Targeting the Wilms Tumor Antigen 1 by TCR Gene
Transfer: TCR Variants Improve Tetramer Binding but Not
the Function of Gene Modified Human T Cells1
Sharyn Thomas,* Shao-An Xue,* Michela Cesco-Gaspere,* Esther San Jose ´,†Daniel P. Hart,*
Vivian Wong,2* Reno Debets,‡Balbino Alarcon,†Emma Morris,* and Hans J. Stauss3*
We have previously described the functional activity of a human TCR specific for an HLA-A2-presented peptide derived from the
Wilms tumor Ag 1 (WT1). Recent studies showed that the expression and function of human TCR was improved by the intro-
duction of an additional disulfide bond between the ?- and ?-chains or by the exchange of the human constant region for murine
sequences. In this study, we analyzed the functional activity of WT1-TCR variants expressed in Jurkat cells and in primary T cells.
The introduction of cysteine residues or murine constant sequences into the WT1-TCR did not result in a global reduction of
mispairing with wild-type TCR chains. Instead, the level of mispairing was affected by the variable region sequences of the
wild-type TCR chains. The analysis of freshly transduced peripheral blood T cells showed that the transfer of modified TCR
constructs generated a higher frequency of Ag-responsive T cells than the transfer of the wild-type TCR. After several rounds of
peptide stimulation this difference was no longer observed, as all transduced T cell populations accumulated ?90% of Ag-
responsive T cells. Although the Ag-responsive T cells expressing the modified TCR bound the HLA-A2/WT1 tetramer more
efficiently than T cells expressing the wild-type TCR, this did not improve the avidity of transduced T cells nor did it result in a
measurable enhancement in IFN-? production and cytotoxic activity. This indicated that the enhanced tetramer binding of
modified WT1-TCR variants was not associated with improved WT1-specific T cell function. The Journal of Immunology, 2007,
of TCR gene therapy was recently demonstrated in the first clinical
trial in melanoma patients (16). The retroviral transfer of a
MART1-specific TCR efficiently generated MART1-specific
CD8?lymphocytes that were used for adoptive T cell therapy.
Infused T cells expanded in vivo and engrafted at high levels in
most melanoma patients. Compared with the impressive clinical
response rate of conventional adoptive T cell therapy with ex-
panded tumor-infiltrating lymphocytes (?50% response rate), the
anti-melanoma activity of the TCR-transduced lymphocytes was
relatively inefficient with only two of 15 patients showing tumor
regression (16, 17). This indicated that the efficiency of TCR gene
therapy should be further improved to achieve better tumor pro-
tection in vivo.
n the past years, several groups have demonstrated that ret-
roviral TCR gene transfer is an attractive strategy to redirect
the Ag specificity of primary T cells (1–15). The feasibility
The inefficient expression of introduced TCR ?- and ?-chains
in T lymphocytes can be one of the rate-limiting steps for TCR
gene therapy. Because TCR surface expression requires asso-
ciation with CD3 ?-, ?-, ?-, and ?-chains, the introduced TCR
competes with the endogenous TCR for a limited number of
CD3 molecules. In addition, the introduced TCR chains may
mispair with endogenous chains, thus further reducing the ex-
pression of relevant TCR ?? heterodimers on the surface of
transduced T cells. Recently, two strategies to reduce TCR mi-
spairing and enhance the association with CD3 molecules have
been described. The introduction of an additional disulfide
bond, which was originally used to produce soluble recombi-
nant TCR molecules, facilitated TCR pairing and expression in
human T cells (18–20). Similarly, hybrid TCR chains in which
the human constant region was exchanged for murine sequences
displayed improved TCR pairing and enhanced association with
the CD3 molecules in human T cells (21).
We have previously isolated from the allogeneic repertoire a
TCR that is specific for a peptide epitope of Wilms tumor Ag 1
(WT1)4presented in the context of HLA-A2 class I molecules (15,
22). We demonstrated that TCR-transduced human T cells effi-
ciently killed human tumor cells in vitro and were able to inhibit
the growth of a human leukemia cell line in NOD/SCID mice.
However, WT1-TCR expression in freshly transduced human T
cells was generally lower than the expression levels of endogenous
TCR chains, and several rounds of in vitro stimulation with WT1
peptides were required to selectively expand the Ag-responsive T
*Department of Immunology and Molecular Pathology, University College London,
Hampstead Campus, Royal Free Hospital, London;†Centro de Biologı ´a Molecular
Severo Ochoa, Consejo Superior de Investigaciones Cientı ´ficas, Universidad Au-
to ´noma de Madrid, Madrid, Spain; and‡Tumor Immunology Group, Department of
Medical Oncology, Eramus Medical Center-Daniel den Hoed Cancer Center, Rotter-
dam, The Netherlands
Received for publication June 8, 2007. Accepted for publication August 23, 2007.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1This work was supported by the European Union-funded ATTACK project and by
the Leukemia Research Fund.
2Current address: Laboratory of Functional Immunogenetics, The Babraham Insti-
tute, Babraham Research Campus, Cambridge, U.K.
3Address correspondence and reprint requests to Prof. Hans J. Stauss. Department of
Immunology and Molecular Pathology, University College London, Hampstead Cam-
pus, Royal Free Hospital, Rowland Hill Street, London, United Kingdom. E-mail
4Abbreviations used in this paper: WT1, Wilms tumor Ag 1; MFI, mean fluorescence
Copyright © 2007 by The American Association of Immunologists, Inc. 0022-1767/07/$2.00
The Journal of Immunology
The goal of this study was to explore whether cysteine modifi-
cations or the insertion of murine constant region sequences can
improve the function of the WT1-TCR when transduced into hu-
man T cells. In one TCR variant, the native disulfide bond between
the ?- and ?-chains was removed and a new disulfide bond was
introduced. This TCR construct was no longer able to form het-
erodimers, and the TCR ?-chain was expressed on the cell surface
in the absence of the ?-chain. In another TCR variant, a new di-
sulfide bond was introduced without changing the native bond.
This TCR modification reduced mispairing and enhanced correct
pairing between the modified chains. Similar results were obtained
with a variant TCR containing the murine constant region se-
quences. Surprisingly, although human T cells expressing the cys-
teine-modified and hybrid TCR displayed enhanced tetramer bind-
ing, this did not result in higher functional avidity or enhanced
Materials and Methods
Media, cells, Abs, tetramer, and peptides
Unless otherwise stated, all culture media were RPMI 1640 (Cambrex)
supplemented with 10% heat inactivated FCS (Sigma-Aldrich), 1% peni-
cillin/streptomycin (Invitrogen Life Technologies), and 1% L-glutamine
(Invitrogen Life Technologies). The cells lines used were the human TCR-
negative Jurkat 76 cell line, the HLA-A2-positive leukemia cell line
K562-A2 that expresses endogenous WT1 protein, and the HLA-A2-pos-
itive lymphoblastoid cell line C1R-A2 that is WT1 negative. The HLA-
A2-positive T2 cell line is deficient in TAP (transporter associated with Ag
processing) and can be efficiently loaded with exogenous peptides. PBMCs
were obtained from volunteer donors from the National Blood Service,
Colindale, London, U.K. Flow cytometry Abs were anti-human PE ??
TCR, allophycocyanin CD3 ?, allophycocyanin IFN ? (BD Biosciences)
and PE V?2.1 (Immunotech). PE-labeled HLA-A2/WT126 tetramers were
obtained from Beckman Coulter and used at 3.3 ?g/ml. The peptides used
in this study were the HLA-A2 binding peptides pWT126 (RMFPNAPYL)
and pWT235 (CMTWNQMNL) and were synthesized by ProImmune as
described previously (15).
Retroviral TCR constructs
All TCR ?- and ?-chain constructs were cloned separately into retroviral
pMP71 vectors. The pMP71 vector containing the WT1-TCR genes was
described previously (15). Cysteine-modified ? and ? TCR chains were
generated by PCR mutagenesis. For the introduction of the disulfide bond
in the constant domain, residue 48 of the V?1.5 TCR chain was changed
from a threonine to a cysteine and residue 57 of the V?2.1 TCR chain was
changed from a serine to a cysteine. For the removal of the endogenous
disulfide bond, residue 95 of the V?1.5 TCR chain and residue 131 of
the V?2.1 TCR chain were changed from cysteines to serines. To gen-
erate the hybrid TCR ?-chain, a fragment encoding the first 141 residues
of the WT1-specific TCR ?-chain was joined to the C-terminal 130 resi-
dues of the murine MDM2-specific TCR ?-chain (23). The hybrid ?-chain
was subsequently cloned upstream of an IRES-GFP (where IRES is inter-
nal ribosome entry site) element in the pMP71 vector using the restriction
sites NotI and SalI. The hybrid TCR ?-chain was similarly produced by
fusing the N-terminal 136 residues of WT1-TCR ?-chain to the C-terminal
168 residues of a murine MDM2-specific TCR ?-chain. The fragment was
then transferred into the pMP71 vector using restriction sites NotI and
BsrGI. The HLA-A2-restricted Tax-TCR (V?12.2 and V?13.1) (24) (spe-
cific for the peptide sequence LLFGYPVYV of HTLV-1) was provided by
Dr. B. Jakobsen (MediGene), and HLA-A2-restricted LMP2 TCR (V?3.1
and V?13.1; specific for the peptide sequence CLGGLLTMV of the LMP2
protein of EBV) was provided by Dr. B. Wilcox (University of Birming-
ham, Birmingham, U.K.).
Transduction of retroviral TCR constructs into Jurkat cells and
primary T cells
For retroviral transduction, 2 ? 106Phoenix amphotropic packaging cells
were cultured in 10-cm culture plates for 24 h at 37°C with 5% CO2in
DMEM supplemented with 10% heat-inactivated FCS, 1% penicillin/strep-
tomycin, and 1% L-glutamine. The culture medium was changed and the
cells were transfected with the vector constructs and pCL-ampho using
calcium phosphate precipitation (Invitrogen Life Technologies). After cul-
turing for 24 h at 37°C with 5% CO2, the DMEM culture medium was
replaced with RPMI 1640 culture medium and incubated for a further 24 h.
The viral supernatant was then harvested. Jurkat cells were split 24 h before
retroviral transduction and PBMCs were activated for 48 h using the anti-
CD3 Ab OKT3 at 30 ng/ml and IL-2 (600 U/ml; Chiron). For retroviral
transductions, retronectin-coated (Takara) 24-well plates were seeded with
cells at 1 ? 106per well in 1 ml, cultured for 30 min, and then transduced
with 500 ?l of the TCR ?-chain viral supernatant and 500 ?l of the TCR
?-chain viral supernatant. For PBMCs the transductions were conducted in
culture medium supplemented with IL-2 at 600 U/ml. After 24 h at 37°C
with 5% CO2, the culture medium for Jurkat cells was replaced and for
PBMCs the replaced medium was supplemented with IL-2 at 100 U/ml.
Flow cytometry analysis was conducted on a BD LSR II flow cytometer
(BD Biosciences) after a further 48-h culture period. FACS data were
analyzed using FACSDiva or WinMDI version 2.9 software.
For the immunoprecipitation experiments Jurkat cells were transduced with
the various TCR ? and ?-chains and then subsequently cloned by limiting
dilution to produce populations of cells that were ?95% TCR positive. For
each immunoprecipitation experiment 3 ? 107Jurkat cells were lysed in 1
ml of Brij 96 lysis buffer containing protease inhibitors (1% Brij 96, 150
mM NaCl, 10 mM Tris-HCl (pH 7,8), 10 mM iodoacetamide, 1 mM
PMSF, 1 ?g/ml leupeptin, and 1 ?g/ml aprotinin). Protein A-Sepharose
together with the anti-V?2.1 Abs were added to the lysates and incubated
for 4 h at 4°C. The immunoprecipitates were resolved in a 7–17% SDS-
polyacrylamide gel, immunotransferred to a nitrocellulose membrane and
incubated sequentially with anti-TCR ? (clone ?F1; Endogen), anti-CD3 ?
(clone M20; Santa Cruz Biotechnology), and anti-CD3 ? (clone 448; Ref.
25) Abs. The membranes were afterward hybridized with streptavidin HRP
(Amersham Biosciences) and developed by ECL (Bio-Rad Laboratories).
Ag-stimulation of TCR-transduced T cells
Transduced primary T cells were stimulated and expanded every 8–10
days. The stimulations were conducted in 24-well plates in 2 ml of culture
medium containing 10% nonheat-inactivated FCS and 10 U/ml IL-2
(Roche) at 37°C with 5% CO2. Each well contained 5 ? 105transduced
cells, 2 ? 105irradiated T2 cells loaded for 2 h with 100 ?M of the
pWT126 (stimulator cells), and 2 ? 106irradiated PBMCs as feeder cells.
IFN-? secretion assays
TCR transduced T cells (1 ? 105) were stimulated with 1 ? 105irradiated
T2 cells loaded for 2 h with pWT126 (relevant peptide) or pWT235 (ir-
relevant peptide). Assays were conducted in triplicates in round-bottom
96-well plates in 200 ?l of culture medium. After 18 h of incubation at
37°C with 5% CO2, the supernatant was harvested and tested for secreted
IFN-? using a human ELISA kit (BD Biosciences) as per the manufactur-
er’s instructions. The data was analyzed using Excel software.
Intracellular IFN-? detection assays
This assay was performed in 96-well round-bottom plates. TCR-transduced
T cells and T2 stimulator cells loaded with relevant (pWT126) or irrelevant
(pWT235) peptide were added at 4 ? 105/well in 200 ?l of culture medium
containing brefeldin A (Sigma-Aldrich) at 1 ?g/ml. After an incubation
period of 4 h at 37°C with 5% CO2, the cells were first stained for surface
CD8 and then fixed, permeabilized, and stained for intracellular IFN-?
using the Fix & Perm kit (Caltag) according to the manufacturer’s instruc-
tions. Samples were acquired on a LSR II flow cytometer and the data was
analyzed using FACSDiva (BD Biosciences).
For the CTL assays, T2 cells, K562-A2 cells or C1R-A2 cells were labeled
with51Cr for 1 h at 37°C with 5% CO2in culture medium and washed three
times.51Cr-labeled T2 cells were then loaded with pWT126 peptide or
pWT235 peptide at decreasing concentrations for 1 h at 37°C with 5% CO2
in culture medium. The CTL assays were conducted in round-bottom 96-
well plates in 200 ?l of culture medium at 37°C with 5% CO2. For different
E:T ratios, peptide-loaded51Cr labeled T2 cells were added to 2-fold di-
lutions of TCR-transduced T cells. For the peptide titration assays, TCR-
transduced cells and peptide-loaded51Cr labeled T2 cells were cultured at
a ratio of 5:1 (E:T). After an incubation period of 4 h at 37°C with 5% CO2,
50 ?l of supernatants was harvested, diluted with 150 ?l of scintillation
fluid, and counted using a Wallac 1450 Microbeta Plus counter. Percentage
specific killing ? experimental51Cr-release ? spontaneous51Cr-release/
maximum51Cr-release ? spontaneous51Cr-release.
5804TCR VARIANTS IMPROVE TCR EXPRESSION BUT NOT T CELL FUNCTION
Expression of wild-type, cysteine mutant, and hybrid TCR
We have generated three variants of the WT1-specific TCR (Fig.
1). In the cysteine variant 1, position 48 in the constant region of
the TCR ?-chain was changed from threonine to cysteine and from
serine to cysteine at position 57 in the constant TCR ?-chain as
described in recent studies (18–20). Cysteine variant 2 was iden-
tical to variant 1, except that the cysteine residues responsible for
the formation of the natural disulfide bond between the TCR ?-
and ?-chains were removed in an attempt to prevent pairing with
wild-type TCR chains. Like the parental TCR, variant 2 could only
form one disulfide bond between the ?- and ?-chain. Finally, the
constant regions of the human WT1 TCR ?- and ?-chains were
replaced with murine sequences to generate the third variant
Wild-type and variant TCR ? and ? genes were inserted sepa-
rately into the retroviral vector MP71 for the transduction of hu-
man T cells. The TCR constructs were introduced into CD3-pos-
itive, TCR-negative Jurkat 76 T cells followed by FACS analysis
using Abs specific for the V?2.1 variable segment used by the
WT1-TCR and anti-CD3 ? Abs. As expected, control Jurkat cells
did not stain with these Abs whereas cells transduced with the
wild-type WT1-TCR, the cysteine-1, and the hybrid TCR double-
stained with anti-TCR and anti-CD3 ? Abs (Fig. 2A), suggesting
that the introduced TCR was assembled with endogenous CD3
components. Surprisingly, Jurkat cells transduced with the cys-
teine-2 TCR construct stained with the anti-TCR Abs but not with
anti-CD3 ? Abs (Fig. 2A), indicating that this cysteine-modified
TCR was expressed on the cell surface without CD3 ?. Staining of
permeabilized cells showed that the cysteine-2 TCR was expressed
intracellularly at similar levels as those of the wild-type and cys-
teine-1 TCRs (Fig. 2B).
We performed immunoprecipitation with anti-V?2.1 Abs fol-
lowed by Western blotting to analyze whether the cysteine-2 TCR
?-chain was associated with TCR ?, CD3 ? and CD3 ?. As ex-
pected anti-V?2.1 immunoprecipitates of the wild-type and cys-
teine-1 TCRs coprecipitated TCR ?, CD3 ?, and CD3 ? (Fig. 3). In
contrast, anti-V?2.1 precipitates of Jurkat cells expressing the cys-
teine-2 TCR failed to coprecipitate TCR ?, CD3 ? or CD3 ?,
suggesting that the TCR ?-chain was expressed without the
?-chain and CD3 molecules. This was confirmed by the demon-
stration that Jurkat cells transduced with only the cysteine-2
?-chain stained with anti-V?2.1 Abs as efficiently as Jurkat cells
transduced with the cysteine-2 TCR ?? combination (Fig. 2A). As
expected, the cysteine-2 TCR was nonfunctional (data not shown)
and was omitted from subsequent experiments.
Analysis of TCR mispairing
We used Jurkat cells to determine whether the cysteine-1 and the
hybrid WT1-TCR were able to pair with the wild-type chains of
the WT1-TCR or two unrelated TCRs, one specific for a Tax pep-
tide of human T cell leukemia virus type 1 (HTLV-1) and the other
for a LMP2 peptide of EBV. The same batches of retroviral vector
preparations for the ? and the ? TCR genes were used for these
mispairing experiments. Coinfection of Jurkat cells with the vec-
tors containing the “matched” ? and ? genes resulted in the surface
expression of the wild-type TCR, the cysteine-1 TCR, and hybrid
this study. The cysteine-1 variant TCR and the cysteine-2 variant TCR
contain an introduced cysteine bond at residues 48 and 57 of the ? and ?
TCR chains, respectively. The cysteine-1 variant TCR also contains the
endogenous bond that has been removed in the cyteine-2 variant TCR. In
the hybrid variant TCR, the human constant region has been replaced with
Schematic representation of the WT1-TCR variants used in
cells. Jurkat cells, negative for endogenous TCRs, were transduced with
vectors encoding the ?- and ?-chains of the wild-type TCR, cysteine-1
TCR, cysteine-2 TCR, and hybrid TCR. In addition, Jurkat cells were
also transduced with only the ?-chain of the cysteine-2 TCR. A, Trans-
duced cells were Ab stained with anti-V?2.1 and anti-CD3 ? Abs, fol-
lowed by FACS analysis to determine cell surface coexpression of TCR
and CD3 ?. B, Transduced cells were permeabilized followed by stain-
ing with anti-V?2.1 Ab and FACS analysis to determine cell surface
and intracellular TCR ? expression.
WT1-TCR variants are expressed in transduced Jurkat
kat cells transduced with three variants of the WT1 TCR were lysed in 1%
Brij 96 and immunoprecipitated with the specific Ab V?2.1. The immu-
noprecipitates were subjected to 7–17% SDS-polyacrylamide gel electro-
phoresis, transferred to nitrocellulose, and immunoblotted with either anti-
? mAb (clone ?F1), anti-CD3 ? mAb (clone M20), or anti-CD3 ?
polyclonal Ab 448. The specific chains are shown by arrows.
Biochemical analysis of TCR-transduced Jurkat cells. Jur-
5805The Journal of Immunology
TCR as well as the Tax-TCR and LMP2-TCR (Fig. 4A). The per-
centage of Jurkat cells expressing “matched” TCR was used as
reference point for subsequent experiments using the same amount
of the retroviral preparations for TCR mispairing experiments.
The wild-type ?- and ?-chains of the WT1-TCR were coex-
pressed with the ?- and ?-chains of the Tax-TCR and LMP2-TCR
to determine the level of cross-pairing between these unmodified
human TCR chains. Fig. 4B shows that the wild-type WT1-TCR
?-chain paired with the Tax-TCR ?-chains at similar levels as the
Tax ?-chain. The pairing between WT1-TCR ? and LMP2-TCR ?
was less efficient than LMP2 ?? pairing. In contrast, the WT1-
TCR ?-chain paired efficiently with the Tax and LMP2 ?-chains
The cysteine-1 ?-chain showed reduced pairing with the
?-chains of the Tax-TCR and LMP2-TCR, whereas no reduc-
tion in pairing was seen with the ?-chain of the wild-type WT1-
TCR (Fig. 4C). The cysteine-1 ?-chain showed no reduction in
pairing with wild-type ?-chains of the WT1, Tax, and LMP2
TCR (Fig. 4C).
The hybrid TCR ?-chain showed no detectable reduction in
pairing with the analyzed ?-chains (Fig. 4D). In contrast, the hy-
brid TCR ?-chain showed poor pairing with the wild-type TCR
?-chain while it paired efficiently with the ?-chains of the Tax and
LMP2 TCRs (Fig. 4D).
Together, these data show that the alterations in the WT1-TCR
constructs reduced pairing with some wild-type TCR chains while
maintaining the pairing efficiency with other wild-type chains.
Finally, we analyzed the level of mispairing of the WT1-TCR
?-chain and the cysteine-1 and hybrid versions with the repertoire
of endogenous TCR ?-chains present in primary human T cells.
Activated human T cells were transduced with the retroviral TCR
? constructs only, followed by anti-V?2.1 staining to detect cells
expressing the introduced ?-chains. Flow cytometry of transduced
T cells revealed that a similar percentage of human T cells ex-
pressed the introduced wild-type, cysteine-1, and hybrid ?-chains
(Fig. 4E). This indicated that the modified ?-chains paired with the
endogenous repertoire of ?-chains as efficiently as the wild-type
Due to the lack of anti-V?1.5 Abs, similar experiments could
not be performed with the modified WT1-TCR ?-chains.
Analysis of TCR-transduced primary T cells
The retroviral vectors encoding the wild-type WT1-TCR, the cys-
teine-1, and the hybrid versions were used to cotransfer the ? and
? genes into primary human T cells. We measured the ability of
transduced bulk T cells to produce IFN-? after stimulation with
WT1 peptides and control peptides. Consistently, early bulk cul-
tures transduced with the modified TCR constructs produced more
of the wild-type (WT) TCR, the cysteine-1 (Cys-1) TCR, and the hybrid (Hyb) TCR. Jurkat cells were also transduced with vectors encoding the
?- and ?-chains of the Tax-TCR and the LMP2-TCR. B, Jurkat cells were cotransduced with vectors encoding the ?-chain of the wild-type TCR
and the ?-chains of the Tax-TCR or the LMP2-TCR. Alternatively, Jurkat cells were cotransduced with vectors encoding the ?-chain of the wild-type
TCR and the ?-chains of the Tax-TCR or the LMP2-TCR. C, Jurkat cells were cotransduced with vectors encoding the ?-chain of the cysteine-1
variant TCR and the ?-chain of the wild-type TCR, the Tax-TCR, or the LMP2-TCR. Alternatively, Jurkat cells were cotransduced with vectors
encoding the ?-chain of the cysteine-1 variant TCR and the ?-chain of the wild-type TCR, the Tax-TCR or the LMP2-TCR. D, Jurkat cells were
cotransduced with vectors encoding the ?-chain of the hybrid variant TCR and the ?-chain of the wild-type TCR, the Tax-TCR, or the LMP2-TCR.
Alternatively, Jurkat cells were cotransduced with vectors encoding the ?-chain of the hybrid variant TCR and the ?-chain of the wild-type TCR,
the Tax-TCR, or the LMP2-TCR. Transduced cells (A–D) were stained with anti-V?2.1 for cells expressing the WT-1 ?-chain or with anti-human
?? TCR for cells expressing the Tax or LMP2 ?-chain and anti-CD3 ? Abs and then FACS analyzed to determine cell surface expression of TCR
and CD3 ?. E, Primary human PBMCs were mock transduced or transduced with vectors encoding the ?-chains of the wild-type TCR, cysteine-1
variant TCR, or hybrid variant TCR. Cells were stained with anti-V?2.1 and anti-CD8 Abs followed by FACS analysis.
WT1-TCR variants can mispair with wild-type TCR chains. A, Jurkat cells were transduced with vectors encoding the ?- and ?-chains
5806TCR VARIANTS IMPROVE TCR EXPRESSION BUT NOT T CELL FUNCTION
IFN-? than cultures transduced with the wild-type TCR (Fig. 5).
This is consistent with recent reports demonstrating the enhanced
effector function of human T cells transduced with cysteine-1 and
hybrid TCR constructs (18, 19, 21).
We explored further the reason for the enhanced Ag response
after transduction with the WT1-TCR variants. It was possible that
individual T cells expressing the TCR variants mounted stronger
peptide-specific effector functions than T cells expressing the wild-
type WT1-TCR. Alternatively, it was possible that the TCR vari-
ants assembled more efficiently in transduced bulk T cells and thus
generated a higher frequency of Ag-responsive T cells, which may
account for the improved peptide-specific response compared with
bulk T cells transduced with the wild-type TCR.
Monitoring the accumulation of Ag-responsive T cells
Staining with anti-V?2.1 Abs and HLA-A2/WT1 tetramers was
used in an attempt to monitor the numbers of T cells expressing the
introduced WT1-TCR ?-chain, and the ?? heterodimer, respec-
tively. Primary human T cells transduced with the wild-type, cys-
teine-1 and hybrid TCRs contained similar percentages of CD8?T
cells expressing the V?2.1?TCR chain (4.5–4.7%), indicating
similar transduction efficiency (Fig. 6A). A small number of tet-
ramer-positive T cells were clearly detectable after transduction
with the cysteine-1 and hybrid TCRs, but not after transduction
with the wild-type TCR (Fig. 6A). This initial observation sug-
gested that the wild-type TCR was unable to assemble detectable
levels of functional ?? heterodimers in primary T cells.
However, repeated peptide-stimulation of T cells transduced
with the wild-type TCR led to the selective accumulation of
more peptide-specific IFN-? when compared with bulk T cells transduced
with wild-type TCR. Primary T cells transduced with wild-type TCR, cys-
teine-1 TCR, and hybrid TCR were stimulated with pWT126. After 8 days
the percentage of CD8?V?2.1?cells was 4.8, 5.8, and 4% for T cells
transduced with wild-type TCR, cysteine-1 TCR, and hybrid TCR, re-
spectively. The percentage of CD8?tetramer?cells was 0.2, 1.2, and
1% for T cells transduced with wild-type TCR, cysteine-1 TCR, and
hybrid TCR, respectively. These cells were stimulated with T2 cells
loaded with relevant and irrelevant peptides. Specific IFN-? release was
determined by subtracting the IFN-? release induced by an irrelevant
peptide from the IFN-? release induced by a relevant peptide.
Bulk T cells transduced with WT-1 TCR variants release
the ?- and ?-chains of the wild-type TCR, cysteine-1 TCR, and hybrid TCR. The percentage of CD8?V?2.1?cells and CD8?tetramer?cells after mock
transduction was 1.4 and 0%, respectively. A, FACS analysis of freshly transduced T cells stained with anti-CD8 Ab in parallel with V?2.1 Ab or
HLA-A2/WT1 tetramer. B, FACS analysis of transduced T cells that have undergone three rounds of peptide stimulation and stained with anti-CD8 Ab
in parallel with V?2.1 Ab or HLA-A2/WT1 tetramer. The MFI of V?2.1 for these cells was 1955, 2235, and 2128 for T cells transduced with wild-type
TCR, cysteine-1 TCR, and hybrid TCR, respectively. C, FACS analysis of transduced T cells that have undergone six rounds of peptide stimulation and
stained with anti-CD8 Ab in parallel with V?2.1 Ab or HLA-A2/WT1 tetramer. The MFI of V?2.1 for these cells was 812, 996, and 1178 for T cells
transduced with wild-type TCR, cysteine-1 TCR, and hybrid TCR, respectively. D, T cells were stimulated with control peptide (gray histograms) or
pWT126 peptide (open histograms) for 4 h followed by intracellular staining with IFN-? cytokine Ab. The percentage of CD8?V?2.1?cells for this
experiment was 89, 87, and 90% for T cells transduced with wild-type TCR, cysteine-1 TCR, and hybrid TCR, respectively.
Phenotype and functional analysis of primary T cells transduced with WT1-TCR constructs. PBMCs were transduced with vectors encoding
5807The Journal of Immunology
CD8?V?2.1?T cells that remained largely tetramer negative, sug-
gesting that tetramer-negative T cells expressed functional WT1-
TCR heterodimers that were able to respond to stimulation with
WT1 peptides (Fig. 6B). After three rounds of peptide stimulation
larger numbers of CD8?V?2.1?T cells accumulated in cultures
transduced with the cysteine-1 and hybrid TCRs (57 and 66%
CD8?V?2.1?T cells, respectively) compared with cultures trans-
duced with the wild-type TCR (35% CD8?V?2.1?T cells; Fig.
6B). In addition, a large proportion of the T cells expressing the
cysteine-1 and hybrid TCR-bound tetramers (22 and 21%),
whereas only 1.4% of T cells expressing the wild-type TCR were
tetramer positive. Additional experiments (data not shown) dem-
onstrated that the differences in the staining profiles between T
cells expressing the wild-type, cysteine-1, and hybrid TCRs were
also detectable using a 10 times higher concentration of tetramer.
After six rounds of peptide stimulation, all cultures transduced
with the different TCR constructs accumulated a large percentage
(88–95%) of CD8?V?2.1?T cells (Fig. 6C). As seen after 3 wk,
many of the CD8?V?2.1?T cells expressing the cysteine-1 and
hybrid TCRs bound HLA-A2/WT1 tetramers, while most
CD8?V?2.1?T cells expressing the wild-type TCR remained tet-
ramer negative (Fig. 6C).
Similar observations were made with transduced T cells from
different donors in independent experiments. Consistently, during
the first three rounds of peptide stimulation CD8?V?2.1?T cells
accumulated rapidly after transduction with the cysteine-1 and hy-
brid TCRs, and many of these T cells bound WT1 tetramers. The
accumulation of CD8?V?2.1?T cells occurred more slowly after
transduction with the wild-type TCR, and most of the T cells were
tetramer negative. Continued Ag selection beyond three rounds of
stimulation resulted in the accumulation of similar high numbers
of CD8?V?2.1?cells in cultures expressing the wild-type, cys-
teine-1, or hybrid TCR, but accumulation of tetramer-positive T
cells only occurred in cultures expressing modified TCRs.
CD8?V?2.1?T cells expressing wild-type, cysteine-1, or hybrid
TCR are functionally equivalent
We used Ag-selected T cell lines containing mostly CD8?V?2.1?
T cells to test whether we could detect functional differences be-
tween the wild-type, cysteine-1, and hybrid TCRs. Intracellular
IFN-? staining demonstrated that the majority of T cells, indepen-
dently of the TCR they expressed, produced IFN-? when stimu-
lated with the WT1 peptides (Fig. 6D). In these T cell lines, the
percentage of IFN-?-positive T cells was nearly the same as the
percentage of V?2.1?T cells, showing that the majority of V?2.1-
expressing T cells were peptide specific. Furthermore, the mean
fluorescent intensity (MFI) of the IFN-? staining was similar in
cells expressing the wild-type, cysteine-1, and hybrid TCRs (MFI
of 1982, 2014, 2020, respectively), suggesting that the three TCR
constructs triggered comparable levels of intracellular IFN-?. An
ELISA of the culture supernatant was used to measure the amount
of IFN-? that was secreted by the Ag selected T cell lines. The
results indicated that the T cell lines expressing the wild-type,
cysteine-1, and hybrid TCRs secreted similar amounts of IFN-?
(Fig. 7A), confirming the results of the intracellular staining.
Next, the cytotoxic activity of T cell lines expressing the dif-
ferent TCR constructs was determined. Using target cells coated
PBMCs were transduced with vectors encoding the ?- and ?-chains of the wild-type TCR, cysteine-1 TCR, and hybrid TCR and were then peptide
stimulated until the different T cell cultures had similar V?2.1 expression. A, T cells were cultured for 18 h with T2 cells loaded with relevant and irrelevant
peptides. IFN-? release was determined by ELISA. The percentage of CD8?V?2.1?cells for this experiment was 93, 86, and 86% for T cells transduced
with wild-type TCR, cysteine-1 TCR, and hybrid TCR, respectively, whereas the percentage of CD8?tetramer?cells was 2, 30, and 32%, respectively. B,
T cells were cultured in a 4-h assay with51Cr-labeled T2 cells loaded with 100 ?M relevant (left-hand panel) and irrelevant (right-hand panel) peptides
at the stated E:T ratios. C, T cell were cultured in a 4 h assay with T2 cells labeled with51Cr and loaded with the stated pWT126 concentration or with
51Cr-labeled T2 cells loaded with 100 ?M irrelevant peptide. Alternatively, T cells were cultured with51Cr-labeled K562-A2 or C1R-A2 cells. The
percentage of CD8?V?2.1?cells for experiments B and C was 87, 90, and 82% for T cells transduced with wild-type TCR, cysteine-1 TCR, and hybrid
Cultures with tetramer-positive cells do not have an increased T cell specific function compared with cultures without tetramer positive cells.
5808 TCR VARIANTS IMPROVE TCR EXPRESSION BUT NOT T CELL FUNCTION
with the WT1 peptides, similar levels of peptide-specific cytotox-
icity were observed for all TCR constructs even at low effector T
cell to target cell ratios (Fig. 7B). Although this showed that T cells
expressing the wild-type, cysteine-1, and hybrid TCRs all dis-
played similar killing activity, these experiments did not analyze
the triggering threshold for the three TCR constructs. To explore
this, cytotoxicity assays were performed against target cells coated
with decreasing concentrations of the WT1 peptide. The results
revealed a nearly identical dose-response for the T cell lines ex-
pressing the wild-type, cysteine-1, and hybrid TCRs (Fig. 7C).
Similarly, the T cells expressing wild-type TCR killed the leuke-
mia cell line K562-A2, which expresses WT1 endogenously, as
efficiently as T cells expressing the cysteine-1 and hybrid TCRs
(Fig. 7C). The WT1-negative cell line C1R-A2 was used as a neg-
ative control in these experiments.
Together, these results showed that although T cell lines ex-
pressing the cysteine-1 and hybrid TCRs consistently contained
much higher numbers of tetramer binding CD8?cells than T cell
lines expressing the wild-type TCR, this was not associated with a
measurable increase in IFN-? production, T cell avidity, or killing
activity against WT1-expressing tumor cells.
We have analyzed the expression and function of three variants of
the WT1 TCR. The introduction of a new disulfide bond and re-
moval of the native bond produced TCR chains that were unable to
pair as heterodimers. The cysteine-2 TCR ?-chain not only failed
to pair with the ?-chain but also with the CD3 ?- and ?-chains.
Despite this lack of pairing, the cysteine-2 ?-chain was expressed
on the cell surface, presumably as a homodimer. We noted that the
human/murine hybrid ?-chain displayed similar properties, as it
was able to reach the cell surface without the TCR ?- and the CD3
?-chain. This was apparent in Jurkat cells transduced with the hy-
brid ? gene together with TCR ? genes where a population of cells
that expressed TCR ? but not CD3 ? was detectable by FACS
analysis (see Fig. 4A). Furthermore, this TCR ??CD3 ??popu-
lation was also apparent in Jurkat cells transduced with only the
hybrid TCR ? gene (data not shown). Thus, the introduction of
murine constant sequences and the double cysteine modification
generated TCR ?-chains that acquired the ability to be expressed
without TCR ? and CD3 ?. Although the cysteine-2 TCR ?-chain
lost the ability to assemble with TCR ? and CD3 ? when present
in transduced cells, this ability was retained by the hybrid TCR
?-chain. This highlights the fact that genetic TCR modifications
can result in unexpected alterations of TCR assembly and CD3
A detailed analysis of mispairing showed that the introduction
of murine constant region sequences or the addition of a single
cysteine (without removal of the native cysteine) decreased pairing
with only some wild-type TCR chains. At present, the molecular
mechanisms that determine the efficiency of TCR pairing are not
known. Recent elegant studies with human TCR genes showed that
“strong” TCR pairs were efficiently expressed while “weak” TCR
pairs were inefficiently expressed on the surface of T cells (26, 27).
As the constant regions of human TCR ? and ? genes are identical,
except for possible differences due to the usage of constant ? 1 and
2, these studies showed that the variable region sequences play a
major role in determining the efficiency of TCR expression (26). It
is likely that efficient ?-? pairing and formation of stable het-
erodimers are features of a “strong” TCR, whereas inefficient pair-
ing is a feature of a “weak” TCR. We postulate that the TCR
constant region modifications explored here more readily disrupt
“weak” TCR ?? combinations, whereas little effect is seen with
“strong” TCR combinations where the variable region sequences
drive efficient ?? pairing that can proceed despite modifications in
the constant region.
Similarly as in previous studies, we observed that T cells trans-
duced with the hybrid and the cysteine-1 TCRs bound tetramers
more efficiently than T cells transduced with the wild-type WT1
TCR (18, 19, 21). Ag-stimulation of T cells transduced with the
modified TCRs resulted in the expansion of V?2.1?tetramer?as
well as V?2.1?tetramer?T cells, indicating that both were Ag
responsive. Ag stimulation of T cells transduced with the wild-type
TCR resulted in the expansion of T cells that were largely unable
to bind tetramer. The expression levels of V?2.1, as determined by
MFI, was generally lower in cells expressing the wild-type TCR
compared with cells expressing the modified TCRs (e.g., Fig. 6, B
and C). Although this difference was relatively modest, it might be
sufficient to alter tetramer staining, which is based on low affinity
interactions that might be susceptible to small changes in the den-
sity of the TCR ligand. In contrast, T cell activation involves mul-
tiple receptor/ligand interactions, including ligation of the TCR,
the CD8 coreceptor, costimulatory molecules such as CD28, and
accessory molecules such as LFA1, which may render this activa-
tion pathway less susceptible to small reductions in the amounts of
TCR expressed by the responding T cells. This could explain our
observation that, despite differences in tetramer staining, T cells
expressing the wild-type, hybrid, and cysteine-1 TCRs showed
similar levels of tumor cell killing and displayed comparable avid-
ity as determined by peptide titration in cytotoxicity assays (Fig.
7C) and IFN-? production (not shown). This observation that the
lack of tetramer binding was not associated with a reduction in T
cell avidity is similar to the demonstration that hepatitis B-specific
CD8?T cells of chronically infected patients showed reduced tet-
ramer binding while retaining high avidity T cell function (28).
This functional similarity of the TCR constructs seems to con-
trast with previous studies where hybrid and cysteine-modified
TCR displayed increased cytokine production and cytotoxicity
when introduced into human T cells. In these studies, T cells were
functionally analyzed immediately after RNA transfection (18, 21)
or after retroviral transduction followed by expansion with CD3/
CD28 beads (19). Importantly, the T cells were not selected using
Ag-stimulation before the functional analysis. In our study, bulk T
cells transduced with the modified TCR constructs showed stron-
ger peptide-specific IFN-? production than bulk T cells transduced
with the wild-type TCR. However, after several rounds of peptide
stimulation all T cell lines accumulated a high percentage of Ag-
responsive T cells, at which stage functional differences between
the modified and the wild-type TCR were no longer detectable.
These data suggest that the TCR modifications provide a major
advantage by increasing the frequency of freshly transduced T
cells expressing sufficient levels of the introduced TCR ?? het-
erodimer to mount peptide-specific immune responses. This is
most likely due to the recently demonstrated reduction of mispair-
ing (18, 19, 21), the improved ?? pairing between the modified
TCR chains (19), and the stable association with CD3 molecules
(21) that together would be expected to increase the frequency of
successful competition with the endogenous TCR chains. In con-
trast, the wild-type TCR chains succeeded less frequently in this
competition, thus reducing the frequency of Ag-responsive T cells
present in a bulk population of transduced PBMCs.
The authors have no financial conflict of interest.
1. Clay, T. M., M. C. Custer, J. Sachs, P. Hwu, S. A. Rosenberg, and
M. I. Nishimura. 1999. Efficient transfer of a tumor antigen-reactive TCR to
5809 The Journal of Immunology
human peripheral blood lymphocytes confers anti-tumor reactivity. J. Immunol. Download full-text
2. Cooper, L. J., M. Kalos, D. A. Lewinsohn, S. R. Riddell, and P. D. Greenberg.
2000. Transfer of specificity for human immunodeficiency virus type 1 into pri-
mary human T lymphocytes by introduction of T-cell receptor genes. J. Virol. 74:
3. Fujio, K., Y. Misaki, K. Setoguchi, S. Morita, K. Kawahata, I. Kato, T. Nosaka,
K. Yamamoto, and T. Kitamura. 2000. Functional reconstitution of class II MHC-
restricted T cell immunity mediated by retroviral transfer of the ?? TCR com-
plex. J. Immunol. 165: 528–532.
4. Kessels, H. W., M. C. Wolkers, M. D. van den Boom, M. A. van der Valk, and
T. N. Schumacher. 2001. Immunotherapy through TCR gene transfer. Nat. Im-
munol. 2: 957–961.
5. Stanislawski, T., R. H. Voss, C. Lotz, E. Sadovnikova, R. A. Willemsen,
J. Kuball, T. Ruppert, R. L. Bolhuis, C. J. Melief, C. Huber, et al. 2001. Cir-
cumventing tolerance to a human MDM2-derived tumor antigen by TCR gene
transfer. Nat. Immunol. 2: 962–970.
6. Clay, T. M., and M. I. Nishimura. 2003. Retroviral transfer of T-cell receptor
genes into human peripheral blood lymphocytes. Methods Mol. Biol. 215:
7. Heemskerk, M. H., M. Hoogeboom, R. A. de Paus, M. G. Kester,
M. A. van der Hoorn, E. Goulmy, R. Willemze, and J. H. Falkenburg. 2003.
Redirection of antileukemic reactivity of peripheral T lymphocytes using gene
transfer of minor histocompatibility antigen HA-2-specific T-cell receptor com-
plexes expressing a conserved ? joining region. Blood 102: 3530–3540.
8. Morgan, R. A., M. E. Dudley, Y. Y. Yu, Z. Zheng, P. F. Robbins, M. R. Theoret,
J. R. Wunderlich, M. S. Hughes, N. P. Restifo, and S. A. Rosenberg. 2003. High
efficiency TCR gene transfer into primary human lymphocytes affords avid rec-
ognition of melanoma tumor antigen glycoprotein 100 and does not alter the
recognition of autologous melanoma antigens. J. Immunol. 171: 3287–3295.
9. Schaft, N., R. A. Willemsen, J. de Vries, B. Lankiewicz, B. W. Essers,
J. W. Gratama, C. G. Figdor, R. L. Bolhuis, R. Debets, and G. J. Adema. 2003.
Peptide fine specificity of anti-glycoprotein 100 CTL is preserved following
transfer of engineered TCR ? ? genes into primary human T lymphocytes. J. Im-
munol. 170: 2186–2194.
10. Tahara, H., K. Fujio, Y. Araki, K. Setoguchi, Y. Misaki, T. Kitamura, and
K. Yamamoto. 2003. Reconstitution of CD8?T cells by retroviral transfer of the
TCR ? ?-chain genes isolated from a clonally expanded P815-infiltrating lym-
phocyte. J. Immunol. 171: 2154–2160.
11. Heemskerk, M. H., M. Hoogeboom, R. Hagedoorn, M. G. Kester, R. Willemze,
and J. H. Falkenburg. 2004. Reprogramming of virus-specific T cells into leu-
kemia-reactive T cells using T cell receptor gene transfer. J. Exp. Med. 199:
12. Chamoto, K., T. Tsuji, H. Funamoto, A. Kosaka, J. Matsuzaki, T. Sato, H. Abe,
K. Fujio, K. Yamamoto, T. Kitamura, et al. 2004. Potentiation of tumor eradi-
cation by adoptive immunotherapy with T-cell receptor gene-transduced T-helper
type 1 cells. Cancer Res. 64: 386–390.
13. Engels, B., E. Noessner, B. Frankenberger, T. Blankenstein, D. J. Schendel, and
W. Uckert. 2005. Redirecting human T lymphocytes toward renal cell carcinoma
specificity by retroviral transfer of T cell receptor genes. Hum. Gene Ther. 16:
14. Morris, E. C., A. Tsallios, G. M. Bendle, S. A. Xue, and H. J. Stauss. 2005. A
critical role of T cell antigen receptor-transduced MHC class I-restricted helper
T cells in tumor protection. Proc. Natl. Acad. Sci. USA 102: 7934–7939.
15. Xue, S. A., L. Gao, D. Hart, R. Gillmore, W. Qasim, A. Thrasher, J. Apperley,
B. Engels, W. Uckert, E. Morris, and H. Stauss. 2005. Elimination of human
leukemia cells in NOD/SCID mice by WT1-TCR gene-transduced human T cells.
Blood 106: 3062–3067.
16. Morgan, R. A., M. E. Dudley, J. R. Wunderlich, M. S. Hughes, J. C. Yang,
R. M. Sherry, R. E. Royal, S. L. Topalian, U. S. Kammula, N. P. Restifo, et al.
2006. Cancer regression in patients after transfer of genetically engineered lym-
phocytes. Science 314: 126–129.
17. Rosenberg, S. A., and M. E. Dudley. 2004. Cancer regression in patients with
metastatic melanoma after the transfer of autologous antitumor lymphocytes.
Proc. Natl. Acad. Sci. USA 101 (Suppl. 2): 14639–14645.
18. Cohen, C. J., Y. F. Li, M. El-Gamil, P. F. Robbins, S. A. Rosenberg, and
R. A. Morgan. 2007. Enhanced antitumor activity of T cells engineered to express
T-cell receptors with a second disulfide bond. Cancer Res. 67: 3898–3903.
19. Kuball, J., M. L. Dossett, M. Wolfl, W. Y. Ho, R. H. Voss, C. Fowler, and
P. D. Greenberg. 2006. Facilitating matched pairing and expression of TCR-
chains introduced into human T-cells. Blood 109: 2331–2338.
20. Boulter, J. M., M. Glick, P. T. Todorov, E. Baston, M. Sami, P. Rizkallah, and
B. K. Jakobsen. 2003. Stable, soluble T-cell receptor molecules for crystallization
and therapeutics. Protein Eng. 16: 707–711.
21. Cohen, C. J., Y. Zhao, Z. Zheng, S. A. Rosenberg, and R. A. Morgan. 2006.
Enhanced antitumor activity of murine-human hybrid T-cell receptor (TCR) in
human lymphocytes is associated with improved pairing and TCR/CD3 stability.
Cancer Res. 66: 8878–8886.
22. Gao, L., I. Bellantuono, A. Elsasser, S. B. Marley, M. Y. Gordon, J. M. Goldman,
and H. J. Stauss. 2000. Selective elimination of leukemic CD34?progenitor cells
by cytotoxic T lymphocytes specific for WT1. Blood 95: 2198–2203.
23. McNicol, A. M., G. Bendle, A. Holler, T. Matjeka, E. Dalton, L. Rettig,
R. Zamoyska, W. Uckert, S. A. Xue, and H. J. Stauss. 2007. CD8?/? ho-
modimers fail to function as co-receptor for a CD8-dependent TCR. Eur. J. Im-
munol. 37: 1634–1641.
24. Garboczi, D. N., U. Utz, P. Ghosh, A. Seth, J. Kim, E. A. VanTienhoven,
W. E. Biddison, and D. C. Wiley. 1996. Assembly, specific binding, and crys-
tallization of a human TCR-?? with an antigenic tax peptide from human T
lymphotropic virus type 1 and the class I MHC molecule HLA-A2. J. Immunol.
25. San Jose, E., A. Borroto, F. Niedergang, A. Alcover, and B. Alarcon. 2000.
Triggering the TCR complex causes the downregulation of nonengaged receptors
by a signal transduction-dependent mechanism. Immunity 12: 161–170.
26. Heemskerk, M. H., R. S. Hagedoorn, M. A. van der Hoorn, L. T. van der Veken,
M. Hoogeboom, M. G. Kester, R. Willemze, and J. H. Falkenburg. 2007. Effi-
ciency of T-cell receptor expression in dual-specific T cells is controlled by the
intrinsic qualities of the TCR chains within the TCR-CD3 complex. Blood 109:
27. Sommermeyer, D., J. Neudorfer, M. Weinhold, M. Leisegang, B. Engels,
E. Noessner, M. H. Heemskerk, J. Charo, D. J. Schendel, T. Blankenstein, et al.
2006. Designer T cells by T cell receptor replacement. Eur. J. Immunol. 36:
28. Reignat, S., G. J. Webster, D. Brown, G. S. Ogg, A. King, S. L. Seneviratne,
G. Dusheiko, R. Williams, M. K. Maini, and A. Bertoletti. 2002. Escaping high
viral load exhaustion: CD8 cells with altered tetramer binding in chronic hepatitis
B virus infection. J. Exp. Med. 195: 1089–1101.
5810 TCR VARIANTS IMPROVE TCR EXPRESSION BUT NOT T CELL FUNCTION