J. Exp. Med.
Volume 189, Number 2, January 18, 1999 301–308
The R ockefeller University Press • 0022-1007/99/01/301/08 $2.00
A Human Minor Histocompatibility Antigen Specific for
B Cell Acute Lymphoblastic Leukemia
By Harry Dolstra,
Francis Brasseur, Ewald Mensink,
Gosse J. Adema,
Theo M. de Witte,
Carl G. Figdor,
and Elly van de Wiel-van Kemenade
Pierre G. Coulie,
Hospital Nijm egen, 6500 HB Nijm egen, The Netherlands; and the
Ludwig Institute for Cancer Research, Université Catholique de Louvain, B-1200 Brussels , Belgium
Departm ent of Hem atology and the
Departm ent of Tum or Im m unology, University
Cellular Genetics Unit and the
Human minor histocompatibility antigens (mHags) play an important role in the induction of
cytotoxic T lymphocyte (CTL) reactivity against leukemia after human histocompatibility
leukocyte antigen (HLA)-identical allogeneic bone marrow transplantation (BMT). As most
mHags are not leukemia specific but are also expressed by normal tissues, antileukemia reactiv-
ity is often associated with life-threatening graft-versus-host disease (GVHD). Here, we de-
scribe a novel mHag, HB-1, that elicits donor-derived CTL reactivity in a B cell acute lym-
phoblastic leukemia (B-ALL) patient treated by HLA-matched BMT. We identified the gene
encoding the antigenic peptide recognized by HB-1–specific CTLs. Interestingly, expression of
HB-1 gene was only observed in B-ALL cells and Epstein-Barr virus–transformed B cells.
HB-1 gene–encoded peptide EEKR GSLHVW is recognized by the CTL in association
with HLA-B44. Further analysis reveals that a polymorphism in the
gle amino acid exchange from His to Tyr at position 8 within this peptide. This amino acid
substitution is critical for recognition by HB-1–specific CTLs. The restricted expression of the
polymorphic HB-1 Ag by B-ALL cells and the ability to generate HB-1–specific CTLs in vitro
using peptide-loaded dendritic cells offer novel opportunities to specifically target the immune
system against B-ALL without the risk of evoking GVHD.
HB-1 gene generates a sin-
histocompatibility antigens • B cell acute lymphoblastic leukemia • tumor immunity
bone marrow transplantation • cytotoxic T lymphocytes • minor
mia. The therapeutic effect can be partly attributed to the
elimination of residual host leukemic cells by donor-
derived T cells, termed the graft-versus-leukemia reactivity
(1). However, this donor-derived T cell reactivity generally
causes GVHD, a life-threatening complication in alloge-
neic bone marrow transplantation (BMT)
of donor-derived CTLs to recognize host minor histocom-
ransplantation with HLA-identical sibling bone mar-
row is successfully used to treat patients with leuke-
(1). The ability
patibility Ags (mHags) as foreign peptides plays an impor-
tant role in both GVHD and graft-versus-leukemia reactiv-
ity (2–4). mHags are HLA-associated peptides generated
from polymorphic regions of proteins present in the target
cells (5–7). Interestingly, as well as mHags with ubiquitous
tissue distribution, mHags with expression restricted to he-
matopoietic cells or leukemic cells have been characterized
Although allogeneic BMT may cure many patients with
leukemia, relapse of the tumor occurs in a significant num-
ber of patients, indicating that in these patients not all tu-
mor cells are sufficiently eliminated or suppressed. There-
fore, there is a continuous search for additional and more
specific immunotherapeutic strategies in the treatment of
leukemia in order to eliminate leukemic cells without in-
kemia; BMT, bone marrow transplantation; C
lymphoblastoid cell line; mHag, minor histocompatibility antigen;
Mo-MuLV, Moloney’s murine leukemia virus; OR F, open reading frame.
Abbre viations use d in this pape r: B-ALL, B cell acute lymphoblastic leu-
, threshold cycle; LCL,
HB-1: A Polymorphic Leukemia-associated Ag
ducing GVHD. We have focused our attention on CTLs
that selectively recognize leukemic but not normal cells.
R ecently, we isolated from a BMT recipient an HLA-B44–
restricted CTL clone recognizing a novel mHag named
HB-1, which shows specificity for B cell acute lymphoblas-
tic leukemia (B-ALL). Using this CTL, we have identified
both the nucleotide sequence encoding the HB-1 Ag and
the polymorphic CTL epitope. Furthermore, we show that
the expression of the
HB-1 gene is restricted to B-ALL and
EBV-transformed B cells.
Materials and Methods
BR L) supplemented with 10% FCS. CTL clone MP1 was isolated
from PBLs of leukemia patient MP after HLA-identical BMT and
grown in IMDM supplemented with 10% pooled human serum,
irradiated EBV transformed-lymphoblastoid cell line (EBV-LCL)
of the patient pre-BMT (10
/ml), irradiated allogeneic PBMCs
/ml), 100 U/ml IL-2 (Glaxo), and 0.4
c DNA Library Construc tion.The cDNA library was constructed
using the Superscript Plasmid System (GIBCO BR L). Total
R NA was isolated from EBV-LCL MP and poly(A)
was prepared by oligo-dT binding (Qiagen). mR NA was con-
verted into cDNA using an oligo-dT primer that contains a NotI
site at its 5
end. The cDNA was ligated to SalI adaptors, digested
with NotI, and ligated into SalI and NotI sites of the expression
with the recombinant plasmids and clones were selected with
ampicillin. This library was divided into 1,500 pools of
cDNA clones. Each pool was amplified for 4 h and plasmid DNA
was extracted with Bio-R ad miniprep kits (Bio-R ad).
Isolation of the HLA-B*4403 Gene.
LCL MP was extracted using the Trizol method (GIBCO BR L).
R everse transcription was performed on 2
ing an oligo-dT primer and Moloney’s murine leukemia virus
(Mo-MuLV) reverse transcriptase (GIBCO BR L). HLA cDNA
was amplified by PCR using 50 pmol HLA-5UTR primer
3UTR primer (5
dNTPs, and 2.5 U Taq polymerase as previously described (12).
After separation of the PCR product on a 1% agarose gel, the
1.2-kb DNA fragment was isolated and cloned into pCR 3 vector
by using the TA cloning kit (Invitrogen). This cloned
gene was sequenced by the dideoxynucleotide chain ter-
Transfe c tion of COS-1 Ce lls and CTL Stim ulation Assay.
fections were performed by the lipofection method. In brief, 1.5
10 COS-1 cells were plated in a flat-bottomed 96-well plate and
incubated for 18 h at 37
C. 200 ng of plasmid pCR 3 containing
HLA-B*4403 gene and
400 ng of plasmid pSV-Sport1 con-
taining a pool of the cDNA library were mixed with 1
fectamine (GIBCO BR L) in 100
tures were incubated for 30 min and 50
wells to the COS-1 cells. After 4 h of incubation, 50
IMDM/20% FCS was added. Transfectants were tested for their
ability to stimulate the production of IFN-
brief, 3,000 CTLs were added to the wells containing transfected
cells or 3
10 stimulator cells in 200
25 U/ml IL-2. After 18 h of incubation at 37
natant was collected and the IFN-
by ELISA (Endogen).
All cell lines were cultured in IMDM (GIBCO
DH10B were electroporated
Total R NA from EBV-
g of total R NA us-
), 50 pmol HLA-
), 0.5 mM
l of lipo-
l IMDM. Transfection mix-
l was added in duplicate
by the CTLs. In
l IMDM/10% FCS and
concentration was determined
l of super-
Produc tion of Trunc ate d and Mutate d c DNA Construc ts.
and mutation constructs were produced by PCR using specific
primers, and amplified products were subsequently cloned into
vector pCR 3 using the unidirectional TA cloning method (Invit-
HB-1 Polym orphism .Total RNA from cells was extracted using
the Trizol method (GIBCO BR L). R everse transcription was
performed on 2
g of total R NA using an oligo-dT primer and
Mo-MuLV reverse transcriptase (GIBCO BR L). HB-1 cDNA
was amplified by PCR using 50 pmol HB1-F5 primer (5
2.5 U Taq polymerase (GIBCO BR L). The PCR was performed
for 33 cycles (1 min at 94
C, 1 min 60
products were digested with NlaIII to discriminate between
Peptides and Cr-release Assay.
a free COOH terminus by Fmoc peptide chemistry using a multi-
ple synthesizer (ABIMED). Peptides (
analytical HPLC) were dissolved in DMSO and stored at
Cr-release assays were performed as previously described (13). In
peptide recognition assays, target cells were preincubated with
various concentrations of peptide for 1 h at 37
l before the addition of effector cells. After 4 h of incuba-
tion at 37
l supernatant was collected and radioactivity
was measured by a gamma counter.
Taqm an™ PCR Assay for HB-1 Expression.
cells was extracted using the Trizol method (GIBCO BR L). R e-
verse transcription was performed on 2
oligo-dT primer and Mo-MuLV reverse transcriptase (GIBCO
BR L). Of the total cDNA volume of 20
each PCR reaction. PCR amplification and real time quantita-
tion analysis were performed using the Taqman™ assay (14, 15).
The following sequences were used as primers and Taqman™
nucleotide [nt] 58–77), 5
(HB1- GSP3, antisense, nt 271–291), 5
sense, nt 188–217), 5
CTCATCT TTGGGCTGTT T TCT TCCGCC-(TAMR A)-3
(Pbgd-probe). The reaction mixture contains 1.25 U AmpliTaq
Gold (PE-Applied Biosystems), 250
and 15 pmol antisense primer in a total reaction volume of 50
The enzyme was activated by heating for 10 min at 95
step PCR procedure of 60 s at 60
for 40 cycles. 6 mM MgCl
and 100 nM probe for the HB-1
PCR , and 5 mM MgCl
and 100 nM probe for the Pbgd PCR ,
was used. The PCR and Taqman™ analysis were performed in the
ABI/PR ISM 7700 Sequence Detector System (PE-Applied Bio-
systems). The system generates a real time amplification plot based
upon the normalized fluorescence signal. Subsequently the thresh-
old cycle (C
), i.e., the fractional cycle number at which the num-
ber of amplified target reaches a fixed threshold, is determined.
is proportional to the initial number of target copies in the
sample (14, 15). We used the expression of the
de am inase
(Pbgd) gene to normalize the HB-1 expression.
used as an endogenous reference to correct for differences in the
amount of total R NA added to the reaction. The Pbgd primers
only allows amplification of cDNA derived from the
keeping gene (16, 17). HB-1 mR NA expression was quantified by
determining calibration functions for HB-1 and Pbgd expression of
a reference cell line. Therefore, 2
), 50 pmol HB1-GSP3 primer
), 0.5 mM dNTPs, and
C, and 1 min 72
Peptides were synthesized with
90% pure as indicated by
C in a volume of
Total R NA from
g of total R NA using an
l was used for
(Pbgd-R), and 5
M dNTPs, 15 pmol sense,
C. A two-
C was applied
C and 15 s at 95
g of R NA of the B-ALL cell
Dolstra et al.
line KM3 was reverse transcribed into cDNA and serially diluted
into water. This cDNA serial dilution was prepared once, stored at
C, and used in all tests performed in this study. The linear cal-
ibration functions between the C
starting quantity (N) were C
21.9 for HB-1 and Pbgd, respectively. HB-1 and
Pbgd mR NA expression in all test samples were quantified using
these calibration functions. At the same level of Pbgd expression
the level of HB-1 expression of test samples was determined as a
percentage of the HB-1 expression in cell line KM3.
and the logarithm of the initial
25.8 and C
Identification of a cDNA Coding for the HB-1 Antigenic Pep-
tide.We generated a cDNA library from an EBV-LCL
that expresses the HB-1 Ag since it was efficiently lysed by
CTL MP1. Using expression cloning, we isolated from this
library a cDNA that upon transfection along with the
HLA-B44 cDNA into COS-1 cells stimulated CTL MP1
to release IFN-? (Fig. 1). The IFN-? release was as high as
that induced by EBV-LCL MP, from which the cDNA li-
brary was generated, and even higher than that induced by
EBV-LCL as well as B-ALL cells of the HLA-B44, HB-
1–positive patient VR . Untransformed CD40-stimulated B
cells of patient VR could not stimulate CTL MP1 to re-
lease IFN-?. Transfection of either of the aforementioned
cDNAs alone failed to induce the production of IFN-? by
CTL MP1, indicating that the isolated cDNA clone en-
codes the HB-1 Ag.
The cDNA encoding the HB-1 Ag consists of 397 nu-
cleotides with no significant homology to sequences pres-
ently recorded in data banks. To localize the region encod-
ing the HB-1 epitope, we transfected COS-1 cells with
truncated HB-1 cDNA constructs in combination with the
HLA-B44 cDNA and tested their capacity of inducing
IFN-? release by the CTLs (Fig. 2). The smallest truncated
construct that still encoded both the translation initiation
codon and the peptide coding region contains the nucle-
otide sequence 100 to 165 (Fig. 2). The three possible
translational reading frames within this sequence did not
code for a peptide according to the described HLA-B44
binding motif, a Glu at position 2 and a Phe or Tyr at posi-
tion 9 or 10 (18, 19). We next synthesized all 9-, 10-, and
11-mer peptides with a Glu residue at position 2 encoded
by the translational reading frames within nucleotide se-
quence 100 to 165, and tested their ability to induce lysis
by CTL MP1 upon loading on HB-1–negative target cells.
The 10-mer EEKR GSLHVW was specifically recognized
by CTL MP1, but not by HLA-B44–restricted, EBNA3C-
specific CTLs (Fig. 3, A and B). CTL MP1 did not recog-
nize the 9-mer peptide EEKR GSLHV (data not shown).
The open reading frame (OR F) encoding the EEKR G-
SLHVW CTL epitope contains a CTG translation initia-
tion codon resulting in a putative 41-amino acid protein
(Fig. 4). Substitution of this CTG into AAG resulted in a
complete loss of the ability to stimulate IFN-? release by
CTL MP1 upon transfection into COS-1 cells (data not
shown). Together, these data led us to conclude that within
the HB-1 sequence the CTG at position 108–110 initiates
the translation of a 41-amino acid protein from which the
EEKR GSLHVW CTL epitope is generated.
The HB-1 Ge ne Is Only Expre sse d by Transform e d B Ce lls.
Interestingly, we observed that CTL MP1 exhibits specific
cytotoxicity towards leukemia- and EBV-transformed B
cells, but not against untransformed B cells, T cells, mono-
cytes, and fibroblasts (8). Therefore, we studied the level of
HB-1 gene expression in a large panel of tumor and non-
malignant cells using real time quantitative reverse tran-
recognized by CTL clone MP1. Production of IFN-? is shown upon
stimulation with the following stimulator cells: B-ALL cells, EBV-trans-
formed B cells, and B cells stimulated with CD40 plus 100 U/ml TNF-?
for 2 d of the HLA-B44, HB-1–positive patients MP and VR , and COS-1
cells cotransfected with the HB-1 cDNA plus HLA-B44 cDNA, and
COS-1 cells transfected either with the HLA-B44 cDNA alone or with
HB-1 cDNA alone. R elease of IFN-? was measured by ELISA.
Identification of the cDNA encoding the antigenic peptide
tion initiation codon and the antigenic peptide recognized by CTL MP1.
HB-1 cDNA deletion constructs were cloned into an expression vector
and cotransfected with the HLA-B44 cDNA into COS-1 cells. Trans-
fected cells and CTL MP1 were incubated for 18 h and the release of
IFN-? was measured by ELISA.
Location of the nucleotide sequence coding for the transla-
HB-1: A Polymorphic Leukemia-associated Ag
scriptase PCR . All B-ALL samples expressed the HB-1
gene at a significant level exceeding 10% of that found in
the reference B-ALL cell line KM3 (Fig. 5 A). In addition,
2 out of 14 B cell lymphomas and 2 out of 5 acute undif-
ferentiated leukemias showed significant HB-1 expression.
In contrast, all T-ALLs, multiple myelomas, acute myeloid
leukemias, and nonhematological solid tumors lacked HB-1
expression. Since the number of HB-1 transcripts in B-ALL
cells is even lower than that of the low copy gene Pbgd we
concluded that HB-1 is a rare mR NA species. This notion
was confirmed by the observation that we could not detect
HB-1 expression by Northern blot analysis, whereas ?-actin
mR NA was readily detected (data not shown).
Analysis of the expression of the HB-1 gene in a panel of
nonmalignant cells revealed significant levels in 90% of the
EBV-transformed B cell lines, whereas no significant HB-1
transcription was observed in all other nonmalignant cell
types (Fig. 5 B). Some B cell and PHA-stimulated T cell
samples express very low levels of HB-1 (?10% of that
found in the reference B-ALL cell line KM3; Fig. 5 B).
PHA-stimulated T cell blasts with an HB-1 expression lev-
els of 8 and 2.5% of that found in the KM3 cell line were
HB-1 antigenic peptide. (A) Cy-
tolytic activity by CTL clone
MP1 against HLA-B44–positive
target cells incubated with 5 ?M
of HB-1 peptide EEKR GSL-
HVW. Controls included the
HLA-B44–positive target cells
incubated either without peptide
or with the EBNA3C 281–290
peptide EENLLDFVR F. (B)
EBNA3C-specific CTL against
HLA-B44–positive target cells
incubated with 5 ?M of
EBNA3C peptide EENLLD-
FVR F. Controls included the
HLA-B44–positive target cells
incubated either without peptide
or with the HB-1 peptide EE-
Identification of the
encoded protein starting from the CTG start codon (underlined) at nucle-
otide positions 108–110. The sequence corresponding to the HLA-B44–
restricted HB-1 peptide is boxed. These sequence data are available from
EMBL/GenBank/DDBJ under accession number AF103884.
Sequence of HB-1 cDNA and of the 41 amino acid–
cells. (A) Measurement of the HB-1 expression obtained by real time quan-
titative reverse transcriptase PCR in tumor cells and cell lines. The follow-
ing cell lines were used: B-ALL (KM3, BV173); B cell lymphoma (Daudi,
R aji, R amos, SU-DHL6); multiple myeloma (R PMI1758); T-ALL (Jurkat,
CEM, HSB-2); and acute myeloid leukemia (Lama, Kasumi, K562, HL60,
KG-1). The expression levels were determined by a calibration function
generated from R NA of the HB-1–positive B-ALL cell line KM3 and ex-
pressed relative to the HB-1 level measured in these cells. The Pbgd gene
was used as standard to correct for R NA quantity and quality. The detec-
tion limit is indicated with a solid line. Samples showed significant HB-1
gene expression if they exceeded 10% of that found in the B-ALL cell line
KM3. This arbitrary threshold is indicated with a dashed line. (B) Measure-
ment of the HB-1 expression obtained by real time quantitative reverse
transcriptase PCR in freshly isolated cells or primary cell cultures.
Expression of the HB-1 gene in tumor cells and nonmalignant
Dolstra et al.
not recognized by CTL MP1 (Fig. 6). Loading of these
PHA-stimulated T cell blasts with the EEKR GSLHVW
peptide results in lysis that is as high as the killing of the
HB-1–positive EBV-LCL. Analysis of HB-1 expression in
normal tissues revealed only low (?10%) transcription
levels in testis samples (Table I). Together, these results
demonstrate that substantial HB-1 mR NA expression is
observed only in B-ALL cells and in EBV-transformed B
cells and that very low mR NA expression of HB-1 does
not result in lysis by CTL MP1.
A Polym orphism in the HB-1 Gene Determ ines CTL Recog-
nition.CTL MP1 lyse EBV-transformed B cells derived
from the patient MP, whereas EBV-transformed B cells
derived from the HLA-identical sibling donor BP are not
recognized (8). Since the HB-1 gene was significantly
expressed by both cell lines, these results suggested the
presence of a polymorphism in this sequence. Analysis of
the HB-1 sequence of the donor revealed that it differs
from that of the patient at only one nucleotide (position
153: C to T), leading to an amino acid change from H to Y
in the HB-1 peptide (Fig. 7 A). The corresponding alleles
were named HB-1H and HB-1Y, respectively.
Studies of the HB-1 gene polymorphism in relatives of
patient MP resulted in a clear correlation between recogni-
tion by CTL MP1 of EBV-LCL of HLA-B44–positive
family members and the expression of the HB-1H allele
(Fig. 7 B). To verify that only expression of the HB-1H
allele leads to recognition by CTL MP1, we transfected
HB-1H or HB-1Y cDNA along with HLA-B44 cDNA into
COS-1 cells. Cells transfected with the HB-1H cDNA
stimulated IFN-? release by the CTL, whereas cells trans-
fected with the HB-1Y cDNA did not (Fig. 8 A). In addi-
tion, EBV-transformed B cells incubated with the peptide
encoded by the HB-1Y allele were not lysed by CTL MP1
(Fig. 8 B). This probably is not the result of defective bind-
ing of the HB-1Y peptide to HLA-B44 molecules, as both
peptides appeared to bind with equal affinity to these HLA
molecules (data not shown). These results demonstrate that
at least two allelic forms of the HB-1 gene exist, and that
only the HB-1H allele encodes the peptide that is recog-
nized by CTL MP1.
Over the past few years, a large number of CTL-defined
tumor Ags and their encoding genes have been identified
in melanoma (20). However, little is known about CTL-
defined Ags encoded by genes with expression restricted to
leukemia. Using biochemical methods to isolate and se-
quence antigenic peptides presented by MHC class I mol-
ecules, the first CTL epitopes on leukemic cells have
recently been identified (6, 7, 21). Expression of the iden-
tified polymorphic Ags HA-1 and HA-2 is restricted to he-
matopoietic cells but they are not leukemia specific (2). Al-
though these mHags have this limited tissue distribution,
mismatching for HA-1 is associated with the occurrence of
severe GVHD (22). To the best of our knowledge, the
polymorphic HB-1 gene product described here is the first
identified leukemia-associated mHag that is only signif-
icantly expressed by leukemia- and EBV-transformed B
cells. In some normal cells, few HB-1 transcripts are pres-
PHA-stimulated T cell blasts of HLA-B44, HB-1–positive individuals.
PHA-stimulated T cell blasts were preincubated either without peptide or
with the HB-1 peptide EEKR GSLHVW. The level of HB-1 gene ex-
pression in EBV-LCL VH and MT was 140 and 122% of that found in
the B-ALL cell line KM3, respectively, and in PHA-stimulated T cell
blasts VH and MT was 8 and 2.5%, respectively. Both individuals express
homozygous the HB-1H allele and the HLA-B44 subtype of VH is
B*4403 and of MT B*4402. The E/T cell ratio was 3:1.
Cytolytic activity by CTL clone MP1 against EBV-LCL and
T able I. Expression of the HB-1 Gene in Norm al Tissues
Type of normal tissuePositive samples/samples tested*
*Expression of the HB-1 gene was determined by real time quantitative
reverse transcriptase PCR of total R NA as described in Materials and
Methods. Samples were scored as positive if their HB-1 gene expression
exceeded 1% of that found in the B-ALL cell line KM3.
HB-1: A Polymorphic Leukemia-associated Ag
ent, but expression is too low for recognition by HB-1–spe-
cific CTLs, as was also observed for a number of other
tumor-associated Ags (20, 23–25). Low levels of transcrip-
tion of the melanocyte-associated Ag gp100 have been
found in nearly all normal tissues and tumor cell lines of
nonmelanocytic origin, but no gp100 protein could be de-
tected by either Western blot or cytotoxicity assays (23).
Similarly, the Ag encoded by the N-acetylglucosam inyl-trans-
ferase V (GnTV) gene was not recognized by specific CTLs
when transcription levels did not exceed 8% of that of the
reference melanoma cell line (24). Finally, MAGE-1–spe-
cific CTLs are unable to recognize tumor cell lines express-
ing low levels (?10%) of the MAGE-1 gene when com-
pared with melanoma cells that were efficiently lysed (25).
These reported data and the observation that PHA-stimu-
lated T cells expressing low levels of the HB-1 gene (2.5
and 8% of that of the reference B-ALL cell line) are not
lysed by CTL MP1 indicate that only significant HB-1
transcription leads to CTL recognition.
The HB-1 antigenic peptide is encoded by a sequence
which starts with a CUG codon resulting in the translation
of a short protein of 41 amino acids. Such an unusual initi-
ation of translation at a CUG has been reported previously
(26). Whether this product is the only protein encoded by
the HB-1 gene or whether there are more proteins en-
coded by alternative OR Fs is currently unknown. Al-
though most eukaryotic mR NAs have a single OR F of
which translation is usually initiated by an AUG codon,
several human genes are bicistronic, encoding two proteins
(27–29). For instance, the gp75 gene in melanoma has two
overlapping OR Fs resulting in two completely different
proteins: (i) gp75 as recognized by IgG antibodies in the se-
rum from melanoma patients, and (ii) a short protein of 24
amino acids from which an antigenic peptide recognized
by CTLs is generated (27). Whether the generation of the
short 41-amino acid protein from the HB-1 gene is the
result of translation of an alternative OR F awaits further
characterization of the gene.
At present we can not exclude the possibility that the
HB-1 gene–encoded protein is a B cell differentiation Ag
that is overexpressed in B-ALL cells and lost in mature B
cells. Alternatively, the malignant transformation of pro-
genitor B cells itself may induce HB-1 expression. This lat-
ter possibility is supported by the finding that the HB-1
gene also shows significant expression in B cells that are
transformed in vitro with EBV. These EBV-transformed B
cell lines express all EBV gene–encoded proteins, whereas
EBV-positive Burkitt’s lymphoma cell lines express only
the EBNA1 protein (30). Since HB-1 is significantly ex-
pressed by EBV-transformed B cell lines and not by Bur-
kitt’s lymphoma cell lines, it is tempting to speculate that
expression of EBV gene–encoded proteins other than EBNA1
may induce HB-1 mR NA transcription in mature B cells.
Studies dealing with expression levels during B cell differ-
entiation and the role of EBV transformation in inducing
HB-1 expression are currently under investigation.
The antigenic peptide EEKR GSLHVW encoded by the
HB-1 gene is recognized in association with HLA-B44, a
common HLA-B allele expressed by 23% of the Caucasian
gene. (A) Sequence of the peptide coding region of the HB-1 gene of pa-
tient MP and donor BP. The nucleotide and amino acid polymorphism
are underlined. (B) Correlation between expression of HB-1 alleles and
cytolytic activity by CTL clone MP1 against EBV-transformed B cell
lines of relatives of patient MP. Filled circles (females) or squares (males)
indicate strong lysis by CTL MP1. Open symbols indicate no lysis. Ex-
pression of HB-1 alleles was determined by reverse transcriptase PCR
amplification and digestion of PCR products with restriction enzyme
The mHag HB-1 is encoded by the HB-1H allele of the HB-1
peptide is critical for recognition by CTL clone MP1. (A) Production of
IFN-? is shown upon stimulation with EBV-transformed B cells of pa-
tient MP and donor BP, and COS-1 cells cotransfected with the HB-1H
or HB-1Y cDNA and HLA-B44 cDNA. (B) Cytolytic activity by CTL
clone MP1 against HLA-B44–positive target cells incubated with the
HB-1H or HB-1Y peptide. The E/T cell ratio was 10:1.
The single amino acid exchange within the HB-1 antigenic
Dolstra et al.
population. Five subtypes of HLA-B44 have been identi-
fied, but the most frequently expressed subtypes are HLA-
B*4402 and -B*4403. These two subtypes differ only by a
single amino acid substitution from Asp (*4402) to Leu
(*4403) in position 156 of the ?2 domain (18, 31, 32).
Both HLA-B*4402 and -B*4403 are able to present the
HB-1 peptide to CTL clone MP1. The consensus peptide
binding motif for HLA-B44 shows a predominance for Glu
at position 2, and Tyr or Phe at position 9 or 10 (18, 19).
The HB-1 peptide contains a Glu at position 2, but in con-
trast to the consensus motif it has a Trp at position 10, indi-
cating that all amino acids with aromatic side chains facili-
tate binding to HLA-B44. A polymorphism in the HB-1
gene resulted in a single amino acid exchange from His to
Tyr at position 8 of the HB-1 peptide. The anchor residues
of the peptide are not involved in this substitution, and
both the HB-1H and HB-1Y peptide bind with similar af-
finity to HLA-B44 molecules. Since the HB-1Y peptide is
not recognized by CTL clone MP1, the polymorphism ap-
pears to influence a TCR contact residue.
The molecular identification and characterization of the
leukemia-associated mHag HB-1 allows the opportunity to
treat leukemia patients with immunotherapy specifically
targeted to the tumor without the risk of inducing GVHD.
For this, patients must be typed for the presence of the
HB-1H allele and their tumor cells must significantly ex-
press the HB-1 gene. The mHag HB-1 might turn out to
be an excellent target against which specific CTLs can eas-
ily be generated from allogeneic donors and adoptively
transferred into BMT recipients with relapsed B-ALL. We
have already succeeded in generating HB-1 specific CTLs
in vitro from peripheral blood of healthy HLA-B44–posi-
tive individuals by stimulating CD8-positive T cells with
peptide-loaded autologous dendritic cells (our unpublished
results). These induced CTLs displayed lysis of both pep-
tide-loaded HB-1–negative target cells, and autologous
HLA-B44–positive EBV-transformed B cells endogenously
expressing the HB-1 Ag. The presence of HB-1–specific
CTLs in the T cell repertoire of HB-1–positive individuals
and the restricted expression of the HB-1 gene by B-ALL
cells may also allow the use of HB-1–encoded Ags for vac-
cination protocols to induce specific immunity in B-ALL
We thank R ick Brouwer for cloning the B*4403 gene, Aukje Zimmerman for help in screening the cDNA
library, and Wendy Unger for generation of the EBNA3C peptide-specific CTL clone. We thank Louis van
de Locht and Ellen Linders for technical assistance with the ABI/PR ISM 7700 Sequence Detector System.
We thank Han Zendman (Department of Pathology, University Hospital Nijmegen, The Netherlands) for
providing R NA samples from primary cell cultures.
This work was supported by grants from the Dutch Cancer Society (KUN 97-1508), the Ank van Vlissin-
gen Foundation, and the Maurits and Anna de Kock Foundation.
Address correspondence to Harry Dolstra, Department of Hematology, University Hospital Nijmegen,
Geert Grooteplein 8, PO Box 9101, 6500 HB Nijmegen, The Netherlands. Phone: 31-24-361-9449; Fax:
31-24-354-2080; E-mail: firstname.lastname@example.org
Received for publication 27 August 1998 and in revised form 5 Novem ber 1998.
1. Horowitz, M.M., R .P. Gale, P.M. Sondel, J.M. Goldman, J.
Kersey, H.J. Kolb, A.A. R imm, O. R ingden, C. R ozman, B.
Speck, et al. 1990. Graft-versus-leukemia reactions after bone
marrow transplantation. Blood. 75:555–562.
2. Goulmy, E. 1996. Human minor histocompatibility antigens.
Curr. Opin. Im m unol. 8:75–81.
3. Perrault, C., D.C. R oy, and C. Fortin. 1998. Immunodomi-
nant minor histocompatibility antigens: the major ones. Im -
m unol. Today. 19:69–74.
4. Pardoll, D. 1997. Taming the sinister side of BMT: Dr. Jekyll
and Mr. Hyde. Nature Med. 3:833–834.
5. Wallny, H.J., and H.G. R ammensee. 1990. Identification of
classical minor histocompatibility antigen as cell-derived pep-
tide. Nature. 343:275–278.
6. Wang, W., L.R . Meadows, J.M.M. den Haan, N.E. Sher-
man, Y. Chen, E. Blokland, J. Shabanowitz, A.I. Agulnik,
R .C. Hendrickson, C.E. Bishop, et al. 1995. Human H-Y: a
male-specific histocompatibility antigen derived from the
SMCY protein. Science. 269:1588–1590.
7. Den Haan, J.M.M., L.M. Meadows, W. Wang, J. Pool, E.
Blokland, T.L. Bishop, C. R einhardus, J. Shabanowitz, R .
Offringa, D.F. Hunt, et al. 1998. The minor histocompatibil-
ity antigen HA-1: a diallelic gene with a single amino acid
polymorphism. Science. 279:1054–1057.
8. Niederwieser, D., A. Grassegger, J. Aubock, M. Herold, D.
Nachbaur, A. R osenmayr, A. Gachter, W. Nussbaumer, S.
Gaggl, M. R itter, and C. Huber. 1993. Correlation of minor
histocompatibility antigen-specific cytotoxic T lymphocytes
with graft-versus-host disease status and analysis of tissue dis-
tribution of their target antigens. Blood. 81:2200–2208.
9. Warren, E.H., P.D. Greenberg, and S.R . R idell. 1998. Cy-
totoxic T-lymphocyte-defined human minor histocompati-
bility antigens with a restricted tissue distribution. Blood. 91:
10. Faber, L.M., S.A.P. van Luxemburg-Heijs, R . Willemze, and
J.H.F. Falkenburg. 1992. Generation of leukemia-reactive
cytotoxic T lymphocyte clones from the HLA-identical bone
marrow donor of a patient with leukemia. J. Exp. Med. 176:
HB-1: A Polymorphic Leukemia-associated Ag
11. Dolstra, H., H. Fredrix, F. Preijers, E. Goulmy, C.G. Figdor,
T.M. de Witte, and E. van de Wiel-van Kemenade. 1997.
R ecognition of a B cell leukemia-associated minor histocom-
patibility antigen by CTL. J. Im m unol. 158:560–565.
12. Ennis, P.D., J. Zemmour, R .D. Salter, and P. Parham. 1990.
R apid cloning of HLA-A,B cDNA by using the polymerase
chain reaction: frequency and nature of errors produced in
amplification. Proc. Natl. Acad. Sci. USA. 87:2833–2837.
13. Van de Wiel-van Kemenade, E., A.A. te Velde, A.J. de Boer,
R .S. Weening, A. Fischer, J. Borst, C.J.M. Melief, and C.G.
Figdor. 1992. Both LFA-1-positive and -deficient T cell
clones require the CD2/LFA-3 interaction for specific cy-
tolytic activation. Eur. J. Im m unol. 22:1467–1475.
14. Higuchi, R ., C. Fockler, G. Dollinger, and R . Watson.
1993. Kinetic PCR analysis: real-time monitoring of DNA
amplification reactions. Biotechnology. 11:1026–1030.
15. Gibson, U.E.M., C.A. Heid, and P.M. Williams. 1996. A
novel method for real time quantitative R T-PCR . Genom e
16. Chretien, S., A. Dubart, D. Beaupain, N. R aich, B. Grand-
champ, J. R osa, M. Goossens, and P.H. R omeo. 1988. Al-
ternative transcription and splicing of the human porphobili-
nogen deaminase gene result either in tissue-specific or in
housekeeping expression. Proc. Natl. Acad. Sci. USA. 85:6–
17. Finke, J., R . Fritzen, P. Ternes, W. Lange, and G. Dolken.
1993. An improved strategy and a useful housekeeping gene
for R NA analysis from formaline-fixed, paraffin embedded
tissues by PCR . Biotechniques. 14:448–453.
18. Fleischhauer, K., D. Avila, F. Vilbois, C. Traversari, C. Bor-
dignon, and H.J. Wallny. 1994. Characterization of natural
peptide ligands for HLA-B*4402 and -B*4403: implications
for peptide involvement in allorecognition of a single amino
acid change in the HLA-B44 heavy chain. Tissue Antigens.
19. DiBrino, M., K.C. Parker, D.H. Margulies, J. Shiloach, R .V.
Turner, W.E. Biddison, and J.E. Coligan. 1995. Identifica-
tion of the peptide binding motif for HLA-B44, one of the
most common HLA-B alleles in the Caucasian population.
Biochem istry. 34:10130–10138.
20. Van den Eynde, B.J., and P. van der Bruggen. 1997. T cell
defined tumor antigens. Curr. Opin. Im m unol. 9:684–693.
21. Den Haan, J.M.M., N.E. Sherman, E. Blokland, E. Huczko,
F. Koning, J.W. Drijfhout, J. Skipper, J. Shabanowitz, D.F.
Hunt, V.H. Engelhard, and E. Goulmy. 1995. Identification
of a graft versus host disease-associated human minor histo-
compatibility antigen. Science. 268:1476–1480.
22. Goulmy, E., R . Schipper, J. Pool, E. Blokland, J.H. Falken-
burg, J. Vossen, A. Grathwohl, G.B. Vogelsang, H.C. van
Houwelingen, and J.J. R ood. 1996. Mismatches of minor
histocompatibility antigens between HLA-identical donors
and recipients and the development of graft-versus-host dis-
ease after bone marrow transplantation. N. Engl. J. Med. 334:
23. Brouwenstijn, N., E.H. Slager, A.B.H. Bakker, M.W.J.
Schreurs, C.W. van der Spek, G.J. Adema, P.I. Schrier, and
C.G. Figdor. 1997. Transcription of the gene encoding mel-
anoma-associated antigen gp100 in tissues and cell lines other
than those of the melanocytic lineage. Br. J. Cancer. 76:1562–
24. Guilloux, Y., S. Lucas, V.G. Brichard, A. van Pel, C. Viret,
E. de Plaen, F. Brasseur, B. Lethe, F. Jotereau, and T. Boon.
1996. A peptide recognized by human cytolytic T lympho-
cytes on HLA-A2 melanomas is encoded by an intron se-
quence of the N-acetylglucosaminyltransferase V gene. J.
Exp. Med. 183:1173–1183.
25. Lethe, B., P. van der Bruggen, F. Brasseur, and T. Boon.
1997. MAGE-1 expression threshold for the lysis of mela-
noma cell lines by a specific CTL. Melanom a Res. 7(Suppl 2):
26. Hann, S.R . 1994. R egulation and function of non-AUG-ini-
tiated proto-oncogenes. Biochim ie. 76:880–886.
27. Descombes, P., and U. Schibler. 1991. A liver-enriched tran-
scriptional activator protein, LAP, and a transcriptional inhib-
itory protein, LIP, are translated from the same mR NA. Cell.
28. Wang, R .F., M.R . Parkhurst, Y. Kawakami, P.F. R obbins,
and S.A. R osenberg. 1996. Utilization of an alternative open
reading frame of a normal gene in generating a novel human
cancer antigen. J. Exp. Med. 183:1131–1140.
29. Shichijo, S., M. Nakao, Y. Imai, H. Takasu, M. Kawamoto,
F. Niiya, D. Yang, Y. Toh, H. Yamana, and K. Itoh. 1998.
A gene encoding antigenic peptides of human squamous cell
carcinoma recognized by cytotoxic T lymphocytes. J. Exp.
30. Klein, G. 1994. Epstein-Barr virus strategy in normal and
neoplastic B cells. Cell. 77:791–793.
31. Tiercy, J.M., N. Djavad, N. R ufer, D.E. Speiser, M. Jeannet,
and E. R oosnek. 1994. Oligotyping of HLA-A2, -A3, and
-B44 subtypes: Detection of subtype incompatibilities be-
tween patients and their serologically matched unrelated
bone marrow donors. Hum . Im m unol. 41:207–215.
32. Fleischhauer, K., N.A. Kernan, B. Dupont, and S.Y. Yang.
1991. The two major subtypes of HLA-B44 differ for a single
amino acid at codon 156. Tissue Antigens. 37:133–137.