| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,¶
* La Jolla Institute for Allergy and Immunology, San Diego, CA 92109;
Tumor Immunology Unit, Weatherall Institute of Molecular Medicine, Oxford University, Oxford, United Kingdom;
Epimmune Incorporated, San Diego, CA 92121;
Laboratory of Viral Diseases, National Institutes of Health, Bethesda, MD 20892; and
¶ Division of Immunology and Infectious Diseases, Queensland Institute of Medical Research, Herston, Queensland, Australia
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
20-fold larger than that of the model RNA viruses typically studied in mice. Based on predictive computational algorithms for peptide binding to HLA supertypes, we synthesized a panel of 2889 peptides to begin to create an immunomic map of human CTL responses to poxviruses. Using this panel in conjunction with CTLs from vaccinia virus-infected HLA transgenic mice, we identified 14 HLA-A*0201-, 4 HLA-A*1101-, and 3 HLA-B*0702-restricted CD8+ T cell determinants distributed over 20 distinct proteins. These peptides were capable of binding one or multiple A2, A3, and B7 supertype molecules with affinities typical of viral determinants. Surprisingly, many of the viral proteins recognized are predicted to be late gene products, in addition to the early intermediate gene products expected. Nearly all of the determinants identified have identical counterparts encoded by modified vaccinia virus Ankara as well as variola virus, the agent of smallpox. These findings have implications for the design of new smallpox vaccines and the understanding of immune responses to large DNA viruses in general. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
200 kb encoding >200 identifiable nonoverlapping open reading frames (ORFs). Recent concerns that variola virus (VARV), the agent of smallpox, could potentially be reintroduced into nature as a bioterrorism weapon, as well as recent outbreaks of zoonotic Orthopoxvirus infections, have led to renewed interest in poxvirus vaccination (6). Eradication of smallpox was based on immunization with vaccinia virus Western reserve (WR) strain (VACV), a poxvirus of unknown origin, with an uncertain relationship to Jenner’s vaccine, and whose identity remains shrouded in mystery. Although smallpox eradication is perhaps the greatest public health accomplishment in human history (6), VACV has a number of serious shortcomings as a modern day vaccine (7). It is therefore of great importance to develop new improved vaccines that protect against smallpox, but with less adverse reactions and diminished potential for transmission from vaccinees to others.
The VACV and VARV genomes are 90% homologous. It is believed that both T and B cell responses directed to identical or highly homologous target sequences play a vital role in heterologous protection against VARV afforded by vaccination with vaccinia virus. Several lines of evidence suggest that cellular immunity makes an important contribution to smallpox immunity (8, 9, 10, 11, 12). Ultimately, it will be impossible to ascertain the precise mechanisms that contribute to protective VARV immunity, and improved vaccines will have to be based on the use of modified forms of live VACV immunization. The efficacy of such vaccines will have to be based on their abilities to recapitulate responses induced by Dryvax and other VACV preparations that were shown to offer good immunity to smallpox.
Thus, in addition to its intrinsic interest in terms of understanding immunodominance in humans to large viruses, it is critical to characterize the HLA-restricted T cell responses to VACV. To date, reports have been limited to the description of just a few determinants presented by a single restricting MHC class I molecule (13, 14, 15). Specifically, Ennis and coworkers (14) reported the identification of two different HLA-A*0201-restricted epitopes induced by vaccinia vaccination and conserved among vaccinia and VARVs. Drexler and coworkers (13) reported the identification of an HLA-A*0201-restricted epitope using a combination of bioinformatics predictions and assay in HLA transgenic mice. The paucity of information stems partly from the daunting size of the poxvirus genome and difficulties in obtaining samples from vaccinees.
To deal with the large poxvirus coding capacity, we have used computer algorithms to predict potential immunogenic peptides. Previous work from numerous systems has established that the greatest single hurdle to immunogenicity for potential viral determinants is their affinity for class I molecules (3). Generally, only 1% of peptides are able to bind any given MHC class I allomorph above the threshold affinity associated with immunogenicity (KD = 500 nM). In the present study, we tackle this problem by using a computational approach to predict HLA-binding peptides encoded in the VACV genome. Although predictive algorithms are not perfect, they successfully identify 80% of class I binding peptides. To this end, we generated a panel of 2889 synthetic peptides.
To surmount problems with obtaining VACV-specific CTLs from human volunteers, we used HLA transgenic mice. Numerous studies have shown that antiviral CTL responses in humans and HLA transgenic mice overlap to a remarkable degree (16, 17, 18, 19, 20). This is consistent with the substantial literature documenting that the Ag-processing machinery is highly conserved between mouse and human cells (21, 22). In this study we describe how this approach has resulted in the identification of 20 new antigenically active determinants that bind to HLA molecules with high affinity.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Peptides used in initial screening experiments were tested without chromatographic purification and were synthesized as crude material by Mimotopes, Pepscan Systems BV, or A and A Labs, as described elsewhere (23). Candidate epitopes identified in pool deconvolution studies, or peptides use as radiolabeled ligands for binding assays, were resynthesized by A and A Labs and purified to >95% homogeneity by reverse-phase HPLC. Purity of these peptides was determined using analytical reverse-phase HPLC and amino acid analysis, sequencing, and/or mass spectrometry. Peptides were radiolabeled with the chloramine T method, as described elsewhere (24).
MHC peptide-binding assays
Quantitative assays to measure the binding affinity of peptides to purified HLA-A2- (A*0201, A*0202, A*0203, A*0206, A*6802), -A3- (A*0301, A*1101, A*3101, A*3301, A*6801), and -B7- (B*0702, B*3501, B*5101, B*5301, B*5401) supertype molecules are based on the inhibition of binding of a radiolabeled standard peptide, and were performed as previously described (23, 24, 25, 26). Briefly, 1–10 nM of radiolabeled peptide was coincubated at room temperature with 1 µM to 1 nM purified MHC in the presence of 1–3 µM human 
2-microglobulin (Scripps Laboratories) and a mixture of protease inhibitors. After a 2-day incubation, binding of the radiolabeled peptide to the corresponding MHC class I molecule was determined by capturing MHC/peptide complexes on Greiner Lumitrac 600 microplates (Greiner Bio-One) coated with the W6/32 Ab, and measuring bound cpm using the TopCount microscintillation counter (Packard Instrument).
For competition assays, the concentration of peptide yielding IC50 of the binding of the radiolabeled peptide was calculated. Peptides were typically tested at six different concentrations covering a 100,000-fold dose range, and in three or more independent assays. Under the conditions used, where (label) < (MHC) and IC50 
(MHC), the measured IC50 values are reasonable approximations of the true Kd values.
Bioinformatic analyses
An unpublished VACV (WR) sequence provided by Dr. Bernard Moss (National Institute of Allergy and Infectious Diseases), which has been the basis of other studies (27, 28), has been used for this study. To identify candidate epitopes for use in ORF-specific peptide pools, each predicted ORF of the vaccinia WR strain was analyzed using previously described algorithms (4) that predict the affinity of peptides for specific HLA class I molecules. The panel of algorithms used allowed prediction of ligands for HLA-A2, -A3, and -B7-supertype molecules. For each gene, peptides predicted to bind with an IC50 < 100 nM were selected for study. To reduce the number of predicted peptides identified in large ORFs, predictions for all supertypes were combined, and the best scoring 40 candidates, regardless of supertype, were selected. As a result, 964 A2-supertype, 1584 A3-supertype, and 341 B7-supertype peptides were selected and synthesized for study.
Mice
HLA-A*0201/Kb, HLA-A*1101/Kb, and HLA-B*0702/Kb transgenic mice used in this study were the F1 generation derived from crossing homozygous transgenic mice (H-2b haplotype) expressing a chimeric gene consisting of the 
1 and 2 domains of HLA and the 3 domain of H-2Kb (17, 18, 19) with BALB/c mice (The Jackson Laboratory). HHD A2 mice (29) were bred at the Weatherall Institute of Molecular Medicine. The other transgenic mice were bred and maintained at the La Jolla Institute for Allergy and Immunology facility (San Diego, CA) following National Institutes of Health guidelines and Institutional Animal Care and Use Committee-approved animal protocols.
Stimulator cells and cell lines in ELISPOT assays
Stimulator cells used for peptide-specific IFN-
release were Jurkat cells transfected with the HLA-A*0201/Kb or the HLA-B*0702/Kb chimeric genes, or the .221A*1101/Kb cell line (17, 18, 30). These cell lines were transfected with the same HLA construct expressed in the corresponding HLA transgenic mice used in this study. The cell line .221A*1101/Kb was derived by transfection of the HLA-A*1101/Kb gene into 3A4-721.221 tumor cells, which lacks expression of HLA-A, -B, or -C class I genes due to gamma-ray-induced deletions in the HLA complex (31). LPS-stimulated B lymphoblasts obtained by cultivating splenocytes in the presence of LPS (8.5 µg/ml) and dextran sulfate (7 µg/ml) (Sigma-Aldrich), for 3 days at 37°C, were also used as stimulator cells. All cells were grown in RPMI complete culture medium (RPMI 1640 medium, 25 nM HEPES (pH 7.4; Invitrogen Life Technologies), supplemented with 10% FBS, 4 mM L-glutamine, 5 x 10–5 M 2-ME, 0.5 mM sodium pyruvate, 0.1 mM MEM nonessential amino acids, 100 µg/ml streptomycin, and 100 IU/ml penicillin).
Viruses
The WR strain of vaccinia virus was obtained from Dr. Bernard Moss (National Institute of Allergy and Infectious Diseases).
Infection and immunizations
HLA transgenic mice were infected i.p. with 2 x 106 PFU of vaccinia WR strain (VACV) in PBS. After 7 days, the mice were sacrificed, and the splenocytes or purified CD8+ T cells were used for ex vivo mouse-IFN- enzyme-linked immunospot (ELISPOT) measurement. For peptide immunization, mice were immunized s.c. with a mixture of peptide (10 µg/mouse) and the helper IA-b-restricted epitope, HBV core 128–140 (140 µg/mouse) (32) in PBS/10% DMSO emulsified in IFA. After 12–14 days, the mice were sacrificed, and CD8+-purified T cells were used for ex vivo ELISPOT measurement.
Ex vivo ELISPOT assays
The ELISPOT assays were performed as previously described (33). Briefly, either 4x 105 splenocytes or 2 x 105 splenic CD8+ T cells (isolated by anti-CD8-coated magnetic beads (Miltenyi Biotec)) were cultured with peptide pulsed or VACV WR-infected stimulator cells. The stimulator cells were 1 x 105 Jurkat-A*0201/Kb, 1 x 105 Jurkat-B*0702/Kb cells, 1 x 104 .221-A*1101/Kb, or 1 x 104 LPS blasts, in flat-bottom 96-well nitrocellulose plates (Immobilon-P membrane; Millipore), which had been precoated with anti-IFN- mAb (BD Pharmingen; 4 µg/ml). After 20-h incubation at 37°C, plates were washed with PBS/0.05% Tween, and wells were incubated with biotinylated anti-IFN- mAb (BD Pharmingen; 1 µg/ml) for 4 h at 37°C. After additional washing, spots were developed by sequential incubation with Vectastain ABC peroxidase (Vector Laboratories) and 3-amino-9-ethylcarbazole solution (Sigma-Aldrich) and counted by computer-assisted image analysis (Zeiss KS ELISPOT Reader).
Each assay was performed in six replicate wells, and the experimental values were expressed as the mean net spots/106 unfractionated splenocytes or CD8+ lymphocytes ± SEM for each peptide. Responses against irrelevant peptides (HCV core 132, DLMGYIPLV for A*0201; HBV env. 378, LLPIFFCLWV for B*0702; and Human Mage3 69, SSLPTTMNY for A*1101) were measured to establish background values that were subtracted from the experimental values. To determine the level of significance, a Student’s t test was performed in which p 
0.05 using the mean of triplicate values of the response against relevant peptides vs the response against irrelevant control peptides. The net number of spots/106 effector cells was calculated as follows: [(number of spots against relevant peptide) – (number of spots against irrelevant control peptide)] x [(1 x 106)/(number of effector cells/well)]. The stimulation index (SI) was calculated as follows: (number of spots against relevant peptide)/(number of spots against irrelevant control peptide).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
In previous studies, we demonstrated that 95% of HLA-restricted determinants bind their relevant HLA molecules with a KD of 500 nM or less (34). Subsequent studies described computer algorithms that allow the prediction of the binding affinity of potential determinants to specific HLA molecules (4). In using these algorithms, different affinity thresholds can be selected. High-affinity thresholds generate highly accurate predictions at the cost of capturing a small fraction of determinants. Lower affinity thresholds capture a progressively larger fraction of the determinants, at the expense of lower accuracy (i.e., more false predictions are generated). In this study, we use predictive thresholds previously established to predict 75% of HLA-restricted determinants (4).
A large fraction of HLA class I allomorphs can be assigned into a relatively few supertypes, each characterized by largely overlapping peptide-binding repertoires reflecting consensus structures in the main peptide binding pockets (35). Predictions were performed for the HLA-A2, -A3, and -B7 supertypes. At least one of these three supertype specificities is present in 85% of the human worldwide population among all ethnic backgrounds (34).
These algorithms were applied to the 258 predicted ORFs in the complete VACV-WR genome. Although not all of these ORFs are known to be expressed, we included all of them, because it has been shown that poorly expressed or aberrant transcripts can yield T cell determinants (36). We synthesized 964 HLA-A2, 1584 HLA-A3, and 341 HLA-B7 potential ligands from the poxvirus proteome based on in silico prediction of high or intermediate binding.
Identification of 21 different VACV-derived determinants
We determined the antigenicity of synthetic peptides using splenocytes obtained from HLA transgenic mice 1 wk after i.p. infection with VACV. The derivation of HLA-A*0201/Kb and HLA-B*0702/Kb mice has been described elsewhere (17, 19). To identify HLA-A3 supertype peptides, we generated HLA-A*1101/Kb transgenic mice (18). Splenocytes were incubated ex vivo with Jurkat HLA-A*0201/Kb-, Jurkat HLA-B*0702/Kb-, and .221-A*1101/Kb-transfected cell lines, respectively, that had been pulsed with pools of the corresponding potentially antigenic peptides (each pool contained 15 peptides), scoring antigenicity by ELISPOT assay for IFN- secretion. The results from individual experiments were highly reproducible, and an average of 2–3 independent experiments were used to identify antigenically active pools.
The results of these initial screens are shown in Fig. 1 for HLA-A*0201/Kb, -A*1101/Kb, and -B*0702/Kb transgenic mice, respectively. Peptide pools generating an average of >20 spot-forming cells (SFCs)/106 cells (shown in black in Fig. 1) and an average SI >1.4 were selected for further characterization experiments. This low SI threshold was purposely selected to allow for a comprehensive identification, and is on the side of inclusiveness in the epitope identification process.
|
Using transfected cell lines, we next identified the individual peptides in each pool that were responsible for the antigenic activity detected. As expected, based on the low criteria for pool positiveness used, as described above, for several of the pools that gave weak responses, no definitive individual peptide could be identified that stimulated a response. However, we also were able to identify 14 HLA-A*0201-, 4 HLA-A*1101-, and 3 HLA-B*0702-restricted determinants that gave reproducibly robust signals (Fig. 2) (SI above 2 in multiple independent experiments). One of these epitopes was previously and independently been identified by Drexler et al. (13).
|
Having identified a number of candidate epitopes, we needed to demonstrate that they were truly recognized by TCD8+ restricted by the presumed HLA gene product. The 14 determinants identified in VACV WR-infected HLA-A*0201 transgenic mice were tested by using APCs expressing various mouse or human MHC class I molecules and varying doses of the peptide determinants to stimulate purified CD8+ T cells from the HLA-A*0201/Kb-immunized mice. Specifically, we used LPS-stimulated B lymphoblasts from the HLA-A*0201/Kb transgenic CB6F1 mice; LPS B cell blasts from CB6F1 mice, which do not express the HLA transgene, and Jurkat HLA-A*0201/Kb, a human cell line transfected with the same HLA-A*0201/Kb construct expressed in the HLA transgenic mice. Additional controls included LPS blasts from the unrelated mouse strain CBA/J (H-2k haplotype), as well as LPS blasts derived from KbDb knockout mice, which do not express any classical MHC class I molecules.
All 14 determinants were able to activate highly purified TCD8+ (Fig. 3, and Table I). Two patterns of MHC restriction were observed. Eleven peptides demonstrated robust antigenicity using HLA-A*0201/Kb Jurkat cells and HLA-A*0201/Kb transgenic B blasts, but not with any of the APCs not expressing the HLA-A*0201 transgene (Fig. 3, A and B, show representative examples). For three other determinants (M1L, 374–383; A17L, 61–70; and B6R, 108–116) a more complex pattern was seen. These peptides were antigenic when presented by either HLA-A*0201/Kb-expressing target cells or CB6F1 B blasts, but not to peptide-pulsed LPS blasts derived from the control H-2k mice or class I knockout mice (see Fig. 3, C and D, for representative data). Thus, these peptides are presented by either HLA-A*0201 or one of the endogenous mouse class I coexpressed in the HLA-A*0201/Kb transgenic mice.
|
|
Immunogenicity of the HLA-A*0201-restricted determinants in various mice strains
We extended these studies to purified TCD8+ obtained from VACV WR-infected mice of various MHC haplotypes for the ability to respond to the antigenic peptides presented by autologous LPS blasts (Table I). As expected, we failed to detect responding TCD8+ to any of the peptides in CBA/J mice, or in H-2Kb–/–Db–/– mice, which lack expression of any MHC class I molecules. Concordant with the results described above, 11 of the 14 (78.6%) HLA-A*0201-restricted peptides were only recognized by mice expressing HLA-A*0201. For nine of these determinants, vigorous responses were also observed in HHD A2 mice, which only express HLA-A*0201 (i.e., not Kb or Db). Interestingly, two peptides (G7L (250–258) and VACWR050 (196–204)) were recognized by HLA-A*0201/Kb TCD8+ but not by HHD (or CB6F1) TCD8+. These data suggest that the presence of mouse class I MHC molecules are required to generate the TCD8+ repertoire capable of responding to these peptides.
The three peptides for which both mouse and human restriction was demonstrated above were recognized by TCD8+ from CB6F1 mice, further confirming that these determinants can be restricted by the endogenous mouse class I molecules as well as the human HLA-A*0201 transgene. Responses to two of the peptides (M1L (374–383) and A17L (61–70)) were generated in both HHD and CB6F1 mice, clearly demonstrating this dual specificity. The B6R (108–116) determinant was immunogenic in CB6F1 and the HLA-A*0201/Kb transgenic mice, but not in HHD mice. As postulated above, this suggests that mouse class I expression was required for the generation of a T cell repertoire capable of recognizing this determinant in the context of HLA-A*0201 as well as in the context of mouse class I. Notably, the sum of TCD8+ responses to the 14 individual peptides in the various mouse strains tested was equal to or exceeded the magnitude of the responses detected against VACV-infected target cells. This suggests that a significant proportion of the total response to the virus can be accounted for by the response to these determinants.
In the same set of experiments, we tested a previously reported HLA-A*0201-restricted determinant, HRP2 (74–82) (14). Consistent with these earlier observations, we observed a significant response, albeit comparatively modest in VACV-infected HHD mice and no response in HLA-A*0201/Kb transgenic mice. This observation can be interpreted in a reverse manner to that posited above, namely that the presence of mouse class I leads to a decrease in the repertoire of T cells capable of recognizing the HRP2 (74–82) determinant in the context of HLA-A*0201.
Identification of determinants restricted by HLA-A*1101 and HLA-B*0702
We extended this approach to characterize determinants recognized by TCD8+ from HLA-A*1101/Kb and HLA-B*0702/Kb transgenic mice. Starting with four determinants recognized most robustly by HLA-A*1101/Kb transgenic mice, we used target cells from various mice strains as well as .221-A*1101/Kb-transfected target cells to determine that two determinants (I3L (116–124) and E7R (130–138)) were restricted by HLA-A*1101 and not CB6F1 class I molecules (Fig. 4A, and data not shown). Consistent with this finding, neither of these determinants induced responses in CB6F1-VACV-infected mice. I3L (116–124) induced a response in VACV-infected HLA-A*1101/Kb mice that did not express mouse class I molecules (Table II).
|
|
Two of the three HLA-B*0702-restricted determinants (J2R (116–124) and D1R (808–817)) strictly required HLA-B*0702/Kb expression for antigenicity, and were not presented by CB6F1-derived LPS blasts (data not shown). The third determinant (A34R (82–90)) exhibited dual restriction in the experiments with different target cells, although the reactivity to CB6F1-pulsed target cells was relatively weak and only detected at the highest Ag dose tested (data not shown). Consistent with these findings, when the three peptides were tested in immunized HLA-B*0702/Kb and wild-type CB6F1 mice, J2R (116–124) and D1R (808–817) were specifically recognized by HLA-B*0702/Kb mice but not by CB6F1, whereas A34R (82–90) was recognized by both transgenic and nontransgenic CB6F1 mice (Table III). HLA-B*0702/Kb transgenic mice that lacked expression of mouse class I molecules were not available for testing.
|
It was important, of course, to test the capacity of the candidate determinants to bind to their respective class I-restricting elements (Table IV). All of the peptides bound their restricting elements with at least intermediate affinity (500 nM). Significantly, 12 of 14 (85.7%) -A2-, 3 of 4 (75%) -A11-, and 2 of 3 (67%) -B7-restricted determinants bound with very high affinity (<20 nM).
|
We also analyzed these HLA determinants for the presence of H-2 binding motifs or H-2 binding capacity. In 6/6 cases of determinants exhibiting dual restriction, the peptide either contained a canonical mouse class I motif or bound with high affinity (100 nM or less) to one of the relevant purified mouse class I molecules (data not shown).
In conclusion, the data presented in this section supports the functional restriction data obtained from TCD8+ activation assays and suggest that the A2- and A3-supertype determinants identified, because of their high degree of cross-reactivity, could be used to elicit or monitor responses in the context of multiple HLA-class I molecules.
Peptide immunization induces CD8+ T cells capable of recognizing VACV-infected human cells
An important potential limitation of using HLA transgenic mice is that mouse and human cells generate distinct peptide repertoires due to differences in the Ag-processing and class I assembly machineries. To address this question, we generated determinant-specific TCD8+ by immunizing HLA-A*0201/Kb transgenic mice with eight representative HLA-A*0201-restricted peptide. Purified TCD8+ were then tested for activation against peptide-pulsed or VACV-infected APCs (Fig. 5). For all determinants tested, vigorous IFN- production was observed when VACV-infected HLA-A*0201/Kb Jurkat were used as stimulator cells. This response was similar in magnitude to that elicited by HLA-A*0201/Kb autologous LPS Blasts (data not shown), whereas no response was observed when CB6F1-derived stimulator cells were used (data not shown). The level of response to the VACV-infected cells was similar to that with peptide-pulsed stimulator cells, suggesting that the HLA-restricted determinants identified in the HLA transgenic system are generated efficiently by natural processing in infected human cells.
|
We have identified determinants derived from 20 different VACV proteins. Table V lists for each Ag its name as defined in the Poxvirus Bioinformatics Resource Center (www.poxvirus.org) (VACV protein name), as well as the name of the viral Ag according to McCraith et al. (27). Table V also lists known structural and biological characteristics of the Ags from which the determinants were derived. Among the variables listed are the predicted protein size and the stage specificity of expression (early/intermediate vs late) (27, 37, 38, 39, 40). A description of the predicted function is also listed, when available, together with a presumed functional category (e.g., virulence factors, virion structure, viral genome regulation). Finally, the Table lists the corresponding determinants, their restriction element, and the percentage homology of the VACV sequence to that found in the genomes of variola major India and Bangladesh and modified vaccinia virus Ankara (MVA) (www.poxvirus.org).
|
|
From the standpoint of vaccines, it is crucial to know the conservation of TCD8+ determinants among the poxvirus family (Table VI). Of the 21 class I-restricted VACV determinants identified, 18 (85.7%) and 17 (80.9%) were identical with homologues in variola major India and MVA viruses, respectively. These results underline the potential relevance of these determinants in the context of vaccination against smallpox, or for monitoring responses resulting from immunization with MVA-based vaccines.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
One pitfall of using synthetic peptide screening to identify determinants is the existence of "mimotopes" determinants that are antigenic but not immunogenic in the context of virus infections. In most cases this arises from limitation in the Ag-processing machinery to generate the peptide from viral gene products. Mimotope cross-reactivity is a severe problem when using synthetic peptides at very high concentrations (>10–8 M). Even highly potent peptides can represent mimotopes, however, as clearly shown by Belz et al. (41). Therefore, the epitopes identified by our study should be considered provisional until confirmed by other approaches that they are naturally processed immunogens.
Despite this caveat, based on past experience with peptides demonstrating in vitro antigenicity in the nanomolar range, it is unlikely that any more than a few of peptides we identify are mimotopes. With this in mind, our findings, in conjunction with those of other studies, indicate that the TCD8+ response to VACV (WR strain) does not focus on a few immunodominant proteins, but rather is spread over at least 25 different gene products, including 20 distinct VACV proteins identified in the present study in conjunction with previous reports that identified five additional immunogenic Ags (14, 28). Thus, at least 10% of possible viral gene products can be recognized by at least 1 of the 3 human class I specificities analyzed. At the same time, however, the bulk of the TCD8+ response focuses on relatively few (1, 2, 3, 4) determinants, which account for 60% of the total response in a given animal. This finding is consistent with our previous findings in C57BL/6 mice, where a single VACV determinant of five determinants identified accounts for approximately one-third of the total response to VACV (28).
Our study also illustrates the complexity of immunodominance patterns. Although HLA-A*0201 serves as a restriction element for 14 determinants, only 3–4 determinants were restricted by B*0702 and A*1101. This cannot be attributed to differences in the performance of the predictive peptide binding algorithms, as determined by empirical measurements of predicted peptide binding to their respective class I molecules. Rather, this might relate, in the case of A*1101/Kb transgenic mice, to the inefficient transport by mouse TAP of the ligands preferred by A*1101, which carry a C-terminal-positive charge. In the case of B*0702, the rather restrictive specificity for peptides with P in position 2 (17) resulted to only 341 predicted HLA-B7-supertype binding peptides being screened in our immunogenicity studies, as opposed to 964 HLA-A*0201- and 1584 HLA-A*1101-predicted peptides. Thus, B*0702 may simply sample a smaller pool of peptides and present proportionally less determinants.
The majority of the epitopes identified in the current study are restricted by HLA-A*0201. Thus, the conclusion that responses to VACV do not focus on a few immunodominant proteins might not be correct in the case of other alleles. In fact, fewer epitopes were identified for B*0702 and A*1101 than for A*0201.
One of our most striking observations is that the sum of the responses against the determinants defined is of similar magnitude to the response against virus-infected cells in each of the transgenic mouse systems investigated. This is consistent with the identification of a large portion of determinants recognized by VACV-specific TCD8+ in the various mice. However, strong caveats need to be raised against this interpretation, which is probably overly simplistic. Aside from the issue of mimotopes, it is likely that the VACV-infected APCs used to measure overall responses fail to optimally present the full range of determinants presented in vivo.
It is noteworthy that 5 of 14 HLA-A*0201 and 3 of 4 HLA-A*1101-restricted determinants were not recognized by TCD8+ from VACV-infected HLA transgenic mice lacking the endogenous mouse MHC molecules. This suggests that the T cell repertoire is significantly influenced by nonrestricting mouse class I allomorphs. This finding has important implications for human immunity, where class I allomorphs are expressed in a large variety of permutations. In contrast to inbred mice, which express two or three class I allomorphs depending on the strain, humans express up to six distinct allomorphs of a total population of thousands of allomorphs present in appreciable frequencies in human populations. This may result in vastly different repertoires of T cells restricted by the same allomorph and, consequently, a lack of any discernable pattern of immunodominance in human responses to poxviruses.
Taken together, the breadth of responses to different VACV gene products together with the variable patterns of immunodominance observed have important implications in the context of natural or nefarious immune evasion strategies. If the immune responses focused on few dominant Ags and predictable patterns of immunodominance were in place, it is conceivable that mutant or doctored viruses lacking the immunodominant Ags or determinants could result from natural evolution or be created in the laboratory. Our findings suggest that this scenario is unlikely to occur.
Previously, it was reported that using inserted genes, early gene expression is associated with greater immunogenicity, possibly relating to restricted expression of early gene products in dendritic cells (42). Surprisingly, we failed to observe any bias toward early viral gene expression. Two factors could contribute to this discrepancy. First, it is possible, that the temporal control of "late" gene products under the control of natural viral promoters is not as stringent as the inserted gene products designed for this purpose. Second, whereas the inserted genes are excluded from virions, many of the natural late gene products are viral structural proteins, which may enter the class I-processing pathway of professional APCs during viral entry (43, 44). Indeed, we found that viral structural proteins are a preferred source of determinants. Less intuitively we found that virulence factors are also frequently recognized. This is probably related to their relatively abundant expression, because it appears that endogenous Ag-processing machinery largely samples from pools of defective ribosomal products, defective forms of viral proteins (45, 46) that are created as a by-product of errors in protein synthesis. The poor recognition of genome regulatory proteins being consistent with this model, inasmuch as the expected low abundance of these proteins should be commensurately represented by less pools of defective ribosomal product synthesis.
Our findings have important implications for diagnostic testing and vaccine development. We have identified a large number of determinants that will permit detailed experimental monitoring of cellular immune responses induced by poxvirus-specific immunization including vaccination of HLA-A2-, -A3-, and -B7-supertype-positive humans, which cover 85% (supertypes) of the human worldwide population, irrespective of ethnic background. These determinants bind their restricting element with high affinity and are candidates for the production of class I tetramer reagents, which we plan to make generally available to the scientific community. Our studies have only examined responses 7 days after immunization. In this respect it will be interesting to examine in future studies whether different epitopes are recognized at later points, or following rechallenge with VACV or heterologous viruses. The good news for vaccines is that there is high conservation of the 21 putative determinants we identify between VACV and its target pathogen VARV, and potential replacement vaccine, MVA. Future studies will address whether these determinants are recognized in humans following vaccination. Experiments are in progress with PBMC of human volunteers vaccinated with VACV or exposed to MVA. It will also be possible to use these peptides to test PBMC of individuals naturally exposed to poxviruses such as Monkeypox, responsible for a recent outbreak in the United States (6).
In conclusion, our study illustrates the complex patterns of immunodominance in responses to a large virus and identifies 21 different new VACV (WR strain)-derived determinants, thus greatly expanding the number of determinants available for further study.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported by Contract HHSN266200400024C and Grant RO1AI56268 from the National Institutes of Health. V.C. and F.M. are recipients of grants from the Cancer Research United Kingdom (C399-A2291) and United Kingdom Medical Research Council. D.C.T. is the recipient of the National Health and Medical Research Council of Australia Howard Florey Centenary Fellowship (224273). 
2 Address correspondence and reprint requests to Dr. Alessandro Sette, La Jolla Institute for Allergy and Immunology, 3030 Bunker Hill Street, Suite 326, San Diego, CA 92109. E-mail address: alex{at}liai.org
3 Abbreviations used in this paper: HBV, hepatitis B virus; HCV, hepatitis C virus; ORF, open reading frame; VARV, variola virus; WR, Western Reserve; VACV, vaccinia virus WR strain; SI, stimulation index; SFC, spot-forming cell; MVA, vaccinia virus Ankara.
Received for publication June 6, 2005. Accepted for publication July 27, 2005.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
-producing T cells after smallpox vaccination. J. Infect. Dis. 185:1657.-1659. [Medline]
2 microglobulin (2m) HLA-A2.1 monochain transgenic H-2Db 2m double knockout mice. J. Exp. Med. 185:2043.-2051. Related articles in The JI:
This article has been cited by other articles:
|
|
 |
|
  E. Assarsson, H.-H. Bui, J. Sidney, Q. Zhang, J. Glenn, C. Oseroff, I. N. Mbawuike, J. Alexander, M. J. Newman, H. Grey, et al. Immunomic Analysis of the Repertoire of T-Cell Specificities for Influenza A Virus in Humans J. Virol., December 15, 2008; 82(24): 12241 - 12251. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. S. Meyer, W. Kastenmuller, G. Gasteiger, M. Franz-Wachtel, T. Lamkemeyer, H.-G. Rammensee, S. Stevanovic, D. Sigurdardottir, and I. Drexler Long-Term Immunity against Actual Poxviral HLA Ligands as Identified by Differential Stable Isotope Labeling J. Immunol., November 1, 2008; 181(9): 6371 - 6383. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Jing, D. H. Davies, T. M. Chong, S. Chun, C. L. McClurkan, J. Huang, B. T. Story, D. M. Molina, S. Hirst, P. L. Felgner, et al. An Extremely Diverse CD4 Response to Vaccinia Virus in Humans Is Revealed by Proteome-Wide T-Cell Profiling J. Virol., July 15, 2008; 82(14): 7120 - 7134. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Oseroff, B. Peters, V. Pasquetto, M. Moutaftsi, J. Sidney, V. Panchanathan, D. C. Tscharke, B. Maillere, H. Grey, and A. Sette Dissociation between Epitope Hierarchy and Immunoprevalence in CD8 Responses to Vaccinia Virus Western Reserve J. Immunol., June 1, 2008; 180(11): 7193 - 7202. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. G. Kenter, M. J.P. Welters, A.R. P.M. Valentijn, M. J.G. Lowik, D. M.A. Berends-van der Meer, A. P.G. Vloon, J. W. Drijfhout, A. R. Wafelman, J. Oostendorp, G. J. Fleuren, et al. Phase I Immunotherapeutic Trial with Long Peptides Spanning the E6 and E7 Sequences of High-Risk Human Papillomavirus 16 in End-Stage Cervical Cancer Patients Shows Low Toxicity and Robust Immunogenicity Clin. Cancer Res., January 1, 2008; 14(1): 169 - 177. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Gasteiger, W. Kastenmuller, R. Ljapoci, G. Sutter, and I. Drexler Cross-Priming of Cytotoxic T Cells Dictates Antigen Requisites for Modified Vaccinia Virus Ankara Vector Vaccines J. Virol., November 1, 2007; 81(21): 11925 - 11936. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Kastenmuller, G. Gasteiger, J. H. Gronau, R. Baier, R. Ljapoci, D. H. Busch, and I. Drexler Cross-competition of CD8+ T cells shapes the immunodominance hierarchy during boost vaccination J. Exp. Med., September 3, 2007; 204(9): 2187 - 2198. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. R. Mothe, B. S. Stewart, C. Oseroff, H.-H. Bui, S. Stogiera, Z. Garcia, C. Dow, M. P. Rodriguez-Carreno, M. Kotturi, V. Pasquetto, et al. Chronic Lymphocytic Choriomeningitis Virus Infection Actively Down-Regulates CD4+ T Cell Responses Directed against a Broad Range of Epitopes J. Immunol., July 15, 2007; 179(2): 1058 - 1067. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Mitra-Kaushik, J. Cruz, L. J. Stern, F. A. Ennis, and M. Terajima Human Cytotoxic CD4+ T Cells Recognize HLA-DR1-Restricted Epitopes on Vaccinia Virus Proteins A24R and D1R Conserved among Poxviruses J. Immunol., July 15, 2007; 179(2): 1303 - 1312. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Assarsson, J. Sidney, C. Oseroff, V. Pasquetto, H.-H. Bui, N. Frahm, C. Brander, B. Peters, H. Grey, and A. Sette A Quantitative Analysis of the Variables Affecting the Repertoire of T Cell Specificities Recognized after Vaccinia Virus Infection J. Immunol., June 15, 2007; 178(12): 7890 - 7901. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Moutaftsi, H.-H. Bui, B. Peters, J. Sidney, S. Salek-Ardakani, C. Oseroff, V. Pasquetto, S. Crotty, M. Croft, E. J. Lefkowitz, et al. Vaccinia Virus-Specific CD4+ T Cell Responses Target a Set of Antigens Largely Distinct from Those Targeted by CD8+ T Cell Responses J. Immunol., June 1, 2007; 178(11): 6814 - 6820. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. F. Kotturi, B. Peters, F. Buendia-Laysa Jr., J. Sidney, C. Oseroff, J. Botten, H. Grey, M. J. Buchmeier, and A. Sette The CD8+ T-Cell Response to Lymphocytic Choriomeningitis Virus Involves the L Antigen: Uncovering New Tricks for an Old Virus J. Virol., May 15, 2007; 81(10): 4928 - 4940. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Jing, T. M. Chong, B. Byrd, C. L. McClurkan, J. Huang, B. T. Story, K. M. Dunkley, L. Aldaz-Carroll, R. J. Eisenberg, G. H. Cohen, et al. Dominance and Diversity in the Primary Human CD4 T Cell Response to Replication-Competent Vaccinia Virus J. Immunol., May 15, 2007; 178(10): 6374 - 6386. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Tang, M. Murtadha, M. Schnell, L. C. Eisenlohr, J. Hooper, and P. Flomenberg Human T-cell responses to vaccinia virus envelope proteins. J. Virol., October 1, 2006; 80(20): 10010 - 10020. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. L. Smith, F. Mirza, V. Pasquetto, D. C. Tscharke, M. J. Palmowski, P. R. Dunbar, A. Sette, A. L. Harris, and V. Cerundolo Immunodominance of Poxviral-Specific CTL in a Human Trial of Recombinant-Modified Vaccinia Ankara J. Immunol., December 15, 2005; 175(12): 8431 - 8437. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
