The possibility that immunodominant human leukocyte antigen (HLA) class 1 major histocompatibility complex (MHC)-restricted cytotoxic T-lymphocyte (CTL) responses might be directed through HLA-C molecules has often been discounted. HLA-C antigens are expressed on the cell surface at approximately 10% of the levels of HLA-A and B , and there is greater homogeneity of the antigen recognition site within HLA-C molecules than within HLA-A or B . However, elution of peptides from HLA-C molecules yields peptides approximately 9 amino acids in length, fully comparable to those eluted from HLA-A and B molecules . In addition, several naturally occurring HLA-C-restricted CTL epitopes have been described .
An HLA-CW*0102-restricted CTL epitope is now described. This is one of two dominant responses detected in a donor in whom eight HIV-specific CTL responses have been identified. This is believed to be the first HLA-CW*0102-restricted epitope identified.
Donor 868 was an asymptomatic HIV-1-seropositive man who acquired HIV infection through homosexual contact. In May 1996 his CD4 count was 230 × 106 cells/l and his viral load was 193 000 RNA copies/ml plasma. Antiretroviral therapy (zidovudine plus didanosine) was commenced at this timepoint. His most recent virus load was 5000–5500 RNA copies/ml (March 1997), and the CD4 cell count at this time was 400–500 × 106/l. His HLA class 1 tissue type [by sequence-specific primer (SSP) polymerase chain reaction (PCR)]  was HLA-A*0101/*24, B*2705/*3501, CW*0102/*0401.
Bulk CTL cultures were generated as previously described . Autologous Epstein-Barr virus-transformed B cells were pulsed with pools of overlapping 15–20mer peptides spanning the HIV-1 Gag and Nef proteins, and assayed against bulk-cultured CTL from donor 868. In addition, bulk CTL were assayed for recognition of peptides previously defined as HIV-specific epitopes  that were restricted by those HLA class 1 molecules expressed by donor 868.
Seven previously described epitopes were recognised (Fig. 1a). A strong response was also seen to a novel epitope within the 20mer p24.4 (amino-acid sequence AFSPEVIPMFSALSEGATPQ; HIV-1LAI Gag residues 163–182). There was similar recognition of two 15mers within this 20mer, which overlapped by 10 amino acids (AFSPEVIPMFSALSE, residues 163–177; and VIPMF SALSEGATPQ, residues 168–182), indicating that the epitope was contained in both 15mers (Fig. 1b). Using truncated peptides, the optimally recognised peptide was the octamer VIPMFSAL (Fig. 1c). The presenting molecule was HLA-CW*0102 (Fig. 1d).
It is likely that HLA-C-restricted CTL responses may be undetected, even when present, and therefore be underestimated for two main reasons. First, inaccuracies associated with HLA typing are exacerbated where HLA-C is involved, partly because of low surface expression, and partly because of serological blanks (CW*12 to CW*17), which are undetectable, probably due to a lack of suitable antisera. In a comparison between serological and SSP-PCR-obtained HLA-C phenotypes, there were discrepancies in 37% of HLA-C alleles, mostly due to the failure of serology to detect alleles . A second possible source of underestimation of HLA-C-restricted CTL responses is the result of close linkage disequilibrium between certain HLA-C alleles and HLA-B alleles, such as HLA-CW*0802 and B*1402. If the HLA-C allele is undetected, this may suggest an apparent HLA-B-restricted response .
No peptides have been eluted from HLA-CW*0102, but predictions of the likely anchor residues of binding peptides can be made using data from the determination of other peptide-MHC class 1 structures , comparison of the amino-acid sequence of different MHC class 1 molecules , and by reference to peptides eluted from other HLA-C molecules . Using the nomenclature of Saper et al. , there are six pockets, A-F, within the peptide-binding groove of the HLA molecule, into which particular residues of the peptide may be accommodated, the residues at position 2 (P2) in the peptide being accommodated in the B pocket, and the residue at the C-terminal position (PC) in the peptide being accommodated in the F pocket. The presumed anchor positions for a peptide epitope presented by HLA-CW*0102 would be at P2 and PC, by analogy with peptides binding to HLA-CW*0301, CW*0401, CW*0602 and CW*0702 , with secondary anchor positions at P5/P6. In these four HLA-C molecules, the PC anchor was determined as a medium-to-large-sized hydrophobic residue (such as Leu or Phe). The residues likely to contribute to the F pocket suggest a very similar PC anchor in HLA-CW*0102-binding peptides. Small (such as Pro) to large hydrophobic residues act as anchor residues in the B pocket of HLA-C-binding peptides from elution data . Therefore, the peptide VIPMFSAL would be consistent with Leu as the carboxy-terminal anchor residue and Ile as the P2 anchor.
In summary, use of molecular MHC class 1 tissue typing methods enabled the identification of the HLA restriction of the response to be reached rapidly and unequivocally. The immunodominance of the HLA-CW*0102-restricted response in bulk-cultured lymphocytes in this donor implied that CTL responses directed through HLA-C molecules are in no way inferior to those directed through HLA-A and B molecules. Full use of molecular tissue typing should be made in order that HLA-C-restricted CTL responses are not underestimated.
Overlapping peptides were supplied by the Medical Research Council AIDS Directed Programme (Dr H. Holmes).
1. Snary D, Barnstable CJ, Bodmer WF, Crumpton MJ: Molecular structure of human histocompatibility antigens: the HLA-C series
. Eur J Immunol
2. Zemmour J, Parham P: Distinctive polymorphism at the HLA-C locus: implications for the expression of HLA-C
. J Exp Med
3. Falk K, Rotzschke IO, Grahovac B, et al.
: Allele-specific peptide ligand motifs of HLA-C molecules
. Proc Natl Acad Sci U S A
4. Rammensee H-G, Friede T, Stevanovic S: MHC ligands and peptide motifs: first listing
5. Bunce M, O'Neill CM, Barnardo MCNM, et al.
: Phototyping: comprehensive DNA typing for HLA-A, -B, -C, -DRB1, -DRB3, -DRB4, -DRBS and DQB1 by PCR with 144 primer mixes utilising sequence specific primers (SSP-PCR)
. Tissue Antigens
6. Nixon DF, Townsend ARM, Elvin JG, Rizza CR, Gallwey J, McMichael AJ: HIV-1 gag-specific cytotoxic T lymphocytes defined with recombinant vaccinia virus and synthetic peptides
7. Brander C, Walker BD: The HLA class I restricted CTL response in HIV infection: identification of optimal epitopes
. In HIV Molecular Immunology Database
1995. Part IV. Edited by Korber BTM, Brander C, Walker BD, Koup R, Moore JP, Haynes BF, Myers G. Los Alamos: Los Alamos National Laboratory, Theoretical Biology and Biophysics; 1995:1–9.
8. Bunce M, Barnardo MCNM, Procter J, Marsh SGE, Vilches C, Welsh KI: High resolution HLA-C typing by PCR-SSP: identification of allelic frequencies and linkage disequilibria in 604 consecutive unrelated random UK Caucasians and a comparison with serology
. Tissue Antigens
9. Madden DR, Garboczi DN, Wiley DC: The antigenic identity of peptide-MHC complexes: a comparison of the conformations of five viral peptides presented by HLA-A2
10. Arnett KL, Parham P: HLA class 1 nucleotide sequences
. Tissue Antigens
11. Saper MA, Bjorkman PJ, Wiley DC: Refined structure of the human histocompatibility antigen HLA-A2 at 2.6Å resolution
. J Mol Biol