CD4+ T cell help guides formation of CD103+ lung-resident memory CD8+ T cells during influenza viral infection

Abstract

Tissue-resident memory T (Trm) cells provide enhanced protection against infection at mucosal sites. Here we found that CD4(+) T cells are important for the formation of functional lung-resident CD8(+) T cells after influenza virus infection. In the absence of CD4(+) T cells, CD8(+) T cells displayed reduced expression of CD103 (Itgae), were mislocalized away from airway epithelia, and demonstrated an impaired ability to recruit CD8(+) T cells to the lung airways upon heterosubtypic challenge. CD4(+) T cell-derived interferon-γ was necessary for generating lung-resident CD103(+) CD8(+) Trm cells. Furthermore, expression of the transcription factor T-bet was increased in "unhelped" lung Trm cells, and a reduction in T-bet rescued CD103 expression in the absence of CD4(+) T cell help. Thus, CD4(+) T cell-dependent signals are important to limit expression of T-bet and allow for the development of CD103(+) CD8(+) Trm cells in the lung airways following respiratory infection.

Figure 1. Intravascular Staining Allows for Identification of Resident and Circulating Influenza-Specific Memory CD8⁺ T Cell Populations

Analysis of the NP³⁶⁶⁺ T cell response 40 days after X31-GP33 infection. Mice were intravenously (i.v.) injected with CD8β antibody prior to sacrifice to distinguish tissue parenchyma versus vascular-associated influenza-specific CD8⁺ T cells. (A) Representative plots of the NP₃₆₆⁺ T cells in the CD8⁺ T cell population and the CD8β⁺ T cells in the NP₃₆₆⁺ T cell population. (B) Representative histograms of CD103, CD69, KLRG1, Ly6C, CD62L, CD127, CXCR3, T-bet, and GzmB expression in CD8β⁺ (gray filled) and CD8β⁻ (black) NP366⁺ T cells. Representative of four independent experiments with four or five mice per group carried out 30–60 days following X31-GP33 infection. Statistical analyses were performed using Prism GraphPad software v6.0. (*p < 0.05; **p < 0.01; ***p < 0.001).

Figure 2. CD4⁺ T Cell Help Critical for the Formation of CD103^hi CD8⁺ Trm Cells

Analysis of the NP₃₆₆⁺ T cell response 40 days after X31-GP33 infection in PBS (helped) and GK1.5 (unhelped) treated mice. Analysis of the NP₃₆₆⁺ T cell response 40 days after X31-GP33 infection. (A) Percentage and (B) numbers of NP₃₆₆⁺T cells and resident and circulating NP₃₆₆⁺ T cells. Intravenously administered anti-CD8β antibody was used to distinguish NP₃₆₆⁺ T cells in tissue parenchyma versus vascular associated. (C) Percentages of NP₃₆₆⁺ cells, which are CD103⁺, CD69⁺, and CXCR3⁺. (D) CD103 and CD69 expression in NP₃₆₆⁺ CD8β⁻ T cells and CXCR3 expression in NP₃₆₆⁺ CD8β⁺ T cells in helped (black) and unhelped (gray filled) mice. Representative of four independent experiments with four or five mice per group carried out 30–60 days following X31-GP33 infection. See also Figure S1.

Figure 3. CD4⁺ T Cell Help Is Important for the Localization of Influenza-Specific Memory CD8⁺ T Cells to Lung Airway Epithelium

Analysis of the localization of P14 GP33⁺ T cells 45 days after X31-GP33 infection in PBS (helped) and GK1.5 (unhelped) treated mice. 50,000 Thy1.1⁺ P14 GP33⁺ T cells were transferred to each mouse 1 day prior to X31-GP33 infection to facilitate identification of influenza-specific CD8⁺ T cells. (A) At a memory time point (day 45), the mice were sacrificed and the lungs fixed and stained for Thy1.1 (red) and EpCAM (airways, green) and analyzed with confocal microscopy. Representative staining showing presence of Th1.1⁺ P14 GP₃₃⁺ T cells embedded in lung airways. (B) Comparison of the total number of P14 GP₃₃⁺ T cells in each image and the percent of P14 GP₃₃⁺ T cells touching the lung airway between helped (PBS treated) and unhelped (GK1.5 treated) mice. Imaris suite was used for analysis of all images. Representative of two independent experiments with four or five mice per group carried out 30–45 days following X31-GP33 infection. See also Figure S2.

Figure 4. Unhelped Influenza-Specific CD8⁺ T Cells Impaired in Ability to Mediate Heterosubtypic Protection

Analysis of the ability of helped and unhelped memory CD8⁺ T cells to mediate heterosubtypic protection. To directly compare the ability of helped and unhelped CD8⁺ T cells to mediate heterosubtypic protection, PBS (helped)- and GK1.5 (unhelped)-treated MD4 mice were infected with X31-GP33. Two weeks later, both groups were treated with GK1.5 to create both helped and unhelped mice that do not have influenza-specific CD4⁺ T or B cells. Both groups were then rechallenged with the heterosubtypic influenza virus PR8-GP33, alongside that of naive mice as unimmunized controls. (A) Viral load at day 2 after PR8-GP33 rechallenge. (B) Number of P14 GP33⁺ T cells in lung airway at day 4 after PR8-GP33 infection. (C) MFI of granzyme B expression of lung P14 GP33⁺ T cells at day 4 after PR8-GP33 infection. Representative of four independent experiments with three to five mice per group. (D) Helped and unhelped MD4 mice were generated as described above. At a memory time point (day 30), P14 GP33⁺ T cells were isolated from the spleens of helped and unhelped X31-GP33 immune mice and equal numbers of congenically marked cells transferred in to helped or unhelped MD4 mice. The next day all recipients were challenged with PR8-GP33 and the number of the transferred P14 GP33⁺ found in the lung airways 6 days later determined. Representative of three independent experiments with three to five mice per group.

Figure 5. CD4⁺ T Cell-Derived IFN-γ Is Important for Formation of CD103^hi CD8⁺ T Cells

Analysis of the temporal requirements for CD4⁺ T cell help for the formation of CD103^hi CD8⁺ Trm cells. Mice were treated with GK1.5 antibody at various time points after X31-GP33 infection. (A) CD103 MFI in CD8β⁻ GP33⁺ T cells at day 45 after X31-GP33 infection. Representative of three independent experiments with four or five mice per group. (B) 1 × 10⁶ P14 GP33⁺ CD8⁺ T cells and 1 × 10⁶ Smarta GP66⁺ CD4⁺ T cells were transferred to mice 1 day prior to WSN-GP33/67 infection and the kinetics of influenza-specific CD4⁺ and CD8⁺ T cell entry into the lung airways determined. Representative of two independent experiments with three to four mice per group. WT:Cd4^−/− and Ifng^−/−:Cd4^−/− mixed bone-marrow chimeras were generated. Following reconstitution, the WT:Cd4^−/− mice were treated with PBS or GK1.5 and all groups infected with X31-GP33 and sacrificed at day 40. (C) Percentages (top) and numbers (bottom) of NP₃₆₆⁺ T cells, and resident versus circulating NP₃₆₆⁺ T cells. Intravenous anti-CD8β antibody was used to distinguish tissue parenchyma versus vascular-associated NP₃₆₆⁺ T cells. (D) Representative plots of CD103⁺CD69⁺ NP₃₆₆⁺ T cells (top). Percentages of NP₃₆₆⁺ cells that are CD103⁺, CD69⁺, and CD103⁺CD69⁺ (bottom). Representative of three independent experiments with four or five mice per group carried out 30–60 days following X31-GP33 infection. See also Figure S3.

Figure 6. T-bet Suppresses Formation of Trm Cells and Is Capable of Directly Binding to Itgae Locus

Analysis of T-bet expression in NP₃₆₆⁺ CD8β⁻ T cells in PBS (helped) and GK1.5 (unhelped) treated mice. (A) Representative histogram of T-bet expression and MFI at day 8 after X31-GP33 infection in helped and unhelped mice. (B) Representative histogram of T-bet expression and MFI at day 45 after X31-GP33 infection in helped and unhelped mice. P14 GP33⁺ T cells were transduced with T-bet overexpressing or an empty vector and transferred to mice infected with X31-GP33 the prior day. Mice were sacrificed 30 days after X31-GP33 infection. (C) Representative plot (top) and percentage (bottom) of resident versus circulating transduced (GFP⁺) T cells. Intravenous anti-CD8β antibody was used to distinguish tissue parenchyma versus vascular-associated GFP⁺ T cells. (D) Percentage of GFP⁺ cells which are CD103⁺, CD69⁺, and CXCR3⁺ (top). CD103 and CD69 expression in GFP⁺ CD8β⁻ T cells and CXCR3 expression in GFP⁺ CD8β⁺ T cells in empty vector control (black) and T-bet overexpressing (gray filled) cells (bottom). Representative of three independent experiments with four or five mice per group. (E) Representative histogram of CD103 expression in empty (black) or T-bet overexpressing (gray filled) cells upon skewing in absence or presence of TGF-β. Representative of three independent in vitro culture experiments. (F) Representative histogram of pSmad2/3 expression in empty (black) or T-bet overexpressing (gray filled) cells following 1 hr culture in the absence or presence of TGF-β. Representative of three independent in vitro culture experiments. (G) Determination of the ability of T-bet to bind to Itgae locus in CD8⁺ T cells at a site in intron 1 in which T-bet is known to bind in CD4⁺ T cells. P14 GP33⁺ T cells were isolated at day 8 following LCMV Armstrong infection and ChIP PCR was performed to assess T-bet enrichment in intron 1 relative to an isotype control. Binding of T-bet to the CD127 3′ UTR and binding of T-bet in intron 1 of the Itgae locus in T-bet^−/− P14 GP33⁺ T cells were used as negative controls. The red text indicates a putative Smad3 binding site and the green text indicates a putative T-bet binding site. Representative of three independent experiments with five to ten pooled mice per group. See also Figure S4.

Figure 7. Genetic Modulation of T-bet Rescues the Ability of Unhelped Trm Cells to Upregulate CD103

50,000 Tbx21^+/+ and Tbx21^+/− P14 GP33⁺ T cells were transferred to mice treated with PBS (helped) or GK1.5 (unhelped) 1 day prior to X31-GP33 infection. Mice were sacrificed at day 40 postinfection. (A) Representative plot of the percentage of resident versus circulating helped and unhelped Tbx21^+/+ and Tbx21^+/− P14 GP33⁺ cells. Intravenous anti-CD8β antibody was used to distinguish tissue parenchyma versus vascular-associated GFP⁺ T cells. (B) The percentage of resident versus circulating helped and unhelped Tbx21^+/+ and Tbx21^+/− P14 GP33⁺ cells. (C) Percentage of P14 GP33⁺ cells that are CD103⁺, CD69⁺, and CXCR3⁺. (D) MFI of CD103 and CD69 expression in P14 GP33⁺ CD8β⁻ T cells and CXCR3 MFI in GP33⁺ CD8β⁺ T cells. Representative of three independent experiments with four or five mice per group carried out 30–60 days following X31-GP33 infection.