Date of Award
Doctor of Philosophy
To overcome the extreme antigenic diversity of HIV, there is increasing interest in developing vaccines that target CD8 T cell responses to conserved regions of the viral proteome. Understanding the immunogenic potential of these domains is therefore critical for the success of this vaccine strategy. CD8 T cell responses to conserved domains restricted by the "neutral" and highly prevalent HLA-A2 allele have not been extensively characterized. Here we describe a novel essentially invariant 10-mer epitope in HIV Gag (p2414-23, RTLNAWVKVV (RV10)) in that it is expressed by the majority HIV isolates from all clades. Here we examined the character of the CD8 T cell response to this subdominant epitope in healthy volunteers and in HIV-infected individuals.
RV10 was identified with in vitro immunized CD8 T cells from eight seronegative HLA-A*0201 (A*02) carriers. In brief, CD8 T cells purified by immunobeads were primed with autologous dendritic cells (DCs) transduced by a lentiviral vector to express Gag. Antigen-specific T cells were further expanded by three weekly re-stimulations with autologous monocytes pulsed with 123 15-mer overlapping peptides (OLPs) spanning across the entire Gag protein. Reactivities to individual OLPs were identified by IFN-ELISPOT assays using matrix peptide pools followed by interrogation at the single peptide level. Of particular interest, OLPs 7908 and/or 7909 encoding RV10 were recognized by CD8 T cells immunized from all donors. RV10 was fine mapped from five cultures using a cassette of truncated peptides within OLP 7908. Interestingly, the 10-mer peptide encodes two well-known HLA-A2-restricted 9-mer epitopes, Gag (p2415-23, TLNAWVKVV (TV9)) and Gag (p2414-22, RLNAWVKVV (RV9)). This 10-mer Gag sequence appears to enjoy a paradoxically high T cell repertoire reserve despite its subdominance as judged by the rarity of reported responses in HLA-A*02+ HIV-infected individuals.
To analyze the character of the CD8 T cell responses to RV10, TV9 and RV9, parallel cultures were generated from four healthy HLA-A*02 carriers. All cultures were stable with distinct populations of tetramer-binding cells. RV10- and TV9-specific T cells were consistently cross-reactive as measured by functional assays such as cytotoxicity and polychromatic flow cytometry. In contrast, they were less reactive to RV9. On the other hand, RV9-specific T cell cultures recognized the cognate peptide efficiently but were minimally cross-reactive to the other two peptides. Interesting to note, the three epitopes sharing the same 8-mer antigenic core display two distinct patterns of reactivities. To show that the three peptides are naturally processed and presented by MHCI molecules of infected cells, CD8 T cells were primed with autologous transduced DCs and re-stimulated with autologous transduced B cells expressing Gag. Tetramer-binding RV10-, TV9- and RV9-specific CD8 T cells were found in cultures from four of four donors, thereby indicating that these reactivities were immunized by epitopic determinants naturally processed and presented by both APCs.
TCRB gene usage by TV9- and RV10-specific CD8 T cells sorted to >98% purity were compared by an unbiased template-switch anchored rt-PCR . While a single TCR can recognize both peptides, no clonal overlap was found between three sets of parallel cultures of TV9 and RV10 cells. These observations suggest that the repertoires for the two peptides are very diverse and furthermore, cross-recognition among the clonotypes is common.
Alanine scanning mutagenesis revealed that substitution of asparagine (N) (position 4 of RV10 and position 3 of TV9) was associated with the most precipitous decline in recognition by RV10- and TV9-T cells, respectively. Moreover, recognition by RV10-specific T cells appeared to be more dependent on tryphtophan in position 6 than their TV9 counterparts. Of note, alanine substitutions at all positions of both peptides did not significantly reduce binding to HLA-A*02. In some instances, binding was somewhat increased. In summary, the arginine residues in these peptides appear to serve as a key TCR contact site, thereby accounting for reciprocal cross-recognition. RV10-specific T cells appeared to recognize the phenylalanine residue as well. This provides compelling albeit indirect evidence that RV10 is a distinct epitope.
Here we have analyzed in detail the character of CD8 T cell responses to a set of three nested HLA-A*0201 epitopes that share a core 8-mer antigenic determinant in common. This finding has implications in T cell-based HIV vaccine design for the one of the most common HLA allele worldwide. Specifically, our data suggest a new mechanism of immunological escape ("CTL escape") by this highly variant virus. The addition of an arginine to the TV9 sequence, a seemingly useful HLA-A*02 epitope, results in immune interference due to the simultaneous presentation of cross-reactive epitopes by the same MHCI molecule, thereby preventing a full-scale activation at priming and over the long-term maintenance of Gag-specific CD8 T cell expansion. Understanding this immunological mechanism may be of great potential importance given the multiplicity of overlapping CD8 epitopes and their tendency to cluster in various HIV protein regions.
Received from ProQuest
Margaret Christina Costanzo
Costanzo, Margaret Christina, "Mobilizing Subdominant HIV-Specific CTLs As A Novel Vaccine Strategy" (2012). Open Access Theses & Dissertations. 2066.