Identification of tumor-associated, MHC class II-
restricted phosphopeptides as targets
Florence R. Depontieua,1, Jie Qianb,1,2, Angela L. Zarlingc, Tracee L. McMillera, Theresa M. Salaya, Andrew Norrisb,
A. Michelle Englishb, Jeffrey Shabanowitzb, Victor H. Engelhardc, Donald F. Huntb,d,3, and Suzanne L. Topaliana,3
aDepartment of Surgery, The Johns Hopkins University School of Medicine, Baltimore, MD 21231; and Departments ofbChemistry anddPathology,
and thecBeirne B. Carter Immunology Center and Department of Microbiology, University of Virginia, Charlottesville, VA 22901
Edited by Emil R. Unanue, Washington University School of Medicine, St. Louis, MO, and approved May 13, 2009 (received for review April 10, 2009)
The activation and recruitment of CD4?T cells are critical for the
development of efficient antitumor immunity and may allow for
the optimization of current cancer immunotherapy strategies.
Searching for more optimal and selective targets for CD4?T cells,
we have investigated phosphopeptides, a new category of tumor-
derived epitopes linked to proteins with vital cellular functions.
Although MHC I-restricted phosphopeptides have been identified,
it was previously unknown whether human MHC II molecules
present phosphopeptides for specific CD4?T cell recognition. We
first demonstrated the fine specificity of human CD4?T cells to
discriminate a phosphoresidue by using cells raised against the
candidate melanoma antigen mutant B-Raf or its phosphorylated
counterpart. Then, we assessed the presence and complexity of
human MHC II-associated phosphopeptides by analyzing 2 autol-
ogous pairs of melanoma and EBV-transformed B lymphoblastoid
lines. By using sequential affinity isolation, biochemical enrich-
ment, mass spectrometric sequencing, and comparative analysis, a
total of 175 HLA-DR-associated phosphopeptides were character-
ized. Many were derived from source proteins that may have roles
in cancer development, growth, and metastasis. Most were ex-
pressed exclusively by either melanomas or transformed B cells,
suggesting the potential to define cell type-specific phosphatome
‘‘fingerprints.’’ We then generated HLA-DR?1*0101-restricted CD4?
T cells specific for a phospho-MART-1 peptide identified in both
tide-pulsed antigen-presenting cells as well as for intact melanoma
cells. This previously undescribed demonstration of MHC II-restricted
phosphopeptides recognizable by human CD4?T cells provides po-
tential new targets for cancer immunotherapy.
tumor antigen ? tumor immunology
immunity in patients, as detected with in vitro immune moni-
toring, and yet they have had limited clinical success (1). One
reason may be the nature of the targeted antigens, the majority
of which are proteins overexpressed in tumor cells but not
essential to maintaining their malignant phenotype.
Phosphorylation is the most common and ubiquitous form of
enzyme-mediated posttranslational protein modification, and
transient phosphorylation of intracellular signaling molecules
regulates cellular activation and proliferation (2). Phosphoryla-
tion cascades are often dysregulated during malignant transfor-
tissues, and distant metastasis (3, 4). Limited but growing
evidence has shown that tumor-associated phosphoproteins pro-
cessed intracellularly through an endogenous pathway can give
rise to phosphopeptides complexed to MHC I molecules, which
are displayed on the cell surface (5, 6). CD8?T cells immunized
to specifically recognize these phosphopeptides are also capable
of recognizing intact human tumor cells, suggesting that phos-
phopeptides may represent a new class of targets for cancer
mmunotherapies directed against currently defined tumor-
associated or tumor-specific antigens can enhance antitumor
immunotherapy (5, 6). In these studies and others, T cell
discrimination of the phosphopeptide versus its nonphosphory-
lated counterpart was observed, indicating that phosphorylation
can influence peptide immunogenicity (5–13). Recent crystal
structural definition of phosphorylated peptide–HLA-A2 com-
plexes demonstrated direct and indirect interactions of the phos-
phoresidue with the MHC molecule, often significantly increasing
the affinity of the phosphopeptide for MHC I. Additionally,
direct interactions with the T cell receptor (14, 15).
Mounting evidence indicates that MHC II-restricted CD4?T
lymphocytes are a critical component of antitumor immunity,
and their activation and recruitment may be required to optimize
cancer immunotherapies (16, 17). A variety of posttranslational
modifications have been identified on naturally processed MHC
class II-associated epitopes. These include N- and O-linked
glycosylation, N-terminal acetylation, nitration, deamidation,
and deimination/citrullination (18). Although an early attempt
to detect phosphorylation on class II MHC peptides met with
failure (5), new technology has now made it possible to observe
this modification as well (19). Here, we demonstrate the exis-
tence of MHC II-associated phosphopeptides on human mela-
noma cells and EBV-transformed B (EBV-B) lymphoblasts, and
we define and compare the sequences of phosphopeptides
complexed to HLA-DR molecules on 2 autologous pairs of
melanoma and B cell cultures. Furthermore, we provide a
previously undescribed demonstration of the ability of human
CD4?T cells to specifically recognize phosphoepitopes dis-
played in the context of MHC II molecules by using the example
of an HLA-DR?1*0101-restricted phospho-melanoma antigen
recognized by T cells-1 (phospho-MART-1) peptide isolated
independently from 2 melanoma cell lines. These findings sug-
gest that tumor-associated phosphopeptides provide targets for
CD4?as well as CD8?T cells, potentially enabling the devel-
opment of new immunotherapeutic strategies.
Results and Discussion
T Cell Recognition of the Candidate Phosphopeptide B-Raf mutpT599.
To assess the potential for human CD4?T cells to specifically
discriminate phosphoresidues, we first explored the candidate
Author contributions: F.R.D., J.Q., A.L.Z., J.S., V.H.E., D.F.H., and S.L.T. designed research;
F.R.D., J.Q., A.L.Z., T.L.M., T.M.S., A.N., A.M.E., and J.S. performed research; F.R.D., J.Q.,
A.L.Z., T.L.M., T.M.S., A.N., A.M.E., J.S., V.H.E., D.F.H., and S.L.T. analyzed data; and F.R.D.,
J.Q., J.S., V.H.E., D.F.H., and S.L.T. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
1F.R.D. and J.Q. contributed equally to this work.
2Present address: Thermo Fisher Scientific, Somerset, NJ 08873.
This article contains supporting information online at www.pnas.org/cgi/content/full/
www.pnas.org?cgi?doi?10.1073?pnas.0903852106 PNAS ?
July 21, 2009 ?
vol. 106 ?
no. 29 ?
by 60% of melanomas (20). The somatic V600E mutation consti-
tutively activates the B-Raf serine-threonine kinase, and hence the
MAPK cascade, presumably by mimicking a phosphorylation
event. V600E is juxtaposed to the known dominant phosphosite
patient peripheral blood mononuclear cells (PBMCs) with a non-
phosphorylated, 29-mer candidate B-Raf mut peptide generated
B-Raf mut-specific HLA-DR?1*0404-restricted CD4?T cells that
did not cross-react with the wild-type peptide (B-Raf wt) and that
B-Raf mut (21). To investigate the fine specificity of MHC II-
restricted CD4?T cells for phosphate moieties, T cells raised
against nonphosphorylated B-Raf mut were assessed for recogni-
tion of synthetic B-Raf mutpT599. These T cells failed to react to the
phosphorylated peptide. Conversely, we used the monophospho-
rylated peptide B-Raf mutpT599to sensitize new CD4?T cell
cultures, which specifically recognized both B-Raf mutpT599and
B-Raf wtpT599but not their nonphosphorylated counterparts (Fig.
S1). By using anti-MHC blocking antibodies and HLA-matched or
mismatched allogeneic antigen-presenting cells (APCs), we char-
acterized HLA-DR11 as the restricting MHC allele for T cell
against synthetic B-Raf mutpT599failed to recognize melanoma
cells, suggesting that the hypothetical epitope might not be gener-
ated by intracellular processing, or that the conformation of the
conformation of a naturally processed epitope (22). Importantly,
however, these experiments provide a previously undescribed dem-
onstration that, similarly to MHC I-restricted CD8?T cells, human
CD4?T cells are capable of specifically recognizing phosphopep-
can be a critical determinant of recognition.
Identification and Characterization of HLA-DR-Associated Phos-
phopeptides. To identify naturally processed tumor-associated,
MHC II-restricted phosphopeptides as potential targets for im-
mune recognition, we affinity-isolated HLA-DR peptide com-
plexes from 2 cultured melanoma lines (1363-mel and 2048-mel)
and their autologous EBV-B cell counterparts (1363-EBV and
2048-EBV). These cell lines were selected because they constitu-
tively express significant levels of common HLA-DR molecules, as
assessed by flow cytometric analysis using the pan-DR mAb L243
(Fig. S2) and HLA allele-specific mAbs. By HLA genotyping,
in 31% of melanoma patients (23); DR?1*0404, found in 6.5% of
patients (23); and DR?4*0103. Notably, 1363-mel and 1363-EBV
contain a single DR molecule, HLA-DR?1*0101, affording an
opportunity to isolate phosphopeptides with unambiguous HLA
allele restriction. Patients 2048 and 1363 share HLA-DR?1*0101,
enabling the possibility of finding commonly expressed peptides on
cell lines from both patients.
A total of 175 phosphopeptides were sequenced from the 4 cell
lines (Table 1, Table S1, and Table S2). Of note, this analysis does
not account for peptides containing mutations, which would not be
detected by existing search algorithms (see Materials and Methods).
Twenty-three phosphopeptides were isolated from 2 or more cell
lines, yielding a total of 150 unique phosphopeptides. Similar to
nonphosphorylated MHC II-associated epitopes, the average
length of the phosphorylated peptides was 16 aa (range, 8–28 aa)
(24). Also characteristic of MHC II epitopes, 78% (117 of 150
sequences) were found within nested sets, defined as groups of
peptides sharing core sequences but having distinct N and C
termini. Most phosphopeptides were specifically expressed by ei-
ther melanomas (Table 1) or EBV-B cells (Table S1), although
some were expressed by both cell types (Table S2). Only 23
phosphopeptide sequences from 7 source proteins were identified
from 1363-mel, whereas a larger number of phosphopeptide se-
2048-mel cells. The smaller number of sequences isolated from
1363-mel likely reflects the significantly lower expression of
HLA-DR molecules by these cells, as well as their expression of a
single DR allele (Fig. S2). Thirty-nine phosphopeptide sequences
from 15 proteins, and 48 phosphopeptides from 20 proteins, were
identified from 1363-EBV and 2048-EBV B cells, respectively. As
might be anticipated because patients 1363 and 2048 share the
HLA-DR?1*0101 allele, phosphopeptides common to both pa-
tients occurred, including those found in both melanomas
(MART-1 and tensin-3; Table 1) or in both EBV-B cell lines
(CD20; Table S1). Experiments are in progress with melanomas
and EBV-B cells generated from other patients to better define
cell type-specific phosphatome ‘‘fingerprints’’ and potential im-
munotherapeutic targets specific for melanomas or EBV-
Phosphosites were assigned unambiguously for 96% of the 175
phosphopeptides sequenced in this study. Phosphopeptides con-
tained only 1 phosphorylated residue, with the exception of a
phosphopeptide derived from frizzled 6 in 1363-mel, which
contained 2 phosphoresidues (Table S2). Among a total of 57
defined phosphosites (accounting for redundancy in nested
peptide sets), phosphate moieties were bound to serine, threo-
nine, or tyrosine residues in 93.0%, 5.3%, and 1.7% of cases,
respectively. Interestingly, these frequencies are similar to those
found in the HeLa cell-derived phosphoproteome (86.4%,
11.8%, and 1.8%, respectively; ref. 2), suggesting that there is no
significant bias in the processing or MHC II binding of peptides
containing a particular phosphoresidue. Of note, studies have
not identified MHC I-associated phosphopeptides containing
phosphotyrosine residues; this may be explained by the generally
low frequency of this posttranslational modification, as well as by
the relatively small number of phosphopeptides isolated in our
earlier study compared with the current report (36 vs. 175
phosphopeptides, respectively; ref. 6). A total of 60% (32 of 53)
of the source proteins for the phosphopeptides described in this
report are known to be phosphorylated (Table S3). However,
only 17 (29.8%) of the 57 defined phosphosites had been
identified previously (Table 1, Table S1, and Table S2). When
analyzed with an algorithm based on experimentally verified
phosphorylation sites in eukaryotic proteins (NetPhos 2.0) (25),
80% (32 of 40) of the previously unknown phosphosites identi-
fied here are highly predicted to be true phosphorylation sites.
The 150 unique phosphopeptides listed in Table 1, Table S1, and
Table S2 are derived from a total of 53 different protein sources
representing all cellular compartments, although transmembrane
peptides processed through the endosomal/lysosomal pathway. For
those proteins located in the plasma membrane, the isolated
phosphopeptides emanate from the cytoplasmic tail region. The
processing of cytosolic and nuclear proteins via the MHC II
pathway does not fit the ‘‘classical’’ model of antigen processing.
However, mounting evidence suggests that autophagy, a stress-
activated process operational in intracellular protein turnover, may
play a critical role in shunting cytoplasmic proteins into the lyso-
somal compartment, thus influencing the MHC II–peptide reper-
toire (26, 27). Importantly, the majority of source proteins for the
phosphopeptides found in this study are known to support vital
biological functions, such as metabolism, cell cycle regulation, and
intracellular signaling, and they may have important roles in cancer
development, growth, and metastasis (Table S3). Thus, the deriv-
Analyzing the abundance of the isolated phosphopeptides re-
1363-mel, small acidic protein and tensin-3 in 2048-mel, and CD20
in 2048-EBV were present at 51–140 copies per cell). In fact, many
phosphopeptides were found to be expressed at less than 6 copies
www.pnas.org?cgi?doi?10.1073?pnas.0903852106 Depontieu et al.
per cell (Table 1, Table S1, and Table S2). This highlights the
exquisite sensitivity of the mass spectrometric methods used for
detection. Because T cell responses may be activated by fewer than
10 peptide–MHC complexes per cell, the phosphopeptides de-
scribed in this report are potentially immunogenic (28–30).
Specific CD4?T Cell Recognition of Phospho-MART-1. To assess the
ability of human CD4?T cells to specifically recognize tumor-
associated phosphopeptides, we selected the MART-1100–111
phosphopeptide (pMART-1, containing pS108) for further
study. As shown in Table 1, a nested set of phosphopeptides
derived from the C terminus of MART-1 was eluted from both
1363-mel and 2048-mel—which share HLA-DR?1*0101—but
not from the autologous EBV-B cell lines. Because of its
selective expression pattern in cells of the melanocytic lineage,
including normal melanocytes and melanoma cells, MART-1
Table 1. Characteristics of HLA-DR-associated phosphopeptides selectively expressed by melanoma cells
1363-mel and 2048-mel
Melanoma antigen recognized by
Matrix-remodeling-associated protein 7
Amino-terminal enhancer of split
Ankyrin repeat domain-containing
AP-3 complex subunit-?-1
Casein kinase II subunit-?
Insulin receptor substrate 2
Interleukin 1 receptor accessory protein
LUC7-like isoform bN
receptor component 1
NF-?B inhibitor-interacting Ras-like
Probable fibrosin-1 long-transcript
protein isoform 2
Small acidic protein
Transmembrane protein 184CPM/C
Protein sources were determined by searching peptide sequences against the nr and refseq databases for human proteins (www.ncbi.nlm.nih.gov/BLAST). ?,
not detected; ?, ?6 copies per cell; ??, 6–50 copies per cell; ???, 51–140 copies per cell; and nm, not measured.
and UK, Unknown.
†pS, pT, and pY correspond to serine, threonine, or tyrosine-associated phosphorylated residues, respectively. Italics indicate the exact site of phosphorylation
could not be determined.
‡Phosphosites searched in the Phospho-ELM database (41). Values in parentheses show the score from the NetPhos 2.0 Server (22) indicating the probability for
the site to be phosphorylated, scale 0–1.0. Higher scores indicate greater confidence in the prediction, with the designated binding threshold at 0.500. In cases
of undefined phosphorylation sites (italicized), both scores are given.
§Abundance of peptide containing Met and Metox, respectively.
Depontieu et al.PNAS ?
July 21, 2009 ?
vol. 106 ?
no. 29 ?
(also termed Melan-A) is an important target of immunother-
apeutic approaches for the treatment of melanoma, including
vaccines and adoptive T cell transfer (31, 32). This transmem-
brane protein, which is localized to melanosomes and functions
to regulate mammalian pigmentation (33), was not known to be
phosphorylated before the current report. However, several
nonphosphorylated MHC class I- and class II-restricted immu-
nogenic epitopes have been identified in MART-1, spanning the
entire protein (34–38). Some have provided the basis for syn-
thetic melanoma peptide vaccines.
Fresh PBMCs from melanoma patient D7, with the MHC II
genotype HLA-DR?1*01, HLA-DQ?1*0501, were repeatedly
stimulated in vitro with the pMART-1 peptide under microculture
conditions. After several rounds of stimulation, the CD4?micro-
culture designated D7-pM specifically secreted IFN-? (Fig. 1) and
GM-CSF in response to the pMART-1 peptide, but not to the
nonphosphorylated MART-1 peptide pulsed onto DR1-expressing
APCs. In titration experiments, pMART-1 could be recognized at
concentrations ?1 ?M. As shown in Fig. 1, the tensin-31434–1445
phosphopeptide (pTensin-3), also eluted from 1363-mel, was not
recognized by D7-pM T cells, nor were other peptides having high
affinity for HLA-DR1, including TPImut and HA. Conversely,
CD4?T cells specific for TPImut or HA secreted cytokines in
response to their cognate epitopes but were not stimulated by
pMART-1. These data suggest that CD4?D7-pM T cells specifi-
cally recognize pMART-1 and that the phosphorylated serine
residue is a critical determinant of recognition.
A nested set of 4 MART-1 phosphopeptides was eluted from
1363-mel and 2048-mel, sharing HLA-DR?1*0101 (Table 1).
Because phosphopeptides complexed to MHC molecules were
affinity-eluted on an anti-HLA-DR column, and because HLA-
DR?1*0101 is the only DR molecule contained in 1363-mel by
genotyping, we hypothesized that HLA-DR?1*0101 was the
restriction element for pMART-1 recognition by CD4?D7-pM
T cells. This was investigated with 2 complementary approaches:
anti-MHC mAbs were used to inhibit T cell recognition of
peptide-pulsed APCs, and a panel of allogeneic APCs with
diverse HLA types was used to present pMART-1 to T cells (Fig.
2). T cell recognition of DR1?2048-EBV pulsed with pMART-1
was significantly inhibited by an mAb directed against HLA-DR
(L243) but not by mAbs against HLA-DQ, HLA-DP, or class I
MHC (Fig. 2A). Furthermore, D7-pM T cells specifically rec-
EBV-B cells sharing HLA-DR?1*0101 (Fig. 2B). Next, we
sought to determine whether D7-pM T cells could also recognize
whole melanoma cells expressing MART-1 and HLA-
DR?1*0101. As shown in Fig. 3, D7-pM T cells recognized all
of the 3 allogeneic melanomas tested that expressed both
HLA-DR?1*0101 and MART-1, but they failed to recognize
DR1-negative or MART-1-negative cells. In the same experi-
ment, DR1-restricted CD4?tumor infiltrating lymphocytes
(TIL) 1558, specific for the TPImut antigen unique to 1558-mel,
recognized 1558-mel but not the other tumors (39). Two repeat
experiments yielded similar results. HLA-DR restriction of
tumor recognition by T cells was confirmed with anti-MHC mAb
blockade. Of interest, although D7-pM CD4?T cells specifically
reacted against intact DR1?MART-1?tumor cells, they failed
to recognize DR1?EBV-B cells pulsed with lysates of
MART-1?tumors. In the same experiment, CD4?TIL 1558
recognized processed 1558-mel lysate and provided a positive
control for exogenous pathway processing (39). Thus, although
in these preliminary experiments. Future work will address the
mechanism by which pMART-1 is processed intracellularly for
presentation in the context of MHC class II molecules.
In summary, our demonstration of phosphopeptides associ-
ated with human MHC II molecules has revealed a new cohort
phopeptide pMART-1. Peptide recognition by D7-pM T cells raised against
pMART-1 (Top) is compared to recognition by CD4?HA-specific cells from the
same patient (Middle) or by TPImut-specific CD4?TIL1558 (Bottom). Peptides
were pulsed onto HLA-DR1?2048-EBV cells. Similar results were obtained by
using DR1?1363-EBV as APC.
CD4?T cells specifically recognize the melanoma-associated phos-
pMART-1 peptide. (A) An anti-HLA-DR mAb inhibits T cell recognition of
allogeneic DR1?APC. T cell IFN-? secretion measured by ELISA. †, Genotype;
EBV-B cell expression of HLA-DR1 was confirmed by flow cytometry. ‡, Stim-
ulation index (SI), ratio of IFN-? secretion in response to pMART-1 versus the
irrelevant HA peptide. SI ?2 is considered significant.
HLA-DR1?1*0101 restriction of D7-pM CD4?T cells specific for the
intact melanoma cells expressing MART-1 and HLA-DR?1*0101. In compari-
son, DR1-restricted CD4?TIL 1558 cells are specific for autologous melanoma
cells expressing the unique tumor antigen TPImut (39). HLA-DR1 expression
was determined by flow cytometry. MART-1 expression was determined by
(32). ND, not done.
D7-pM T cells specific for pMART-1 peptide recognize allogeneic
www.pnas.org?cgi?doi?10.1073?pnas.0903852106Depontieu et al.
and therefore targetable by cancer immunotherapies. Although
tumor-associated, MHC class II-restricted phosphopeptides
were previously thought to be degraded within the MHC II
processing environment, recent technological advances have
enabled us to identify a large number of phosphoepitopes from
both melanoma and EBV-B cell lines. These may provide
optimal targets for immunotherapy because intracellular phos-
phoproteins associated with dysregulated signaling pathways
play an important role in supporting the malignant cell pheno-
type and in providing escape mechanisms from antineoplastic
agents. MART-1, a commonly expressed melanoma antigen
containing MHC I- and MHC II-restricted epitopes, has been a
focus of cancer immunotherapeutics for more than a decade and
has now been revealed as a phosphoprotein recognizable by
phosphopeptide-specific CD4?T cells derived from a melanoma
patient. Similarly, 2 phosphoproteins shown here to be sources for
MHC II-restricted peptides—tensin-3 and insulin receptor sub-
phosphopeptides (6). These findings suggest opportunities for
potent, polyvalent tumor-specific immunity.
Materials and Methods
in SI Materials and Methods. Human tissues were obtained through protocols
approved by the Institutional Review Boards of the National Cancer Institute,
Medicine. HLA genotypes of patients and cultured cell lines were determined
by the NIH Warren Grant Magnuson Clinical Center HLA Laboratory (Be-
thesda, MD) by using sequence-specific PCR techniques.
Isolation of HLA-DR-Associated Peptides. To prepare cells for extraction of
MHC–peptide complexes, growing cultures were harvested, and spent me-
dry cell pellets were snap frozen on dry ice and stored at ?80 °C for subse-
quent lysis. HLA-DR–peptide complexes were immunoaffinity-purified from
melanoma and EBV-B cells, and the associated peptides were extracted ac-
cording to methods for MHC I-associated peptides (6), with minor modifica-
tions (SI Materials and Methods).
Phosphopeptide Enrichment. Immunoaffinity-purified peptides were con-
A phosphopeptide standard, angiotensin II phosphate (100 fmol), was spiked
into all samples before the esterification step to monitor the efficiency of the
Phosphopeptide Sequence Analysis by Tandem Mass Spectrometry. Phos-
phopeptides were analyzed by nanoflow HPLC-microelectrospray ionization
coupled to either a hybrid linear quadrupole ion trap-Fourier-transform ion
cyclotron resonance (LTQ-FT) mass spectrometer (Thermo Fisher Scientific) or
an LTQ mass spectrometer (Thermo Fisher Scientific) modified to perform
was connected with polytetrafluoroethylene tubing (0.06-inch o.d. and 0.012-
inch i.d.; Zeus Industrial Products) to the end of an analytical HPLC column
(360-?m o.d. and 50 ?m i.d.) containing 7 cm of C18 reverse-phase packing
material (5-?m particles; YMC). Phosphopeptides were eluted to the mass spec-
trometer at a flow rate of 60 nL/min with a gradient: A ? 0.1 M acetic acid
in H2O; 0–60% B in 20 min, 60–100% B in 5 min. Parameters used to acquire
ETD/MS/MS spectra in a data-dependent mode on the modified LTQ instrument
have been described previously (40).
the 175 sequences reported here were unambiguous. None of the sequences
were assigned by software. Approximate copy/cell numbers for each phos-
phopeptide were determined by comparing peak areas of the observed
parent ions to that of angiotensin II phosphate (DRV[pY]IHPF, 100 fmol;
immobilized metal affinity chromatography. Specific functional and intracel-
lular localization information for source proteins was determined from the
phopeptides were searched in the Phospho-ELM database (http://phospho.
elm.eu.org) (41) to determine whether the phosphoresidue was previously
described or novel. Phosphopeptide sequences were analyzed in the NetPhos
2.0 Server (www.cbs.dtu.dk/services/NetPhos) (25), to obtain a prediction
score indicating the confidence in occurrence of the phosphorylation site.
Generation of Human Phosphopeptide-Specific CD4?T cells. Peptide-specific
CD4?T cells were raised by repetitive in vitro stimulation of PBMCs from
melanoma patients, as described previously (21). For candidate B-Raf phos-
phopeptides, patients were selected whose melanomas harbored the com-
mon T1799A (V600E) mutation. For pMART-1, selected patients expressed the
HLA-DR?1*0101 allele. Briefly, PBMCs were cultured in flat-bottom 96-well
plates at 2 ? 105cells per well in RPMI 1640 plus 10% heat-inactivated human
AB serum. GM-CSF (200 units/mL) and IL-4 (100 units/mL; PeproTech) were
Recombinant IL-2 (120 IU/mL) was added to lymphocyte cultures on day 7 and
added at 25 ng/mL at day 7 and were replenished every 4–7 days. Thereafter,
T cells were restimulated every 10–14 days with irradiated autologous PBMCs
or EBV-B cells pulsed with phosphopeptide, 1 ? 105feeder cells per well.
Long-term CD4?T cell cultures were maintained in 120 IU/mL IL-2 and 20%
conditioned medium from lymphokine-activated killer cell cultures (B-Raf-
specific T cells) or 120 IU/mL IL-2, 25 ng/mL IL-7, and 25 ng/mL IL-15 (pMART-
a unique HLA-DR1-restricted mutant TPI epitope, were used as controls in
some experiments (39).
T Cell Recognition Assays. To assess specific peptide recognition, 0.2 ? 105to
1.0 ? 105T cells per well were cocultured overnight in flat-bottom 96-well
plates with 1 ? 105EBV-B cells that had been prepulsed for 16–24 h with
peptides at 20 ?M (pB-Raf experiments) or 25 ?M (pMART-1 experiments).
Culture supernatants were then harvested, and GM-CSF or IFN-? secretion by
Systems). When allogeneic EBV-B cells were used as APCs to determine the
MHC restriction of peptide-specific T cells, excess peptide was washed off
before combining APCs with T cells. Intact tumor cell recognition was tested
(1 ? 105per well) for 20 h. In some assays, mAbs directed against MHC
molecules were used to inhibit T cell reactivity, including W6/32 (IgG2a,
American Type Culture Collection), B7/21 (IgG1, anti-HLA-DP; Becton Dickin-
son), and SPVL3 (IgG2a, anti-HLA-DQ; Beckman Coulter). Final concentrations
of mAb in blocking assays were 2.5 ?g/mL for B7/21 and 20 ?g/mL for the
others. Surface expression of HLA-DR1 on tumor and EBV-B cells was con-
firmed by flow cytometry after staining with biotinylated anti-DR1, 10, 103
(One Lambda) and counterstaining with streptavidin-phycoerythrin.
Note Added in Proof. We have recently become aware of work by Meyer VS et
al. (43) identifying MHC II-associated phosphopeptides from a human mela-
noma line and a B lymphoblastoid line.
ACKNOWLEDGMENTS. We thank Dr. Drew Pardoll for his critical review of the
at The Johns Hopkins University School of Medicine (F.R.D., T.L.M., T.M.S., and
1. Guinn BA, et al. (2007) Recent advances and current challenges in tumor immunology
and immunotherapy. Mol Ther 15:1065–1071.
2. Olsen JV, et al. (2006) Global, in vivo, and site-specific phosphorylation dynamics in
signaling networks. Cell 127:635–648.
3. Haluska FG, et al. (2006) Genetic alterations in signaling pathways in melanoma. Clin
Cancer Res 12:2301s–2307s.
4. Oka M, Kikkawa U (2005) Protein kinase C in melanoma. Cancer Metastasis Rev
major histocompatibility complex class I molecules in vivo. J Exp Med 192:1755–1762.
6. Zarling AL, et al. (2006) Identification of class I MHC-associated phosphopeptides as
targets for cancer immunotherapy. Proc Natl Acad Sci USA 103:14889–14894.
Depontieu et al.PNAS ?
July 21, 2009 ?
vol. 106 ?
no. 29 ?
7. Andersen MH, et al. (1999) Phosphorylated peptides can be transported by TAP Download full-text
molecules, presented by class I MHC molecules, and recognized by phosphopeptide-
specific CTL. J Immunol 163:3812–3818.
carcinoma of the lung-specific cytotoxic T lymphocytes is derived from a mutated
elongation factor 2 gene. Cancer Res 58:5144–5150.
9. Larson JK, Otvos L, Jr, Ertl HC (1992) Posttranslational side chain modification of a viral
epitope results in diminished recognition by specific T cells. J Virol 66:3996–4002.
therapeutic effect of a phosphorylated synthetic peptide of the 70K snRNP protein
administered in MR/lpr mice. Eur J Immunol 33:287–296.
stimulatory activity, serum stability and conformation of synthetic peptides carrying T
helper cell epitopes. Biochim Biophys Acta 1313:11–19.
12. van Stipdonk MJ, et al. (1998) T cells discriminate between differentially phosphory-
lated forms of alphaB-crystallin, a major central nervous system myelin antigen. Int
13. Yadav R, et al. (2003) The H4b minor histocompatibility antigen is caused by a
combination of genetically determined and posttranslational modifications. J Immu-
peptides and MHC class I: A molecular basis for the presentation of transformed self.
Nat Immunol 9:1236–1243.
15. Petersen J, et al. (2009) Phosphorylated self-peptides alter human leukocyte antigen
class I-restricted antigen presentation and generate tumor-specific epitopes. Proc Natl
Acad Sci USA 106:2776–2781.
16. Gerloni M, Zanetti M (2005) CD4 T cells in tumor immunity. Springer Semin Immuno-
17. Pardoll DM, Topalian SL (1998) The role of CD4? T cell responses in antitumor
immunity. Curr Opin Immunol 10:588–594.
18. Engelhard VH, Altrich-Vanlith M, Ostankovitch M, Zarling AL (2006) Post-translational
modifications of naturally processed MHC-binding epitopes. Curr Opin Immunol
19. Depontieu F, et al. (2008) Tumor-associated MHC II-restricted phosphopeptides: New
targets for immune recognition. FASEB J 22:1079.1071.
20. Davies H, et al. (2002) Mutations of the BRAF gene in human cancer. Nature 417:949–
of mutated B-RAF in melanoma patients harboring the V599E mutation. Cancer Res
22. Viner NJ, Nelson CA, Deck B, Unanue ER (1996) Complexes generated by the binding
of free peptides to class II MHC molecules are antigenically diverse compared with
those generated by intracellular processing. J Immunol 156:2365–2368.
23. Marincola F, Stroncek D, Simonis T (1998) in HLA 1998, eds Gjertson DW, Terasaki PI
(American Society for Histocompatibility and Immunogenetics, Lenexa, KS), pp 276–
24. Lippolis JD, et al. (2002) Analysis of MHC class II antigen processing by quantitation of
peptides that constitute nested sets. J Immunol 169:5089–5097.
25. Blom N, Gammeltoft S, Brunak S (1999) Sequence and structure-based prediction of
eukaryotic protein phosphorylation sites. J Mol Biol 294:1351–1362.
26. Nimmerjahn F, et al. (2003) Major histocompatibility complex class II-restricted pre-
sentation of a cytosolic antigen by autophagy. Eur J Immunol 33:1250–1259.
intracellular source proteins. Proc Natl Acad Sci USA 102:7922–7927.
28. Kageyama S, Tsomides TJ, Sykulev Y, Eisen HN (1995) Variations in the number of
peptide-MHC class I complexes required to activate cytotoxic T cell responses. J Immu-
29. Irvine DJ, Purbhoo MA, Krogsgaard M, Davis MM (2002) Direct observation of ligand
recognition by T cells. Nature 419:845–849.
30. Engelhard VH, Brickner AG, Zarling AL (2002) Insights into antigen processing gained
by direct analysis of the naturally processed class I MHC associated peptide repertoire.
Mol Immunol 39:127–137.
31. Coulie PG, et al. (1994) A new gene coding for a differentiation antigen recognized by
autologous cytolytic T lymphocytes on HLA-A2 melanomas. J Exp Med 180:35–42.
32. Kawakami Y, et al. (1994) Cloning of the gene coding for a shared human melanoma
antigen recognized by autologous T cells infiltrating into tumor. Proc Natl Acad Sci
33. Hoashi T, et al. (2005) MART-1 is required for the function of the melanosomal matrix
protein PMEL17/GP100 and the maturation of melanosomes. J Biol Chem 280:14006–
34. Bioley G, et al. (2006) Melan-A/MART-1-specific CD4 T cells in melanoma patients:
Identification of new epitopes and ex vivo visualization of specific T cells by MHC class
II tetramers. J Immunol 177:6769–6779.
by frequently expressed MHC class II alleles. Clin Immunol 121:54–62.
36. Larrieu P, Ouisse LH, Guilloux Y, Jotereau F, Fonteneau JF (2007) A HLA-DQ5 restricted
Cancer Immunol Immunother 56:1565–1575.
37. Marincola FM, Rivoltini L, Salgaller ML, Player M, Rosenberg SA (1996) Differential
comparison to healthy donors: Evidence of in vivo priming by tumor cells. J Immu-
nother Emphasis Tumor Immunol 19:266–277.
38. Zarour HM, et al. (2000) Melan-A/MART-1(51–73) represents an immunogenic HLA-
DR4-restricted epitope recognized by melanoma-reactive CD4(?) T cells. Proc Natl
Acad Sci USA 97:400–405.
39. Pieper R, et al. (1999) Biochemical identification of a mutated human melanoma
antigen recognized by CD4(?) T cells. J Exp Med 189:757–766.
40. Chi A, et al. (2007) Analysis of phosphorylation sites on proteins from Saccharomyces
41. Diella F, Gould CM, Chica C, Via A, Gibson TJ (2008) Phospho.ELM: A database of
phosphorylation sites–update 2008. Nucleic Acids Res 36:D240–D244.
42. Kawakami Y, et al. (1997) Production of recombinant MART-1 proteins and specific
antiMART-1 polyclonal and monoclonal antibodies: Use in the characterization of the
human melanoma antigen MART-1. J Immunol Methods 202:13–25.
43. Meyer VS, et al. (May 5, 2009) Identification of natural MHC class II presented phos-
www.pnas.org?cgi?doi?10.1073?pnas.0903852106 Depontieu et al.