CTL recognize short peptide fragments presented by class I MHC molecules. In this study, we examined the effect of phosphorylation on TAP transport, binding to class I MHC molecules, and recognition by CTL of peptide fragments from known phosphorylated oncogene proteins or virus phosphoproteins. We show that phosphopeptides can be efficiently transported from the cytosol to the endoplasmic reticulum by the TAP. Furthermore, we show that phosphorylation can have a neutral, negative, or even a positive effect on peptide binding to class I MHC. Finally, we have generated phosphopeptide-specific CTL that discriminate between the phosphorylated and the nonphosphorylated versions of the peptide. We conclude that phosphopeptide-specific CTL responses are likely to constitute a subset of the class I MHC-restricted CTL repertoire in vivo.

Major histocompatibility complex class I molecules present cytosolic peptide Ags for recognition by CTL. Peptide fragments are generated in the cytosol by proteasome-mediated proteolytic cleavage of normal endogenous proteins, as well as of viral proteins, and products of endogenous genes that have undergone molecular alterations during malignant transformation. Any single class I MHC allele has a peptide repertoire dominated by oligopeptides of 8–10 amino acids length, which share a sequence motif corresponding to two or more essential amino acid side chains, the so-called peptide anchor residues (1, 2, 3). The solution of x-ray crystallographic structures of MHC class I-peptide complexes (4, 5, 6) revealed that depressions or pockets in the MHC peptide-binding groove preferentially accommodate particular amino acids corresponding to the anchor residues of the peptide ligand.

Little is known about the effect on class I MHC-restricted Ag presentation of naturally occurring posttranslational modifications of peptide Ags in vivo. These modifications may affect Ag processing, MHC binding, and interaction with the TCR. Evidence that posttranslational modifications may play a role in Ag presentation and T cell recognition has been accumulating over the last few years. Recently, it has been described that class I (7, 8, 9, 10) and class II (11) MHC-restricted T cells can recognize posttranslationally modified peptides specifically (12, 13).

Phosphorylation is a tightly regulated posttranslational cytosolic event. However, inflammation, intracellular infection, cellular activation, and malignant transformation may cause deregulation of phosphorylation, perhaps leading to generation of sufficient amounts of phosphopeptides for class I MHC-restricted presentation to occur. For example, it is known that following chromosomal translocation in leukemia, certain protooncogenes acquire novel kinase activities leading to the appearance of new transformation-specific phosphorylation (14, 15). Similarly, novel phosphopeptide Ags could be derived from phosphorylated oncogenes exhibiting deregulated phosphorylation upon malignant transformation. Such proteins are phosphorylated on tyrosine, serine, and threonine residues by different cellular kinases, and up-regulation of phosphorylation at specific sites has been associated with transformation (16, 17, 18). Finally, a large number of phosphorylated viral proteins have been characterized, which also might generate phosphopeptide fragments for MHC-restricted presentation.

In this study, we have analyzed the effect of phosphorylation on the ability of peptides to be transported by TAP, binding to class I molecules, and CTL recognition. We conclude that posttranslationally modified cytosolic phosphopeptides could be presented to the immune system by class I MHC molecules and recognized by CTL in a phosphorylation-specific manner.

The TAP-deficient cell lines RMA-S and T2 were maintained in RPMI 1640 (Life Technologies, Gaithersburg, MD) supplemented with 10% FCS and glutamine. T2 cells transfected with HLA-B*0702 (T2-B7) (19), HLA-B*3501 (T2-B35) (20), and HLA-B*2705 (T2-B27) (21) were kindly supplied by A. McMichael (Oxford University, Oxford, U.K.), and T2 cells transfected with H-2Kk (T2-Kk) were kindly supplied by P. Cresswell (Yale University, New Haven, CT). For the T2-B7 and T2-B27 cell lines, the medium was supplemented with 100 μg/ml hygromycin and 250 μg/ml G418, respectively.

EBV-transformed B lymphoblastoid cell lines (BCL)3 were established by coculturing PBMC with supernatant from the EBV-infected marmoset cell line B95.8 in the presence of cyclosporin A. Approximately 3 × 107 PBMC were resuspended in 5 ml R10 and 5 ml B95.8 supernatant and incubated for 4 h at 37°C. The cells were then resuspended in fresh R10 (8 ml), incubated at 37°C, and the following day cyclosporin A was added to the culture (1 μg/ml). Two EBV-transformed BCL lines were established: JS-BCL from the donor JS expressing HLA-A2, -A3, -B7, -B60, and -Cw3, and KIG-BCL from the donor KIG expressing HLA-A2, -A3, -B7, and -B15(62).

The nonphosphorylated peptides were from Research Genetics (Huntsville, AL) and provided at >90% purity, as analyzed by HPLC and mass spectroscopy analysis.

Phosphopeptides containing phosphoserine or phosphotyrosine residues were synthesized using a phosphorylated Fmoc-serine or Fmoc-tyrosine building blocks (FmocSer(PO(Obzl)OH)OH or Fmoc-Tyr(PO(Obzl)OH)OH); Calbiochem, Lisztweg, Germany) in the manual solid-phase Fmoc synthesis of phosphopeptides. All phosphopeptides were HPLC purified and characterized by mass spectroscopy and nuclear magnetic resonance. In addition to oncogene- or virus protein-derived phosphopeptides, the following peptides were synthesized for the TAP assays: 417 [TVNKTERAY], and 417-S [TVNKTESAY], as well as the phosphopeptide 417-P [TVNKTES(−PO32-)AY]. Before use in the TAP assay, some peptides were radiolabeled with Na125I catalyzed by chloramine-T.

Assembly assays for binding of the synthetic peptides to class I MHC molecules metabolically labeled with [35S]methionine were conducted essentially as described (22, 23). This assay is based on the peptide-dependent stabilization of class I molecules after lysis of peptide transporter-deficient cell lines (T2 or RMA-S) and the subsequent detection of correctly folded, stable MHC heavy chains by immunoprecipitation with conformation-dependent Abs. To prevent dephosphorylation of peptides during the assembly assay, phosphatase inhibitors were added (1 mM Na3VO4 and 5 mM NaF). After electrophoresis, gels were exposed to PhosphorImager screens, and peptide binding was quantitated using the Imagequant PhosphorImager program (Molecular Dynamics, Sunnyvale, CA).

Assays for the TAP-mediated translocation across the ER membrane of the radiolabeled peptide 417, the serine-substituted peptide 417-S, as well as a phosphorylated (417-P) peptide analogue in LCL721 cells (LCL; human) or RMA cells (murine) were performed as described (24, 25). T2 cells or RMA-S cells were used to demonstrate the TAP dependence of transport. Briefly, 5 × 106 cells per sample were washed in incubation buffer (130 mM KCl, 10 mM NaCl, 1 mM CaCl2, 2 mM EGTA, 2 mM MgCl2, and 5 mM HEPES, pH 7.3) before permeabilization with 2.5 IU/ml streptolysin O (Wellcome Diagnostics, Dartford, U.K.). The cell suspension was then added to 0.1 μg of iodinated peptide (sp. act., 30 μCi/μg), with or without ATP (Boehringer Mannheim, Mannheim, Germany) in a final volume of 100 μl. Translocation was performed at 37°C over 5 min before ice-cold lysis buffer (1% Triton X-100, 500 mM NaCl, 5 mM MgCl2, 50 mM Tris-HCl, pH 7.5) was added to each tube. After 30 min of lysis (4°C), the nuclei were removed by centrifugation (12,000 rpm, 10 min). Then the N-glycosylated peptides were precipitated by the addition of Con A-Sepharose beads (Pharmacia, Piscataway, NJ) and washed five times with ice-cold lysis buffer, and the counts were associated with the Con A-Sepharose pellet quantitated by gamma counting. For experiments with phosphorylated peptides, the phosphatase inhibitor NaF (10 mM) was included in all buffers after having established that this did not affect translocation of peptide 417.

Peptide-specific CTL lines were generated by immunization of C57BL/6 mice, as described (26). Briefly, RMA-S cells (108) were pulsed with the phosphopeptide K3-P (10 μM) at 26°C for 18 h. The cells were irradiated (60 Gy) and inoculated i.p. into C57BL/6 mice. Mice were sacrificed 3 wk later and the spleen cells were restimulated in vitro with syngeneic spleen cells (4 × 106) pulsed with K3-P for 5 days in RPMI 1640 supplemented with 10% FCS.

Peptide-specific CTL were established in vitro as primary responses using autologous dendritic cells (DC) as APC. DC were developed from PBMC from healthy donors, as described previously (27). Briefly, PBMC were resuspended at 4 × 106 cells/ml in R10 medium in six-well plates (Nunc, Naperville, IL) at 3 ml/well. After 2 h of incubation at 37°C, the nonadherent cells were washed away, and 3 ml R10 medium supplemented with 500 U/ml GM-CSF (Leucomax, Sandoz, Germany) and 25 U/ml IL-4 (PeproTec, Boston, MA) was added to each well. Every third day, 0.5 ml medium was replaced with fresh R10 medium containing GM-CSF and IL-4 to a final concentration of 500 U/ml and 25 U/ml, respectively. After 9 days, 50 U/ml IL-1α was added to the culture, and DC were harvested the next day. DC were resuspended at 5 × 106 cells/ml in RPMI without serum containing 40 μg/ml peptide and 3 μg/ml human β2-microglobulin for 4 h and subsequently irradiated (30 Gy). The peptide-pulsed, irradiated DC (2 × 105/well) were used to stimulate fresh lymphocytes (5 × 106/well) in 24-well plates (Costar) in 2 ml RPMI medium with 20% AIM V medium (Life Technologies) and 5% human serum supplemented with 5 ng/ml IL-7 (PeproTech). On day 10, and weekly thereafter, the responder population was restimulated. Irradiated, peptide-pulsed autologous PBMC were used to restimulate the T cell line at a stimulator:responder ratio of 1:5, with addition of IL-2 (40 U/ml; PeproTech) on the following day. CTL clones were established from bulk cultures by limiting dilution in 96-well plates using peptide-pulsed, irradiated PBMC as feeder cells in the presence of 40 U/ml IL-2 and 1 μg/ml PHA.

Conventional 51Cr release assays for CTL-mediated cytotoxicity were used to test the specificity of CTL lines against peptide-pulsed target cells. Target cells were autologous EBV-transformed B cell lines (assays with human cells) or T2 cells transfected with H-2Kb or H-2Db (murine assays), as described (7). For a typical CTL assay, 106 target cells in 50 μl R10 were labeled with 51Cr (100 μCi; Amersham, Arlington Heights, IL) and in some cases pulsed with peptide (4 μg) in a round-bottom well of a 96-well plate at 37°C for 60 min. The washed target cells were plated out in 96-well plates with CTL at various E:T ratios and incubated at 37°C for 4 h, before 100 μl of medium was aspirated and 51Cr release counted in a gamma counter (Cobra 5005; Packard Instruments, Meriden, CT). The maximum 51Cr release was determined in separate wells by addition of 100 μl 10% Triton X-100, and spontaneous release was determined by the addition of 100 μl R10 only to target cells. Specific lysis was calculated using the following formula: ((experimental release − spontaneous release)/(maximum release − spontaneous release)) × 100.

Assays for direct translocation of phosphopeptides by TAP were conducted as described (24, 25) by adding radiolabeled peptides and phosphopeptides, which contain an N-linked glycosylation sequon to streptolysin O-permeabilized cells. Fig. 1 shows the result of an assay for translocation of the peptides 417, the serine substituted 417-S, and the phosphorylated 417-P using the human LCL cells in the presence or absence of ATP for each peptide. All peptides are transported across the ER membrane in a strictly ATP-dependent manner by both the human (Fig. 1) and murine (results not shown) TAP, as determined by the resulting N-glycosylation and recovery of the iodinated peptides with Con A-Sepharose. Peptide 417 as well as the serine-substituted 417-S were very efficiently translocated with recoveries of 4–5% for LCL, consistent with previously published data (28). Phosphorylation of 417-S (giving 417-P) resulted in translocation of similar high efficiency in LCL. These data show that phosphorylated peptides can be efficiently transported from the cytosol into the ER by TAP.

FIGURE 1.

Translocation of phosphopeptide by human TAP. The index peptide (417), a serine-substituted analogue (417-S), as well as a phosphorylated (417-P) version thereof, all containing an acceptor site for ER-type N-linked glycosylation, were iodinated and added to streptolysin O-permeabilized human LCL cells in the presence or absence of ATP. Peptides transported into the ER were recovered with Con A-Sepharose from detergent lysates and quantified by gamma counting. The results indicated are the amounts of translocated peptide recovered, expressed as a fraction of the total input peptide.

FIGURE 1.

Translocation of phosphopeptide by human TAP. The index peptide (417), a serine-substituted analogue (417-S), as well as a phosphorylated (417-P) version thereof, all containing an acceptor site for ER-type N-linked glycosylation, were iodinated and added to streptolysin O-permeabilized human LCL cells in the presence or absence of ATP. Peptides transported into the ER were recovered with Con A-Sepharose from detergent lysates and quantified by gamma counting. The results indicated are the amounts of translocated peptide recovered, expressed as a fraction of the total input peptide.

Close modal

Based on an extensive binding study4 with 73 potential MHC-binding peptides all containing a natural phosphorylation site from viral phosphoproteins as well as from phosphorylated oncogene products that display altered phosphorylation upon malignant transformation, nine peptides that showed strong binding were selected for phosphorylation. Four peptides were derived from human oncogene phosphoproteins and four peptides were from viral phosphoproteins. In addition, the model peptide K3 (FAPSNYPAL) (7) was included in the study. K3 is an analogue of the H-2Kb-restricted immunodominant Sendai virus NP324–332 epitope FAPGNYPAL.

The tyrosine-phosphorylated version of the CRKL-derived peptide OP42-P (YAQPQTTTPL) bound strongly to HLA-A*0201 (Fig. 2,A), illustrating that phosphorylation did not prevent the efficient binding of this peptide to class I MHC. However, we also observed several instances in which the phosphorylation of nonanchor tyrosine, serine, or threonine residues had either an intermediate or a severe negative effect on peptide binding to class I .MHC alleles. Thus, the phosphorylated c-Jun peptide OP27-P (GLLKLASPEL) only bound weakly to HLA-A*0201 compared with the nonphosphorylated peptide (Fig. 2 B), and the phosphorylated version of the p53378–386 peptide OP39-P did not show any binding to HLA-B*2705, whereas OP39 (SRHKKLMFK) bound better than the positive control epitope (KRWIIMGLNK) (29).

FIGURE 2.

Phosphopeptide binding to HLA-A*0201 (A and B) or HLA-B*2705 (C). The binding of the phosphorylated CRKL207–386 peptide OP42-P (Y(−PO32-)AQPQTTTPL) (▴) was compared with the nonphosphorylated OP42 (YAQPQTTTPL) (▵) and the positive control peptide C1 (○) (A). The binding of the phosphorylated c-Jun67–76 peptide OP27-P (GLLKLAS(−PO32-)PEL) (▪) was compared with the nonphosphorylated OP27 (GLLKLASPEL) (□) and the positive control peptide C1 (○) (B). The binding of the phosphorylated p53378–386 peptide OP39-P (S(−PO32-)RHKKLMFK) (▾) was compared with the nonphosphorylated OP39 (SRHKKLMFK) (▿) and the positive control peptide C4 (⋄) (C).

FIGURE 2.

Phosphopeptide binding to HLA-A*0201 (A and B) or HLA-B*2705 (C). The binding of the phosphorylated CRKL207–386 peptide OP42-P (Y(−PO32-)AQPQTTTPL) (▴) was compared with the nonphosphorylated OP42 (YAQPQTTTPL) (▵) and the positive control peptide C1 (○) (A). The binding of the phosphorylated c-Jun67–76 peptide OP27-P (GLLKLAS(−PO32-)PEL) (▪) was compared with the nonphosphorylated OP27 (GLLKLASPEL) (□) and the positive control peptide C1 (○) (B). The binding of the phosphorylated p53378–386 peptide OP39-P (S(−PO32-)RHKKLMFK) (▾) was compared with the nonphosphorylated OP39 (SRHKKLMFK) (▿) and the positive control peptide C4 (⋄) (C).

Close modal

We then tested the binding of nine different phosphopeptide versions of class I MHC-binding peptides, in which modifications lay outside the primary anchor residues, for their ability to bind to seven different class I MHC molecules. The binding of each peptide was compared with a positive control peptide (Table I), and a binding score was assigned to the different peptides.

Table I.

Positive control peptides

MHC AlleleNameSequenceProteinaScoreC50 (μM)bRef.
A*0201 C1 ILKEPVHGV HIV-1 pol476–484 +++ 1 μM 30 
B*0702 C2 QPRAPIRPI EBV EBNA 3C881–889 +++ 0.8 μM 31 
B*3501 C3 MPLETQLAI Malaria LFA-177–85 +++ 0.9 μM 32 
B*2705 C4 KRWIIMGLNK HIV-1 gag263–272 +++ 8 μM 29, 33 
H-2Kb C5 FAPGNYPAL Sendai virus NP324–332 +++ 0.05 μM 34 
H-2Db C5 FAPGNYPAL Sendai virus NP324–332 +++ 0.005 μM 34 
H-2Kk C6 SDYEGRLI Influenza virus NP50–57 >20 μM 35 
MHC AlleleNameSequenceProteinaScoreC50 (μM)bRef.
A*0201 C1 ILKEPVHGV HIV-1 pol476–484 +++ 1 μM 30 
B*0702 C2 QPRAPIRPI EBV EBNA 3C881–889 +++ 0.8 μM 31 
B*3501 C3 MPLETQLAI Malaria LFA-177–85 +++ 0.9 μM 32 
B*2705 C4 KRWIIMGLNK HIV-1 gag263–272 +++ 8 μM 29, 33 
H-2Kb C5 FAPGNYPAL Sendai virus NP324–332 +++ 0.05 μM 34 
H-2Db C5 FAPGNYPAL Sendai virus NP324–332 +++ 0.005 μM 34 
H-2Kk C6 SDYEGRLI Influenza virus NP50–57 >20 μM 35 
a

The value range listed in subscript indicate the position of the peptide in the sequence.

b

The C50 value is the concentration of the peptide required for half maximal binding.

Thus, peptides binding with the same high affinity as the positive control peptides were assigned the score +++, whereas peptides without measurable binding were assigned the value −. In general, phosphopeptides could be divided into three groups in which phosphorylation had no effect, an intermediate negative effect, or a strong negative effect on binding to class I MHC. The following three phosphopeptides belonged to the group of peptides that maintained efficient class I MHC binding upon phosphorylation: OP42-P (CRKL207–216, YAQPQTTTPL) bound to HLA-A*0201, OP46-P (BCR174–182, KPFYVNVEF) bound to both HLA-B*0701 and HLA-B*3501, and the Sendai virus peptide analogue K3-P (FAPSNYPAL) bound to both H-2Db and H-2Kb with high affinity (Table II). Of the peptides whose binding is blocked by phosphorylation, the influenza M287–94 peptide P49 (DDSHFVSI) is the only one that is phosphorylated on two sites in vivo (41). However, neither the diphosphorylated version nor any of the two monophosphorylated versions of the peptide showed any measurable binding to H-2Kk. The fact that phosphorylation completely abrogates class I MHC binding of some peptides examined is a strong indication that dephosphorylation does not occur during the peptide-binding assay.

Table II.

Binding of phosphopeptides to class I MHC molecules

Peptide NameSequenceaProteinScoreC50 (μM)Ref.b
HLA-A*0201      
OP42 YAQPQTTTPL CRKL207–216 +++ 36 
OP42-P YAQPQTTTPL  ++++ 0.5  
OP27 GLLKLASPEL c-JUN67–76 +++ 37 
OP27-P GLLKLASPEL  >20  
HLA-B*0702      
OP42 YAQPQTTTPL CRKL207–216 >20 36 
OP42-P YAQPQTTTPL  >20  
OP46 KPFYVNVEF BCR174–182 +++ 0.8 38 
OP46-P KPFYVNVEF  ++  
P26 SPIVPSFDM Influenza virus NP473–481 ++ 39 
P26-P SPIVPSFDM  −   
HLA-B*3501      
OP42 YAQPQTTTPL CRKL207–216 +++ 0.9 36 
OP42-P YAQPQTTTPL  >20  
OP46 KPFYVNVEF BCR174–182 +++ 1.6 38 
OP46-P KPFYVNVEF  ++  
P26 SPIVPSFDM Influenza virus NP473–481 +++ 0.5 39 
P26-P SPIVPSFDM  −   
HLA-B*2705      
OP39 SRHKKLMFK p53378–386 ++++ 0.5 40 
OP39-P SRHKKLMFK  − −  
H-2Kb      
K3 FAPSNYPAL Sendai virus NP324–332 +++ 0.2 
K3-P FAPSNYPAL (analogue) +++ 0.2  
H-2Db      
K3 FAPSNYPAL Sendai virus NP324–332 +++ 0.01 
K3-P FAPSNYPAL (analogue) +++ 0.01  
H-2Kk      
P9 SDEKAASPI Influenza virus NP467–475 ++ >20 39 
P9-P SDEKAASPI  >20  
P10 IDEKLSEIL RSV RNA pol138–146 +++ 0.6 16 
P10-P IDEKLSEIL  − −  
P49 DDSHFVSI Influenza virus M287–94 +++ 41 
P49-PP DDSHFVS −   
P49-P1 DDSHFVSI  −   
P49-P2 DDSHFVS −   
Peptide NameSequenceaProteinScoreC50 (μM)Ref.b
HLA-A*0201      
OP42 YAQPQTTTPL CRKL207–216 +++ 36 
OP42-P YAQPQTTTPL  ++++ 0.5  
OP27 GLLKLASPEL c-JUN67–76 +++ 37 
OP27-P GLLKLASPEL  >20  
HLA-B*0702      
OP42 YAQPQTTTPL CRKL207–216 >20 36 
OP42-P YAQPQTTTPL  >20  
OP46 KPFYVNVEF BCR174–182 +++ 0.8 38 
OP46-P KPFYVNVEF  ++  
P26 SPIVPSFDM Influenza virus NP473–481 ++ 39 
P26-P SPIVPSFDM  −   
HLA-B*3501      
OP42 YAQPQTTTPL CRKL207–216 +++ 0.9 36 
OP42-P YAQPQTTTPL  >20  
OP46 KPFYVNVEF BCR174–182 +++ 1.6 38 
OP46-P KPFYVNVEF  ++  
P26 SPIVPSFDM Influenza virus NP473–481 +++ 0.5 39 
P26-P SPIVPSFDM  −   
HLA-B*2705      
OP39 SRHKKLMFK p53378–386 ++++ 0.5 40 
OP39-P SRHKKLMFK  − −  
H-2Kb      
K3 FAPSNYPAL Sendai virus NP324–332 +++ 0.2 
K3-P FAPSNYPAL (analogue) +++ 0.2  
H-2Db      
K3 FAPSNYPAL Sendai virus NP324–332 +++ 0.01 
K3-P FAPSNYPAL (analogue) +++ 0.01  
H-2Kk      
P9 SDEKAASPI Influenza virus NP467–475 ++ >20 39 
P9-P SDEKAASPI  >20  
P10 IDEKLSEIL RSV RNA pol138–146 +++ 0.6 16 
P10-P IDEKLSEIL  − −  
P49 DDSHFVSI Influenza virus M287–94 +++ 41 
P49-PP DDSHFVS −   
P49-P1 DDSHFVSI  −   
P49-P2 DDSHFVS −   
a

Phosphorylated residues are underlined, and phosphopeptides are designated with “−P.”

b

All peptides contain a natural phosphorylation site.

It is interesting to note that phosphorylation of peptide residue 1 in P26 blocks peptide binding to HLA-B*0702 and HLA-B*3501, and similarly phosphorylation of residue 1 in peptide OP39 blocks binding to HLA-B*2705. The side chain of the N-terminal peptide residue does not usually interact directly with the class I MHC molecule. However, charge-pairing effects between the α-amino terminus of the peptide and the phosphate group may prevent peptide binding.

Both the nonphosphorylated and the phosphorylated version of the synthetic peptide K3 (FAPSNYPAL) bound with high affinity to both H-2Db and H-2Kb. Spleen cells from C57BL/6 mice immunized with phosphopeptide K3-P-pulsed RMA-S cells were harvested after 3 wk later and tested for lytic activity against peptide-pulsed T2-Kb and T2-Db cells. As seen in Fig. 3 A, the spleen cell-derived CTL showed a high degree of specificity against the phosphopeptide. However, some CTL activity against the nonphosphorylated peptide was also observed, either due to the coexistence of nonphosphorylated peptide-specific CTL or due to cross-reacting CTL. Furthermore, the CTL response against the phosphopeptide was H-2Kb restricted, because only T2-Kb but not T2-Db cells pulsed with K3-P were lysed.

FIGURE 3.

Peptide recognition by murine CTL raised against K3-P. Splenocyte-derived CTL from C57BL/6 mice immunized with K3-P-pulsed RMA-S cells were used as effectors against T2-Kb (A) or T2-Db (B) target cells pulsed with K3-P (•), K3 (○), or without peptide (▵).

FIGURE 3.

Peptide recognition by murine CTL raised against K3-P. Splenocyte-derived CTL from C57BL/6 mice immunized with K3-P-pulsed RMA-S cells were used as effectors against T2-Kb (A) or T2-Db (B) target cells pulsed with K3-P (•), K3 (○), or without peptide (▵).

Close modal

Having confirmed the efficient binding to HLA-A2 (see Fig. 2,A), the CRKL-derived phosphopeptide OP42-P peptide was also used to generate human CTL in vitro primary responses. Autologous DC were generated from HLA-A2-positive PBMC in the presence of IL-4 and GM-CSF. Next CTL were stimulated in vitro using irradiated OP42-P-pulsed dendritic cells as APC in the presence of IL-2 and IL-7. After five in vitro restimulations, the peptide specificity was tested in standard 51Cr release assays using peptide-pulsed autologous EBV-transformed KIG-BCL as target cells. Target cells pulsed with OP42-P were lysed (Fig. 4 A), whereas target cells pulsed with OP42 or without peptide were not recognized. Thus, the resulting CTL line was specific for the phosphopeptide OP42-P.

FIGURE 4.

Peptide recognition by human CTL raised against OP42-P (A) or against nonphosphorylated OP42 (B). HLA-A2-restricted human in vitro primary CTL raised against the phosphopeptide OP42-P (Y(−PO32-)AQPQTTTPL) (A) or the nonphosphorylated peptide OP42 (B) were tested in a 51Cr release assay using autologous EBV-transformed BCL pulsed with OP42 (○), OP42-P (•), or without peptide (▵).

FIGURE 4.

Peptide recognition by human CTL raised against OP42-P (A) or against nonphosphorylated OP42 (B). HLA-A2-restricted human in vitro primary CTL raised against the phosphopeptide OP42-P (Y(−PO32-)AQPQTTTPL) (A) or the nonphosphorylated peptide OP42 (B) were tested in a 51Cr release assay using autologous EBV-transformed BCL pulsed with OP42 (○), OP42-P (•), or without peptide (▵).

Close modal

CTL were also raised in vitro against the nonphosphorylated version of the peptide OP42, which binds with high affinity to HLA-A2, using a HLA-A2-positive donor. This CTL cell line was cloned, and the specificity of the resulting clones was examined. From Fig. 4 B it is seen that a CTL clone raised against OP42-lysed target cells pulsed with OP42, and that it did not recognize target cells pulsed with phosphorylated OP42-P or without peptide. Thus, a peptide-specific CTL clone was established that did not cross-react with the phosphorylated peptide. Furthermore, these data indicate that no dephosphorylation of peptide occurred during the CTL assay.

These results clearly demonstrate the ability to generate phosphopeptide-specific class I MHC-restricted CTL responses, and suggest that the phosphate group is involved in a specific contact with TCR.

In this study, we demonstrate that phosphopeptides can be transported from the cytosol into the ER by TAP, and that a class I MHC-binding peptide can be phosphorylated without affecting its binding to MHC class I. Thus, it is likely that phosphopeptides are presented by class I MHC molecules in vivo. Furthermore, we demonstrate that at the cell surface phosphopeptides presented by class I MHC molecules can elicit phosphopeptide-specific CTL. Also, CTL raised against a nonphosphorylated peptide recognize target cells pulsed with the nonphosphorylated peptide, without recognizing target cells pulsed with the phosphorylated peptide. Thus, T lymphocytes discriminate between phosphorylated and nonphosphorylated peptides. In future studies, we will address whether class I MHC-restricted T cell responses against endogenously expressed phosphopeptides are part of the natural T cell repertoire in vivo. Also, it cannot be ruled out that proteasome degradation or cytosolic phosphatases prevent the presentation of phosphopeptides in vivo.

We have investigated CTL responses against phosphopeptides in two model systems. We observed a HLA-A2-restricted phosphopeptide-specific primary response against OP42-P in vitro. We also immunized mice with the phosphopeptide K3-P analogue of the dominant Kb-restricted epitope from Sendai virus NP. Even though this phosphopeptide binds with high affinity to H-2Kb or H-2Db, we only observed a Kb-restricted CTL response against the peptide, thus resembling the natural H-2Kb-restricted response in C57BL/6 mice against Sendai virus. Our group has previously described that glycosylation of K3 peptide on the same serine residue at position 4 with N-acetylglucosamine (GlcNAc) does not effect peptide binding to H-2Kb or H-2Db, and that it is possible to elicit both H-2Kb- and H-2Db-restricted glycopeptide-specific CTL response (7). However, in our previous study, CTL were restimulated several times in vitro, which may have induced the Db-restricted response. Recently, we have solved the crystal structure of a glycosylated version of K3 (FAPS(O-β-GlcNAc)NYPAL) in complex with H-2Db, and it was revealed that the carbohydrate residue is pointing out of the groove (42).

Four of the phosphopeptides examined did not show any binding to MHC when phosphorylated. The molecular basis for this inhibition is unclear because all modifications were made at positions outside the primary anchor residues. Two of the four affected peptides were phosphorylated at position 1, which could possibly affect the formation of the conserved bonding in the A pocket of the class I MHC-binding groove. The other two affected peptides were H-2Kk restricted, and binding to this allele, which requires a negatively charged position 2 anchor residue, may be adversely affected by the addition of a further highly negatively charged group. The negative influence of phosphorylation on class I MHC binding is of immunological interest, because it provides a novel potential immune escape mechanism. For example, the influenza NP473–481-derived phosphopeptide P26 did not show any appreciable binding to HLA-B*0702 and HLA-B*3501, whereas the nonphosphorylated version of the same peptide bound with high affinity to both alleles. Thus, if the source protein is phosphorylated in an infected cell, resulting in the generation of the phosphorylated peptide by the Ag-processing apparatus, it cannot be presented at the cell surface by the MHC molecule, thus allowing the virus to avoid immune elimination.

To date, only a few examples of posttranslationally modified natural class I MHC-restricted epitopes have been described. Skipper et al. identified a HLA-A*0201-restricted tyrosinase-derived melanoma peptide, in which the asparagine in position 3 was posttranslationally modified to aspartic acid (9). Meadows et al. identified a modified HLA-A*0201-restricted peptide epitope from the human minor histocompatibility Ag H-Y, describing T cells that only recognize posttranslationally altered forms of this peptide (10). One of these modifications involves attachment to the peptide of a second cysteine residue via a disulfide bond.

Our group has previously shown that class I MHC molecules can bind synthetic glycopeptides and that the glycopeptides elicit a strong CTL response that is glycopeptide specific (7). Furthermore, we have recently shown that glycopeptides serve as natural substrates for the TAP molecule, and that class I MHC molecules present a small amount of glycosylated peptides in vivo (43). Van Stipdonk et al. have recently described that murine class II MHC-restricted T cells raised against differentially phosphorylated forms of αB-crystallin discriminate between the phosphorylated and the nonphosphorylated peptides (44).

In this study, we have investigated the effect of phosphorylation on peptide binding to class I MHC molecules and the recognition of CTL. We have examined three different kinds of phosphoproteins as sources for the cellular production of phosphopeptides. First, because known phosphorylation sites in some proteins become constitutively phosphorylated during oncogene activation, the ensuing phosphopeptide fragments offer themselves as candidates for T cell Ags in line with overexpressed, nonmutated tumor Ags. Examples of such proteins include c-Jun and the tumor suppressor gene product p53. c-Jun becomes activated by phosphorylation of the two Ser residues 63 and 73, and has been reported to be phosphorylated in both leukemia and lung cancer cells (37). Phosphorylation of the p53 tumor suppressor gene product has been described for a number of serine residues, among these Ser9, Ser15, Ser37, Ser315, Ser392, and Ser378 (40, 45, 46, 47).

Second, upon the critical event leading to activation of the transforming phenotype, some protooncogenes characteristically acquire novel kinase activities leading to the appearance of novel phosphorylation sites. Thus, as a consequence of the (9, 22) translocation (the Philadelphia chromosome) that is associated with all chronic myeloid leukemia cases and 20% of adult acute lymphoblastic leukemia, the c-abl protooncogene becomes activated by its translocation to chromosome 22 giving rise to a BCR-ABL fusion protein (48, 49). A crucial difference between the fusion protein and the normal c-abl is that only the chimeric protein exhibits tyrosine kinase activity (14), resulting in autophosphorylation of a number of tyrosine residues (including Tyr177) of the BCR-ABL fusion protein (50, 51), as well as causing a constitutive increase in tyrosine phosphorylation of other cellular proteins. One of the best-described proteins is the CRKL, which is overexpressed and hyperphosphorylated on Tyr207 in Philadelphia chromosome-positive chronic myeloid leukemia cells (36, 52). Each of these new sites of phosphorylation could give rise to tumor-specific phosphopeptide neo-Ags. Furthermore, it is conceivable that transformation-associated posttranslational modification of a cellular protein could alter the pattern of peptide fragments that are generated from it.

Finally, viral phosphoproteins with known phosphorylation sites have also been included in the study. For example, the EBV protein EBNA-2 is phosphorylated at Ser469 and Ser470 (53), EBV LMP-1 at Ser313 and Thr324 (54), and EBV BZLF1 at Ser173 (55). Other viruses with multiple known phosphorylation sites include adenovirus, HIV, and influenza virus.

1

This work was supported by the Danish Medical Research Council, the Novo Nordisk Foundation, the “Direktør E. Danielsen og Hustrus” Foundation, and the Danish Cancer Society. T.E. is a Wellcome Trust Senior Fellow in Basic Biomedical Science.

3

Abbreviations used in this paper: BCL, B lymphoblastoid cell line; BCR, breakpoint cluster region; ER, endoplasmic reticulum; GlcNAc, N-acetylglucosamine; LCL, lymphoblastoid cell line; NP, nucleoprotein.

4

M. H. Andersen, L. Tan, I. Søndergaard, J. Zeuthen, T. Elliott, and J. S. Haurum. Poor correspondence between predicted and experimental binding of peptides to class I MHC molecules. Submitted for publication

1
Rotzschke, O., K. Falk, K. Deres, H. Schild, M. Norda, J. Metzger, G. Jung, H. G. Rammensee.
1990
. Isolation and analysis of naturally processed viral peptides as recognized by cytotoxic T cells.
Nature
348
:
252
2
Falk, K., O. Rotzschke, S. Stevanovic, G. Jung, H. G. Rammensee.
1991
. Allele-specific motifs revealed by sequencing of self-peptides eluted from MHC molecules.
Nature
351
:
290
3
Jardetzky, T. S., W. S. Lane, R. A. Robinson, D. R. Madden, D. C. Wiley.
1991
. Identification of self peptides bound to purified HLA-B27.
Nature
353
:
326
4
Bjorkman, P. J., M. A. Saper, B. Samraoui, W. S. Bennett, J. L. Strominger, D. C. Wiley.
1987
. Structure of the human class I histocompatibility antigen, HLA-A2.
Nature
329
:
506
5
Saper, M. A., P. J. Bjorkman, D. C. Wiley.
1991
. Refined structure of the human histocompatibility antigen HLA-A2 at 2.6 Å resolution.
J. Mol. Biol.
219
:
277
6
Madden, D. R..
1995
. The three-dimensional structure of peptide-MHC complexes.
Annu. Rev. Immunol.
13
:
587
7
Haurum, J. S., G. Arsequell, A. C. Lellouch, S. Y. Wong, R. A. Dwek, A. J. McMichael, T. Elliott.
1994
. Recognition of carbohydrate by major histocompatibility complex class I-restricted, glycopeptide-specific cytotoxic T lymphocytes.
J. Exp. Med.
180
:
739
8
Haurum, J. S., L. Tan, G. Arsequell, P. Frodsham, A. C. Lellouch, P. A. Moss, R. A. Dwek, A. J. McMichael, T. Elliott.
1995
. Peptide anchor residue glycosylation: effect on class I major histocompatibility complex binding and cytotoxic T lymphocyte recognition.
Eur. J. Immunol.
25
:
3270
9
Skipper, J. C., R. C. Hendrickson, P. H. Gulden, V. Brichard, A. Van Pel, Y. Chen, J. Shabanowitz, T. Wolfel, C. L. Slingluff Jr, et al
1996
. An HLA-A2-restricted tyrosinase antigen on melanoma cells results from posttranslational modification and suggests a novel pathway for processing of membrane proteins.
J. Exp. Med.
183
:
527
10
Meadows, L., W. Wang, J. M. den Haan, E. Blokland, C. Reinhardus, J. W. Drijfhout, J. Shabanowitz, R. Pierce, A. I. Agulnik, C. E. Bishop, et al
1997
. The HLA-A*0201-restricted H-Y antigen contains a posttranslationally modified cysteine that significantly affects T cell recognition.
Immunity
6
:
273
11
Michaelsson, E., V. Malmstrom, S. Reis, Å. Engstrom, H. Burkhardt, R. Holmdahl.
1994
. T cell recognition of carbohydrates on type II collagen.
J. Exp. Med.
180
:
745
12
Michaelsson, E., J. Broddefalk, A. Engstrom, J. Kihlberg, R. Holmdahl.
1996
. Antigen processing and presentation of a naturally glycosylated protein elicits major histocompatibility complex class II-restricted, carbohydrate-specific T cells.
Eur. J. Immunol.
26
:
1906
13
Dudler, T., F. Altmann, J. M. Carballido, K. Blaser.
1995
. Carbohydrate-dependent, HLA class II-restricted, human T cell response to the bee venom allergen phospholipase A2 in allergic patients.
Eur. J. Immunol.
25
:
538
14
Clark, S. S., J. McLaughlin, W. M. Crist, R. Champlin, O. N. Witte.
1987
. Unique forms of the abl tyrosine kinase distinguish Ph1-positive CML from Ph1-positive ALL.
Science
235
:
85
15
Xia, Y., L. Y. Hwang, M. H. Cobb, R. Baer.
1994
. Products of the TAL2 oncogene in leukemic T cells: bHLH phosphoproteins with DNA-binding activity.
Oncogene
9
:
1437
16
Villanueva, N., J. Navarro, E. Mendez, I. Garcia Albert.
1994
. Identification of a protein kinase involved in the phosphorylation of the C-terminal region of human respiratory syncytial virus P protein.
J. Gen. Virol.
75
:
555
17
Milne, D. M., L. E. Campbell, D. G. Campbell, D. W. Meek.
1995
. p53 is phosphorylated in vitro and in vivo by an ultraviolet radiation-induced protein kinase characteristic of the c-Jun kinase, JNK1.
J. Biol. Chem.
270
:
5511
18
Milne, D. M., D. G. Campbell, F. B. Caudwell, D. W. Meek.
1994
. Phosphorylation of the tumor suppressor protein p53 by mitogen-activated protein kinases.
J. Biol. Chem.
269
:
9253
19
Alexander, J., J. A. Payne, R. Murray, J. A. Frelinger, P. Cresswell.
1989
. Differential transport requirements of HLA and H-2 class I glycoproteins.
Immunogenetics
29
:
380
20
Dong, T., D. Boyd, W. Rosenberg, N. Alp, M. Takiguchi, A. McMichael, S. Rowland Jones.
1996
. An HLA-B35-restricted epitope modified at an anchor residue results in an antagonist peptide.
Eur. J. Immunol.
26
:
335
21
Anderson, K. S., J. Alexander, M. Wei, P. Cresswell.
1993
. Intracellular transport of class I MHC molecules in antigen processing mutant cell lines.
J. Immunol.
151
:
3407
22
Elvin, J., C. Potter, T. Elliott, V. Cerundolo, A. Townsend.
1993
. A method to quantify binding of unlabeled peptides to class I MHC molecules and detect their allele specificity.
J. Immunol. Methods
158
:
161
23
Tan, L., M. H. Andersen, T. Elliott, J. S. Haurum.
1997
. An improved assembly assay for peptide binding to HLA-B*2705 and H-2Kk class I MHC molecules.
J. Immunol. Methods
209
:
25
24
Neefjes, J. J., F. Momburg, G. J. Hämmerling.
1993
. Selective and ATP-dependent translocation of peptides by the MHC-encoded transporter.
Science
261
:
769
25
Neisig, A., J. Roelse, A. J. Sijts, F. Ossendorp, M. C. Feltkamp, W. M. Kast, C. J. Melief, J. J. Neefjes.
1995
. Major differences in transporter associated with antigen presentation (TAP)-dependent translocation of MHC class I-presentable peptides and the effect of flanking sequences.
J. Immunol.
154
:
1273
26
Gao, X. M., A. J. McMichael.
1996
. Cytotoxic T lymphocytes specific for murine type II collagen do not trigger arthritis in B10 mice.
Clin. Exp. Immunol.
103
:
89
27
Romani, N., S. Gruner, D. Brang, E. Kampgen, A. Lenz, B. Trockenbacher, G. Konwalinka, P. O. Fritsch, R. M. Steinman, G. Schuler.
1994
. Proliferating dendritic cell progenitors in human blood.
J. Exp. Med.
180
:
83
28
Neefjes, J., E. Gottfried, J. Roelse, M. Gromme, R. Obst, G. J. Hämmerling, F. Momburg.
1995
. Analysis of the fine specificity of rat, mouse and human TAP peptide transporters.
Eur. J. Immunol.
25
:
1133
29
Phillips, R. E., S. Rowland-Jones, D. F. Nixon, F. M. Gotch, J. P. Edwards, A. O. Ogunlesi, J. G. Elvin, J. A. Rothbard, C. R. Bangham, C. R. Rizza, A. J. McMichael.
1991
. Human immunodeficiency virus genetic variation that can escape cytotoxic T cell recognition.
Nature
354
:
453
30
Henderson, R. A., A. L. Cox, K. Sakaguchi, E. Appella, J. Shabanowitz, D. F. Hunt, V. H. Engelhard.
1993
. Direct identification of an endogenous peptide recognized by multiple HLA-A2.1-specific cytotoxic T cells.
Proc. Natl. Acad. Sci. USA
90
:
10275
31
Hill, A., A. Worth, T. Elliott, S. Rowland Jones, J. Brooks, A. Rickinson, A. McMichael.
1995
. Characterization of two Epstein-Barr virus epitopes restricted by HLA-B7.
Eur. J. Immunol.
25
:
18
32
Hill, A. V., J. Elvin, A. C. Willis, M. Aidoo, C. E. Allsopp, F. M. Gotch, X. M. Gao, M. Takiguchi, B. M. Greenwood, A. R. Townsend, et al
1992
. Molecular analysis of the association of HLA-B53 and resistance to severe malaria.
Nature
360
:
434
33
Nixon, D. F., A. R. Townsend, J. G. Elvin, C. R. Rizza, J. Gallwey, A. J. McMichael.
1988
. HIV-1 gag-specific cytotoxic T lymphocytes defined with recombinant vaccinia virus and synthetic peptides.
Nature
336
:
484
34
Kast, W. M., L. Roux, J. Curren, H. J. Blom, A. C. Voordouw, R. H. Meloen, D. Kolakofsky, C. J. Melief.
1991
. Protection against lethal Sendai virus infection by in vivo priming of virus-specific cytotoxic T lymphocytes with a free synthetic peptide.
Proc. Natl. Acad. Sci. USA
88
:
2283
35
Gould, K. G., H. Scotney, G. G. Brownlee.
1991
. Characterization of two distinct major histocompatibility complex class I Kk-restricted T-cell epitopes within the influenza A/PR/8/34 virus hemagglutinin.
J. Virol.
65
:
5401
36
De Jong, R., J. ten Hoeve, N. Heisterkamp, J. Groffen.
1997
. Tyrosine 207 in CRKL is the BCR/ABL phosphorylation site.
Oncogene
14
:
507
37
Smeal, T., B. Binetruy, D. Mercola, A. Grover Bardwick, G. Heidecker, U. R. Rapp, M. Karin.
1992
. Oncoprotein-mediated signaling cascade stimulates c-Jun activity by phosphorylation of serines 63 and 73.
Mol. Cell. Biol.
12
:
3507
38
Pendergast, A. M., L. A. Quilliam, L. D. Cripe, C. H. Bassing, Z. Dai, N. Li, A. Batzer, K. M. Rabun, C. J. Der, J. Schlessinger, et al
1993
. BCR-ABL-induced oncogenesis is mediated by direct interaction with the SH2 domain of the GRB-2 adaptor protein.
Cell
75
:
175
39
Kistner, O., K. Muller, C. Scholtissek.
1989
. Differential phosphorylation of the nucleoprotein of influenza A viruses.
J. Gen. Virol.
70
:
2421
40
Takenaka, I., F. Morin, B. R. Seizinger, N. Kley.
1995
. Regulation of the sequence-specific DNA binding function of p53 by protein kinase C and protein phosphatases.
J. Biol. Chem.
270
:
5405
41
Holsinger, L. J., M. A. Shaughnessy, A. Micko, L. H. Pinto, R. A. Lamb.
1995
. Analysis of the posttranslational modifications of the influenza virus M2 protein.
J. Virol.
69
:
1219
42
Glithero, A., J. Tormo, J. S. Haurum, G. Arsequell, G. Valencia, J. Edwards, S. Springer, A. R. Townsend, Y. L. Pao, M. R. Wormald, et al
1999
. Crystal structures of two H-2Db/glycopeptide complexes suggest a molecular basis for CTL cross-reactivity.
Immunity
10
:
63
43
Haurum, J. S., I. Bjerring Høier, G. Arsequell, A. Neisig, G. Valencia, J. Zeuthen, J. Neefjes, T. Elliott.
1999
. Presentation of cytosolic glycosylated peptides by human class I major histocompatibility complex molecules in vivo.
J. Exp. Med.
190
:
145
44
Van Stipdonk, M. J. B., A. A. Willems, S. Amor, C. Persoon-Deen, P. J. Travers, C. J. Boog, J. M. van Noort.
1998
. T cells discriminate between differentially phosphorylated forms of alfaB-crystallin, a major central nervous system myelin antigen.
Int. Immunol.
10
:
943
45
Levine, A. J., J. Momand, C. A. Finlay.
1991
. The p53 tumor suppressor gene.
Nature
351
:
453
46
Zhang, W., C. McClain, J. P. Gau, X. Y. Guo, A. B. Deisseroth.
1994
. Hyperphosphorylation of p53 induced by okadaic acid attenuates its transcriptional activation function.
Cancer Res.
54
:
4448
47
Milczarek, G. J., J. Martinez, G. T. Bowden.
1997
. p53 phosphorylation: biochemical and functional consequences.
Life Sci.
60
:
1
48
De Klein, A., A. G. van Kessel, G. Grosveld, C. R. Bartram, A. Hagemeijer, D. Bootsma, N. K. Spurr, N. Heisterkamp, J. Groffen, J. R. Stephenson.
1982
. A cellular oncogene is translocated to the Philadelphia chromosome in chronic myelocytic leukemia.
Nature
300
:
765
49
Clark, S. S., J. McLaughlin, M. Timmons, A. M. Pendergast, Y. Ben Neriah, L. W. Dow, W. Crist, G. Rovera, S. D. Smith, O. N. Witte.
1988
. Expression of a distinctive BCR-ABL oncogene in Ph1-positive acute lymphocytic leukemia (ALL).
Science
239
:
775
50
Lu, D., J. Liu, M. Campbell, J. Q. Guo, N. Heisterkamp, J. Groffen, E. Canaani, R. Arlinghaus.
1993
. Tyrosine phosphorylation of P160 BCR by P210 BCR-ABL.
Blood
82
:
1257
51
Liu, J., M. Campbell, J. Q. Guo, D. Lu, Y. M. Xian, B. S. Andersson, R. B. Arlinghaus.
1993
. BCR-ABL tyrosine kinase is autophosphorylated or transphosphorylates P160 BCR on tyrosine predominantly within the first BCR exon.
Oncogene
8
:
101
52
Straus, D. J..
1997
. HIV-associated lymphomas.
Curr. Opin. Oncol.
9
:
450
53
Grasser, F. A., S. Gottel, P. Haiss, B. Boldyreff, O. G. Issinger, N. Mueller Lantzsch.
1992
. Phosphorylation of the Epstein-Barr virus nuclear antigen 2.
Biochem. Biophys. Res. Commun.
186
:
1694
54
Moorthy, R. K., D. A. Thorley Lawson.
1993
. All three domains of the Epstein-Barr virus-encoded latent membrane protein LMP-1 are required for transformation of rat-1 fibroblasts.
J. Virol.
67
:
1638
55
Kolman, J. L., N. Taylor, D. R. Marshak, G. Miller.
1993
. Serine-173 of the Epstein-Barr virus ZEBRA protein is required for DNA binding and is a target for casein kinase II phosphorylation.
Proc. Natl. Acad. Sci. USA
90
:
10115