Abstract
An individual's CD8+-cytotoxic-T-lymphocyte (CTL) response to Epstein-Barr virus (EBV) latent cycle antigens focuses on a small number of immunodominant epitopes often presented by just one of the available HLA class I alleles; for example, HLA-A11-positive Caucasians frequently respond to two immunodominant HLA A11 epitopes, IVTDFSVIK (IVT) and AVFDRKSDAK (AVF), within the nuclear antigen EBNA3B. Here, we reexamine the spectrum of EBV strains present in the highly HLA-A11-positive Chinese population for sequence changes in these epitopes relative to the Caucasian type 1 prototype strain B95.8. The IVT epitope was altered in 61 of 64 Chinese type 1 viruses, with four different sequence variants being observed, and the AVF epitope was altered in 46 cases with six different sequence variants; by contrast, all 10 Chinese type 2 viruses retained the prototype 2 epitope sequences. All but one of the type 1 epitope variants were poorly recognized by IVT- or AVF-specific CTLs in pulse-chase assays of peptide-mediated target cell lysis. More importantly, we screened HLA-A11-positive Chinese donors carrying viruses with known epitope mutations for evidence of epitope-specific CTL memory by enzyme-linked immunospot assays: none of the type 1 variants tested, nor the type 2 prototype, appeared to be immunogenic in vivo. The data remain consistent with the possibility that, during virus-host coevolution, pressure from the host CTL-mediated immune response has given A11 epitope-loss viruses a selective advantage.
The principal role of the major histocompatibility complex (MHC) class I-restricted CD8+-cytotoxic-T-lymphocyte (CTL) response is to control infection by intracellular pathogens, in particular viruses. This reflects the ability of such CTLs to recognize antigenic peptides, derived from the intracellular breakdown of endogenously expressed viral proteins and presented on the surface of infected cells as a complex with MHC class I molecules (25). For any one type of virus in the context of one MHC class I allele, the number of viral peptides that are presented at the surface and induce detectable responses is surprisingly small. Thus, the host CTL response even to viruses with a relatively large coding capacity is often focused on two or three immunodominant epitopes presented in the context of one or more MHC alleles (28). Consequently, for viruses which establish chronic infections and which have error-prone replication mechanisms, viral progeny carrying mutations in immunodominant epitopes may enjoy a selective advantage in the host. In humans, this is best illustrated by the evolution of CTL escape mutants of human immunodeficiency virus in infected individuals progressing to AIDS (9, 22).
The situation is quite different for agents such as herpesviruses which, in contrast to the above, are genetically stable and persist in the immune host by suppressing viral antigen expression rather than by quasispeciation. Such viruses have coevolved with their host species over thousands if not millions of years (18, 19), and one does not know to what extent CTL-mediated immune pressure has played a role in shaping such long-term virus evolution. Opportunities to address this question have to rely upon the study of contemporary virus strains and are only likely to be informative where special circumstances prevail, for instance, where particular MHC class I alleles are unusually frequent in the host population and where those alleles are known to be capable of eliciting strong CTL responses against defined viral epitopes.
Some years ago we and our collaborators fortuitously came upon just such circumstances in the course of studying CTL responses to Epstein-Barr virus (EBV), a herpesvirus widely distributed in human populations. We first identified two unusually strong HLA-A11-restricted epitopes within the Epstein-Barr nuclear antigen EBNA3B as encoded by the type 1 Caucasian prototype strain B95.8; these were the immunodominant IVTDFSVIK epitope (EBNA3B codons 416 to 424, called IVT) and the next most dominant AVFDRKSDAK epitope (EBNA3B codons 399 to 408, called AVF) (7, 8). Interestingly, these sequences were often mutated in EBV strains prevalent in the populations of lowland Papua New Guinea (6) and southern China (7), both areas where more than 50% individuals carry the HLA A11 allele (15). Furthermore, when the IVT mutations were examined, the corresponding peptide showed reduced affinity for the HLA A11 peptide binding groove as a result of amino acid changes in the primary anchor positions, 2 and 9, where amino acid side chains interact with binding pockets in the groove (16, 29). Also EBV-transformed B-lymphoblastoid-cell lines (LCLs) carrying virus strains with these mutations were less efficiently recognized by CTLs specific for the wild-type (i.e., B95.8 sequence) IVT epitope (6, 7, 16). Preliminary evidence suggested that the AVF mutants were also poorly recognized by wild-type AVF effectors (7), although this was not investigated in detail. This was interesting in that EBV's ability to infect and subsequently persist within a naive individual appears to depend on colonization of the B-cell system via a latent growth-transforming infection in which all the EBV latent proteins, including EBNA3B, are expressed (23). This growth-transforming infection elicits a strong primary CTL response (4, 26), and it was suggested that, at the point of horizontal transmission, viruses carrying IVT/AVF epitope mutations might have had a selective advantage in these highly HLA-A11-positive populations. This suggestion engendered considerable debate (10) and prompted a number of further studies (3, 11, 13) that called into question the significance of the changes observed.
In this and in an accompanying report (20), we readdress some of the key issues through a more detailed study of EBV strains prevalent in the southern Chinese population. In this paper, we describe a more comprehensive analysis of the IVT and AVF epitope variants prevalent in this population (versus those seen in Caucasian and African populations with lower HLA A11 prevalence) and then we examine both extended sets of variants, first for their antigenicity with respect to recognition by wild-type epitope-specific CTLs and second for their immunogenicity in the context of natural EBV infection in vivo.
MATERIALS AND METHODS
Virus strains and epitope sequencing.
Sequence analysis was carried out on 74 EBV strains from Chinese donors in Hong Kong and Canton, China. These included 31 virus isolates established by spontaneous LCL outgrowth from peripheral blood mononuclear cell (PBMC) cultures (from 17 healthy donors and 14 nasopharyngeal carcinoma [NPC] patients as recently described [21]) and 43 strains that were directly amplifiable from EBV DNA in the PBMCs of a different set of healthy donors. For comparison, we also analyzed 28 spontaneous LCL isolates from African donors (23 healthy donors and 5 Burkitt's lymphoma patients, all from Uganda) and 56 EBV strains from healthy Caucasian donors in the United Kingdom (36 spontaneous LCL isolates and 20 directly amplified from PBMCs).
All strains were analyzed alongside the B95.8 (Caucasian, prototype 1) and Ag876 (African, prototype 2) reference isolates and were initially characterized as type 1 or type 2 based on PCR amplification across type-specific sequences within EBNA2 and EBNA3C (24). To determine the IVT and AVF epitope sequences in the above-described strains, DNA was prepared from LCL pellets or from at least 2 × 107 PBMCs by standard methods. For LCL-derived DNA, the relevant region of the EBNA3B gene was amplified by PCR with the primers E3B8 (5′-CGCCAGTGCACCGGGAGACCC-3′, B95.8 coordinates 96331 to 96351) and E3B9 (5′-CAAAGGTTGCCATGGCTCCAG-3′, B95.8 coordinates 96873 to 96853). In the cases of PBMC-derived DNA, where the much lower EBV genome loads require more rounds of amplification (2), the relevant sequences were amplified by a nested PCR with the following primers: round 1, E3B8.2 (5′-AAGAAGGACCACACTCATATACG-3′, B95.8 coordinates 96298 to 96320) and E3B9.2 (5′-TTTTCAAGAAGGTCTAGCAT-3′, B95.8 coordinates 96962 to 96943); round 2, E3B8 and E3B9. All amplifications were performed under the following conditions: 40 cycles of 94°C for 30 s, 45°C for 90 s, and 72°C for 120 s. The PCR products were gel purified with a QIAquick gel extraction kit (Qiagen, Crawley, West Sussex, United Kingdom) and then directly sequenced with a BigDye, version 3.0, PCR sequencing kit (Applied Biosystems, Warrington, United Kingdom) with the E3B9 primer.
Cytotoxicity assays of epitope antigenicity.
Cytotoxicity assays of epitope antigenicity used effector CTL clones derived from HLA-A11-positive Caucasian donors CM and IM105 (both infected with virus strains encoding B95.8-like IVT and AVF epitopes) by in vitro stimulation of their PBMCs with the autologous B95.8 virus-transformed LCL; several IVT-specific and AVF-specific clones from each donor were tested. The target cells used for peptide loading were from an HLA-A11-positive Ag876 virus-transformed LCL already shown not to be recognized by the above-described clones. Standard cytotoxicity (chromium release) assays were conducted for 5 h at an effector-to-target cell ratio of 1:1. In a first series of experiments, targets were pretreated for 1 h with a range of concentrations (10−6 to 10−14 M) of the wild-type IVT or AVF peptides, of individual IVT or AVF variant peptides, or of the type 2 IVT or AVF peptides. Cytotoxicity assays were then conducted in the continued presence of the peptide at the same concentration. In a second series of experiments, targets were pretreated for 1 h with one of the same set of IVT and AVF peptides or their sequence variants either at the minimum concentration required to give maximum lysis in the previous set of assays or, in the case of peptides not recognized by the above-described assay, at 10−6 M. The cells were then washed well and incubated at 37°C in peptide-free medium for periods of up to 8 h before being used as targets in a standard cytotoxicity assay.
Elispot assays of epitope immunogenicity.
Healthy EBV antibody-positive Chinese donors were screened for A11 epitope responses without prior knowledge of their HLA A11 status or of the epitope sequences in their resident EBV strain. Parallel assays were conducted on healthy EBV antibody-positive Caucasian donors already known to be HLA A11 positive but without prior knowledge of their resident EBV strain. Cryopreserved PBMCs from the above-described donors were resuscitated and used as effectors in standard enzyme-linked immunospot (Elispot) assays of gamma interferon release (12, 27). Cells were seeded in replicate wells at 4 × 105, 2 × 105, and 105 cells/well with peptide at a 2 × 10−6 M concentration. Each donor's PBMCs were tested against the wild-type IVT and AVF peptides, the individual IVT and AVF peptide variants, and the type 2 IVT and AVF peptides; in addition, these experiments included the EBV latent membrane protein LMP2-derived peptide SSCSSCPLSK, a subdominant HLA-A11-restricted epitope whose sequence is known to be conserved between Caucasian and Chinese virus strains (14). Gamma interferon-producing cells were counted after development of the Elispot assay, and results were expressed as the number of spot-forming cells per 106 PBMCs.
RESULTS
A11 epitope variation in Chinese virus strains.
In total, 74 EBV strains of Chinese (Hong Kong or Canton) donor origin were analyzed, either as spontaneous LCL isolates from healthy donors and NPC patients or as sequences directly amplifiable from healthy donor PBMCs. On the basis of standard type-specific amplification assays at defined EBNA2 and EBNA3C gene loci, 64 of these strains were classified as uniformly type 1, 7 strains were classified as uniformly type 2, and 3 strains were classified as intertypic recombinants with type 1 EBNA2 and type 2 EBNA3C sequences (reference 21 and data not shown). The same viruses were then sequenced around the IVT/AVF epitope region in EBNA3B, giving sequences that were either type 1-like or type 2-like in accord with the original virus typing results. Note that all three intertypic recombinants carried type 2 sequences at the EBNA3B and 3C loci (21) and so, for the purpose of the present paper, are grouped with the other type 2 isolates. Within the epitopes themselves, we observed a number of coding changes, and these are shown in Table 1 relative to the type 1 B95.8 and the type 2 Ag876 prototype sequences.
TABLE 1.
Variant IVT and AVF epitope sequences in Chinese EBV strains
| Epitope | Type | Sequence name | Codon sequence (amino acid) at positiona:
|
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | |||
| IVT | 1 | wt1b | ata (I) | gta (V) | act (T) | gac (D) | ttt (F) | agt (S) | gta (V) | atc (I) | aag (K) | |
| IVT/L2 | tta (L) | |||||||||||
| IVT/L5 | ttg (L) | |||||||||||
| IVT/N9 | aat (N) | |||||||||||
| IVT/T9 | acg (T) | |||||||||||
| IVT/R9 | agg (R) | |||||||||||
| 2 | wt2c | gtg (V) | ctt (L) | ata (I) | ||||||||
| AVF | 1 | wt1 | gcg (A) | gtg (V) | ttt (F) | gac (D) | cga (R) | aag (K) | tca (S) | gat (D) | gca (A) | aaa (K) |
| AVF/P1 | ccg (P) | |||||||||||
| AVF/A2 | gcg (A) | |||||||||||
| AVF/N4 | aac (N) | |||||||||||
| AVF/S1F2 | tcg (S) | ttc (F) | ||||||||||
| AVF/P1L2 | ccg (P) | ttg (L) | ||||||||||
| AVF/S1L2 | tcg (S) | ttg (L) | ||||||||||
| 2 | wt2 | tcg (S) | tac (Y) | cca (P) | act (T) | |||||||
Complete sequences are given for Caucasian B95.8 prototype 1 for each epitope. For variants, only sequences containing variations from the prototype sequence are shown. Variations are shown in boldface type.
wt1, Caucasian B95.8 prototype 1 sequence.
wt2, African Ag876 prototype 2 sequence.
Among type 1 Chinese virus strains, we observed four different patterns of nucleotide change in the IVT epitope-coding sequence, all causing amino acid substitutions in the 9-mer epitope. These involved a V→L change at residue 2 (IVT/L2), an F→L change at residue 5 (IVT/L5), and a K→N change or a K→T change at residue 9 (IVT/N9, IVT/T9); the IVT/L5 variant has not been described before. Note that Table 1 also includes the IVT/R9 variant sequence observed in a rare Chinese isolate in earlier work (7). At the AVF epitope, six different sequence variants were identified among type 1 strains, and again, all led to amino acid changes. These involved an A→P change at position 1 (AVF/P1), a V→A change at position 2 (AVF/A2), a D→N change at position 4 (AVF/N4), and three different double substitutions at residues 1 and 2 (AVF/S1F2, P1L2, and S1L2); of these, the AVF/A2 and AVF/N4 variants have not been described before. Interestingly, the 10 type 2 viruses within the panel of Chinese strains exactly followed the prototype Ag876 sequence at both epitopes; these type 2 versions of the IVT and AVF peptides have two and four signature changes, respectively, relative to B95.8, none of which involve the primary anchor positions 2 and 9/10 for HLA A11 binding.
Table 2 shows the particular combinations of IVT and AVF epitope sequences and the numbers of EBV strains with these combinations. The three most common combinations among type 1 Chinese strains were AVF wild type plus IVT/L2, AVF/N4 plus IVT/N9, and AVF/S1F2 plus IVT/L2. One of these common combinations (involving the AVF/N4 variant) and two other rarer combinations, AVF/A2 plus IVT/L5 and AVF/S1F2 plus IVT/N9, have not been observed before. Note that the panel of 64 type 1 virus strains included three with a prototype B95.8 sequence at both epitopes; these strains (all from Hong Kong donors) were also identical to B95.8 at several other EBV latent gene loci (reference 20 and data not shown) and almost certainly represent the introduction of Caucasian strains as a result of the recent European colonization of Hong Kong. Otherwise there were no significant differences between the range of variants seen in Hong Kong versus Canton, nor between strains from healthy donors versus NPC patients, nor between spontaneous LCL isolates versus sequences amplified from PBMCs, except that strains with wild-type AVF sequences were relatively underrepresented in the spontaneous LCL panel (20, 21).
TABLE 2.
EBV strains of different geographic origins with particular IVT and AVF epitope sequence combinations
| Virus type | AVF epitope | IVT epitope | No. of strains from donors who are:
|
||
|---|---|---|---|---|---|
| Chinese | Caucasian | African | |||
| 1 | wt1a | wt1 | 3 | 28 | 23 |
| wt1 | IVT/N9 | 5 | |||
| wt2b | IVT/L2 | 15 | 2 | ||
| AVF/N4 | IVT/N9 | 18 | |||
| AVF/S1F2 | IVT/L2 | 14 | |||
| AVF/S1F2 | IVT/N9 | 1 | |||
| AVF/P1 | IVT/L2 | 4 | |||
| AVF/P1L2 | IVT/N9 | 4 | |||
| AVF/A2 | IVT/L5 | 2 | |||
| AVF/S1L2 | IVT/T9 | 2 | |||
| AVF/S1L2 | IVT/N9 | 1 | 2 | ||
| AVF/S1L2 | IVT/R9 | 1 | |||
| AVF/S1L2 | IVT/L2 | 9 | |||
| AVF/S1L2 | IVT/L2R9 | 4 | |||
| 2 | wt2 | wt2 | 10 | 5 | 5 |
wt1, Caucasian B95.8 prototype 1 epitope sequence.
wt2, African Ag876 prototype 2 epitope sequence.
We then extended the IVT and AVF epitope sequencing to include 56 EBV strains (51 type 1, 5 type 2) from healthy Caucasian donors, since earlier reports, based on small numbers of isolates, had given conflicting results in this context (3, 7). As shown in Table 2, a majority of type 1 Caucasian viruses (28 of 51) had the prototype B95.8 sequence at both epitopes. However, a significant minority were variant at IVT; of the four different variant sequences observed, two (IVT/L2 and IVT/N9) had been seen among the present Chinese strains, one (IVT/R9) had been reported in an earlier study of Chinese viruses (7) and one (IVT/L2R9) was novel. Many of the Caucasian strains with a variant IVT also carried a particular AVF variant sequence, AVF/S1L2; this was the only AVF variant seen in Caucasian viruses and was again represented, albeit rarely, in the present Chinese virus panel. By contrast, sequencing of 23 type 1 African isolates showed that every one followed the B95.8 prototype at both IVT and AVF loci. It was also noticeable that all type 2 virus strains within the Caucasian and African panels exactly followed the Ag876 prototype sequence across both epitopes, just as had the type 2 Chinese strains.
Antigenicity of A11 epitope variants.
Given that the above work had identified new sequence variants of both the IVT and AVF epitopes and that earlier studies of epitope antigenicity had largely focused on the then-known IVT variants (3, 6, 7, 16), we elected to screen both sets of variants for recognition by CTLs generated from Caucasian donors carrying type 1 viruses with wild-type (i.e., B95.8-like) IVT and AVF sequences. In each case we conducted two types of CTL assay, one in which target cells were exposed to a range of concentrations of the relevant synthetic peptide just before and during the 5-h assay itself and one in which target cells were preexposed to the relevant peptide at the minimum concentration eliciting maximal lysis in that first type of assay and then washed free of extraneous peptide and set up as targets in a 5-h assay either immediately or after a chase period of up to 8 h in normal medium.
Figure 1A shows the results obtained using two wild-type IVT-specific CTL clones tested against peptides representing the wild-type IVT sequence, the different IVT variants seen in type 1 viruses, and the prototype 2 sequence. A consistent pattern of results was seen with these and with other such clones. The wild-type, IVT/L2, and IVT/R9 peptides showed similar levels of maximal lysis and similar titration curves, with target cell lysis still detectable at 10−10 M peptide concentrations. The IVT/N9 and IVT/T9 peptides gave intermediate levels of lysis, but only at concentrations of 10−7 M or above, while the novel IVT/L5 epitope and type 2 IVT sequences were essentially never recognized. Figure 1B shows the corresponding data obtained from the same CTL clones when tested on pulse-chased targets. While there was killing of the wild-type, IVT/L2, IVT/R9, and, to some extent, IVT/N9 peptide-loaded targets following a 1-h chase, only the targets bearing the wild-type peptide were killed significantly once the chase had been extended to 3 h; indeed, recognition of the wild-type peptide was still observed for up to 8 h (data not shown). This confirms observations made earlier that even those IVT variants which allowed efficient CTL recognition when present during the assay are much less efficient in pulse-chase assays, reflecting the lower stability of the complexes they form with HLA A11 (16).
FIG. 1.
Recognition of wild-type (wt) and variant IVT peptides in cytotoxicity assays by CTL clones CM c6 and IM52 c1 raised against the wild-type 1 (B95.8-like) IVT epitope sequence. Target cells were an HLA-A11-positive LCL transformed with the prototype 2 Ag876 virus strain loaded with the different IVT peptides as shown. wt1, Caucasian B95.8 prototype 1 epitope sequence; wt2, African Ag876 prototype 2 sequence. (A) Peptide titration assays in which target cells were exposed to the indicated peptide concentrations and the peptides were present throughout the 5-h assay period. (B) Peptide pulse-chase assays in which target cells were preexposed for 1 h at 37°C to the peptides either at the minimum concentration required to give maximal lysis in the assay for which results are shown in panel A or, when that concentration could not be defined, at 10−6 M. The cells were then washed well and incubated at 37°C in peptide-free medium for chase times of up to 8 h before being used as targets in the standard 5-h cytotoxicity assay. Results are expressed as percentages of specific lysis seen at an effector-to-target cell ratio of 1:1. Note that the immediate fall in levels of lysis seen at chase time of 0 h for the L2 and R9 variants compared to wild-type IVT was observed in several such assays.
Figure 2 presents parallel data from experiments with two wild-type AVF-specific CTL clones, again representative of data seen with several clones of this kind. In the continuous presence of peptide, the novel AVF/A2 variant was equivalent to the wild-type sequence in its maximal lysis and titration curve while the AVF/P1, AVF/S1F2, AVF/S1L2, and novel AVF/N4 variant peptides mediated lower levels of maximum lysis, and upon titration, recognition was lost at a 10-fold-higher concentration. By comparison, the AVF/P1L2 was even less antigenic and the type 2 sequence was again not recognized (Fig. 2A). The corresponding pulse-chase assays indicate that the AVF/A2 variant formed A11 complexes which were as stable as the wild-type sequence, still mediating half the maximal lysis after 8 h, whereas lysis of targets preloaded with the other variants fell significantly within a 2-h chase and was undetectable after 8 h (Fig. 2B).
FIG. 2.
Recognition of wild-type (wt) and variant AVF peptides in cytotoxicity assays by CTL clones CM c4 and IM105 c5 raised against the wild-type 1 (B95.8-like) AVF epitope sequence. Target cells were an HLA-A11-positive LCL transformed with the prototype 2 Ag876 virus strain and loaded with the different AVF peptides as shown. wt1, Caucasian B95.8 prototype 1 epitope sequence; wt2, African Ag876 prototype 2 sequence. Peptide titration assays (A) and peptide pulse-chase assays (B) were conducted as described in the legend to Fig. 1. Results are expressed as percentages of specific lysis seen at an effector-to-target cell ratio of 1:1.
Immunogenicity of A11 epitope variants.
The availability of Elispot assays of peptide-induced gamma interferon release now allows the PBMCs of EBV-infected individuals to be screened for epitope-specific T cells in a way which is quicker and more accurate than assays based on the stimulation of limiting-dilution PBMC cultures and the screening of in vitro-expanded clones for cytotoxic activity (27). We therefore used this approach to investigate whether healthy Chinese donors, positive for the HLA A11 allele and carrying an endogenous EBV strain with known IVT and AVF epitope sequences, show evidence of a response to those sequences.
Table 3 (upper section) summarizes the results obtained from 28 healthy EBV-infected Chinese donors who were tested by Elispot assays and who also proved to be HLA-A11-positive by retrospective HLA typing. In the majority of these donors, we were able to identify the IVT and AVF epitope sequences of their resident EBV strains by PCR amplification of viral DNA from PBMC preparations (these sequences contributed to the data already shown in Table 2). All donors were screened for reactivity against the full range of seven IVT peptides (wild type, five variants, and the type 2 sequence) and eight AVF peptides (wild type, six variants, and the type 2 sequence). Included in the same assay was a third epitope sequence, SSC, which elicits a weak A11-restricted response in some individuals. This is derived from the EBV latent membrane protein LMP2, and its sequence is conserved in all Chinese virus strains analyzed to date (14). Table 3 records those instances where a positive Elispot was obtained against any of the screened peptides.
TABLE 3.
A11 epitope responses in Chinese and Caucasian donors carrying EBV strains with defined epitope sequences
| A11-positive donor origin and designation | Epitope sequencea
|
Sequence(s) with positive Elispot reactivity result forb:
|
|||
|---|---|---|---|---|---|
| AVF | IVT | AVF | IVT | SSC | |
| Chinese | |||||
| C27 | wt1c | L2 | wt1, A2, N4 | ||
| C30 | wt1 | L2 | wt1, N4 | ||
| C37 | wt1 | L2 | wt1, A2, N4 | ||
| C39 | wt1 | L2 | wt1, N4 | ||
| C48 | wt1 | L2 | wt1, N4 | ||
| C55 | wt1 | L2 | wt1, N4 | ||
| C56 | wt1 | L2 | wt1, N4, P1 | ||
| C58 | wt1 | L2 | wt1, N4, P1 | SSC | |
| C32 | N4 | N9 | |||
| C35 | N4 | N9 | |||
| C36 | N4 | N9 | SSC | ||
| C206 | N4 | N9 | SSC | ||
| C12 | S1F2 | L2 | |||
| C54 | S1F2 | L2 | SSC | ||
| C59 | S1F2 | L2 | |||
| C218 | S1F2 | L2 | |||
| C13 | S1L2 | T9 | SSC | ||
| C213 | S1L2 | T9 | |||
| C46 | P1L2 | N9 | |||
| C45 | A2 | L5 | SSC | ||
| C210 | P1 | L2 | |||
| C53 | wt2d | wt2 | SSC | ||
| C204 | wt2 | wt2 | |||
| C205 | wt2 | wt2 | |||
| C29 | ND | ND | |||
| C31 | ND | ND | |||
| C43 | ND | ND | SSC | ||
| C50 | ND | ND | |||
| Caucasian | |||||
| Cau 1 | wt1 | wt1 | wt1, A2, N4, P1, S1L2 | wt1, L2, R9, N9 | |
| Cau 2 | wt1 | wt1 | wt1, A2, N4, P1 | wt1, L2, R9, N9 | |
| Cau 3 | wt1 | wt1 | wt1, A2, N4 | wt1, L2, R9, N9 | |
| Cau 4 | wt1 | wt1 | wt1, L2, R9, N9 | ||
| Cau 5 | wt1 | wt1 | wt1, A2, N4, P1, S1L2, S1F2 | wt1, L2, R9, N9, T9 | |
| Cau 6 | wt1 | L2 | wt1, A2, N4, P1, S1L2, S1F2 | ||
| Cau 7 | wt1 | N9 | wt1, A2, N4, P1, S1L2, S1F2 | ||
| Cau 8 | S1L2 | L2 | wt1, A2, N4, P1, S1L2 | ||
| Cau 9 | wt2 | wt2 | |||
| Cau 10 | wt2 | wt2 | |||
Sequence of the AVF and IVT epitopes in the donor's resident EBV strain. Note that the SSC epitope is nonpolymorphic in EBV strains. ND, AVF and IVT sequences could not be amplified from the PBMC sample.
Summary of Elispot assay results showing only those peptides eliciting a response. IVT peptides tested: L2, L5, N9, R9, T9, Caucasian B95.8 prototype 1, and African Ag876 prototype 2. AVF peptides tested: P1, A2, N4, S1F2, P1L2, S1L2, Caucasian B95.8 prototype 1, and African Ag876 prototype 2.
wt1, Caucasian B95.8 prototype 1 epitope sequence.
wt2, African Ag876 prototype 2 epitope sequence.
None of the 28 donors gave a response to the IVT series of peptides. These nonresponders include 13 individuals known to be carrying an IVT/L2 variant virus, 5 individuals carrying an IVT/N9 variant, 2 individuals carrying an IVT/T9 variant, 1 individual carrying an IVT/L5 variant, and 3 individuals carrying a type 2 virus strain. This strongly suggests that, at least, the IVT/L2 and IVT/N9 sequence variants are nonimmunogenic in vivo. The absence of a response to the IVT/T9, IVT/L5, and type 2 epitopes is again consistent with their nonimmunogenicity, but the numbers of individuals tested were small. However, it is interesting that one individual in each of these groups (C13, C45, and C53) lacked IVT reactivity but did make a detectable response to the subdominant SSC epitope.
Turning to assays with the AVF series of peptides, all eight HLA-A11-positive donors who carried a virus with a wild-type AVF sequence showed evidence of reactivity against AVF and also against the AVF/A2, AVF/N4, and (in two cases) AVF/P1 peptides. By contrast, four donors carrying the AVF/N4 sequence and four donors carrying the AVF/S1F2 sequence showed no detectable reactivity against their resident epitopes or against wild-type AVF. Furthermore, we detected no response in smaller numbers of individuals carrying AVF/P1L2, AVF/S1L2, AVF/A2, or type 2 AVF sequences. Again, several of these nonresponders to AVF peptides did show evidence of a response to the SSC epitope. Note that these assays also included another 17 Chinese donors (with a similar range and distribution of IVT and AVF variants to the above) who proved to be HLA A11 negative by retrospective HLA typing; none of these gave any reactivity to any of the IVT series, the AVF series, or to the SSC peptide (data not shown).
Table 4 (upper section) presents the detailed counts from the Elispot assays described above, expressed as spot-forming cells per 106 PBMCs. These data derive from the 14 of 28 HLA-A11-positive Chinese donors whose test results are shown in Table 3, i.e., those whose resident EBV strain had been sequenced across the IVT and AVF epitope regions and who gave a response to one or more of the epitope peptides. The counts show that individuals infected with a wild-type AVF strain (C27 to C58) possess memory CTLs which, under the conditions of the Elispot assay, often recognize the AVF/A2 and AVF/N4 variants just as well as the wild-type AVF sequence; cross-recognition of the AVF/P1 variant was weaker and only detectable when absolute responses to the wild-type peptide were unusually high. Where SSC reactivity was observed, this tended to involve individuals who lacked a detectable response to AVF and the levels of SSC response were generally low.
TABLE 4.
Elispot quantitation of epitope responses in HLA-A11-positive Chinese and Caucasian donors carrying EBV strains with defined A11 epitope sequencesa
| Donor origin and designation | Response to AVF peptide sequence:
|
Response to IVT peptide sequence:
|
Response to SSC | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| wt1b | A2 | N4 | P1 | S1L2 | S1L2 | P1L2 | wt2c | wt1 | L2 | R9 | N9 | T9 | L5 | wt2 | ||
| Chinese | ||||||||||||||||
| C27 | 212 | 175 | 120 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| C30 | 120 | 166 | 128 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| C37 | 90 | 92 | 78 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| C39 | 80 | 100 | 74 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| C48 | 135 | 163 | 105 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| C55 | 121 | 142 | 95 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| C56 | 490 | 517 | 480 | 55 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| C58 | 400 | 427 | 400 | 75 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 35 |
| C45 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 50 |
| C36 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 80 |
| C206 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 50 |
| C13 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 50 |
| C54 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 242 |
| C53 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 60 |
| Caucasian | ||||||||||||||||
| Cau 1 | 52 | 52 | 30 | 10 | 14 | 0 | 0 | 0 | 744 | 752 | 800 | 14 | 0 | 0 | 0 | 0 |
| Cau 2 | 166 | 170 | 150 | 15 | 0 | 0 | 0 | 0 | 600 | 630 | 640 | 10 | 0 | 0 | 0 | 0 |
| Cau 3 | 32 | 30 | 24 | 0 | 0 | 0 | 0 | 0 | 170 | 200 | 185 | 75 | 0 | 0 | 0 | 0 |
| Cau 4 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 148 | 160 | 166 | 8 | 0 | 0 | 0 | 0 |
| Cau 5 | 195 | 195 | 175 | 50 | 115 | 80 | 0 | 0 | 455 | 375 | 495 | 375 | 250 | 0 | 0 | 0 |
| Cau 6 | 288 | 386 | 190 | 72 | 62 | 55 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Cau 7 | 100 | 90 | 95 | 60 | 50 | 30 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Cau 8 | 300 | 240 | 218 | 128 | 130 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
Results are expressed as spot-forming cells per 106 PBMCs in Elispot assays of gamma interferon release involving the full range of AVF and IVT peptide sequences and the SSC peptide. The AVF and IVT sequences present in the donor's resident EBV strain are identified in each case by boldface type; note that the SSC epitope is nonpolymorphic in Chinese and Caucasian EBV strains.
wt1, Caucasian B95.8 prototype 1 epitope sequence.
wt2, African Ag876 prototype 2 epitope sequence.
For comparison, Tables 3 and 4 (lower sections) show corresponding data from HLA-A11-positive Caucasian donors; 10 such individuals, all EBV antibody positive, were screened for A11 epitope responses and their resident EBV strains were sequenced across the IVT and AVF epitopes. Eight donors (Cau 1 to 8) carried a type 1 virus strain, and all eight donors showed an IVT and/or AVF response, whereas two (Cau 9 and 10, shown in Table 3 only) carried a type 2 strain and failed to make a response to either epitope. This latter result confirms the nonimmunogenicity of the type 2 epitope sequences. Of the eight type 1 virus-infected Caucasians, five (Cau 1 to 5) carried a wild-type IVT sequence and gave a response which, in the Elispot assay, was cross-reactive against the IVT/L2, IVT/R9, and, to some extent, IVT/N9 variants. However, the other three individuals (Cau 6 to 8) carried IVT/L2 or IVT/N9 sequences and never mounted a detectable response, confirming these variants as nonimmunogenic. Seven of the type 1 virus-infected Caucasians (Cau 1 to 7) carried a wild-type AVF sequence, and all but one mounted an AVF-specific response that showed either a similar pattern of cross-reactivity against AVF variants as seen with Chinese donors or a broader pattern that included AVF/P1, AVF/S1L2, and AVF/S1F2. Only one HLA-A11-positive donor (Cau 8) was identified carrying an EBV strain with the typically Caucasian AVF/S1L2 variant sequence. Interestingly, this donor did mount an epitope-specific response that in the Elispot assay showed stronger reactivity against the wild-type AVF peptide than against AVF/S1L2 itself. It is conceivable that this donor was coinfected with a wild-type AVF-coding virus strain, but the presence of such a virus could never be detected by PCR amplification from this donor's throat washings or PBMCs (data not shown).
DISCUSSION
The present work set out to provide a more coherent body of evidence regarding A11 epitope polymorphism among EBV strains and the immunological implications of such polymorphism. Using a much larger panel of viruses than that studied earlier (7), this work first confirms that, among type 1 EBV strains in the Chinese population, where 55% of individuals are HLA-A11-positive, the IVT epitope is almost always different from the B95.8 prototype sequence and the AVF epitope is different in around 75% of cases. We identified four patterns of IVT sequence variation and six patterns of AVF variation among the 64 type 1 viruses analyzed. This level of variety speaks against a significant influence of founder effects in determining contemporary levels of IVT and AVF polymorphism among Chinese strains. It is interesting to contrast this with the situation in Papua New Guinea, where a single combination of epitope changes, IVT/T9 and AVF/S1L2, has been seen in 26 of 27 type 1 virus strains drawn both from highly HLA-A11-positive lowland populations and from highland populations, where HLA A11 incidence is very low (3, 6, 7). Founder effects may have been more important in this context, particularly since both lowland and highland Papua New Guinea strains also share a number of markers at other EBNA3A, 3B, 3C loci that distance them as a group from Chinese, African, and Caucasian isolates (11).
We also sought to determine the IVT and AVF epitope status of EBV strains in the sub-Saharan African population (which completely lack the HLA A11 allele) and in the Caucasian population (where around 11% individuals are HLA A11 positive), since published data were very limited on this point (3, 7). We found that all African type 1 viruses followed the prototype B95.8 sequence at both epitopes, as did the majority of Caucasian type 1 strains. However, this wild-type pattern is not as dominant in Caucasians as first suggested (7). At least four IVT variants exist among Caucasian strains, and most of these are found in linkage with a single AVF variant; thus, there is a degree of epitope variation even in a population where A11 allele frequency is relatively low. Almost all of the Caucasian variant sequences have been observed in Chinese EBV strains, but the combinations of IVT and AVF sequences seen in Caucasian strains are often unique (Table 2). The existence of such sequence variants was already known (3), but the frequency of such variants in the Caucasian population is not as high as suggested in that earlier report. In both Chinese and Caucasian type 1 viruses, all nucleotide changes in the epitope regions alter the amino acid sequence, very frequently at key anchor residues for HLA A11 binding. By contrast, Chinese, African, and Caucasian type 2 viruses all appear to retain the prototype Ag876 sequence at both the IVT and AVF loci. This suggests that, if immune pressure is selecting for epitope diversification in type 1 viruses, then the corresponding type 2 sequences do not elicit an equivalent pressure.
Our next objective was to examine the antigenicity of the type 1 IVT and AVF variants in CTL detection assays with T-cell clones raised from Caucasian donors against the B95.8 strain wild-type epitopes. In the first type of assay, wild-type and variant peptides were tested across a wide range of concentrations for their ability to mediate target cell lysis, with the peptides being present throughout the assay period. This revealed efficiencies of cross recognition ranging from zero (for the type 2 IVT and AVF peptides and for the novel IVT/L5 variant) through intermediate (for several of the IVT and AVF variants including the novel AVF/N4 sequence) to equality with the wild-type peptide (for the IVT/L2, IVT/R9, and novel AVF/A2 variants). Others have already studied the recognition of IVT variants in peptide titration assays with broadly similar results (3, 16); the differences in detail probably reflect the different fine specificities of the individual IVT-specific CTL clones used (5). It is apparent, however, that assays conducted in the continued presence of peptide can reveal cross-recognition even when the variant peptides have low inherent affinity for the HLA restricting allele, provided that the main T-cell receptor contact residues are conserved (2, 16). We therefore turned to a second assay involving peptide pulsing of targets followed by increasingly longer chase times in peptide-free medium. This allows one to follow the stability of antigenic complexes over time, a potentially important parameter given the correlation often reported between stability of an MHC-epitope complex and epitope strength (1, 17). These experiments showed that recognition of all of the IVT variants (including IVT/L2 and IVT/R9) and of all but one of the AVF variants (the exception being AVF/A2) was lost more rapidly than that of the wild-type peptide. These results extend the earlier work of Levitsky et al. (16) with IVT variants in pulse-chase assays and show for the first time that most AVF variants follow a similar trend.
It is important, however, that nonrecognition in vitro by CTLs primed against wild-type IVT and AVF epitopes does not necessarily mean that these variant epitopes are nonimmunogenic in vivo when expressed in the context of a natural virus infection. For instance, the novel IVT/L5 peptide retains the key anchor residues at positions 2 and 9 for efficient HLA A11 binding (29) and its nonrecognition by wild-type IVT effectors may simply reflect the importance of position 5 as a T-cell receptor contact residue. A similar argument can also be advanced for the novel AVF/N4 variant as well as for the type 2 versions of both IVT and AVF, neither of which were recognized at all by type 1 epitope-specific effectors in the previous assays. The issue of in vivo immunogenicity has barely been studied in earlier work. The only evidence comes from experiments in which PBMCs from three Chinese donors (each carrying a defined combination of IVT and AVF epitopes) were stimulated in vitro with autologous LCL cells carrying the endogenous EBV strain; no epitope-specific responses were detected in the resultant in vitro-expanded T-cell lines (7). However, this is now known to be a much less sensitive screen for EBV antigen- or epitope-specific reactivities than are Elispot assays of peptide-induced gamma interferon release (27).
We therefore studied Chinese donors in Elispot assays with the full panel of wild-type and variant peptides and used PCR amplification on aliquots of the same PBMCs to determine both the donor's HLA type and the IVT and AVF sequences of the resident EBV strain. The results provide the first clear indication that the variant IVT and AVF epitope sequences identified in Chinese virus strains are indeed nonimmunogenic in vivo. Thus, the most common IVT variant, IVT/L2, which was very well recognized by wild-type epitope-specific effectors in peptide sensitization assays, was encoded by the resident EBV strain in 13 HLA-A11-positive Chinese donors studied in Elispot assays, yet none of these donors had detectable IVT/L2-specific memory. In addition, five HLA-A11-positive individuals infected by EBV strains encoding another common variant, IVT/N9, and three individuals carrying type 2 EBV also showed no evidence of a response. By contrast, all five Caucasian donors carrying an EBV strain with a wild-type IVT sequence displayed strong epitope-specific memory. Interestingly, these cells showed cross-reactivity against the IVT/L2, IVT/R9, and IVT/N9 variant peptides under the high peptide concentration conditions employed in the Elispot assay. Once again, however, Caucasians carrying such variant type 1 or wild-type 2 IVT sequences did not mount a detectable response, supporting the evidence from Chinese donors that these sequences are nonimmunogenic.
Turning to the immunogenicity of AVF variants, none of the 16 Chinese donors tested carrying variant sequences at this locus gave evidence of an epitope-specific response. These included four donors with the novel AVF/N4 sequence and four with the AVF/S1F2 sequence (both peptides that were in the intermediate range of recognition by wild-type AVF-specific CTLs in vitro) as well as three donors with the type 2 sequence. This again strongly suggests that most if not all AVF variants found in Chinese populations are nonimmunogenic, although more donors need to be studied before firm conclusions can be drawn about some of the less common variants. This is particularly the case for the AVF/S1L2 variant, which is rare in Chinese donors but relatively common in Caucasians and where the one relevant Caucasian donor tested (Cau 8) did mount a response, whereas the two available Chinese donors did not. It should be stressed, however, that a significant minority of Chinese donors carry a wild-type AVF sequence, and all eight such individuals tested gave clear evidence of a response, as did six of seven Caucasians in this category. As seen with the IVT response, memory CTLs to the wild-type AVF epitope cross-reacted with some of the peptide variants under the conditions of the Elispot assays, notably with AVF/A2 and AVF/N4 and, in some cases, with the three other variants (P1, S1L2, and S1F2) that had behaved like AVF/N4 in CTL detection assays.
Viewing HLA A11 epitope responses as a whole, it is interesting that in the 10 individuals (8 Chinese and 2 Caucasian) carrying a wild-type AVF but variant IVT sequence, AVF responses tended to be stronger than those seen in the 5 individuals (all Caucasian) carrying a wild-type sequence at both epitopes (means of 204 and 89 spot-forming cells/106 PBMCs, respectively). This suggests that the AVF epitope may elicit slightly stronger responses in the absence of competition from IVT, though never reaching the higher levels of response typically induced by wild-type IVT itself (mean, 423 spot-forming cells/106 PBMCs). Similarly, all examples of a detectable SSC response came from Chinese rather than Caucasian donors (despite the fact that this epitope is conserved in Caucasian EBV strains), and all but one of these SSC-responders carried a virus strain that was variant at both AVF and IVT, implying that immunogenicity of the SSC sequence is increased in the absence of competition from the other A11 epitopes. Importantly, however, even in the absence of such competition, SSC responses are almost always low. We infer that SSC is an inherently weak immunogen and, as such, would not be subject to the intensity of pressure for change in HLA-A11-positive populations as would IVT and AVF.
In summary, we have tested a representative panel of HLA-A11-positive Chinese donors carrying defined IVT and AVF variant sequences and found no evidence that any of the four IVT variants or the six AVF variants identified can elicit a detectable response in these individuals. We also provide the first indication, albeit from a small number of donors, that the type 2 versions of IVT and AVF are not immunogenic in vivo. Though far from constituting proof, these findings are consistent with the view that immune pressure from the HLA-A11-restricted CTL response has selected for diversification of the type 1 IVT and AVF epitope sequences among Chinese virus strains, whereas type 2 strains (which all carry conserved prototype sequences) may not have been subject to the same pressure. Such results encouraged us to look further at the possible biological significance of A11 epitope changes with a complementary approach. This is described in the accompanying paper (20), in which the A11 epitope changes are viewed in the broader context of sequence diversification among Chinese virus strains across the EBNA2, 3A, 3B, and 3C genes.
Acknowledgments
This work was supported by Cancer Research United Kingdom and by the award to R.S.M. of a Medical Research Council Clinical Research Fellowship.
REFERENCES
- 1.Brooks, J. M., R. A. Colbert, J. P. Mear, A. Leese, and A. B. Rickinson. 1998. HLA-B27 subtype polymorphism and CTL epitope choice: studies with EBV peptides link immunogenicity with stability of the B27:peptide complex. J. Immunol. 161:5252-5259. [PubMed] [Google Scholar]
- 2.Brooks, J. M., D. S. G. Croom-Carter, A. M. Leese, R. J. Tierney, G. Habeshaw, and A. B. Rickinson. 2000. Cytotoxic T lymphocyte responses to a polymorphic Epstein-Barr virus epitope identify healthy carriers with coresident viral strains. J. Virol. 74:1801-1809. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Burrows, J. M., S. R. Burrows, L. M. Poulsen, T. B. Sculley, D. J. Moss, and R. Khanna. 1996. Unusually high frequency of Epstein-Barr virus genetic variants in Papua New Guinea that can escape cytotoxic T-cell recognition: implications for virus evolution. J. Virol. 70:2490-2496. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Callan, M. F. C., L. Tan, N. Annels, G. S. Ogg, J. D. K. Wilson, C. A. O'Callaghan, N. Steven, A. J. McMichael, and A. B. Rickinson. 1998. Direct visualization of antigen-specific CD8(+) T cells during the primary immune response to Epstein-Barr virus in vivo. J. Exp. Med. 187:1395-1402. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.de Campos Lima, P.-O., V. Levitsky, M. P. Imreh, R. Gavioli, and M. G. Masucci. 1997. Epitope-dependent selection of highly restricted or diverse T-cell receptor repertoires in response to persistent infection by Epstein-Barr virus. J. Exp. Med. 186:83-89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.de Campos-Lima, P.-O., R. Gavioli, Q. J. Zhang, L. E. Wallace, R. Dolcetti, M. Rowe, A. B. Rickinson, and M. G. Masucci. 1993. HLA-A11 epitope loss isolates of Epstein-Barr virus from a highly A11+ population. Science 260:98-100. [DOI] [PubMed] [Google Scholar]
- 7.de Campos-Lima, P.-O., V. Levitsky, J. Brooks, S. P. Lee, L. F. Hu, A. B. Rickinson, and M. G. Masucci. 1994. T cell responses and virus evolution: loss of HLA A11-restricted CTL epitopes in Epstein-Barr virus isolates from high A11-positive populations by selective mutation of anchor residues. J. Exp. Med. 179:1297-1305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Gavioli, R., M. G. Kurilla, P. O. de Campos-Lima, L. E. Wallace, R. Dolcetti, R. J. Murray, A. B. Rickinson, and M. G. Masucci. 1993. Multiple HLA-A11-restrictec cytotoxic T-lymphocyte epitopes of different immunogenicities in the Epstein-Barr virus-encoded nuclear antigen 4. J. Virol. 67:1572-1578. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Goulder, P. J., C. Brander, Y. Tang, C. Tremblay, R. A. Colbert, M. M. Addo, E. S. Rosenberg, T. Nguyen, R. Allen, A. Trocha, M. Altfeld, S. He, M. Bunce, R. Funkhouser, S. I. Pelton, S. K. Burchett, K. McIntosh, B. T. Korber, and B. D. Walker. 2001. Evolution and transmission of stable CTL escape mutations in HIV infection. Nature 412:334-338. [DOI] [PubMed] [Google Scholar]
- 10.Khanna, R. 1998. Geographic (not immunological) constraints define long-term evolutionary dynamics of Epstein-Barr virus. EBV Rep. 5:127-131. [Google Scholar]
- 11.Khanna, R., R. W. Slade, L. M. Poulsen, D. J. Moss, S. R. Burrows, J. Nicholls, and J. M. Burrows. 1997. Evolutionary dynamics of genetic variation in Epstein-Barr virus isolates of diverse geographical origins: evidence for immune pressure-independent genetic drift. J. Virol. 71:8340-8346. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lalvani, A., R. Brookes, S. Hambleton, W. J. Britton, A. V. Hill, and A. J. McMichael. 1997. Rapid effector function in CD8+ memory T cells. J. Exp. Med. 186:859-865. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lee, S. P., S. Morgan, J. Skinner, W. A. Thomas, S. R. Jones, J. Sutton, R. Khanna, H. C. Whittle, and A. B. Rickinson. 1995. Epstein-Barr virus isolates with the major HLA B35.01-restricted cytotoxic T-lymphocyte epitope are prevalent in a highly B35.01-positive African population. Eur. J. Immunol. 25:102-110. [DOI] [PubMed] [Google Scholar]
- 14.Lee, S. P., R. J. Tierney, W. A. Thomas, J. M. Brooks, and A. B. Rickinson. 1997. Conserved cytotoxic T lymphocyte (CTL) epitopes within Epstein-Barr virus (EBV) latent membrane protein 2, a potential target for CTL-based tumour therapy. J. Immunol. 158:3325-3334. [PubMed] [Google Scholar]
- 15.Lee, T. D., T. M. Zhao, R. Mickey, Y. P. Sun, G. Lee, C. X. Song, D. Z. Cheng, S. Zhou, S. Q. Ding, D. X. Cheng, et al. 1988. The polymorphism of HLA antigens in the Chinese. Tissue Antigens 32:188-208. [DOI] [PubMed] [Google Scholar]
- 16.Levitsky, V., Q. J. Zhang, J. Levitskaya, M. G. Kurilla, and M. G. Masucci. 1997. Natural variants of the immunodominant HLA A11-restricted CTL epitope of the EBV nuclear antigen-4 are nonimmunogenic due to intracellular dissociation from MHC class I:peptide complexes. J. Immunol. 159:5383-5390. [PubMed] [Google Scholar]
- 17.Levitsky, V., Q.-J. Zhang, J. Levitskaya, and M. G. Masucci. 1996. The life span of major histocompatibility complex-peptide complexes influences the efficiency of presentation and immunogenicity of two class I-restricted cytotoxic T lymphocyte epitopes in the Epstein-Barr virus nuclear antigen 4. J. Exp. Med. 183:915-926. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.McGeoch, D. J., and A. J. Davison. 1999. The molecular evolutionary history of the herpesviruses, p. 441-465. In E. Domingo, R. Webster, and J. Holland (ed.), Origins and evolution of viruses. Academic Press, London, United Kingdom.
- 19.McGeoch, D. J., A. Dolan, and A. C. Ralph. 2000. Toward a comprehensive phylogeny for mammalian and avian herpesviruses. J. Virol. 74:10401-10406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Midgley, R. S., A. I. Bell, D. J. McGeoch, and A. B. Rickinson. 2003. Latent gene sequencing reveals familial relationships among Chinese Epstein-Barr virus strains and evidence for positive selection of A11 epitope changes. J. Virol. 77:11517-11530. [DOI] [PMC free article] [PubMed]
- 21.Midgley, R. S., N. M. Blake, Q.-Y. Yao, D. S. G. Croom-Carter, S.-T. Cheung, S.-F. Leung, A. T. C. Chan, P. J. Johnson, D. Huang, A. B. Rickinson, and S. P. Lee. 2000. Novel intertypic recombinants of Epstein-Barr virus in the Chinese population. J. Virol. 74:1544-1548. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Phillips, R. E., S. Rowland-Jones, D. F. Nixon, F. M. Gotch, J. P. Edwards, A. O. Ogunlesi, J. G. Elvin, J. A. Rothbard, C. R. M. Bangham, C. R. Rizza, and A. J. McMichael. 1991. Human immunodeficiency virus genetic variation that can escape cytotoxic T-cell recognition. Nature 354:453-459. [DOI] [PubMed] [Google Scholar]
- 23.Rickinson, A. B., and E. Kieff. 2001. Epstein-Barr virus, p. 2575-2627. In D. M. Knipe and P. M. Howley (ed.), Fields virology. Lippincott Williams and Wilkins, Philadelphia, Pa.
- 24.Sample, J., L. Young, B. Martin, T. Chatman, E. Kieff, and A. Rickinson. 1990. Epstein-Barr virus types 1 and 2 differ in their EBNA-3A, EBNA-3B, and EBNA-3C genes. J. Virol. 64:4084-4092. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Shastri, N., S. Schwab, and T. Serwold. 2002. Producing nature's gene-chips: the generation of peptides for display by MHC class I molecules. Annu. Rev. Immunol. 20:463-493. [DOI] [PubMed] [Google Scholar]
- 26.Steven, N. M., A. M. Leese, N. Annels, S. Lee, and A. B. Rickinson. 1996. Epitope focusing in the primary cytotoxic T-cell response to Epstein-Barr virus and its relationship to T-cell memory. J. Exp. Med. 184:1801-1813. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Tan, L. C., N. Gudgeon, N. E. Annels, P. Hansasuta, C. A. O'Callaghan, S. Rowland-Jones, A. J. McMichael, A. B. Rickinson, and M. F. C. Callan. 1999. A re-evaluation of the frequency of CD8(+) T cells specific for EBV in healthy virus carriers. J. Immunol. 162:1827-1835. [PubMed] [Google Scholar]
- 28.Yewdell, J. W., and J. R. Bennink. 1999. Immunodominance in major histocompatibility complex class I-restricted T lymphocyte responses. Annu. Rev. Immunol. 17:51-88. [DOI] [PubMed] [Google Scholar]
- 29.Zhang, Q. J., R. Gavioli, G. Klein, and M. G. Masucci. 1993. An HLA-A11-specific motif in nonamer peptides derived from viral and cellular proteins. Proc. Natl. Acad. Sci. USA 90:2217-2221. [DOI] [PMC free article] [PubMed] [Google Scholar]