| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Paediatrics, Nuffield Department of Medicine, Peter Medawar Building for Pathogen Research, Oxford, United Kingdom;
Human Immunology Section, Vaccine Research Center, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892;
HIV Pathogenesis Programme, Doris Duke Medical Research Institute, University of Natal, Durban, South Africa;
Partners AIDS Research Center, Massachusetts General Hospital, Boston, MA 02129;
¶ Department of Genitourinary Medicine, High Wycombe General Hospital, High Wycombe, United Kingdom;
|| The Harrison Clinic, Radcliffe Infirmary Hospital, Oxford, United Kingdom;
# Department of Microbiology, University of Washington School of Medicine, Seattle, WA 98195;
** Imperial College School of Medicine, St. Mary’s Hospital, London, United Kingdom; and
Great Ormond Street Hospital for Sick Children, London, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
The HLA region is believed to be the most polymorphic in the human genome, with 1403 different HLA class I alleles described to date, encoding 1122 different proteins (
www.ebi.ac.uk/imgt/hla/docs/release.html
). This polymorphism is focused primarily on the HLA residues that form the peptide-binding groove and thus that define the peptide-binding motif for a given molecule (8, 9, 10, 11, 12). However, there is a degree of degeneracy in HLA peptide binding, whereby multiple distinct class I alleles can bind similar or even identical peptides. Alleles sharing similar peptide-binding motifs have therefore been grouped into so-called HLA supertypes (8), of which 10 have been described, encompassing the vast majority of known HLA alleles (13, 14, 15). The compression of most of the diversity of the class I loci into just 10 supertypes has suggested that an epitope-based vaccine approach to counter a variety of pathogens, including HIV-1, hepatitis B, hepatitis C, and malaria, is possible (15). A vaccine containing a small number of highly promiscuous peptide epitopes or "supertopes" (15) from just the six most frequent supertypes is estimated to provide an average population coverage of >99%, irrespective of ethnicity (8). However, although naturally occurring supertopes exist, and have been well-characterized in terms of binding and/or cross recognition in vitro (5, 16, 17, 18, 19, 20), little is known of the functional consequences in vivo of presentation of an epitope on one class I molecule as opposed to another.
The studies described here address the role of the B7 supertype in the immune control of C-clade HIV-1. We focus on this supertype as HLA-B alleles play the dominant role in the immune control of intracellular pathogens including HIV (21, 22), and B7 is the most prevalent HLA-B supertype (8). This study was conducted on a large cohort of HIV C-clade-infected individuals of Zulu/Xhosa ethnic origin, from KwaZulu Natal (KZN), South Africa. South Africa is affected by more HIV-1 infections than any other country, and the worst affected province is KZN, where antenatal seroprevalence rates have reached 41% (http://avert.org/safricastats.htm).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Characterization of B7 supertype HIV-specific T cell responses was undertaken in a cohort of 515 study subjects of Zulu/Xhosa ethnic origin, from KZN, South Africa, chronically infected with C-clade HIV-1, as previously described (21).
Detection of B7 supertype responses
Subjects were screened for HIV-specific CTL responses by IFN-
ELISPOT assay, using 410 18-mer peptides spanning the entire expressed HIV genome, based on the 2001 C-clade consensus (www.hiv.lanl.gov), as previously described (21). Associations between recognition of individual peptides and expression of HLA-B*0702, B*4201, and/or B*8101 were sought using the Fisher’s exact test, as described (21). All statistical comparisons were subjected to Bonferroni corrections to eliminate the problem of multiple comparisons.
Viral sequencing
Sequence data were obtained from population sequencing of either proviral DNA or viral RNA. For the provirus, genomic DNA was extracted from whole blood using the Puregene DNA isolation kit (Gentra), and was amplified by nested PCR, using Gag, RT, Vpr, and Nef primers as described (23, 24). All sequencing was conducted using the BigDye ready reaction termination mix V3 (Applied Biosystems), using additional Gag-, RT-, Int-, and Nef-specific primers as described (23, 24). Isolation and sequencing of viral RNA was conducted as described (25). Total HIV-1 viruses sequenced were as follows: Gag: 542 (104 B*4201, 63 B*8101, 47 B*0702, 16 B*3910, and 312 other); RT: 160 (36 B*4201, 12 B*8101, 18 B*0702, and 94 other); Vpr: 161 (35 B*4201, 12 B*8101, 18 B*0702, and 96 other); Nef: 215 (41 B*4201, 20 B*8101, 21 B*0702, and 133 other). Neighbor-joining trees of all the sequences together with reference sequences from the Los Alamos Database (www.hiv.lanl.gov) were constructed to verify the viral clades and to eliminate the possibility of contamination with laboratory HIV strains. All residue numbers are taken against the HXB2 reference sequence.
Tetramer synthesis and cell sorting
HLA-B*4201 and B*8101 monomer plasmids where synthesized from B*0702 template cells by site directed mutagenesis (Stratagene) as per manufacturer’s instructions. B*4201 and B*8101 TL9 tetramers were then synthesized as previously described (26). These tetramers were used to sort TL9-specific CTL viably from cryopreserved PBMC of B*4201- or B*8101-expressing individuals with detectable TL9-specific responses, at 70 pounds per square inch using a modified FACSAria (BD Biosciences). Postsort purity was consistently >99%.
Clonotype analysis
mRNA was extracted from at least 5000 viably sorted cells using the Oligotex kit (Qiagen), and subjected to a non-nested, template switch-anchored RT-PCR using a 3' TCRB C region primer (5'-GCTTCTGATGGCTCAAACACAGCGACCTC-3') as described previously (27).Amplified products were ligated into pGEM-T Easy vector (Promega), cloned by transformation of competent DH5 Escherichia coli, and sequenced as described (28). A minimum of 55 clones was generated and analyzed per sample, disregarding pseudogenes and "nonfunctional" sequences that could not be resolved. Nucleotide comparisons were used to establish clonal identity. Data analysis was performed using Sequencher Version 4.2 (Gene Codes).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
This study focuses primarily on the three most common B7 supertype HLA alleles in the Zulu-Xhosa study population, HLA*B4201, B*8101, and B*0702, which occur at phenotypic frequencies of 18.6, 9.6, and 8.2%, respectively. These are closely related alleles, with no amino acid differences in the residues forming their B and F primary binding pockets (www.ebi.ac.uk/imgt/hla/). To determine the CTL epitopes restricted by these B7 supertype alleles, 515 subjects with chronic C-clade HIV-1 infection were analyzed for IFN- responses to a panel of 410 overlapping 18-mer peptides, spanning the entire expressed HIV genome, as previously described (21). Using this method, 14 peptides were identified that are recognized by CTL restricted by at least one of the B7 supertype alleles (Table I). Eleven of the 14 peptides contain optimally defined epitopes restricted by one or more of these alleles (21). These epitopes have strong preference for Pro at position 2, and a hydrophobic aliphatic residue (L, M, I, and A) at the C-terminal position (Table I). No strong preferences at the other peptide-binding pockets (A, D, C, and E, binding peptide position 1, 3, 6, and 7, respectively) have been reported (10), or were evident from the epitopes targeted in this study. Two epitopes, RM9-Nef and FL9-Vpr, were completely promiscuous, with CTL responses restricted by all three alleles detected (Table I). An additional four epitopes, TL9-Gag, SM9-RT, TL-10-Nef, and GL9-Gag, are recognized by CTL restricted by two of three of the B7 supertype alleles (Table I).
|
To investigate the comparative ability of CTL responses directed at identical epitopes, but restricted by different HLA alleles within the B7 supertype, to exert selection pressure on the virus, we next sought to identify HLA-associated polymorphisms within the epitopes (23, 24, 29, 30).
For each of the six promiscuous B7 supertype epitopes, we observed significant differences in the selection pressure exerted by CTL of differing HLA restriction. The two most striking examples involved the epitopes TL9-Gag and RM9-Nef (Table II). TL9-Gag is recognized by both HLA-B*4201- and B*8101-restricted CTL (Table I, Fig. 1). Additionally, this epitope is the dominant response in individuals expressing HLA-B*3910 (data not shown), a further B7 supertype allele (12), but less prevalent in the Durban cohort (phenotypic frequency 2.5%). There is significantly greater variation within TL9-Gag in individuals expressing each of these alleles, B*4201 (22 of 104; p = 0.007, Fisher’s exact test), B*8101 (39 of 62; p = 1.7 x 10–17) and B*3910 (6 of 16; p = 0.0009), compared with those lacking these alleles (32 of 312) (Table II, Fig. 1). No association between B*0702 and variation within the epitope was observed. This is not unexpected because, although TL9 is known to be an HLA-B*0702-restricted epitope (www.hiv.lanl.gov), only a very low frequency (5%) of B*0702 responders to TL9 were detected in this cohort. However, despite TL9 being the dominant response for B*4201, B*8101, and B*3910, there are clear differences in the frequency of polymorphism associated with each allele. In particular, polymorphisms arising in association with HLA-B*8101 were significantly more pronounced than with B*4201 (p = 1.7 x 10–7), and tended to be so with B*3910 (p = 0.06).
|
|
Sequence data from the remaining four promiscuous epitopes are summarized in Fig. 1. In each instance, variation within the epitope is only associated with CTL restricted by one of the HLA alleles. In the case of FL9-Vpr, which, like RM9-Nef, is targeted with equal frequency through all three alleles, only HLA-B*0702 is significantly associated with variation within the epitope (18 of 18 compared with 65 of 96 of individuals lacking B*0702, B*4201, or B*8101; p = 0.002). For SM9-RT, equally targeted by B*0702- and B*8101-positive individuals, and TL10-Nef, equally targeted by B*0702- and B*4201-positive individuals, only B*0702 is associated with polymorphism. Finally, for the B*0702/4201-targeted epitope GL9-Gag, there is a strong "negative association" between expression of B*0702 and conservation of the apparent consensus sequence (43 of 47 B*0702s vs 242 of 504; p = 2.12 x 10–9). We hypothesize that, as has been demonstrated previously, this association arose through positive selection pressure for a stable escape mutation, which spread at the population level until it replaced the consensus amino acid (24). However, irrespective of the mechanism, there are no associations detected between B*4201 and GL9-Gag mutation, again suggesting a disparity between CTL targeting the same epitopes but restricted by different HLA alleles.
In each of the promiscuous epitopes it presents, HLA-B*4201 is the allele least associated with escape mutation. However, it is important to note that this does not indicate a functional defect in B*4201-restricted CTL, as there are strong associations between B*4201 and variation within two of the additional B*4201-restricted epitopes identified in this study, LI9-Int and YL9-RT (p = 1.0 x 10–5 and p = 1.3 x 10–5, respectively; data not shown).
HLA-associated polymorphisms within epitopes are escape mutations and can be HLA specific
The existence of strong associations between HLA allele expression and sequence polymorphism is highly suggestive of escape mutation (23, 24, 29). To confirm this, we next addressed the question of whether the HLA-linked sequence polymorphisms observed affect CTL recognition. Initially focusing on TL9-Gag, the commonly occurring variants (Table II) were tested for recognition by both B*8101- and B*4201-restricted CTL in IFN- ELISpot assays (Fig. 2, A–D, show representative examples from 41 B*4201 and B*8101 individuals tested). As anticipated, where the mutations occur at nonanchor residues (31), the pattern of recognition varied between individuals. However, in each individual some of the variants are recognized very poorly or not at all, demonstrating that they can interfere with TCR recognition. Of note, there was no difference between the patterns of variant recognition by B*8101-TL9 CTL and B*4201-TL9 CTL that would explain the higher level of variation observed in association with B*8101. Focusing on the Nef-RM9 epitope, again, peptide titration of the common variants confirms that these mutations affect CTL recognition, irrespective of the HLA restriction element (Fig. 2, E–H). Thus, data from these two promiscuous B7 supertype epitopes tested support the hypothesis that the HLA-associated variants observed within the B7 supertype epitopes can affect CTL recognition and are escape mutations. In both cases, the lower frequency of epitope variants in B*4201-positive subjects is not explained by the ability of B*4201-TL9-Gag CTL to recognize all the commonly occurring variants.
|
Taken together, in showing a difference in the ability of CTL restricted by different HLA to select for escape mutation, and the occurrence of characteristic HLA-specific escape mutations, these data provide evidence of in vivo qualitative differences between CTL specificities directed toward identical promiscuous epitopes, but presented by different HLA class I alleles.
Selection pressure is not determined solely by the frequency, magnitude, or immunodominance of CTL responses
A potential explanation for the inconsistency with which HLA alleles are associated with variation within their target epitopes would be that the CTL responses restricted by different HLA do not occur at equal frequency for a given epitope. Clearly, if individuals expressing a particular HLA respond to a particular epitope only rarely, then there is unlikely to be a strong association between that allele and epitope variation. However, in the cases described there are no significant differences in the frequency of responders restricted by different HLA alleles (Table I). Indeed, HLA-B*4201 is the most common restriction element of CTL responses targeting RM9-Nef, FL9-Vpr, and TL10-Nef, but is the allele least associated with polymorphism within each of these epitopes.
An alternative explanation is that the observed differences in patterns of escape mutation are not related to frequency of responders per se, but rather occur as a consequence of variation in magnitude and/or immunodominance of CTL responses restricted by the different HLA. The relative magnitudes of the CTL responses, measured by production of IFN-, to the six promiscuous B7 epitopes are shown in Fig. 3A. There are significant differences in the magnitude of TL9-specific responses, with CTL restricted by B*8101 and B*3910 producing more IFN- (mean spot forming units per million PBMC, 1815 and 2266, respectively) than those restricted by B*4201 (1285 spot forming units/million PBMC; p = 0.02 and p = 0.001, respectively). Furthermore, they differ in terms of immunodominance, with TL9 contributing a significantly greater proportion of the total response in B*8101-positive responders than in B*4201-positive responders (median 29 vs 22%, p = 0.05; Fig. 3B). This may help to explain the fact that both B*8101 and B*3910 are more strongly associated with escape mutation in TL9 than B*4201. However, no significant differences exist between the relevant B7 supertype alleles with respect to the magnitude or immunodominance of CTL responses specific for the other five promiscuous peptides (Fig. 3). Therefore, although frequency, magnitude, and immunodominance of CTL responses can vary according to HLA restriction element, the observed differences in the efficacy of CTL targeting identical peptides are not fully explained by these considerations.
|
To further investigate the observed differences between CTL targeting identical epitopes, but restricted by different HLA, we next examined their TCR usage, because previous studies have indicated a potential role for the TCR in selection of escape mutations (27, 32). We focused on the immunodominant TL9-Gag CTL response in B*4201- and B*8101-positive subjects, where a significantly higher level of escape mutation is found in the B*8101 positives (p = 1.7 x 10–7), to determine whether distinctive TCR usage was characteristic of the TL9-Gag response restricted by the different alleles. PBMC from B*4201- and B*8101-positive HIV-1-infected TL9 responders were stained with B*4201 and B*8101 TL9 tetramers, sorted to >99% purity by flow cytometry, and analyzed for TCRB gene expression as described previously (32). A conserved pattern of TCRBV gene usage is apparent, with four of five B*8101-positive and four of six B*4201-positive individuals using TCRBV 12.3 in their dominant clonotypes (Tables III and IV). However, there are clear differences observed in the CDR3s associated with each allele. HLA-B*8101-restricted TL9-specific CTL are significantly more polyclonal than those restricted by B*4201, possessing a mean of nine unique TCR clonotypes compared with three for B*4201 TL9-specific CTL (p = 0.02). An additional striking feature is the fact that the dominant clonotypes in three of six B*4201-restricted TL9-specific CTL populations use identical TCRs (comprising between 85 and 97% of total diversity), so-called "public" clonotypes (33). However, there are no shared public clonotypes between B*8101- and B*4201-restricted TL9-specific CTL, suggesting that these two molecules form distinct peptide/MHC complexes with the TL9 peptide. Indeed, despite a greater TCR diversity, only two of five B*8101 individuals possess identical TCRs, and in one this represents only a very minor clonotype (1% of total diversity; Tables III and IV). Thus, the restricting HLA allele can have a profound effect on the TCR usage, a factor that itself may be important in determining the efficacy of a given CTL response (32, 34).
|
|
Previously studies have proposed that the antiviral pressure exerted by CTL in vivo is related to their functional avidity (35, 36). Therefore, we next examined whether the apparent differences in selection pressure observed between HLA alleles presenting the same peptide could relate to differences in functional avidity of the CTL they restrict. Focusing on the immunodominant epitope TL9-Gag, we assessed the functional avidity of both B*4201- and B*8101-restricted TL9-Gag-specific CTL by using serial peptide concentrations to determine the peptide concentration required to give half maximal recognition (SD50), as measured by IFN- ELISPOT. Although there is some degree of variation, B*8101-restricted TL9-Gag responses are of significantly higher functional avidity than those restricted by B*4201, with a median EC50 of 7 ng/ml compared with 90 ng/ml (p = 0.01, Mann-Whitney U test; Fig. 4). This result is consistent with the observed substantially higher level of sequence variation observed within TL9-Gag in B*8101-positive subjects compared with B*4201-positive subjects.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
The reasons behind these apparent functional differences are unclear. In some cases, such as TL9-Gag, variation in the frequency of escape mutation may be linked to differences in response magnitude and immunodominance. However, no differences in magnitude, immunodominance, or, indeed, frequency of CTL responses are observed for the other five epitopes, suggesting that other factors, such as the nature of the peptide/MHC complex formed by the different HLA molecules, may play an important role in determining the quality of the CTL response elicited. Although HLA-B*0702, B*4201, and B*8101 possess the same primary anchor motif (Table I), HLA-B*0702 is known to have an additional preference for certain amino acids at positions 1, 3, and 7, binding into the A, D, and E pockets, respectively (39). The secondary binding motifs of HLA-B*4201 and B*8101 have not been characterized to date; however, some amino acid differences between B*0702, B*4201, and B*8101 occur at residues thought to influence the D and E pocket (HLA residues 114, 147, 152, and 156 in particular (39)). It is therefore possible that the common variants observed in this study, particularly those occurring at position 1, 3, and/or 7 (seen in TL9-Gag, RM9-Nef (Table II) and GL9-Gag and SM9-RT (data not shown)), have a differential effect on the binding of those epitopes to the three different HLA alleles examined in this study. Indeed, such considerations may explain why, within Nef-RM9, B*0702-restricted CTL select mutations at position 1 preferentially, whereas B*8101-RM9 CTL drive escape at position 6 in the epitope. In addition, subtle differences in the secondary binding pockets of these B7 supertype alleles may alter the conformation of the peptide within the binding groove, generating distinct peptide/MHC complexes. Support for this hypothesis comes from TCR analysis, showing that the HLA restriction element can profoundly affect the TCR usage of CTL specific for the same TL9-Gag epitope. B*4201-positive subjects whose Gag-TL9-specific CTL were analyzed displayed a characteristically narrow TCR repertoire, with three of six possessing identical CDR3 sequences in 85–97% of clones sequenced. In contrast, B*8101-restricted TL9-specific CTL were characterized by a much broader TCR repertoire and no such dominant public clonotype was observed. Finally, we show that B*8101-restricted TL9-specific CTL, most strongly associated with escape mutation, have a higher functional avidity that those restricted by B*4201. This is consistent with the observation that B*8101-restricted TL9-specific CTL responses are of a higher magnitude than those restricted by B*4201, as magnitude and functional avidity of CTL responses have been shown to be positively correlated (40). In addition, it is in line with earlier studies in the SIV/macaque model, which found that rapidly escaping epitopes are targeted by CTL with a higher functional avidity than those epitopes undergoing little or no escape mutation (35, 36). Taken together, these data show the existence of clear qualitative differences between CTL targeting the same epitope but restricted by different, albeit closely related, HLA alleles, and suggest that these differences are determined in large part by the nature of the presenting allele.
Examination of promiscuous epitopes presented by the B58 supertype suggests that these findings may be more generally applicable. The dominant B58 epitope in acute infection, TW10 (TSTLQEQIAW, Gag 240–249; Ref. 41) is presented both by HLA-B*57 and B*5801 alleles, but is associated with a lower frequency of escape in the B*5801-positive subjects (47 of 48 vs 60 of 78, p = 0.0014; Ref. 42). Again, the nature of the selection pressure differs, with B*57 but not B*5801 being particularly associated with mutation at P8 and P9 within TW10 (p = 0.0023; Ref. 42). The dominant epitope targeted in chronic infection in HLA-B*57-positive subjects, but not in B*5801-positive subjects, is the Gag epitope KF11 (KAFSPEVIPMF, Gag 162–172; Ref. 41). In recent studies comparing B*5701 with B*5703, alleles differing by two amino acids and for which in both cases KF11 is the immunodominant epitope in chronic infection, the frequency of escape mutation is substantially higher in the B*5703 positives (23 of 33 B*5703 vs 1 of 17, p = 0.0001; Refs. 24 and 43 ; Yu and Lichterfeld, unpublished observations). Similar to the B4201/8101-TL9 TCR data, comparison of the TCR sequences for the B*5703-KF11 and B*5701-KF11 CTL, demonstrates clear differences in TCR usage, with a highly conserved CDR3 region for the 
-chain in B*5701-positive subjects but not in B*5703-positive subjects (Yu and Lichterfeld, unpublished observations). Thus, functional differences between CTL targeting identical peptides but restricted by different but closely related HLA, are also apparent within the B58 supertype.
The TCR sequencing data demonstrate perhaps most clearly that it is the peptide/MHC complex, as opposed to the peptide in isolation, that is critical in defining the CTL response generated. Earlier studies in the SIV-macaque model noted that early escape in the Mamu-A*01 Tat-SL8 epitope was linked to a narrow TCR repertoire, and late escape in the Mamu-A*01 Gag-CM9 epitope was seen in association with a more diverse TCR repertoire (32), leading to the suggestion that breadth of TCR repertoire has an impact on escape mutation. The data shown here, however, suggest that the relationship between TCR usage and escape mutation is more complex, as the allele most strongly associated with escape in TL9-Gag, B*8101, displays the broader TCR repertoire. Indeed, due to alloreactivity, an individual’s TCR repertoire can even be skewed or limited by the other HLA class I alleles they possess. For example, HLA-B*08-restricted CTL specific for the EBV peptide FLRGRAYGL are associated with a single public TCR, unless isolated from individuals that also possesses B*4402/03, who use different TCRs, as the public clonotype is alloreactive with B*4402/03 (44). Because studies of this type have been hitherto technically problematic, it has been difficult to establish patterns linking TCR data to escape; further data of this type are likely to illuminate this area in the future.
The broader significance of these studies is that they underline once again the raison d’être of HLA diversity; that the differences existing between even closely related HLA alleles are the result of positive selection by human pathogens, and are therefore functionally significant. In HIV infection, outcome can depend upon a single amino acid difference, such as exists between B*3501 (not associated with rapid progression) and B*3502 or B*3503 (associated with rapid progression) (45). Similarly, B*5801 is associated with slow progression in the HIV epidemic in South Africa, whereas B*5802 (differing by three amino acids) is associated with rapid progression (21). Other recent examples from outside HIV further illustrate the same point (46, 47) and indicate that, in general, the critical detail of mechanisms of CTL-mediated control of pathogens such as HIV will result only from large cohort studies of subjects who have been HLA typed at high resolution.
Finally, it is important to note that ex vivo cytokine production, particularly IFN-, is frequently used to measure the immunogenicity of CTL vaccines (48). Moreover, a recently developed HIV "supertype" vaccine was validated based on the ability of individuals with different HLA alleles from the same supertypes (including B7), to respond to predicted "supertopes" in ex vivo IFN- ELISPOT assays (5). The data shown here, however, suggest that the ability to produce IFN- in response to exogenous peptide stimulation does not necessarily correlate to in vivo activity. Indeed, in this study, six of six of the promiscuous B7 epitopes identified by IFN- production elicit CTL responses that differ in their in vivo activity (using the criterion of selection pressure), depending on the restricting HLA allele. Therefore, caution is clearly required both in translating the ability of any given epitope to induce an ex vivo IFN- response into in vivo efficacy, and also in assuming that supertopes identified in this fashion will elicit equally effective CTL responses across the HLA supertype in question.
In conclusion, highly promiscuous supertopes may appear at first glance to be promising candidates for an epitope-based CTL vaccine against HIV-1. However, these data show that CTL targeting an identical epitope peptide may differ substantially in effectiveness when restricted by distinct, albeit closely related HLA alleles. It would be theoretically possible, therefore, for the same peptide to elicit an effective "driver"-type CTL response (37) in individuals expressing one HLA allele, and yet to prime ineffective "passenger"-type (37) CTL in individuals expressing a different HLA allele. At best, this would simply reduce the effective population coverage provided by any given supertope. At worst, however, it might compromise the effectiveness of a vaccine in some HLA groups by eliciting ineffective but immunodominant responses that occupy the immunological space that would otherwise be taken up by potentially beneficial CTL responses (1, 49). Any supertype-based CTL vaccine, therefore, irrespective of pathogen target, should take into account the fact that not all CTL elicited by a given epitope peptide will be equally effective.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported by the National Institutes of Health (Contract NO1-A1-15422, 2RO1AI46995-06) and the Wellcome Trust (to A.L. and P.J.R.G.). D.A.P. is a Medical Research Council (U.K.) Clinical Scientist; B.D.W. is a Doris Duke Distinguished Clinical Science Professor; P.J.R.G. is an Elizabeth Glaser Pediatric AIDS Foundation Scientist. 
2 Address correspondence and reprint requests to Dr. Alasdair Leslie, University of Oxford, Department of Paediatrics, Peter Medawar Building, South Parks Road, Oxford OX1 3SY, United Kingdom. E-mail address: alleslie8{at}yahoo.co.uk
Received for publication April 26, 2006. Accepted for publication July 4, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
interferon. J. Virol. 76: 4251-4259. 
1 helix of the HLA-B heavy chain can have an overriding influence on peptide-binding specificity. J. Immunol. 158: 1660-1669. [Abstract] T cell receptor repertoires against hen egg white lysozyme (HEL) in nontransgenic versus HEL transgenic mice. J. Exp. Med. 180: 861-872. This article has been cited by other articles:
|
|
 |
|
  L. R. McKinnon, R. Capina, H. Peters, M. Mendoza, J. Kimani, C. Wachihi, A. Kariri, M. Kimani, M. Richmond, S. K. Kiazyk, et al. Clade-Specific Evolution Mediated by HLA-B*57/5801 in Human Immunodeficiency Virus Type 1 Clade A1 p24 J. Virol., December 1, 2009; 83(23): 12636 - 12642. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Freeman, R. L. Warren, J. R. Webb, B. H. Nelson, and R. A. Holt Profiling the T-cell receptor beta-chain repertoire by massively parallel sequencing Genome Res., October 1, 2009; 19(10): 1817 - 1824. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Geldmacher, I. S. Metzler, S. Tovanabutra, T. E. Asher, E. Gostick, D. R. Ambrozak, C. Petrovas, A. Schuetz, N. Ngwenyama, G. Kijak, et al. Minor viral and host genetic polymorphisms can dramatically impact the biologic outcome of an epitope-specific CD8 T-cell response Blood, August 20, 2009; 114(8): 1553 - 1562. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. L. Ferre, P. W. Hunt, J. W. Critchfield, D. H. Young, M. M. Morris, J. C. Garcia, R. B. Pollard, H. F. Yee Jr, J. N. Martin, S. G. Deeks, et al. Mucosal immune responses to HIV-1 in elite controllers: a potential correlate of immune control Blood, April 23, 2009; 113(17): 3978 - 3989. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Kearney, F. Maldarelli, W. Shao, J. B. Margolick, E. S. Daar, J. W. Mellors, V. Rao, J. M. Coffin, and S. Palmer Human Immunodeficiency Virus Type 1 Population Genetics and Adaptation in Newly Infected Individuals J. Virol., March 15, 2009; 83(6): 2715 - 2727. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. T. Tan, E. Loggi, C. Boni, A. Chia, A. J. Gehring, K. S. R. Sastry, V. Goh, P. Fisicaro, P. Andreone, C. Brander, et al. Host Ethnicity and Virus Genotype Shape the Hepatitis B Virus-Specific T-Cell Repertoire J. Virol., November 15, 2008; 82(22): 10986 - 10997. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. A. Goepfert, W. Lumm, P. Farmer, P. Matthews, A. Prendergast, J. M. Carlson, C. A. Derdeyn, J. Tang, R. A. Kaslow, A. Bansal, et al. Transmission of HIV-1 Gag immune escape mutations is associated with reduced viral load in linked recipients J. Exp. Med., May 12, 2008; 205(5): 1009 - 1017. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Daucher, D. A. Price, J. M. Brenchley, L. Lamoreaux, J. A. Metcalf, C. Rehm, E. Nies-Kraske, E. Urban, C. Yoder, D. Rock, et al. Virological Outcome after Structured Interruption of Antiretroviral Therapy for Human Immunodeficiency Virus Infection Is Associated with the Functional Profile of Virus-Specific CD8+ T Cells J. Virol., April 15, 2008; 82(8): 4102 - 4114. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Liu, J. McNevin, H. Zhao, D. M. Tebit, R. M. Troyer, M. McSweyn, A. K. Ghosh, D. Shriner, E. J. Arts, M. J. McElrath, et al. Evolution of Human Immunodeficiency Virus Type 1 Cytotoxic T-Lymphocyte Epitopes: Fitness-Balanced Escape J. Virol., November 15, 2007; 81(22): 12179 - 12188. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Harari, C. Cellerai, F. B. Enders, J. Kostler, L. Codarri, G. Tapia, O. Boyman, E. Castro, S. Gaudieri, I. James, et al. Skewed association of polyfunctional antigen-specific CD8 T cell populations with HLA-B genotype PNAS, October 9, 2007; 104(41): 16233 - 16238. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
