Tissue-resident memory T cell reactivation by diverse antigen-presenting cells imparts distinct functional responses

Abstract

CD8+ tissue-resident memory T cells (TRM cells) are poised at the portals of infection and provide long-term protective immunity. Despite their critical roles, the precise mechanics governing TRM cell reactivation in situ are unknown. Using a TCR-transgenic Nur77-GFP reporter to distinguish "antigen-specific" from "bystander" reactivation, we demonstrate that lung CD8+ TRM cells are reactivated more quickly, yet less efficiently, than their counterparts in the draining LNs (TLN cells). Global profiling of reactivated memory T cells revealed tissue-defined and temporally regulated recall response programs. Unlike the reactivation of CD8+ TLN cells, which is strictly dependent on CD11c+XCR1+ APCs, numerous antigen-presenting partners, both hematopoietic and non-hematopoietic, were sufficient to reactivate lung CD8+ TRM cells, but the quality of TRM cell functional responses depended on the identity of the APCs. Together, this work uncovers fundamental differences in the activation kinetics, mechanics, and effector responses between CD8+ memory T cells in peripheral vs. lymphoid organs, revealing a novel tissue-specific paradigm for the reactivation of memory CD8+ T cells.

Figure 1.

Lung T_RM cells are reactivated more rapidly but less efficiently than T_LN cells in medLNs. (A) P14/OT-I Nur77-GFP immune chimeras were generated by cotransferring naive P14⁺ (Thy1.1⁺) and OT-I⁺ (Ly5.1⁺) Nur77-GFP CD8⁺ T cells (5 × 10⁴ cells each) into C57BL/6 (Th1.2⁺/Ly5.2⁺) mice 1 d before i.n. coinfection with recombinant influenza X31-gp33 and X31-ova. 30 d later, the P14/OT-I Nur77-GFP immune chimeras were reinfected with PR8-gp33 i.n., and reactivation of P14⁺ lung T_RM and medLN T_LN cells was assessed at 0, 12, 24, 48, and 120 h.p.i. (B) Representative flow plots of Nur77-GFP expression in P14⁺ lung T_RM cells and medLN T_LN cells. In vivo labeling with αCD8β antibody was used to distinguish CD8⁺ T_RM cells in the lung parenchyma from CD8⁺ T cells in the vasculature. (C and D) Frequency of Nur77-GFP⁺ CD8β⁻ P14⁺ cells (i.e., top left quadrant of flow plots) in B was quantified. Data are expressed as mean ± SD. (E and F) As negative controls, Nur77-GFP expression in OT-I⁺ and P14⁺ cells was examined after PR8-gp33 and PR8 2° infection, respectively. (G) To limit the contribution of T_CIRC cells, a low dose of αThy1.1 (clone 19E12)–depleting antibody was administered i.p., and the frequency of Nur77-GFP⁺ P14⁺ T_RM cell reactivation was examined. N.D., not detected. Antigen-activated P14⁺ lung T_RM (blue) and medLN T_LN (red) cells are distinguished from the bystander-activated OT-I⁺ lung T_RM (green) and medLN T_LN (gold) cell samples by color. Data shown are representative of two independent experiments (n = 3–5 mice/group).

Figure S1.

Intravascular staining to distinguish parenchymal T_RM cells from vasculature T_LN cells. (A) Flu-immunized mice were injected i.v. with fluorescently labeled αCD8β antibody 5 min before sacrifice. Blood, lung, and medLN were collected and stained with αCD8α antibody ex vivo. Shown are gated on total CD8α⁺ T cells. (B) Expression of CD69 and CD103 in medLN, lung parenchyma (CD8a⁺ CD8β⁻), and lung vasculature (CD8a⁺ CD8β⁺) are shown.

Figure S2.

GzmB and CD98 are markers of bystander inflammation, and recruitment of circulating T cells is minimal at 48 h.p.i. (A and B) P14⁺ immunized mice were infected with PR8-gp33 (TCR-activation) or PR8 (bystander-activation) i.n., and the expression of GzmB (A) and CD98 (B) at 72 h.p.i. in the lung (top panel) and medLN (bottom panel) is shown. (C) P14⁺ T_RM and T_LN cells were sorted from resting P14⁺ chimera and co-cultured with control- or gp33-pulsed splenocytes, and Nur77-GFP expression was accessed 24 h later. (D and E) To limit the contribution of T_CIRC cells to the readout of Nur77-GFP in cells in the lung, FTY720 (D) and a low dose of αThy1.1-depleting antibody (E) were administered i.p., and the effects on P14⁺ T_RM cell reactivation was examined 48 h.p.i. n.s., not significant. Data are expressed as mean ± SD.

Figure 2.

CD8⁺ memory T cells in the lung and medLNs display tissue-defined functional programs that are temporally regulated following reactivation in situ. (A) P14⁺ and OT-I⁺ memory CD8⁺ T cells were purified at 0, 24, 48, and 168 h after 2° infection from the lung and medLNs, and RNA-seq was performed. Activated P14⁺ T cells were sorted based on Nur77-GFP⁺ expression, and bystander-activated OT-I T cells were sorted based on CD98^hi expression. (B) PCA of global gene expression showed five distinct clusters reflected by dotted ellipses. (C) Heatmap of differentially expressed TCR-activated genes and bystander-activated genes (log₂FC ≥2 and P < 0.01). Seven modules were distinguished and several representative genes are shown. TCR-activated P14⁺ Nur77-GFP⁺ lung T_RM (blue) and medLN T_LN (red) are distinguished from the bystander-activated (Nur77-GFP⁻ CD98^hi) OT-I⁺ lung T_RM (green) and medLN T_LN (gold) samples by color. Squares, 0 h.p.i; triangles, 24 h.p.i.; diamonds, 48 h.p.i; circles, 168 h.p.i. (D) T_RM and T_LN identity genes were identified by comparing 0 h resting P14⁺ memory CD8⁺ T cells from the lung and medLNs_, based on 1.5-log₂FC and P < 0.05, and the temporal changes of these identity genes following reactivation in T_RM and T_LN cell populations are shown in this heatmap.

Figure S3.

T_RM and T_LN identity genes are maintained following reactivation. T_RM and T_LN identity genes were identified by comparing 0-h resting P14⁺ T_RM cells with 0-h resting P14⁺ T_LN cells based on 1.5-log₂FC and P < 0.05, as represented in this volcano plot. 982 T_RM identity genes and 520 T_LN identity genes were identified.

Figure 3.

Reactivated T_RM cells are in close proximity to different infected cell types. (A) Influenza-immunized mice were infected with PR8-GFP and PR8 (as control), and the lungs were analyzed by flow cytometry 48 h.p.i. (B) The pie chart shows the types of infected GFP⁺ cells identified as a fraction of total GFP⁺ cells (pie chart: alveolar macrophages [CD45⁺ CD169⁺ SiglecF⁺], B cells [CD45⁺ B220⁺], DCs and macrophages [CD45⁺ SiglecF⁻ B220⁻ CD11c⁺ Ly6C⁻], monocytes [CD45⁺ SiglecF⁻ B220⁻ CD11c⁺ Ly6C⁺], and epithelial cells [CD45⁻ EPCAM⁺]). (C–E) P14⁺ immune chimeras were rechallenged with PR8-gp33, and the localization of activated P14⁺ Nur77-GFP⁺ cells in the lung (C and D) and medLN (E) at 48 h.p.i. was analyzed by immunofluorescence confocal imaging. Arrows indicate TCR-activated regions of P14⁺ Nur77-GFP⁺ cells and their interactions with infected cells and different immune subsets. Data shown are representative of two to three independent experiments (n = 3–5 mice/group). All scale bars indicate 50 µm.

Figure S4.

Depletion validation and Zbtb46⁺ cells are required for medLN T_LN cell, but not lung T_RM cell, reactivation. P14⁺ immune chimeras were generated in various genetic hosts, and DT was administered i.t. 1 d before 2° infection to deplete DTR-expressing cells; for αGR-1 depletion, 200 μg αGR-1 (RB6-9C5) was administered i.p. daily starting the day before 2° rechallenge. (A) Validation of depletion strategies across different immunized-genetic hosts 48 h after the initial treatments in the lung or bronchoalveolar lavage. (B) Frequency of Nur77-GFP⁺ P14⁺ T_RM cells in the lung and T_LN cells in the medLN were quantified at 48 h.p.i. in Zbtb46-DTR hosts. Statistical analysis was performed using Student’s t test (two tailed); ****, P < 0.0001. (C) Flu viral titer was obtained by quantitative PCR of whole lung tissue for PR8 polymerase acidic protein (PA) 48 h.p.i. Data are expressed as mean ± SD.

Figure 4.

CD11c⁺ XCR1⁺ cells are strictly required for medLN CD8⁺ T_LN reactivation, but conventional APCs are dispensable for lung CD8⁺ T_RM reactivation. (A) P14⁺ immune chimeras were generated in various genetic hosts as shown. For DTR mice, DT was administered i.t. 1 d before 2° infection to deplete DTR-expressing cells; for αGR-1 depletion, 200 μg αGR-1 (RB6-9C5) was administered i.p. daily starting the day before 2° rechallenge. For LCMV rechallenge experiments, flu-immunized mice were 2° infected with Armstrong strain of LCMV i.n. or LCMV-GFP. (B–D) Quantification of the frequency of Nur77-GFP⁺ P14⁺ T_LN cells in the medLN (B) and P14⁺ T_RM cells in the lung (C) at 48 h after 2° PR8-gp33 infection (B and C) or 48 h after 2° LCMV infection (D). Data shown are a collection of two or more independent experiments (n = 3–5 mice/group). (E and F) P14⁺ immune chimeras were rechallenged i.n. with LCMV-ZsGreen (E), and the colocalizations of ZsGreen signal with various immune cells in the medLN was quantified by Imaris (F). Top scale bars indicate 150 µm, and bottom scale bars indicate 20 µm. Data are expressed as mean ± SEM. Statistical analysis was performed using Student’s t test (two tailed) comparing immunized C57BL/6 to various treatment groups. ****, P < 0.0001.

Figure 5.

CD8⁺ lung T_RM cells can be reactivated by both hematopoietic and nonhematopoietic APCs. (A) P14⁺ immune chimeras were generated in different H-2Db^−/− BMCs as outlined. For CD11c-DTR donor bone marrow groups, DT was administered 500 ng/mouse (i.p.) 1 d before 2° infection. (B) Representative flow plots of Nur77-GFP expression in the P14⁺ lung T_RM cells at 48 h.p.i. and quantified in C. (D) P14⁺ immune chimeras were generated in TAP1^−/− BMCs similar to A and the frequency of reactivated medLN T_LN and lung T_RM cells are quantified. (E) P14⁺ T_RM cells were sorted from steady-state memory lung at >30 d p.i. and co-cultured with epithelial cells (EPCAM⁺), endothelial cells (CD31⁺), CD11c⁺ MHCII⁺ cells, and alveolar macrophages (CD169⁺) sorted from influenza-immunized mice that were 2°-reinfected with PR8-gp33 24 h prior. Flow plots of Nur77-GFP expression in P14⁺ T_RM cells following overnight co-culture. Data shown are expressed as mean ± SD, representative of two independent experiments (n = 3–5 mice/group). Statistical analysis was performed using Student’s t test (two tailed). **, P < 0.01.

Figure S5.

CD8⁺ lung T_RM cells are reactivated by both hematopoietic and nonhematopoietic APCs, and consequentially, their 2° expansion in situ is independent of cDCs. (A) Validation of H-2Db expression in CD45⁺ and CD45⁻ compartments of BMCs. (B) Schematic of ex vivo co-culture. P14⁺ T_RM cells were sorted from steady-state memory lung at >30 d p.i. and co-cultured with epithelial cells (EPCAM⁺), endothelial cells (CD31⁺), CD11c⁺ MHCII⁺ cells, and alveolar macrophages (CD169⁺) sorted from influenza-immunized mice that were 2°-reinfected with PR8-gp33 24 h prior. (C) P14⁺ immune chimeras were generated in C57BL/6 or CD11c-DTR hosts and treated with either FTY720 (i.p., daily 2 days prior) or 100 ng DT (i.t., every 2–3 d) the day before 2° reinfection with PR8-GP33 i.n. Shown are representative flow plots of the expression of Ki67 on P14⁺ cells 96 h.p.i. (D) P14⁺ immune chimeras were generated in C57BL/6 or CD11c-DTR hosts and treated with either FTY720 (i.p.; daily 2 d prior) or 100 ng DT (i.t.; every 2–3 d) the day prior to 2^o reinfection with PR8-GP33 i.n. 96 h.p.i. the expression of Ki67 on P14⁺ cells were examined and quantified. After infection., the expression of Ki67 on P14⁺ cells were examined and quantified.

Figure 6.

Activation by hematopoietic and nonhematopoietic cells induces differential functional output in T_RM cells. (A) P14⁺ immune chimeras were generated in C57BL/6 or CD11c-DTR hosts and treated with either FTY720 (i.p., daily 2 d prior) or 100 ng DT (i.t., every 2–3 d) the day before 2° reinfection with PR8-GP33 i.n. 96 h.p.i. the expression of Ki67 on P14⁺ cells was examined and quantified. Data are expressed as mean ± SEM. (B) Activated P14⁺ T_RM cells were sorted based on Nur77-GFP⁺ expression in C57BL/6 TAP1^+/+, TAP1^BM−/−, TAP1^{nonhemo−/−} BMCs, and RNA-seq was performed. PCA of global gene expression showed two distinct clusters reflected by dotted ellipses. (C) Heatmap of differentially expressed genes between the control and TAP1^−/− BMCs (≥ or ≤1.5 log₂FC and P < 0.05), filtered on all the differentially expressed genes of TCR and bystander T_RM cell signature from Fig. 2 C. Three main modules were observed. (D) Genes from each module were parsed for enrichment of biological pathways using Enrichr. Top pathways from Gene Ontology Biological Processes with corresponding adjusted P values are shown. Relevant gene names are highlighted in heatmap in C.