated and
1%-y-treated cell lines is shown in the figure.
TABLE I - ATP- AND TAP-DEPENDENT PEPTIDE TRANSPORT
IN BLAND LCL CELLS
Cell line
cpm
- ATP
= SE’
+ ATP
666 5 131 5,977 2 89
T1
595 5 81
401 f 65
T2
BL28
3,444 2 256 6,181 f 359
E95B-BL28 3,584 & 256 9,481 f 116
IARC 139 2,276 2 200 16,168 2 1,473
ATP’IATP- ratio:
8.9
0.7
1.8
2.6
7.1
‘Values represent the mean 2 SE of 2-7 independent experiments where each LCL/BL pair was tested in parallel.2Ratio
between recovered peptide library in the presence and absence of
exogenous ATP.
BL28 showed a similar kinetics, but the ATP-dependent
transport remained below 20% of that achieved in the LCL
even after prolonged incubation. In contrast, significant peptide translocation occurred in E95B-BL28 when the assay was
run for 20-25 min. In accordance with the significant upregulation of TAP1 and TAP2 demonstrated in Western blots,
cytokine treatment increased the ATP-dependent peptide
transport to LCL-like levels in both BL lines (Fig. 5).
Cytokine treatment does not restore presentation of the
endogenous EBNA4 and does not enhance presentation
of the pre-formed E441,5424epitope
Cytokine treatment induced a significant increase of surface
HLA A l l expression as detected by FACS analysis of BL cells
stained with the All-specific MAb A U F 5.13 (Fig. 6). We
tested, therefore, whether the higher A1 1 expression together
with the restored T A P function were sufficient to reconstitute
presentation of the resident EBNA4 in E95B-BL28 or enhanced antigen presentation in cells infected with the VaccEBNA4 and Vacc-MetIVT recombinants. As illustrated in
Figure 7, cytokine-treated E95B-BL28 remained resistant to
lysis by E 4 4 L M 2 4 - ~ p e CTLs,
~ i f i ~ and there was no detectable
FIGURE
6 - Cytokine-induced up-regulation of class I antigens.
BL28 and E95B-BL28 were cultured for 48 hr in medium alone or
medium containing 500 IU/ml IFN-y. Total MHC class I expression was detected by using the MAb W6/32, while HLA-A11
expression was assessed by the MAb AUF 5.13. Arrows indicate
the histogram corresponding to the cytokine-treated cells. Shaded
areas indicate the negative control stained with the FITCconjugated antibody alone.
improvement in the presentation of the epitope following
infection with high multiplicities of Vacc-EBNA4 or VaccMetlVT.
Maturation and transpofl of MHC class I are normal in BL cells
Failure to achieve efficient peptide loading in the ER,
whether due to inefficient peptide translocation or lack of
appropriate molecular chaperones, results in the accumulation
of “empty” class I molecules that fail to mature to fully
sialylated forms and are degraded intracellularly before they
can reach the cell surface (Ljunggren et al., 1990). To test
whether a similar phenomenon could explain the defective
antigen presentation, the maturation of class I heavy chain was
compared in BL28, E95B-BL28 (not shown), and in the
corresponding LCL IARC 139 (Fig. 8). The cells were labeled
with [%]Met for 10 rnin and chased at 37°C for u p to 120 min.
At each time point, class I molecules were immunoprecipitated from the cell extracts with the W6/32 MAb and half of
the immunoprecipitate was digested with Endo H. After 0 rnin
of chase all heavy chains were Endo H-sensitive in both BL28
and IARC 139. The maturation of heavy chains to sialylated
Endo H-resistant forms was virtually complete after approximately 3 W O rnin of chase in IARC 139. A slight delay in
maturation was detected in BL28, but only a minor fraction of
Endo H-sensitive molecules were present after 60 rnin of
chase and virtually all class I molecules had become Endo
H-resistant after 90 min. The results are consistent with the
formation of stable class I complexes in the E R and their
normal transport to the Golgi complex.
DISCUSSION
Our results confirm previous studies suggesting that an
impaired antigen-presenting capacity may contribute to the
escape of tumor cells from CTL surveillance. W e have shown
that the EBV-positive E95B-BL28 cell line is unable to present
a n HLA A1 1-restricted epitope from EBNA4, though it
expresses relatively high levels of the protein encoded by the
256
FRISAN ETAL.
BL28
A
DHSp
E95B-BL28
mw
46-
50 -
*
nd
-
lARCl39
uninfected
46-
BL28
endoHmin
0
1
+
-
+
10
-
+
30
-
+
60
-
+
90
-
+
120
B
L
100
I
i#
8
'"1
80
C
IARC 139 Endo H '
+ IARC139 Endo H '
60
(P
f
40
"
1FN-y
-
+
-
+
:
0
FIGURE7-Presentation of the E441M24
epitope in IFN-ytreated BL cells. Untreated and IFN-y-treated cells were infected
for 4 hr with Vacc-EBNA4 at an m.0.i. of 10 or Vacc-MetIVT at an
CTL clone
m.0.i. of 0.3 before labeling. The E441M24-specific
BK289 was used as effector. The percent relative lysis was
calculated as described in Figure 2. Mean k SE of 3 experiments.
nd, Not done. *The DHsp LCL carries a mutated EBV strain and
does not present the endogenous E441M24
epitope.
resident virus (Fig. 1) and is efficiently lysed after preincubation with the relevant synthetic peptide (Fig. 2u). Both
E95B-BL28 and the parental EBV-negative BL28 could present the E441w24epitope after infection with a vaccinia recombinant that over-expressed the full-length EBNA4 protein
(Fig. 2u). However, only 2MO% maximal lysis could be
obtained even at an m.0.i. 10-fold higher than that required for
100% maximal lysis of a control LCL of normal B-cell origin
(Fig. 26). The defect was not overcome by expression of the
pre-formed epitope (Fig. 3), suggesting the impairment of
functions down-stream of the production of antigenic peptides.
Defects of peptide transport have been proposed to provide
a route of immunoescape in several types of tumor. Restifo el
al. (1993) have shown that the poor antigen-presenting capacity of small-cell carcinoma lines correlates with impaired
transport of class I molecules from the ER to the cell surface,
suggesting that peptides may not be available for binding to
MHC. In line with this possibility, the transcription of the
TAPl and TAP2 components of the peptide transporter was
down-regulated in the tumors and class I expression could be
restored, along with antigen-presenting capacity, by treatment
with IFN-y. Down-regulation of the TAPl and, to a minor
extent, TAP2 proteins was also reported in BL lines that have
retained in vim the phenotypic characteristics of the original
tumor (Rowe et al., 1995), and down-regulation of TAPl was
demonstrated by immunohistochemical methods in a signiticant proportion of cervical carcinoma biopsies (Cromme et aL,
1994). Khanna et al. (1994) have shown that low levels of TAPl
mRNA are expressed in the EBV-negative BWO cell line that
cannot present an HLA B8-restricted epitope from EBNM.
W
8
-
*
W
20
0
10
U EL28 EndoH'
+ BL28EndoH'
30
60
Time (min)
90
120
FIGURE
8 - Assembly and transport of class I molecules in LCL
and BL cells. (a) Cells were ulsed labeled with 100 FCi [%]Met
and chased at the indicatecftimeito monitor class I exocytosis.
Detergent extracts were immuno recipitated with the MAb W6/
32, and the mock- or Endo HJgested class I molecules were
resolved by SDS-PAGE. The Endo H-sensitive (lower band and
Endo H-resistant (u per band) class I heavy chains are c early
visible in the autoraJoograms. The positions of the m.w. markers
are indicated. (b) The intensity of the class I-specific bands was
quantitated by densitometry and the percentage of Endo H-sensitive (Endo Hs) and Endo H-resistant (Endo Hr) class I heavy
chains was calculated for each chase time.
I'
The involvement of the transporter was confirmed by the
capacity to overcome the defect in cells transfected with a
minigene encoding the epitope preceded by an ER localization
signal. This was further substantiated by the partial reconstitution of antigen-presenting capacity in the EBV-negative BJAB
and BLAl cell lines transfected with the EBV-encoded LMPl,
in parallel with the up-regulation of TAPl and -2 expression
(Rowe et al., 1995). We have confirmed the observation that
TAP1 and TAP2 are down-regulated in BLcells in a phenotypedependent fashion, and we have extended this finding by
demonstrating that the decreased expression of TAPS correlates with impaired peptide transport in a streptolysin-0
translocation assay (Figs. 4,5; Table I). Our results are also in
line with the previously reported capacity of LMPl to modulate TAP expression since the LMP1-positive E95B-BL28 cell
line, that has acquired an LCL-like phenotype following EBV
conversion, expressed significantly higher levels of TAPl and
TAP2 compared with the EBV-negative BL28 and behaved
like the corresponding LCL in peptide-translocation assays.
Impairment of transporter function is likely to play an important role in the immune escape of BL cells. However, this
defect cannot fully explain the poor antigen presentation
highlighted in this report since reconstitution of TAP expression and function by treatment with IFN-y was not sufficient to
restore presentation of the resident EBNA4 in E95B-BL28
DEFECTS OF ANTIGEN PROCESSING IN BL CELLS
and did not enhance presentation of the over-expressed
full-length protein or pre-formed epitope in cells infected with
recombinant vaccinia viruses (Fig. 7).
Failure to load antigenic peptides onto class I molecules in
the ER could result in the formation of unstable complexes
that are destroyed before they can reach the cell surface. The
effects of inefficient peptide loading have been extensively
investigated in the TAP-deficient T2 cell line (Salter and
Cresswell, 1986). Shortage of peptides or inefficient association with TAPSwere shown to lead to the accumulation of free
heavy chains that can be detected by sequential immunoprecipitation with anti-p2m and anti-heavy chain antibodies. Under
these conditions, the single N-linked glycan of the heavy chain
remains Endo H-sensitive for at least 3 hr after synthesis,
consistent with the failure to transport the unstable complexes
from the ER to the trans-Golgi and further to the cell surface
(Salter and Cresswell, 1986). According to these criteria, the
defect carried by the BL lines appears to be substantially
different from that observed in the in vitro selected mutants.
The rate of class I maturation was similar in BL28 and in the
corresponding LCL, IARC 139 (Fig. 8), suggesting that the
majority of class I molecules expressed in the tumor cells form
stable complexes with p2m and peptides and are normally
transported to the cell surface. Similar results were reported
by Rowe et al. (1999, who failed to demonstrate any accumulation of free class I heavy chains by sequential immunoprecipitation of extracts from 7 BL lines. These findings are in line
with our previous observation that the low class I antigen
expression of phenotypically representative BL lines is mainly
due to transcriptional down-regulation of class I genes and
correlates with the absence of specific transcription factors
required for class I promoter activity (Imreh et al., 1995).
The results presented here suggest that the defective antigen presentation of BL28 and E95B-BL28 is related to events
occurring after the generation of antigenic peptides but before
TAP-dependent peptide transport and MHC class I assembly
and maturation. It is noteworthy that low levels of presentation
were achieved when the relevant antigen was over-expressed
following infection with high multiplicities of the VaccEBNA4 and Vacc-MetIVT viruses. Similarly, in the work of
Rowe et al. (1995), presentation of a B35-restricted epitope
was observed after infection of BL lines with high multiplicities
of vaccinia recombinants for 18 hr and was therefore dependent on levels of expression much higher that those observed
in regular EBV-infected cells. One possible explanation for
these findings is that peptides produced by proteasomal
degradation of intracellular antigens may be destroyed in the
257
cytosol before entering the ER lumen. This could result in the
selective loss of peptides derived from proteins expressed at
low levels or with a short metabolic turnover, while peptides
derived from highly expressed proteins may still be present in
sufficient amounts to assure the correct assembly of the
relatively few class I molecules expressed in BL cells. Two
mechanisms can be envisaged. It has been shown that cytosolic
proteases effectively degrade synthetic peptide substrates (Roelse et al., 1994). Conceivably, such proteases may be expressed
at higher levels or in different combinations in BL cells
compared to LCLs. Alternatively, cytosolic peptide chaperones may be expressed at lower levels in BL cells. Recent
evidence suggests that HSPs play a role in antigen processing/
presentation by binding antigenic peptides (Srivastava et al.,
1994). Immunization, with purified hsp70 and gp96 preparations from mouse and rat tumors were shown to induce
rejection of the corresponding live tumors in syngeneic hosts,
while HSPs isolated from normal tissues did not elicit immunity against any of the tumors tested. Cytosolic HSPs, such as
hsp70, may act by shielding cytosolic peptides from further
degradation to single amino acids or by preventing their
unproductive coupling to other proteins (reviewed in Li and
Srivastava, 1994). Indeed, preliminary results suggest that
BL28 and E95B-BL28 express significantly lower levels of
hsp70 compared to regular LCLs. The significance of this
observation for the defective antigen presentation of the tumor
cells is presently under investigation.
Taken together, our findings add a new dimension to the
complexity of the defects that results in impaired antigenpresenting capacity in tumor cells. The accumulation of
multiple defects leading to a non-immunogenic phenotype may
provide a selective advantage for in vivo growth, in spite of the
continued expression of tumor antigens that can be the target
of specific rejection responses.
ACKNOWLEDGEMENTS
We thank Dr. J. Trawsdale for the kind gift of TAP1- and
TAPZspecific rabbit anti-sera. This work was supported by
grants from the Swedish Cancer Society, the Children’s Cancer
Foundation, the Karolinska Institute and the Magnus Bergwal
Stiftelse, Stockholm, Sweden, and by the Lucille P. Markey
Charitable Trust and the University of Virginia Cancer Center
Support Grant. T.F. was supported by a fellowship from the
Concern 11 Foundation, Los Angeles, CA, and J.L. was
supported by a fellowship from the Cancer Research Institute,
New York and the Concern Foundation, Los Angeles.
REFERENCES
CROMME,
F.V., AIREY,
J., HEEMELS,M.T., PLOEGH,
H.L., KEATING,
P.J., STERN,P.L., MEIJER,C.J. and WALBOOMERS,
J.M., Loss of
transporter protein, encoded by TAP-1 gene, is highly correlated with
loss of HLA expression in cervical carcinomas. J. a p . Med., 179,
335-340 (1994).
DE CAMPOS-LIMA,
P.O., LEVITSKY,
V., BROOKS,
J., LEE,S.P., Hu, L.F.,
A.B. and MASUCCI,M.G., T cell responses and virus
RICKINSON,
evolution: loss of HLA A1 1-restricted CTL epitopes in Epstein-Barr
virus isolates from highly A1 1-positive populations by selective mutation of anchor residues./. u p . Med., 179,1297-1305 (1994).
GACZYNSKA,
M., ROCK,K.L. and GOLDBERG,
A.L., y-Interferon and
expression of MHC genes regulate peptide hydrolysis by proteasomes.
Narure (Lond.),365,264-267 (1993).
HEEMELS.
M.T., S C H U M A C HT.N.,
~ R , WONIGEIT,
K. and PLOEGH,
H.L.,
Peptide translocation by variants of the transporter associated with
antigen processing. Science, 262,2059-2063 (1993).
IMREH, M.P., ZHANG,
0.J.. DE CAMPOS-LIMA,
P.O., IMREH,S., KRAUSA,
P.. BROWNING.
M.. KLEIN. G. and MASUCCI.M.G.. Mechanisms of
allele-selective’down-regulation
of HLA class I in Burkitt’s lymphoma.
Inr. J. Cancer, 62,9&96 (1995).
KELLY,A., POWIS,S.H., KERR, L.-A,. MOCKRIDGE,
I., ELLIOTT,T.,
1
~~
~
B., ZIEGLER,
A., TROWSDALE,
J. and
BASTIN,J., UCHANSKA-ZIEGLER,
TOWNSEND,
A., Assembly and function of the two ABC transporter
proteins encoded in the human histocompatibility complex. Nature
(Lond.),355,641-644 (1992).
KHANNA,R., BURROWS,
S.R., ARGAET,V. and Moss, D.J., Endoplasmic reticulum signal sequence facilitated transport of eptide epitopes
restores immunogenicity of an antigen processing Befective tumour
cell line. Inr. Imrnunol., 6,639-645 (1994).
LI, Z . and SRIVASTAVA,
P.K., A critical contemplation on the role of
heat shock proteins in transfer of antigenic eptides during antigen
presentation. Behring Inst. Mitt., 9 4 , 3 7 4 7 (1994).
LJUNGGREN,
H.G., STAM,N.S., O H L ~ N
C.,, NEEFJES,
J.J., HOGLUND,
P.,
T.N.M., TOWNSEND,
A.,
HEEMELS,M.T., BASTIN,J., SCHUMACHER,
KARRE,K. and PLOEGH,H.L., Em ty MHC class I molecules come out
in the cold. Narure (Lond.),346,4%480
(1990).
MANN,K.P., STAUNTON,
D . and THORLEY-LAWSON,
D.A., EpsteinBarr virus-encoded protein found in the plasma membranes of
transformed ce1ls.J. Virol., 55,710-720 (1985).
MOMBURG,
F., ROELSE,J., HOWARD,
J.C., BUTCHER,
G.W., HAMMERG.J. and NEEFJES,J.J., Selectivity of MHC-encoded peptide
LING,
258
FRlSAh' ETAL.
trans orters from human, mouse and rat. Nature (Lond.), 367,648-651
(1994.
MURRAY,
R.J., KURILLA, M.G., GRIFFIN,
H.M., BROOKS,
J.M., MACKETT, M., ARRAND,
J.R., ROWE,M., BURROWS,
S.R., Moss, D.J., KIEFF,
E. and RICKINSON,
A.B., Human cytotoxic T cell responses against
Epstein-Barr virus nuclear antigens demonstrated using recombinant
vaccinia viruses. Proc. naf.Acad. Sci. (Wash.), 87,2906-2910 (1990).
RESTIFO,N.P., ESQUIVEL,
F., KAWAKAMI,
Y., YEWDELL,
J.W., MULE,
J.J., ROSENBERG,
S.A. and BENNINK,
J.R., Identification of human
cancers deficient in antigen processing. J. exp. Med., 177, 265-272
( 1993).
RICKINSON,
A., MURRAY,
R., BROOKS,J., GRIFFIN,H., Moss, D. and
MASUCCI,M., T cell recognition of Epstein-Barr virus associated
lymphomas. In: A. McMichael and L. Franks (eds.), A new look at
tumor immunology, pp. 53-80, CSHL Press, New York (1992).
ROELSE, J., GROMME,M., MOMBURG,
F., HAMMERLING,
G. and
NEEFJES,
J., Trimming of TAP-translocated pe tides in the endoplasmic reticulum and in the cytosol during recycying. J. exp. Med., 180,
1591-1597 (1994).
ROWE,M. and 12 OTHERS, Restoration of endogenous antigen process-
ing in Burkitt's lymphoma cells by Epstein-Barr virus latent membrane
protein-]: coordinate up-regulation of peptide trans rters and HLAclass 1 antigen expression. Europ. J. Immunol., 25,13%-1384 (1995).
SALTER,R.D. and CRESSWELL,
P., Impaired assembly and trans
HLA-A and -B antigens in a mutant T X B cell hybrid. EMBXoft
943-949 (1986).
SRIVASTAVA,
P.K., UDONO,H., BLACHARE,
N.E. and LI, Z., Heat shock
proteins transfer peptides durin antigen processing and CI'L priming.
mmunogenefics, 39,93-98 (199f).
TORSTEINSDOTTIR,
S., MASUCCI,
M.G., EHLIN-HENRIKSSON,
B., BRAUTBAR, C., BASSAT,
B., KLEIN,G. and KLEIN, E., Differentiation dependent sensitivity of human B-cell derived lines to major histocompatibility complex-restricted T-cell cytotoxicity.Proc. nar. Acad. Sci. (Wash.),
83,5620-5624 (1986).
TOWNSEND,
A., ROTHBARD,
J., GOTCH,F., BAHADUR,
B., WRAITH,D.
and MCMICHAEL,
A,, The epitopes of influenza nucleoprotein recognized by cytotoxic T lymphocytes can be defined with short synthetic
peptides. Cell, 44,959-968 (1986).
TOWNSEND,
A. and TROWSDALE,
J., The transporters associated with
antigen presentation. Semin. Cell Biol., 4,53-61 (1993).
if
Документ
Категория
Без категории
Просмотров
2
Размер файла
908 Кб
Теги
884
1/--страниц
Пожаловаться на содержимое документа