CXCR3 chemokine receptor-ligand interactions in the lymph node optimize CD4+ T helper 1 cell differentiation

Abstract

Differentiation of naive CD4(+) T cells into T helper (Th) cells is a defining event in adaptive immunity. The cytokines and transcription factors that control Th cell differentiation are understood, but it is not known how this process is orchestrated within lymph nodes (LNs). Here we have shown that the CXCR3 chemokine receptor was required for optimal generation of interferon-γ (IFN-γ)-secreting Th1 cells in vivo. By using a CXCR3 ligand reporter mouse, we found that stromal cells predominately expressed the chemokine ligand CXCL9 whereas hematopoietic cells expressed CXCL10 in LNs. Dendritic cell (DC)-derived CXCL10 facilitated T cell-DC interactions in LNs during T cell priming while both chemokines guided intranodal positioning of CD4(+) T cells to interfollicular and medullary zones. Thus, different chemokines acting on the same receptor can function locally to facilitate DC-T cell interactions and globally to influence intranodal positioning, and both functions contribute to Th1 cell differentiation.

Figure 1. CXCR3 expression is upregulated rapidly in draining LNs (dLNs) and correlates with IFN-γ expression

(A) Experimental protocol. Antigen-pulsed DCs were injected 24 hr before i.v. OT II cells. (Phenotype of transferred DCs Figure S1). CD62L blocking antibody was given 2 hr following OTII cell transfer and every 24 hr. At times indicated, dLNs (popliteal) and non-dLNs (bracial) were harvested to assess T cell activation. Time course of (B) cell numbers (C) CD44 and (D,E) CXCR3 cell surface induction on transferred OTII cells. (F,G) Correlation between IFN-γ⁺ cells and CXCR3 mean fluorescence intensity (MFI). Data are representative of 3 independent experiments (n=4–6).

Figure 2. CXCR3 is required for optimal Th1 differentiation following transfer of OVA-pulsed DCs

WT and Cxcr3^−/− OTII cells were co-transferred into WT hosts that received OVA-pulsed DCs. 60 hr post T cell transfer dLNs were harvested. (A) Plots of IFN-γ and TNF-α production by WT and Cxcr3^−/− OTII cells. (B) Frequency of IFN-γ⁺, IL-2⁺, TNF-α⁺ WT (square) and Cxcr3^−/− (triangle) OTII cells. (C) Total numbers of polyfunctional (producing IL-2, TNF-α and IFN-γ) WT and Cxcr3^−/− OTII cells in dLNs. (D) MFI of CD40L, CD69, CD25 at 24–36 hr post T cell transfer of WT and Cxcr3^−/− OTII cells. In (B–D) the line connects paired WT and Cxcr3^−/− OTII cells transferred into the same WT host. (E) CMFDA labeled cell proliferation following WT (grey) and Cxcr3^−/− (dashed) OTII cell co-transfer. No treat mice (without CD62L blocking were included to show peaks with continual T cell entry). (F) Frequency of TNF-α⁺IFN-γ⁺ WT (grey) and Cxcr3^−/− (open) OTII cells in mice without, or with CD62L blocking alone or with FTY720. (G) Kinetics of WT and Cxcr3^−/− OTII cell induction of IFN-γ⁺TNF-α⁺. Data are representative of 3 independent experiments (n=6–8). (See Figure S2A,B for transfer of WT and Cxcr3^−/− cells into separate hosts and Figure S2C for in vitro polarization of WT and Cxcr3^−/− OTII cells).

Figure 3. DC-derived CXCL10 optimizes Th1 responses and identification of chemokine-expressing DCs during T cell priming

(A,B) Pulsed WT, Cxcl9^−/− and Cxcl10^−/− DCs were transferred to hosts prior to OTII T cells. 60 hr post T cell transfer, dLN T cells were harvested. (A) Plots of IFN-γ and TNF-α production of WT OTII cells activated by DCs of indicated genotype. (B) Frequency of TNF-α⁺IFN-γ⁺ WT OTII cells activated by WT (black), Cxcl9^−/− (open), or Cxcl10^−/− (grey) DCs. (C) Schematic of REX3 Tg construct indicating insertion of RFP into the Cxcl9 locus and BFP into the Cxcl10 locus of the RP-24-164O11 BAC. Open box non-coding exons and black box coding exons of Cxcl9 and Cxcl10 genes; red box RFP ORF, blue box BFP ORF, and stripped box SV40 poly A site; FRT Flippase Recognition Target and loxP Cre recombinase site. (D) CMFDA labeled DCs expanded from REX3 Tg were pulsed, stimulated and injected into WT mice. Plot and bar graph indicating frequency of REX3 negative (open), CXCL10-BFP only (grey), and CXCL10-BFP and CXCL9-RFP double positive (stripe) DCs. (E) Tracked REX3⁺⁺ (expressing CXCL10-BFP and CXCL9-RFP) and REX3⁻ (REX3 negative) DCs MFI of CD40, CD86 at 24–36 hr post OTII cell transfer. Histogram and MFI plots are shown. (F–I) DCs expanded from REX3 Tg mice injected in WT mice. 24 hr later, WT and Cxcr3^−/− OTII cells were transferred i.v. MP-IVM of exposed dLN was performed at 6–8 hr post T cell transfer. (F) Intravital multiphoton micrograph of representative REX3⁺ DC interactions with WT and Cxcr3^−/− OTII cells. Numbered arrowheads indicate long-lived interactions between REX3⁺ DCs and WT (green) or Cxcr3^−/− (red) OTII cells. White lines indicate tracks of WT and Cxcr3^−/− cell centroids engaging in short-lived interactions with REX3⁺ DCs. Elapsed time in minutes:seconds. Bar indicates 30 µm. (G) Mean 3D track velocity, confinement ratio and arrest coefficient of WT and Cxcr3^−/− OTII cells. Dashed boxes around WT (green) Cxcr3^−/− (red) OTII cells indicate the frequency of cells in the specified region; 3D track velocity >10 µm/min; confinement ration >0.4; arrest coefficient <0.2 (not arrested). (H) Frequency of WT and Cxcr3^−/− OTII cells in long term (open) short-lived (grey) or without (stripe) DC interactions. (I) Short-lived interactions between REX3+ DCs and either WT or Cxcr3^−/− OTII cells. Duration of contacts from movies imaged 6–8 hr post T cell transfer. Only interactions where the initiation and termination of the interactions were observed were analyzed. Data are representative of 3 independent experiments (n=3–6). (See Figure S3A,B for confirmation that C57BL/6 mice do not produce CXCL11 protein; Figure S3C,D for RNA expression of Cxcl9 and Cxcl10 in whole dLN or non-dLNs in mice receiving OTII cells or not; Figure S3E–G for confirmation of Cxcl10^−/− DC tracking and function; Figure S3H,I for BAC Tg construction and PCR genotyping of REX3 Tg mice; Figure S3J–M; Movie S1 for representative imaging performed 6–8 hr post T cell transfer and Figure S4 for quantification of imaging performed at 24–27 hr post T cell transfer).

Figure 4. CXCR3 is required for optimal Th1 differentiation following OVA and TLR-ligand immunization

(A) Experimental protocol. Host mice were immunized with OVA and LPS and PolyI:C prior to i.v. co-transfer of WT and Cxcr3^−/− OTII cells. CD62L blocking antibody 2 hr following OTII cell. 60 hr post T cell transfer, dLNs T cells were harvested to assess cytokine production. (B) Plots and (C) fold change of IFN-γ and TNF-α production by WT and Cxcr3^−/− OTII cells. (D) Total numbers of polyfunctional (producing IL-2, TNF-α and IFN-γ) WT and Cxcr3^−/− OTII cells in dLNs. Line connects paired WT and Cxcr3^−/− OTII cells transferred into the same WT host. (F–G) LN sections from REX3 Tg mice. CXCL9-RFP, red; CXCL10-BFP, blue; B220 and CD19 immunostaining white. (E) Unimmunized REX3 Tg LN. (F) REX3 Tg dLNs post immunization. Right panels shows grey scale of B220 and CD19 immunostaining (top), CXCL9-RFP (middle) and CXCL10-BFP (lower) of LN harvested at 36 hr post T cell transfer. Bar indicates 500 µm. Data are representative of 3 independent experiments (n=3–6).

Figure 5. CXCL9 and CXCL10 have non-redundant roles in promoting OTII cell IFN-γ responses following host immunization

(A–B) BM chimeras created with (A) REX3 Tg BM into WT hosts and (B) WT BM into REX3 Tg hosts. Reconstituted mice were immunized and transferred with OTII cells. 24–36 hr post T cell transfer, dLNs reporter protein expression (CXCL9-RFP, red; CXCL10-BFP, blue; B220 and CD19 immunostaining white). Bar indicates 500 µm. Higher magnification images are from regions indicated (I, 2, 3). (C) WT, Cxcl9^−/− and Cxcl10^−/− host mice were immunized into the footpad 24 hr prior to adoptive transfer of WT OTII cells. At 60 hr post T cell transfer dLNs were harvested and restimulated to assess cytokine production. Plots of IFN-γ and TNF-α production by WT OTII cells transferred into WT, Cxcl9^−/− or Cxcl10^−/− hosts. (D) Frequency of TNF-α⁺IFN-γ⁺ transferred OTII cells in WT (black), Cxcl9^−/− (open) and Cxcl10^−/− (grey) hosts 60 hr following immunization. (E) BM chimeras of WT hosts reconstituted with WT (black), Cxcl9^−/− (open), Cxcl10^−/− (grey) BM. Chimeras were immunized and transferred with WT OTII cells. At 60 hr post T cell transfer, dLNs were harvested to assess cytokine production. Fold change of frequency of TNF-α⁺IFN-γ⁺ transferred cells in indicated BM chimeras is shown. (F) BM chimeras of WT (black), Cxcl9^−/− (open), Cxcl10^−/− (grey) hosts reconstituted with WT BM. As in E, dLN cytokine production. Fold change of frequency of TNF-α⁺IFN-γ⁺ transferred cells in indicated BM chimeras is shown. Data are representative of 2–3 independent experiments (n=4–8).

Figure 6. CXCR3 ligands determine intranodal location of newly activated OTII cells following immunization

(A) Immunized and unimmunized WT hosts received Actin-GFP WT or Cxcr3^−/− OTII cells. T cell location 24–36 hr post T cell transfer. (B–C) Unimmunized (top panels) and immunized (lower panels; 2 representative LNs shown) WT hosts received co-transferred labeled WT (CMTMR, orange) and Cxcr3^−/− (CMFDA, green) OTII cells. T cell location 24–36 hr post T cell transfer. Regions of LN were determined as indicated in Figure S5. (B) Representative snapshots of WT and Cxcr3^−/− location in dLNs. (C) Quantification of WT and Cxcr3^−/− OTII cell location in dLN 24–36 hr post T cell transfer. † indicates p<0.05; * indicates p<0.001 between WT and Cxcr3^−/− OTII cells. (D–E) Immunized and unimmunized WT, Cxcl9^−/−, and Cxcl10^−/− hosts received transferred WT Actin-GFP OTII cells. T cell location 24–36 hr post T cell transfer. (D) Representative snapshots of WT and Cxcr3^−/− location in dLNs. (E) Quantification of WT OTII cell locations in indicated hosts with and without immunization. * indicates p<0.05 between WT and Cxcl9^−/− hosts; † indicates p<0.05 between WT and Cxcl10^−/− hosts. Bar indicates 200 µm.

Figure 7. CXCR3 is required for maximal endogenous antigen-specific Th1 cell differentiation during infection

8 days following i.v. LCMV infection, splenocytes from mice were harvested, restimulated, and tetramer enriched for detection of (A) LCMV gp66 tetramer-specific cells and (B,C) IFN-γ and TNF-α production. (D,E) BM chimeras of mixed WT and Cxcr3^−/− BM in Rag1^−/− hosts were infected with LCMV and harvested for detection of IFN-γ and TNF-α production from LCMV-tetramer-positive cells. Data are representative of 2 independent experiments (n=4).