First and foremost among the many factors that influence epitope presentation are the degradation of Ag, which results in peptide liberation, and the presence of HLA class I molecules able to present the peptides to T lymphocytes. To define the regions of HIV-1 Nef that can provide multiple T cell epitopes, we analyzed the Nef sequence and determined that there are 73 peptides containing 81 HLA-binding motifs. We tested the binding of these peptides to six common HLA molecules (HLA-A2, -A3, -A24, -B7, -B8, and -B35), and we showed that most of them were efficient binders (54% of motifs), especially peptides associating with HLA-A3, -B7/35, and -B8 molecules. Nef peptides most frequently recognized by T cells of HIV-1-infected individuals were 90–97, 135–143, 71–81, 77–85, 90–100, 73–82, and 128–137. The frequency of T cell recognition was not directly related to the strength of peptide-HLA binding. The generation of Nef epitopes is crucial; therefore, we investigated the digestion by the 20S proteasome of a large peptide, Nef66–100. This fragment was efficiently cleaved, and NH2-terminally extended precursors of epitope 71–81 were recognized by T cells of an HIV-1-infected individual. These results suggest that a high frequency of T cell recognition may depend on proteasome cleavage.

Presentation of antigenic peptides to CTL by HLA class I molecules is a multistage process. Within APCs, several important steps govern presentation, including degradation of Ag, which is accomplished by cytosolic proteolytic systems that deliver peptides to TAP for selection and association with MHC class I molecules (1). Antigenic peptides are generated mostly by multicatalytic particles, or proteasomes, that liberate products with 8–11 amino acids. These products are suitable in size for MHC class I ligands, and they can be either standard-length products, or COOH-terminal-length variants, or NH2-terminally extended precursor fragments (2, 3, 4). Although the specificity of proteasomes is well documented, it is not currently possible to deduce major cleavage sites from the analysis of a protein sequence (5). Only in vitro digestion of large synthetic peptides with purified proteasomes can indicate whether an epitope can be produced or destroyed (3, 6). In fact, among numerous factors that influence this degradation, the residues that flank a potential epitope can have a profound positive or negative effect on the quantity of peptides liberated and on the peptide presentation to T cells (3).

In MHC class I peptide presentation, the nature of the HLA alleles that select and bind peptides is important because only HLA-peptide complexes of sufficient binding affinity and stability are loaded and presented to TCRs (7, 8, 9). Degradation of Ag, binding of liberated peptides to MHC molecules, and epitope recognition by T cells must all be assessed to provide insight into the dynamic interplay that exists between Ag and T cell response, as suggested by Koup (10).

To study the antigenic potential of a protein and to identify regions that could provide multiple T epitopes, we selected the HIV-1 Nef protein. Its relatively small size (∼200 amino acids), the knowledge of its structure in solution (11), and the high frequency of CTL responses directed against this protein in HIV-1-infected individuals are the factors that governed this choice.

In the present report, we analyze the distribution of HLA-specific anchor residues, the HLA-binding capacity of relevant peptides, and the frequency of epitope presentation by each of six common HLA alleles. To examine Nef processing, we investigated the proteasome cleavage of the epitope-rich region 66–100. Relationships between Nef structure, processing, and CTL epitope recognition were then sought.

Peptides.

Synthetic peptides of 8–11 residues containing putative anchor-binding motifs for HLA-A2, -A3, -A24, -B7, -B8, and -B35 were selected from the Nef HIV-1/Bru sequence. They were synthesized by Chiron Mimotopes (Victoria, Australia) or Neosystem (Strasbourg, France). They were supplied by Agence Nationale de la Recherche sur le SIDA. Lyophilized peptides were dissolved in DMSO and diluted to 1 mg/ml in water (final concentration of DMSO, 10%), aliquoted, and stored at −20°C.

HLA molecules.

HLA molecules were purified from EBV-transformed B cell lines. After cell lysis in PBS with 1% Nonidet P-40 and protease inhibitors, HLA molecules were retained on affinity columns using anti-HLA Igs, eluted, and then frozen at −80°C. HLA molecules were denatured in PBS containing 12.5 mM NaOH (pH 11.7) and 1.5 M urea for 1 h at 4°C. HLA H chains and β2-microglobulin (β2m)3 were separated from endogenous peptides on a Sephadex G25 column (PD10; Amersham Pharmacia Biotech, Uppsala, Sweden) equilibrated in PBS containing 0.05% Tween 20, 2 mM EDTA, and 0.1% Nonidet P-40, as previously reported (12). Then 2 μg/ml exogenous β2m (Sigma) and 6 mM 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (Sigma) were added just before addition of exogenous peptide.

HLA-peptide binding.

Aliquots of HLA H chains (1 μg in 50 μl) were incubated with different concentrations (10−4 and 10−6 M) of exogenous peptides in Eppendorf microtubes (Eppendorf-Netheler-Hinz GmbH, Hamburg, Germany) for 1 h at room temperature and then for 24 h at 4°C. Reassembled HLA molecules were further incubated for 90 min at 37°C in wells of microtiter plates coated with anti-HLA Abs (13). Correctly folded HLA complexes were revealed with anti-β2m Ig coupled to alkaline phosphatase, with 4-methyl-umbelliferyl phosphate (M-8883; Sigma, St. Louis, MO) as substrate. Fluorescence generated was measured at 360/460 nm in a Microfluor reader (Victor 1420; Wallac, Turku, Finland).

Lymphocytes donors.

PBMC from 76 HIV-1-seropositive individuals were isolated by density gradient centrifugation (separation medium; Flow, Irvine, U.K.) and used after freezing and thawing. Cohorts were established with the approval of the local ethics committee (Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale of Cochin hospital), and all participants gave their written informed consent for the constitution of cell banks.

Enzyme-linked immunospot (ELISPOT) assay.

The IFN-γ ELISPOT assay was adapted from Scheibenbogen et al. and described by Dalod et al. (14). Ninety-six-well nitrocellulose plates (Millipore, Bedford, MA) were coated with 2 μg/ml mouse anti-human IFN-γ mAb (number 1598-00; Genzyme, Rüsselheim, Germany). PBMCs, either freshly isolated or thawed, were cultured overnight in complete medium (RPMI 1640 supplemented with Glutamax, nonessential amino acids, sodium pyruvate (1 mM), HEPES buffer (10 mM), penicillin (100 U/ml), streptomycin (100 mg/ml) (Life Technologies, Paisley, U.K.), and 10% FCS (PAN Biotech, Aidenbach, Germany)) and plated in triplicate at serial dilutions (3 × 105–104 cells/well). Appropriate stimuli were then added, and the plates were incubated for 20 h at 37°C in 5% CO2. After washing, the cells were incubated with 100 μl rabbit polyclonal anti-human IFN-γ Ab diluted 1:250 (IP500; Genzyme), then with a biotinylated anti-rabbit Ig G diluted 1:500 (Boehringer Mannheim, Mannheim, Germany), and finally with alkaline phosphatase-labeled extravidin (Sigma). Spots were developed by adding chromogenic alkaline phosphatase substrate (Bio-Rad, Hercules, CA), and colored spots were counted in a stereomicroscope. A result was considered to be significant when the numbers of spots were at least twice the background value (value given by negative peptides) and were proportional to the numbers of plated cells. Frequencies of IFN-γ spot-forming cells (SFC) were calculated. Positive controls for IFN detection consisted of six wells containing 300–1000 cells stimulated with 50 ng/ml PMA and 500 ng/ml ionomycin. This strong mitogenic stimulus verified that freezing and thawing did not introduce artifacts, and it constituted an indirect check of overall T cell viability. Negative controls for immune recognition consisted of epitopes derived from various viruses (for example, peptide Tax11–19 from human T cell lymphotropic virus-1, which associates with HLA-A2); they never elicited a significant response compared with PBMCs incubated in medium alone. Positive controls for immune recognition were epitopes of EBV or influenza virus. By using purified T cell subsets, we verified that IFN-γ was secreted by CD8+ cells.

Proteasome digestion.

Peptide 66–100 from Nef HIV-1/Bru had the following sequence: VGFPVTPQVPLRPMTYKAAVDLSHFLKEKGGLEGL. It was synthesized by Neosystem. The peptide (150 μg) was incubated at 37°C with 30 μg of 20S proteasome (539150; Calbiochem, La Jolla, CA) in 200 μl assay buffer containing 20 mM Tris-HCl (pH 8) and 0.5 mM EDTA for times varying from 4 to 24 h. Cleavage products were separated by reversed phase HPLC (Perkin-Elmer, Norwalk, CT) on a C18 column (Nucleosil; 10 μm, 250 × 4.0 mm; Macherey-Nagel, Hoerdt, France) using a linear gradient of 0.1% trifluoroacetic acid in water and 0.08% trifluoroacetic acid in acetonitrile, 1% of the latter in 5 min, 1–10% in 5 min, 10–35% in 50 min, and 35–60% in 10 min at 0.8 ml/min flow rate. Chromatogram was recorded at 214 nm (759 Å; Applied Biosystems, Roissy, France). Major peaks (absorbance above 0.15 absorbance unit (AU) were collected and lyophilized.

Detection of epitopes in proteasome digests.

Thawed PBMCs of patient Z037 were cultured overnight and plated in a 96-well nitrocellulose plate at 4 × 105 cells/well in 100 μl of complete medium. Lyophilized fractions were dissolved in complete medium, and 100 μl/well were added on cells. ELISPOT IFN-γ were revealed as described before after overnight incubation.

Mass spectrometry analysis and Edman’s sequencing.

Mass analysis was performed on a linear matrix-assisted laser desorption ionization time-of-flight Bruker (Billerica, MA) instrument using α-cyano-4-hydroxycinnamic acid as matrix. Peptides were sequenced by automated Edman degradation using an Applied Biosystems Procise CLC protein sequencer equipped with a PTH ABI 140D analyzer (Applied Biosystems, Foster City, CA).

Eight to eleven-mer Nef peptides with binding motifs specific for six common HLA molecules, as defined by Sidney and Sette (15, 16), are shown in Fig. 1. Twenty-two of these peptides were for HLA-A2 (L, I, M, V, or A at position 2 and V, L, or A at the C-terminal), 18 peptides for HLA-A3 with a motif requiring the presence of positively charged residues (L, V, A, I, or T at position 2 and K, R, or Y at the C-terminal), 13 peptides for HLA-A24 (Y or F at position 2 and L, F, Y, or I at the C-terminal), 23 peptides for HLA-B7/B35 (P or S at position 2 and a hydrophobic residue at the C-terminal), and 5 peptides with HLA-B8-specific motif (basic residue R or K, or Q at positions 3 and 5, and hydrophobic residue at the C-terminal). Thus, we found 81 motifs, and we defined 73 peptides containing one or two HLA-binding motifs (Fig. 1). Nef regions containing multiple motifs are, in decreasing order: 126–153 (15 motifs in 28 residues), 65–112 (25 motifs in 48 residues), 180–204 (13 motifs in 25 residues), and 9–23 (7 motifs in 15 residues). If we consider the distribution of various motifs, HLA-A2, -A3, or -B7/35-specific motifs are observed in all regions, whereas HLA-A24-specific motifs are mainly located in the central region, which contains a high frequency of hydrophobic and aromatic residues.

FIGURE 1.

HLA-binding motifs in Nef protein. Nef sequence showing the positions of peptides with binding motifs for HLA-A2, -A3, -A24, -B7, -B8, and -B35 molecules. The first and last amino acids of each peptide are represented; HLA restriction is in bold type. ∗, Peptide 190–198, with HLA-A2 motif and with Leu at position 191, is a variant of peptide 190–198 from Nef HIV-1/Bru sequence.

FIGURE 1.

HLA-binding motifs in Nef protein. Nef sequence showing the positions of peptides with binding motifs for HLA-A2, -A3, -A24, -B7, -B8, and -B35 molecules. The first and last amino acids of each peptide are represented; HLA restriction is in bold type. ∗, Peptide 190–198, with HLA-A2 motif and with Leu at position 191, is a variant of peptide 190–198 from Nef HIV-1/Bru sequence.

Close modal

We then tested the binding of all these peptides to the six HLA molecules. We used purified HLA H chains and synthetic peptides as described in Materials and Methods. For each HLA molecule, we tested a reference epitope that yielded 100% binding (see legend of Fig. 2). The most significant results were obtained at a peptide concentration of 10−6 M. We defined as high binders the peptides yielding 50–100% binding and as moderate binders the peptides yielding 20–50% binding. We also considered as low binders the peptides yielding <20% binding at 10−6 M and >20% at 10−4 M.

FIGURE 2.

Nef sequence showing the positions of HLA-binder peptides and the strength of binding. From Table I, peptides with high reactivity (≥50% binding at 10−6 M) are represented with a bold line, peptides with moderate reactivity (20 to <50% at 10−6 M) with a normal line, and peptides with low reactivity (<20% at 10−6 M and significant binding at 10−4 M) with a dotted line. Binder-rich regions 66–100 and 126–146 are delimited by arrows. ∗, Peptide 136–146 was described as an HLA-A2 binder (17 ) but was not tested by us. ∗∗, Peptide 190–198 yielded high HLA-A2 binding when containing Leu191 and no binding when containing Phe191 (Nef Bru sequence).

FIGURE 2.

Nef sequence showing the positions of HLA-binder peptides and the strength of binding. From Table I, peptides with high reactivity (≥50% binding at 10−6 M) are represented with a bold line, peptides with moderate reactivity (20 to <50% at 10−6 M) with a normal line, and peptides with low reactivity (<20% at 10−6 M and significant binding at 10−4 M) with a dotted line. Binder-rich regions 66–100 and 126–146 are delimited by arrows. ∗, Peptide 136–146 was described as an HLA-A2 binder (17 ) but was not tested by us. ∗∗, Peptide 190–198 yielded high HLA-A2 binding when containing Leu191 and no binding when containing Phe191 (Nef Bru sequence).

Close modal

We found 11 high, 16 moderate, and 16 low binders. In addition, peptide 136–146 was a binder described by Lucchiari-Hartz et al. (17). Finally, 44 HLA binders were detected for 81 motifs (54%). A large majority of them (32/41 = 78%) were concentrated in region 66–146 (this region represents 39% of Nef protein size). Most high binders correspond to peptides complexed to HLA-A3, -B7/35, and -B8 molecules and lower binders to peptides complexed to HLA-A24 and -A2 molecules.

Nineteen synthetic peptides from binder-rich regions 66–100 and 126–146 were tested for their capacity to stimulate PBMCs from HIV-1-infected patients ex vivo. The induction of IFN-γ secretion was revealed by ELISPOT. Only significant positive results (more than twice the background value for each individual) were retained. An epitope may be defined by peptide recognition by PBMCs from only one patient. Seventeen peptides were positive. Frequencies of T cell recognition are given in Table I. A high T cell recognition frequency (T cell recognition observed in at least 50% of the donors tested) was obtained with eight peptide-HLA complexes yielding high (three of eight), moderate (three of eight), or low (two of eight) HLA binding, respectively. Moderate T cell recognition frequency (25 to <50%) was observed in six peptide-HLA complexes yielding high (one of six), moderate (three of six), or low (two of six) HLA binding. Low T cell recognition frequency was observed in eight peptide-HLA complexes yielding high (two of eight), moderate (five of eight), or low (one of eight) HLA binding. The frequency of T cell recognition was not directly related to the strength of HLA binding, which shows the influence of other factors, notably those related to Ag processing, epitope variations, or available repertoire. The peptides most frequently recognized were, in decreasing order, 90–97, 135–143, 71–81, 77–85, 90–100, 73–82, and 128–137. We determined six new epitopes: 71–79 and 71–81 presented by HLA-B7/B35, 83–91 by HLA-A2, 126–135 by HLA-A24, 130–139 by HLA-B35, and 136–144 by HLA-A3.

Table I.

HLA binding and T cell recognition of peptides from regions 66–100 and 126–146a

HLAPeptideSequenceBinding (%)Frequency of T Cell Recognition
10−4 M10−6 M(responder/tested)(%)
A2 75–83 PLRPMTYKA 43 NTb  
 83–91 AAVDLSHFL 70 12 (3/18) 17 
 90–100 FLKEKGGLEGL 55 15 (8/12) 67c 
 136–145 PLTFGWCYKL 35 18 (9/28) 32 
 137–145 LTFGWCYKL 60 42 NT  
A3 73–82 QVPLRPMTYK 100 100 (9/15) 60 
 84–92 AVDLSHFLK 100 75 (4/12) 33 
 86–94 DLSHFLKEK 34 30 (0/5)  0 
 136–144 PLTFGWCYK 26 (3/12) 25 
 137–144 LTFGWCYK 67 NT  
A24 67–76 GFPVTPQVPL 39 16 NT  
 80–87 TYKAAVDL 26 13 (0/6)  0 
 89–97 HFLKEKGGL 36 17 NT  
 126–135 NYTPGPGVRY 100 40 (3/10) 30 
 134–143 RYPLTFGWCY 64 21 (5/120 42 
 138–145 TFGWCYKL 61 NT  
B7 68–76 FPVTPQVPL 70 36 (1/13)  8 
 71–79 TPQVPLRPM 47 28 (1/10) 10 
 71–81 TPQVPLRPMTY 54 (9/12) 75 
 77–85 RPMTYKAAV 100 20 (7/10) 70 
 128–135 TPGPGVRY 28 NT  
 128–137 TPGPGVRYPL 72 100 (8/16) 50 
 130–139 GPGVRYPLTF 87 61 (0/10)  0 
 135–143 YPLTFGWCY 50 34 (2/13) 15 
B35 68–76 FPVTPQVPL 100 65 (1/12)  8 
 71–79 TPQVPLRPM 70 42 (1/10) 10 
 71–81 TPQVPLRPMTY 60 35 (5/10) 50 
 74–81 VPLRPMTY 61 46 (5/16) 31 
 77–85 RPMTYKAAV 49 18 (0/3)  0 
 128–137 TPGPGVRYPL 67 52 (1/9) 11 
 130–139 GPGVRYPLTF 29 21 (1/11)  9 
 135–143 YPLTFGWCY 62 45 (11/14) 79 
B8 90–97 FLKEKGGL 100 85 (12 /14) 86 
HLAPeptideSequenceBinding (%)Frequency of T Cell Recognition
10−4 M10−6 M(responder/tested)(%)
A2 75–83 PLRPMTYKA 43 NTb  
 83–91 AAVDLSHFL 70 12 (3/18) 17 
 90–100 FLKEKGGLEGL 55 15 (8/12) 67c 
 136–145 PLTFGWCYKL 35 18 (9/28) 32 
 137–145 LTFGWCYKL 60 42 NT  
A3 73–82 QVPLRPMTYK 100 100 (9/15) 60 
 84–92 AVDLSHFLK 100 75 (4/12) 33 
 86–94 DLSHFLKEK 34 30 (0/5)  0 
 136–144 PLTFGWCYK 26 (3/12) 25 
 137–144 LTFGWCYK 67 NT  
A24 67–76 GFPVTPQVPL 39 16 NT  
 80–87 TYKAAVDL 26 13 (0/6)  0 
 89–97 HFLKEKGGL 36 17 NT  
 126–135 NYTPGPGVRY 100 40 (3/10) 30 
 134–143 RYPLTFGWCY 64 21 (5/120 42 
 138–145 TFGWCYKL 61 NT  
B7 68–76 FPVTPQVPL 70 36 (1/13)  8 
 71–79 TPQVPLRPM 47 28 (1/10) 10 
 71–81 TPQVPLRPMTY 54 (9/12) 75 
 77–85 RPMTYKAAV 100 20 (7/10) 70 
 128–135 TPGPGVRY 28 NT  
 128–137 TPGPGVRYPL 72 100 (8/16) 50 
 130–139 GPGVRYPLTF 87 61 (0/10)  0 
 135–143 YPLTFGWCY 50 34 (2/13) 15 
B35 68–76 FPVTPQVPL 100 65 (1/12)  8 
 71–79 TPQVPLRPM 70 42 (1/10) 10 
 71–81 TPQVPLRPMTY 60 35 (5/10) 50 
 74–81 VPLRPMTY 61 46 (5/16) 31 
 77–85 RPMTYKAAV 49 18 (0/3)  0 
 128–137 TPGPGVRYPL 67 52 (1/9) 11 
 130–139 GPGVRYPLTF 29 21 (1/11)  9 
 135–143 YPLTFGWCY 62 45 (11/14) 79 
B8 90–97 FLKEKGGL 100 85 (12 /14) 86 
a

Peptides were tested at 10−4 and 10−6 M. Results are expressed as the percentage of strongest binding reached by a reference CTL epitope that was, respectively, Influenza virus matrix58–66 (GILGFVFTL) for HLA-A2, Nef73–82 for HLA-A3, HIV Env590–598 (RYLKDQQLL) for HLA-A24, Nef128–137 for HLA-B7, Nef68–76 for HLA-B35, and HIV Env591–598 (YLKDQQLL) for HLA-B8. Recognition of Nef peptides by T cells was tested by ELISPOT as specified in Materials and Methods, and the percentage of positivity is shown. Epitopes are stored in the Los Alamos database (21 ).

b

NT, Not tested.

c

High recognition frequencies (≥50%) are represented in bold type.

The large number of epitopes represented in Fig. 3 and the high frequency of T cell recognition of some of them possibly result from an efficient processing of the Nef regions 66–100 and 126–146 with liberation of adequate peptides. As the processing of the second Nef region has already been studied by Lucchiari-Hartz et al. (17), our study focused on in vitro cleavage of the first region, Nef66–100. A recombinant 20S proteasome with mainly chymotrypsin-like activity was used. Degradation was controlled by reversed phase HPLC chromatography after various incubation times at 37°C; 24 h incubation was the optimum time (data not shown). The major peak of the HPLC chromatogram corresponding to Nef66–100 (eluted after 67 min) greatly decreased after 24 h incubation, whereas numerous other peaks appeared between 30 and 55 min. Fractions corresponding to peaks with absorbance higher than 0.15 AU were collected. A partial HPLC chromatogram of the peptides that separated at 40–70 min is represented in Fig. 4,A. We chose to study the generation of epitope Nef71–81 because of its central position in Nef66–100 sequence and the presence of flanking sequences that may influence processing. As shown in Fig. 4,B, HIV-1-seropositive patient Z037 presented a very strong response against this epitope (2800 SFC/106 PBMC). T cells of this patient were used to screen the presence of this epitope in HPLC fractions. Peptides eluted before 40 min were not recognized (data not shown). As shown in Fig. 4,C, fractions C and D induced strong reactivity (>1000 SFC/106 PBMC). Analysis of peptide contents of these fractions by mass spectrometry and Edman sequencing revealed the presence of peptide 69–81 in C and 68–81 in D (Table II). These two species are NH2-extended forms of the Nef71–81 epitope. Fraction A was not recognized by Z037 T cells, and only Nef72–81 was characterized but was too short for inducing a stimulation. Fractions B and E were recognized by a smaller number of effector T cells (∼500 SFC/106 PBMC), and analysis of their content revealed that major species were Nef88–100 (fraction B) and Nef84–100 (fraction E). T cells from patient Z037 recognized Nef90–100 with an intensity of 570 SFC/106 PBMCs, and they probably also recognized extended forms of this epitope (Nef88–100 and Nef84–100). However, demonstration of cleavage by 20S proteasome at position 100 would require the use of another COOH-terminally extended peptide. There was a relationship between the number of T effector cells recognizing the optimal peptide Nef71–81 (2800 SFC/106 pulsed PBMCs) and the number of T cells recognizing the NH2-extended forms Nef68–81 and Nef69–81 in fractions C and D (>1000 SFC/106 pulsed PBMCs). Similarly, peptide Nef90–100 activated 570 T effector cells, and fractions B and E, containing NH2-extended forms of this epitope, activated ∼500 effectors. In summary, we have characterized the efficient generation by 20S proteasome of different forms of Nef71–81, which is an epitope having a high frequency of T cell recognition in HIV-infected patients.

FIGURE 3.

Frequency of recognition of Nef epitopes in central region 66–146. Frequency of T cell recognition classified as high (≥50%) is represented by a bold line, moderate (25 to <50%) by a normal line, and low (>0 to <25%) by a dotted line. Only peptides recognized by T cells of HIV-infected patients were reported.

FIGURE 3.

Frequency of recognition of Nef epitopes in central region 66–146. Frequency of T cell recognition classified as high (≥50%) is represented by a bold line, moderate (25 to <50%) by a normal line, and low (>0 to <25%) by a dotted line. Only peptides recognized by T cells of HIV-infected patients were reported.

Close modal
FIGURE 4.

Generation of epitopes by 20S proteasome cleavage of Nef66–100 large peptide. A, Peptide Nef66–100 was incubated at 37°C with 20S proteasome for 24 h in Tris buffer plus 0.5 mM EDTA. A sample digest was fractionated by HPLC. Part of the elution profile of a C18 column is shown. Peaks with absorbance >0.15 AU were collected and analyzed by mass spectrometry and Edman’s degradation. Found sequences are reported above the peaks. Experimentally detected cleavage sites are indicated by arrows above the Nef66–100 sequence. B, Reactivity of T cells from patient Z037 against the five HLA-matched epitopes from Nef region 66–100, tested by ELISPOT assay. C, Reactivity in ELISPOT IFN-γ of T cells from Z037 against fractions A–E and mock (collected out of peaks).

FIGURE 4.

Generation of epitopes by 20S proteasome cleavage of Nef66–100 large peptide. A, Peptide Nef66–100 was incubated at 37°C with 20S proteasome for 24 h in Tris buffer plus 0.5 mM EDTA. A sample digest was fractionated by HPLC. Part of the elution profile of a C18 column is shown. Peaks with absorbance >0.15 AU were collected and analyzed by mass spectrometry and Edman’s degradation. Found sequences are reported above the peaks. Experimentally detected cleavage sites are indicated by arrows above the Nef66–100 sequence. B, Reactivity of T cells from patient Z037 against the five HLA-matched epitopes from Nef region 66–100, tested by ELISPOT assay. C, Reactivity in ELISPOT IFN-γ of T cells from Z037 against fractions A–E and mock (collected out of peaks).

Close modal
Table II.

Peptide sequences detected by mass spectrometry and Edman sequencing in Nef66–100 digesta

HPLC PeakMeasured Mass [M + H]+Edman Sequencing of the First Amino AcidsValidated SequenceTheoretical Mass [M + H]+Position in Nef
1200.37 PQVPLRP… PQVPLRPMTY 1202.46 72–81 
1414.38 SHFLKEK… SHFLKEKGGLEGL 1415.63 88–100 
1498.54 PVTPQVP… PVTPQVPLRPMTY 1499.82 69–81 
1645.96 FPVTPQV… FPVTPQVPLRPMTY 1646.99 68–81 
 2002.06 VGFPVTP… VGFPVTPQVPLRPMTYKA 2002.43 66–83 
1812.14 AVDLSHF… AVDLSHFLKEKGGLEGL 1814.09 84–100 
1801.68 VGFPVTP… VGFPVTPQVPLRPMTY 1803.18 66–81 
HPLC PeakMeasured Mass [M + H]+Edman Sequencing of the First Amino AcidsValidated SequenceTheoretical Mass [M + H]+Position in Nef
1200.37 PQVPLRP… PQVPLRPMTY 1202.46 72–81 
1414.38 SHFLKEK… SHFLKEKGGLEGL 1415.63 88–100 
1498.54 PVTPQVP… PVTPQVPLRPMTY 1499.82 69–81 
1645.96 FPVTPQV… FPVTPQVPLRPMTY 1646.99 68–81 
 2002.06 VGFPVTP… VGFPVTPQVPLRPMTYKA 2002.43 66–83 
1812.14 AVDLSHF… AVDLSHFLKEKGGLEGL 1814.09 84–100 
1801.68 VGFPVTP… VGFPVTPQVPLRPMTY 1803.18 66–81 
a

After digestion of Nef66–100 by 20S proteasome, peptides were separated by HPLC on a C18 column. In fractions containing peaks A–F of the HPLC chromatogram with an absorbance >0.15 AU, products were characterized by mass spectrometry and Edman sequencing up to the first seven N-terminal amino acids. Complete peptide sequences are shown in the fourth column. Within the sequences of NH2-terminally extended precursors, epitopes 71–81 and 90–100 are represented in bold type.

The wealth of a protein in potential epitopes, their localization in the protein sequence, their presentation by common HLA molecules, and the frequency of T cell recognition are critical for vaccine development. Here, we present an extensive cartography of Nef peptides containing HLA-binding motifs that are specific for six common HLA molecules: -A2, -A3, -A24, -B7, -B8, and -B35 (Fig. 1). In Nef, such peptides can be identified along the whole sequence, but notably in the central region 66–146(66–146), which is characterized by numerous hydrophobic residues, proline, and some basic residues. Peptides that truly bound to these HLA molecules were remarkably concentrated in regions 66–100 and 126–146 (Fig. 2).

We chose to study further the presentation of peptides from these two regions and their T cell recognition. A high proportion of peptides containing HLA-A3- and HLA-B7/B35-specific motifs were good binders, which is consistent with the high frequency of T cell recognition of several epitopes presented by these HLA molecules. For HLA-A2, a lower binding is also consistent with the lower frequency of Nef epitopes identified (Table I). Among the peptides that can bind to the 6 HLA molecules tested (Fig. 2), 19 were tested against T cells of 76 HIV positive patients; 17 of these induced stimulation. The frequency of T recognition was especially high for epitopes 90–97 (presented by HLA-B8), 135–143 (B35), 71–81 (B7 and B35), 77–85 (B7), 90–100 (A2), 73–82 (A3), and 128–137 (B7). This result suggests that few recognition-inhibiting variations occur in these Nef segments. It may also reveal that cellular processing liberates high quantities of these peptides.

Although several enzymatic systems are involved in protein digestion, proteasomes probably generate the majority of cytosolic peptides (18). Niedermann et al. showed that processing of polypeptides by proteasomes is conserved in evolution (19). In APCs, IFN-γ induction alters proteasomal proteolysis in such a way that the generation of peptides with structural features of MHC class I ligands might be optimized (4, 5). Cleavage after basic residues is possible after this IFN-γ treatment, which enhances trypsin-like activity, but the majority of changes may be quantitative rather than qualitative. In the present report, by studying digestion of peptide Nef66–100 by a 20S proteasome having a chymotrypsin-like activity, we identified the efficient production of NH2-extended forms of epitope 71–81. When T cells of HIV-1-infected individuals were tested, peptide 71–81 was recognized by 75% of HLA-B7-typed donors and by 50% of HLA-B35-typed donors (Table I). Mass spectrometry and Edman’s sequencing showed that two NH2-extended forms, Nef68–81 and Nef69–81, were efficiently produced by 20S proteasome (Table II). In the case of epitope Nef71–81, HLA-B7 and -B35 bindings were low and moderate, respectively. Dominance of T cell response could be explained by a high concentration of liberated peptides and could therefore be very dependent on proteasome cleavage. It is interesting to note that another epitope, Nef74–81, was recognized by 31% of HLA-B35-typed patients, and this epitope could also have Nef68–81 and Nef69–81 as precursors. However, HLA-B7-typed patients did not recognize Nef74–81 because of its failure to bind to this HLA allele. Of the other peptides from the same region, peptide 68–76 bound very efficiently to HLA-B7, but its frequency of T recognition was low, which could be consistent with a very low production by proteolytic systems. In Nef region 123–152, Lucchiari-Hartz et al. (17) identified naturally processed peptides and the generation by 20S proteasomes from T1 cells of five epitopes from the fragment Nef123–152 (128–135, 128–137, 135–143, 136–145, and 136–146), presented by HLA-A2 and HLA-B7 molecules. We tested epitopes 135–143 and 128–137, which were frequently recognized by T cells of our HIV-1-infected patients. All these data suggest that ranking of immunodominance of T cell responses to these peptides may depend on proteasome cleavage.

By using minigenes that code for 17–22-mer peptides and include one or several epitopes, different authors have shown the effect of COOH-terminal flanking residues on epitope liberation (3). According to data of Shimbara et al. (20), cleavage patterns of large peptides obtained with isolated proteasomes revealed that flanking Ala, basic, or hydrophobic residues enhance digestion, whereas flanking Gly, Pro, or acidic residues may inhibit it or prevent random cleavage and, thus, contribute to the efficient production of HLA-B7 ligands. This rule can be applied to Pro-rich Nef regions 68–85 and 128–143 where a large number (nine) of HLA-B7/B35-specific epitopes was detected. The abundance of hydrophobic and aromatic residues in central Nef region increases the chances of cleavage. Degradation of the 123–152 region by proteasome produces several fragments that are epitopes (17), and degradation of the 66–100 region probably generates numerous epitopes or their precursors.

It was then possible to define the relationships that may exist between Nef structure and epitope liberation. The structure of HIV-1 Nef in solution, described by Grzesiek et al. (11), comprises one polyproline helix (residues 69–78), three other helices (residues 81–94, 105–118, and 194–198), five β-sheet strands (residues 100–102, 126–128, 134–137, 142–146, and 181–186), and two solvent-exposed long disordered loops (NH2 terminus 1–67 and 146–179). An abundance of hydrophobic, aromatic, and basic residues is compatible with high concentration of HLA-binding motifs, structural formations such as β-sheets or helices, and efficient proteasome digestion. In fact, regions 66–100 and 126–146 are both rich in HLA-binding motifs and well-structured segments of helices (69–78 and 81–94), or β-sheets (126–128, 134–137, 142–146), and their degradation by proteasomes delivers epitopes (this study and Ref. 17). In contrast, from the Los Alamos database (21), no CTL epitopes were identified in the coiled loop region 20–67, which is very rich in small residues (Ala, Thr, and Ser) and in charged residues (Arg, Lys, Asp, and Glu), or in region 146–179, which is very rich in negatively charged (Asp and Glu) residues. These two regions are poor in HLA-binding residues and are thus unlikely to be a source of CTL epitopes.

It remains to be confirmed whether the potential of any protein to generate CTL epitopes can be predicted more precisely by analyzing both secondary and tertiary structures.

We are grateful to Nelly Bonilla for technical assistance and to Renaud Fortuner for editing the English text.

1

This work was supported by the Agence Nationale de la Recherche sur le SIDA. W.C. was supported by the Ministère de l’Education Nationale, de la Recherche et de la Technologie. A.B. was the recipient of a postdoctoral fellowship from SIDACTION foundation.

3

Abbreviations used in this paper: β2m, β2-microglobulin; AU, absorbance unit; ELISPOT, enzyme-linked immunospot; SFC, spot-forming cells.

1
Koopmann, J. O., G. J. Hämmerling, F. Momburg.
1997
. Generation, intracellular transport and loading of peptides associated with MHC class I molecules.
Curr. Opin. Immunol.
9
:
80
2
Paz, P., N. Brouwenstijn, R. Perry, N. Shastri.
1999
. Discrete proteolytic intermediates in the MHC class I antigen processing pathway and MHC I-dependent peptide trimming in the ER.
Immunity
11
:
241
3
Niedermann, G., E. Geier, M. Lucchiari-Hartz, N. Hitziger, A. Ramsperger, K. Eichmann.
1999
. The specificity of proteasomes: impact on MHC class I processing and presentation of antigens.
Immunol. Rev.
172
:
29
4
York, I. A., A. L. Goldberg, X. Y. Mo, K. L. Rock.
1999
. Proteolysis and class I major histocompatibility complex antigen presentation.
Immunol. Rev.
172
:
49
5
Rock, K. L., A. L. Goldberg.
1999
. Degradation of cell proteins and the generation of MHC class I-presented peptides.
Annu. Rev. Immunol.
17
:
739
6
Luckey, C. J., G. M. King, J. A. Marto, S. Venketeswaran, B. F. Maier, V. L. Crotzer, T. A. Colella, J. Shabanowitz, D. F. Hunt, V. H. Engelhard.
1998
. Proteasomes can either generate or destroy MHC class I epitopes: evidence for nonproteasomal epitope generation in the cytosol.
J. Immunol.
161
:
112
7
Sette, A., A. Vitiello, B. Reherman, P. Fowler, R. Nayersina, W. M. Kast, C. J. M. Melief, C. Oseroff, L. Yuan, J. Ruppert, et al
1994
. The relationship between class I binding affinity and immunogenicity of potential cytotoxic T cell epitopes.
J. Immunol.
153
:
5586
8
Van der Burg, S. H., M. J. W. Visseren, R. M. P. Brandt, W. M. Kast, C. J. M. Melief.
1996
. Immunogenicity of peptides bound to MHC class I molecules depends on the MHC-peptide complex stability.
J. Immunol.
156
:
3308
9
Busch, D. H., E. G. Pamer.
1998
. MHC class I/peptide stability: implications for immunodominance, in vitro proliferation, and diversity of responding CTL.
J. Immunol.
160
:
4441
10
Koup, R. A..
1994
. Virus escape from CTL recognition.
J. Exp. Med.
180
:
779
11
Grzesiek, S., A. Bax, M. Clore, A. M. Gronenborn, J. S. Hu, J. Kaufman, I. Palmer, S. J. Stahl, P. T. Wingfield.
1996
. The solution structure of HIV-1 Nef reveals an unexpected fold and permits delineation of the binding surface for the SH3 domain of Hck tyrosine protein kinase.
Nat. Struct. Biol.
3
:
340
12
Gnjatic, S., B. Bressac-de-Paillerets, J. G. Guillet, J. Choppin.
1995
. Mapping and ranking of potential cytotoxic T epitopes in the p53 protein: effect of mutations and polymorphism on peptide binding to purified and refolded HLA molecules.
Eur. J. Immunol.
25
:
1638
13
Couillin, I., F. Connan, B. Culmann-Penciolelli, E. Gomard, J. G. Guillet, J. Choppin.
1995
. HLA-dependant variations in human immunodeficiency virus Nef protein alter peptide/HLA binding.
Eur. J. Immunol.
25
:
728
14
Dalod, M., M. Dupuis, J. C. Deschemin, D. Sicard, D. Salmon, J. F. Delfraissy, A. Venet, M. Sinet, J. G. Guillet.
1999
. Broad, intense anti-human immunodeficiency virus (HIV) ex vivo CD8+ responses in HIV type 1-infected patients: comparison with anti-Epstein-Barr virus responses and changes during antiretroviral therapy.
J. Virol.
73
:
7108
15
Sidney, J., H. M. Grey, R. T. Kubo, A. Sette.
1996
. Practical, biochemical and evolutionary implications of the discovery of HLA class I supermotifs.
Immunol. Today
17
:
261
16
Sette, A., J. Sidney.
1999
. Nine major HLA class I supertypes account for the vast preponderance of HLA-A and -B polymorphism.
Immunogenetics
50
:
201
17
Lucchiari-Hartz, M., P. M. van Endert, G. Lauvau, R. Maier, A. Meyerhans, D. Mann, K. Eichmann, G. Niedermann.
2000
. Cytotoxic T lymphocyte epitopes of HIV-1 nef: generation of multiple definitive major histocompatibility complex class I ligands by proteasomes.
J. Exp. Med.
191
:
239
18
Coux, O., K. Tanaka, A. L. Goldberg.
1996
. Structure and functions of the 20S and 26S proteasomes.
Annu. Rev. Biochem.
65
:
801
19
Niedermann, G., R. Grimm, E. Geier, M. Maurer, C. Realini, C. Gartmann, J. Soll, S. Omura, M. C. Rechsteiner, W. Baumeister, K. Eichmann.
1997
. Potential immunocompetence of proteolytic fragments produced by proteasomes before evolution of the vertebrate immune system.
J. Exp. Med.
185
:
209
20
Shimbara, N., K. Ogawa, Y. Hidaka, H. Nakajima, N. Yamasaki, S. Niwa, N. Tanahashi, K. Tanaka.
1998
. Contribution of proline residue for efficient production of MHC class I ligands by proteasomes.
J. Biol. Chem.
273
:
23062
21
HIV Molecular Immunology database of Los Alamos. Los Alamos, New Mexico. http://hiv-web.lanl.gov/immuno/index.html.