Vol. 2. 593-600. Marc/i 1996 Clinical Cancer Research 593
Lack of T-Cell-mediated Recognition of the Fusion Region of the
pmLIRAR-a Hybrid Protein by Lymphocytes of Acute Promyelocytic
Leukemia Patients’
Said Dermime, Carla Bertazzoli,
Edoardo Marchesi, Fernando Ravagnani,
Kurt Blaser, Gian Marco Corneo,
Enrico Pogliani, Giorgio Parmiani, and
Carlo Gambacorti-Passerini2
Division of Experimental Oncology D and Blood Bank, Istituto
Nazionale Tumori, Via Venezian 1, 20133 Milan. Italy [S. D.. C. B..
E. M.. F. R., 0. P.. C. G-P.l: Swiss Institute of Allergy and Asthma
Research. Davos, Switzerland [K. B.]: and Section of Hematology.
University of Milan, S. Gerardo Hospital, Monza, Italy [G. M. C.. E. P.1
ABSTRACT
In previous studies, it was shown that the fusion regionof the pmIIRAR-a protein, expressed by acute promyelo-
cytic leukemia (APL) cells, can be specifically recognized in
vitro by donor (D. E.) CD4 T cells in a HLA class II DR11-restricted fashion. We present here the results on the rec-
ognition of several pm1IRAR-a peptides by APL patients
expressing HLA DR1 1. The in vitro immunization of peniph-
eral blood lymphocytes from four patients in remission
(S. R., F. R., M. M., P. G.) with BCR1/25, a 25-men pml/
RAR-a, did not elicit either a polyclonal or a clonal immuneresponse specific to the peptide. We then generated new
donor anti-pmlIRAR-a CD4� 1-cell clones. These cloneswere tested for their recognition of BCR1/25. One clone
(C3/5, CD3�, CD4�, CD8) was selected for further analy-sis. Clone C3/5 showed specific proliferation, cytotoxicity,and cytokine (tumor necrosis factor a, granulocyte-mac-rophage colony-stimulating factor) production when chal-
lenged with autologous lymphoblastic cell lines pulsed with
peptide BCR1/25. C3/5 cells developed specific proliferation
and cytotoxicity when challenged with peptide-pulsed lym-phoblastic cell lines and peripheral blood lymphocytes from
the four DR1 1 � APL patients. APL blasts, available onlyfrom patients F. R. and P. G., were not lysed by C3/5 andwere unable to present peptide BCR1/25. Incubation of APL
cells with IFN-�j failed to induce HLA class II molecules and
recognition by the C3/5 clone. Since APL cells do not express
HLA class II molecules, we tested in two donors (D. E. and
C. H. R.) and in patients S. R. and P. G. whether the use of9-mer peptides (BCR1/9) would generate a CD8IHLA class
I-restricted response. No peptide-specific 1-cell line or clone
Received 6/15/95; revised 10/18/95: accepted 1 1/28/95.I This work was supported in part by the Italian Association tir Cancer
Research. the Italy-United States Program on Therapy ofTumors (Grant
531 23.12.87). and a donation from CEVA SpA, Monza.
2 To whom requests for reprints should be addressed. Phone: 39.2.239-0207: Fax: 39.2.236-2692.
could be generated from both donors and patients. These
findings are discussed in relation to possible therapeutic
approaches to the immunotherapy of APL.
INTRODUCTION
The pml/RAR-cs fusion protein, specifically expressed in
APL3 ( 1-4) cells. creates a new tumor-specific amino acid
sequence. It has been shown that a 25-men peptide (BCRI/25),
encompassing the fusion region, contains an antigenic site (ab-
sent in the normal parent molecules) recognizable by CD4�
T-cell clones of a healthy donor (D. E.) in an HLA DR1 I-
restricted fashion on presentation by autologous APCs (5. 6).
Antipeptide CD4 clones also recognize pml/RAR-a-transfected
LCLs (6).
The possible recognition of the pml/RAR-a protein by 1
lymphocytes of APL patients would open the possibility of
directing the patient’s immune system against the leukemic cells
through the recognition of a tumor-specific/transformation-re-
lated molecule (5). Although binding motifs for most DR mol-
ecules were not known at the time of the study, our previous
data (6) indicated that DR1 1 functions as a restriction element in
this case. For this reason, we focused on the analysis of DR 1 1 �
of DRI 1 APL patients.
In the present work, we studied the in vitro immune re-
sponse of peripheral blood lymphocytes from four HLA DR1 I
APL patients to several pml/RAR-a peptides.
PATIENTS AND METHODS
Patients
Four APL patients were studied after informed consent was
obtained. Three patients (F. R., S. R., and P. G.) were in first
remission, and one patient (M. M.) was in second remission.
Two patients (S. R. and M. M.) underwent autologous bone
marrow transplantation: the conditioning regimen included
high-dose chemotherapy but not total-body irradiation. At the
time of the study, patients had not received treatment for at least
6 months. and their peripheral counts were within normal
ranges.
I The abbreviations used are: APL. acute promyelocytic leukemia: LCL,lymphoblastoid cell line; APC. antigen-presenting cell; PBMC. periph-
eral blood mononuclear cell: LAK. lymphokine-activated killer: IL.
interleukin; HPLC, high-performance liquid chromatography: HS. hu-
man serum; dThd, thymidine; PBL. peripheral blood lymphocyte:
TNF-a. tumor necrosis factor a; GM-CSF, granulocyte-macrophage
colony-stimulating factor; MAb, monoclonal antibody; PHA, phytohe-
magglutinin: TT, tetanus toxoid; SI. stimulation index; TCR. T-cell
receptor.
on April 8, 2021. © 1996 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
594 Immune Recognition of the pml/RAR-a Protein
PBMCs
Peripheral blood was obtained from two healthy donors
(D. E. and C. H. R.) and four APL patients in remission (S. R.,
F. R., M. M., and P. G.) after informed consent. PBMCs were
isolated by centnifugation on Ficoll gradients. The HLA typing
was as follows: DE (A2,23, B49,35, DR1 1,13), CHR (A2,l I,
B35,60, DR1,4), SR (Al,24. B51, DR8,ll), FR (A23, B49,5l,
DR4,l 1), MM (A23,l0, B27,49, DRIO,l I), and PG (Al,2, Bl5,
DRI,l I).
Cell Lines
Lymphoblastoid cell lines (DE.LCL, SR.LCL, FR.LCL,
MM.LCL, PG.LCL, and CHR.LCL) were prepared from PBMCs
as previously described (6). LAK cell lines were prepared by
culturing PBMCs with recombinant IL-2 (l0� units/ml; Euro-
Cetus By, Amsterdam, the Netherlands). The Daudi cell line
was cultured in RPMI 1640 (Biowhittaker, Walkersville, MD)
plus 10% FCS (Biological Industries, Kibbutz Beth Haemek,
Israel). FR.APL cells were obtained from bone marrow or
peripheral blood of an APL patient (F. R.) and frozen in liquid
nitrogen. They were thawed and cultured for I 8 to 24 h before
use in proliferation or cytotoxicity tests.
Peptides
The following peptides were used: BCRI/25, a 25-men
(NSNHVASGAGEAAIETQSSSSEEIV) peptide, and a pool of
nine different 9-men peptides (VASGAGEAA, ASGAGEAAI,
SGAGEAAIE, GAGEAAIET, AGEAAIETQ, GEAAIETQS,
EAAIETQSS, AAIETQSSS, and AIETQSSSS) collectively in-
dicated as BCRI/9. They encompass the BCR1-type fusion
region of the pml/RAR-a protein (4). The peptides were syn-
thesized by the University of Wisconsin Biotechnology Center
(Madison, WI) or American Peptide Company (Sunnyvale, CA)
and purified using high-performance liquid chromatography to a
minimum purity of 95%.
Activation of Lymphocytes
The protocol described by Chen et a!. (7) was followed. In
brief, 50 x 106 fresh PBMCs (l0� for patient P. G. and donor
C. H. R.) were incubated in a humidified atmosphere at 37#{176}Cin
5% CO, with 20 p.M corresponding peptides at 5 X 106 cells/mI
serum-free RPMI 1640 medium for I h. Fresh autologous PB-
MCs (50 X 106 cells in 10 ml RPM! 1640 medium + 10% HS)
were added to each APC culture in culture flasks (75 cm2). The
flasks were incubated upright for 7 days. Primed lymphocytes
were then harvested and restimulated with irradiated (3000 rad)
autologous PBMC-pulsed peptides (20 �J.M) at a lymphocyte:
APC ratio of I :3. The next day, IL-2 (10 units/ml) was added,
and the percentage of HS was increased to 10%. Five days later,
one half of the medium from each culture was replaced with
RPMI 1640 + 10% HS + 10 units/ml IL-2. The derived T-cell
lines were left in culture for 7 days before cloning on day 21
poststimulation. They were maintained in culture by periodic
stimulation (every 7-10 days) using irradiated (8000 rad) auto-
logous LCL-pulsed peptide (20 �.LM) and IL-2 (10 units/mi). To
test for proliferative activity of the derived T-cell lines, cells
were washed and placed in U-bottomed 96-well plates (5 X l0�
cells/well). They were stimulated with irradiated (8000 nad)
autologous LCLs (l0� cells/well) in the presence or absence of
the appropriate peptides. LCL cells were washed three times in
serum-free RPMI 1640 medium, incubated with 20 �LM peptides
as indicated above, and then irradiated. After 48 h of culture, the
plates were labeled with [3H]dThd, further cultured for 18 h,
harvested, and counted in a beta counter.
Generation of Lymphocyte Clones
Primed 1-cell lines were cloned on day 21 by limiting
dilution (20, 10, 5, 2, and 1 cell/well/200 pA RPM! 1640 + 10%HS) in 96-well U-bottomed plates. Autologous LCLs, PBLs,
and allogeneic PBLs (25 X l0� cells/well) were used as a feeder
layer. Autologous PBLs and LCLs were incubated separately in
serum-free RPM! 1640 medium (5 X 106 cells/ml) with 20 �iM
of the same mixture of peptides, used to prime lymphocytes (see
above), for 30 mm at 37#{176}Cbefore irradiation (3000 rad for PBLs
and 8000 rad for LCL cells). X-irradiated (3000 rad) allogeneic
PBLs and IL-2 (50 units/mI) were also added to the wells. The
plates were incubated for 14 days; 100 jil fresh medium and 50
units/ml IL-2 were replaced in each well every 3 days. Wells
with positive signs of growth were selected for expansion (see
‘ ‘Results’ ‘ ). The probability of clonality for each well contain-
ing growth was calculated as described previously (6). Growing
clones were transferred to 96-well flat-bottomed plates, and each
clone was stimulated with 25 X l0� irradiated (8000 rad) autol-
ogous LCLs pulsed with the appropriate peptides (20 jiM) and
25 X l0� (3000 rad) irradiated allogeneic PBLs. !L-2 (50
units/mI) was added to the wells after I day. This activation
cycle was repeated after 7-10 days of culture at lymphocyte:
autologous LCL:allogeneic PBL ratios of I : 1 : I . unless other-
wise indicated.
Screening of Clones
The following assays were carried out.
Proliferation Assay. 1-cell clones were first screened
for proliferative activity 28 days after cloning. Each well was
washed off to remove IL-2 and split into three. The first one was
expanded for further culture. The second well was stimulated
with irradiated (8000 rad) autologous LCLs ( l0� cells/well)
pulsed with the appropriate peptides (20 �1M). The third well was
incubated with irradiated LCLs without peptides. Cells were
cultured for 48 h, incubated for an additional I 8 h with
[3HJdThd (1 p�Ci/well), and harvested. In subsequent expeni-
ments, the incubation time was reduced to 24 h (see “Results”).
Cytotoxic Assay. T-cell clones were first screened for
cytotoxic activity 6 weeks after cloning: one third of the cell
suspension from each clone was admixed with 5tCr-labeled
autologous LCLs (10� cell/well) as a control, and the other third
was incubated with LCLs pulsed with peptide (20 pM). The
assay was performed as previously described (6). The incuba-
tion time was 4-6 h. Spontaneous release never exceeded 15%.
Anti-TNF-a monoclonal antibody (Farmitalia, Milan, Italy) and
TNF-a (EuroCetus BV) were used in some assays (see � ‘Results”).
Cytokine Production and Detection Assays
Cells from clone C3/5 (3 X 106 cells) were cultured in
48-well plates (1.5 X 106 cells/mi) with irradiated (8000 rad)
autologous DE.LCL (3 X 106 cells) in the presence on absence
on April 8, 2021. © 1996 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
SR/9 SR/25 IDE/9/25
Po1�cloma1 T cell lines
Clinical Cancer Research 595
of 20 J.LM peptide (25-men). C3/5 cells (3 X 10#{176})were also
cultured in 2 ml RPMI 1640 + 10% HS in the absence of LCLs.
The supernatant from each culture (2 ml) was collected by
centnifugation after I and 2 days. immediately frozen in liquid
nitrogen. and stored at -80#{176}C until used. The cytokines assayed
in the present study were: IL-2, IL-3, IL-4, IL-6, IFN-y, TNF-a,
and GM-CSF. IL-2 activity was measured by [3HJdThd uptake
of IL-2-dependent cytotoxic T lymphocyte line cells as previ-
ously described (Ref. 8. 1 unit biological activity was referred as
half-maximal proliferation of cytotoxic T lymphocyte line
cells). In brief, serial dilutions of IL-2 or supernatant from each
culture were incubated in 200 �il for 24 h, followed by an
overnight pulse of I p.Ci/well of I3HldThd. IL-3 was assayed by
using a Quantikine Human IL-3 Immunoassay (Research and
Diagnostic Systems, Minneapolis, MN) with a detection limit of
30 pg/mi. IL-4 and IFN--y were measured using the sandwich
ELISA as described in detail elsewhere (9). The mouse MAb
8F I 2 ( 10) and biotinylated 3H4, generously provided by C. H.
Heusser (Ciba-Geigy Ltd.. Basel. Switzerland), were used for
IL-4, and MAbs 43-I I and biotinylated 45-15. kindly supplied
by S. Alkan (Ciba-Geigy Ltd.. were used for IFN--y. The
sensitivities of the lL-4 and IFN--y ELISA were 30 and 40
pg/nil. respectively. IL-6 and TNF-cs were analyzed using kits
from Chromogenix AB (Molndal. Sweden), with a detection
limit of 4 pg/mI as given by the manufacturer. The human
GM-CSF was assayed using Innotest human GM-CSF from
Innogenetics NV (Antwerp, Belgium) with a detection limit of
8 pg/ml.
PHA and TI Assays
PBLs were simultaneously thawed, washed, and seeded at
l0� cells/well. PHA (Murex, Temple Hile. England) was added
at I �ig/rnl, and proliferation was assessed as [3H]dThd uptake
(6-h incubation) after 72 h. For proliferation to Ti’ ( 10 p.g/ml:
Connaugh Lab), 2 x l0� PBLs/well were cultured for 6 days
and labeled with l3HIdThd tc�r the last 18 h of culture.
Immunofluorescence Assay
This assay was performed as previously described ( I I).
The MAbs used were in the form of diluted ascites: anti-CD3.
anti-CD4, anti-CD8, anti-TCR-ct�3, anti-TCR-y�, anti-HLA
class I (W6/32). and anti-HLA class II (Dl-12). Samples were
analyzed using flow cytometry (FACs IV: Becton Dickinson,
Sunnyvale. CA) with a logarithmic signal amplification.
Screening for HLA-binding Motifs
A list including the binding motifs for 35 HLA class I
specifIcities was kindly supplied by Dr. H. G. Rammensee
(Heidelberg. Germany) and includes the following specificities:
Al. 0201/5. 3. 11. 24. 31. 33, 68: B7, 8, 2702/5, 3501/3, 3701,
3901/2. 40. 4402/3. 5101/2/3, 5201, 53, 58. 60, 61, 62, 78;
CwO3Ol. 0401, 0602, 0702.
HLA Stabilization Assay
The ability of peptides to bind HLA molecules was as-
sessed by the peptide-mediated increase in the amount of HLA
expression on the membrane of the 12 ( I 2). L72 1 .22l( 1 3), and
ST-EMO ( 14) cell lines. Cw6 and Cw0702 transfectants of T2
20000
17500
15000
�- 12500
U
C10
� 7500
5000
2500
0
Fig. I Proliferative response of three polyclonal T-cell lines ISRI9,
SR125 ((mm an APL patient in remission), and DE/9/25 (from a healthy
donor)l against BCRI/9. BCR1/25, or BCRI/9 + BCR1/25 peptides,respectively. Each T-celI line (5 X l0� cells/well) was cultured for 48 h
with irradiated (80(X) rad) autologous EBV-transformed LCLs (l0�
cells/well) in the presence (� or absence (m) of the corresponding
peptide (�, lymphocytes cultured in RPMI 1640 + 10% HS in the
absence of both LCLs and peptides. They were incubated ftr an
additional 18 h with i�HidThd and harvested. LCLs were pulsed with
peptides in serum-free medium for 30 mm before irradiation. cpm
values represent the mean of six replicates. Bars. SD.
and L721.22l were also used (kindly supplied by Dr. D. Schen-
del, Munich Germany). Cells were washed in serum-free RPMI
1640 and incubated with 20-100 pg/ml peptide for 18 h at
37#{176}C.Expression of HLA class I molecules was evaluated by
indirect immunofluorescence with W6/32 MAb or with mono-
specific MAbs GAP-A3, 4E (anti-B/C), or BB7.2 (anti-A2).
Positivity was defined as an increase by at least 50% in the
median channel of fluorescence.
RESULTS
Polyclonal 1-Cell Lines. The results for the proliferative
activity of 1-cell lines, obtained from an APL patient in nemis-
sion (S. R.) and a healthy donor (D. E.), are presented in Fig. I.
SR/9 and SR/25 were obtained by culture in the presence of
BCRI/9 or BCRI/25; DE/9/25 was derived from cultures in the
presence ofboth BCR1/9 and BCRI/25 peptides. SR 1-cell lines
did not show a specific proliferative activity against autologous
peptide-pulsed LCLs. A low but reproducible SI (SI cpm test
well/cpm control well) of 2.2 was obtained for the DE/9/25
T-cell line. These lines consisted of a mixture of CD4 and
CD8-a/�3 lymphocytes. No specific cytotoxic activity against
peptide-pulsed LCLs was obtained in these cell lines (data not
shown).
1-Cell Clones. The 1-cell lines SR/9, SRJ25, and DE/
9/25 were cloned by limiting dilution. Two hundred fifteen
clones (with a clonal probability of >90%) were obtained; 52,
41, and 122 clones were generated from SR/9, SR/25, and
on April 8, 2021. © 1996 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
100
90
U) 80...4
In�1)
p-I
60U
p.-’
11�.
..1
U 40InD.
(J�
�\. 20
10
0
4000
,-% 3500
,� 3000
�i 2500
�2000
C 1500
.9 �i� 600L
4J� 400
0 200
C.)
0
A B C
Target cells
0 1 2
Days in culture
596 Immune Recognition of the pmlIRAR-cs Protein
Fig. 2 Cytotoxic activity of clone C3/5 (E. effector cells) against
autologous DE.LCL (T, target cells) in the presence of BCR1/9 (B) and
BCRI/25 (C) or absence (A) of peptide. LCLs at 5 X 106 cells/ml were
incubated with 51Cr for 90 mm with or without peptides (20 p.M). After
washing. the peptides were readded to the target (10� cells/well) at the
same concentration.
DE/9/25. respectively. Among the 52 SR/9 and the 41 SR/25
clones tested, no specific cytotoxic or proliferative activity was
obtained. Proliferation against LCLs pulsed with control pep-
tides or with PBS was observed in several clones, suggesting
autoreactivity or anti-EBV reactivity (data not shown). Among
the 122 clones obtained from the DE/9/25 T-ceil line, 59 were
selected according to their specific proliferation (SI >2) or
cytotoxicity (specific lysis >10%). Fifty-five clones showed
specific proliferation to the peptide pool (SI ranged from 2 to
16.47) with fourciones showing specific lysis also [B9/5 (17%),
C3/l (20%), C3/5 ( 19%), and C4/l (13%)1. An additional four
clones showed specific lysis only [A4/5 (19%), E2/l (10.5%),
E3/l (23%), and F2/l ( 15%)]. The phenotype of these clones
was uniformly CD3�, CD4�, CD8, TCR-a/�, and TCR-
-y/�l� . One clone (C3/5), which showed the highest and the most
stable results, was selected for further analysis. In subsequent
experiments the proliferative and cytotoxic activity of the C3/5
clone against BCRI/25 and BCR1/9 was evaluated. BCR1/9 did
not induce specific proliferation (data not shown), indicating
that BCR1/25 was the peptide recognized in the peptide mixture
initially used for stimulation. The results for the cytotoxic ac-
tivity of C3/5 against autologous LCLs pulsed separately with
BCR1/25 and BCR1/9 are shown in Fig. 2.
T-cell clones were also produced, with the same stimula-
tion protocol and using combined or separate stimulation with
BCRI/25 and BCRI/9, from three additional DR1 I � APL pa-
tients (F. R., M. M., and P. G.) and with BCR 1/9 from donors
C. H. R.. L. R., and B. C. A total of 1015 clones (155 from
patient F. R., 106 from patient M. M., 251 from patient P. G.,
Fig. 3 Specific cytokine production by the C3/5 clone. Cells ( 1 .5 X
l06/ml) were cultured with irradiated (8000 rad) autologous DE.LCL
(1.5 X 106 cells/mI) in the presence (#{149}and V) or absence (0 and 7) of
the BCR1/25 peptide (20 pM). Supernatants were harvested on days 1
and 2 and assayed for the presence of TNF-a (0 and #{149})and GM-CSF
(V and Y).
198 from donor C. H. R., 227 from donor L. R., and 188 from
donor B. C.) was obtained and screened. No specific prolifera-
tive or cytotoxic activity was observed, while a specific re-
sponse was obtained from donors’ PBLs (all being HLA-A2�)
in response to the influenza peptide GILGFVFTL (data not
shown).
Specific Cytokine Production by the C3/5 Clone in
Response to Autologous Peptide-pulsed LCLs. To assess
cytokine production, cells from the C3/5 clone were stimulated
with autologous DE.LCL in the presence or absence of the
BCR1/25 peptide (20 p.M), and cytokine production was mea-
sured after 1 and 2 days of culture. !L-2, IL-3, !L-4, IL-6, IFN-�,
TNF-a, and GM-CSF were assayed. IL-2, IL-3, IL-4, IL-6, and
IFN--y levels were undetectable or below the detection limit of
the assay used (data not shown). C3/5 cells cultured with autol-
ogous DE.LCL-pulsed peptide for I day secreted significant
amounts ofTNF-a (933, 1 189 pg/ml) and GM-CSF (3395, 3301
pg/ml) in two separate experiments compared with those stim-
ulated with DE.LCL in the absence of peptide (26 pg/mi or
below the detection limit for TNF-a and 184, 1 26 pg/ml for
GM-CSF). TNF-a and GM-CSF production peaked at day 1 and
remained stable at day 2 (Fig. 3). To test whether TNF-ct
produced by the C3/5 clone could be involved in its cytolytic
activity, TNF-a and anti-TNF-a MAb were used. The addition
of anti-TNF-a at different concentrations 0. 1, 1 , and 10 p.g/ml
to the target cells DE.LCL in the presence or absence of the
peptide, did not inhibit the cytolytic activity exerted by the C3/5
clone (Fig. 4). The antibody was instead active in inhibiting the
cytotoxic activity of TNF-cs against the TNF-a-sensitive WEHI
164 cell line (data not shown). TNF-a, at 1-1000 units/ml, had
no cytolytic activity on DE.LCL cells when added for the same
on April 8, 2021. © 1996 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
60
U) �.p1
Ui
�40
0
�200’)
90
80
.� 78LII
�6#{216}’
0 50..-4
.�40.0�3g.
0’)
./ �
0������
./
10
03:1 6:1 12�1 24:1
Ef’f’ector Target ratio
Fig. 4 Effect of MAb directed against TNF-a produced by the C3/5clone on the cytotoxic activity. Cells from the C3/5 clone were chal-
lenged with autologous DE.LCL in the absence of peptide (0) and in the
presence of 20 �iM BCRL/25 peptide (#{149}).Anti-TNF-a at 0. 1 �.g/ml (V),1 �ig/m1 (Y), and 10 p.g/ml (�I1) was mixed with LCL-pulsed peptide
(BCR1/25) before the effector cells were added.
duration of the cytotoxic assay (data not shown). The results
indicate that cytotoxic activity and TNF-a release probably
represent two separate functions in clone C3/5, although both
are specifically induced by BCR1/25.
Reactivity of C3/5 toward HLA-DR11 Allogeneic APLs,
LCLs, and PBLs. Five aliogeneic LCLs [SR. FR, MM, PG
(sharing DR1 1 with DE and obtained from APL patients) and
CHR (DR1 1 �)} were subjected to a cytotoxic test in the pres-
ence or absence of BCR1/25 (20 �.LM) at different E:T ratios. The
results are shown in Fig. 5. The C3/5 clone specifically lysed
DR1 1-matched SR. FR, and MM [and PG (data not shown)j
LCLs pulsed with BCR1/25, as well as the autologous DE.LCL.
CHR cells failed to present the peptide to the C3/5 clone. The
ability of the C3/5 clone to proliferate to the autologous
DE.LCL and HLA-DR1 1-matched SR. FR, MM, and PG.LCL
pulsed with different peptide concentrations was also investi-
gated. A preliminary experiment was carried out to determine
the optimal stimulation period giving the highest proliferative
activity of clone C3/5 at this stage of culture. Cells were
cultured for 24, 48, and 72 h with autologous irradiated (8000
rad) DE.LCL in the presence (2 and 20 p.M) or absence of the
BCR!/25 peptide, incubated for an additional 18 h with
[3HJdThd, and harvested. As shown in Fig. 6A, the 24 h- time
point gave the highest cpm count (P < 0.05 using the t test
analysis). The proliferative activity of the C3/5 clone against
autologous and allogeneic LCL-pulsed peptide (0, 5, 20, and 50
ELM) was subsequently studied at the 24-h time point (Fig. 6B).
All DR1 1 � peptide-pulsed LCLs stimulated the C3/5 clone in a
dose-dependent fashion (PG.PBL induced an uptake of 38,500
i.e
0�
3:1 6:1 12:1 24:1
E?f’ector: Target ratio
Fig. 5 Cytotoxic activity of clone C3/5 against autologous DE.LCL in
the absence (0) or presence (5) of peptide, and allogeneic SR.LCL (V
and V), FR.LCL (LI and U), MM.LCL (A and A), and CHR.LCL ( c�and #{149}) cells in the presence (#{149},U, A, and #{149}) or absence (0, LI. A, and
� ) of 20 pM BCR1/25.
± 2,300 cpm). Although some discrepancies between prolifer-
ation and cytotoxicity were observed (SR and MM.LCL induced
a good proliferative response but produced low levels of cyto-
toxicity, while FR.LCL induced little proliferation but was very
efficiently lysed by C3/5), all four APL patients’ LCLs were
able to present peptide BCR1/25 to C3/5. The ability of LCLs
derived from APL patients to present peptide BCR1/25 to C3/5
suggests that the failure in raising anti-pm1IRAR-ct clones from
the four APL patients was probably not due to a defect in
antigen presentation capability. To further investigate this point,
this experiment was replicated using patients’ PBLs (instead of
LCLs) as APCs. The results are presented in Table I , and
confirm that the antigen presentation capability in the APL
patients studied was grossly intact.
Since the C3/5 clone was specifically able to lyse FR.LCL
cells pulsed with the BCRI/25 peptide in a HLA class II (DR1 I)
restriction fashion, we further tested whether this clone could
exert a cytotoxic activity against FR.APL cells. Because APL
cells are usually DR , FR.blasts (�85% leukemic cells) were
also cultured with IFN--y (a cytokine known to induce or en-
hance the expression of HLA class II molecules) before being
used. The results are present in Fig. 7. C3/5 cells did not lyse
FR.APL blasts and the culture of FR.APL cells with IFN-�y (1 0�
units/mi) for 48 h before the experiment did not enhance the
cytolytic activity. This was possibly due to the failure of IFN-’-y
to induce HLA class II molecules on FR.APL blasts as assayed
by immunofluonescence (data not shown). FR.APL cells were
also unable to present peptide BCR1/25 even after IFN--y pre-
incubation. To test whether the interaction of C3/5 cells with
peptide-pulsed FR.LCL could kill the APL blasts by a bystander
Clinical Cancer Research 597
on April 8, 2021. © 1996 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
Time in culture (hours)
80
70
U).� 60
p-I �#{216}0
�4#{216}.
>5,
10
0
22500
20000
17500
E 150000.0 12500
� 10000wE
5000
2500
0
45000
40000
36800
E0.I) 25000
� 2000001
E isooo
10000
6000
0
Stimulators (LOL)
Fig. 6 Proliferative response of the C3/5 clone against the BCR1/25
peptide. In A, the C3/5 clone (5 X iO� cells/well) was cultured for 24,
48, and 72 h with irradiated (8000 rad) autologous DE.LCL (l0�
cells/well) in the absence (�) or presence of 2 p.M (In) and 20 p.M (U)of peptide BCR1/25 (UI. lymphocytes cultured in medium alone). In B,
the C3/5 clone (5 X l0� cells/well) was cultured for 24 h with irradiated(8000 rad) autologous DE.LCL or ailogeneic SR.LCL, FR.LCL, and
MM.LCL (l0� cells/well) in the presence of 5 p.M (�), 20 p.M (5), and
50 p.M (U) or in the absence (Eli) of BCR1/25 peptide. LCLs at 5 X 106
cells/ml were incubated with the peptide for 30 mm before irradiation,
cultured for an additional 1 8 h with [3H]dThd, and harvested. cpm
values represent the means of three replicates. Bars, SD.
effect, peptide-pulsed FR.LCL cells were admixed with labeled
FR.APL cells at a I : I ratio before incubation with the C3/5
effector clone. Even under these conditions, C3/5 cells were
unable to lyse FR.APL cells (Fig. 7). FR.APL cells proved
otherwise sensitive to lysis as evidenced by the ability of LAK
cells to nonspecifically lyse these cells (Fig. 7). Identical results
were obtained using leukemic cells from patient P. G. and
increasing the incubation time to 6 h (data not shown).
DE SR FR MM
598 Immune Recognition of the pmIIRAR-a Protein
Table 1 Antigen pr esentation capability of A PL patients’ PBL�
Type of APCs -L- BCR 1/25 - BCR 1/25
DE.LCL J�(05 ± 162 3,161 ± 320
DE.PBL 18,083 ± 2,607 6,067 ± 360SR.PBL 16,556 ± 3,995 5,638 ± 941
FR.PBL 18,650 ± I 13 4,645 ± 847
MM.PBL 15,178 ± 3,659 4,210 ± 266PG.PBL 18,025 ± 1,856 3,541 ± 188
“ C3/5 cells (50,000/well) were cocultured with 10� irradiated
APCs incubated with BCR 1/25 (25 p.M) or PBS. Plates were culturedfor 48 h, labeled with [3H]dThd for additional 18 h, and harvested.Values represent the mean X SD cpm obtained in three replicates. C/S
cells incubated in medium alone incorporated 195 ± 70 cpm; irradiated
APCs incorporated 419-841 cpm.
. 1r�;:�:=r-�w:�_ -V
3:1 6:1 12�1 24:1.
EP?ector Target ratio
Fig. 7 Cytotoxic activity of clone C3/5 against allogeneic FR.APL
blasts before and after incubation with IFN-�y (10� units/ml, 48 h) in the
presence (#{149},#{149},Y, and A) or absence ( 0 , 0, E, V. and A) of the
BCRI/25 peptide. The target cells used were: FR.LCL (0 and #{149}),FR.APL (V and Y), FR.APL + FR.LCL (E and U), FR.APL + IFN-’y(A and A), and LAK + FR.APL ( <)).
PHA and TI Responses of Patients’ PBLs. Because of
the negative results obtained with lymphocytes from APL pa-
tients, a PHA response assay was performed to test the T-cell
response in these patients. Results are presented in Table 2 and
show that, compared to PBLs from two normal donors (D. E.
and C. H. R.), patients’ PBLs showed a dramatically reduced
response. These results were surprising to us, since all patients
were studied while in remission and off treatment for >6 months
(F. R. was in complete remission and off therapy for >2 years).
Comparable results were obtained when response to a specific
antigen (IT) was evaluated. Six-day proliferative responses to
IT averaged 293,959 ± 42,793 (SD) cpm in three donors
(including D. E. and C. H. R.) and 49,602 ± 15,047 cpm in the
patient group (P = 0.002).
on April 8, 2021. © 1996 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
Clinical Cancer Research 599
4 C. Gambacorti-Passerini, unpublished results.
Table 2 PHA response in four APL patients and two healthy donors
Type of cells With PHA Without PHA
DE.PBL 20,528 ± 2613 107 ± 87
SR.PBL 3,281 ± 48 228 ± 97
FR.PBL 8,381 ± 476 171 ± 56MM.PBL 1,930 ± 100 214 ± 59
PG.PBL 9,150 ± 1,987 326 ± 83
C1-LR.PBL 26,547 ± 2,438 335 ± 101a Cells were thawed, washed, and seeded at l0� cells/well (see
‘ ‘Patients and Methods’ ‘ ). Values represent mean cpm ± SD from sixreplicates.
Screening of the pmIIRAR-a Fusion Region for HLA
Class I-binding Motifs. Although donor lymphocytes can
recognize pml/RAR-a peptides in a DR-restricted fashion, at-
tempts at generating class I-restricted CTLs in two donors (and
in two patients) failed. We recently obtained the binding motifs
for the 35 most frequent class I alleles and matched these motifs
against the sequence of the pmiIRAR-a fusion region. No good
matching was identified,4 suggesting the lack of binding motifs
for the class I HLA molecules most frequently encountered in
the general population. To provide partial confirmation to these
negative findings, BCR1/9 peptides were separately used in
HLA stabilization experiments. Different cell lines (see “Pa-
tients and Methods’ ‘) were used, enabling us to assess binding
to HLA-A2,3, B5101,63, and Cw0701,0702. Appropriate posi-
tive control peptides (natural ligands) were included and in-
duced an increase in the mean channel of fluorescence of 150%
(A0201 ), 500% (A0301 ), and 269% (Cw0702). All BCR 1/9
peptides induced increments of <20% (on no increase at all)
oven values obtained with control unrelated peptides.
DISCUSSION
The data presented here demonstrate specific proliferation,
cytotoxicity, and cytokine production by HLA-DR1 1 � donor
CD4 1-cell clones upon challenge with HLA-DR1 I LCLs or
PBLs pulsed with the BCR1/25 peptide; no BCR1/9 peptide-
specific CD8IHLA class I-restricted response could be gener-
ated in both donors and patients. We also showed that these
clones could not be obtained in four HLA-DR1 I � APL patients,
and that APL blasts from DRI 1 patients were not recognized by
a DR1 1-restricted CD4 1-cell clone and were unable to present
peptide BCR1/25. These results raise several questions regard-
ing the possible recognition of the APL-specific pml/RAR-a
molecule by I cells. On one side, our results confirm previous
findings that donor I cells can recognize a peptide encompass-
ing the fusion region of the pmIIRAR-a molecule (6). Previous
data indicate that the cellular processing of the pml/RAR-a
protein can produce peptides that stimulate anti-BCR1/25 clones
(6). Here, we show that BCR1/25-specific 1-cell clones can also
exert a specific cytotoxic activity and release cytokines upon
stimulation with the peptide. We also present evidence that
cytotoxic activity and TNF-a release by these clones are sepa-
rate functions, although both seem to be BCR1/25 specific.
Finally, LCLs and PBLs derived from APL patients expressing
the relevant restriction element (HLA-DR I 1 ) are able to present
BCR1/25 similar to autologous LCLs.
On the other side, two different sets of data have to be
considered. First, repeated attempts at generating anti-BCR1/25
1-cell clones in four DRI 1 � APL patients were not successful.
These results cannot be attributed to an insufficient antigen-
presenting capability in these patients, since their LCLs and
PBLs induced proliferation and cytotoxicity in the C3/5 clone in
the presence of BCR1/25 similar to autologous LCLs. However,
it is possible that the frequency and/or the responsiveness of
anti-BCRI/25 lymphocytes was particularly low in these pa-
tients, even if PBLs were recovered when the patients were in
remission with normal peripheral blood counts and 1-cell sub-
sets values. The results of the PHA and IT response assays
strongly support this conclusion and indicate that a generalized
impairment of the cellular immune system, already reported in
cancer patients (15), can be present even during long-term
remissions. The molecular analysis of the observed hyporespon-
siveness is presently under investigation in our laboratory.
The possible deletion or suppression of anti-pmiIRAR-a-
specific T-cell clones by APL cells is another possibility that
deserves further investigation as well as ways to increase the in
vivo frequency of anti-pmlIRAR-a precursors in APL patients.
The failure to generate anti-BCRI/25-specific clones in APL
patients represents the most relevant problem to be solved
before therapeutic approaches can be considered.
Second, in both APL patients and healthy donors we were
unable to raise CD8, HLA class-I restricted 1-cell clones. This
is of particular importance given the usual and strong suppres-
sion of the DR gene expression in APL cells. Although our
strategy of pulsing APCs with short peptides could not have
been successful for technical reasons, it produced positive ne-
sults in other systems (16) and against known immunogenic
peptides. It has been demonstrated that inhibitory peptides com-
petitively block the binding to HLA molecules of stimulatory
peptides (1 7). Since a mixture of 9-men peptides was used in our
experiments, peptide cross-inhibition cannot be excluded. The
use of different stimulation protocols (e.g., using pml/RAR-a-
transfected cells) or APCs (like dendritic cells) could produce
positive results. These results can also be explained by the
observed lack of effective binding motifs for HLA class I
molecules in the pm1IRAR-a junction. A recent article by
Bocchia et al. (18), analyzing HLA binding to two pm1IRAR-a
peptides, is consistent with such an hypothesis.
Thus, it is possible that in APL, the pmlIRAR-cz protein is
recognized by T lymphocytes, but only when presented by a
restriction element (HLA-DR1 1) that is not present on the APL
cells themselves. The recognition of putative DR� APL stem
cells in the form of colony-forming units could still be possible.
However, the existence of these cells has yet to be demonstrated
(19), and true APL colonies are difficult to obtain (20). Alter-
natively, the induction of DR molecules on APL cells could
permit an efficient immune recognition (20).
The lack of expression of DR molecules by APL cells and
the resulting inability of DRI 1 � APL cells to present pml/
RAR-a peptides could also be reflected by the lack of clinical
significance of DR I 1 expression in APL patients undergoing
bone marrow transplantation. In collaboration with the Intema-
on April 8, 2021. © 1996 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
600 Immune Recognition of the pml/RAR-a Protein
6. Gambacorti-Passerini, C., Grignani, F., Arienti, F, Pandolfi, P. P.,Pelicci, P. G., and Parmiani, G. Human CD4 lymphocytes specifically
tional Bone Marrow Transplant Registry, an analysis was con-
ducted in I 89 APL patients in first remission who underwent an
ailogeneic related bone marrow transplant. The patients were
grouped into HLA-DR1 1 � (n 44) and HLA-DR1 1 (n =
145). No significant difference in survival or disease-free sun-
vival was evident between the two groups. Also, the proportion
of DRI 1 patients (23%) does not differ significantly from that
seen in the general population.4 In case the pmlIRAR-a junction
could not be presented by the HLA class I molecule, then the
only chance that this fusion protein can be recognized in APL
cells by I lymphocytes resides in the restoration of DR expres-
sion through pharmacological (2 1 ) or genetic manipulation, or
in the targeting of the still hypothetical DR� APL stem cells.
Our studies demonstrate that the BCR1/25 peptide corre-
sponding to the fusion region of the pm1IRAR-a protein present
in APL cells can be specifically recognized by donor (but not by
patient) CD4 1 cells in a HLA class II (DR1 1)-restricted fash-
ion; CD8IHLA class I-restricted I cells were not generated,
possibly due to the lack of HLA-binding motifs in the pml/
RAR-cx fusion region. Finally. APL blasts were not lysed by
anti-pml/RAR-a cytotoxic T cell clones due to the lack of
expression of the DR1 I molecule by the APL blasts studied.
Further immunotherapeutic approaches to APL will need to
overcome the two main problems encountered in this study: the
lack of DR expression in APL cells (20) and the poor immune
status of APL patients.
ACKNOWLEDGMENTS
We thank Dr. H. G. Rammensee for providing the HLA-binding
motifs and for critical reading of the manuscript, and Dr. M. M.Horowitz, Dr. P. A. Rowlings, and Dr. M. M. Bortin (International BoneMarrow Transplant Registry, Milwaukee, WI) for providing survival
data analysis on transplanted APL patients. We also thank the Antony
Nolan Foundation (London, England) which kindly provided the CHRcells, Dr. M. T. Illeni for HLA class I typing, Dr. F. Poli (Ospedale
Maggiore, Milan, Italy) for DR typing, and T. Blesken for technicalassistance in the cytokine detection assays.
REFERENCES
1. Rowley, J. D., Golomb, H. M., and Dougherty, C. 15/17 transloca-tion, a consistent chromosomal change in acute promyeiocytic leukemia.
Lancet, 1: 549-550, 1977.
2. dethe, H.. Chomienne, C., Lanotte, M.. Degos, L., and Dejean, A.The t(15;l7) translocation of acute promyelocytic ieukaemia fuses theretinoic acid receptor a gene to a novel transcribed locus. Nature
(Lond.), 347: 558-561, 1990.
3. Borrow. J., Goddard, A. D., Sheer, D., and Solomon, E. Molecular
analysis of acute promyeiocytic leukemia breakpoint cluster region onchromosome 17. Science (Washington DC), 249: 1577-1580, 1990.
4. Grignani. F., Fagioli, M., Alcalay, M., Longo, L., Pandolfi. P. P.,
Donti, E., Biondi, A., Lo Coco, F., Grignani, F., and Pelicci, P. G. Acutepromyelocytic leukemia: from genetics to treatment. Blood, 83: 10-25,
1994.
5. Gambacorti-Passenini, C. Immunogenicity of fusion proteins: an cx-ample of tumor-specific/transformation-related antigens. Int. J. Clin.Lab. Res., 23: 186-191, 1993.
recognize a peptide representing the fusion region of the hybrid proteinpml/RARa present in acute promyelocytic ieukaemia cells. Blood, 81:
1369-1375, 1993.
7. Chen, W., Peace, D. J., Rovira, D. K., You, S-G., and Cheever, M. A.
T-cell immunity to the joining region of p210BCRABL protein. Proc.Nati. Acad. Sci. USA, 89: 1468-1472, 1992.
8. Gillis, S., Ferm, M. M., Ou, W., and Smith, K. A. 1-cell growth
factor: parameters of production and quantitative microassay for activ-ity. J. Immunol., 120: 2027-2032, 1978.
9. Carballido, J. M., Carballido-Perrig, N., Terres, 0., Heusser, C. H.,and Blaser, K. Bee venom phospholipase A2 specific T cell clones fromhuman allergic and non allergic individuals: cytokine patterns change inresponse to the antigen concentration. Eur. J. Immunol., 22: 1357-1363,1992.
10. Andersson, U., Andersson, J., Lindfors, A., Wagner, K., Moller, G.,and Heusser, C. H. Simultaneous production of interleukin 2, interleukin4 and intenferon--y by activated human blood lymphocytes. Eur. J.
Immunol., 20: 1591-1596, 1990.
11. Gambacorti-Passerini, C., Rivoltini, L., Supino, R.. Mariani, M.,and Parmiani, 0. Differential lysis of melanoma clones by autologousrecombinant interleukin 2-activated lymphocytes. Relationship withspontaneous resistance to doxorubicin. mt. J. Cancer, 42: 544-548,
1988.
12. Salter, R. D., and Cresswell, P. Impaired assembly and transport of
HLA-A and B antigens mutants TxB cell hybrid. EMBO J., 5, 943-949,1986.
13. Steinle, A., and Schendel, D. J. HLA class I alleles of LCL 721 and174xCEM. T2 (T2). Tissue Antigens, 43: 268-270, 1994.
14. Dc la Salle, H., Hanau, D., Fricker, D., Urlacher, A., Kelly, A.,Salamero, J., Powis, S. H., Donato, L., Bausinger, H., Laforet, M., Jeras,M., Spehner, D., Bieber, T., Faikenrodt, A., Cazenave, J. P., Trowsdale,T., and Tongio, M. M. Homozygous human TAP peptide transportermutation in HLA class I deficiency. Science (Washington DC), 265:
237-241, 1994.
15. Mizoguchi, H., O’Sheea, J., Longo, D., Loeffler, C., McVicar, D.,and Ochoa, A. Alterations in signal transduction molecules in 1 lym-phocytes from tumor-bearing mice. Science (Washington DC), 258:
1795-1798, 1992.
16. Peace, D. J., Smith, J. W., Chen, W., You, 5-0., Cosand, W. L.,Blake, J., and Cheever, M. A. Lysis of ras oncogene-transformed cells
by specific cytotoxic I lymphocytes elicited by primary in vitro immu-nization with mutated Ras peptide. J. Exp. Med., 179: 473-479, 1994.
17. Gedde-Dahl, 1., III, Fossum, B., Eriksen, J. A., Thorsby, E., and
Gaudemack, G. T cell clones specific for p21 ras-derived peptides:characterization of their fine specificity and HLA restriction. Eur. J.Immunol., 23: 754-760, 1993.
18. Bocchia, M., Wenworth, P. A., Southwood, S., Sidney, J., McGraw,K., Scheinberg, D., and Sette, A. Specific binding ofleukemia oncogenefusion protein peptides to HLA class I molecules. Blood, 85: 2680-2685, 1995.
19. Turhan, A. G., Lemoine, F. M., Debert, C., Bonnet, M. L., Baillou,C., Picard, F., Macintyre, E. A., and Varet, B. Highly purified primitivehematopoietic stem cells are PML-RARA negative and generate non-
clonal progenitors in acute promyelocytic leukemia. Blood, 85: 2 154-
2161, 1995.
20. Jinnai, I. In vitro growth response to G-CSF and GM-CSF by bonemarrow cells of patients with acute myeloid leukemia. Leuk. Res., 14:
227-240, 1990.
21. Khanna-Gupta, A., Kolibata, K., Zibeilo, T., and Berliner, N. NB4
cells show bilineage potential and an aberrant pattern of neutrophilsecondary granule protein gene expression. Blood, 84: 294-302, 1994.
on April 8, 2021. © 1996 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
1996;2:593-600. Clin Cancer Res S Dermime, C Bertazzoli, E Marchesi, et al. promyelocytic leukemia patients.pml/RAR-alpha hybrid protein by lymphocytes of acute Lack of T-cell-mediated recognition of the fusion region of the
Updated version
http://clincancerres.aacrjournals.org/content/2/3/593
Access the most recent version of this article at:
E-mail alerts related to this article or journal.Sign up to receive free email-alerts
Subscriptions
Reprints and
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Permissions
Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)
.http://clincancerres.aacrjournals.org/content/2/3/593To request permission to re-use all or part of this article, use this link
on April 8, 2021. © 1996 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from