An inhibitor of HIV-1 protease modulates proteasome activity, antigen presentation, and T cell responses

Abstract

Inhibitors of the protease of HIV-1 have been used successfully for the treatment of HIV-1-infected patients and AIDS disease. We tested whether these protease inhibitory drugs exerted effects in addition to their antiviral activity. Here, we show in mice infected with lymphocytic choriomeningitis virus and treated with the HIV-1 protease inhibitor ritonavir a marked inhibition of antiviral cytotoxic T lymphocyte (CTL) activity and impaired major histocompatibility complex class I-restricted epitope presentation in the absence of direct effects on lymphocytic choriomeningitis virus replication. A potential molecular target was found: ritonavir selectively inhibited the chymotrypsin-like activity of the 20S proteasome. In view of the possible role of T cell-mediated immunopathology in AIDS pathogenesis, the two mechanisms of action (i.e., reduction of HIV replication and impairment of CTL responses) may complement each other beneficially. Thus, the surprising ability of ritonavir to block the presentation of antigen to CTLs may possibly contribute to therapy of HIV infections but potentially also to the therapy of virally induced immunopathology, autoimmune diseases, and transplantation reactions.

Figure 1

Effect of ritonavir on CTL activity in vivo. (A and B) Inhibition of footpad swelling after LCMV infection: (A) C57BL/6 mice were infected in the footpad on day 0 (300 pfu of LCMV-WE), and footpad swelling was measured over time. Ritonavir (1.25 mg/mouse/day) or placebo control (same volume of PBS and 10% alcohol) were administered i.p. from day 0. In this and subsequent experiments, readings are from two footpads from 2–3 mice per group and are representative of 3–4 separate experiments. (B) Footpad swelling induced by LCMV infection was measured on day 8 in mice treated with varying doses of ritonavir. (C and D) Inhibition of ex vivo CTL activity. Ritonavir-treated (1.25 mg/mouse/day) or placebo-treated C57BL/6 mice were infected with 200 pfu of LCMV-WE i.v. At 8 days postchallenge, splenocytes were tested for lysis of EL4 cells (H-2^b) prepulsed with 500 nM peptide GP_33–41 (KAVYNFATC; C) or NP_396–407 (FQPQNGQFI; D). E/T, effector to target ratio. (E) Inhibition of CD8⁺ T cell expansion in vivo. Mice were infected with 200 pfu of LCMV-WE i.v. and treated with ritonavir or placebo as above. Splenocytes were stained with fluorescein isothiocyanate-conjugated anti-CD8 (PharMingen), and the percentage of positive cells was determined by flow cytometry. (F) Effect of ritonavir on LCMV viral load. Mice were infected with 200 pfu of LCMV i.v. and treated with ritonavir or placebo as in C. Spleen virus titer was determined on day 8.

Figure 2

(A) Effect of ritonavir on MHC class I-restricted presentation in vitro as determined by CTL lysis of LCMV-WE-infected targets. (B) GP_33–41-specific CTLs were prepared by secondary restimulation in vitro from spleens of LCMV-primed mice. LCMV-WE infection of MC57 fibroblasts (H-2^b) was performed at a multiplicity of infection of 0.04, ritonavir was added at a concentration of 5 μg/ml, and the cells were cultured for 36 h. Effect of ritonavir on lysis of LCMV–GP-transfected targets. Targets were LCMV–GP-transfected MC57 cells as described in ref. . Effectors were splenocytes from C57BL/6 mice at 8 days after 200 pfu LCMV-WE infection. Targets were treated with ritonavir or left untreated as in A.

Figure 3

Effect of ritonavir on MHC class I presentation to human T cell clones. (A) M113 melanoma cells were cultured for 24 h with various concentrations of ritonavir, resuspended, and fixed in paraformaldehyde. The same concentrations of indinavir were used as a control (placebo). Tumor necrosis factor α production by HLA-A2-restricted T cells specific for the MART-1 epitope 26–35 (M77-84; ref. 18) was measured in the supernatant and is expressed as the percentage of maximal release obtained with untreated stimulator M113 cells. (B) M113 cells were cultured for 24 h in ritonavir (0.1–5 μg/ml) and then used as stimulators in a proliferation assay with M77-84 T cells as responders. Proliferation was measured over 48 h with thymidine added at 1 μCi/well for the last 18 h.

Figure 4

Chymotrypsin-like activity of the proteasome is inhibited by ritonavir. Hydrolysis of fluorogenic substrates (A) 100 μM Suc-LLVY-MCA, (B) 100 μM Z-GGL-MCA, and (C) 400 μM Bz-VGR-MCA by isolated 20S proteasomes from murine B8 fibroblast cells were plotted vs. indicated concentrations of ritonavir and the protea- some inhibitor LLnL. The results of 1-h digests with 500 ng of 20S proteasome in a final volume 100 μl are shown (29). Values were in a linear range of detection and are the means of triplicates with SE of <3% for all data points. (D) Comparison of four HIV-1 protease inhibitors. Hydrolysis of Suc-LLVY-MCA by murine 20S proteasomes was tested as described in B under identical conditions, except that a different batch of purified proteasomes was used. (E) Comparison of four protease inhibitors: effect on CTL lysis ex vivo. C57BL/6 mice (2–3 animals per group) that had been primed previously with LCMV-WE were rechallenged (2 × 10⁶ pfu of LCMV-WE i.v.) and treated with either placebo or 4 mg of ritonavir, saquinavir, indinavir, or nelfinavir orally once daily for 4 days. Splenocytes then were tested directly ex vivo for lysis of NP_396–404-pulsed EL4 targets. Lysis of unpulsed targets was <3%. (F) Ritonavir does not affect the presentation of the proteasome-independent influenza NP epitope 50–57 (30). LK^d cells were infected with a vaccinia recombinant expressing a fusion protein of ubiquitin and influenza NP (UbMNP). Ritonavir treatment was for 36 h at 10 μM and lactacystin treatment was for 30 min at 100 μM before cells were infected with 10 pfu/cell of recombinant vaccinia virus. Labeling with 100 μCi ⁵¹Cr occurred for 90 min during infection; next, recombinant proteins were allowed to be expressed for 4 h at 37°C before usage as target cells in a 4-h cytolytic assay by using H-2K^k/NP epitope 50–57-specific polyclonal CTLs as effectors. (G) Accumulation of ubiquitin conjugates in ritonavir-treated cells. B8 murine fibroblast cells were treated for 3 h as indicated, and cellular lysates were analyzed by Western blots probed with an ubiquitin conjugate-reactive antiserum (Dako). The densitometric evaluation of enhanced chemiluminescence-exposed films was normalized to the amount of total protein on Ponceau-stained blots. (H) Inhibition of IκBα degradation in ritonavir-treated cells. NFS5.3 murine B cells were starved for 30 min and metabolically labeled with [³⁵S]-methionine/[³⁵S]-cysteine for 1 h in the presence or absence of inhibitors. Cells were treated with a final concentration of 45 μg/ml lipopolysaccharide before chasing for indicated time periods. Immunoprecipitation was performed with an IκBα-specific polyclonal antibody (Santa Cruz Biotechnology). IkBα bands were quantified on a BAS1500 Imager (Fuji) and normalized to proteasome subunit MC3 precipitated from the same lysate.