Patients with HER-2/neu–expressing breast cancer remain at risk for relapse following standard therapy. Vaccines targeting HER-2/neu to prevent relapse are in various phases of clinical testing. Many vaccines incorporate the HER-2/neu HLA-A2–binding peptide p369–377 (KIFGSLAFL), because it has been shown that CTLs specific for this epitope can directly kill HER-2/neu–overexpressing breast cancer cells. Thus, understanding how tumors process this epitope may be important for identifying those patients who would benefit from immunization. Proteasome preparations were used to determine if p369–377 was processed from larger HER-2/neu–derived fragments. HPLC, mass spectrometry, cytotoxicity assays, IFN-γ ELISPOT, and human breast cancer cell lines were used to assess the proteolytic fragments. Processing of p369–377 was not detected by purified 20S proteasome and immunoproteasome, indicating that tumor cells may not be capable of processing this Ag from the HER-2/neu protein and presenting it in the context of HLA class I. Instead, we show that other extracellular domain HER-2/neu peptide sequences are consistently processed by the proteasomes. One of these sequences, p373–382 (SLAFLPESFD), bound HLA-A2 stronger than did p369–377. CTLs specific for p373–382 recognized both p373–382 and p369–377 complexed with HLA-A2. CTLs specific for p373–382 also killed human breast cancer cell lines at higher levels than did CTLs specific for p369–377. Conversely, CTLs specific for p369–377 recognized p373–382. Peptide p373–382 is a candidate epitope for breast cancer vaccines, as it is processed by proteasomes and binds HLA-A2.

Breast cancer is one of the leading causes of cancer death for women worldwide and is the leading cancer by site for females in the United States (1, 2). This year alone it is estimated that the United States will have ∼226,870 new cases of invasive breast cancer and 39,510 deaths from breast cancer in females (1). Breast cancer is generally divided into three subtypes: estrogen receptor/progesterone receptor–positive, HER-2/neu receptor–positive, and triple-negative breast cancers. Despite recent advances in therapy for primary breast tumors, relapse remains a significant concern, particularly for the HER-2/neu and triple-negative subsets. Tumors that recur in these patients are often more aggressive, metastatic, and chemoresistant and are the leading cause of death among women with breast cancer. Survival rates for patients with recurrent breast cancer are ∼50%, whereas the 5-y survival rate for patients with primary breast cancer is almost 90% (3).

Vaccines are being developed to prevent the recurrence of breast cancer. The majority of vaccines being tested in clinical trials are designed to augment the immune response against tumor Ags, which are often overexpressed in breast tumors (4, 5). HER-2/neu, which is expressed in 20–40% of invasive breast tumors, is one Ag commonly targeted with vaccines (6). HER-2/neu is a transmembrane domain–spanning receptor with an intracellular tyrosine kinase–signaling domain. The HER2-signaling domain is activated upon heterodimerization with other members of the HER family (HER1, HER3, or HER4) or upon homodimerization (7, 8). Overexpression of this oncogenic protein leads to cellular transformation through increased proliferation and survival (7).

Although many different HER-2/neu vaccination strategies have been or are being investigated, vaccines formulated with synthetic cytotoxic T cell–activating peptide epitopes are among the most advanced (913). For example, one vaccine that has generated significant interest is composed of the adjuvant, GM-CSF, mixed with the HLA-A2–binding HER-2/neu extracellular domain (ECD)–derived peptide p369–377 (E75) (4, 5). Fisk and colleagues (14) identified p369–377 in 1995 on the basis of the presence of HLA-A2 anchoring motifs. Experiments using peptide-elicited CD8+ T cell clones or HLA-A2+ transgenic mice provided strong evidence of natural processing of p369–377 and HLA-A2 presentation (14, 15). Some controversy, however, emerged in 1998, when Zaks and Rosenberg (16) failed to demonstrate that vaccine-elicited p369–377–specific T cells could readily recognize HLA-A2+ HER-2/neu+ tumors, casting some doubt on the utility of this epitope in eliciting therapeutic T cells. However, human studies by Knutson and colleagues (17, 18) further supported the prior studies suggesting natural presentation in HLA-A2.

Subsequent to this early work, Mittendorf and colleagues (5) conducted exploratory Phase I and II clinical trials in early-stage node-negative and node-positive breast cancer patients. In the most recent 24-mo landmark analysis, Mittendorf (4) reported summary results of those trials. Of 182 evaluable patients, 24-mo disease-free survival was 94.3% in the vaccine group and 86.8% in the control. Although not reaching statistical significance (p = 0.08) when considering all evaluable patients, further analysis revealed statistically significant benefit in subsets of patients. Phase II and III clinical trials investigating p369–377 in combination with GM-CSF, with or without trastuzumab, are ongoing (http://www.clinicaltrials.gov) to further determine clinical benefit.

Considering the potential for success of the p369–377 vaccine in preventing recurrence, our group became interested in developing biomarkers that may further enhance selection of patients who would benefit from the vaccine. Because p369–377 is displayed in the context of HLA class I, our interest in the current study focused on the role of the proteasome or immunoproteasome in processing p369–377. These proteasomes are two large multisubunit complexes that are important for the processing of the vast majority of HLA class I presented epitopes (19). Once processed, the peptides released from the proteasomes are translocated into the endoplasmic reticulum via TAP1 and TAP2 and loaded onto class I molecules. Although p369–377 is an epitope that is empirically predicted to be processed by the proteasomes (A.M. Henle and K.L Knutson, unpublished observations), to the best of our knowledge, no studies have specifically shown this biochemically, which was the focus of the current study. Thus, using purified human 20S proteasome and immunoproteasome, we asked whether p369–377 could be processed from longer HER-2/neu peptides, allowing it to be available for binding HLA class I molecules. Our data suggest that p369–377 is not processed by either the 20S proteasome or the immunoproteasome. However, other peptides derived from the ECD of HER-2/neu were consistently processed by both enzymes. One newly identified peptide that is processed by both the immunoproteasome and the proteasome is p373–382, a strong HLA-A2 binder that is able to prime CD8+ T cells for HER-2/neu+ breast tumor cell recognition and lysis.

The breast cancer cell lines SK-BR3, HCC1419, MCF7, and BT20 were purchased from the American Type Culture Collection, immediately resuscitated and expanded in RPMI 1640 medium containing 10% FBS, and supplemented with 25 mM HEPES, 1.5 g/l sodium bicarbonate, 0.1 mM MEM nonessential amino acids, 1 mM sodium pyruvate, 2 mM l-glutamine, and 50 mM 2-ME (Invitrogen), at 37°C and 5% CO2. Prior to assays, the cell lines were incubated overnight with 500 U/ml IFN-γ to ensure maximum HLA class I expression. The esophageal adenocarcinoma cell line FLO-1 was a gift from Dr. Harry H. Yoon (Mayo Clinic, Rochester, MN). For verification, all tumor cell lines were tested for HLA-A status, using a Biotest HLA-A SSP kit (Dreieich, Germany), and were tested for HER-2/neu status, using RT-PCR [primer sequences: F: (5′-GCTCTTTGAGGACAACTATGCCC-3′) R: (5′-GCCCTTACACATCGGAGAACAG-3′)]. Tumor cells were frozen as seeding stocks. For the assays, the cells were thawed rapidly in a 37°C water bath and expanded in RPMI 1640 medium, as described above. Because HER-2/neu expression has been associated with loss of class I expression in some models (20, 21), the HLA-A2 and HER-2/neu status of each cell line was verified by flow cytometry following exposure to IFN-γ, as we have previously described (22), using mouse anti-human HLA-A2 FITC and anti HER-2/neu FITC Abs. Cognate isotypes were, respectively, mouse IgG2b,κ FITC isotype and mouse IgG1,κ FITC isotype (BD Pharmingen).

The peptides used in this study were the OVA peptide, SIINFEKL; the HLA-A2–binding influenza matrix protein M1 peptide p58–66 (pFLU), GILGFVFTL; and the HER-2/neu peptides p369–377, KIFGSLAFL; p368–376, KKIFGSLAF; p372–380, GSLAFLPES; p364–373, FAGCKKIFGS; p373–382, SLAFLPESFD; p364–382, FAGCKKIFGSLAFLPESFD; and p362–384, QEFAGCKKIFGSLAFLPESFDGD. All peptides were manufactured by the Mayo Clinic Proteomics Core or Elim Biopharmaceuticals and were >80% pure, as assessed by HPLC and mass spectrometric analysis. The HER-2/neu ECD was purified by trastuzumab affinity chromatography from culture supernatants of baby hamster kidney cell–produced ECD, as previously described by us and others (23, 24).

Potential HLA-A2–binding peptides from large HER-2/neu peptide sequences were predicted using a matrix pattern that calculates how peptides bind to the HLA-A2 motif based on previously published peptide motifs. The prediction algorithm is “SYFPEITHI: database for MHC ligands and peptide motifs” (25). It can be accessed via www.syfpeithi.de.

The C-Term 3.0 and 20S prediction networks on the Netchop 3.1 Server (www.cbs.dtu.dk/services/NetChop/), as well as proteasome and immunoproteasome models 1, 2, and 3 on the Proteasome Cleavage Prediction Server (http://imed.med.ucm.es/Tools/pcps/index.html), were used to identify the cleavage of potential HLA-A2–binding peptides from large HER-2/neu peptide sequences. Results from all possible cleavage prediction methods were compiled for the study.

The lymphoblastic cell line T2 (American Type Culture Collection) was maintained at 37°C and 5% CO2 in IMDM supplemented with 20% FBS. For stabilization assessment, T2 cells were washed, counted, and resuspended at 1 × 106 cellsper milliliter in IMDM in a 24-well plate and pulsed with a serial dilution of 0.2 − 100 μM peptide for 16 h at room temperature. Subsequently, cells were washed 3× in PBS/0.5% BSA and stained with anti–HLA-A2 FITC Ab (clone BB7.2; BD Pharmingen) or isotype control mouse IgG2b,κ FITC Ab (BD Pharmingen) and analyzed by flow cytometry on a FACScan with CellQuest software (BD Biosciences). Influenza peptide (pFLU, GILGFVFTL; see Ref. 26) served as a positive control. SIINFEKL served as the negative control peptide.

Whole-blood samples obtained from Leuko Reduced Separator Chambers (Trima) from the Department of Transfusion Medicine at Mayo Clinic were lysed in ammonium-chloride-potassium buffer and separated over Ficoll-Paque (GE Healthcare). A total of 1 × 106 cells from the PBMC fraction were stained with mouse anti-human HLA-A2 FITC BB7.2 Ab (BD Pharmingen) or mouse IgG2b,κ FITC Ab (BD Pharmingen) and analyzed with a FACScan flow cytometer (BD Biosciences). PBMC samples that stained positive for HLA-A2 were further enriched with the CD8+ T Cell Isolation Kit II (Miltenyi Biotec) and separated by AutoMACS (Miltenyi Biotec) per the manufacturer’s protocol.

Enriched CD8+ T cells were plated in 2 ml media in six-well plates at 2 × 106 cells per milliliter. Peptide was added at a concentration of 10 μg/ml. Remaining CD8 PBMCs were frozen for future restimulations. On days 3 and 5, human rIL-2 was added to the wells at 50 U/ml, and human rIFN-γ was added at 100 U/ml. On day 7, cryopreserved PBMCs were thawed, washed three times, pulsed with 10 μg/ml peptide for 2 h, and irradiated to 4000 rads. The PBMCs were washed and suspended in fresh media and then added to the CD8+ T cells. One additional restimulation was done on day 14. The cells were used for assays on days 21–28.

An IFN-γ ELISPOT assay was performed as described previously (18). Briefly, cultured CD8+ T cells were incubated at 37°C for 48 h with breast cancer cell lines or PBMCs that had been pulsed with 100 μM peptide at 37°C and 5% CO2 for 16 h, media alone, or PMA–ionomycin. Respective wells containing target cells were, in some assays, incubated with blocking Abs for 1 h at a concentration of 1 μg/100 μl RPMI + 10% FBS media prior to CD8+ T cell addition. All Abs were from BD Pharmingen (purified mouse anti-human HLA-A2 BB7.2; purified mouse anti-human HLA-ABC W6/32; purified mouse IgG2b,κ isotype control; and purified IgG2a,κ isotype control).

Tumor cells were lysed using an impedance-based approach as previously described (22, 27). Then 100 μl RPMI 1640 (Mediatech) with 10% FBS was added to each well in an E-Plate 16 (Roche). Background impedance on the plate was measured on the xCELLigence RTCA SP instrument (Roche) at 37°C and 5% CO2. Cells from human tumor cell lines (SK-BR3, MCF7, BT20, FLO, and HCC1419) were harvested with trypsin, counted, washed, and resuspended in RPMI 1640 with 10% FBS and 500 U/ml human IFN-γ (PeproTech) at 5 × 104 cells per milliliter. IFN-γ was used to ensure maximal upregulation of HLA class I molecules. A total of 100 μl tumor cells was added to each well of the E-Plate 16, which was then placed in the xCELLigence RTCA DP. Impedance was measured every 5 min for ∼20 h at 37°C and 5% CO2 (until the cells adhered to the gold electrodes on the bottom of each well and proliferated to ∼1 × 104 cells per well). T cells were counted and resuspended at a concentration of 5 × 105 cells per milliliter in RPMI 1640 + 10% FBS media. Then 100 μl T cells or media alone was added to the respective wells. The E-plate 16 was placed in the xCELLigence RTCA SP, and impedance measurements were recorded every 5 min for >10 additional hours at 37°C and 5% CO2. T cell–mediated death of tumor cells was monitored in real time and was indicated by a decrease in cell index. Data were analyzed with RTCA Software 1.2 (Acea Biosciences). Results were normalized 1–2 h after T cell addition. Cytolytic activity was calculated as the percentage of cytolysis 10 h after the normalization time (= [CIno effector − CIeffector]/CIno effector × 100). For HLA class I blocking, Abs were added at a concentration of 1 μg/100 μl RPMI 1640 + 10% FBS media.

Purified human 20S proteasome, immunoproteasome, PA28 activator α subunit, and synthetic lactacystin were obtained from Boston Biochem. Proteasome and immunoproteasome enzymes were assayed for activity with the fluorescent Suc-LLVY-AMC substrate (Boston Biochem) per the manufacturer’s protocol. Cleavage assays were performed in 400 μl reaction buffer (25 mM HEPES, 0.5 mM EDTA, pH 7.6; Boston Biochem). PA28 activator (stock 20 μM) was added to the reaction buffer, to a final concentration of 14.5 nM. Human proteasome or immunoproteasome enzymes (stock 2 μM) were added to the reaction buffer, to a final concentration of 1.45 nM. A 10 mM peptide substrate (HER-2/neu, 19 mer or 23 mer) was diluted in the reaction mix 1:1000. Some samples had lactacystin added to the reaction, to a final concentration of 10 μM, to inhibit the proteasome activity. Blanks without proteasome or immunoproteasome were run in parallel. Samples were incubated in a 37°C water bath for 30–60 min. Preliminary studies suggested no difference in the cleavage products detected between 30 min and 1 h (data not shown). Samples were then filtered using prewashed YM-30 30,000-kDa MWCO microcon filters (Millipore) at 14,000 × g for 15 min. Filter retentates were collected by inverted centrifugation at 3000 × g for 3 min.

Cleavage products (5 μl reaction mixture) from the HER-2/neu peptide incubations were separated and identified using an Agilent 1100 HPLC system interfaced to an Agilent MSD/TOF mass spectrometer. The products were resolved on an Agilent Zorbax 300SB-C18 column (1 × 150 mm; 3.5 μ; 45°C), using mobile phases containing H2O, acetonitrile, isopropanol, and formic acid [Pump A: 98:1:1:0.1 and Pump B: 10:80:10:0.1 (v/v/v/v)]. The separation gradient was 5–60% B over 23 min; 80% B for 5 min; with a 7-min equilibration at 5% B, using a flow rate of 60 μL/min. Mass spectra were collected in positive mode, using an electrospray ionization interface over an m/z range of 300–2800. Other instrument parameters used were spray voltage, 3500 V; fragmentor, 175 V; skimmer, 65 V; radio frequency octopole, 250 V; gas temperature, 325°C; gas flow, 5 L/min; and nebulizer gas, 30 ψ. Raw spectra were obtained and peaks transformed to molecular masses using Agilent MassHunter Qualitative Analysis with BioConfirm software (version B.02.00). The observed masses were then compared with the theoretical masses of the known HER-2/neu peptide sequences and assigned a sequence.

Statistical analyses were performed using GraphPad Prism 5. Data were analyzed using one-way ANOVA, Tukey, Mann–Whitney, or Student t tests as stated in the figure legends, and the results were considered statistically significant if p < 0.05.

A 19 aa sequence (p364–382, FAGCKKIFGSLAFLPESF) derived from the ECD of HER-2/neu was synthesized to study whether the proteasome and immunoproteasome could cleave the HLA-A2 HER-2/neu epitope, p369–377. Processing studies using longer peptides are common in the cancer epitope discovery field and provide greater detection in vitro compared with cleavage assays using full-length rHER-2/neu protein (2830). HER-2/neu p369–377 is embedded in the synthesized 19 mer, with an extra 5 aa on both the N- and the C-termini.

Purified proteasome and immunoproteasome enzymes were individually incubated with the 19-mer substrate and PA28 activator. Cleaved products were analyzed via liquid chromatography and mass spectrometry (Fig. 1A). Several peptides between 8 and 10 aa in length, the appropriate length for binding to HLA class I molecules, were detected in the cleaved samples. However, p369–377 was not detected in any of the samples (Fig. 1B). These results suggest that p369–377 is not a reaction product of the proteasome or immunoproteasome.

FIGURE 1.

The proteasome and immunoproteasome fragment synthetic HER-2/neu–derived peptides into multiple shorter peptides, not including p369–377. (A) Total ion chromatograms for the 19 mer in reaction buffer, 19 mer + PA28 activator + immunoproteasome, 19 mer + PA28 activator + proteasome, and 19 mer + PA28 activator + proteasome + lactacystin. *Starting 19-mer FAGCKKIFGSLAFLPESFD; ** and *** indicate deletion products from the peptide synthesis. The numbered peaks are notable peptides identified by accurate monoisotopic mass as fragment peptides of the 19 mer and correspond to the sequences in Table I. Synthesized 9-mer p369–377 is also shown. Other peaks observed on the baseline are background products from the assay. (B) Extracted ion chromatograms for HER-2/neu p369–377 KIFGSLAFL-OH fragment for 19 mer + PA28 activator + immunoproteasome and 19 mer + PA28 activator + proteasome. HER-2/neu p369–377 was not detected.

FIGURE 1.

The proteasome and immunoproteasome fragment synthetic HER-2/neu–derived peptides into multiple shorter peptides, not including p369–377. (A) Total ion chromatograms for the 19 mer in reaction buffer, 19 mer + PA28 activator + immunoproteasome, 19 mer + PA28 activator + proteasome, and 19 mer + PA28 activator + proteasome + lactacystin. *Starting 19-mer FAGCKKIFGSLAFLPESFD; ** and *** indicate deletion products from the peptide synthesis. The numbered peaks are notable peptides identified by accurate monoisotopic mass as fragment peptides of the 19 mer and correspond to the sequences in Table I. Synthesized 9-mer p369–377 is also shown. Other peaks observed on the baseline are background products from the assay. (B) Extracted ion chromatograms for HER-2/neu p369–377 KIFGSLAFL-OH fragment for 19 mer + PA28 activator + immunoproteasome and 19 mer + PA28 activator + proteasome. HER-2/neu p369–377 was not detected.

Close modal
Table I.
Peptide fragments processed from HER-2/neu p364–382, FAGCKKIFGSLAFLPESFD
Label on FigureHER-2/neu Amino Acid NumberRetention Time (min)PeptidePeptide Generated by ImmunoproteasomePeptide Generated by ProteasomeHLA-A*0201 Binding ScoreaCleavage Predicted by IP/P Serversb
371–381 20.4 FGSLAFLPESF NA 
364–374 15.45 FAGCKKIFGSL NA 
366–376 15.45 GCKKIFGSLAF NA 
372–382 18.4 GSLAFLPESFD NA 
373–382 18.5 SLAFLPESFD 13 
371–380 19.1 FGSLAFLPES 
372–381 19.1 GSLAFLPESF 
364–373 12.8 FAGCKKIFGS 10 
374–382 18.2 LAFLPESFD 
373–381 19.2 SLAFLPESF 16 
− 369–377 ND KIFGSLAFL − − 28 
375–382 16.06 AFLPESFD NA 
364–382 19.14 FAGCKKIFGSLAFLPESFD-NH4 NA NA NA NA 
364–382 19.14 FAGCKKIFGSLAFLPESFD-COOH NA NA NA NA 
** NA 19.54 FAGKKIFGSLAFLPESFD-NH4 NA NA NA NA 
** NA 19.54 FAGKKIFGSLAFLPESFD-COOH NA NA NA NA 
*** NA 14.78 FAGKKIFGSL NA NA NA NA 
*** NA 14.78 GKKIFGSLAF NA NA NA NA 
Label on FigureHER-2/neu Amino Acid NumberRetention Time (min)PeptidePeptide Generated by ImmunoproteasomePeptide Generated by ProteasomeHLA-A*0201 Binding ScoreaCleavage Predicted by IP/P Serversb
371–381 20.4 FGSLAFLPESF NA 
364–374 15.45 FAGCKKIFGSL NA 
366–376 15.45 GCKKIFGSLAF NA 
372–382 18.4 GSLAFLPESFD NA 
373–382 18.5 SLAFLPESFD 13 
371–380 19.1 FGSLAFLPES 
372–381 19.1 GSLAFLPESF 
364–373 12.8 FAGCKKIFGS 10 
374–382 18.2 LAFLPESFD 
373–381 19.2 SLAFLPESF 16 
− 369–377 ND KIFGSLAFL − − 28 
375–382 16.06 AFLPESFD NA 
364–382 19.14 FAGCKKIFGSLAFLPESFD-NH4 NA NA NA NA 
364–382 19.14 FAGCKKIFGSLAFLPESFD-COOH NA NA NA NA 
** NA 19.54 FAGKKIFGSLAFLPESFD-NH4 NA NA NA NA 
** NA 19.54 FAGKKIFGSLAFLPESFD-COOH NA NA NA NA 
*** NA 14.78 FAGKKIFGSL NA NA NA NA 
*** NA 14.78 GKKIFGSLAF NA NA NA NA 

Numbers and asterisks indicate peptide labels in Fig. 1A.

a

The SYFPEITHI server was used to predict nonamer and decamer peptide binding to HLA-A*0201.

b

The 20S and C-term 3.0 prediction methods on the Netchop 3.1 server and the Proteasome Cleavage Prediction Server with models 1, 2, and 3 for the proteasome and immunoproteasome enzymes were used to predict whether the smaller peptides could be processed by the enzymes from the larger 23-mer sequence, irrespective of our in vitro data.

+, The peptide was produced by the respective enzyme; –, lack of peptide detection in processed samples; NA, not applicable: Peptide is either a deletion product, is starting material, or is too large for binding predictions to HLA-A*0201; ND, not detected.

Because peptides other than p369–377 were detected as degradation products, we questioned whether any of these peptides may serve as potential candidates for HER-2/neu cancer immunotherapies. Thus, an HLA binding prediction server, SYFPEITHI, was used to predict the ability of these other peptides to bind HLA-A2 molecules (Table I). Several peptides scored higher than 10, suggesting that these epitopes may bind HLA-A2. For comparison, p369–377 had a score of 28, indicating that it is predicted to bind strongly to HLA-A2, a finding consistent with prior studies (5). Proteasome and immunoproteasome cleavage servers were also used to compare algorithm predictions with the in vitro results. The cleavage predictions from the algorithms did not always correspond to the in vitro biochemical data (Table I). Specifically, the algorithm predictions aligned with the in vitro cleavage data in 6 of 12 (50%) peptides (Table I).

Because the algorithms predicted that many degradation peptides may serve as targets that could be displayed in HLA class I on the surface of breast cancer cells, T2 HLA-A2 stabilization assays were performed. Influenza matrix protein M1 peptide was used as the positive control because it is known to bind HLA-A2 strongly and result in HLA-A2 stabilization on the surface of T2 cells. SIINFEKL was used as the negative control, as it does not bind HLA-A2. Four peptides (p368–376, KKIFGSLAF; p372–380, GSLAFLPES; p364–373, FAGCKKIFGS; and p373–382, SLAFLPESFD) identified via the in vitro proteasome and immunoproteasome assays were synthesized and tested for their binding to HLA-A2. Only the p369–377 epitope and another epitope, p373–382, resulted in increased surface levels of HLA-A2, indicating that these peptides are able to bind A2 and stabilize the complex (Fig. 2A and data not shown [nonbinding HER-2/neu peptides]). Titration studies with each peptide were conducted to compare the peptide affinity for HLA-A2. The p373–382 epitope bound HLA-A2 strongly, resulting in HLA-A2 stabilization at concentrations as low as 200 nM. In contrast, p369–377 did not begin to stabilize HLA-A2 until the peptide was at a concentration of ∼6–7 μM (Fig. 2B). Pulsing with 50 μM of the p373–382 peptide consistently resulted in increased surface expression of the HLA-A2 molecule on T2 cells, at levels similar to those of the positive control pFLU peptide (Fig. 2C). The p369–377 epitope stabilized HLA-A2 at ∼50% of the maximal stabilization seen with pFLU HLA-A2 upregulation. The concentration of peptide that induced half-maximal stabilization of HLA-A2 was ∼0.94 μM for pFLU, ∼1.8 μM for p373–382, and ∼9.6 μM for p369–377 (Fig. 2D).

FIGURE 2.

HER-2/neu p373–382 stabilizes HLA-A2 on T2 cells. (A) Shown are composite histograms of HLA-A2 levels on T2 cells pulsed with 50 μM various peptides [Influenza peptide, pFLU, served as a positive control, cells pulsed with OVA (SIINFEKL) peptide as a negative control (nonbinding peptide)]. Unpulsed T2 cells were stained with isotype Ab. One representative experiment of 10 is shown. (B) Shown is the fold change in HLA-A2 staining intensity on T2 cells pulsed with a serial dilution of 0.195 − 100 μM peptide. Results are calculated as the ratio of sample fluorescence relative to fluorescence obtained in the absence of peptide, and presented as the mean and SEM of 10 experiments. (C) Shown are the mean (± SEM, n = 10) fluorescence intensities of T2 cells pulsed with 50 μM peptide relative to the maximal response seen with pFLU. ***p < 0.001, calculated using one-way ANOVA with the Tukey multiple comparison test. (D) Shown are EC50 (i.e., half-maximal effective concentration) curves depicting concentration-dependent HLA-A2 stabilization over various concentrations of peptide (n = 10). Inset data show extrapolated EC50 values calculated using a one-site saturation model.

FIGURE 2.

HER-2/neu p373–382 stabilizes HLA-A2 on T2 cells. (A) Shown are composite histograms of HLA-A2 levels on T2 cells pulsed with 50 μM various peptides [Influenza peptide, pFLU, served as a positive control, cells pulsed with OVA (SIINFEKL) peptide as a negative control (nonbinding peptide)]. Unpulsed T2 cells were stained with isotype Ab. One representative experiment of 10 is shown. (B) Shown is the fold change in HLA-A2 staining intensity on T2 cells pulsed with a serial dilution of 0.195 − 100 μM peptide. Results are calculated as the ratio of sample fluorescence relative to fluorescence obtained in the absence of peptide, and presented as the mean and SEM of 10 experiments. (C) Shown are the mean (± SEM, n = 10) fluorescence intensities of T2 cells pulsed with 50 μM peptide relative to the maximal response seen with pFLU. ***p < 0.001, calculated using one-way ANOVA with the Tukey multiple comparison test. (D) Shown are EC50 (i.e., half-maximal effective concentration) curves depicting concentration-dependent HLA-A2 stabilization over various concentrations of peptide (n = 10). Inset data show extrapolated EC50 values calculated using a one-site saturation model.

Close modal

Although the p373–382 peptide was released from the 19-mer peptide by both the proteasome and the immunoproteasome and can bind HLA-A2, the epitope was located on the C-terminal end of the 19 mer, and therefore cleavage between two adjacent amino acids was not possible to measure. However, in both the proteasome and the immunoproteasome samples in Fig. 1, the potential for processing of the 19 mer into the p373–382 decamer was detected by monitoring C-terminal end conversion from an amide to a free acid (not shown).

To determine if p373–382 could be internally processed, a 23-mer peptide (p362–384, QEFAGCKKIFGSLAFLPESFDGD) from HER-2/neu was synthesized. The in vitro proteasome and immunoproteasome assays were repeated with the 23 mer. Although proteolysis was generally less efficient with the longer peptide, liquid chromatography and mass spectrometry showed that several peptide epitopes were detected in these samples (Supplemental Fig. 1A, Supplemental Table I). Again, the p369–377 epitope was not detected in the proteasome or the immunoproteasome samples (Supplemental Fig. 1B). Extracted ion chromatograms showed that the p373–382 decamer was processed in both the proteasome and, to a lesser extent, the immunoproteasome samples (Fig. 3). Although lower levels of the decamer were detected in the lactacystin-inhibited reactions compared with the uninhibited proteasome reaction, generation of p373–382 was still present, suggesting the proteasome uses both lactacystin-resistant and -sensitive catalysis to liberate the peptide. Overall comparison shows that the peptide products generated by proteolysis from the 23-mer peptide were similar to the products obtained with the 19-mer peptide (Figs. 3B, 3C).

FIGURE 3.

HER-2/neu p373–382 is processed from HER-2/neu 23 mer by the immunoproteasome and proteasome. (A) Extracted ion chromatograms examining for the HER-2/neu p373–382 fragment from the 23 mer in reactions containing (top to bottom) reaction buffer, immunoproteasome, proteasome, and proteasome + lactacystin. Also shown is purified HER-2/neu 10 mer p373–382. (B) Predominant cleavage products from digestion of human HER-2/neu 23-mer peptide 362–384 with 20S proteasome and immunoproteasome. Peptide fragments were detected by HPLC interfaced with electrospray ionization time-of-flight mass spectrometry. Numbers above amino acid residues indicate the number of cleavages occurring at that site. Numbers below designate amino acid residue positions in the HER-2/neu protein. (C) Predominant cleavage products from digestion of human HER-2/neu 19-mer peptide 364–382 with 20S proteasome and immunoproteasome.

FIGURE 3.

HER-2/neu p373–382 is processed from HER-2/neu 23 mer by the immunoproteasome and proteasome. (A) Extracted ion chromatograms examining for the HER-2/neu p373–382 fragment from the 23 mer in reactions containing (top to bottom) reaction buffer, immunoproteasome, proteasome, and proteasome + lactacystin. Also shown is purified HER-2/neu 10 mer p373–382. (B) Predominant cleavage products from digestion of human HER-2/neu 23-mer peptide 362–384 with 20S proteasome and immunoproteasome. Peptide fragments were detected by HPLC interfaced with electrospray ionization time-of-flight mass spectrometry. Numbers above amino acid residues indicate the number of cleavages occurring at that site. Numbers below designate amino acid residue positions in the HER-2/neu protein. (C) Predominant cleavage products from digestion of human HER-2/neu 19-mer peptide 364–382 with 20S proteasome and immunoproteasome.

Close modal

Thus far, the results suggest that p373–382 can be processed by the proteasome and immunoproteasome and that it binds HLA-A2 strongly, suggesting potential clinical relevance. To address this, blood was obtained from human donors, and short-term CD8+ T cell lines were generated against pFLU, p369–377, and p373–382. IFN-γ ELISPOT analysis using autologous peptide pulsed PBMCs as targets demonstrated specificity of the lines (Fig. 4A). The pFLU T cell line recognized only pFLU and none of the other peptide pulsed PBMC targets. The p369–377 pulsed PBMCs were recognized by the p369–377 CD8+ T cells and, unexpectedly, also by the p373–382 T cells. Similarly, the p373–382 pulsed PBMCs were recognized by the p373–382–specific CD8+ T cells but also by the p369–377–specific T cells. Thus, the results show that T cells responsive to p373–382 exist in the human T cell repertoire.

FIGURE 4.

HER-2/neu p373–382–specific CD8+ T cells have increased recognition of human breast cancer cell lines. (A and B) Shown are the mean (± SEM, n = 3) numbers of ELISPOTs obtained in response to pFLU, p369–377, or p373–382 peptide-specific CD8+ T cells derived from short-term culture to (A) autologous PBMCs pulsed with pFLU, p369–377, or p373–382 or (B) irradiated tumor cells. Results are from a single donor and are representative of similar results obtained with three other donors. (C) Shown are the mean (± SEM, n = 4) % lysis values calculated from the impedance-based lysis assay at 10 h following the addition of T cells. Also shown are the expression levels of HER-2/neu and HLA-A2 genotyping results for each of the cell lines. Results are pertinent to (B) and (C). HER-2/neu overexpression levels in each cell line and HLA-A2 expression are indicated below the x-axis. Expression levels were determined by flow cytometry and RT-PCR. The p values in (B) and (C) were calculated using an unpaired Student t test. (D) Shown are the mean (± SEM, n = 3) numbers of IFN-γ ELISPOTs obtained in response to pFLU or p373–382 peptide-specific CD8+ T cells derived from short-term culture to autologous PBMCs pulsed with pFLU, p373–382, HER-2/neu ECD, or OVA protein. The experiment was repeated twice in triplicate, with both experiments yielding similar results.

FIGURE 4.

HER-2/neu p373–382–specific CD8+ T cells have increased recognition of human breast cancer cell lines. (A and B) Shown are the mean (± SEM, n = 3) numbers of ELISPOTs obtained in response to pFLU, p369–377, or p373–382 peptide-specific CD8+ T cells derived from short-term culture to (A) autologous PBMCs pulsed with pFLU, p369–377, or p373–382 or (B) irradiated tumor cells. Results are from a single donor and are representative of similar results obtained with three other donors. (C) Shown are the mean (± SEM, n = 4) % lysis values calculated from the impedance-based lysis assay at 10 h following the addition of T cells. Also shown are the expression levels of HER-2/neu and HLA-A2 genotyping results for each of the cell lines. Results are pertinent to (B) and (C). HER-2/neu overexpression levels in each cell line and HLA-A2 expression are indicated below the x-axis. Expression levels were determined by flow cytometry and RT-PCR. The p values in (B) and (C) were calculated using an unpaired Student t test. (D) Shown are the mean (± SEM, n = 3) numbers of IFN-γ ELISPOTs obtained in response to pFLU or p373–382 peptide-specific CD8+ T cells derived from short-term culture to autologous PBMCs pulsed with pFLU, p373–382, HER-2/neu ECD, or OVA protein. The experiment was repeated twice in triplicate, with both experiments yielding similar results.

Close modal

To address whether T cells generated by p373–382 recognize HER-2/neu–expressing breast cancer cells, a panel of HER-2/neu and HLA-A2–expressing breast cancer cell lines was used in an ELISPOT assay with the three different T cell lines (Fig. 4B). With IFN-γ ELISPOT, the activities of the lines were compared. Despite equivalency of the cell lines, in terms of numbers of specific cells, as assessed in the ELISPOT using autologous targets, T cells generated by p373–382 generally recognized breast cancer cells at higher levels than did p369–377–specific T cells. BT20 cells served as the negative control cell line, as they are HLA-A2 negative. These results indicate that cancer cells may be processing and presenting the p373–382 peptide at their surface in the context of HLA-A2. The CD8+ T cell lines were also tested for cytotoxicity. Similar to the IFN-γ ELISPOT results, the p373–382 T cells lysed the cancer cell lines at higher levels than did the p369–377 T cells. The BT20 (HLA-A2 negative) and FLO (HER-2/neu negative and HLA-A2 positive) cell lines served as negative controls (Fig. 4C).

To further demonstrate that p373–382 is a naturally processed epitope from HER-2/neu, we pulsed autologous PBMCs with HER-2/neu p373–382, pFLU, HER-2/neu ECD, and OVA protein and assayed for IFN-γ+ ELISPOT with HER-2/neu p373–382–specific CD8+ T cells and pFLU-specific CD8+ T cells. CD8+ T cells specific for p373–382 had a strong response against p373–382 pulsed PBMCs and against HER-2/neu ECD pulsed PBMCs (Fig. 4D). The amount of nonspecific activity displayed by the T cells was minimal, as control pFLU-specific T cells had little recognition of targets pulsed with HER-2/neu peptide and OVA, and similarly, p373–382–specific T cells recognized control OVA and pFLU pulsed targets at low levels.

To address whether the IFN-γ response of p373–382–generated T cells was HLA class I restricted, T cell lines were generated against all three peptides, pFLU, p369–377, and p373–382, followed by IFN-γ ELISPOT testing in the presence or absence of a blocking anti–HLA class I H chain Ab. The pFLU T cells did not recognize p373–382 pulsed PBMC targets. As expected, both p369–377 and p373–382–generated T cells recognized p373–382 pulsed PBMCs with IFN-γ release (Fig. 5A). The addition of anti-HLA class I H chain (anti–HLA-ABC) Ab resulted in a significant 90% reduction in the numbers of cells responding with IFN-γ. The same cell line was tested several days later, replacing anti–HLA-ABC with specific anti–HLA-A2 blocking Ab. As shown in Fig. 5B, anti–HLA-A2 blocked IFN-γ release to the same extent as anti–HLA-ABC, suggesting that HLA-A2 is the primary mediator of reactivity. Also shown in Fig. 5A and 5B, the number of p369–377 T cells releasing IFN-γ in response to p373–382 pulsed targets decreased completely with the addition of anti–HLA-ABC and anti–HLA-A2 Abs, further supporting cross-reactivity of the p369–377 and p373–382 T cell lines. Consistent with the ELISPOT analysis, blocking of HLA class I on the surface of SKBR3 breast cancer cells nearly abolished lysis of the tumor cells by both p369–377 T cells and p373–382 T cells (Fig. 5C).

FIGURE 5.

IFN-γ and lytic responses of p373–382–generated T cells are HLA class I restricted. (A and B) Shown are the mean (± SEM, n = 3) numbers of p373–382–specific ELISPOTs obtained in response to p369–377 or p373–382 peptide-specific CD8+ T cells derived from short-term culture to autologous PBMCs pulsed with p369–377 or p373–382 in the presence or absence of (A) HLA-ABC blocking Ab or (B) HLA-A2 blocking Ab. One representative experiment of three is shown. (C) Shown are the mean (± SEM, n = 4) % lysis values calculated from the impedance-based lysis assay at 10 h following the addition of T cells. The target cells were pretreated with or without 10 μg/ml HLA-ABC blocking Ab or HLA-A2 blocking Ab and appropriate isotype control Abs 1 h prior to T cell addition. Similar results were obtained with T cells generated from three other donors. The p values were calculated using a nonparametric Mann–Whitney U test.

FIGURE 5.

IFN-γ and lytic responses of p373–382–generated T cells are HLA class I restricted. (A and B) Shown are the mean (± SEM, n = 3) numbers of p373–382–specific ELISPOTs obtained in response to p369–377 or p373–382 peptide-specific CD8+ T cells derived from short-term culture to autologous PBMCs pulsed with p369–377 or p373–382 in the presence or absence of (A) HLA-ABC blocking Ab or (B) HLA-A2 blocking Ab. One representative experiment of three is shown. (C) Shown are the mean (± SEM, n = 4) % lysis values calculated from the impedance-based lysis assay at 10 h following the addition of T cells. The target cells were pretreated with or without 10 μg/ml HLA-ABC blocking Ab or HLA-A2 blocking Ab and appropriate isotype control Abs 1 h prior to T cell addition. Similar results were obtained with T cells generated from three other donors. The p values were calculated using a nonparametric Mann–Whitney U test.

Close modal

HER-2/neu peptide p369–377 was one of the earliest tumor Ag–derived cytotoxic T cell epitopes to have been identified (14). On the basis of successfully moving this epitope into human clinical use, we chose to determine whether this epitope is cleaved for HLA class I presentation by the immunoproteasome or proteasome. Despite a positive prediction result with the proteasome and immunoproteasome cleavage servers, cleavage of the peptide was not detected in the in vitro proteasome and immunoproteasome assays. In contrast, an HPLC comigrating peptide, HER-2/neu p373–382, was consistently observed. This peptide was able to generate Ag-specific CD8+ CTLs from the peripheral blood of HLA-A2+ normal donors that effectively killed HER-2/neu–expressing tumor cells. HER-2/neu p369–377–specific CD8+ CTLs were also generated from the peripheral blood of HLA-A2+ normal donors, using the p369–377 peptide, but these did not recognize HLA-A2+ breast cancer cells as strongly as did p373–382–specific CTLs. These results suggest that HER-2/neu–positive cancer cells produce p373–382 via the proteasome and immunoproteasome. HER-2/neu p373–382 may be an effective candidate to induce HER-2/neu–specific effector CD8+ T cell responses in vivo in patients.

As noted, HER-2/neu p369–377 was not detected in the in vitro proteasome and immunoproteasome assays. Despite an abundance of prior work demonstrating its potential therapeutic utility, the reason that p369–377 was not detected is unclear, but it could be because the proteasome and immunoproteasome, which both frequently cleave at residues in the middle of this sequence, rapidly destroyed the epitope upon production. The observation, however, that products spanning the K–K bond in the substrate peptides accumulated among the reaction products, suggests that C-terminal processing does not occur. Our data do not rule out the possibility that p369–377 is generated in the cell by other mechanisms; such mechanisms involve serine and cysteine proteases, as well as aminopeptidases, which trim the N-terminal amino acids of peptides in the cytosol and ER (31), all of which may act upon longer HER-2/neu sequences to yield the p369–377 epitope.

Unlike the case with other epitopes, the processing of HER-2/neu p373–382 was not completely blocked in the in vitro assays in the presence of the proteasome inhibitor lactacystin. However, lactacystin inhibits only the trypsin, chymotrypsin, and caspase-like activities of the proteasome. The proteasome also has other lactacystin-resistant catalytic activities, including peptidyl-glutamyl peptide hydrolyzing and branched-chain amino acid–preferring activities, which could aid in processing of HER-2/neu p373–382 and account for some of the product detected in the presence of the inhibitor (29, 3133).

When tested in stabilization assays, the affinity of p373–382 for binding to HLA-A2 molecules was comparable to that of the positive control peptide from influenza matrix protein, pFLU. The observation that p369–377 stabilized HLA-A2 is consistent with previous reports (34, 35). However, the EC50 for binding of p369–377 to HLA-A2 is significantly higher than that for both pFLU and p373–382, indicating that p373–382 has an increased binding affinity for HLA-A2 molecules. This stronger affinity for HLA-A2 may provide an explanation for the enhanced recognition of breast cancer cell lines by HER-2/neu p373–382–specific CTLs; the p373–382 peptide may remain bound to the HLA-A2 molecules on cancer cells for a longer time, sustaining presentation to CTLs. This suggestion is consistent with prior studies showing a strong correlation between peptide:HLA complex stability and immunogenicity (3639).

Evidence is provided for cross-reactivity of CTL lines generated by each of the two peptides that overlap in their sequences by 5 aa. HER-2/neu p373–382–specific CTLs recognized PBMCs pulsed with p373–382 and, to a slightly lesser extent, PBMCs pulsed with p369–377. Vice versa, HER-2/neu p369–377–specific CTLs recognized PBMCs pulsed with p369–377 and PBMCs pulsed with p373–382. A significant number of data show that T cells can be cross reactive (4045). One recent study elegantly showed that a single autoimmune TCR is able to recognize more than a million different decamer peptides in the context of a single MHC class I molecule (45). In that study, several epitopes identified were much better agonists than the wild-type index peptide, despite having only a few amino acids in common. Several mechanisms of cross-reactivity have been identified that could explain the cross-reactivity between p373–382 and p369–377 observed in the current study (46). Although the traditional view would be that these shared 5 aa would not occupy the same position in the closed binding cleft in HLA class I molecules, recent data suggest fluidity and motion that may allow the different peptide:HLA complexes to assume (i.e., tune) similar three-dimensional structures that can be seen by the TCRs (42).

The binding of HER-2/neu p373–382 to HLA-A2 partially coincides with the identified anchor residues at positions 2 and 9 for nonamers (47). Leucine is a dominant amino acid residue that occupies position 2 of HLA-A2 motifs, and it is present in the second position of p373–382 while absent from position 2 of p369–377. Valine and leucine are frequently found in position 9 of restricted nonomer epitopes. A leucine is found in position 9 of the p369–377 epitope, but neither leucine nor valine is found in position 9 of p373–382. Rather, p373–382 contains a terminal aspartic acid that is very rarely observed in the terminal position of an HLA-A2 epitope. Our search of the epitope database at the Immune Epitope Database and Analysis Resource (http://www.immuneepitope.org), at the time of manuscript preparation, reveals only ∼6 HLA-A2 epitopes with terminal aspartic acid residues. Arguably, much of the information that has been used to populate these and other databases has been based on prediction algorithms derived from a limited set of data obtained from elution studies or studies in which crystal structures of HLA class I peptide complexes have been solved (48). Despite that, however, it is clear from higher throughput studies that leucine and valine are favored at position 9 in nonomeric epitopes (48).

What has become more appreciated in recent times is that HLA-A2 can present longer epitopes, from 10 to 15 aa. In a recent study, Scull and colleagues (49) found that a large fraction of HLA-A2 peptides were longer than 9 aa. Furthermore, many of the longer peptides with the favored position 2 leucine anchor had a variety of different amino acids at putative anchor residue position 9, including charged amino acids such as threonine, lysine, and aspartic acid. Relevant to this study, several eluted longer peptides with a position 2 leucine also contained aspartic acids at position 10. In general, over the years, epitope identification has been an inherently biased approach focusing on characterizing nonamers. Although it is likely that concepts derived from nonamers could in some ways extend to longer peptides, emerging data suggest critical differences exist. Myers and colleagues (50) recently showed that HLA-A2 molecules on glioblastoma tumor cells express a novel 12-mer peptide frameshift. Although a bona fide consensus HLA-A2 nonamer was fully contained within the 12 mer, molecular modeling revealed that binding of the nonamer to HLA-A2 was different at most of the amino acid positions. The similarities appeared to be largely confined to the terminal NH-2 moiety, as well as the position 1 aa binding to the HLA molecule. Relevant to the current study, however, were the observations that the T cells generated against the 12 mer were cross reactive with T cells generated against the embedded 9 mer, suggesting the presence of areas shared by the 9 mer and 12 mer that are conserved enough for T cell recognition.

Predictive algorithms continue to improve and will likely shed additional light on longer peptides that may be relevant to vaccine-based therapeutic or prevention strategies. Several algorithms have now been developed that combine various factors that affect binding and that go beyond identifying motifs and, rather, discover contributions of all amino acids to binding. In more recent times, stabilized matrix method-based algorithms (e.g., SMMPMBEC and SMM) that address interactions between amino acids within an epitope, rather than assuming independence, have been developed that consistently outperform those algorithms that assume independence (e.g., SYFPEITHI) (25, 48, 51, 52). The stabilized matrix methods also outperform those based on neural networks (48, 51, 52). SMMPMBEC (51) and SMM (48) prediction methods show that p373–382 is predicted to bind with an IC50 of 2.7 μM and 0.482 μM, respectively. Our observation of an EC50 of 1.8 μM, using the T2 stabilization assay, is in good agreement with the predictions.

In closing, our investigations show that the commonly employed HLA-A2–binding HER-2/neu–derived peptide, p369–377, is not processed from HER-2/neu protein fragments by the proteasome or immunoproteasome, as previously suggested. Rather, an overlapping 10-mer peptide, p373–382, was processed. Although p373–382 does not have a typical HLA-A2–binding amino acid sequence, binding studies demonstrate HLA-A2–binding activity. Furthermore, the peptide appears to be naturally processed and is able to elicit peptide-specific T cells that cross react with p369–377–elicited T cells. The latter finding of cross-reactivity may explain the clinical utility of p369–377.

We thank the Mayo Proteomics Research Center for helpful discussions.

This work was supported by grants from the Susan G. Komen for the Cure Foundation and National Institutes of Health, National Cancer Institute Grant R01-CA113861 (to K.L.K.) and by National Institutes of Health, National Cancer Institute Grant R01 CA152045 (to K.L.K. and R.C.).

The online version of this article contains supplemental material.

Abbreviation used in this article:
ECD

extracellular domain.

1
American Cancer Society
.
2012
.
Cancer Facts & Figures.
American Cancer Society
,
Atlanta, GA
.
2
U.S. Cancer Statistics Working Group
.
2010
.
United States Cancer Statistics: 1999–2007 incidence and mortality web-based report.
U.S. Department of Health and Human Services, CDC and National Cancer Institute. Available at
: http://www.cdc.gov/uscs. Accessed April 27, 2012.
3
Weiss
R. B.
,
Woolf
S. H.
,
Demakos
E.
,
Holland
J. F.
,
Berry
D. A.
,
Falkson
G.
,
Cirrincione
C. T.
,
Robbins
A.
,
Bothun
S.
,
Henderson
I. C.
,
Norton
L.
Cancer and Leukemia Group B
.
2003
.
Natural history of more than 20 years of node-positive primary breast carcinoma treated with cyclophosphamide, methotrexate, and fluorouracil-based adjuvant chemotherapy: a study by the Cancer and Leukemia Group B.
J. Clin. Oncol.
21
:
1825
1835
.
4
Mittendorf
E. A.
,
Clifton
G. T.
,
Holmes
J. P.
,
Clive
K. S.
,
Patil
R.
,
Benavides
L. C.
,
Gates
J. D.
,
Sears
A. K.
,
Stojadinovic
A.
,
Ponniah
S.
,
Peoples
G. E.
.
2012
.
Clinical trial results of the HER-2/neu (E75) vaccine to prevent breast cancer recurrence in high-risk patients: from US Military Cancer Institute Clinical Trials Group Study I-01 and I-02.
Cancer
118
:
2594
2602
.
5
Mittendorf
E. A.
,
Holmes
J. P.
,
Ponniah
S.
,
Peoples
G. E.
.
2008
.
The E75 HER2/neu peptide vaccine.
Cancer Immunol. Immunother.
57
:
1511
1521
.
6
Disis
M. L.
,
Smith
J. W.
,
Murphy
A. E.
,
Chen
W.
,
Cheever
M. A.
.
1994
.
In vitro generation of human cytolytic T-cells specific for peptides derived from the HER-2/neu protooncogene protein.
Cancer Res.
54
:
1071
1076
.
7
Hudis
C. A.
2007
.
Trastuzumab—mechanism of action and use in clinical practice.
N. Engl. J. Med.
357
:
39
51
.
8
Ménard
S.
,
Tagliabue
E.
,
Campiglio
M.
,
Pupa
S. M.
.
2000
.
Role of HER2 gene overexpression in breast carcinoma.
J. Cell. Physiol.
182
:
150
162
.
9
Benavides
L. C.
,
Sears
A. K.
,
Gates
J. D.
,
Clifton
G. T.
,
Clive
K. S.
,
Carmichael
M. G.
,
Holmes
J. P.
,
Mittendorf
E. A.
,
Ponniah
S.
,
Peoples
G. E.
.
2011
.
Comparison of different HER2/neu vaccines in adjuvant breast cancer trials: implications for dosing of peptide vaccines.
Expert Rev. Vaccines
10
:
201
210
.
10
Czerniecki
B. J.
,
Koski
G. K.
,
Koldovsky
U.
,
Xu
S.
,
Cohen
P. A.
,
Mick
R.
,
Nisenbaum
H.
,
Pasha
T.
,
Xu
M.
,
Fox
K. R.
, et al
.
2007
.
Targeting HER-2/neu in early breast cancer development using dendritic cells with staged interleukin-12 burst secretion.
Cancer Res.
67
:
1842
1852
.
11
Koski
G. K.
,
Koldovsky
U.
,
Xu
S.
,
Mick
R.
,
Sharma
A.
,
Fitzpatrick
E.
,
Weinstein
S.
,
Nisenbaum
H.
,
Levine
B. L.
,
Fox
K.
, et al
.
2012
.
A novel dendritic cell-based immunization approach for the induction of durable Th1-polarized anti-HER-2/neu responses in women with early breast cancer.
J. Immunother.
35
:
54
65
.
12
Sharma
A.
,
Koldovsky
U.
,
Xu
S.
,
Mick
R.
,
Roses
R.
,
Fitzpatrick
E.
,
Weinstein
S.
,
Nisenbaum
H.
,
Levine
B. L.
,
Fox
K.
, et al
.
2012
.
HER-2 pulsed dendritic cell vaccine can eliminate HER-2 expression and impact ductal carcinoma in situ.
Cancer
118
:
4354
4362
.
13
Disis
M. L.
,
Wallace
D. R.
,
Gooley
T. A.
,
Dang
Y.
,
Slota
M.
,
Lu
H.
,
Coveler
A. L.
,
Childs
J. S.
,
Higgins
D. M.
,
Fintak
P. A.
, et al
.
2009
.
Concurrent trastuzumab and HER2/neu-specific vaccination in patients with metastatic breast cancer.
J. Clin. Oncol.
27
:
4685
4692
.
14
Fisk
B.
,
Blevins
T. L.
,
Wharton
J. T.
,
Ioannides
C. G.
.
1995
.
Identification of an immunodominant peptide of HER-2/neu protooncogene recognized by ovarian tumor-specific cytotoxic T lymphocyte lines.
J. Exp. Med.
181
:
2109
2117
.
15
Lustgarten
J.
,
Theobald
M.
,
Labadie
C.
,
LaFace
D.
,
Peterson
P.
,
Disis
M. L.
,
Cheever
M. A.
,
Sherman
L. A.
.
1997
.
Identification of Her-2/Neu CTL epitopes using double transgenic mice expressing HLA-A2.1 and human CD.8.
Hum. Immunol.
52
:
109
118
.
16
Zaks
T. Z.
,
Rosenberg
S. A.
.
1998
.
Immunization with a peptide epitope (p369-377) from HER-2/neu leads to peptide-specific cytotoxic T lymphocytes that fail to recognize HER-2/neu+ tumors.
Cancer Res.
58
:
4902
4908
.
17
Knutson
K. L.
,
Schiffman
K.
,
Cheever
M. A.
,
Disis
M. L.
.
2002
.
Immunization of cancer patients with a HER-2/neu, HLA-A2 peptide, p369-377, results in short-lived peptide-specific immunity.
Clin. Cancer Res.
8
:
1014
1018
.
18
Knutson
K. L.
,
Schiffman
K.
,
Disis
M. L.
.
2001
.
Immunization with a HER-2/neu helper peptide vaccine generates HER-2/neu CD8 T-cell immunity in cancer patients.
J. Clin. Invest.
107
:
477
484
.
19
Kloetzel
P. M.
,
Ossendorp
F.
.
2004
.
Proteasome and peptidase function in MHC-class-I-mediated antigen presentation.
Curr. Opin. Immunol.
16
:
76
81
.
20
Vertuani
S.
,
Triulzi
C.
,
Roos
A. K.
,
Charo
J.
,
Norell
H.
,
Lemonnier
F.
,
Pisa
P.
,
Seliger
B.
,
Kiessling
R.
.
2009
.
HER-2/neu mediated down-regulation of MHC class I antigen processing prevents CTL-mediated tumor recognition upon DNA vaccination in HLA-A2 transgenic mice.
Cancer Immunol. Immunother.
58
:
653
664
.
21
Herrmann
F.
,
Lehr
H. A.
,
Drexler
I.
,
Sutter
G.
,
Hengstler
J.
,
Wollscheid
U.
,
Seliger
B.
.
2004
.
HER-2/neu-mediated regulation of components of the MHC class I antigen-processing pathway.
Cancer Res.
64
:
215
220
.
22
Asiedu
M. K.
,
Ingle
J. N.
,
Behrens
M. D.
,
Radisky
D. C.
,
Knutson
K. L.
.
2011
.
TGFbeta/TNF(alpha)-mediated epithelial-mesenchymal transition generates breast cancer stem cells with a claudin-low phenotype.
Cancer Res.
71
:
4707
4719
.
23
Dela Cruz
J. S.
,
Lau
S. Y.
,
Ramirez
E. M.
,
De Giovanni
C.
,
Forni
G.
,
Morrison
S. L.
,
Penichet
M. L.
.
2003
.
Protein vaccination with the HER2/neu extracellular domain plus anti-HER2/neu antibody-cytokine fusion proteins induces a protective anti-HER2/neu immune response in mice.
Vaccine
21
:
1317
1326
.
24
Taylor
C.
,
Hershman
D.
,
Shah
N.
,
Suciu-Foca
N.
,
Petrylak
D. P.
,
Taub
R.
,
Vahdat
L.
,
Cheng
B.
,
Pegram
M.
,
Knutson
K. L.
,
Clynes
R.
.
2007
.
Augmented HER-2 specific immunity during treatment with trastuzumab and chemotherapy.
Clin. Cancer Res.
13
:
5133
5143
.
25
Rammensee
H.
,
Bachmann
J.
,
Emmerich
N. P.
,
Bachor
O. A.
,
Stevanović
S.
.
1999
.
SYFPEITHI: database for MHC ligands and peptide motifs.
Immunogenetics
50
:
213
219
.
26
Nijman
H. W.
,
Houbiers
J. G.
,
van der Burg
S. H.
,
Vierboom
M. P.
,
Kenemans
P.
,
Kast
W. M.
,
Melief
C. J.
.
1993
.
Characterization of cytotoxic T lymphocyte epitopes of a self-protein, p53, and a non-self-protein, influenza matrix: relationship between major histocompatibility complex peptide binding affinity and immune responsiveness to peptides.
J. Immunother. Emphasis Tumor Immunol.
14
:
121
126
.
27
Erskine
C. L.
,
Henle
A. M.
,
Knutson
K. L.
.
2012
.
Determining optimal cytotoxic activity of human Her2neu specific CD8 T cells by comparing the Cr51 release assay to the xCELLigence system
.
J. Vis. Exp.
(
66
):
e3683
.
doi:10.3791/3683
.
28
Textoris-Taube
K.
,
Henklein
P.
,
Pollmann
S.
,
Bergann
T.
,
Weisshoff
H.
,
Seifert
U.
,
Drung
I.
,
Mügge
C.
,
Sijts
A.
,
Kloetzel
P. M.
,
Kuckelkorn
U.
.
2007
.
The N-terminal flanking region of the TRP2360-368 melanoma antigen determines proteasome activator PA28 requirement for epitope liberation.
J. Biol. Chem.
282
:
12749
12754
.
29
Valmori
D.
,
Gileadi
U.
,
Servis
C.
,
Dunbar
P. R.
,
Cerottini
J. C.
,
Romero
P.
,
Cerundolo
V.
,
Lévy
F.
.
1999
.
Modulation of proteasomal activity required for the generation of a cytotoxic T lymphocyte-defined peptide derived from the tumor antigen MAGE-3.
J. Exp. Med.
189
:
895
906
.
30
Chapiro
J.
,
Claverol
S.
,
Piette
F.
,
Ma
W.
,
Stroobant
V.
,
Guillaume
B.
,
Gairin
J. E.
,
Morel
S.
,
Burlet-Schiltz
O.
,
Monsarrat
B.
, et al
.
2006
.
Destructive cleavage of antigenic peptides either by the immunoproteasome or by the standard proteasome results in differential antigen presentation.
J. Immunol.
176
:
1053
1061
.
31
Rock
K. L.
,
Farfán-Arribas
D. J.
,
Shen
L.
.
2010
.
Proteases in MHC class I presentation and cross-presentation.
J. Immunol.
184
:
9
15
.
32
Ho
Y. K.
,
Bargagna-Mohan
P.
,
Wehenkel
M.
,
Mohan
R.
,
Kim
K. B.
.
2007
.
LMP2-specific inhibitors: chemical genetic tools for proteasome biology.
Chem. Biol.
14
:
419
430
.
33
Orlowski
M.
,
Cardozo
C.
,
Michaud
C.
.
1993
.
Evidence for the presence of five distinct proteolytic components in the pituitary multicatalytic proteinase complex. Properties of two components cleaving bonds on the carboxyl side of branched chain and small neutral amino acids.
Biochemistry
32
:
1563
1572
.
34
Peoples
G. E.
,
Goedegebuure
P. S.
,
Smith
R.
,
Linehan
D. C.
,
Yoshino
I.
,
Eberlein
T. J.
.
1995
.
Breast and ovarian cancer-specific cytotoxic T lymphocytes recognize the same HER2/neu-derived peptide.
Proc. Natl. Acad. Sci. USA
92
:
432
436
.
35
Rongcun
Y.
,
Salazar-Onfray
F.
,
Charo
J.
,
Malmberg
K. J.
,
Evrin
K.
,
Maes
H.
,
Kono
K.
,
Hising
C.
,
Petersson
M.
,
Larsson
O.
, et al
.
1999
.
Identification of new HER2/neu-derived peptide epitopes that can elicit specific CTL against autologous and allogeneic carcinomas and melanomas.
J. Immunol.
163
:
1037
1044
.
36
van der Burg
S. H.
,
Visseren
M. J.
,
Brandt
R. M.
,
Kast
W. M.
,
Melief
C. J.
.
1996
.
Immunogenicity of peptides bound to MHC class I molecules depends on the MHC-peptide complex stability.
J. Immunol.
156
:
3308
3314
.
37
Hall
F. C.
,
Rabinowitz
J. D.
,
Busch
R.
,
Visconti
K. C.
,
Belmares
M.
,
Patil
N. S.
,
Cope
A. P.
,
Patel
S.
,
McConnell
H. M.
,
Mellins
E. D.
,
Sonderstrup
G.
.
2002
.
Relationship between kinetic stability and immunogenicity of HLA-DR4/peptide complexes.
Eur. J. Immunol.
32
:
662
670
.
38
Hernández
J.
,
Schoeder
K.
,
Blondelle
S. E.
,
Pons
F. G.
,
Lone
Y. C.
,
Simora
A.
,
Langlade-Demoyen
P.
,
Wilson
D. B.
,
Zanetti
M.
.
2004
.
Antigenicity and immunogenicity of peptide analogues of a low affinity peptide of the human telomerase reverse transcriptase tumor antigen.
Eur. J. Immunol.
34
:
2331
2341
.
39
Borbulevych
O. Y.
,
Baxter
T. K.
,
Yu
Z.
,
Restifo
N. P.
,
Baker
B. M.
.
2005
.
Increased immunogenicity of an anchor-modified tumor-associated antigen is due to the enhanced stability of the peptide/MHC complex: implications for vaccine design.
J. Immunol.
174
:
4812
4820
.
40
Regner
M.
2001
.
Cross-reactivity in T-cell antigen recognition.
Immunol. Cell Biol.
79
:
91
100
.
41
Wooldridge
L.
,
Laugel
B.
,
Ekeruche
J.
,
Clement
M.
,
van den Berg
H. A.
,
Price
D. A.
,
Sewell
A. K.
.
2010
.
CD8 controls T cell cross-reactivity.
J. Immunol.
185
:
4625
4632
.
42
Borbulevych
O. Y.
,
Piepenbrink
K. H.
,
Gloor
B. E.
,
Scott
D. R.
,
Sommese
R. F.
,
Cole
D. K.
,
Sewell
A. K.
,
Baker
B. M.
.
2009
.
T cell receptor cross-reactivity directed by antigen-dependent tuning of peptide-MHC molecular flexibility.
Immunity
31
:
885
896
.
43
Frankild
S.
,
de Boer
R. J.
,
Lund
O.
,
Nielsen
M.
,
Kesmir
C.
.
2008
.
Amino acid similarity accounts for T cell cross-reactivity and for “holes” in the T cell repertoire.
PLOS One
3
:
e1831
.
44
Popovic
J.
,
Li
L. P.
,
Kloetzel
P. M.
,
Leisegang
M.
,
Uckert
W.
,
Blankenstein
T.
.
2011
.
The only proposed T-cell epitope derived from the TEL-AML1 translocation is not naturally processed.
Blood
118
:
946
954
.
45
Wooldridge
L.
,
Ekeruche-Makinde
J.
,
van den Berg
H. A.
,
Skowera
A.
,
Miles
J. J.
,
Tan
M. P.
,
Dolton
G.
,
Clement
M.
,
Llewellyn-Lacey
S.
,
Price
D. A.
, et al
.
2012
.
A single autoimmune T cell receptor recognizes more than a million different peptides.
J. Biol. Chem.
287
:
1168
1177
.
46
Yin
Y.
,
Mariuzza
R. A.
.
2009
.
The multiple mechanisms of T cell receptor cross-reactivity.
Immunity
31
:
849
851
.
47
Falk
K.
,
Rötzschke
O.
,
Stevanović
S.
,
Jung
G.
,
Rammensee
H. G.
.
2006
.
Allele-specific motifs revealed by sequencing of self-peptides eluted from MHC molecules. 1991.
J. Immunol.
177
:
2741
2747
.
48
Peters
B.
,
Tong
W.
,
Sidney
J.
,
Sette
A.
,
Weng
Z.
.
2003
.
Examining the independent binding assumption for binding of peptide epitopes to MHC-I molecules.
Bioinformatics
19
:
1765
1772
.
49
Scull
K. E.
,
Dudek
N. L.
,
Corbett
A. J.
,
Ramarathinam
S. H.
,
Gorasia
D. G.
,
Williamson
N. A.
,
Purcell
A. W.
.
2012
.
Secreted HLA recapitulates the immunopeptidome and allows in-depth coverage of HLA A*02:01 ligands.
Mol. Immunol.
51
:
136
142
.
50
Myers
C. E.
,
Hanavan
P.
,
Antwi
K.
,
Mahadevan
D.
,
Nadeem
A. J.
,
Cooke
L.
,
Scheck
A. C.
,
Laughrey
Z.
,
Lake
D. F.
.
2011
.
CTL recognition of a novel HLA-A*0201-binding peptide derived from glioblastoma multiforme tumor cells.
Cancer Immunol. Immunother.
60
:
1319
1332
.
51
Kim
Y.
,
Sidney
J.
,
Pinilla
C.
,
Sette
A.
,
Peters
B.
.
2009
.
Derivation of an amino acid similarity matrix for peptide: MHC binding and its application as a Bayesian prior.
BMC Bioinformatics
10
:
394
.
52
Buus
S.
,
Lauemøller
S. L.
,
Worning
P.
,
Kesmir
C.
,
Frimurer
T.
,
Corbet
S.
,
Fomsgaard
A.
,
Hilden
J.
,
Holm
A.
,
Brunak
S.
.
2003
.
Sensitive quantitative predictions of peptide-MHC binding by a ‘Query by Committee’ artificial neural network approach.
Tissue Antigens
62
:
378
384
.

K.L.K. and A.M.H. have a patent pending describing some of the new epitopes found in this study, including p373–382. Furthermore, the Mayo Clinic has licensed, with consideration, to Tapimmune (Seattle, WA), the rights to develop p373–382 as a therapy. The other authors have no financial conflicts of interest.