NK cell-intrinsic FcεRIγ limits CD8+ T-cell expansion and thereby turns an acute into a chronic viral infection

Abstract

During viral infection, tight regulation of CD8+ T-cell functions determines the outcome of the disease. Recently, others and we determined that the natural killer (NK) cells kill hyperproliferative CD8+ T cells in the context of viral infection, but molecules that are involved in shaping the regulatory capability of NK cells remain virtually unknown. Here we used mice lacking the Fc-receptor common gamma chain (FcRγ, FcεRIγ, Fcer1g-/- mice) to determine the role of Fc-receptor and NK-receptor signaling in the process of CD8+ T-cell regulation. We found that the lack of FcRγ on NK cells limits their ability to restrain virus-specific CD8+ T cells and that the lack of FcRγ in Fcer1g-/- mice leads to enhanced CD8+ T-cell responses and rapid control of the chronic docile strain of the lymphocytic choriomeningitis virus (LCMV). Mechanistically, FcRγ stabilized the expression of NKp46 but not that of other killer cell-activating receptors on NK cells. Although FcRγ did not influence the development or activation of NK cell during LCMV infection, it specifically limited their ability to modulate CD8+ T-cell functions. In conclusion, we determined that FcRγ plays an important role in regulating CD8+ T-cell functions during chronic LCMV infection.

Fig 1. FcεRIγ is an important contributor to NK cell–mediated regulation of virus-specific CD8⁺ T cells.

(A) Surface and intracellular FcεRIγ expression by natural killer (NK) cells from spleens of Fcer1g^+/– and Fcer1g^–/– mice that had been infected intravenously (i.v.) with 2 × 10⁴ plaque-forming units (PFU) of the Docile strain of the lymphocytic choriomeningitis virus (LCMV-Docile). Cells were analyzed 48 hours after infection (n = 4). (B) The histogram shown depicts the proliferation capacity of P14 cells in spleen represented as CFSE dilution in Fcer1g^+/– and Fcer1g^–/– mice. 10⁷ CFSE-labelled splenocytes from P14 x CD45.1 mice were adoptively transferred into Fcer1g^+/– and Fcer1g^–/– mice. After one day mice were i.v. infected with 200 PFU of LCMV-WE strain and the CFSE dilution was assessed in spleen at day 4 (n = 4). (C) Schematic of the experimental setup. (D) Splenocytes (10⁴) from WT P14 or Ifnar^–/– x P14 mice were adoptively transferred into Fcer1g^+/– or Fcer1g^–/– mice one day earlier, then the mice were i.v. infected with 2 × 10⁴ PFU of LCMV-Docile. In the upper panel, shown are representative histograms for the frequencies of WT P14 or Ifnar^–/– P14 cells at day 6 post-infection. In the lower panel, the bar graph represents total number of transferred WT P14 or Ifnar^–/– P14 cells in the blood at the indicated days after infection (n = 4). (E) 10⁴ splenocytes from P14 × Ifnar^–/– mice were transferred into Fcer1g^+/– or Fcer1g^–/– mice that had been treated with isotype antibody or anti NK1.1 antibody at day 3 and 1 before i.v infection with 2 × 10⁴ PFU of LCMV-Docile. The graph shows the total number of transferred P14 cells in blood at day 15 post-infection (n = 3–4). Data are shown as mean ± SEM. Significant differences between the two groups were detected by unpaired two-tailed t-tests and are indicated as follows: ns, not significant; * p<0.05; ** p<0. 01; *** p<0.001; **** p<0.0001.

Fig 2. FcεRIγ has no impact on NK cell activation but does affect the expression of NKp46.

Fcer1g^+/– and Fcer1g^–/– mice were left untreated or were infected i.v. with 2 x 10⁴ PFU of LCMV-Docile. Mice were put to death on day 2 (d2) after infection and NK cells in the spleen were analyzed for various markers by flow cytometry. (A) Representative fluorescence-activated cell sorting (FACS) plots for the frequencies of NK cells (left panel). The bar graph in right panel shows total number of NK cells in naïve and LCMV-Docile infected mice (n = 4). (B) Frequency of various markers in intracellularly stained NK cells from naïve and LCMV infected mice (n = 3–5). Data are pooled from two independent experiments. (C) Surface expression of TRAIL on splenic NK cells from naïve and LCMV-infected mice (n = 3–4). (D) Intracellular staining of PKC-θ on splenic NK cells from naïve and LCMV-infected mice (n = 3–4). (E) Representative histograms for various cell surface markers on NK cells from naïve and LCMV-infected mice (n = 3–4). Experimental data are representative of three independent experiments. Data are shown as mean ± SEM. Significant differences between the two groups were detected by unpaired two-tailed t-tests and are indicated as follows: * p<0.05; ** p<0. 01.

Fig 3. FcεRIγ stabilizes NKp46 on NK cells.

(A) Surface analysis of various markers on splenic NK cells of naïve Fcer1g^+/–, Fcer1g^–/–, and Jh^–/– mice (n = 4). Shown histogram is a representative of three experiments. (B) Bar graph showing the mRNA expression of NCR1, FcεRIγ, and CD3ζ as determined by RT-PCR from purified NK cells isolated from naïve spleens of Fcer1g^+/– and Fcer1g^–/– mice (n = 3). (C) Intracellular expression of CD3ζ on splenic NK cells from naïve or i.v infected Fcer1g^+/– and Fcer1g^–/– mice with 2 x 10⁴ PFU of LCMV-Docile for 36 hours (n = 3–4). (D) Histogram depicting surface and intracellular staining of NKp46 on splenic NK cells from naïve or i.v infected Fcer1g^+/– and Fcer1g^–/– mice with 2 x 10⁴ PFU of LCMV-Docile for 36h (n = 4). The histograms are representative of two independent experiments. (E) Representative histogram for surface NKp46 expression on splenic NK cells from Fcer1g^+/– and Fcer1g^–/– mice treated ex-vivo with 20μg/ml MG-132 for 48 hours as indicated (n = 3) (left panel). In the right panel, the shown is median fluorescence intensity (MFI) for the same experiment (n = 3). Data are shown as mean ± SEM. Significant differences between the two groups were detected by unpaired two-tailed t-tests and are indicated as follows: ns, not significant; * p<0.05; ** p<0. 01; *** p<0.001.

Fig 4. FcεRIγ curtails CD8⁺ T-cell functions during chronic LCMV infection.

Fcer1g^+/+ and Fcer1g^–/– mice were infected i.v. with 2 × 10⁴ PFU of LCMV-Docile and were bled at various time points or put to death on day 8 after infection. (A) The left representative FACS plot showing the frequency of glycoprotein (GP)33-Tet⁺ CD8⁺ T cells of total leukocytes in blood 8 days after infection. The right panel shows graphs of the kinetics for the frequency and number of CD8⁺ T cells (middle; n = 3–12) and virus-specific GP33-Tet⁺ CD8⁺ T cells in blood at the indicated time points (right; n = 3–12). Data are pooled from 3 independent experiments. (B) Frequency and total number of CD8⁺ T cells from spleen and liver on day 8 after infection (n = 4). (C) Representative FACS plots and graphs showing the frequency and total number of GP33-Tet⁺ CD8⁺ T cells in spleens and livers on day 8 after infection (n = 4). (D) Representative histogram showing the expression of PD1 and KLRG1 on GP33-Tet⁺ CD8⁺ T cells in spleens and livers on day 8 after infection (n = 4) (E) FACS plots (left panel) and graphs (right panel) depict the percentage and total numbers of CD8⁺ T cells producing interferon (IFN)-γ and tumor necrosis factor (TNF)-α in spleens and livers on day 8 after infection. The cells were stimulated in-vitro for 5 hours in the presence or absence of GP33 peptide (n = 4). (F) The bar graph represents the number of activated P14 cells after 4 hours of transfer in NK cells-depleted naïve C57BL6/J (B6/J) mice or Fcer1g^+/– and Fcer1g^–/– mice which were i.v infected with 200 PFU of LCMV-WE strain 3 days before the transfer (n = 6–7). Data are pooled from 2 independent experiments. The detailed protocol is described in materials and methods (in-vivo killer assay). Data are shown as mean ± SEM. Significant differences between the two groups were detected by unpaired two-tailed t-tests and are indicated as follows:: ns, not significant; * p<0.05; ** p<0. 01; *** p<0.001; **** p<0.0001.

Fig 5. FcεRIγ exacerbates viral control during the course of chronic LCMV infection.

Several groups of Fcer1g^+/+ and Fcer1g^–/– mice were infected i.v with 2 × 10⁴ PFU of LCMV-Docile, were bled or killed at diverse time points, and were analyzed for certain variables. (A) Kinetics of viral titers in serum at the indicated time points after infection (n = 4–8). Data are pooled from 3 independent experiments. (B) Kinetics of viral titers in various organs at the indicated time points after infection (n = 3–4). (C) Viral titers in various organs on day 28 after infection (n = 7–8). Data are pooled from 2 independent experiments. (D) Levels of aspartate aminotransferase (AST), alanine aminotransferase (ALT), and lactate dehydrogenase (LDH) measured in serum on day 12 after infection (n = 4). (E) Percentage of body weight is shown at various days after infection (n = 5). (F) Representative immunofluorescence for liver histological sections from Fcer1g^+/– and Fcer1g^–/– mice stained for LCMV nucleoprotein (green) and CD8⁺ T cells (red) at day 12 after infection. One slide representative of 4 slides is shown. Scale bar, 200μm. Data are shown as mean ± SEM. Significant differences between the two groups were detected by unpaired two-tailed t-tests and are indicated as follows: ns, not significant; * p<0.05; ** p<0.01; *** p<0. 001; **** p<0.0001.

Fig 6. FcεRIγ of natural killer cells exerts an intrinsic effect on virus control.

(A) Schematic of experimental setup. Fcer1g^+/– and Fcer1g^–/– mice were injected intraperitoneally with 200ug of anti-NK1.1 or isotype antibody on day -3 and day -1 and were infected i.v with 2 × 10⁴ PFU of LCMV-Docile at day 0. The mice were bled on days 8, 12, 20, and 32 after infection and were put to death on day 32 after infection. (B) The upper panel shows representative FACS plots for the frequency of glycoprotein (GP)33-Tet⁺ CD8⁺ T cells in the spleens on day 32 after infection. The lower panel shows graphs indicating the frequencies of CD8⁺ T cells and GP33-Tet⁺ CD8⁺ T cells in murine spleens on day 32 after infection (n = 6–10). (C) The FACS plots (upper panel) and graphs (lower panel) show the percentages of CD8⁺ T cells producing IFN-γ and TNF-α from splenocytes on day 32 after infection. These cells were stimulated in-vitro for 5 hours in the presence of GP33 peptide (n = 6–10). (D) Kinetics of viral titers in serum at indicated time points (n = 7–10). (E) Viral titers from various organs on day 32 after infection (n = 7–10). Data are pooled from two independent experiments (B-E). Data are shown as mean ± SEM. Significant differences between the groups were detected by unpaired two-tailed t-tests and are indicated as follows: ns, not significant; * p<0.05; ** p<0.01; *** p<0. 001; **** p<0.0001.