The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mata, M.
Right arrow Articles by Paterson, Y.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mata, M.
Right arrow Articles by Paterson, Y.
The Journal of Immunology, 1998, 161: 2985-2993.
Copyright © 1998 by The American Association of Immunologists

The MHC Class I-Restricted Immune Response to HIV-gag in BALB/c Mice Selects a Single Epitope That Does Not Have a Predictable MHC-Binding Motif and Binds to Kd Through Interactions Between a Glutamine at P3 and Pocket D1

Marielena Mata*, Paul J. Travers{dagger}, Qiang Liu2,*, Fred R. Frankel* and Yvonne Paterson3,*

* Department of Microbiology, University of Pennsylvania Medical School, Philadelphia, PA; and {dagger} Department of Crystallography, Birkbeck College, London, United Kingdom


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Using a strain of Listeria monocytogenes that stably expresses and secretes HIV gag to deliver this Ag to the MHC class I pathway of Ag processing, we have identified the immunodominant CTL epitope to gag in the BALB/c mouse and shown that it is Kd restricted. The specific motif for the peptides that bind the MHC class I molecule H-2 Kd is believed to be a nonamer with residues tyrosine or phenylalanine in the second amino acid position and leucine or isoleucine in the carboxyl-terminal or ninth amino acid position as dominant anchoring positions. Surprisingly, the identified gag peptide, AMQMLKETI, does not contain an anchoring aromatic residue in position two although competition assays with other Kd-restricted epitopes indicated that it binds to Kd with comparable affinity. Using a theoretical molecular dynamics approach to probe the stability of peptide binding to MHC class I molecules, we show that the absence of an appropriate anchor residue at P2 in AMQMLKETI is compensated by favorable interactions of the glutamine at P3 with pocket D of Kd. These findings were verified experimentally, demonstrating the predictive power of this theoretical approach in analyzing MHC class I/peptide interactions. These studies also indicate that CTL epitope prediction that relies on dominant peptide motifs may not always identify the correct epitope.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Through an intricate and highly regulated mechanism, endogenous Ags are processed in the cytoplasm of the cell (1, 2, 3) and transported to the endoplasmic reticulum, where they bind the MHC class I molecule (4) before their export to the cell surface where they can be recognized by CD8+ T cells.

Falk et al. have described a series of motifs based on elution of peptides from class I molecules and peptide sequencing of bulk populations (5). These motifs, which we will call the Rammensee motifs, have been used to predict epitopes within Ags for many purposes, including the development of peptide vaccines (6, 7). In the mouse haplotype H-2d, for example, the motifs for all the loci have been identified, and, in particular, the Kd motif has been described to be a nonamer with residues tyrosine or phenylalanine in the second amino acid position and leucine, isoleucine, or valine at the carboxyl-terminal or ninth amino acid position as anchoring residues (8).

Listeria monocytogenes is a Gram-positive, facultative intracellular bacterium that enters the macrophage upon phagocytosis. It then escapes from the phagolysosome and lives in the cytoplasm of the cell (9). As a consequence, secreted L. monocytogenes Ags are processed and presented by both the class I and class II Ag presentation pathways. This results in strong cell-mediated immune responses, with the induction of both CD8+ CTL and CD4+ T cells (10, 11), thought to be of the Th1 phenotype (12, 13). It has recently been proposed that, in HIV infection, strong cell-mediated responses consisting of T helper cells of the type 1 phenotype and CTL may be protective (14, 15, 16).

We had previously described a genetic method to modify permanently the chromosome of L. monocytogenes so that HIV gene products can be expressed as secreted proteins under the control of a copy of the strong promoter of the hemolysin (hly)4 gene, which encodes listeriolysin-O (17). Furthermore, we demonstrated that mice immunized with one of these constructs, Lm-gag, mount a strong, specific, long-lasting CTL response against the HIV-1 gag protein and that the response is directed predominantly against an epitope present in the p24 portion of the protein.

In this article we have identified the immunodominant CTL epitope contained in the p24 portion of the protein and have shown that it is Kd restricted. Surprisingly, this peptide, gag 197–205 AMQMLKETI, does not contain the anchoring tyrosine residue in position two as predicted by the Rammensee motif (-Y------I/L) and displayed by all known Kd epitopes (8). Furthermore, other peptides in p24 that contain H-2d motifs do not appear to be selected for immune recognition, including some that contain tyrosine or phenylalanine in the second position, or to bind to Kd. In addition, competition assays suggest that AMQMLKETI binds to Kd with an affinity comparable to that of hly 91–99, a Kd-restricted L. monocytogenes epitope containing the Rammensee motif. In the absence of a crystal structure for Kd, we have used homology modeling to construct a model of this molecule and examined the contribution of individual residues to the stability of the Kd/AMQMLKETI complex using a molecular dynamics approach. By comparison with a similar analysis of the immunodominant hemolysin peptide GYKDGNEYI and the analogues AMAMLKETI and AYAMLKETI also bound to Kd, we show that the lack of an aromatic residue at position two in AMQMLKETI appears to be compensated by favorable interactions mediated by the glutamine residue at position three with pocket D in the Kd molecule. We have verified these predictions by showing that AMQMLKETI and AYAMLKETI effectively compete with hly 91–99 for recognition by hly-specific CTL whereas AMAMLKETI is a poor inhibitor.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Bacterial strains

L. monocytogenes (Lm-wt) strain 10403S (11) was the wild-type organism used in these studies. It has an LD50 of approximately 3 x 104 when injected i.v. or i.p. into BALB/c mice (18). Lm-gag refers to either one of two recombinant strains of L. monocytogenes, each of which carries one copy of the HIV-1 strain HXB gag gene stably integrated into the listerial chromosome and both of which secret the gag gene product as determined by Western blotting of secreted proteins (Ref. 17; Z.-J. Yao and Y. Paterson, unpublished observations). All strains were grown in brain/heart infusion (BHI) medium (Difco Labs, Detroit, MI).

Cell lines

P815 mouse mastocytoma cells (H-2d) and tk- L cells (H-2k) transfected with H-2d class I molecules, L-Dd, L-Kd (kind gift from Dr. Eisenhlor, Thomas Jefferson University, Philadelphia, PA) (19), L-Ld, and the parental cell line L-tk- (kind gift from Dr. T. Hansen, Washington University, St. Louis, MO) (20) were grown in RPMI 1640 media supplemented with 10% FBS, 100 mM HEPES, 100 U/ml penicillin, 100 µg/ml streptomycin and 2 mM L-glutamine. Appropriate H-2d class I expression was verified by FACS analysis (data not shown).

Vaccinia constructs

The vaccinia viruses used in this paper were vVK1 (Vac-gag) (provided by B. Moss, National Institutes of Health, Bethesda, MD), NYCBH (Vac+), vAbT141 (p55), vAbT228 (p17), and vAbT286 (p24) (kindly provided by D. Panicali, Therion Biologics Corporation, Cambridge, MA) (21).

Viral stocks were grown in monolayer HeLa S3 cells and titrated from confluent monolayer cultures using a plaque assay with BSC-1 cells.

Generation and expansion of gag-specific CTLs

Six- to eight-week-old BALB/c female mice (H-2d) (Charles River Laboratories substrain AnNCrIBR, Charles River laboratories, Raleigh, NC) were immunized by i.p. inoculation with 106 or 107 live Lm-gag, or with 6 x 103 live Lm-wt (about 0.2 LD50), the wild-type strain. For some experiments, effectors were derived from mice boosted with a similar dose 2 to 4 wk later. Splenocytes, obtained from mice 1 to 2 wk after their last immunization with Lm-gag, were cultured after a third of the splenocytes were infected with five plaque-forming units per cell of Vac-gag (VVK1) or vAbT141 (p55). Splenocytes were cocultured for 7 days at 37°C in 40 ml of Iscove’s modified DMEM supplemented with 10% FBS, 100 U/ml penicillin, and 100 µg/ml streptomycin.

Alternatively splenocytes obtained from mice immunized with Lm-gag or Lm-wt were stimulated in vitro with 2 µM of peptides gag 197–205 or hly 91–99 (6), respectively, for 5 days in six-well plates at a concentration of 1 x 107 cells/ml (3 ml/well).

51Cr release assays

P815, L-Kd, L-Dd, L-Ld, and L-tk- target cells were washed twice with PBS, resuspended at a concentration of 107 cells/ml in Eagle’s minimal essential medium supplemented with 2.5% FBS and 100 mM HEPES, and infected with 16.5 plaque-forming units of vaccinia virus per cell. After adsorption for 2 h at 37°C, the infected cells were resuspended in RPMI 1640 medium supplemented with 10% FBS and antibiotics, and incubated at 37°C for 2 h. The cells were then pelleted, resuspended in 100 µl medium containing 50 µCi Na51CrO4 per 5 to 6 x 105 cells, and incubated at 37°C for 1 h.

In vitro-stimulated splenocytes were incubated at various ratios with 2 x 104 51Cr-labeled target cells suspended in RPMI 1640 medium (total volume of 200 µl/well) in round-bottom, 96-well plates for 4 h at 37°C. Released 51Cr was determined in 100 µl of supernatant, and the percentage of specific lysis was calculated as (experimental cpm-spontaneous cpm)/(total cpm-spontaneous cpm) x 100. All assays were performed in triplicate, and the mean and SD were calculated.

In assays where peptides were used as a source of Ag and as competitors, uninfected target cells were labeled with Na51CrO4, washed, and then incubated in 150 µl for 1 to 2 h at 37°C with peptides at various concentrations, before addition of splenocytes.

Overlapping peptides

Twenty-two overlapping 20-mer peptides with 10-amino acid overlaps, spanning residues 133 to 362 of the HIV-1SF2 gag p24 protein, were provided by the MRC AIDS Reagent Project, London, U.K. (Repository reference: ADP 788.1-22). HIV-1SF2 gag differs in amino acid sequence with the HXB strain gag, which was used to make the Lm-gag vaccine, only at two positions in the p24 fragment. One change is from I to L at position 138 and the other is an E to D change at position 132. Both of these substitutions would conserve the motifs associated with H-2d MHC class I presentation (6, 7).

Peptide synthesis

Peptides gag 193–201, gag 197–205A199, and gag 197–205Y198A199 were purchased as crude cleavage products from Biopolymer Synthesis Center, California Institute of Technology, Pasadena, CA. All other peptides were synthesized in our laboratory using DuPont’s RaMPS multiple peptide synthesis system procedure on Wang Resin (p-Alkoxy-benzyl alcohol resin) (DuPont Chemical Company, Wilmington, DE) using F-moc (9-fluorenylmethyloxycarbonyl) nitrogen terminal-protected amino acids (BACHEM Biotech Company, Philadelphia, PA). Couplings were conducted using DPC (N-N-diisopropylcarbodiimide), HOBT (1-hydroxybenzotriazole) in DMF (N,N-dimethylformamide) solvent and were monitored by ninhydrin reaction in 100°C water for 5 min. Simultaneous resin cleavage and side-chain deprotection were achieved by high concentration TFA (all chemical reagents purchased from Aldrich (Milwaukee, WI) and Sigma Chemical Company (St. Louis, MO)). All crude products were purified to 98% purity by reverse-phase HPLC (Perkin-Elmer, Norwalk, CT, Series 400) using a Vydac C18 column, and their m.w. was verified by matrix-assisted laser desorption ionization (MALDI)-mass spectroscopy.

Molecular modeling and calculations

All molecular modeling operations of the interaction of peptides with Kd were performed using SYBYL Molecular Modeling Software (Tripos Associates, St. Louis, MO) and a Silicon Graphics display system (Silicon Graphics, San Diego, CA). Structures of peptides bound to the {alpha}1 and {alpha}2 domains of Kd were built by homology modeling using a data set of known crystallographic structures of peptide bound to human or mouse MHC class I molecules deposited in the Brookhaven Protein Data Bank. Structures of peptides of more than 9 residues or without a hydrophobic amino acid residue at the carboxyl terminus were excluded, leaving 11 structures in the data set, five of peptides bound to mouse H-2 and six of peptides bound to human HLA. Lack of sequential homology between the Kd-binding peptides and those in the homology data set was treated by adopting for the test peptide the conserved backbone coordinates between the first three and last two residues in the peptide data set. The SYBYL LOOP closing algorithm was then used to build the middle four residues. All structures were fully minimized using the Powell algorithm, in which the coordinates of the structure are modified in an iterative manner and the conformational energy is calculated. Further changes in the coordinates are based on reducing the energy by a path of steepest descent. This method, as is the case with all minimization algorithms, is limited to relaxing the structure into the local energy minimum.

Conformational energies were calculated, for minimization and molecular dynamics calculations, using an AMBER force field (22, 23) as the sum of torsional, electrostatic, van der Waals, and hydrogen bonding interaction energies between nonbonded atoms separated by at least three bonds. Because bonded atoms move, the impact of bond angle bending and bond stretching on the total energy are also included as specific energy terms in AMBER. Molecular dynamics analyses of the stability of peptide binding to Kd were performed on hydrated (approximately 2,500 water molecules) and minimized structures at constant temperature. Water is treated explicitly in these calculations; therefore, a dielectric constant of 1.0 was used for electrostatic interactions. Only MHC atoms within 6.5 Å of the peptide residues were allowed to move during the simulations. Preliminary molecular dynamics simulations, which were conducted for 12 ps for AMQMLKETI/Kd showed that dissociation of the peptide from Kd occurred within 6 ps. Molecular dynamic simulations were also performed on the VSV-8/Kb and SEV-9/Kd complexes (24) in addition to AMQMLKETI/Kd to compare the behavior of the modeled structure of the peptide/H-2 Kd complex with the x-ray crystallographically determined complexes. The behavior of the complexes during this period was essentially similar, and dissociation occurred for all three molecules within 6 ps. Subsequent simulations were thus restricted to this time frame.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
The immune response to gag delivered by Lm-gag strain in BALB/c mice (H-2d) is Kd restricted

To determine the MHC restriction of the Ag-specific response to gag delivered by Lm-gag, a series of L cell lines transfected with each of the class I molecules, Dd, Kd, and Ld, were used as target cells in 4-h 51Cr release assays (Fig. 1Go). Only L cells transfected with Kd infected with vaccinia virus expressing gag were lysed, suggesting that the epitope was Kd restricted. In our previous studies (17), we had shown that the immunodominant region of gag was contained within the p24 protein fragment. To confirm that the Kd-restricted epitope was indeed contained within p24, we compared the CTL activity against L-Kd cells and P815 cells when infected with VVp55, VVp24, and VVp17 (Fig. 2Go). The CTL activity against both cells lines was comparable, and lysis was observed only when the cells were infected with p55 or p24, indicating that the epitope contained within p24 is Kd restricted.



View larger version (18K):
[in this window]
[in a new window]
  		FIGURE 1. Gag-specific CTLs are MHC class I Kd restricted. BALB/c female mice were immunized with 1 x 107 CFU Lm-gag. After 1 to 2 wk, splenocytes were isolated, and primary cultures were set up with vaccinia-gag (p55) constructs for 1 wk to provide effector cells. L cells transfected with Kd, Dd, or Ld and P815 (H-2d) cell lines were used as target cells in standard 4-h 51Cr release assays after 4 h infection with vaccinia-gag. Percent specific lysis is defined in Materials and Methods. The error bars represent the SD between triplicate assays for each measurement.

 


View larger version (17K):
[in this window]
[in a new window]
  		FIGURE 2. The Kd-restricted epitope is contained within gag p24. Effector cells were prepared as in Figure 1Go. L cells transfected with Kd and P815 (H-2d) cells were used as target cells in standard 4-h 51Cr release assays after 4 h infection with vaccinia constructs expressing p55, p24, p17, or the parental vaccinia strain.

 
The H-2 Kd-restricted gag epitopes do not contain a strong anchor residue at position two

In previous studies (17), two 20-mer peptide sequences within p24 were observed to induce lysis of P815 target cells by Lm-gag-induced splenocytes whereas a number of peptides of appropriate length containing the Ld, Dd, or Kd motifs did not target P815 cells for lysis (data not reported). Long peptides have been shown to be effective in CTL assays because they either contain truncated contaminants of appropriate length from the peptide synthesis procedure (25, 26) or because they are cleaved to shorter peptides by extracellular proteases such as angiotensin converting enzyme (ACE) found in FBS in culture media (27, 28). It seemed likely, therefore that these sequences contained the Kd-restricted CTL epitopes of appropriate size. However, while the 20-mer peptides, gag 193–212 and 233–252, did contain a number of potential H-2d carboxyl-terminal anchor residues, none was positioned at an appropriate distance from other required dominant anchor residues usually associated with H-2d binding, as described by Falk et al. (5) (Table IGo). Therefore, we tested four nonamer peptides contained within gag 193–212 and 233–252 that contain the appropriate carboxyl-terminal residue, either leucine or isoleucine, and three other peptides contained within p24 that had the other anchoring residue, a tyrosine or a phenylalanine at the second position from the amino-terminal, for their ability to induce lysis of target cells by Lm-gag-induced splenocytes (Fig. 3Go, A and B). Two nonamers, gag 197–205 (AMQMLKETI) contained within gag 193–212 and gag 239–247 (TTSTLQEQI) contained within gag 233–252, induced significant lysis of the target cells, suggesting that these are the epitopes recognized in the H-2 Kd context.


View this table:
[in this window]
[in a new window]
  		Table I. The presence of the Kd-restricted epitope motif (5) within the p24 fragment of HIV gag, the peptide sequences 193–212 and 233–252, and other known Kd-restricted peptides1

 


View larger version (22K):
[in this window]
[in a new window]
  		FIGURE 3. Two nonamers contained within p24 induce lysis of target cells. BALB/c mice were immunized with 1 x 107 CFU Lm-gag and boosted after 3 wk with a similar dose. One week after the last immunization, splenocytes were isolated, and primary cultures were set up with vaccinia-gag constructs for 1 wk. 51Cr release assay was performed as in Figure 1Go, except that P815 cells were incubated with 0.05 µM of peptide. A, Peptides were contained within larger peptides that induced lysis and had a leucine or isoleucine at the carboxyl terminal. B, Peptides were chosen from the p24 sequence based on the presence of an aromatic residue at the second position.

 
To compare the relative affinity of the identified nonamer epitopes and the longer peptides, CTL analyses at limiting dilutions of peptide were performed at an E:T ratio of 30:1 (Fig. 4Go). In both cases, the shorter peptides show higher activity than their longer counterparts, suggesting that the nonamers are of a more appropriate size to bind to Kd and that the longer peptides are not the ideal length for recognition by gag-specific CTL. In the case of gag 193–212, while activity is lost at about one nM for the longer peptide, the shorter peptide targets cells for CTL activity at concentrations as low as 10 pM. On the other hand, gag 233–252 is recognized only at the highest concentration (10 µM) while the shorter peptide gag 239–247 shows activity similar to that of gag 193–212, the other 20-mer, and loses its ability to target cells for lysis at one nM. From these data, we conclude that gag 197–205 is the immunodominant epitope in the immune response to HIV-gag delivered by L. monocytogenes in BALB/c mice.



View larger version (27K):
[in this window]
[in a new window]
  		FIGURE 4. The nonamers are more efficient than the 20-mers in inducing specific lysis. BALB/c female mice were immunized with 1 x 107 CFU Lm-gag and boosted 2 wk later. One week after the last immunization, splenocytes were isolated, and primary cultures were set up with vaccinia-gag constructs for 1 wk. P815 (H-2d) cells were used as target cells in a standard 4-h 51Cr release assays after 1 h incubation with peptides at varying concentrations.

 
Gag epitope 197–205 binds to Kd with an affinity comparable to other known epitopes

Since the immunodominant epitope identified did not contain the full Rammensee motif, we compared the ability of this peptide to bind Kd to that of other antigenic epitopes that do contain the motifs by comparing the ability of these peptides to inhibit Kd-restricted CTL recognition of another epitope in the presence of competing peptide. This assay has been shown to be a more sensitive measure of MHC class I binding than other widely used assays, such as the RMA-S assay, which measures peptide-induced stability of cell surface MHC class I molecules (29). We were particularly interested in hly 91–99 (Table IGo), the Kd-restricted immunodominant epitope in the L. monocytogenes-secreted virulence factor hemolysin, for a number of reasons. First, Pamer and associates successfully identified this as an immunodominant epitope for the L. monocytogenes response by synthesizing the eleven nonamer peptides contained in the hly sequence that fitted the Kd-restricted Rammensee motifs and then testing their ability to target P815 cells for lysis by a CTL line derived from BALB/c mice immunized with L. monocytogenes (6). Further studies, using peptide-binding competition assays, showed that hly 91–99 along with two other hly peptides had the highest binding affinity for H-2 Kd of all 11 hly peptides (30). In addition, immunization with Lm-gag results in strong CTL responses to hly 91–99, as well as to gag 197–205, eliminating possible variability due to different ways of presenting the Ags to the immune system. Finally, as an immunodominant epitope of L. monocytogenes, hly 91–99 is a potential competitor in vivo for gag 197–205 when delivered by Lm-gag since hemolysin is produced copiously under the same conditions as gag when delivered by Lm-gag.

To compare the ability of hly 91–99 or gag 197–205 to inhibit lysis by the other peptide, effector cells specific for either peptide were prepared from splenocytes from mice immunized with Lm-gag and expanded in vitro with either gag 197–205 or hly 91–99. CTL assays were performed in which the same peptide with which splenocytes had been expanded was used as the antigenic peptide at varying concentrations in the presence of the competitor peptide at 2 µM. Both peptides were equally able to compete for the Kd molecules, as shown by a decrease in lysis (a 2.5 log shift of the curve) in the presence of the competitor peptide (Fig. 5Go, A and B).



View larger version (24K):
[in this window]
[in a new window]
  		FIGURE 5. Gag and hly epitopes bind to Kd with similar affinities. BALB/c female mice were immunized with 6 x 103 CFU Lm-wt (A) or 107 CFU Lm-gag (B). After 1 wk, splenocytes were isolated, and primary cultures were set up using 2 µM of antigenic peptide to restimulate the splenocytes. P815 (H-2d) cells were used as target cells in standard 4-h 51Cr release assays after 1 h incubation with the corresponding antigenic peptide at varying concentrations and the competitor peptide at 2 µM. A, Primary culture was set up with hly 91–99, and gag 197–205 was the competitor. B, Primary culture was set up with gag 197–205, and hly 91–99 was the competitor.

 
To determine whether other gag peptides with partial Kd motifs, i.e., that contain the anchoring aromatic residue in position two, were able to bind to Kd molecules, we determined their ability to inhibit lysis induced by gag 197–205 in competition assays where the antigenic peptide was used at 0.05 µM in the presence of competitor peptides at increasing concentrations up to and including 10 µM (Fig. 6Go). In contrast to Kd-binding peptides hly 91–99 or NP148–157(R-156) (31), none of the tested peptides were able to compete with gag 197–205, even at concentrations of 10 µM, suggesting that these peptides do not bind to Kd or bind with an affinity lower than can be detected by this assay.



View larger version (28K):
[in this window]
[in a new window]
  		FIGURE 6. Other gag peptides with partial Kd motifs fail to compete with gag 197–205 for Kd binding. Effector cells were prepared as in Figure 5GoB. P815 (H-2d) cells were used as target cells in a standard 4-h 51Cr release assays after 1 h incubation with the antigenic peptide at 0.05 µM and varying concentrations of either hly 91–99 or NP 147–158R-156 or other gag peptides as competitors.

 
The binding of gag 197–205 to Kd is stabilized by the presence of a glutamine at position three

Currently, there are no crystal structures of the mouse H-2 Kd molecule. Therefore, to determine how gag 197–205, AMQMLKETI, binds to Kd in the absence of an aromatic residue at position two, we constructed the Kd molecule by homology modeling and compared the binding of AMQMLKETI to that of hly 91–99, GYKDGNEYI (Fig. 7Go, a and c). Although a methionine at position two in the gag peptide could fit into the B pocket of the Kd, it did not fill the pocket as well as the tyrosine in that position in GYKDGNEYI. Whereas only one MHC residue (Gln63) was within 3.5 Å of the side chain of Met198 of the gag peptide, additional residues (Tyr7, Ala24, Gln63, Tyr92, Phe99, Tyr159) were within 3.5 Å of Tyr92 of the hly peptide. This suggested that the stability of the AMQMLKETI/Kd complex must be derived from compensating interactions by other residues. To determine the source of this stability, we performed molecular dynamics analyses of both peptide/Kd complexes. The atoms in the hydrated complexes were subjected to Newtonian motion for 6 ps, maintaining the temperature of the system at 293°K and constant volume. Quite early in these dynamics runs (at about 2 ps), the peptides began to disassociate from the peptide binding groove at the carboxyl terminus, followed by the charged amino terminus, remaining anchored at the end of 6 ps by only the tyrosine at position two in the case of GYKDGNEYI, or glutamine at position three in the case of AMQMLKETI (Fig. 7Go, b and d). It is of interest that the dissociation of AMQMLKETI from the Kd peptide binding groove is no more extensive than for GYKDGNEYI (Fig. 7Go, b and d), consistent with our experimental data that these peptides have a comparable affinity for Kd.



View larger version (65K):
[in this window]
[in a new window]
  		FIGURE 7. Molecular dynamics simulation of Kd binding peptides. Computer graphics representation plotted using MOLSCRIPT (32) of the peptides hly 91–99 (GYKDGNEYI), gag 197–205 (AMQMLKETI), gag 197–205A199 (AMAMLKETI), and gag 197–205Y198A199 (AYAMLKETI) bound to the modeled structure of Kd. Only the {alpha}1 domain and peptide binding groove (25 Å wide, 11 Å deep) of Kd are shown; the {alpha}2 domain of Kd has been removed in this representation for easier viewing. a and c show the minimized peptide/MHC complexes at the start of the molecular dynamics simulation. The starting structures of gag 197–205A199/Kd and gag 197–205Y198A199/Kd are not shown since they were constructed from AMQMLKETI/Kd and are the same except for the different side chains at 198 and 199. b and d–f show the conformations of the peptide/Kd complexes after 6 ps of molecular dynamics.

 
To verify that the dissociation of AMQMLKETI and GYKDGNEYI from Kd during these simulations was not a result of inaccuracies in our model of the peptide/Kd complexes, we also subjected the known crystal structures of VSV-8 and SEV-9 (24) complexed to Kb to similar molecular dynamics simulations. The Kb complexes appear to be less stable than the Kd complexes. In both cases the peptides dissociated from the Kb molecule starting at the carboxyl terminus, as did the Kd complexes, but losing contact with Kb at the amino terminus and at pocket C with the Tyr at P5 for VSV-8 and P6 for SEV-9 within 6 ps (data not shown).

We next established that the glutamine at position three was, indeed, acting as the major anchor residue for AMQMLKETI. We changed this residue to an alanine and, after minimization, explored its dissociation from Kd by molecular dynamics for 6 ps. This change resulted in the complete dissociation of the peptide from the Kd molecule (see Fig. 7Goe), confirming that the interaction of AMQMLKETI with Kd is mainly stabilized by this residue binding to pocket D. Finally, to determine whether the glutamine at P3 was as efficient an anchor as Tyr at P2, we examined the stability of the AYAMLKETI/Kd complex (Fig. 7Gof). At the end of a 6-ps molecular dynamics simulation, this peptide remained anchored by the Tyr at P2. This change appeared to restore stability to AMAMLKETI in the absence of Gln at P3 (Fig. 7Go, e and f) but remained less closely associated with Kd than AMQMLKETI (Fig. 7Go, f and d).

Gag 197–205Q199 and 197–205Y198 are better competitors than 197–205A199 for hly 91–99 CTL recognition, as theoretically predicted

The results of the molecular dynamics simulations described above predict that AMQMLKETI and AYAMLKETI have an equivalent affinity for Kd that is higher than AMAMLKETI. To verify this, we used synthetic analogues with these sequences as competitors for the recognition of effector CTL specific for hly 91–99. The assays were performed essentially as described for Figure 5Go, using hly 91–99 as the antigenic peptide at varying concentrations in the presence of the competitor peptide at 2 µM. Both AMQMLKETI and AYAMLKETI were equally able to compete for the Kd molecules, as shown by a decrease in lysis (a 2.0 log shift of the curve) compared with the titration curve in the absence of inhibitor. In contrast, AMAMLKETI did not compete with hly 91–99 in this assay (Fig. 8Go), as anticipated from our molecular dynamics findings.



View larger version (24K):
[in this window]
[in a new window]
  		FIGURE 8. Gag 197–205 and gag 197–205Y198A199 epitopes bind to Kd with higher affinities than gag 197–205A199. BALB/c female mice were immunized with 1 x 106 CFU Lm-gag and boosted 1 wk later. One week after the last immunization, splenocytes were isolated, and primary cultures were set up using 2 µM of hly 91–99 to restimulate the splenocytes. P815 (H-2d) cells were used as target cells in standard 4-h 51Cr release assays after 1 h incubation with the hly 91–99 at concentrations from 10 µM to 10-6 µM and the competitor peptides at 0.1 µM.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
We had previously shown that a L. monocytogenes construct expressing and secreting HIV gag was able to generate a CTL response to the gag Ag that recognized two 20-amino acid-long peptides (17). In this study we have shown that the response is Kd restricted (Figs. 1Go and 2Go) and have further defined the epitopes by identifying two nine-amino acid sequences within the 20-mers (Fig. 3Go) that are recognized at concentrations lower than the 20-mer peptides (Fig. 4Go). One of the peptides, gag 197–205, could target cells for lysis at picomolar concentrations and was 105 times more potent than the other short peptide, suggesting that it is the immunodominant epitope for this response.

The motifs described for peptides binding class I molecules not only specify the number of residues but also describe anchoring positions where specific residues are located. In this case, for Kd, the motif contains tyrosine or phenylalanine in position two and isoleucine, leucine, or valine in position nine (5). It is of interest then that, although we have identified two epitopes that fit the motif in size, both fail to meet the requirements for the anchoring position at position two since the immunodominant epitope, gag 197–205, contains a methionine at this position whereas a threonine is present in the case of gag 239–247. It is important to emphasize, at this time, that these are the first identified epitopes that do not meet the Kd motif criteria (reviewed in 8 although there have been reports describing epitopes for other class I molecules, including HLA-A (33), Kb (34), and Kk (35), that do not conform to their corresponding described motifs. Two questions then arise from this finding. The first is why, in a sequence containing several classical Rammensee Ld and Dd motif peptides, in addition to other partial Kd motifs (Table IGo) that do contain the appropriate anchoring amino acid at position two, the emerging epitopes do not meet these criteria. This selection could take place at two potential levels: the processing pathway, including MHC binding, or the T cell recognition of the peptide in the context of Kd. There are no peptides with complete Kd Rammensee motifs in the gag p24 sequence, and those partial Kd motif peptides with a tyrosine at position two do not bind to MHC as determined by competition assays (Fig. 6Go). Therefore, control may occur at MHC binding in the case of the generation of Kd-restricted epitopes. However, this explanation is not sufficient to account for the lack of recognition of Ld- and Dd-restricted epitopes in this response.

There are several checkpoints at which epitope selection may occur in the MHC class I-restricted pathway. Peptide generation, which takes place in the cytoplasm and appears to be the most critical step to determine immunogenicity (36), can be regulated by latent membrane protein (LMP)-2 and LMP-7, two proteins that modulate proteasome specificity (37, 38, 39, 40). The generated peptides then must be transported to the endoplasmic reticulum by TAP (41, 42), which could play a role in peptide selectivity, although to a lesser degree (36, 43, 44, 45). Finally, the peptide is required to bind to the MHC molecule with an affinity high enough to stabilize the class I molecule and remain bound on the surface of the cell (46, 47). All these steps have been the result of evolutionary changes for a highly regulated mechanism of defense. During this process, sequences flanking the epitope particularly at the carboxyl terminal (48, 49, 50) appear to play an important role in epitope selection. Thus the peptides we have identified may have been generated in preference to all the other sequences to be presented by H-2d MHC molecules by simply being the only epitopes to emerge from all of these checkpoints for recognition by CD8+ T cells. In addition, the immunodominant epitope, 197–205, is capable of inducing lysis of P815 cells at concentrations as low as 1 to 10 pM. Therefore, although gag 197–205 lacks the predictable tyrosine or phenylalanine at position 2, it nevertheless appears to bind Kd with an affinity similar to that of other conventional Kd-binding peptides such as hly 91–99 (Fig. 5Go, A and B), a peptide that was identified as an immunodominant epitope by searching the hemolysin sequence for peptides with the Rammensee motif (6).

A second question then arises as to the source of this stability in the absence of an aromatic residue at position 2. Molecular dynamics is a powerful tool for examining receptor/ligand interactions and particularly for thermodynamic properties and is particularly suited to study time-dependent phenomena (51). However, this technique has not been widely exploited for studying MHC/peptide interactions, possibly because early attempts to use this approach (52) before the structure of MHC-bound peptides had been solved by x-ray crystallography, incorrectly predicted an {alpha}-helical conformation for an MHC class I bound peptide. Better and more reliable predictions have been made since the known x-ray structures of MHC-bound peptides have been used as starting points for these analyses (53, 54, 55). Indeed, the stability of binding of a set of HLA-B27 peptide analogues, as determined by the buried surface areas of the peptide residues after molecular dynamics perturbation, could be qualitatively correlated with the ability of each analogue to catalyze the assembly of HLA-B27 in vitro (55).

We thus performed a theoretical analysis of the stability of the interaction of gag 197–205 with Kd, using molecular dynamics. This simulation of the dissociation of this peptide from the MHC class I molecule indicated that, within a 6-ps simulation, it remained more firmly anchored to the Kd molecule than other peptide/MHC complexes, such as hly 91–99 to Kd (Fig. 7Go, A-D) and VSV8 or SEV9 to Kb (data not shown). In addition, the binding of the glutamine residue at position three to pocket D of Kd appeared to convey stability to the gag peptide/Kd interaction (Fig. 7Go, C and D). Indeed, changing this residue to an alanine resulted in total dissociation of the peptide during the molecular dynamics simulation (Fig. 7GoE), indicating that the interaction of glutamine with pocket D was crucial to the stability of the complex and compensated for the absence of an anchor residue at position two in the peptide. Restoration of binding of this analogue (AMAMLKETI) to Kd could be restored by providing an anchor residue at position two, by there substituting a tyrosine for methionine (Fig. 7GoF). The confirmation of these findings in competition assays using these analogues as competitors for hly 91–99 recognition is an important experimental verification of the molecular dynamics approach.

Although secondary anchor residues have been reported for other MHC class I molecules, including HLA-A2 (56) and Kb (57), position three has not previously been shown to be important in stabilizing peptide binding to Kd. Indeed, in a recent compilation of 26 Kd-binding peptides, all contained a tyrosine at position two except one which had a phenylalanine at this position whereas the residues found at position three were most variable, and glutamine was found in only four peptides (8).

In this study we have clearly shown that a peptide that does not contain the Rammensee motif is recognized as the major epitope in an immune response to gag delivered by Listeria in H-2d mice. Several groups have been using peptide-binding motifs to identify possible epitopes in states of disease to develop peptide vaccines and therapies (6), often with the help of computer algorithms (7). The data here presented suggest that these theoretical methods can fail to identify dominant epitopes that do not meet the criteria described by these motifs and indicate that caution should be used in their application. Even a predictive algorithm that is more broadly based on empirical measurements of the contribution of individual residues to the half time of dissociation of a peptide binding to MHC class I (58) failed to identify AMQMLKETI as the immunodominant epitope within the gag p24 sequence. In contrast, molecular dynamics analyses of putative peptide epitopes associated with MHC class I appears to be a useful tool in designing peptide analogues with differing affinities for MHC molecules.


   Acknowledgments
 
We thank Dr. Eric Pamer, Yale University, for helpful discussions and sharing reagents with us.


   Footnotes
 
1 This work was supported by National Institutes of Health Grants AI-36657 and GM-31841. P.J.T. is a Royal Society University Research fellow. Part of this work was conducted while Y.P. was on sabbatical leave at Birkbeck College. Back

2 Current address: DuPont Industries, Wilmington, DE. Back

3 Address correspondence and reprint requests to Dr. Yvonne Paterson, Department of Microbiology, 323 Johnson Pavilion, 3610 Hamilton Walk, University of Pennsylvania, Philadelphia, PA 19104. Back

4 Abbreviations used in this paper: hly, hemolysin;, Lm-gag, L. monocytogenes strain expressing HIV-gag; Lm-wt, L. monocytogenes wild-type strain. Back

Received for publication January 23, 1998. Accepted for publication May 22, 1998.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Bennink, J. R., J. W. Yewdell, W. Gerhard. 1982. A viral polymerase involved in recognition of influenza virus-infected cells by a cytotoxic T cell clone. Nature 296:75.[Medline]
  2. Gooding, L. R., K. A. O’Connell. 1983. Recognition by cytotoxic T lymphocytes of cells expressing fragments of the SV40 tumor antigen. J. Immunol. 131:2580.[Abstract]
  3. Townsend, A. R., J. Bastin, K. Gould, G. G. Brownlee. 1986. Cytotoxic T lymphocytes recognize influenza haemagglutinin that lacks a signal sequence. Nature 324:575.[Medline]
  4. Townsend, A., C. Ohlen, J. Bastin, H.-G. Ljunggren, L. Foster, K. Karre. 1989. Association of class I major histocompatibility heavy and light chains induced by viral peptides. Nature 340:443.[Medline]
  5. Falk, K., O. Rotzsche, S. Stevanovic, G. Jung, H.-G. Rammensee. 1991. Allele-specific motifs revealed by sequencing of self-peptides eluted from MHC molecules. Nature 351:290.[Medline]
  6. Pamer, E. G., J. T. Harty, M. J. Bevan. 1991. Precise prediction of a dominant class I MHC-restricted epitope of Listeria monocytogenes. Nature 353:852.[Medline]
  7. Meister, G. E., C. G. Roberts, J. A. Berzofsky, A. S. De Groot. 1995. Two novel T cell epitope prediction algorithms based on MHC-binding motifs: comparison of predicted and published epitopes from Mycobacterium tuberculosis and HIV protein sequences. Vaccine 13:581.[Medline]
  8. Rammensee, H.-G., T. Friede, S. Stevanovic. 1995. MHC ligands and peptide motifs: first listing. Immunogenetics 41:178.[Medline]
  9. Falkow, S., R. R. Isberg, D. A. Portnoy. 1992. The interaction of bacteria with mammalian cells. Annu. Rev. Cell Biol. 8:333.
  10. De Libero, G., S. H. E. Kaufmann. 1986. Antigen-specific Lyt-2+ lymphocytes from mice infected with the intracellular bacterium Listeria monocytogenes. J. Immunol. 137:2688.[Abstract]
  11. Bishop, D. K., D. J. Hinrichs. 1987. Adoptive transfer of immunity to Listeria monocytogenes: the influence of in vitro stimulation on lymphocyte subset requirements. J. Immunol. 139:2005.[Abstract]
  12. Hsieh, C.-S., S. E. Macatonia, C. S. Tripp, S. F. Wolf, A. O’Garra, K. M. Murphy. 1993. Development of TH1 CD4+ T cells through IL-12 produced by Listeria-induced macrophages. Science 260:547.[Abstract/Free Full Text]
  13. Hsieh, C.-S., S. E. Macatonia, A. O’Garra, K. M. Murphy. 1993. Pathogen-induced Th1 phenotype development in CD4+ {alpha}ß-TCR transgenic T cells is macrophage dependent. Int. Immunol. 5:371.[Abstract/Free Full Text]
  14. Atlan, H., M. J. Gersten, P. L. Salk, J. Salk. 1993. Can AIDS be prevented by cell vaccination?. Immunol. Today 14:200.[Medline]
  15. Clerici, M., G. M. Shearer. 1993. A Th1->Th2 switch is a critical step in the etiology of HIV infection. Immunol. Today 14:107.[Medline]
  16. Salk, A., P. A. Bretscher, P. L. Salk, M. Clerici, G. Shearer. 1993. A strategy for prophylactic vaccination against HIV. Science 260:1270.[Free Full Text]
  17. Frankel, F. R., S. Hegde, J. Lieberman, Y. Paterson. 1995. Induction of cell-mediated immune responses to human immunodeficiency virus type 1 Gag protein by using Listeria monocytogenes as a live vaccine vector. J. Immunol. 155:4775.[Abstract]
  18. Schafer, R., D. A. Portnoy, S. A. Brassell, Y. Paterson. 1992. Induction of a cellular immune response to a foreign antigen by a recombinant Listeria monocytogenes vaccine. J. Immunol. 149:53.[Abstract]
  19. Eisenlohr, L. C., I. Bacik, J. R. Bennink, K. Bernstein, J. W. Yewdell. 1992. Expression of a membrane protease enhances presentation of endogenous antigens to MHC class I-restricted T lymphocytes. Cell 71:963.[Medline]
  20. Lee, D. R., R. J. Rubocki, W. R. Lie, T. H. Hansen. 1988. The murine MHC class I genes, H-2Dq and H-2Lq, are strikingly homologous to each other, H-2Ld, and two genes reported to encode tumor-specific antigens. J. Exp. Med. 168:1719.[Abstract/Free Full Text]
  21. Koup, R. A., J. L. Sullivan, P. H. Levine, F. Brewster, S. Mahr, G. Mazzara, S. McKenzie, D. Panicali. 1989. Antigenic specificity of antibody-dependent cell-mediated cytotoxicity directed against human immunodeficiency virus in antibody-positive sera. J. Virol. 63:584.[Abstract/Free Full Text]
  22. Weiner, S. J., P. A. Kollman, D. A. Case, U. C. Singh, C. Ghio, G. Alagona, Jr S. Profeta, P. Weiner. 1984. A new force field for molecular mechanical simulation of nucleic acids and proteins. J. Am. Chem. Soc. 106:765.
  23. Weiner, S. J., D. T. Nguyen, P. A. Kollman, D. A. Case. 1986. An all atom force field for stimulations of proteins and nucleic acids. J. Comput. Chem. 7:230.
  24. Fremont, D. H., M. Matsumura, E. A. Stura, P. A. Peterson, I. A. Wilson. 1992. Crystal structures of two viral peptides in complex with murine MHC class IH-2Kb. Science 257:919.[Abstract/Free Full Text]
  25. Rotzsche, O., K. Falk, K. Deres, H. Schild, M. Norda, J. Metzger, G. Jung, H.-G. Rammensee. 1990. Isolation and analysis of naturally processed viral peptides as recognized by cytolytic T cells. Nature 348:252.[Medline]
  26. Schumacher, T. N., M. L. DeBruijn, L. N. Vernie, W. M. Kast, C. J. Melief, J. J. Neefjes, H. L. Ploegh. 1991. Peptide selection by MHC class I molecules. Nature 350:703.[Medline]
  27. Kozlowski, S., M. Corr, T. Takeshita, L. F. Boyd, C. D. Pendleton, R. N. Germain, J. A. Berzofsky, D. H. Margulies. 1992. Serum angiotensin-1 converting enzyme activity processes a human immunodeficiency virus 1 gp 160 peptide for presentation by major histocompatibility complex class I molecules. J. Exp. Med. 175:1417.[Abstract/Free Full Text]
  28. Sherman, L. A., T. A. Burke, J. A. Biggs. 1992. Extracellular processing of peptide antigens that bind class I major histocompatibility molecules. J. Exp. Med. 175:1221.[Abstract/Free Full Text]
  29. Feltkamp, M. C. W., M. P. M. Pierboom, R. E. M. Toes, F. Ossendorp, J. Schegget, C. J. M. Melief, W. M. Kast. 1995. Competition inhibition of cytotoxic T-lymphocyte (CTL) lysis, a more sensitive method to identify candidate CTL epitopes than induction of antibody-detected MHC class I stabilization. Immunol. Lett. 47:1.[Medline]
  30. Wipke, B. T., S. C. Jameson, M. J. Bevan, E. G. Pamer. 1993. Variable binding affinities of listeriolysin-O for the H-2Kd class I molecule. Eur. J. Immunol. 23:2005.[Medline]
  31. Ikonomidis, G., Y. Paterson, F. Kos, D. Portnoy. 1994. Delivery of a viral antigen to the class I processing and presentation pathway by L. monocytogenes. J. Exp. Med. 180:2209.[Abstract/Free Full Text]
  32. Kraulis, P.. 1991. MOLSCRIPT: a program to produce both detailed and schematic plots of protein structure. J. Appl. Crystallogr. 24:946.
  33. Kast, W. M., R. M. Brandt, J. Sydney, J. W. Drijfhout, R. T. Kubo, H. M. Grey, C. J. Melief, A. Sette.. 1994. Role of HLA-A motifs in identification of potential CTL epitopes in human papillomavirus type 16 E6 and E7 proteins. J. Immunol. 152:3904.[Abstract]
  34. Sadovnikova, E., X. Zhu, S. M. Collins, J. Zhou, K. Vousden, L. Crawford, P. Beverley, H. J. Stauss.. 1994. Limitations of predictive motifs revealed by cytotoxic T lymphocyte epitope mapping of the human papilloma virus E7 protein. Int. Immunol. 6:289.[Abstract/Free Full Text]
  35. Calin-Laurens, V., M. C. Trescol-Biemont, D. Gerlier, C. Rabourdin-Combe.. 1993. Can one predict antigenic peptides for MHC class I-restricted cytotoxic T lymphocytes useful for vaccination?. Vaccine 11:974.[Medline]
  36. Deng, Y., J. W. Yewdell, L. C. Eisenlohr, J. R. Bennink.. 1997. MHC affinity, peptide liberation, T cell repertoire, and immunodominance all contribute to the paucity of MHC class I-restricted peptides recognized by antiviral CTL. J. Immunol. 158:1507.[Abstract]
  37. Fehling, H. J., W. Swat, C. Laplace, R. Kuhn, K. Rajewsky, U. Muller, H. von Boehmer.. 1994. MHC class I expression in mice lacking the proteasome subunit LMP-7. Science 265:1234.[Abstract/Free Full Text]
  38. Goldberg, A. L., K. L. Rock. 1992. Proteolysis, proteasomes and antigen presentation. Nature 357:375.[Medline]
  39. Monaco, J. J.. 1992. A molecular model of MHC class I-restricted antigen processing. Immunol. Today 13:173.[Medline]
  40. Van Kaer, L., P. Ashton-Rickardt, M. Eichelberger, M. Gaczynska, K. Nagashima, K. Rock, A. Goldberg, P. Doherty, S. Tonegawa. 1994. Altered peptidase and viral-specific response in LMP2 mutant mice. Immunity 1:533.[Medline]
  41. Monaco, J. J.. 1992. Major histocompatibility complex-linked transport proteins and antigen processing. Immunol. Res. 11:125.[Medline]
  42. Townsend, A., J. Towsdale. 1993. The transporter associated with antigen presentation. Semin. Cell Biol. 4:53.[Medline]
  43. Momburg, F., J. Roelse, J. Howard, G. Butcher, G. Hammerling, J. Neefjes. 1994. Selectivity of MHC-encoded peptide transporters from human, mouse and rat. Nature 367:648.[Medline]
  44. Powis, S., E. Deverson, W. Coadwell, A. Ciruela, N. Huskisson, H. Smith, G. Butcher, J. Howard. 1992. Effect of polymorphism of an MHC-linked transporter on the peptides assembled in a class I molecule. Nature 357:211.[Medline]
  45. Sijts, A. J. A. M., A. Neisig, J. Neefjes, E. G. Pamer. 1996. Two Listeria monocytogenes CTL epitopes are processed from the same antigen with different efficiencies. J. Immunol. 156:685.
  46. Rock, K. L., C. Gramm, B. Benacerraf. 1991. Low temperature and peptides favor the formation of class I heterodimers on RMA-S cells at the cell surface. Proc. Natl. Acad. Sci. USA 88:4200.[Abstract/Free Full Text]
  47. Schumacher, T. N., M. T. Heemels, J. J. Neefjes, W. M. Kast, C. J. Melief, H. L. Ploegh. 1990. Direct binding of peptide to empty MHC class I molecules on intact cells and in vitro. Cell 62:563.[Medline]
  48. Buchholz, D., P. Scott, N. Shastri. 1995. Presentation without proteolytic cleavage of endogenous precursors in the MHC class I antigen processing pathway. J. Biol. Chem. 270:6515.[Abstract/Free Full Text]
  49. Eisenlohr, L. C., J. W. Yewdell, J. R. Bennink. 1992. Flanking sequences influence the presentation of an endogenously synthesized peptide to cytotoxic T lymphocytes. J. Exp. Med. 175:481.[Abstract/Free Full Text]
  50. Shastri, N., T. Serwold, F. Gonzalez. 1995. Presentation of endogenous peptide/MHC class I complexes is profoundly influenced by specific C-terminal flanking residues. J. Immunol. 155:4339.[Abstract]
  51. Rosenfeld, R., S. Vajda, C. DeLisi. 1995. Flexible docking and design. Annu. Rev. Biophys. Biomol. Struct. 24:677.[Medline]
  52. Rognan, D., M. J. Reddehase, U. H. Koszinowski, G. Folker. 1992. Molecular modeling of an antigenic complex between a viral peptide and a class I MHC glycoprotein. Proteins 13:70.[Medline]
  53. Lim, J.-S., S. Kim, H.-G. Lee, K.-Y. Lee, T.-J. Kwon, K. Kim. 1996. Selection of peptides that bind to the HLA-A2.1 molecule by molecular modeling. Mol. Immunol. 33:221.[Medline]
  54. Rognan, D., L. Scapozza, G. Folkers, A. Daser. 1994. Molecular dynamics simulation of MHC-peptide complexes as a tool for predicting potential T cell epitopes. Biochemistry 33:11476.[Medline]
  55. Rognan, D., L. Scapozza, G. Folkers, A. Daser. 1995. Rational design of nonnatural peptides as high-affinity ligands for the HLA-B*2705 human leukocyte antigen. Proc. Natl. Acad. Sci. USA 92:753.[Abstract/Free Full Text]
  56. Ruppert, J., J. Sidney, E. Celis, R. T. Kubo, H. M. Grey, A. Sette. 1993. Prominent role of secondary anchor residues in peptide binding to HLA-A2.1 molecules. Cell 74:929.[Medline]
  57. Shibata, K., M. Imarai, G. M. Van Bleek, S. Joyce, S. G. Nathenson. 1992. Vesicular stomatitis virus antigenic octapeptide N52-59 is anchored into the groove of the H-2Kb molecule by the side chains of three amino acids and the main-chain atoms of the amino terminus. Proc. Natl. Acad. Sci. USA 89:3135.[Abstract/Free Full Text]
  58. Parker, K. C., M. A. Bednarek, J. E. Coligan. 1994. Scheme for ranking potential HLA-A2-binding peptides based on independent binding of individual peptide side-chains. J. Immunol. 152:163.[Abstract]



This article has been cited by other articles:


Home page
 J. Immunol.Home page
Z. Li, M. Zhang, C. Zhou, X. Zhao, N. Iijima, and F. R. Frankel
Novel Vaccination Protocol with Two Live Mucosal Vectors Elicits Strong Cell-Mediated Immunity in the Vagina and Protects against Vaginal Virus Challenge
J. Immunol., February 15, 2008; 180(4): 2504 - 2513.
[Abstract] [Full Text] [PDF]

Home page
 J. Virol.Home page
J. Lin, Y. Zhi, L. Mays, and J. M. Wilson
Vaccines Based on Novel Adeno-Associated Virus Vectors Elicit Aberrant CD8+ T-Cell Responses in Mice
J. Virol., November 1, 2007; 81(21): 11840 - 11849.
[Abstract] [Full Text] [PDF]

Home page
 CVIHome page
E. A. Reap, S. A. Dryga, J. Morris, B. Rivers, P. K. Norberg, R. A. Olmsted, and J. D. Chulay
Cellular and Humoral Immune Responses to Alphavirus Replicon Vaccines Expressing Cytomegalovirus pp65, IE1, and gB Proteins
Clin. Vaccine Immunol., June 1, 2007; 14(6): 748 - 755.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
C. Ranasinghe, S. J. Turner, C. McArthur, D. B. Sutherland, J.-H. Kim, P. C. Doherty, and I. A. Ramshaw
Mucosal HIV-1 Pox Virus Prime-Boost Immunization Induces High-Avidity CD8+ T Cells with Regime-Dependent Cytokine/Granzyme B Profiles
J. Immunol., February 15, 2007; 178(4): 2370 - 2379.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
J. N. Kochenderfer, C. D. Chien, J. L. Simpson, and R. E. Gress
Synergism between CpG-Containing Oligodeoxynucleotides and IL-2 Causes Dramatic Enhancement of Vaccine-Elicited CD8+ T Cell Responses
J. Immunol., December 15, 2006; 177(12): 8860 - 8873.
[Abstract] [Full Text] [PDF]

Home page
 Infect. Immun.Home page
L.-M. Chen, G. Briones, R. O. Donis, and J. E. Galan
Optimization of the Delivery of Heterologous Proteins by the Salmonella enterica Serovar Typhimurium Type III Secretion System for Vaccine Development.
Infect. Immun., October 1, 2006; 74(10): 5826 - 5833.
[Abstract] [Full Text] [PDF]

Home page
 J. Virol.Home page
X. Zhao, M. Zhang, Z. Li, and F. R. Frankel
Vaginal Protection and Immunity after Oral Immunization of Mice with a Novel Vaccine Strain of Listeria monocytogenes Expressing Human Immunodeficiency Virus Type 1 gag.
J. Virol., September 1, 2006; 80(18): 8880 - 8890.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
L. B. Arruda, D. Sim, P. R. Chikhlikar, M. Maciel Jr, K. Akasaki, J. T. August, and E. T. A. Marques
Dendritic Cell-Lysosomal-Associated Membrane Protein (LAMP) and LAMP-1-HIV-1 Gag Chimeras Have Distinct Cellular Trafficking Pathways and Prime T and B Cell Responses to a Diverse Repertoire of Epitopes
J. Immunol., August 15, 2006; 177(4): 2265 - 2275.
[Abstract] [Full Text] [PDF]