Determination of a major histocompatibility complex class I restricting simian immunodeficiency virus Gag241-249 epitope

Several major histocompatibility complex class I (MHC-I) alleles such as HLA-B*57 have been shown to be associated with lower viral loads and better prognosis in HIV-1 infections, and MHC-I-restricted epitope-specific effective cytotoxic T lymphocyte (CTL) responses are found to play an important role in this reduction of viral loads ^[1-3]. Characterization of these effective CTLs could contribute to the development of an effective AIDS vaccine.

We have developed a prophylactic vaccine using a Sendai virus vector expressing simian immunodeficiency virus mac239 (SIVmac239) Gag (SeV-Gag) and have shown its protective efficacy against SIVmac239 challenge in a group of Burmese rhesus macaques (Macaca mulatta) sharing an MHC-I haplotype 90-120-Ia ^[4]. Involvement of SIVmac239 Gag_241-249 (SSVDEQIQW) epitope-specific CTL responses in this viral control have been indicated ^[5]. Interestingly, the SIVmac239 Gag_241-249 epitope is located in a region corresponding to the HLA-B*57-restricted HIV-1 Gag_240-249 epitope, TW10 (TSTLQEQIAW), and TW10-specific CTL responses have also been indicated to exert strong suppressive pressure on HIV-1 replication resulting in lower viral loads ^[6,7]. An SIVmac239 Gag_241-249-specific CTL escape mutation has been shown to result in a loss of viral fitness similarly with a TW10-specific CTL escape mutation ^[5]. In the present study, for further analysis of SIVmac239 Gag_241-249-specific CTL function, we have tried to determine the MHC-I that restricts this CTL epitope.

Among eight MHC-I alleles consisting of MHC-I haplotype 90-120-Ia ^[4,8], expression of three alleles, Mamu-A*90120-4, Mamu-A*90120-5, and Mamu-B*90120-6, was predominant at RNA levels. We cloned cDNAs of these three alleles and established HLA-A/B/C-negative human 721.221 cell lines ^[9] expressing these cDNAs, respectively. These cells were pulsed with 10 nmol/l of Gag_241-249 peptide and used as target cells for the CTL assay using an SIVmac239 Gag_241-249-specific CTL clone as the effector. Measurement of cytotoxicity in standard ⁵¹Cr release assay ^[5] revealed specific killing of Gag_241-249-pulsed cells expressing Mamu-A*90120-5, indicating restriction of this CTL epitope by the Mamu-A*90120-5 molecule (Fig. 1a).

Fig. 1

Both of the Mamu-A*90120-5-restricted SIVmac239 Gag_241-249 epitope and the HLA-B*57-restricted HIV-1 TW10 epitope are considered to have the same anchor residues, serine (S) at position 2 and tryptophan (W) at the carboxyl terminus. Comparison of amino acid sequences of antigenic peptide-binding domains (α1 and α2 domains) in Mamu-A*90120-5 with those in HLA-B*5701 revealed limited similarities (154/182 = 84.6%) between these two (Fig. 1b). This might be compatible with previous reports indicating that human and macaque MHC-I molecules with divergent peptide-binding grooves can bind similar or identical peptides ^[10,11]. MHC-I molecules form a peptide-binding groove including B-pocket and F-pocket that play a key role in determination of the binding peptide motif for its specific binding to the MHC-I. Mamu-A*90120-5 and HLA-B*5701 showed similarity in eight of 11 residues at 7, 9, 24, 25, 34, 45, 63, 66, 67, 70, and 99, which are considered to be anchor residues involved in B-pocket binding and in seven of eight residues at 77, 80, 81, 116, 123, 143, 146, and 147 involved in F-pocket binding ^[11-13].

In addition, TW10 epitope-specific CTLs, HLA-B*57-restricted HIV-1 Gag_147-155 [ISW9 (ISPRTLNAW)] epitope-specific CTLs have also been indicated to exert strong selective pressure on HIV-1 ^[14]. The SIVmac239 Gag_149-157 amino acid sequence corresponding to the HIV-1 Gag_147-155 epitope region is LSPRTLNAW, showing a difference at the amino terminus, and CTL responses specific for a peptide including the SIVmac239 Gag_149-157 amino acid sequence were not induced by SeV-Gag vaccination in Mamu-A*90120-5-positive macaques (data not shown). Interestingly, the SIVmac239 Gag 148th proline (P) and 149th leucine (L) correspond to the HIV-1 Gag 146th P and the 147th L, respectively that have been indicated to be selected in HIV-1-infected humans possessing HLA-B*57. Selection of the former 146th P has been shown to result in escape from ISW9-specific CTL recognition by disturbance in antigen processing ^[14]. Thus, it is speculated that the SIVmac239 Gag_149-157-derived peptide may not be presented by Mamu-A*90120-5 even if it has an ability to bind this peptide.

Both SIVmac239 Gag_241-249-specific CTLs and HIV-1 TW10-specific CTLs have been indicated to exert strong suppressive pressure on SIV/HIV-1 replication and select for a mutation resulting in escape from their recognition at the cost of viral fitness. Thus, this Gag region may be a promising CTL target for viral control, and SIVmac239 infection in Mamu-A*90120-5-positive macaques could be a unique model for examining viral replication in the presence of those CTLs targeting this region like TW10-specific CTLs. Finally, we obtained a phycoerythrin-conjugated Gag_241-249 epitope-Mamu-A*90120-5 tetramer for specific detection of Gag_241-249-specific CTLs. This could be useful for the analysis of Gag_241-249-specific CTL responses in Mamu-A*90120-5-positive macaques infected with SIVmac239.

Acknowledgements

The present work was supported in part by grants from the Ministry of Education, Culture, Sports, Science, and Technology, grants from the Japan Health Sciences Foundation, and grants from the Ministry of Health, Labor, and Welfare in Japan.