Download Class I MHC Molecule Protein in Association with an Allogeneic

This information is current as
of June 12, 2017.
Potent Cytolytic Response by a CD8+ CTL
Clone to Multiple Peptides from the Same
Protein in Association with an Allogeneic
Class I MHC Molecule
Shigeki Kageyama, Theodore J. Tsomides, Naomi Fukusen,
Ioannis A. Papayannopoulos, Herman N. Eisen and Yuri
Sykulev
References
Subscription
Permissions
Email Alerts
This article cites 35 articles, 19 of which you can access for free at:
http://www.jimmunol.org/content/166/5/3028.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2001 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 12, 2017
J Immunol 2001; 166:3028-3034; ;
doi: 10.4049/jimmunol.166.5.3028
http://www.jimmunol.org/content/166/5/3028
Potent Cytolytic Response by a CD8ⴙ CTL Clone to Multiple
Peptides from the Same Protein in Association with an
Allogeneic Class I MHC Molecule1
Shigeki Kageyama,2* Theodore J. Tsomides,3* Naomi Fukusen,4*
Ioannis A. Papayannopoulos,5* Herman N. Eisen,* and Yuri Sykulev6†
N
atural ligands for the Ag-specific receptor (TCR) on
CD8⫹ CTL are complexes of short peptide fragments
bound to MHC class I molecules. These peptides are
normally generated by proteosomes from cytosolic proteins in the
target cells (1) and are transferred into the lumen of the endoplasmic reticulum (ER)7 by transporter molecules. In the ER, peptides
bind nascent class I MHC molecules to form peptide-MHC
(pMHC) complexes that are translocated to the cell surface for
survey by CTL (2).
Clone 2C was originally derived from an H-2b mouse immunized with H-2d cells expressing Kd, Ld, and Dd class I MHC
*Center for Cancer Research and Department of Biology, Massachusetts Institute of
Technology, Cambridge, MA 02139; and †Department of Microbiology and Immunology, Kimmel Cancer Institute, Thomas Jefferson University, Philadelphia, PA
19107
Received for publication September 15, 2000. Accepted for publication December
18, 2000.
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.
1
This work was supported by National Institutes of Health Research Grants AI44477
and CA60686 (to H.N.E.) and AI3254 (to Y.S.) and The W. W. Smith Charitable
Trust Award (to Y.S.).
2
Current address: Fuji Photo Film Co. Ltd., Asaka Research Laboratories, Asaka,
Saitama, Japan 351-8585.
3
Current address: Department of Medicine, Maine Medical Center, Portland, ME
04102.
4
Current address: Department of Biochemistry and Nutrition, National Institute of
Public Health, Tokyo, Japan 108-8638.
5
Present address: AstraZeneca R&D Boston, Worcester, MA 01605.
6
Address correspondence and reprint requests to Dr. Yuri Sykulev, Department of
Microbiology and Immunology, Kimmel Cancer Institute, Thomas Jefferson University, Philadelphia, PA 19107. E-mail address: [email protected]
7
Abbreviations used in this paper: ER, endoplasmic reticulum; pMHC, peptideMHC; OGDH, oxoglutarate dehydrogenase; TFA, trifluoroacetic acid; MALDI, matrix-assisted laser desorption-ionization; QC, glutamine cyclase.
Copyright © 2001 by The American Association of Immunologists
proteins (3). 2C CTL specifically responds to Ld in association
with the naturally processed peptide LSPFPFDL (p2Ca), isolated
from spleen and other tissues (4), or the longer natural peptide
VAITRIEQLSPFPFDL (p2Cb), isolated from the same source and
containing the entire sequence of p2Ca (5). Both peptides are derived from a ubiquitous intracellular protein, oxoglutarate dehydrogenase (OGDH) (5, 6). A synthetic p2Cb peptide was digested
in vitro with cellular extracts containing proteosomes and was
found to produce active fragments, suggesting that it might be a
natural precursor of smaller active peptides (5).
The synthetic peptide QLSPFPFDL (QL9), extending by one
amino acid from the N terminus of p2Ca in the murine OGDH
sequence, was ⬃100-fold more potent in sensitizing Ld⫹ target
cells for specific lysis by 2C CTL than naturally occurring p2Ca
(7). Because this peptide has not been isolated to date from tissue
extracts (5, 8), it is unclear whether it is naturally processed and
presented on target cells as an epitope for 2C CTL or behaves as
a heteroclitic ligand. To address this issue, we tested p2Ca, p2Cb,
and QL9 using three experimental approaches: peptide binding to
Ld on live Ld⫹ cells, in vitro digestion of p2Cb by purified 20S
proteosomes, and the endogenous expression of minigenes encoding these peptides in target cells. We show that QL9 can be produced in vitro from the longer natural precursor p2Cb by 20S
proteosomes and that cells transfected with the QL9 minigene become susceptible to specific lysis by 2C CTL. We also found that
the N-terminal glutamine residue of QL9 is rapidly converted to
pyroglutaminyl in the low pH trifluoroacetic acid (TFA) solution,
the medium generally used to extract peptides from tissues. This
conversion precludes the verification of QL9 by Edman degradation, which may explain our inability to identify this peptide in
tissue extracts.
In addition to QL9, several longer peptides sharing a common
C-terminal sequence with the QL9 end were cleaved from p2Cb by
20S proteosomes. Synthetic analogs of these peptides bind Ld to
0022-1767/01/$02.00
Downloaded from http://www.jimmunol.org/ by guest on June 12, 2017
CTL clone 2C recognizes the allogeneic class I MHC molecule Ld in association with peptides derived from ␣-ketoglutarate
dehydrogenase (oxoglutarate dehydrogenase (OGDH)), a ubiquitous intracellular protein. One of these peptides, QLSPFPFDL
(QL9), elicits more vigorous cytolytic responses than two previously identified naturally processed peptides with overlapping
sequences, LSPFPFDL (p2Ca) and VAITRIEQLSPFPFDL (p2Cb), from OGDH. In this study, we show that QL9 forms a more
stable complex with cell surface Ld than does p2Ca or p2Cb and is processed from the longer, naturally occurring peptide p2Cb
by 20S proteosomes in vitro. The N-terminal cyclized pyroglutaminyl QL9 (pyroQL9), a form of QL9 to which it is converted at
the low pH used for peptide isolation from tissue extracts, is even more active than QL9 in cytotoxicity assays with 2C CTL.
Overall, the results indicate that along with the abundant natural peptides p2Ca and p2Cb, the QL9 and other OGDH peptides
of various lengths, sharing a conserved C-terminal sequence, are also processed and presented with Ld as allogeneic ligands for
T cells expressing 2C TCR. All these peptides, each available in a low amount, could act in concert at the cell surface, resulting
in a high density of cognate ligands that accounts for the exceptionally potent cytolytic response by 2C CTL. The Journal of
Immunology, 2001, 166: 3028 –3034.
The Journal of Immunology
form peptide-Ld complexes recognizable by 2C CTL. All these
data indicate that the strong allogeneic response of 2C CTL to Ld⫹
target cells is mediated by numerous peptides presented on target
cells in the context of Ld. Although most of these peptides are
derived from OGDH, the presence of other Ld-binding peptides
from different sources that might serve as ligands for 2C CTL
cannot be excluded. Recognition of multiple ligands by the same
TCR could help to account for the exceptional vigor of some allogeneic T cell responses.
Materials and Methods
Cells
Peptides
Peptides were synthesized using conventional tBoc chemistry (Biopolymers Laboratory, Massachusetts Institute of Technology, Cambridge, MA),
and many of them were purified by HPLC. Peptide concentrations were
determined by quantitative amino acid analyses and/or bicinchoninic acid
assay.
Peptide radiolabeling
The monoiodinated radioactive form of the mouse CMV peptide (125I1pMCMV) was prepared as previously described (11). Briefly, Na127I and
Na125I were mixed to give a 4- to 5-fold molar excess of total iodide of
predetermined specific activity over peptide. The reaction was conducted in
the presence of Iodobeads (Pierce, Rockford, IL) for 30 min at pH 7.0. The
mono- and di-iodinated peptide derivatives and unlabeled peptide were
separated by HPLC. Fractions containing the monoiodinated peptide were
lyophilized and resuspended in H2O. The specific activities of the radiolabeled peptides were determined by counting aliquots in a gamma counter
(Packard, Downers Grove, IL).
Cytotoxicity assay
Cr-labeled target cells (2 ⫻ 104 cells/well) were incubated in triplicate
with peptide in K medium for 30 – 60 min (37°C) in 96-well round-bottom
microtiter plates before addition of 2C CTL at a CTL:target cell ratio of
5:1. After 4 h at 37°C in a CO2 incubator, plates were centrifuged (200 ⫻
g, 5 min), and 100-␮l aliquots from each supernatant were counted in a
gamma counter. Percent specific lysis was determined as: 100 ⫻ (51Cr
release in the experimental supernatant ⫺ spontaneous release)/(total release in detergent ⫺ spontaneous release). Spontaneous release controls
contained a peptide plus 51Cr-labeled target cells but no 2C CTL. Total
release controls contained 51Cr-labeled target cells in the presence of 10%
detergent (Nonidet P-40). In the absence of peptide, the target cell lysis by
CTL was ⬍1% for C1R-Ld cells and did not exceed 10% for T2-Ld cells.
Construction of episomal vectors with minigene DNA
Episomal plasmids were constructed using the expression vector p8901 as
previously described (13, 14). In brief, two oligonucleotide fragments with
sequences that overlap the desired peptide were annealed, extended with
Klenow DNA polymerase, and amplified by PCR. The PCR product was
cloned into the BamHI-NotI site of the p8901 vector, and the identity of the
resulting plasmid was confirmed by direct sequencing.
Minigene transfection
C1R-Ld and T2-Ld cells were transfected with p8901 plasmids encoding
p2Ca or p2Cb or their variants as well as with a plasmid containing the
adenovirus E3/19-kDa protein ER translocation signal sequence RYMIL
GLLALAAVCSA(A) followed by the QL9 sequence as previously described (13–15). The transfected cells are referred to as C1R-Ld(p2Ca),
C1R-Ld(p2Cb), and C1R-Ld(sig-QL9), respectively (see Table I).
Acid extraction and HPLC fractionation of low molecular mass
material from minigene-transfected C1R-Ld cells
Peptides were isolated from minigene-transfected C1R-Ld cells as previously described (8). In brief, 109 cells were washed with K medium and
immediately frozen in liquid nitrogen for storage at ⫺70°C until subjected
to acid extraction. Thawed cells were combined with 10 ml of 0.7% TFA
and immediately homogenized with 50 strokes in a Dounce homogenizer
(Kontes, Vineland, NJ). After 30 min on ice, the homogenate was centrifuged (31,000 ⫻ g) for 60 min at 4°C, and the supernatant was subjected
to ultrafiltration (2,600 ⫻ g, 4°C) using a Centricon 10 membrane (Mr
cut-off, 10 kDa; Amicon, Beverly, MA). Filtrates were freeze dried and
dissolved in 0.1% TFA for reversed-phase HPLC, using a C18 column
(218TP104; Vydac, Hesperia, CA) and a gradient (0.067%/min) of solvent
B in solvent A, where solvent A is 0.1% TFA and solvent B is acetonitrile
containing 0.085% TFA. Fractions (⬃1.0 ml) collected at 1-min intervals
were dried by Speed-Vac (Savant, Farmindale, NY), dissolved in water,
and assayed for their ability to sensitize T2-Ld cells for specific lysis by
2C CTL.
Processing of p2Cb by 20S proteosomes
20S proteosomes were purified as previously described (1). Before digestion the synthetic peptide p2Cb was HPLC purified to homogeneity. The
purified peptide (0.35 ␮g/ml) was incubated with exhaustively dialyzed
20S proteosomes (0.32 ␮g/ml) in a final volume of 15 ␮l of 13 mM TrisHCl (pH 8.0) containing 5 mM DTT. After 2 h at 37°C, the digestion was
stopped by adding 285 ␮l of 0.2% TFA. The reaction mixture was then
subjected to ultrafiltration using a Centricon 10 membrane (Amicon) to
separate peptides from high molecular mass material as described above.
An aliquot of the isolated peptides was examined by mass spectrometry to
determine the molecular masses of the peptide fragments.
51
Mass spectrometric analysis
Samples were analyzed by matrix-assisted laser desorption-ionization
(MALDI) time-of-flight mass spectrometry (16) on a linear Vestec VT2000
instrument (Vestec/PE Biosystems, Houston, TX), operating at 25-kV accelerating voltage (17). To each sample the MALDI matrix, ␣-cyano-4hydroxycinnamic acid (Aldrich, Metuchen, NJ), was added as a 10 mg/ml
solution in 60% water-40% acetonitrile. Spectra were calibrated with a
mixture of peptides of known sequence in the appropriate mass range (external calibration). All mass spectral data were acquired and processed
using a data system developed at the Massachusetts Institute of Technology
mass spectrometry facility.
Competition assay to measure peptide binding
Results
Peptide binding to Ld on live cells was measured as previously described
(12). In brief, T2-Ld cells were suspended at 107 cells/ml in either K medium or PBS supplemented with 0.1% BSA containing protease inhibitors
aprotinin (2 ␮g/ml), PMSF (100 ␮g/ml), EDTA (5 mM), and iodoacetamide (20 ␮g/ml). Cells (2.5 ⫻ 106) were incubated in microfuge tubes
with a fixed concentration of HPLC-purified monoiodinated pMCMV
(125I1-pMCMV) and various concentrations of unlabeled peptides in a total
volume of 500 ␮l. After 2 h at 37°C, the cells were rapidly pelleted, washed
twice with chilled K medium, and transferred to tubes containing 400 ␮l of
oil (84% silicon oil, density 1.050; 16% paraffin oil, density 0.838, v/v) to
separate cell-bound and unbound 125I-peptide by centrifugation. Radioactivity in cell pellets and supernatants was measured in a gamma counter.
Nonspecific binding of 125I1-pMCMV, measured in the presence of a 500fold molar excess of unlabeled pMCMV, was subtracted to determine the
amount of radioactivity specifically bound to the cells.
Expression of p2Cb in Ld⫹ target cells renders the latter
susceptible to TCR-mediated lysis by 2C CTL
Since p2Cb is thought to be a longer natural precursor of p2Ca (5),
we investigated processing and presentation of p2Ca and p2Cb in
live cells using the minigene expression system. To reduce the
confounding effect of peptides from endogenous mouse OGDH,
we chose the human cell line C1R-Ld as a recipient target cell. It
has been previously shown that the synthetic human analog VAIT
RIEHVSPFPFDL of p2Cb was ⬎10 times less potent in the cytotoxicity assay, while the p2Ca human analog VSPFPFDL can be
presented by Ld to 2C CTL with similar efficacy (our unpublished
observations). Nevertheless, neither untransfected C1R-Ld cells
Downloaded from http://www.jimmunol.org/ by guest on June 12, 2017
The CD8⫹ CTL 2C from the original clone was maintained as described
previously (3). T2-Ld is a human cell line T2 that lacks the TAP transporter
genes and was transfected with the Ld gene (9). C1R, a mutant human cell
line that does not express HLA-A2, was derived from the B lymphoblastoid
cell line (10). C1R cells transfected with the gene for Ld are called C1R-Ld.
T2-Ld and C1R-Ld (gifts from P. Cresswell, New Haven, CT) were maintained in K medium (RPMI 1640 supplemented with 10% heat-inactivated
FCS, 10 mM HEPES, 2 mM L-glutamine, 100 U/ml penicillin, 100 ␮g/ml
streptomycin, and 50 ␮M 2-ME) in the presence of 250 and 400 ␮g/ml of
the antibiotic G418, respectively.
3029
3030
A FAMILY OF PEPTIDES INDUCES A STRONG ALLOGENEIC RESPONSE
Table I. Lysis of CIR-Ld cells expressing mouse OGDH peptides by 2C CTLa
Transfectants
Peptide Sequenceb
Specific
Lysisc
C1R-Ld(p2Ca)
C1R-Ld(p2Cb)
C1R-Ld(sig-p2Ca)
C1R-Ld(p2Ca-LLKEAQ)
C1R-Ld(p2Cb-LLKEAQ)
C1R-Ld(sig-p2Cb)
C1R-Ld(sig-p2Ca-LLKEAQ)
C1R-Ld
(met)LSPFPFDLd
VAITRIEQLSPFPFDL
Signal peptidee-LSPFPFDL
(met)LSPFPFDLLLKEAQ
VAITRIEQLSPFPFDLLLKEAQ
Signal peptide-VAITRIEQLSPFPFDL
Signal peptide-LSPFPFDLLLKEAQ
None
⫾
⫹⫹⫹
⫺
⫺
⫹
⫹
⫾
n.c.
The CTL assay was performed as described in Materials and Methods. Duration of the assay was 4 – 6 h; E:T ⫽ 25:1.
The amino acid sequence of the peptides expressed in C1R-Ld cells are shown. In some constructs, sequences of p2Ca (LSPFPFDL) and p2Cb (VAITRIEQLSPFPFDL) (both
underlined) were extended by six amino acids from C-terminal end matching the sequence of OGDH. Methionine may not be cleaved off from the N terminus of p2Ca and p2Ca-LLKEAQ
due to the fine specificity of methionine aminopeptidase; acetylation of the N-terminal methionine is unlikely to occur with these peptide sequences (36).
c
The level of specific lysis for various transfectants was: ⫺, 5%; ⫾, 8 –9; ⫹, 10%; ⫹⫹⫹, 25%; n.c., lysis of C1R-Ld in the absence of peptide was used as a negative control;
this did not exceed 6 ⫾ 2%. The comparison is based on the assumption that various minigenes are expressed at a similar level in the transfectants.
d
Cleavage of the N-terminal initiating methionine by methionine aminopeptidase in the cytoplasm of the transfected cells depends on the radius of gyration of the first amino acid and on the
nature of the third amino acid of the nascent peptide (36). Thus, most likely, endogenously generated p2Ca contained methionine at the N-terminal position and is referred to as (met)p2Ca.
e
RYMILGLLALAAVCSA(A) is the sequence of ER translocation signal peptide from adenovirus E3/19-kDa protein (15).
a
b
Alanine scan of QL9 reveals alternative Ld-binding motif
To determine which amino acid residues stabilize the QL9-Ld
complex, we systematically replaced with alanine the amino acid
residues in every position of QL9 and measured the binding of the
A low molecular mass fraction extracted from C1R-Ld(p2Cb)
cells contains peptides recognized by 2C CTL
To confirm that p2Cb was expressed and processed in the C1RLd(p2Cb) cells, we extracted peptides from these cells with 0.1%
TFA, separated them by HPLC, and measured the cytolytic activity of each fraction. As shown in Fig. 2, peptides in one broad peak
at a retention time of ⬃56 –58 min were effective in eliciting cytotoxicity. Retention times of synthetic peptides separated under
the same conditions were p2Ca and QL9 at 52 min, pyroQL9 (see
below) at 56 min, and p2Cb at 58 min. These data indicate that the
active peptide fraction isolated from C1R-Ld(p2Cb) most likely
contains p2Cb and/or pyroQL9, but not p2Ca or QL9.
QL9 binds to cell surface Ld with high affinity and is the most
potent OGDH peptide by cytotoxicity
The activities of synthetic OGDH peptides in CTL assays and the
apparent binding constants for the reaction of these peptides with
recombinant Ld protein have been previously determined (5, 7, 12,
18 –21). Although the affinities of p2Ca, p2Cb, and QL9 measured
here with live Ld⫹ cells were somewhat lower, QL9 still bound
more strongly to cell surface Ld than did p2Ca or p2Cb peptides
(see Fig. 3A and Table II). Also, the concentration of QL9 in the
extracellular medium required for half-maximal lysis (SD50) of
T2-Ld target cells was ⬃50- to 100- and 1000-fold lower than the
concentrations required for p2Ca and p2Cb, respectively (Fig. 3B
and Table II), consistent with the higher affinity of 2C TCR for
QL9-Ld than for p2Ca-Ld (7).
FIGURE 1. Specific lysis of C1R-Ld cells transfected with minigenes coding
for naturally occurring OGDH-derived peptides. A, 51Cr-labeled C1R-Ld transfectants were incubated with 2C CTL at the indicated E:T cell ratio for 4 h at 37°C.
The specific lysis was based on released 51Cr in the culture supernatant. Data of a
representative experiment are shown. The transfectants were C1R-Ld (p2Cb; F)
and C1R-Ld (p2Ca; 䡺). Untransfected C1R-Ld (E) were used as a control. Mean
values and error bars (SD) are based on triplicate determinations at every E:T cell
ratio. ⴱ, p ⬍ 0.05; ⴱⴱ, p ⬍ 0.01; ⴱⴱⴱ, p ⬍ 0.001 (determined from comparison of
the specific lysis of C1R-Ld (p2Cb) and C1R-Ld (p2Ca), by unpaired t test). B, The
clonotypic anti-TCR Ab (1B2) blocks lysis of C1R-Ld(p2Cb) by 2C CTL. 51Crlabeled C1R-Ld(p2Cb) cells were incubated with 2C CTL in the presence of the
indicated concentrations of 1B2 Ab for 4 h at 37°C. The E:T cell ratio was 25:1.
The dotted line indicates specific lysis in the absence of the Ab.
Downloaded from http://www.jimmunol.org/ by guest on June 12, 2017
nor the cells transfected with a minigene coding for human p2Cb
were lysed to a measurable extent by 2C CTL (data not shown). In
contrast, transfection of C1R-Ld cells with murine p2Cb rendered
these cells susceptible to specific lysis (Table I and Fig. 1A). Lysis
of C1R-Ld(p2Cb) cells by 2C CTL was TCR mediated, since the
clonotypic anti-TCR Ab 1B2, in a concentration-dependent manner, blocked the cytolysis (Fig. 1B). When the signal peptide was
introduced upstream of p2Cb to promote peptide translocation into
the ER (14), the specific lysis of C1R-Ld(signal-p2Cb) cells was
not enhanced. Extension of p2Cb from the C-terminal end by 6 aa
residues matching the murine OGDH sequence also led to less
efficient lysis of the target cells, indicating less efficient processing
and presentation of this longer peptide (Table I). C1R-Ld cells
transfected with the plasmid containing p2Ca were not, however,
lysed by 2C CTL.
The Journal of Immunology
3031
mono-substituted QL9 derivatives to Ld on T2-Ld cells. As indicated by the binding constant values in Table II, positions P7 and
P9 are critical for QL9 binding to Ld. These data are consistent
with the results of a similar analysis of the p2Ca-Ld reaction (22)
and with x-ray crystallographic analysis of the Ld protein (23).
They indicate that QL9, like p2Ca, uses an alternative Ld binding
motif (22, 24), depicted in bold in Table II.
PyroQL9 binds more strongly to Ld and exhibits higher activity
in the CTL assay than QL9
The reversible cyclization of N-terminal glutamine in QL9 to pyrrolidone carboxylate is not only catalyzed by glutamine cyclase
(QC) (25), but also occurs nonenzymatically, especially at low pH
(26). To prepare pyroQL9 we incubated freshly dissolved QL9 in
0.1% TFA followed by HPLC separation. QL9 and pyroQL9 were
eluted at 52 and 56 min, respectively, and their identities in the
eluted fractions were confirmed by amino acid analysis and Edman
degradation (data not shown). As is evident from Fig. 3 and Table
II, pyroQL9 was more active than QL9 in both binding to Ld and
cytotoxicity assays. Consistent with its higher activity in the cytolytic assay, pyroQL9 bound 3-fold more strongly to cell surface
Ld (Table I and Fig. 1).
QL9 and longer OGDH peptides with a conserved C terminus
are generated in vitro from p2Cb by purified 20S proteosomes
Because p2Cb and pyroQL9 were not resolved by HPLC under the
standard conditions used, we analyzed the products generated from
incubating purified p2Cb with 20S proteosomes by means of mass
spectrometry. From their molecular masses, several peptide fragments matching the C-terminal sequence of p2Cb could be identified (Fig. 4); these molecular masses correspond to TRIEQLSPF
PFDL (TL13), RIEQLSPFPFDL (RL12), IEQLSPFPFDL (IL11),
EQLSPFPFDL (EL10), and pyroQL9 (QL9 minus NH3; Fig. 4).
Interestingly, we could not detect a peptide with molecular mass
corresponding to p2Ca. The results indicate that proteolytic cleavage of p2Cb yields several fragments including QL9, which had
been converted to pyro-QL9 upon exposure to TFA during its
isolation.
Expression of QL9 in the endoplasmic reticulum of Ld⫹ target
cells leads to specific lysis of these cells by 2C CTL
To demonstrate that endogenously generated QL9 can be presented effectively on the cell surface in association with Ld, we
FIGURE 3. Presentation of OGDH peptides to 2C CTL. A, Peptide
binding to cell surface Ld. A total of 2.5 ⫻ 106 T2-Ld cells was combined
with 125I1-pMCMV (19.6 nM) and the unlabeled, noniodinated test peptide
at various concentrations in 500 ␮l of RPMI 1640 containing 0.1% BSA
and a mixture of protease inhibitors. After 2 h at 37°C, cell-bound and free
125
I1-MCMV were separated to measure cell-bound radioactivity. The
amount of bound reference peptide (cpm) corrected for nonspecific binding
is plotted as a function of the unlabeled peptide concentration. B, Percent
specific lysis of 51Cr-labeled T2-Ld cells in the presence of various concentrations of the indicated peptides (for peptide structures, see Table II).
The results of a representative experiment are shown. Arrows indicate the
peptide concentration required for half-maximal lysis (SD50).
transfected a minigene-coding signal-QL9 peptide into C1R-Ld⫹
and found that 2C CTL lysed the sig-QL9 transfectants (Fig. 5).
Because synthetic pyroQL9 and QL9 are both very active in cytotoxicity assays (see above), the low level of specific lysis observed in these experiments indicates that only a very small
amount of QL9 was available at the cell surface.
Discussion
Although the MHC moiety of an allo-MHC-peptide complex plays
a critical role in allogeneic T cell responses (27), the MHC-bound
peptide can also determine the specificity of a T cell-mediated
alloreaction (8, 28, 29). The octapeptide p2Ca and the 16-mer
p2Cb were the first naturally occurring peptides found to be critical
for recognition of allogeneic MHC class I molecules by CTL (5,
8). Since p2Cb contains the entire sequence of p2Ca, it was reasonable to ask whether it is the natural precursor of p2Ca (5). To
address this, we found that expression of p2Cb in Ld⫹ target cells
renders the latter susceptible to specific lysis by 2C CTL. However, HPLC separation of peptides extracted from these target cells
did not provide evidence for the presence of p2Ca. Instead, the
retention time of the active CTL assay fractions eluted by HPLC
indicated that these fractions contained p2Cb and/or pyroQL9. Because both peptides have very similar retention times, we simulated p2Cb processing in vitro with 20S proteosomes and analyzed
the fragmented peptide products by mass spectrometry. We found
several fragments with various molecular masses (Fig. 4), one of
Downloaded from http://www.jimmunol.org/ by guest on June 12, 2017
FIGURE 2. HPLC fractionation of peptides derived from C1R-Ld transfected with p2Cb minigene, CIR-Ld(p2Cb). Low molecular mass fraction
(⬍10 kDa) isolated from homogenized C1R-Ld(p2Cb) cells was fractionated by reversed-phase chromatography. Fractions collected at 1-min intervals were tested in the cytotoxicity assay using T2-Ld as target (see
Materials and Methods for details).
3032
A FAMILY OF PEPTIDES INDUCES A STRONG ALLOGENEIC RESPONSE
Table II. Binding of OGDH peptides and the alanine-substituted analogs
of QL9 analogs to Ld protein on live cells at 37°C and peptide
concentrations required for half-maximal lysis of Ld⫹ target cells by
2C CTLa
Peptideb
Name
Sequence
p2Ca
LSPFPFDL
p2Cb
VAITRIEQLSPFPFDL
QL9
QLSPFPFDL
pyro-QL9
pyro-QLSPFPFDL
MCMV
YPHFMPTNL
127
127
I-MCMV
I-YPHFMPTNL
7 ⫻ 10
9 ⫻ 104
4 ⫻ 107
1 ⫻ 108
3 ⫻ 108
4 ⫻ 108
5
ALSPFPFDL
4 ⫻ 104
QASPFPFDL
5 ⫻ 104
QLAPFPFDL
1 ⫻ 105
QLSAFPFDL
1 ⫻ 105
QLSPAPFDL
5.5 ⫻ 106
QLSPFAFDL
2 ⫻ 105
QLSPFPADL
⬍4 ⫻ 104
QLSPFPFAL
3 ⫻ 105
QLSPFPFDA
⬍4 ⫻ 104
. . .VAITRIEQLSPFPFDLLLKE. . .
OGDH
SD50 (nM)
1–0.1
200
0.005–0.03
0.001–0.01
Inactive
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
a
Clone 2C of CD8⫹ CTL is positively selected by Kb (syngeneic) class I MHC
but also recognizes an allogeneic MHC class I protein Ld in association with peptides
from OGDH (for review, see Ref. 27).
b
OGDH-derived peptides (5, 8) and their analogs are aligned to match the sequence of OGDH, ubiquitous intracellular protein (5). The alternative FDL motif for
Ld-binding peptides (24) is depicted in bold.
c
Binding of the indicated peptides to cell surface Ld was measured by competition
binding of the unlabeled peptides and radiolabeled MCMV to Ld on T2-Ld cells (12).
which (m/z 1046) matches the molecular mass of pyroQL9. Further
evidence suggesting that QL9 is naturally processed derives from
the finding that expression of this peptide in the ER results in
specific lysis of transfected cells by 2C CTL (Fig. 5). All of these
results indicate that QL9 could be naturally processed and contributes to the allogeneic response of 2C CTL against Ld⫹ target
cells. In accord with this conclusion, we found that QL9 is not only
the most potent OGDH peptide in the cytolytic assay, but that it
also binds more strongly to Ld on live cells than the other OGDHderived peptides (Table II).
Assuming that the transfected DNA is expressed equally well
within various transfectants, we note that the p2Cb minigene
downstream of the signal sequence resulted in much weaker sensitization of the target cells than the minigene without the signal
sequence (Table I). Although p2Cb has been shown to be readily
transported from the cytosol to the ER (30), it might be too long to
be effectively “trimmed” in the ER to produce an optimal peptide
epitope (31). It is possible that p2Cb could be retrotranslocated
from the ER to the cytosol to be shortened and then recycled back
to the ER (31). Such recycling would result in a lower yield of
optimal length peptide and could explain the inefficient lysis of
C1R-Ld(sig-p2Cb) target cells.
Almost complete lack of sensitization of the target cells carrying
the p2Ca minigene was perhaps surprising, since this is a naturally
processed peptide, and its synthetic analog readily renders Ld⫹
targets susceptible to specific lysis by 2C CTL (8). An explanation
for this apparent disparity could be that peptides of an optimal
length are produced in the ER rather than in the cytosol (32, 33).
Optimal length peptides lack the flanking residues that are thought
to be essential for their association with cytosolic chaperones that
facilitate TAP-dependent peptide translocation into the lumen of
the ER (33). The above considerations may also explain our failure
to detect presentation by transfected cells carrying the minigeneencoded QL9 without the signal sequence (data not shown).
FIGURE 4. MALDI-time-of-flight mass spectrum analysis of the mixture of peptides produced upon digestion of p2Cb with purified 20S proteosomes. Several peptides were detected, the molecular masses of which
can be assigned to contiguous segments of the p2Cb sequence, as discussed
in the text.
That the presumed expression of p2Ca in the ER of transfected
C1R-Ld cells did not render the cells susceptible to lysis by 2C
CTL (see Table I) is at odds with experiments demonstrating that
optimal peptides expressed in the ER are usually effectively presented on the cell surface (14, 15). A low copy number of this
minigene in the transfected cells may also account for the lack of
detectable peptide presentation. In addition, p2Ca forms very
short-lived complexes with Ld protein (half-life is ⬃5 min; I. Vturina, Y. Sukulev, and H. N. Eisen, unpublished observation), and
consequently, the epitope density (i.e., number of cognate pMHC
complexes per cell) required to elicit cytotoxicity would not be
achieved despite a continuous flux of p2Ca-Ld complexes to the
cell surface. These results also suggest that p2Ca may not be responsible alone for the strong allogeneic response of 2C CTL
against Ld⫹ targets cells (such as P815 cells) under physiological
conditions.
Why has QL9 not been identified in tissue extracts as a natural
peptide? We have shown here that the N-terminal glutamine of this
peptide is rapidly converted to pyroglutamine in TFA, which is
commonly used to extract peptides from tissues. This correlates
with the finding that one of the peptide fragments extracted with
FIGURE 5. 2C CTL specifically lyse Ld⫹ target cells transfected with a
minigene encoding sig-QL9. 2C CTL and 51Cr-labeled C1R-Ld cells transfected with the sig-QL9 minigene were incubated at various E:T cell ratios
for 4 h at 37°C. The experiment shown is representative of three independent titrations. ⴱⴱ, p ⬍ 0.01; ⴱⴱⴱ, p ⬍ 0.001 (determined from comparison
of the specific lysis of transfected and untransfected target cells, by unpaired t test).
Downloaded from http://www.jimmunol.org/ by guest on June 12, 2017
QL9-A1
QL9-A2
QL9-A3
QL9-A4
QL9-A5
QL9-A6
QL9-A7
QL9-A8
QL9-A9
Affinity (M⫺1)c
The Journal of Immunology
C-terminal sequence, these peptides display the same epitope to
the 2C TCR.
In summary, the data discussed above indicate that QL9 and
several longer OGDH-derived peptides might be naturally processed from p2Cb and presented on Ld⫹ target cells. Although
complexes of these peptides with the Ld protein bind the 2C TCR
with relatively high affinity, they are short-lived. Even the most
stable of these complexes, QL9-Ld, has a half-life of about 20 min
(20). We have previously shown that fewer than 10 QL9-Ld complexes/target cell can render these cells susceptible to specific lysis
by 2C CTL (20). However, the magnitude of the observed lysis
induced by a few cognate pMHC complexes is far lower than that
of the lysis of the Ld⫹ mouse mastocytoma cell line P815, which
is commonly used as an allogeneic target for 2C CTL (3, 8). To
date, only p2Ca was isolated from cell surface Ld molecules on
P815 cells (8). It is possible that the other OGDH peptides discussed here are also presented on the cell surface, but at much
lower densities, which would make it difficult to prove their identities by direct isolation. Based on these considerations, we submit
that QL9 and other OGDH peptides acting in concert with p2Ca
and p2Cb on the surface of the Ld⫹ target cells elevate the total
density of 2C-recognized epitopes and thus account for the exceptionally vigorous lysis of the target cells by T cells expressing the
2C TCR.
Acknowledgments
We thank Dr. Ike Eisenlohr for valuable discussions of this manuscript,
Prof. Klause Biemann for help with the interpretation of the mass spectrometry data, Mimi Rasmussen and Carol McKinley for excellent technical support, and Richard F. Cook and colleagues at the Massachusetts
Institute of Technology Biopolymers Laboratory for peptide synthesis and
amino acid analyses.
References
1. Goldberg, A. L., and K. L. Rock. 1992. Proteolysis, proteasomes and antigen
presentation. Nature 357:375.
2. Monaco, J. J. 1992. A molecular model of MHC class-I-restricted antigen processing. Immunol. Today 13:173.
3. Kranz, D., D. H. Sherman, M. V. Sitkovsky, M. S. Pasternack, and H. N. Eisen.
1984. Immunoprecipitation of cell surface structures of cloned cytotoxic T lymphocytes by clone-specific antisera. Proc. Natl. Acad. Sci. USA 81:573.
4. Wu, M. X., T. J. Tsomides, and H. N. Eisen. 1995. Tissue distribution of natural
peptides derived from a ubiquitous dehydrogenase, including a novel liver-specific peptide that demonstrates the pronounced specificity of low affinity T cell
reactions. J. Immunol. 154:4495.
5. Udaka, K., T. J. Tsomides, P. Walden, N. Fukusen, and H. N. Eisen. 1993. A
ubiquitous protein is the source of naturally occurring peptides that are recognized by a CD8⫹ T-cell clone. Proc. Natl. Acad. Sci. USA 90:11272.
6. Koike, K., Y. Urata, and S. Goto. 1992. Cloning and nucleotide sequence of the
cDNA encoding human 2-oxoglutarate dehydrogenase (lipoamide). Proc. Natl.
Acad. Sci. USA 89:1963.
7. Sykulev, Y., A. Brunmark, T. J. Tsomides, S. Kageyama, M. Jackson,
P. Peterson, and H. N. Eisen. 1994b. High-affinity reactions between antigenspecific T-cell receptors and peptides associated with allogeneic and syngeneic
major histocompatibility complex class I proteins. Proc. Natl. Acad. Sci. USA
91:14487.
8. Udaka, K., T. J. Tsomides, and H. N. Eisen. 1992. A naturally occurring peptide
recognized by alloreactive CD8⫹ cytotoxic T lymphocytes in association with a
class I MHC protein. Cell 69:989.
9. Alexander, J., J. A. Payne, J. A. Murrey, J. A. Frelinger, and P. Cresswell. 1989.
Differential transport requirements of HLA and H-2 class I glycoproteins. Immunogenetics 29:380.
10. Hogan, K. T., N. Shimojo, S. F. Walk, V. H. Engelhard, W. L. Maloy,
J. E. Coligan, and W. E. Biddison. 1988. Mutations in the ␣2 helix of HLA-A2
affect presentation but do not inhibit binding of influenza virus matrix peptide.
J. Exp. Med. 168:725.
11. Tsomides, T. J., and H. N. Eisen. 1993b. Stoichiometric labeling of peptides by
iodination on tyrosyl or histidyl residues. Anal. Biochem. 210:129.
12. Kageyama, S., T. J. Tsomides, Y. Sykulev, and H. N. Eisen. 1995. Variations in
the number of peptide-MHC class I complexes required to activate cytotoxic T
cell responses. J. Immunol. 154:567.
13. Gammon, M. C., M. A. Bednarek, W. E. Biddison, S. S. Bondy, J. D. Hermes,
G. E. Mark, A. R. Williamson, and H. J. Zweerink. 1992. Endogenous loading of
HLA-A2 molecules with an analog of the influenza virus matrix protein-derived
peptide and its inhibition by an exogenous peptide antagonist. J. Immunol. 148:7.
Downloaded from http://www.jimmunol.org/ by guest on June 12, 2017
TFA from the digest of p2Cb produced by purified 20S proteosomes has a molecular mass corresponding to pyroQL9 but not to
QL9. Since p2Cb and pyroQL9 were not resolved by HPLC under
the standard conditions used (see above), naturally produced QL9
in tissue extracts would not necessarily have been detected
chromatographically.
The conversion of N-terminal glutamine to pyroglutamine is
catalyzed by QC, which is present in various tissues, including
spleen (25). Whether QC converts the N-terminal glutamine of
QL9 to its cyclized form under physiological conditions is not
clear.
Besides the fragment with a molecular mass of pyroQL9 found
in the 20S proteosome digest, there were other fragments corresponding to OGDH peptides of various lengths, all sharing with
p2Cb the same C-terminal end: TRIEQLSPFPFDL (TL13),
RIEQLSPFPFDL (RL12), IEQLSPFPFDL (IL11), and EQLSPF
PFDL (EL10; see Fig. 4). That these peptides might also play a
role in the recognition of Ld⫹ target cells by 2C CTL is consistent
with our findings demonstrating that synthetic analogs of the various OGDH peptides, including p2Cb, bind soluble Ld protein to
form pMHC complexes that are recognized by 2C TCR on the
surface of live 2C cells (Ref. 21, and Y. Sykulev, A. Brunmark and
H. N. Eisen, unpublished observations) and that such peptides have
considerable activity in cytotoxicity assays. In support of this,
C1R-Ld cells sensitized with highly purified p2Cb induced a calcium flux in 2C CTL that was detectable within about 1 min; this
time is too short to be accounted for by processing of p2Cb into a
shorter active peptide fragment that could be responsible for the
observed response (5).
How are all of these multiple length peptides accommodated in
the Ld-binding groove to produce biologically active peptide-MHC
complexes? One explanation lies in the critical role of an alternative Ld-binding motif in the C-terminal sequence shared by all
active OGDH peptides (22, 24). We have shown that amino acids
at P9 and P7 are essential for the binding of QL9 to Ld (Table II).
A similar conclusion has been reached by others who found that
homologous amino acid residues (P8 and P6) are responsible for
the binding of p2Ca to Ld (22, 24). It is remarkable that even the
C-terminal tetrapeptide PFDL of p2Ca, but not the N-terminal
fragment LSPF, is still able to interact with Ld, albeit weakly, and
that the soluble PFDL-Ld complex binds to 2C TCR with
measurable affinity (Ref. 34, and Y. Sykulev, A. Brunmark, and
H. N. Eisen, unpublished observations). It seems that Ld-bound
OGDH peptides use the same register in the binding groove,
with the peptide’s C terminus anchored in the F pocket of the Ld
molecule (23).
Based on the crystal structure of the Ld protein (23, 35), Speir et
al. (23) built a model of the QL9-Ld complex. This model confirmed that the P9 and P7 amino acid residues of the peptide are
essential for stability of this peptide-MHC complex. The modeled
structure of QL9-Ld also explains why different OGDH peptides
sharing a common C terminus form Ld-peptide complexes that all
bind 2C TCR with relatively high affinities ranging from 105 to 107
M⫺1 (Ref. 21, and Y. Sykulev, A. Brunmark, and H. N. Eisen,
unpublished observations). The presence of bulky amino acids
Trp73, Trp97, and Tyr99 in a prominent ridge on the floor of the Ld
molecule binding groove forces bound peptides to bulge out toward the solvent. This peptide bulging increases the proximity of
the peptide’s C terminus to the 2C TCR surface, resulting in increased electrostatic interactions between the negatively charged
penultimate aspartic acid of the peptide and the positively charged
residues of the TCR ␤-chain (23). Because negatively charged Asp
in the penultimate position is present in all OGDH peptides sharing
3033
3034
A FAMILY OF PEPTIDES INDUCES A STRONG ALLOGENEIC RESPONSE
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
tamyl-peptides: presence in pituitary, brain, adrenal medulla, and lymphocytes.
J. Biol. Chem. 262:8532.
Orlowski, M., and A. Meister. 1971. Enzymology of Pyrrolidone Carboxylate
Acid. Academic Press, New York.
Eisen, H. N., Y. Sykulev, and T. J. Tsomides. 1997. The antigen-specific T-cell
receptor and its reactions with peptide-MHC complexes. In Advances in Protein
Chemistry, Vol. 49. E. Haber, ed. Academic Press, San Diego, p. 1.
Malarkannan, S., F. Gonzalez, V. Nguyen, G. Adair, and N. Shastri. 1996. Alloreactive CD8⫹ T cells can recognize unusual, rare, and unique processed peptide/MHC complexes. J. Immunol. 157:4464.
Obst, R., N. Netuschil, K. Klopfer, S. Stevanovic, and H. G. Rammensee. 2000.
The role of peptides in T cell alloreactivity is determined by self-major histocompatibility complex molecules. J. Exp. Med. 191:805.
Shepherd, J. C., T. N. M. Schumacher, P. G. Ashton-Rickardt, S. Imaeda,
H. L. Ploegh, C. A. Janeway, Jr., and S. Tonegawa. 1993. TAP1-dependent peptide translocation in vitro is ATP dependent and peptide selective. Cell 74:577.
Roelse, J., M. Gromme, F. Momburg, G. Hammerling, and J. Neefjes. 1994.
Trimming of TAP-translocated peptides in the endoplasmic reticulum and in the
cytosol during recycling. J. Exp. Med. 180:1591.
Snyder, H. L., J. W. Yewdell, and J. R. Bennink. 1994. Trimming of antigenic
peptides in an early secretory compartment. J. Exp. Med. 180:2389.
Paz, P., N. Brouwenstijn, R. Perry, and N. Shastri. 1999. Discrete proteolytic
intermediates in the MHC class I antigen processing pathway and MHC I-dependent peptide trimming in the ER. Immunity 11:241.
Gillanders, W., H. L. Hanson, R. J. Rubocki, T. H. Hansen, and J. M. Connolly.
1997. Class I-restricted cytotoxic T cell recognition of split peptide ligands. Int.
Immunol. 9:81.
Balendiran, G. K., J. C. Solheim, A. C. M. Young, T. H. Hansen,
S. G. Nathenson, and J. C. Sacchettini. 1997. The three-dimensional structure of
an H-2Ld-peptide complex explains the unique interaction of Ld with ␤2m and
peptide. Proc. Natl. Acad. Sci. USA 94:6880.
Moerschell, R. P., Y. Hosokawa, S. Tsunasawa, and F. Sherman. 1990. The
specificities of yeast methionine aminopeptidase and acetylation of amino-terminal methionine in vivo: processing of altered iso-1-cytochromes c created by
oligonucleotide transformation. J. Biol. Chem. 265:19638.
Downloaded from http://www.jimmunol.org/ by guest on June 12, 2017
14. Zweerink, H. J., M. C. Gammon, U. Utz, S. Y. Sauma, T. Harrer, J. C. Hawkins,
R. P. Johnson, A. Sirotina, J. D. Hermes, B. D. Walker, et al. 1993. Presentation
of endogenous peptides to MHC class I-restricted cytotoxic T lymphocytes in
transport deletion mutant T2 cells. J. Immunol. 150:1763.
15. Anderson, K., P. Cresswell, M. Gammon, J. Hermes, A. Williamson, and
H. Zweerink. 1991. Endogenously synthesized peptide with an endoplasmic reticulum signal sequence sensitizes antigen processing mutant cells to class I-restricted cell-mediated lysis. J. Exp. Med. 174:489.
16. Karas, M., and F. Hillenkamp. 1988. Laser desorption ionization of proteins with
molecular masses exceeding 10,000 daltons. Anal. Chem. 60:2299.
17. Papayannopoulos, I. A., and K. Biemann. 1992. Fast atom bombardment and
tandem mass spectrometry of synthetic peptides and byproducts. Pept. Res. 5:83.
18. Sykulev, Y., A. Brunmark, M. Jackson, R. J. Cohen, P. A. Peterson, and
H. N. Eisen. 1994. Kinetics and affinity of reactions between an antigen-specific
T-cell receptor and peptide-MHC complexes. Immunity 1:15.
19. Dutz, J. P., T. J. Tsomides, S. Kageyama, M. H. Rasmussen, and H. N. Eisen.
1994. A cytotoxic T lymphocyte clone can recognize the same naturally occurring self peptide in association with a self and a nonself class I MHC protein. Mol.
Immunol. 31:967.
20. Sykulev, Y., M. Joo, I. Vturina, T. J. Tsomides, and H. N. Eisen. 1996. Evidence
that a single peptide-MHC complex on a target cell can elicit a cytolytic T cell
response. Immunity 4:565.
21. Sykulev, Y., Y. Vugmeyster, A. Brunmark, H. Ploegh, and H. Eisen. 1998. Peptide antagonism and T cell receptor interactions with peptide-MHC complexes.
Immunity 9:475.
22. Al-Ramadi, B. K., M. T. Jelonek, L. F. Boyd, D. H. Margulies, and
A. L. Bothwell. 1995. Lack of strict correlation of functional sensitization with
the apparent affinity of MHC/peptide complexes for the TCR. J. Immunol. 155:
662.
23. Speir, J., K. Garcia, A. Brunmark, M. Degano, P. Peterson, L. Teyton, and
I. Wilson. 1998. Structural basis of 2C TCR allorecognition of H-2Ld peptide
complexes. Immunity 8:553.
24. Robinson, R. A., and D. R. Lee. 1996. Studies of tum-peptide analogs define an
alternative anchor that can be utilized by Ld ligands lacking the consensus P2
anchor. J. Immunol. 156:4266.
25. Busby, W. H., Jr., G. E. Quackenbush, J. Humm, W. W. Youngblood, and
J. S. Kizer. 1987. An enzyme(s) that converts glutaminyl-peptides into pyroglu-