Presentation of rVV to HIV-specific CTL
The presentation of new recombinants (rVV-gagA1, rVV-gagA2, rVV-gagC, rVV-gagD) to CTL and the resulting lysis as measured in Cr-release assays was checked using two cell lines and the clone shown in Fig. 2. The 24-14 cell line (derived from patient 868) was specific for an HLA-B*27-restricted epitope (amino acids 263–272), which was invariant between the different isolates, and resulted in comparable levels of lysis in targets infected with the different recombinants (Fig. 2a). An HLA-B*0801-restricted 24–13 (amino acids 259–267) variant-specific clone from patient 008 caused similar levels of lysis when incubated with HLA-B*0801-positive targets infected with the rVV constructs (Fig. 2b). This clone has been shown to recognize variant 24–13 peptides with either D or E at position 2. In contrast, a CTL line specific for an HLAA*0201-restricted Gag epitope (amino acids 77–85; also from patient 868) failed to recognize targets infected with either of the constructs expressing Gag p55 cloned from A clade isolates. Sequence analysis of this epitope in different clades suggested that the lack of recognition may have been due to a phenylalanine substituted for tyrosine at the third position of this epitope in Gag p55 expressed by both of the A clade rVV. The variation within this epitope in the D clade rVV (isoleucine at position 5) was well recognized by patient 868 CTL and corresponded to a natural variant found in his provirus (unpublished data). The above data confirms that target cells infected with rVV expressing p55 from A, C and D clades of HIV-1 were able to present Gag antigens in an appropriate fashion to CTL.
Cross-clade recognition of p55 by CTL
To test whether the response to Gag was cross-reactive between clades we cultured CTL from individuals infected with different clades of HIV-1 restimulated with autologous virus, and then tested for recognition of autologous Epstein-Barr virus-transformed B cells (BCL) infected with the different rVV. Although this restimulation system may not have been the most sensitive method of inducing CTL in culture, it did avoid introducing any specificity bias that may have been induced by the use of peptides or vaccinia for restimulation, and should therefore better reflect CTL specificity in vivo. Out of six Ugandan patients tested, five (U003, T009, T010, T014 and T029) had inducible CTL responses that were able to recognize the rVV expressing all four different clades of HIV-1 Gag (Fig. 3a). The remaining patient (T007) made a response that was specific for rVV-gagA2 (HIV-1). The response in T007 was more than clade-specific as autologous target cells infected with rVV-gagA1 were not recognized. This patient appeared to have a response that was completely specific for A2 Gag p55, and his CTL were therefore probably recognizing an epitope that was different in A1 and A2 Gag. This patient was shown by HMA to be infected with A or C clade virus. One patient (U003), also infected with A or C clade virus, although having CTL showing some cross-reactivity between clades, preferentially recognized A1 Gag.
None of the patients showed any HIV-2 Gag cross-reactivity when lysis of HIV-2 rVV-gag-infected target cells was compared with lysis of autologous B cells infected with rVV-PB2 or uninfected cells.
Out of seven patients tested in the United Kingdom (five out of seven infected with B clade HIV-1), the majority had CTL responses (> 10% greater lysis than that of the same target cells infected with an irrelevant rVV) that recognized two or more clades (Fig. 3b). Patient EW, who was infected with HIV-1 in East Africa and therefore may have been infected with non-B clade virus, made a CTL response recognizing A and C clade Gag only. Patient 36M had cross-reactive CTL that were capable of recognizing all clades except D. Patient 46D had CTL which primarily recognized autologous target cells expressing rVV A2 Gag with a weak response to clade D Gag. Patient pt84 had cross-reactive CTL but target cells infected with rVV-gagD were seen best. In patient pt91, high background lysis of uninfected target cells or target cells infected with rVV-PB2 was seen but specific recognition of target cells expressing rVV B Gag was observed. Both patients 065 and 0120 made cross-reactive CTL responses to autologous target cells expressing p55 from all the HIV-1 clades tested.
None of these patients showed any HIV-2 Gag cross-reactivity when lysis of rVV HIV-2 Gag target cells was compared with lysis of autologous cells infected with rVV-PB2 or uninfected cells.
Further characterization of the cross-reactive CTL response in three Ugandan patients
In three Ugandan patients (U003, T009 and T029), further characterization of the cross-reactive CTL response was carried out using individual peptides and peptide pools from HIV-1SF2 p24 and HIV-1SF2 p17 described above. Patient U003 had made a cross-reactive CTL response with a preference for A1 (Fig. 3a). When tested with pools of peptides this patient made a highly specific response to the pool of peptides p24-12 to p24-17 (Fig. 4a). When CTL from patient U003 were tested with autologous target cells pulsed with individual peptides, a clear response was seen to peptide p24-17 (Fig. 4b). When sequence analysis was performed, several amino-acid changes were seen within this 20-mer in the different clades (Fig. 4c). Since CTL from this individual appeared to have a preference for A1 Gag, this suggests that either the cross-reactive epitope recognized was at the aminoterminus of the peptide, which is relatively invariant in HIV-1 (but with two major substitutions in HIV-2), and that there was a second A1-specific response that was not seen in our assays because we were using SF2-based peptides, or that the epitope was at the carboxy-terminus of the peptide (which is more variable) with the A1 specificity determined by the change of AE to KL. In either case, HIV-2 non-cross-reactivity was accounted for by the changes seen in the HIV-2 sequence. This patient had class I HLA-A*1/A*3001, B*71/B*50, C*03/C*04 and it must be assumed that an epitope contained within p24-17 was being presented by one of these molecules. To define the MHC restriction of this epitope, matched BCL that share single class I molecules with U003, were tested for their ability to present p24-17 to U003 CTL. It was observed that only autologous BCL and a BCL sharing HLA-B*71 were recognized when prepulsed with p24-17. It is therefore likely that HLA-B*71 was the restricting element for p24-17.
CTL from patient T009 were cross-reactive and able to recognize autologous target cells infected with rVV expressing p55 from all clades tested (Fig. 3a). When CTL from this patient were tested with autologous target cells pulsed with pools of peptides, a clear response was seen to pool p24-18 to p24-22 (Fig. 4d). Furthermore, a specific response to peptide p24-22 was seen (Fig. 4e). When sequence analysis of this epitope within the different clades was performed, the peptide was seen to be invariant at the amino-terminal and it seemed likely that the peptide epitope recognized was within this area of the peptide (Fig. 4f). Such positioning would suggest that the non-recognition of HIV-2 Gag by CTL from this patient was due to the M→L change at position 4 of the 20-mer. Alternatively, the epitope could be at the carboxy-terminus and T009 CTL could tolerate the glutamic acid and serine substitutions at positions 13 and 15, respectively. The failure to recognize HIV-2 Gag rVV could then be accounted for by the glutamine and leucine substitutions at positions 16 and 20. This patient was tissue-typed and had class I HLA-A*30.02/A*6802, B*7/B*13, C*0602/ C*0702v, and it was again assumed that the epitope contained within p24-22 was being presented by one of these molecules. A panel of class I-matched BCL were used to define the MHC restriction of this epitope. Autologous BCL and a cell line sharing only HLA-B7 were able to present p24-22 to T009 CTL, and it was concluded that HLA-B7 was the restricting element for this peptide.
CTL from patient T029 were cross-reactive and able to recognize autologous target cells infected with rVV expressing p55 from all clades tested (Fig. 3a). When CTL from this patient were tested with autologous target cells pulsed with pools of peptides, a clear response was seen to pool p24-12 to p24-17 (Fig. 4g).
Furthermore, a specific response to p24-16 was seen (Fig. 4h). A comparison of this SF2 sequence with the p55 sequences from rVV showed that the central residues of the peptide were invariant between the HIV-1 sequences (Fig. 4i). If the CTL from T029 were specific for an epitope in the centre of this 20-mer, then the substitution of a phenylalanine and arginine at positions 11 and 12 of the peptide with a glutamine and serine in the HIV-2 Gag sequence would explain why this rVV was not recognized. This patient had class I HLA-A*30.02/A*6602, B*5801/B*81, C*04/C*0701, and it was assumed that the epitope contained within peptide p24-16 was being presented by one of these molecules.
Thus, the above experiments demonstrated preliminary characterization of cross-reactive peptide epitopes in relatively conserved parts of the Gag protein, which should ideally be included in a future vaccine.
We have shown extensive cross-reactivity in CTL responses in HIV-infected individuals between four of the major HIV-1 clades. Three of these clades are the predominant subtypes in regions of the world where there is greatest need for prevention of infection/disease, and where a vaccine is most likely to be tested. The first vaccines that are tested are likely to be based on B clade HIV-1, since most experimental laboratory work has been based on the B clade because it is prevalent in the developed world.
The data presented here demonstrate that responses of clonal CTL populations or of cultured long-term CTL lines that have been propagated using single specific peptides, may be very specific in their activity and may not reflect the broader response in vivo. However, it seems that most patients infected with diverse clades of HIV-1 have the capability to mount cross-reactive CTL responses that recognize the Gag proteins of clades of HIV-1 other than those with which they were infected. A few patients appear to make responses that are specific for Gag from a single clade. However, it has been reported  that most individual patients infected with HIV-1 make CTL responses specific for several peptide epitopes derived from, for example, Nef, Pol, and Env, as well as Gag. It seems likely, therefore, that the cross-reactivity that we are measuring is the minimum that might be found in any one patient. We may also be missing isolate-specific CTL responses. For example, if we had used a single rVV expressing p55 Gag from clade A1 virus when testing patient T007, we would not have seen any Gag-specific CTL response. Clearly this patient had CTL that recognized an epitope that was present in autologous virus and rVV A2 but was absent in rVV A1. No recombinant construct is going to be able to reflect absolutely the virus with which a patient is infected unless recombinants specific for each patient are prepared. Patients with a particular HLA class I type (such as HLA-B8) may make CTL responses that are specific for peptides from variable regions of the virus , and such individuals might be expected to show less cross-reactivity.
Despite initial reports , we failed to demonstrate the presence of CTL that were cross-reactive between HIV-1 and HIV-2. We were unable to test expression of HIV-2 Gag in cells infected with HIV-2 rVV-gag in the same way as the HIV-1 rVV-gag constructs were tested. However, the same stocks of HIV-2 rVV-gag are being used in ongoing experiments in West Africa where CTL from a large number of individuals infected with HIV-2 have been shown to recognize autologous target cells infected with HIV-2 rVV-gag. Some cross-reactive CTL have recently been reported in exposed but uninfected individuals who may have been exposed to both HIV-1 and HIV-2 . In this case, CTL were restimulated using specific peptides and may, or may not, reflect a normal in vivo response. Our results make the described, although not confirmed, putative protection offered by prior infection with HIV-2 to infection with HIV-1  unlikely to be mediated by CTL.
Many of our Ugandan patients may have been exposed to A, D and C clade virus (or may have been infected with recombinant viruses), but both patients EW and 065 had a documented single exposure to virus, so it seems unlikely that cross-reactivity is a function of multiple exposure. This is however a preliminary report, and we need to study many more patients to estimate the frequencies of cross-reactive CTL, to examine specificities other than to p55 Gag and to examine CTL activity in individuals infected with defined recombinants. We also need to elucidate whether these CTL responses are cross-reactive at the clonal level or whether, as seems more likely, we are measuring polyclonal responses that reflect the in vivo situation. In this respect we feel that a natural infection with a quasispecies of a dynamic virus such as HIV-1 may lead to good polyclonal cross-reactive cellular immune responses, whereas a vaccine based on a single sequence (in a recombinant vector or a DNA vaccine for example), may induce less cross-reactivity. A vaccine based on a live attenuated virus, as has been suggested by Desrosiers and coworkers [29,30], would be expected to induce cross-reactive responses.
In order to evaluate the CTL responses induced by vaccine preparations we need to continue to identify cross-reactive peptides restricted through common HLA class I molecules in genetically diverse populations that can be used in vaccine trials. Such peptides should be should be included in vaccine preparations and will be used in restimulation strategies in order to quantify CTL responses.
The authors thank the patients and staff at the Uganda Virus Research Institute and the AIDS Support Organization clinics for their cooperation and M. Bunce and M. Barnado from the Transplant Centre, Churchill Hospital, Oxford for help with the molecular tissue typing; also B. Biryahwaho and P. Tuguma for support in Uganda.
1. Walker C, Moody D, Stites D, Levy J: CD8 lymphocytes can control HIV infectionin vitroby suppressing virus replication
2. Koup R, Safrit JT, Cao Y: Temporal association of cellular immune responses with the initial control of viraemia in primary human immunodeficiency virus type 1 syndrome
. J Virol
3. Safrit J, Koup R: The immunology of primary HIV infection: which immune responses control HIV replication?Curr Opin Immunol
4. Carmichael A, Jin X, Sissons P, Borysiewicz L: Quantitative analysis of the human HIV-1 specific cytotoxic T-lymphocyte (CTL) response at different stages of infection: differential CTL responses to HIV-1 and Epstein Barr virus in late disease
. J Exp Med
5. Ariyoshi K, Cham F, Berry N, Jaffar S, Sabally S, Corrah T, Whittle H: HIV-2-specific cytotoxic T-lymphocyte activity is inversely related to proviral load
6. Moss P, Rowland-Jones S, Frodsham P, et al.
: Persistent high frequency of HIV specific cytotoxic T cells in peripheral blood of infected donors
. Proc Natl Acad Sci USA
7. Cheynier R, Langlade-Demoyen P, Marescot MR, et al.
: Cytotoxic T-lymphocyte responses in the peripheral blood of children born to HIV-1 infected mothers
. Eur J Immunol
8. Aldhous MC, Watret KC, Mok JY, Bird AG, Froebel KS: Cytotoxic T-lymphocyte activity and CD8 subpopulations in children at risk of HIV infection
. Clin Exp Immunol
9. Rowland-Jones S, Nixon D, Aldous M, et al.
: HIV specific CTL activity in an HIV-exposed but uninfected infant
10. Langlade-Demoyen P, Ngo N, Ferchal F, Oksenhendler E: HIVnef-specific cytotoxic T-lymphocytes in non-infected heterosexual contact of HIV-infected patients
. J Clin Invest
11. Rowland-Jones S, Sutton J, Ariyoshi K, et al.
: HIV-specific cytotoxic T cells in HIV-exposed but uninfected Gambian women
. Nature Med
12. Gallimore A, Cranage M, Cook N, et al.
: Early suppression of SIV replication by CD8+nef-specific cytotoxic T cells in vaccinated macaques
. Nature Med
13. Rubsamen-Waigmann H, von Briesen H, Holmes H, et al.
: Standard conditions of virus isolation reveal biological variability of HIV type 1 in different regions of the world
. AIDS Res Hum Retroviruses
14. Louwagie J, McCutchan IE, Peeters M, et al.
: Phylogenetic analysis ofgaggenes from 70 international HIV-1 isolates provides evidence for multiple genotypes
15. Zachar V, Goustin A, Zacharova V, et al.
: Genetic polymorphism of envelope V3 region of HIV type 1 subtypes A, C and D from Nairobi, Kenya
. AIDS Res Hum Retroviruses
16. Weber J, Fenyo EM, Beddows S, Kaleebu P, Bjorndal A: Neutralisation serotypes of human immunodeficiency virus type I field isolates are not predicted by genetic subtype. The WHO network for HIV isolation and characterisation
. J Virol
17. Moore J, Trkola A: HIV type 1 coreceptors, neutralization sub-types, and vaccine development
. AIDS Res Hum Retroviruses
18. McAdam S, Klenerman P, Tussey L, et al.
: Immunogenic variants that bind to HLA-B8 but fail to stimulate CTL responses
. J Immunol
19. Phillips R, Rowland-Jones S, Nixon D, et al.
: Human immunodeficiency virus genetic variation that can escape cytotoxic T cell recognition
20. Weber J, Cheinsong-Popov R, Callow D, et al.
: Immunogenicity of the yeast recombinant p17/p24:ty virus-like particles (p24-VLP) in healthy volunteers
21. Bachmann MH, Delwart EL, Shpaer EG, Lingenfelter P, Singal R, Mullins JI: Rapid genetic characterization of HIV type 1 strains from four World Health Organization-sponsored vaccine evaluation sites using a heteroduplex mobility assay
. AIDS Res Hum Retroviruses
22. Bunce M, Fanning GC, Welsh KI: Comprehensive, serologically equivalent DNA typing for HLA-B by PCR using sequence specific primers (PCR-SSP)
. Tissue Antigens
23. Bunce M, Barnado MC, Welsh KI: Improvements in HLA-C typing using sequence-specific primers (PCR-SSP) including definition of HLA-Cw9 and Cw10 and a new allele HLA-‘Cw7/8v'
. Tissue Antigens
24. Moss B, Flexner C: Vaccinia virus expression vectors
. Annu Rev Imunol
25. Nixon D, Townsend A, Elvin J, Rizza C, Gotch F, McMichael A: HIV-1gag-specific CTL defined with recombinant vaccinia viruses and synthetic peptides
26. Autran B, Letvin NL: HIV epitopes recognised by cytotoxic T-lymphocytes
27. Gotch F, Nixon D, Gallimore A, McAdam S, McMichael A: Cytotoxic T lymphocyte epitopes shared between HIV-1, HIV-2, and SIV
. J Med Primatol
28. Marlink R, Kanki P, Thior I, et al.
: Reduced rate of disease development after HIV-2 infection as compared to HIV-1
29. Desrosiers RC: Safety issues facing development of a live-attenuated, multiply deleted HIV-1 vaccine
. AIDS Res Hum Retroviruses
30. Wyand M, Manson K, Garcia M, Montefiori D, Desrosiers R: Vaccine protection by a triple deletion mutant of simian immunodeficiency virus
. J Virol
Keywords:© Lippincott-Raven Publishers.
Africa; CD8; cellular immunity; vaccine