CCR5 deficiency drives enhanced natural killer cell trafficking to and activation within the liver in murine T cell-mediated hepatitis

Abstract

Natural killer (NK) cells are innate immune cells that are enriched in the liver, but the processes underlying NK cell trafficking to the liver and cellular activation within the liver of patients with T cell-mediated liver diseases remain poorly defined. Concanavalin A (Con A) hepatitis is a murine model mimicking many aspects of human T cell-mediated liver diseases. Here we demonstrate that severe hepatitis in CCR5-deficient (KO) mice is associated with increased hepatic NK cell recruitment driven by enhanced hepatic production of CCL5 acting via CCR1 and by enhanced hepatic NK cell activation relative to that observed in wild-type mice after Con A administration. Furthermore, NK cell depletion ameliorated severe hepatitis in CCR5 KO mice but did not alter hepatitis in wild-type mice after Con A treatment. We propose that in the setting of CCR5 deficiency NK cells assume a profound effector role in Con A hepatitis via enhanced CCL5-CCR1 driven hepatic recruitment in addition to augmented cytokine-driven NK cell activation to produce interferon-gamma. These results highlight the potential profound impact of altered chemokine receptor expression on the innate immune response in the setting of T cell-mediated hepatitis.

Figure 1

A: Time course of NK cell recruitment into the liver of WT (□; n = 5) and CCR5 KO (▪; n = 4) mice during Con A-induced hepatitis; **P < 0.01 versus all other groups. B: Representative FACS histogram depicting increased NK cell influx into the liver of CCR5 KO mice compared with WT mice at 8 hours after Con A administration. C: The effect of anti-asialo-GM1 Ab (n = 6) or control Ab (n = 5) treatment on ALT levels in WT and CCR5 KO mice 8 hours after Con A treatment; **P < 0.01 versus control Ab-treated WT mice; ^##P < 0.01 versus anti-asialo Ab-treated CCR5 KO mice. D: Representative H&E staining of liver sections showing widespread/confluent hepatocellular necrosis (white arrows) and inflammatory cell infiltrates throughout the liver in Con A-treated CCR5 KO mice relative to patchy hepatocellular necrosis and mild inflammatory cell infiltrates in CCR5 KO mice pretreated with asialo-GM1 Ab all at 8 hours after Con A treatment. E: Representative FACS histogram depicting NK cell depletion after anti-asialo-GM1 mAb treatment in naïve CCR5 KO mice.

Figure 2

Correlation of severe hepatitis in CCR5 KO mice with enhanced hepatic IFN-γ production. A: Enzyme-linked immunosorbent assay determination of hepatic levels of IFN-γ in WT (□; n = 5) and CCR5 KO mice (▪; n = 4) during Con A-induced hepatitis. ^#P < 0.05 versus naïve WT; *P < 0.05 versus Con A-treated WT (90 minutes) and naïve groups; **P < 0.01 versus all WT groups. B: Number of IFN-γ-producing NK cells in the liver of WT (□; n = 6) and CCR5 KO (▪; n = 5) mice after Con A administration. *P < 0.05 versus Con-treated WT (90 minutes) and naïve groups; ^#P < 0.05 versus Con-treated WT (90 minutes) and naïve groups; **P < 0.01 versus all groups. C: Representative FACS dot plot demonstrating increased NK cell intracellular IFN-γ in the liver of CCR5 KO mice after 8 hours of Con A treatment relative to WT mice. D: Number of IFN-γ-producing CD4⁺ T cells in the liver of WT (□; n = 5) and CCR5 KO (▪; n = 4) mice 8 hours after Con A administration. *P < 0.05 versus naïve WT and naïve CCR5 KO mice; **P < 0.01 versus all other groups. E: Representative FACS histogram demonstrating reduced CD4⁺ T cell recruitment in the liver of CCR5 KO mice after anti-asialo-GM1 Ab treatment relative to CCR5 KO given control antibody.

Figure 3

A: Plasma ALT levels 8 hours after Con A administration in CCR5 KO mice pretreated with anti-IFN-γ mAb (▪; n = 4) relative to CCR5 KO mice given control Ab (▪; n = 4). ***P < 0.001 versus control Ab-treated CCR5 KO mice. B: Representative H&E staining of liver sections depicting widespread/confluent hepatocellular necrosis (arrows) and inflammatory cell infiltrates throughout the liver in Con A-treated CCR5 KO mice (arrows) relative to patchy hepatocellular necrosis and mild inflammatory cell infiltrates in CCR5 KO mice pretreated with anti-IFN-γ mAb all at 8 hours after Con A treatment. C: Effect of anti-IL-4 mAb treatment on hepatic IFN-γ levels in CCR5 KO 8 hours after Con A administration (▪; n = 4) relative to CCR5 KO mice given control Ab (▪; n = 4). *P < 0.05 versus control Ab-treated CCR5 KO mice. D: Number of NK cells in the liver 8 hours after Con A administration in CCR5 KO mice pretreated with anti-IL-4 mAb (▪; n = 4) relative to CCR5 KO mice given control Ab (▪; n = 4). *P < 0.05 versus control Ab-treated CCR5 KO mice. E: Representative FACS histogram depicting reduced NK cell recruitment in the liver of Con A-treated CCR5 KO mice at the 8 hour time point following pretreatment with anti-IL-4 mAb relative to Con A-treated CCR5 KO mice given control Ab. F: Number of IFN-γ-producing NK cells in the liver 8 hours after Con A administration in CCR5 KO mice pretreated with anti-IL-4 mAb (▪; n = 4) relative to CCR5 KO mice given control Ab (▪; n = 4). *P < 0.05 versus control Ab-treated CCR5 KO mice. G: Number of IFN-γ-producing NK cells in the liver 90 minutes after Con A administration in CCR5 KO mice pretreated with anti-IL-4 mAb (▪; n = 4) relative to CCR5 KO mice given control Ab (▪; n = 4). *P < 0.05 versus control Ab-treated CCR5 KO.

Figure 4

A: Number of NK cells in the liver 8 hours after Con A administration in CCR5 KO mice pretreated with anti-IFN-γ mAb (▪; n = 4) relative to CCR5 KO mice given control Ab (▪; n = 4). *P < 0.05 versus control Ab-treated CCR5 KO mice. B: Number of IFN-γ-producing NK cells in the liver 8 hours after Con A administration in CCR5 KO mice pretreated with anti-IFN-γ mAb (▪; n = 4) relative to CCR5 KO mice given control Ab (▪; n = 4). *P < 0.05 versus control Ab-treated CCR5 KO mice.

Figure 5

CCL5 promotes NK cell recruitment into the liver of CCR5 KO mice. A: Enzyme-linked immunosorbent assay determination of hepatic levels of CCL5 in WT (□; n = 5) and CCR5 KO mice (▪; n = 4) 8 hours after Con A-induced hepatitis. *P < 0.05 versus all other groups. B: Hepatic levels of CCL3 in WT (□; n = 5) and CCR5 KO mice (▪; n = 4) as determined by enzyme-linked immunosorbent assay at 8 hours after Con A-induced hepatitis. ^#P < 0.05 versus their respective naive groups. C: Immunohistochemical localization of CCL5 in paraffin-embedded liver sections from WT and CCR5 KO mice 8 hours after Con A administration (magnification, ×40). D: Number of NK cells in the liver 8 hours after Con A administration in CCR5 KO mice pretreated with anti-CCL5 mAb (▪; n = 4) relative to CCR5 KO mice given control Ab (▪; n = 4). *P < 0.05 versus control Ab-treated CCR5 KO mice. E: Number of IFN-γ-producing NK cells in the liver 8 hours after Con A administration in CCR5 KO mice pretreated with anti-CCL5 mAb (▪; n = 4) relative to CCR5 KO mice given control Ab (▪; n = 4). *P < 0.05 versus control Ab-treated CCR5 KO mice.

Figure 6

Schematic summary of events of how NK cells may promote severe hepatitis in CCR5-deficient mice following Con A administration. IL-4, in conjunction with IFN-γ, produced by activated hepatic NKT cells directly promotes early (ie, 90 minutes post-Con A) NK cell activation after Con A treatment (as reflected by increased NK cell IFN-γ production) and then indirectly mediate later (ie, 8 hours post-Con A) hepatic NK cell recruitment to and activation within the liver mediated by increased hepatic levels of hepatocyte-derived CCL5 acting via CCR1 expressed on NK cells. Activation of antigen-presenting cells during the development of Con A hepatitis may result in the increased production and release of IL-12 and IL-18 within the liver, and both of these cytokines may subsequently promote further NK cell activation and IFN-γ production within the liver. It is the combination of these events that drives the development of severe liver injury in CCR5-deficient mice following Con A administration.