Processing and presentation of exogenous HLA class I peptides by dendritic cells from human immunodeficiency virus type 1-infected persons

Abstract

Dendritic cells (DCs) loaded with viral peptides are a potential form of immunotherapy of human immunodeficiency virus type 1 (HIV-1) infection. We show that DCs derived from blood monocytes of subjects with chronic HIV-1 infection on combination antiretroviral drug therapy have increases in expression of HLA, T-cell coreceptor, and T-cell activation molecules in response to the DC maturation factor CD40L comparable to those from uninfected persons. Mature DCs (mDCs) loaded with HLA A*0201-restricted viral peptides of the optimal length (9-mer) were more efficient at activating antiviral CD8(+) T cells than were immature DCs or peptide alone. Optimal presentation of these exogenous peptides required uptake and vesicular trafficking and was comparable in DCs derived from HIV-1-infected and uninfected persons. Furthermore, DCs from HIV-1-infected and uninfected persons had similar capacities to process viral peptides with C-terminal and N-terminal extensions through their proteasomal and cytosolic pathways, respectively. We conclude that DCs derived from HIV-1-infected persons have similar abilities to process exogenous peptides for presentation to CD8(+) T cells as those from uninfected persons. This conclusion supports the use of DCs loaded with synthetic peptides in immunotherapy of HIV-1 infection.

FIG. 1.

Levels of expression of HLA DR, HLA class I (ABC), CD83, and CD86 in iDCs and mDCs derived from CD14⁺ monocytes cultured in IL-4 and granulocyte-monocyte colony-stimulating factor for 10 non-HIV-infected [HIV(−)] and 8 HIV-1-infected [HIV(+)] persons. Values are MFI ± SE. Numbers above bars show percent positive DCs ± SE. For HIV(−) persons, P values for comparison of MFI for iDCs and mDCs were as follows: 0.002 for HLA DR, 0.01 for HLA ABC, 0.001 for CD83, and 0.01 for CD86. For HIV(+) persons, these P values were 0.02 for HLA DR, 0.04 for HLA ABC, 0.02 for CD83, and 0.02 for CD86. P > 0.05 for comparison of expression of each of these markers in HIV(−) versus HIV(+) persons.

FIG. 2.

Comparison of autologous iDCs and mDCs for induction of IFN-γ in CD8⁺ T cells from HLA A*0201-positive non-HIV-1-infected and HIV-1-infected subjects. The DCs from three non-HIV-1-infected donors were loaded with FLU M1_58-66 (A) or EBV BMLF1_280-288 (B) and were used to induce a dose-dependent IFN-γ response in highly enriched autologous CD8⁺ T cells. P ≤ 0.02 for mDCs loaded with FLU M1₅₈₋₆₆ peptide or EBV BMLF1_280−288 compared to no DCs (none) or iDCs. The iDCs and mDCs from six HIV-1-infected subjects receiving combination antiretroviral therapy (median CD4⁺ T-cell count, 642/μl [range, 226 to 987/μl]; median plasma HIV RNA load, <50 copies/ml [range, <50 to 2,501 copies/ml]) were loaded with HIV-1 p17_77-85 (C) or HIV-1 RT_476-484 (D) and were used to stimulate enriched autologous CD8⁺ T cells. P ≤ 0.02 for mDCs compared to no DCs or iDCs loaded with either the HIV-1 p17_77-85 peptide or the HIV-1 RT_476-484 peptide.

FIG. 3.

Production of IFN-γ by CD8⁺ T-cell lines specific for EBV BMLF1_280-288 or CMV strain AD169 pp65_495-503 stimulated by allogeneic DCs derived from seven uninfected persons [HIV(−) DCs] or eight HLA A*0201-positive HIV-1-infected subjects (median CD4⁺ T-cell count, 530/μl [range, 226 to 987/μl]; median plasma HIV RNA load, <50 copies/ml [range, <50 to 2,501 copies/ml]) [HIV(+) DCs]. (A) For the T-cell response to stimulation with EBV BMLF1_280-288 alone versus peptide-loaded iDCs or mDCs, P ≤ 0.01 for both HIV(−) and HIV(+) persons. For stimulation with peptide-loaded iDCs versus mDCs, P = 0.001 for HIV(−) persons and P = 0.005 for HIV(+) persons. For comparison of T-cell responses for HIV(−) versus HIV(+) persons, P > 0.05 (not significant). (B) For the T-cell response to stimulation with CMV pp65_495-503 alone versus peptide-loaded HIV(−) iDCs or mDCs, P = 0.05 or P = 0.01, respectively; for the response to stimulation with CMV pp65_495-503 alone versus peptide-loaded HIV(+) iDCs or mDCs, P = 0.02 and P = 0.0002, respectively; for the response to stimulation with peptide-loaded iDCs versus mDCs, P = 0.02 for HIV(−) and P = 0.0003 for HIV(+) persons; for comparison of T-cell responses for HIV(−) versus HIV(+) persons, P > 0.05.

FIG. 4.

Effects of metabolic inhibitors on processing of exogenous 9-mer peptides by autologous iDCs and mDCs for IFN-γ production in HLA A*0201-positive non-HIV-1-infected and infected persons. The APCs from uninfected persons were pretreated with metabolic inhibitors and loaded with FLU M1_58-66 (A) or EBV BMLF1_280-288 (B). Lactacystin did not inhibit IFN-γ production stimulated by FLU M1_58-66- or EBV BMLF1_280-288-expressing APCs. In contrast, brefeldin A and cycloheximide inhibited as much as 100% of IFN-γ production stimulated by FLU M1_58-66- or EBV BMLF1_280-288-expressing APCs. APCs from an HLA A*0201-positive HIV-1-infected subject (CD4⁺ T-cell count, 627/μl; plasma HIV RNA load, 6,484 copies/ml) were pretreated with inhibitors and loaded with HIV-1 p17_77-85 (C) or HIV-1 RT_476-484 (D). Lactacystin did not inhibit IFN-γ production stimulated by either peptide. In contrast, brefeldin A and cycloheximide inhibited as much as 65% of IFN-γ production stimulated by HIV-1 RT_476-484- or HIV-1 p17_77-85-expressing APCs. These data are representative of four experiments with different non-HIV-1-infected and HIV-1-infected persons.

FIG. 5.

Effects of lactacystin, brefeldin A, and cycloheximide on lysis of EBV BMLF1_280-288- or HIV-1 RT_476-484-stimulated iDCs and mDCs by CTLs from HLA A*0201-positive non-HIV-1-infected or HIV-1-infected subjects. The iDCs or mDCs were loaded with peptides in the presence or absence of lactacystin (10 μM), brefeldin A (5 μg/ml), or cycloheximide (1 mg/ml) and used as targets in a CTL assay. The effector cells were CD8⁺ T cells from three HLA-A*0201-positive non-HIV-1-infected donors (A) and three HLA-A*0201-positive HIV-1-infected donors (B). The recognition of EBV BMLF1_280-288-expressing DCs by CTLs at an E/T ratio of 20 was sensitive to the effect of brefeldin A (P = 0.01 for iDCs; P = 0.04 for mDCs) or cycloheximide (P = 0.001 for iDCs; P = 0.0001 for mDCs) but not lactacystin (P > 0.05) as judged by comparison to non-inhibitor-treated DCs from three non-HIV-1-infected donors. The recognition of HIV-1 RT_476-484-expressing DCs by CTLs at an E/T ratio of 20 was sensitive to the effect of brefeldin A (P = 0.03 for iDCs; P = 0.01 for mDCs) or cycloheximide (P = 0.01 for iDCs; P = 0.001 for mDCs) but not lactacystin (P > 0.05) as judged by comparison to non-inhibitor-treated DCs from three HIV-1-infected donors. The same patterns of CTL lysis were detected by using an E/T ratio of 40 or 10 (data not shown).

FIG. 6.

Effects of metabolic inhibitors on the HLA class I processing pathway of the N and C termini of the EBV BMLF1_280-288 peptide. The iDCs (A) and mDCs (B) from an HLA A*0201-positive non-HIV-1-infected donor were treated with the indicated concentrations of lactacystin, brefeldin A, or cycloheximide for 30 min, washed, and then loaded with the 9-mer EBV BMLF1_280-288 peptide alone (G-L: GLCTLVAML) or with its N-terminal flanking sequence (AIQNAGLCTLVAML) or C-terminal flanking sequence (GLCTLVAMLEETIFW). CD8⁺ T cells were then cultured with DCs in the 18-h Elispot assay. The data show that lactacystin blocked the presentation of the C-terminal extension but not of the N-terminal extension and 9-mer EBV peptides to CD8⁺ T cells by iDCs and mDCs in a dose-dependent fashion. In contrast, brefeldin A and cycloheximide blocked the presentation of all three peptides (C-terminal extension, N-terminal extension, and 9-mer peptides). These data are representative of three experiments with different subjects.

FIG. 7.

Effects of metabolic inhibitors on the HLA class I processing pathway of the N and C termini of the HIV-1 p17_77-85 peptide in DCs from an HLA A*0201-positive HIV-1-infected donor (the same donor whose APCs were used for the experiment for which results are shown in Fig. 4C). The iDCs (A) and mDCs (B) were treated with the indicated concentrations of lactacystin, brefeldin A, or cycloheximide for 30 min, washed, and then loaded with the 9-mer HIV-1 p17 peptide alone (S-L: SLYNTVATL) or with its N-terminal flanking sequence (GSEELRSLYNTVATL) or C-terminal flanking sequence (SLYNTVATLYCVHQR). CD8⁺ T cells were then cultured with DCs in the 18-h Elispot assay. The data show that lactacystin blocked the presentation of the C-terminal extension but not of the N-terminal extension and 9-mer HIV-1 peptides to CD8⁺ T cells by iDCs and mDCs in a dose-dependent fashion. In contrast, brefeldin A and cycloheximide blocked the presentation of all three peptides (C-terminal extension, N-terminal extension, and 9-mer peptides). These data are representative of three experiments with different subjects.