Type I Interferon Receptor on NK Cells Negatively Regulates Interferon-γ Production

Abstract

NK cells are a key antiviral component of the innate immune response to HSV-2, particularly through their production of IFN-γ. It is still commonly thought that type I IFN activates NK cell function; however, rather than requiring the type I IFN receptor themselves, we have previously found that type I IFN activates NK cells through an indirect mechanism involving inflammatory monocytes and IL-18. Here, we further show that direct action of type I IFN on NK cells, rather than inducing IFN-γ, negatively regulates its production during HSV-2 infection and cytokine stimulation. During infection, IFN-γ is rapidly induced from NK cells at day 2 post-infection and then immediately downregulated at day 3 post-infection. We found that this downregulation of IFN-γ release was not due to a loss of NK cells at day 3 post-infection, but negatively regulated through IFN signaling on NK cells. Absence of IFNAR on NK cells led to a significantly increased level of IFN-γ compared to WT NK cells after HSV-2 infection in vitro. Further, priming of NK cells with type I IFN was able to suppress cytokine-induced IFN-γ production from both human and mouse NK cells. We found that this immunosuppression was not mediated by IL-10. Rather, we found that type I IFN induced a significant increase in Axl expression on human NK cells. Overall, our data suggests that type I IFN negatively regulates NK cell IFN-γ production through a direct mechanism in vitro and during HSV-2 infection.

Keywords: HSV; Human NK cells; IFN-γ; NK cells; type I IFN.

Figure 1

No significant difference in total number of vaginal NK cells between day 2 and day 3 post-infection. WT mice were infected with HSV-2 ivag. Vaginal tissue was isolated and processed at baseline through to d3 p.i. and stained for CD45, CD3, and NK1.1. NK cells were gated as CD45+, CD3– and NK1.1+. Representative flow plots are shown in (A) and total number of NK cells is shown in (B; n = 4). n.s., not significant; ^*p < 0.05.

Figure 2

Increased proportion of CD27-CD11b+ NK cells at day 2 post-infection. WT mice were infected with HSV-2 ivag and vaginal cells were isolated at baseline through to day 3 post-infection. Vaginal cells were stained for CD45, CD3, NK1.1 CD27, and CD11b. Cells were first gated on CD45+, CD3–, and NK1.1+ to determine the NK cell population. NK cells were then examined for CD27 and CD11b expression. Representative flow plots are shown in (A). Proportion of double-negative NK cells (CD27–CD11b–; Q4) is shown in (B; n = 4). Proportion of CD27+ (CD27+CD11b–; Q1) NK cells is shown in (C; n = 4). Proportion of double-positive (CD27+CD11b+; Q2) NK cells is shown in (D; n = 3). Proportion of CD11b+ (CD27–CD11b+; Q3) NK cells is shown in (E; n = 4). ^*p < 0.05; ^**p < 0.01; ^***p < 0.001.

Figure 3

Absence of type I IFN receptor on NK cells allows for a significant increase in IFN-γ production from NK cells. Rag2^−/−γc^−/− splenocytes were isolated and infected with HSV-2 MOI 3. Infected Rag2^−/−γc^−/− splenocytes were then co-cultured with isolated WT or Ifnar^−/− NK cells for 48 h. As controls, Rag2^−/−γc^−/− splenocytes, WT, and Ifnar^−/− NK cells were each infected in separate wells for 48 h. After 48 h of incubation, supernatants were collected and assayed for IFN-γ protein production (n = 3 repeated once with similar results). ^***p < 0.001.

Figure 4

Type I IFN pre-treatment suppresses cytokine-induced IFN-γ production from NK cells. NK cells were isolated from WT spleens and pre-treated with 100 U of IFN-β for 12 h. After the 12 h pre-treatment, NK cells were stimulated with the indicated doses of IL-15 (A; n = 3, repeated once with similar results), 1 ng/mL IL-12 (B; n = 3), or 25 ng/mL IL-18 (C; n = 3, repeated once with similar results) for 24 h. Supernatants were collected and assayed for IFN-γ production. WT splenocytes were pre-treated with increasing doses of IFN-β for 4 h. After 4 h of pre-treatment, splenocytes were stimulated with 250 ng/mL of IL-15 for 24 h. Supernatants were collected and assayed for IFN-γ production (D; n = 3, repeated once with similar results). NK cells were isolated from WT spleen and stimulated with either 200 ng/mL IL-15 alone, different doses of IFN-β alone, or IL-15 and IFN-β at the same time for 24 h. Supernatants were assayed for IFN-γ (E; n = 3). WT splenocytes were isolated and stimulated with either 100 U IFN-α, 100 U IFN-β, 250 ng/mL IL-15, or a combination of IFN-α, IFN-β, and IL-15 at the same time. After 24 h of stimulation, supernatants were collected and assayed for IFN-γ levels (F; n = 3). n.s., not significant, ^*p < 0.05, ^**p < 0.01, ^***p < 0.001.

Figure 5

Type I IFN pre-treatment suppresses IL-15-induced human NK cell IFN-γ production. CD56+ cells were isolated from PBMCs and pre-treated with 100 U of IFN-β for 12 h. After 12 h, NK cells were then stimulated with 250 ng/mL IL-15 for 24 h. Supernatants were collected and assayed for human IFN-γ levels (A; n = 4). PBMCs were isolated and pre-treated with 100 U IFN-β for 12 h. After 12 h of treatment, PBMCs were stimulated with 250 ng/mL IL-15 for 24 h. Supernatants were collected and assayed for IFN-γ levels (B; n = 3). After 18 h of stimulation, cells were stained with a fixable viability dye. The proportion of viable cells is graphically shown (C,D, n = 3). Original flow plots are also shown for NK cells and PBMCs (E,F, respectively) ^***p < 0.001.

Figure 6

Type I IFN does not significantly alter levels of IL-10. Isolated WT splenocyte NK cells (A; n = 3, repeated once with similar results) and splenocytes (B; n = 3) were stimulated with media or IFN-β for 24 h. Supernatants were assayed for IL-10 production. WT and Ifnar^−/− mice were infected with HSV-2 ivag. Vaginal lavages were collected at baseline through to d3 p.i. and examined for IL-10 levels (C; n = 3). n.s., not significant.

Figure 7

Significantly increased expression of Axl on NK cells after type I IFN stimulation. PBMC-isolated NK cells and PBMCs were stimulated with 100 U of IFN-β for 18 h. After stimulation, cells were stained with anti- CD56, CD3, Tyro3, Axl, and Mer. Cells were first gated as CD56+CD3– NK cells and then examined for expression of Tyro3 (A; n = 3), Axl (B; n = 3), or Mer (C; n = 3), ^*p < 0.05, ^**p < 0.01.