Chemokines control naive CD8+ T cell selection of optimal lymph node antigen presenting cells

Abstract

Naive antiviral CD8(+) T cells are activated in the draining LN (DLN) by dendritic cells (DCs) presenting viral antigens. However, many viruses infect LN macrophages, which participate in initiation of innate immunity and B cell activation. To better understand how and why T cells select infected DCs rather than macrophages, we performed intravital microscopy and ex vivo analyses after infecting mice with vaccinia virus (VV), a large DNA virus that infects both LN macrophages and DCs. Although CD8(+) T cells interact with both infected macrophages and DCs in the LN peripheral interfollicular region (PIR), DCs generate more frequent and stable interactions with T cells. VV infection induces rapid release of CCR5-binding chemokines in the LN, and administration of chemokine-neutralizing antibodies diminishes T cell activation by increasing T cell localization to macrophages in the macrophage-rich region (MRR) at the expense of PIR DCs. Similarly, DC ablation increases both T cell localization to the MRR and the duration of T cell-macrophage contacts, resulting in suboptimal T cell activation. Thus, virus-induced chemokines in DLNs enable antiviral CD8(+) T cells to distinguish DCs from macrophages to optimize T cell priming.

Figure 1.

Histological characterization of the MRR of the LN. (A and B) Frozen LN sections from WT mice given 70-kD tetramethylrhodamine-labeled dextran (red) 30 min before excision. (A) ER-TR7 staining (grey) identifies the following nodal regions: MRR, macrophage-rich region; PIR, peripheral interfollicular region; CA, capsule; SCS, subcapsular sinus; CR, cortical ridge; B, B cell follicle; T, T cell zone. (B) Mosaic-tiled confocal images of whole LNs stained with different Abs (grey, left images), and higher magnification images of staining of the MRR (right images). (C) Number of dextran⁺ cells per node per cell type as determined by node dissociation followed by flow cytometry. Cells were gated on FITC-dextran positivity and divided into CD11b⁺ DCs (CD11c⁺CD11b⁺ cells), CD11b⁻ DCs (CD11c⁺CD11b⁻ cells), and macrophages (CD11c⁻CD11b⁺ cells). Macrophages were further gated on F4/80 to identify medullary macrophages (F4/80⁺). Dextran⁺ cells belonging to each population were quantified using flow cytometric percentages and total node counts. (D) MPM images from CD11c-eYFP mice given dextran. Bottom panels show higher magnification. CD11c⁺ cells (green), collagen (second harmonic generation, blue), dextran⁺ cells (red). (E) High-magnification confocal image of whole-mounted CD11c-eYFP LNs after dextran administration. CD11c⁺ cells (green), dextran⁺ cells (red). Scale bars are shown in micrometers. We made similar observations in at least three additional experiments per image.

Figure 2.

LN SCS and medullary macrophages are infected by VV. (A) Flow cytometric analysis of single-cell LN suspensions from saline-injected mice. CD169⁺CD11c^lo macrophages can be further divided into CD11b⁺F4/80⁻ SCS or CD11b⁺F4/80⁺ medullary macrophages. (B) LNs from VV-infected mice 10 hpi. Cells were gated based on expression of a vaccinia-expressed fluorescent protein, into DCs and CD169⁺ macrophages, and finally on F4/80⁺ or F4/80⁻ cells. (C) Numbers of vaccinia-infected cells in different nodal populations. Circles indicate individual LNs, and bars show mean. (D) Whole-mount confocal images of LNs 10 hpi, with VV-NP-S-mCherry (red). Macrophages, green (FITC-labeled dextran); F4/80, grey. Scale bars are shown in micrometers. Data are representative of three experiments.

Figure 3.

CD8⁺ T cells interact with LN DCs and macrophages after viral infection. (A) LNs from CD11c-eYFP mouse (green DCs) excised 10 hpi with VV-NP-S-mCherry (red) were examined as whole mounts by confocal microscopy. OT-I cells were labeled with Cell Tracker Blue before transfer. (B) MPM images of the inguinal LN, as in A. (C) MPM images from the top 30 µm of the inguinal node of CD11c-eYFP mice given fluorescent-dextran (red) to label macrophages and OT-I cells (blue). Uninfected nodes (left) 10 hpi with nonfluorescent backbone virus (VV-β-gal; control, middle left) or VV-Ova (middle right). (right) A schematic of T cell localization in relation to macrophages after VV infection. DCs (green), MRR (red), OT-I cells (blue; right). Scale bars are shown in micrometers. We made similar observations in two additional experiments.

Figure 4.

CD8⁺ T cell interactions with macrophages are largely nonproductive. (A) LN OT-I cells dividing in response to s.c. delivery of VV-NP-S-eGFP. 5 × 10⁶ OT-I CFSE-labeled OT-I cells were transferred into CD11c-DTR-eGFP mice. 12 h later, mice were treated with DTx or PBS (untreated). After an additional 12 h, mice were infected with VV (untreated and DTx) or left uninfected (naive). Cells were analyzed at 48 hpi for division. Far right panel shows overlays. Untreated mice, black lines; DTx-treated mice, red lines; uninfected DTx-treated, filled grey lines. Numbers = OT-I cells recovered. (B) Graph showing percentage of dividing OT-I cells. Dots represent individual nodes. (C–E) Activation marker profile of T cells. Histograms were only gated on dividing cells. Untreated mice, black lines; DTx-treated mice, red lines; uninfected DTx-treated, filled grey lines. Numbers in top right corner indicate time after infection. (F) IFN-γ production (y axes) versus division (indicated by CFSE dilution; x axes) at 2 days after infection with VV-Ova. (G) Graphical representation of data shown in F. Dots represent individual nodes. (H) 2 × 10⁶ OT-I cells were transferred into CD11c-DTR-eGFP mice (expressing CD45.2) that were treated with DTx (middle) or PBS (right) before intradermal infection with VV-SIINFEKL in the ear pinnae. 4 days after infection, ears were removed and analyzed for the percentage of OT-I cells present (CD8⁺CD45.1⁺ cells). (I) Graphical representation of data in H. All experiments were performed at least three times with three to six animals/group yielding similar results.

Figure 5.

Infected macrophages present antigen in DC-ablated mice. (A) VV-NP-S-eGFP–infected cells (green) just under the collagenous capsule of the node (second harmonic generation (blue)). Numbers of infected cells in nonablated (untreated) and DC-ablated (DTx-treated) mice (right). (B) Confocal microscopy of a frozen section of a DC-ablated LN 10 hpi with VV-NP-S-mCherry (red) showing only the red and green channels (left), only the red and grey channels (middle), or all channels (right). FITC-dextran (green) and anti-CD11b staining (white) identify macrophages. (C) Same as shown in B, except staining with F4/80 (grey) to identify medullary macrophages. (D) Numbers of infected cells of each cell type in the node. Cells were gated into CD11c⁺CD169⁻ (DCs) or CD11c^dimCD169⁺ (CD169⁺ macrophages), then CD169⁺ macrophages were further gated based on F4/80 expression. Background (blue) is caused by low cell recovery and macrophage autofluorescence. (E) Flow cytometric analysis of LN single-cell suspensions from untreated or DTx-treated animals 6 hpi with VV-venus-ubiquitin-SIINFEKL. Cells were stained with 25D-1.16 recognizing K^b-SIINFEKL complexes. Histograms are gated on infected cells (Venus eYFP⁺). Infected with virus lacking SIINFEKL (grey shaded histograms), infected without 25D-1.16 stain (black lines), infected untreated mice (blue lines), and infected DTx-treated mice (red lines). Scale bars are shown in micrometers. Results are shown from one experiment of two (A–C) or four (D and E).

Figure 6.

CD8⁺ T cells stably interact with VV-infected cells in DC-ablated mice. (A) MPM images of untreated or DTx-treated CD11c-DTR-eGFP animals that were given 1.0 × 10⁷ Cell Tracker Red (red) labeled OT-I cells before DC depletion. Images acquired 6 hpi with VV-SIINFEKL (nonfluorescent). The macrophage rich-region (MRR) was identified by in vivo uptake of FITC-dextran (green). (B) Percentage of OT-I cells per 63X field (238 × 238 × 85 µm) located in the MRR in untreated or DTx-treated mice. Results were analyzed with an unpaired t test. (C) Time-lapse MPM images of OT-I cells (red) in the LN of untreated (top panels) or DC-ablated (bottom panels) mice 6–10 hpi with VV-NP-S-eGFP (green). White circles indicate stable contacts. For untreated mice, 10⁶ OT-I cells were transferred; for DTx-treated, 10⁷ (it was necessary to transfer more OT-I cells into DTx-treated mice than untreated mice due to decreased homing to the LN following DTx-mediated ablation). (D) Calculation of contact times between OT-I cells and VV-infected cells over the course of a 30 min imaging session. (E) Plot of OT-I cells’ movement (step size) between individual frames of a 30 min movie in untreated vs. DC-ablated mice. We classified movement under 0.5 µm between frames as a pause step. (F) Calculation of the percentage of T cells arresting in each condition. (G) Color-mapped plot of T cell tracks during a 30-min movie. Each box on grid is 20 × 20 µm. Increasing track speed is colored from purple to red (bottom bar). (H) Calculation of average OT-I cell speed in untreated or DTX-treated mice. Results are shown from one experiment of three to six with similar results. No differences between untreated and DTx-treated mice were statistically significant according to unpaired Student’s t tests. Time is shown in minutes. Scale bars are shown in micrometers.

Figure 7.

CD8⁺ T cells rapidly scan infected non-DCs. (A) MPM time series of T cells interacting with infected DCs, uninfected DCs, or infected non-DCs. CD11c-eYFP mice were given 10⁷ Cell Tracker Blue-labeled OT-I cells (blue) and were infected s.c. with VV-NP-S-mCherry 12 h. later. Six hpi, inguinal LNs were imaged over sequential 20–30 min periods using MPM. The tracks of individual T cells during a 20-min imaging period are plotted (right). (B) T cell speeds, (C) track straightness (T cell displacement/track length), and (D) T cell arrest calculated for each type of APC interaction over 30 min. Open circles, infected DCs; closed triangles, uninfected DCs; closed circles, macrophages. Mean and SEM is shown. Statistics were performed using an unpaired Student’s t test. Shown is one experiment of two analyzed with 10–20 nodes per group.

Figure 8.

VV infection induces rapid DC chemokine secretion. Mice were infected s.c. with VV-Ova and LN harvested at 6 or 12 hpi Chemokine protein levels were determined from clarified node homogenates via Bioplex assay. Dots represent individual mice. Data are shown from two of four independent experiments.

Figure 9.

Chemokine-based CD8⁺ T cell homing to DCs. (A) MPM images of animals that were given 1.0 × 10⁷ Cell Tracker Red (red)–labeled OT-I cells in the presence or absence (untreated) of chemokine-neutralizing Abs against CCL3, CCL4, and CCL5. Images acquired 6–8 hpi with VV-Ova (nonfluorescent). MRR delineated by white lines. (B) Percentage of OT-I cells per 63× field (238 × 238 × 85 µm) located in the MRR in untreated or antibody-treated mice. Results were analyzed with an unpaired Student’s t test. (C) MPM images from the top 30 µm of the inguinal node of CD11c-eYFP mice given fluorescent-dextran (red) to label macrophages and OT-I cells (blue). Mice were given chemokine neutralizing Abs; nodes imaged at 8 hpi with nonfluorescent VV-Ova. (right) A schematic of T cell and macrophage localization after VV infection with CCR5-ligand blockade. Scale bars are shown in micrometers. (D and E) OT-I cells (red) in the MRR (green) after s.c. administration of recombinant CCL3 at 2.25 hpi with VV-SIINFEKL (nonfluorescent; E) tracks of OT-I cells in the presence (left) or absence (right) of rCCL3 2–3 hpi. Tracks are colored according to mean track speed (slowest [purple) to fastest [red]). (F) Distribution of CCR5KO OT-I cells (red, left) or WT OT-I cells (red, right) 6 hpi with VV-NP-S-eGFP (green). (G) CCR5KO OT-I cells (red) in the MRR (visualized using the intrinsic autofluorescence of macrophages) at 6 hpi (virus=green, collagen=blue) (H) CD69 expression by LN OT-I cells 15 hpi with VV-Ova in the presence or absence of chemokine-neutralizing Abs. Data were averaged from three independent experiments normalized using the highest mean fluorescence intensity in an individual experiment as 100. (I) IFN-γ production by OT-I cells in the node 48 hpi with VV-Ova. Data were compiled from two independent experiments and normalized to the highest mean fluorescence intensity per experiment. (J) Same as described in H, but with CCR5KO OT-I cells. All experiments were performed at least twice with three to six mice/group.