Natural killer cells act as rheostats modulating antiviral T cells

Abstract

Antiviral T cells are thought to regulate whether hepatitis C virus (HCV) and human immunodeficiency virus (HIV) infections result in viral control, asymptomatic persistence or severe disease, although the reasons for these different outcomes remain unclear. Recent genetic evidence, however, has indicated a correlation between certain natural killer (NK)-cell receptors and progression of both HIV and HCV infection, implying that NK cells have a role in these T-cell-associated diseases. Although direct NK-cell-mediated lysis of virus-infected cells may contribute to antiviral defence during some virus infections--especially murine cytomegalovirus (MCMV) infections in mice and perhaps HIV in humans--NK cells have also been suspected of having immunoregulatory functions. For instance, NK cells may indirectly regulate T-cell responses by lysing MCMV-infected antigen-presenting cells. In contrast to MCMV, lymphocytic choriomeningitis virus (LCMV) infection in mice seems to be resistant to any direct antiviral effects of NK cells. Here we examine the roles of NK cells in regulating T-cell-dependent viral persistence and immunopathology in mice infected with LCMV, an established model for HIV and HCV infections in humans. We describe a three-way interaction, whereby activated NK cells cytolytically eliminate activated CD4 T cells that affect CD8 T-cell function and exhaustion. At high virus doses, NK cells prevented fatal pathology while enabling T-cell exhaustion and viral persistence, but at medium doses NK cells paradoxically facilitated lethal T-cell-mediated pathology. Thus, NK cells can act as rheostats, regulating CD4 T-cell-mediated support for the antiviral CD8 T cells that control viral pathogenesis and persistence.

Figure 1. NK cells influence T-cell-dependent pathology and viral persistence during LCMV infection.

a–d, C57BL/6 mice treated with IgG2a (Control) or anti-NK1.1 (ΔNK) were infected with low (5 × 10⁴ p.f.u.), medium (2 × 10⁵ p.f.u.), or high (2 × 10⁶ p.f.u.) doses of LCMV. a, Weight loss (mean ± s.e.m.) during infection (n = 3–43 per group per day), *P < 0.05, **P < 0.01. b, Haematoxylin & eosin (H&E) staining of lung (×400) at day 15 p.i. c, Survival after high- or medium-dose infection. d, Viral titres (mean ± s.e.m.) after low (day 7, n = 3 per group), medium (day 15, n = 9–15 per group), or high (day 8, n = 3 per group) dose infection. Dotted line represents limit of detection. PowerPoint slide

Figure 2. LCMV-specific T-cell responses enhanced in NK-cell-depleted mice.

a–c, Number (mean ± s.e.m.) of LCMV-specific T cells measured by IFN-γ expression (a, c) or tetramer-binding (b) at various days p.i. (a, c) (medium dose, n = 3 per group per day) or at day 5 p.i. (b) (various doses, n = 3–11 per group). d, Co-production of TNF and IL-2 by gated IFN-γ⁺ CD4 T cells (day 8 p.i., medium dose) after in vitro stimulation with GP61 peptide. e, Viral titres in liver (n = 3–11 per group per day, medium dose). f, CFSE dilution by donor (Thy1.1⁺) CD4 T cells in uninfected (Uninf.) and infected (medium dose, day 6 p.i.) control and ΔNK Thy1.2⁺ host mice. Control versus ΔNK mice, *P < 0.05, **P < 0.01. PowerPoint slide

Figure 3. Role of CD4 T cells in NK suppression of antiviral CD8 T cells, viral control and immunopathology.

a–e, Prior to medium- (a, b) or high-dose (c–e) infection, mice were injected with isotype (Con, control), anti-NK1.1 (ΔNK), anti-CD4 (ΔCD4), or both (ΔNKΔCD4). a, b, Viral titres in spleen (a) and number of cytokine-producing NP_396–404-specific CD8 T cells (n = 5 per group, ± s.e.m.) in spleen at day 7 of medium-dose infection (b). c, d, H&E staining of lung (×100) (c) and survival at day 12 of high-dose infection (n = 3–5 per group) (d). PowerPoint slide

Figure 4. NK cells rapidly eliminate activated CD4 T cells.

a–f, In vivo cytotoxicity assays were performed as described in Methods using donor cells from NK-cell-depleted wild-type (Ly5.1⁺) (a–f) or Fas^lpr (lpr) (d) mice 4 days p.i. with medium-dose LCMV (a–e) or other viruses (f). Recovery of donor T cells was examined 5 h after transfer into control or NK-depleted (ΔNK) Ly5.2⁺ mice (a, n = 21–28 per group; b–f, n = 3–5 per group). Host mice were uninfected or inoculated with medium-dose LCMV (day 3 p.i.) (a, c, d, f), various doses of LCMV (day 3 p.i.) (b), or PV (day 2), MHV (day 3) and polyI:C (pI:C) (day 1) (e). a, b, d–f, Lysis (mean ± s.e.m.) of donor T cells in infected relative to uninfected recipient mice; control versus ΔNK mice, *P < 0.05. WT, wild type. c, Annexin V reactivity of donor (Ly5.1⁺) and host (Ly5.1⁻) CD4 T cells. g, Proposed model connecting NK-cell killing of CD4 T cells to CD8 T cells and infection outcome in the presence of NK cells. PowerPoint slide