(A) Upper panel: The DR1-tetramer with peptide PG13 but not IP13 or transferrin receptor (TFR) bound to AC-25 T cells. Lower panel: The PG13 tetramer did not bind a control DQ7 restricted clone specific for a different region of Gag. (B) Relative binding of PP16, PG13, and TFR-containing tetramers showing brightest staining with the PP16 tetramer, representative of two replicate experiments. All staining was performed at 25C for 1.5 hours. 50,000 events were collected per condition. MFI = Mean Fluorescence Intensity (C) Competitive inhibition of PG13 tetramer binding by unlabeled DR1-PP16 monomeric complex. AC-25 T cells were stained with low dose PG13 tetramer (40 μg/ml) and monomers were added as noted. TFR only = cells stained with negative control tetramer. Mean fluorescence intensity from two replicate experiments was normalized and averaged. Error bars represent the standard error.

Top: Structure of native gag PP16 peptide bound to HLA-DR1, from the crystal structure of the complex (Zavala-Ruiz et al., 2004). The view shows the C-terminal region of the peptide binding site, viewed from above. The dashed bonds indicate the continuation of the peptide backbone to the terminal proline residue; this was not visualized in the crystal structure. The P9 residue serine is buried in the site, and the other residues are partially or completely exposed. The glycine at P11 allows the peptide backbone to bend sharply, forming a hairpin that orients the P13 residue threonine for interaction with TCR.
Bottom: Hypothetical model of the Pro11 peptide bound into the same site, with the proline at P11 indicated in a common conformation. Note that this peptide cannot assume the hairpin conformation shown above because the proline 5-membered ring severely constrains the phi angle (curved arrow) to ~60 degrees. This angle is free to rotate for glycine.

(A)The Th clone from AC-25 was stimulated with varied concentrations of PP16 or PG13. The left panel shows IFN-γ secretion in an intracellular cytokine staining assay with at least 50,000 events collected/condition. The middle panel shows proliferation by ³H-thymidine incorporation, performed in triplicate. The right panel shows serine esterase release in the supernatant 4 hours after stimulation. PP16 elicited responses while PG13 did not. Each of the experiments was repeated at least twice. (B) In a control experiment peptide PG13 was able to stimulate IFN-γ secretion measured by ELISPOT assay from a DR4 restricted clone from subject 161J to a comparable degree to the minimum epitope EVIPMFSALS for the 161J Th cell clone. The data represent the average of seven experiments. Responses to PG13 and EVIPMFSALS were not statistically significantly different (P > 0.05), while PP16 responses were greater: ** p < 0.05, PG13 and EVIPMFSALS vs. PP16; * p < 0.05, EVIPMFSALS vs. PP16.

In all of the assays the AC-25 clone was stimulated with autologous B-LCL pulsed for one hour with a suboptimal concentration of agonist peptide PP16, 0.1 μg/ml for IFN-γ secretion and proliferation and 1.6 μg/ml for serine esterase release. Cells were then washed 10 μg/ml of antagonist peptide PG13 or control influenza matrix peptide PLKAEIAQRLEDV (Flu) was added. The control peptide binds DR1 but does not activate the HIV-specific clone. The left panel shows IFN-γ secretion in an intracellular cytokine staining assay with at least 50,000 events collected/condition. The middle panel shows proliferation by CFSE dilution. The right panel shows serine esterase release in the supernatant 4 hours after stimulation. All experiments were performed at least twice. * p < 0.05, PG13 vs. Flu peptide epitope.

The full-length peptide with a glycine to proline substitution at P11 was tested for antagonist activity in an intracellular cytokine staining assay for IFN-γ secretion. Potential antagonist peptides (10 μg/ml) were added to B-LCL pre-pulsed with a suboptimal concentration of PP16 (0.1 μg/ml). At least 50,000 events were collected per condition, and the null DR1-binding influenza peptide (Flu) was included as a negative control in the antagonist assay. The data represent the average of three independent experiments. * p < 0.05, PG13 vs. Flu peptide epitope.