Chemokine CCL19 promotes type 2 T cell differentiation and allergic airway inflammation

Abstract

Background:

Allergic asthma is driven largely by allergen-specific type 2 T helper (Th2) cells, which develop in regional lymph nodes upon the interaction of naïve CD4⁺ T cells with allergen-bearing dendritic cells (DCs) that migrate from the lung. This migration event is dependent on CCR7 and its chemokine ligand, CCL21. However, is has been unclear whether the other CCR7 ligand, CCL19, has a role in allergic airway disease.

Objective:

We sought to define the role of CCL19 in Th2 differentiation and allergic airway disease.

Methods:

CCL19-deficient mice were studied in an animal model of allergic asthma. DCs or fibroblastic reticular cells (FRCs) from wildtype (WT) and CCL19-deficient mice were cultured with naïve CD4⁺ T cells, and cytokine production was measured by ELISA. Recombinant CCL19 was added to CD4⁺ T cell cultures, and gene expression was assessed by RNA-sequencing and qPCR. Transcription factor activation was assessed by flow cytometry.

Results:

Lungs of CCL19-deficient mice had less allergic airway inflammation, reduced airway hyperresponsiveness, and less IL-4 and IL-13 production compared with lungs of CCL19-sufficient animals. Naïve CD4⁺ T cells co-cultured with CCL19-deficient DCs or FRCs produced lower amounts of type 2 cytokines than did T cells co-cultured with their WT counterparts. Recombinant CCL19 increased phosphorylation of STAT5 and induced expression of genes associated with Th2 and IL-2 signaling pathways.

Conclusion:

Our results reveal a novel, Th2-inducing function of CCL19 in allergic airway disease and suggest that strategies to block this pathway might help to reduce the incidence or severity of allergic asthma.

Graphical Abstract

Capsule summary:

Chemokine CCL19 produced by dendritic cells and lymph node fibroblastic reticular cells promotes Th2 cell differentiation by activating STAT5 in naïve CD4⁺ T cells upon activation, thereby promoting allergic airway inflammation.

INTRODUCTION

Allergic asthma is a complex disease stemming from maladaptive immune responses against a variety of environmental allergens, including pollen, house dust mite and mold in house dust¹. Inhalation of such environmental allergens triggers the development of type 2 immune responses, characterized by type 2 helper T (Th2) cells that produce IL-4, IL-5 and IL-13². Together, these cytokines drive eosinophilic inflammation, mucus production and airway hyperresponsiveness (AHR), which is a cardinal feature of asthma that contributes to the difficulty in breathing (dyspnea) experienced by asthmatics. Thus, a comprehensive understanding of asthma requires knowledge of the mechanisms the give rise to Th2 differentiation. All T helper cell lineages differentiate from naïve CD4⁺ T cells that are stimulated by antigen-bearing dendritic cells (DCs), but the nature of those T helper cells depend largely on the cytokines to which they are exposed during development. DCs are an important source of such cytokines, which include IL-12 that drives Th1 differentiation; IL-1β, IL-6 and TGF-β that induce Th17 cells; and TGF-β and retinoic acid that promote regulatory T cell (Treg) differentiation. However, it is still unclear whether soluble factors derived from DCs are critical for Th2 differentiation. IL-4 is both necessary and sufficient to direct Th2 differentiation³, but this cytokine is primarily produced by T cells, not DCs, and DCs lacking the Il4 gene can still drive Th2 differentiation^4–6. TSLP and IL-33 derived from epithelial cells can upregulate OX40L, thereby augmenting the ability of DCs to induce Th2 cells^{3, 6, 7}, but DCs are capable inducing Th2 differentiation without such additional factors⁸. Moreover, although production of OX40L has been demonstrated in GM-CSF-induced bone marrow derived DCs (BMDCs), expression of Tnfsf4 (the gene encoding OX40L) in tissue DCs is very low (ImmGene; http://rstats.immgen.org/Skyline/skyline.html). Therefore, the ability of DCs to induce Th2 cells, even without their production of cytokines such as IL-12, IL-6, IL-1β or TGF-β, suggests that naïve CD4⁺ T cells might be intrinsically programmed to differentiate into Th2 cells and do not require exogenous, Th2-polarizing factors⁹.

The migration of antigen-bearing conventional DCs (cDCs) to regional lymph nodes (LNs) is an important step in the induction of adaptive immune responses. This migration is critically dependent on the CC-motif chemokine receptor 7 (CCR7) and its chemokine ligands that recruit mature Ccr7-expressing DCs and naïve T cells to T cell areas of LNs¹⁰. The CCR7 ligand CC-motif chemokine 21 (CCL21) is produced by lymphatic endothelial cells, high endothelial venules and by fibroblastic reticular cells (FRCs) in LNs and spleen, whereas the other CCR7 ligand, CCL19, is produced by FRCs and DCs^{11, 12}. CCR7 and CCL21 are required for the accumulation of naïve T cells and DCs in tissue-draining LNs^{13, 14}, but CCL19 is dispensable for this regard^15–17. Despite their defect in DC migration, mice lacking Ccr7 have enhanced allergic responses in mouse models of asthma¹⁸, probably because they have a defect in regulatory T cell development and are therefore unable to control the inflammation^{10, 19}. A similar phenotype is also observed in paucity of lymph node T cells (plt) mice, which lack both CCL19 and CCL21^{20, 21}. However, it has remained unclear what role, if any, CCL19 has in allergic airway inflammation.

In this study, we examined the role of CCL19 in Th2 cell differentiation and in environmentally relevant mouse models of asthma. We found that mice lacking CCL19 had reduced allergic airway inflammation and AHR compared with their CCL19-sufficient counterparts. In vitro, CCL19 promoted Th2 differentiation from naïve T cells through the activation of STAT5 phosphorylation in CD4⁺ T cells. These findings identify specific roles of CCL19 in the development of allergic asthma.

METHODS

Mice

C57BL/6J, Ccl19^–/– (B6.129X1-Ccl19^tm1/Cys/J), CD45.1 (B6.SJL-Ptprc^a Pepc^b/BoyJ), Il4^–/– (B6.129P2-Il4^tm1Cgn/J), OT-II TCR transgenic (B6.Cg(TcraTcrb)425Cbn/J) and C57-BL/6-plt/plt mice (B6N.DDD-plt/NknoJ) were purchased from Jackson Laboratories. Ccr7^gfp/gfp mice (C57BL/6-Ccr7^tm1.1Dnc) were generated as previously described²². All mice were backcrossed to C57BL/6J mice. CD45.1 OT-II mice were generated by crossing CD45.1 and OT-II TCR transgenic mice. Mice were bred and housed in specific pathogen-free conditions at the NIEHS and were used between 6 and 12 weeks of age in accordance with guidelines provided by the Institutional Animal Care and Use Committees. Both male and female mice were used in this study.

Allergic sensitization and mouse model of asthma

Sterile, filtered house dust extracts (HDE) were prepared from dust as described previously²³. For allergic sensitization, mice were lightly anesthetized with isoflurane and given 2 oropharyngeal (o.p.) aspirations, 1 week apart, of 100 μg LPS-free ovalbumin (OVA) (Worthington Biomedical) with 10 μL HDE (OVA/HDE) in a total volume of 50 μL in PBS²⁴. The sensitized mice were challenged 1 week after the second sensitization by exposing them to an aerosol of 1% OVA (grade V, Millipore Sigma) in PBS for 1 hour. In some experiments, mice were sensitized on two occasions that were 1 week apart with o.p. aspiration of 1 μg house dust mite (HDM) (Dermatophagoides pteronyssinus, Greer), and then challenged with o.p. aspiration of 10 μg HDM. For DC migration assays, mice were sensitized with o.p. aspiration of 10 μM PHK26 dye (Sigma Aldrich) with either 100 ng/mL LPS (Sigma Aldrich) or OVA/HDE. Following euthanasia with intraperitoneal injection of sodium pentobarbital (Vortech Pharmaceuticals) at 48 hours after challenge, bronchoalveolar lavage fluid (BALF) and lung tissue were collected. BALF leukocytes were loaded onto glass slides using Cytospin (Thermo Fisher Scientific) and were stained with hematoxylin and eosin (H & E) for light microscopy. Left lobes of mouse lungs were formalin fixed, and sections of paraffin-embedded tissues were stained with either H & E, or Alcian blue. For morphometric quantification of airway mucus burden, Alcian blue-stained sections of the left lung lobe were scanned with Olympus VS120 whole slide scanner, and Alcian blue-stained area was quantitated using Visiopharm image analysis software (Visiopharm) on airway-targeted regions of interest (ROI) drawn along the epithelial basal lamina. All airways (proximal and distal) in a section were included in the analysis. Right lobes were incubated in 500 μL of complete RPMI1640 (cRPMI-10) containing 10% fetal bovine serum (FBS) (Hyclone), 10 mM Hepes, 50 μM β-mercaptoethanol, penicillin and streptomycin (Thermo Fisher Scientific) for 24 hours. IFN-γ, IL-4, IL-13 and IL-17 in culture supernatants were measured by ELISA (BD Bioscience), and IL-5 was measured by BioPlex (BioRad). OVA-specific IgG₁, IgG_2a and IgE in sera were measured by ELISA (BD Bioscience, BioLegend). AHR was measured by FlexiVent (Scireq) with inhalation of increasing doses (0, 12.5, 25, 50 and 100 mg/mL) of methacholine (Sigma Aldrich), a muscarinic receptor agonist, at 48 hours after challenge. To evaluate Th2 responses in mediastinal lymph nodes (mLNs) following sensitization, 1×10⁷ lymphocytes isolated from LNs and spleens of CD45.1 OT-II mice were adoptively transferred to recipient mice by intravenous injection through tail veins prior to o.p. aspiration of OVA/HDE. Four days later, mLNs were harvested, and 1×10⁶ mLN cells were cultured for 2 days in 200 μL of cRPMI-10 in a 96 well flat bottom plate. Cytokines in culture supernatants were analyzed by ELISA. The results of individual mice are presented as dots in graphs.

Bone marrow (BM) transplantation

BM cells were prepared from femurs, tibiae, humeri and sternum of donor mice. Following red blood cell lysis using ACK buffer, 1×10⁷ total BM cells were injected intravenously into sex-matched, gamma-ray (9 Gy)-irradiated recipient mice. Previous studies have shown that this method leads to >95% reconstitution of donor BM cells^{25, 26}. BM chimera mice were used in experiments 8 weeks after BM transplantation.

Flow cytometric analysis and sorting

Cells were diluted to 10⁶/100 μL and incubated with a non-specific blocking reagent cocktail of anti-mouse CD16/CD32 (2.4G2) and normal mouse and rat serum (Jackson ImmunoResearch) as described previously²⁷. Cell surface molecules were stained with following antibodies (Abs): BUV395-anti-CD4 (RM4–5), and -CD11b (M1/70), eFluor 450-anti-CD31 (390) and -MHC class-II I-A^b (AF6.120.1), BV510-anti-CD11b, CD44 (1M7), CD45 (30-F11), CD88 (20/70), CD103 (2E7) and I-A^b, BV711-anti-CD45.2 (104), TCR Cβ (H57–597), Alexa Fluor 488-anti-CD3ε (145–2C11), -CD19 (605), FITC-anti-CD44, CD45.1 (A20), PerCP-Cy5.5-anti-CD11c (N418 and HL3) and CD25 (3C7), PE-anti-Siglec-F (E50–2440), -CD88 and -Podoplanin (8.1.1), PE-Dazzle594-anti-F4/80 (BM8), -CD115 (AFS98), -CD3ε and -CD19 (605), APC-anti-CD3ε, CD4, CD88, CD317 (129c1) and -Siglec-F, biotinylated anti-CD3ε, -CD19, -CD45 and -F4/80 (BD Biosciences, BioLegend or Thermo Fisher Scientific). Staining with biotinylated Abs was followed by Alexa Fluor 488- or eFluor605-conjugated streptavidin. For intracellular cytokine analysis, mLN cells were incubated with 50 ng/mL PMA, 500 ng/mL ionomycin (Sigma) and brefeldin A (BD Bioscience) for 4 hours in a CO₂ incubator. After surface molecule staining with Abs, cells were fixed and permeabilized using a Fixation/Permeabilization kit according to the manufacturer’s instruction (BD Bioscience) then stained with PE-conjugated anti-IL-4 (11B11) and PerCP-Cy5.5-conjugated anti-IL-13 (W17010B)- Abs. For transcription factor analysis, cultured CD4⁺ T cells were stained for surface molecule staining with Abs, fixed and permeabilized using a Fixation/Permeabilization kit (eBioscience) then stained with PE-conjugated anti-GATA3 (L50–823) Abs. Stained cells were analyzed on LSR-Fortessa flow cytometer (BD Biosciences), and the data were analyzed using FACS Diva (BD Bioscience) and FlowJo (Tree Star) software. For purification, stained cells were sorted using FACS ARIA-II cell sorters (BD Biosciences). Singlet cells were analyzed or purified, and dead cells stained with eFluor 780-conjugated Live/Dead dye (Thermo Fisher Scientific) were excluded from analysis.

Isolation of DCs

Lungs, spleens and mLNs were removed from mice after euthanasia with intraperitoneal injection of sodium pentobarbital. Lungs were perfused by PBS injection into right ventricle. Minced tissues were digested with Liberase TM (Sigma Aldrich), collagenase XI, hyaluronidase and DNase I (Sigma Aldrich) for 30 min. (spleen and mLNs) or 60 min. (lung) as described previously²⁸. To enrich DCs, low density cells from spleen or lung were collected by gradient centrifugation using 14.5 % or 16 % Nycodenz (Accurate Chemical), respectively. Total splenic DCs were purified using an automated magnet-activated cell sorter (AutoMACS, Miltenyi) after incubation with MACS beads-conjugated anti-CD11c Abs (Miltenyi). cDC1 (lung CD103⁺ or spleen CD8a⁺) and CD11b⁺ cDC2 were purified by flow cytometry-based cell sorting.

Isolation of FRCs

MLNs were excised from untreated mice and digested with Liberase TM, dispase (Sigma Aldrich), and DNase I for 30 min. Following red blood cell lysis, total LN cells were cultured in complete DMEM (cDMEM-10) containing 10 % FBS and pyruvate at 1–3×10⁶/mL. Non-adherent cells were removed 3 days later, and the media were changed twice per week. After 3–4 weeks culture, cells were harvested using 0.25 % trypsin, and podoplanin⁺CD45^–CD31^– FRCs were sorted by flow cytometry. Purified FRCs were maintained in vitro culture with cDMEM-10 media.

CD4⁺ T cell culture

Naïve CD4⁺ T cells were purified from skin-draining LNs and spleens by AutoMACS using an Ab cocktail containing anti-mouse CD8α (53–6.7), CD8β (53–5.8), CD11b (M1/70), CD11c (HL3), CD16/32 (2.4G2), CD19 (6D5), CD25 (PC61), CD44 (IM7), CD45R-B220 (RA3–6B2), CD49b (DX5), MHC class-II I-A^b (AF6.120), and Ly-6C/G (RB6–8C5) as previously described²⁸. Naïve CD4⁺ T cells (5×10⁴ cells/well) were cultured in a CO₂ incubator for 5 days in 200 μL Iscove’s modified Dulbecco’s medium (IMDM) containing 10 % FBS (Hyclone or Thermo Fisher Science), 50 μM β-mercaptoethanol, penicillin and streptomycin in 96-well flat-bottom plates (BD Biosciences)²⁹ coated with anti-CD3 (145–2C11) (1 μg/mL) and -CD28 Abs (37.51) (1 μg/mL) (Biolegend) with or without recombinant mouse CCL19 (rCCL19) (R&D Systems). In some experiments, γ-ray irradiated FRCs (10⁴ cells/well) were added to the naïve CD4⁺ T cell culture. In some experiments, naïve CD4⁺ T cells were cultured in 96-well U-bottom plates (BD Biosciences) with or without 10 nM OVA_323–339 and cDCs (5×10³ cells/well). Unless specified, cells were harvested and washed 5 days after culture, and viable cells were counted using Luna-FL cell counter (Logos Biosystems). To elicit effector T cell responses, cultured T cells were incubated (1×10⁵ cells/200 μL/well) for 24 h in a 96-well flat bottom plate coated with anti-CD3ε (1 μg/mL) and -CD28 (1 μg/mL) Abs. To measure CCL19 production, 2×10⁵ irradiated FRCs were cultured with 2×10⁵ naïve CD4⁺ T cells in a 24 well plate coated with anti-CD3ε (1 μg/mL) and - CD28 (1 μg/mL) Abs for 3 days. Cytokines or CCL19 in the supernatant of incubated T cells or FRCs were measured by ELISA.

Analysis of transcription factor activation

Lymphocytes from LNs and spleen or purified naïve CD4⁺ T cells were incubated for 1 hour in cRPMI-10 medium in 96-well flat bottom plate with immobilized anti-CD3ε (1 μg/mL) and CD28 Abs (1 μg/mL) with or without various amounts of recombinant mouse (rm)CCL19 (R&D Systems). In some experiments, naïve CD4⁺ T cells were cultured for 16 hours in cRPMI-10 containing 1% FBS in 96-well flat bottom plates with immobilized anti-CD3ε (0.1 μg/mL) and CD28 Abs (1 μg/mL) with or without rCCL19. After washing, cultured CD4⁺ T cells were incubated for 1 hour in cRPMI containing 1% FBS with or without 1 ng/mL (24 IU) human rIL-2 (BioLegend) together with 10 μg/mL anti-mouse IL-2 Abs (JES6–1A12, BioLegend)³⁰. Harvested cells were fixed and permeabilized with Phosflow Lyse/Fix Buffer I and Phosflow Perm Buffer III (BD Bioscience) according to the manufacturer’s instructions. For lymphocyte analysis, cells were stained with BUV395-anti-CD4 (RM4–5), eFluor450-anti-I-A^b (AF6.120), FITC-anti-TCR Cβ (H57−597), PE-anti-pERK1/2 (20a), and APC-pSTAT1 (4a), -pSTAT4 (38-pSTAT4), -pSTAT5 (47) or pSTAT6 (S71–773) Abs. pSTAT⁺ or pERK⁺ cells among TCR⁺CD4⁺I-A^b– cells were assessed by flow cytometry. For naïve CD4⁺ T cell analysis, cells were stained with PerCP-Cy5.5-anti-CD4 (GK1.5), PE-anti-pERK1/2, and APC-pSTAT5 Abs then pSTAT5⁺ and pERK1/2⁺ cells among CD4⁺ cells were analyzed by flow cytometry.

Quantitative PCR

Naïve CD4⁺ T cells from C57BL/6 were cultured in cRPMI-10 medium in 96-well flat-bottom plates (BD Biosciences) coated with anti-CD3ε (1 μg/mL) and −CD28 Abs (1 μg/mL) with or without 1 ng/mL rmCCL19 (R&D Systems). RNA of cultured CD4⁺ T cells was isolated using NucleoSpin RNA XS kit (Takara Bio) according to the manufacturer’s instructions. mRNA was converted to cDNA using oligo dT primers and a SuperScript III First Strand kit (Thermo Fisher Science). Quantitative PCR (qPCR) amplification was performed using SYBR Green Master Mix (Thermo Fisher Science) on a Mx3000P QPCR system (Agilent Technologies) using the primers shown in Table E1. Relative expression of each gene (arbitrary units) was normalized to that of the house-keeping genes, Ppih or Gapdh.

Transcriptome analysis

RNA from cultured CD4⁺ T cells of C57BL/6 mice was isolated using NucleoSpin RNA XS kit (Takara Bio). Reverse-stranded RNA-Seq libraries with unique barcode adapters were constructed from total RNA using TruSeq RNA sample prep kit (Illumina) according to the manufacturer’s instructions. Multiplexed libraries were sequenced by the NIEHS Epigenomics and DNA Sequencing Core Laboratory on a NovaSeq 6000 (Illumina) as single-end 76-mers. The data were processed using RTA version 3.3.3. Reads were mapped to the mm10 reference genome via STAR version 2.5 (parameters --outMultimapperOrder Random --outSAMattrIHstart 0 --outFilterType BySJout --alignSJoverhangMin 8 --limitBAMsortRAM 55000000000 --outSAMstrandField intronMotif --outFilterIntronMotifs RemoveNoncanonical)³¹. Counts per gene were determined by Subread featureCounts v1.5.0-p1 (parameters: -s2) for a set of gene models defined by RefSeq transcripts (NCBI Curated) as downloaded from the UCSC Table Browser (http://genome.ucsc.edu/cgi-bin/hgTables) as of January 10, 2020³². Differentially expressed genes (DEGs) were identified via DESeq2 v1.28.1 [R 4.0.2] at an FDR threshold of 0.05 with lfcShrink via ‘apeglm’³³. Depth tracks were generated by STAR version 2.5 (parameters --outWigType bedGraph –outWigStrand Stranded --outWigNorm RPM). Pathways of DEGs were analyzed using Ingenuity Pathway Analysis (IPA) (Qiagen). The RNA-seq expression heatmap was generated by ComplexHeatmap (v2.0.0) [R 3.6.1]³⁴ with row-scaling of variance-stabilizing transformation scores as calculated by DESeq2. Trimmed Mean of the M-values (TMM)-normalized counts per million (CPM) values were calculated with EdgeR v3.30.3 [R 4.0.2]³⁵.

Expression of human CCL19, human CCR7, mouse Ccl19 and mouse Ccr7 genes in the lung and LN cells were analyzed using single cell RNA-Seq (scRNA-Seq) datasets (Supplementary materials) downloaded from Cell X Gene Browser (https://cellxgene.cziscience.com)^36–39. Violin plots were generated with R package ggplot2 v3.3.5.

Statistics

Data are presented as mean ± SEM. Statistical differences between groups were calculated by one-way or two-way ANOVA, Kruskal-Wallis test, or two-tailed, unpaired t-test using GraphPad Prism software. Outliers identified by GraphPad Prism ROUT method (Q=1%) were removed from analysis. P<0.05 was considered as statistically significant.

RESULTS

CCL19 promotes allergic airway inflammation

To test the role of CCL19 in the development of allergic airway inflammation, we compared allergic responses of CCL19-sufficient (Ccl19^+/–) and CCL19-deficient (Ccl19^–/–) mice in an environmentally relevant model of asthma. In this model, mice were sensitized on two occasions through the airway using adjuvant-containing HDE together with highly purified OVA (OVA/HDE) (Fig. 1A). Upon subsequent challenge by exposure to aerosolized OVA, these animals develop multiple features reminiscent of allergic asthma, including AHR²³. AHR contributes to the difficulty in breathing (dyspnea) experienced by asthmatics and can be modeled in live mice by measuring their changes in airway resistance in response to increasing doses of inhaled methacholine. AHR was seen in both Ccl19^+/– mice and Ccl19^–/– mice after sensitization and challenge, but this physiologic response was significantly attenuated in Ccl19^–/– mice (Fig. 1B and Fig. E1A). No differences were seen in mice that were not sensitized prior to challenge, indicating that the reduced AHR in Ccl19^–/– mice stems from their attenuated adaptive immune response.

Fig. 1. Allergic airway inflammation in CCL19 KO mice.

(A) Timeline for mouse model of asthma. Ccl19^+/– and Ccl19^–/– male mice were sensitized with OVA/HDE twice, then challenged once with OVA aerosol. (B) AHR. Line plot shows mean airway resistance values for various concentrations of methacholine. Combined results from two experiments are shown. Data were analyzed by one-way ANOVA (n=10–17). (C) Cell numbers for the indicated leukocyte types in BALF as determined by differential microscopy following the indicated treatments. Combined results from two experiments are shown. Data were analyzed by one-way ANOVA (n=10–13). (D) OVA-specific IgE, IgG1 and IgG2a in sera. Data were analyzed by one-way ANOVA (n=10–13). (E) Histological analysis of lungs after OVA challenge. Paraffin sections of lung were stained with either H&E or Alcian blue, and examined by microscopy. Representative photographs for each experimental group are shown (original magnification 200X). Complied quantitative results for mucus producing cells is shown in a right panel. Data were analyzed by one-way ANOVA (n = 6). (F) Cytokine production. Lungs were cultured in the presence of OVA, and cytokines in culture supernatants measured by ELISA or BioPlex. Data were analyzed by one-way ANOVA (n=10–13). Combined results from two experiments are shown.

In this OVA/HDE model of asthma, mice also develop allergen-specific type 2 immune responses, and eosinophilic inflammation of the airway²³. Compared with Ccl19^+/– mice, Ccl19^–/– mice had statistically significant reductions in eosinophils and lymphocytes in the BALF following allergic sensitization and challenge (Fig. 1C), whereas there were no significant differences in neutrophils, whose accumulation in this model is dependent on IL-17 produced by Th17 cells²⁴. No differences in cellular inflammation were seen between WT and Ccl19^+/– mice (Fig. E1B), suggesting that a single copy of functional Ccl19 is sufficient for development of allergic airway inflammation. The decreased eosinophilia in Ccl19^–/– mice was due to a diminished adaptive immune response because animals undergoing OVA challenge only (no sensitization) had very little airway inflammation, and there were no Ccl19 genotype-specific differences in airway leukocytes at baseline (Fig. 1C). Similarly diminished cellular inflammation in Ccl19^–/– mice in comparison with WT mice was also seen in a different asthma model in which house dust mite (HDM) was used as the allergen (Fig. E1C and D). Thus, Ccl19^–/– mice have a defect in type 2 inflammation following allergen sensitization and challenge.

Immunoglobulin isotype class switch to IgE and IgG₁ is promoted by IL-4, while IgG_2a class switch is promoted by IFN-γ^{40, 41}. OVA-specific IgE and IgG₁ in sera was attenuated in Ccl19^–/– mice compared with Ccl19^+/– mice, whereas OVA-specific IgG_2a was very low in all groups tested (Fig. 1D), suggesting that type 2 response is diminished in Ccl19^–/– mice. On the other hand, mucous cell metaplasia, another feature of type 2 inflammation, was comparable between Ccl19^+/– and Ccl19^–/– mice by histological analysis (Fig. 1E). These results suggest that although CCL19 promotes airway inflammation and AHR, it is dispensable for mucus hyperproduction.

As eosinophilia, AHR, and IgE and IgG₁ production are all promoted by type 2 cytokines after undergoing allergic airway inflammation, we evaluated production of IL-4, IL-5 and IL-13 in mouse lungs from OVA/HDE-sensitized and OVA-challenged mice. Ccl19^–/– mice had significantly lower levels of IL-4 and IL-13 than did their similarly treated Ccl19^+/– counterparts (Fig. 1F). IL-5 level was slightly lower in Ccl19^–/– mice, but the difference was not statistically significant, possibly because the amounts of this cytokine were very high and thus not strongly influenced by CCL19. IL-17 and IFN-γ were only moderately reduced in Ccl19^–/– mice and were not significantly different from Ccl19^+/– mice (Fig. 1F). Together, these data indicate that CCL19 promotes type 2 allergic airway inflammation in a mouse model of asthma.

The diminished airway inflammation and AHR in CCL19-deficient mice was unexpected because previous reports had shown that CCR7-deficient mice and plt mice have comparable or even more severe inflammation than WT animals^{18, 20}. We confirmed that CCR7-deficient mice and plt mice also developed strong inflammation in the OVA/HDE model of asthma (Fig. E1E and F). Thus, the phenotype of CCL19-deficient mice is qualitatively distinct phenotype those of CCR7-deficient mice and plt mice.

CCR7, the receptor of CCL19, is highly produced by naïve T cells, but not by innate lymphoid cells (ILCs) (Fig. E2A–C and E3A, B). We therefore questioned whether the diminished allergic airway inflammation in CCL19-deficient mice is due to impaired Th2 generation. Following OVA/HDE allergic sensitization, production of IL-13 from lung-draining mLNs was lower in Ccl19^–/– mice compared with WT mice (Fig. 2A). IL-5 was also significantly lower in mLNs of Ccl19^–/– mice (Fig. 2B). IL-4 production also trended lower in mLNs of Ccl19^–/– mice, but this difference was not statistically significant, possibly because only very low levels of IL-4 were observed (Fig. 2A). As similar results were seen when intracellular IL-4 and IL-13 were assessed in mLN CD4⁺ T cells by flow cytometry, the diminished production of Th2 cytokines seems to be due to reduction of Th2 cells after allergic sensitization (Fig. 2C). These data suggest that CCL19 acts to promote Th2 cell differentiation.

Fig. 2. Th2 responses and cDC migration in mLNs.

(A, B) Cytokine production from mLN cells of Ccl19^+/+ and Ccl19^–/– mice 4 days after sensitization with OVA/HDE. Data were analyzed by one-way ANOVA (n=6) (A) or (n=3–5) (B). (C) Flow cytometric analysis of cytokine producing cells among CD3⁺CD4⁺ T cells in mLNs of Ccl19^+/+ and Ccl19^–/– mice 4 days after sensitization with OVA/HDE. Representative cytograms (left panels) and compiled data (right panels) are shown. Data were analyzed by one-way ANOVA (n=5). (D-H) Migration of cDCs from the lung to mLNs. Mice were instilled with PKH dye and LPS. PKH-labeled cDC1 and cDC2s in mLNs and lung were evaluated by flow cytometry. Shown are data for CCR7-deficient mice (D), plt mice (E), and for CCL19-deficient mice at 24 h post-LPS (F) and at 1, 3, or 6 days post-LPS (G, H). Data were analyzed by unpaired t-test (n=3–4) (D-F) or by one-way ANOVA (n=3) (G, H). N.S.: not significant. Male and female mice were used.

To examine whether the diminished Th2 responses seen in CCL19-deficient mice was related to defective migration of lung cDCs to mLNs, we instilled the fluorescent dye, PKH26, into the airways of mice to label lung resident cDCs. The labeled cDCs were tracked their migration to LNs following lung cDC activation by inhalation of LPS, which, like HDE, induces Ccr7 expression (Fig. E3C). As expected, very few cDCs migrated to mLNs in CCR7-deficient mice or plt/plt mice (Fig. 2D and E). However, cDCs in lungs of Ccl19^–/– mice migrated as efficiently as did their counterparts in WT or Ccl19^+/– mice upon inhalation of either LPS or OVA/HDE (Fig. 2F–H and E3D). These results exclude defective cDC migration as a cause of the reduced allergic airway inflammation in Ccl19^–/– mice and implied that the reduced responsiveness of these mice in the asthma model might instead be related to a different role for CCL19 in priming allergen-specific Th2 responses.

DC-derived CCL19 promotes Th2 differentiation

Previous reports have shown that IL-4 is sufficient and necessary to direct Th2 differentiation³, and in agreement with previous reports^4–6, we found that IL-4 produced by T cells is necessary for Th2 differentiation. By contrast, IL-4 production by cDCs is very low and dispensable for Th2 differentiation (Fig. E3E–H). The TNF super family member 4 gene (Tnfsf4) that encodes OX40L has been reported to promote Th2 responses^{3, 6, 7}. However, we found that cDCs express only very low levels of Tnfsf4 at steady state and that these levels are not increased after allergic sensitization (Fig. E3I). In agreement with a previous report¹⁸, we verified that Ccl19 is expressed in cDCs isolated from the lung and spleen, but not in macrophages (Fig. 3A). By contrast, Ccl21 was not expressed by cDCs. These results, and the diminished Th2 responses in CCL19-deficient mice, prompted us to test for a direct role of cDC-derived CCL19 in Th2 differentiation. Although naïve CD4⁺ T cells proliferated following treatment with anti-CD3 Abs, addition of cDCs to the cultures augmented the magnitude of this proliferation (Fig. 3B). T cell production of the type 2 cytokines, IL-4 and IL-13, was strongly promoted by cDCs, whereas IFN-γ production was not (Fig. 3B). To investigate the contribution of CCL19 to cDC-driven Th2 differentiation, we co-cultured naïve CD4⁺ T cells from OVA-specific TCR transgenic OT-II mice with splenic cDCs from untreated WT or Ccl19^–/– mice that were loaded with OVA_323–339 peptides. cDCs from WT mice were more effective at driving Th2 differentiation than were peptide-loaded cDCs from Ccl19^–/– mice (Fig. 3C). The inefficient induction of GATA3⁺ Th2 differentiation by Ccl19^–/– cDCs was restored by addition of rCCL19 to the co-culture of cDCs and naïve CD4⁺ T cells (Fig. 3D). We next tested the role of CCL19 in Th2 induction by lung cDCs. Naïve CD4⁺ T cells from OT-II mice were co-cultured with cDCs isolated from lungs (Fig. E4A) of either WT or Ccl19^–/– mice following their treatment with OVA/HDE. Although CD103⁺ cDC1s and CD11b⁺ cDC2s from WT or Ccl19^–/– mice could both induce T cell proliferation and Th2 cytokine production, WT cDCs were more effective in these regards (Fig. 3E). These experiments show that CCL19 derived from cDCs promotes Th2 differentiation and suggests that the diminished allergic airway inflammation seen in Ccl19^–/– mice could be due, at least in part, to the lack of CCL19 production by cDCs.

Fig. 3. cDC-derived CCL19 promotes Th2 differentiation.

(A) Expression of Ccl19 and Ccl21 in cDC subsets and in macrophages from the spleen and lung. mRNA expression was measured by qPCR in triplicate test, and Ccl19 and Ccl21 mRNA were normalized to Ppih expression. (B) Naïve CD4⁺ T cells from C57BL/6 mice were cultured with splenic DCs (gray bars) or without DCs (white bars) with the indicated amounts of immobilized anti-CD3 and CD28 Abs. Proliferation of CD4⁺ T cells after 5 days culture. Cytokines produced by cultured CD4⁺ T cells, as measured by ELISA. Data were analyzed by two-way ANOVA (n=3). (C) Naïve CD4⁺ T cells from OT-II mice were cultured with 10 nM OVA_323–339 peptides and splenic cDCs from WT or Ccl19^–/– mice. Proliferation of CD4⁺ T cells was estimated after 5 days culture. Cytokines produced by cultured CD4⁺ T cells were measured by ELISA. Data were analyzed by one-way ANOVA (n=3). (D) Flow cytometric analysis of GATA3⁺ CD4⁺ T cells. Naïve CD4⁺ T cells from OT-II mice were co-cultured with splenic cDCs from WT or Ccl19^–/– mice and 10 nM OVA_323–339 peptides with or without 1 ng/mL rCCL19. GATA3⁺ cells in CD3⁺CD4⁺ T cells were estimated. Representative histograms and compiled data are shown. Data were analyzed by one-way ANOVA (n=4). (E) Naïve CD4⁺ T cells from OT-II mice were cultured with lung cDCs from WT mice (gray bars) or Ccl19^–/– mice (white bars) that received HDE/OVA through the airways. Proliferation of CD4⁺ T cells after 5 days of culture. Cytokines produced by cultured CD4⁺ T cells were measured by ELISA. Data were analyzed by two-way ANOVA (n=3). Male and female mice were used.

Non-hematopoietic stromal cell-derived CCL19 promotes allergic airway inflammation

CCL19 is also produced by non-hematopoietic stromal cells such as FRCs in lymphoid tissues¹². Our analysis of published scRNA-Seq data verified that human CCL19 is expressed in fibroblasts in both lung and LNs (Fig. E5A and B), and mouse Ccl19 expressed in lung fibroblasts (Fig. E5C). Thus, stromal cell-derived CCL19 might also contribute the development of allergic airway inflammation. To assess the contribution of CCL19 derived from hematopoietic and non-hematopoietic cells in the OVA/HDE model of asthma, we generated reciprocal BM chimera mice using Ccl19^+/– and Ccl19^–/– mice as both donors and recipients (Fig. 4A). Ccl19^+/– mice that had received either Ccl19^+/– or Ccl19^–/– mouse BM after allergen sensitization and challenge developed eosinophilia. By contrast, Ccl19^–/– recipient mice developed less eosinophilia, regardless of whether they had received Ccl19^+/– or Ccl19^–/– BM (Fig. 4B). In agreement with this result, allergic sensitization with OVA/HDE increase the production of Th2 cytokines, IL-4 and IL-13, in Ccl19^+/– recipients compared with unsensitized Ccl19^+/– control mice upon challenge, but not in similarly treated Ccl19^–/– recipient mice (Fig. 4C). Clear differences between recipient genotypes were not seen when IL-17 and IFN-γ production were measured. These results suggest that CCL19 derived from non-hematopoietic stromal cells is required for development of allergic airway inflammation in mouse model of asthma.

Fig. 4. Stroma cell-derived CCL19 promotes airway inflammation.

(A) BM chimeric mice were generated by reciprocal adoptive transfer of BM cells between CCL19-sufficient (Ccl19^+/– HT) and -deficient (Ccl19^–/– KO) male mice following γ-ray irradiation. Mice were sensitized with OVA/HDE then challenged with OVA aerosol. (B, C) Cell numbers for the indicated leukocytes in BALF (B), and cytokines in culture supernatants of lungs (C) from mice after challenge. Combined results from two experiments are shown. Data were analyzed by Kruskal-Wallis test (n=6–20). Statistical significance of sensitized BM chimera mice against unsensitized HT or KO control mice is shown.

Our results prompted us to examine the role of CCL19-producing non-hematopoietic stromal cells in Th2 differentiation. FRCs are non-hematopoietic cells in residing in secondary lymphoid tissues, such as LNs and spleen^{12, 18}. These cells can produce a variety of cytokines and chemokines, including CCL19. Because our analysis of scRNA-Seq data for lung and LN cells revealed that fibroblasts in the lung also express human CCL19 and mouse Ccl19 (Fig. E5A–C), we first examined whether Th2 differentiation occurs the lung or mLNs. When measured shortly after allergic sensitization, the Th2 cytokines, IL-4 and IL-13, were increased in mLNs, but not in the lung (Fig. 5A), suggesting that mLN cells have an important role in Th2 differentiation. We therefore tested the impact of CCL19 produced by mLN FRCs on Th2 induction. Although it is reported that FRCs (Podoplanin⁺CD45^–CD31^–) (Fig. E4B) require shear stress for optimal production of CCL19⁴², FRCs from mLNs of WT mice, but not Ccl19^–/– mice, nonetheless maintained some Ccl19 mRNA expression ex vivo (Fig. 5B). To study the contribution of FRC-derived CCL19 on Th2 differentiation, we added FRCs from mLNs either WT and Ccl19^–/– mice to cultures of naïve CD4⁺ T cells that were stimulated with immobilized anti-CD3 and -CD28 Abs. Under this culture condition, FRCs from WT mice, but not Ccl19^–/– mice, co-cultured with CD4⁺ T cells produced CCL19 proteins (Fig. 5C). We deliberately withheld cDCs from the co-cultures of T cells with FRCs to avoid confounding results with CCL19 production by cDCs. In this culture system, WT FRCs enhanced the production of IL-4 and IL-13 from CD4⁺ T cells compared with control cultures with no FRCs, whereas Ccl19^–/– FRCs did not (Fig. 5D). These results suggest that CCL19 produced by FRCs can augment the differentiation of naïve CD4 T cells into Th2 cells.

Fig. 5. CCL19 from FRCs promotes Th2 differentiation.

(A) Th2 cytokine production from mLNs or lung after allergic sensitization with OVA/HDE. MLN or lung cells from C57BL/6 mice that received OT-II cells were cultured 4 days after sensitization, and cytokines in the culture supernatant were analyzed by ELISA. Data were analyzed by two-way ANOVA (n=12). Statistical significance of after sensitization against unsensitized mice is shown. (B) Ccl19 gene expression in cultured mLN FRCs from WT or Ccl19^–/– mice, detected by qPCR using Ccl19atg and Gapdh primers. Data were analyzed by unpaired t-test (n=3). (C, D) Naïve CD4⁺ T cells from WT mice were cultured with or without mLN FRCs from WT or Ccl19^–/– mice with immobilized anti-CD3 and CD28 Abs. (C) CCL19 in the culture supernatant was analyzed by ELISA. (D) Proliferation of and cytokine production from CD4⁺ T cells after 5 days culture. Combined results from two experiments are shown. Data were analyzed by one-way ANOVA (C: n=3–4, D: n=6). Male and female mice were used.

CCL19 alters the expression of Th2 differentiation pathway genes

To investigate CCL19-induced pathways leading to Th2 differentiation, we performed bulk RNA-Seq on naïve CD4⁺ T cells activated by immobilized anti-CD3 and CD28 Abs in the presence or absence of rCCL19. At 6 hours post-CCL19 treatment, 630 genes were upregulated by that chemokine, 538 genes were downregulated, and 14,737 genes were unchanged (FDR 0.05) (Fig. 6A). IPA of DEGs revealed that rCCL19 affected genes in multiple pathways including the ‘Th2 Pathway’ (Fig. 6B). Genes in this pathway whose expression was increased by CCL19 included Il2, Il4, Il5, Il13 and Gfi1 (Fig. 6C and Fig. E6A). The ‘IL-2 Signaling’ pathway was also influenced by rCCL19 (Fig. E6B, E6C). Upregulated genes in this pathway include Il2 and the IL-2 receptor genes Il2ra, Il2rb and Il2rg (Fig. 6D). As IL-2 is essential for Th2 differentiation^{3, 43}, these data suggest CCL19 skews cells towards a Th2 lineage in part by increasing the sensitivity of naïve T cells to IL-2 (Fig. E3B). Interestingly, the Th1-promoting transcription factor genes, Runx3, Stat4, and Mapk3, were all lower in rCCL19-treated cells, although Tbx21 was higher (Fig. 6C). Together, these results suggest that CCL19 upregulates the expression of Th2-promoting genes and suppresses some Th1-promoting genes. Surprisingly, expression of Gata3, which encodes the master regulator of Th2 differentiation, was not increased at this early time point in rCCL19-treated CD4⁺ T cells, nor was that of another Th2-promoting factor, Stat5b (Fig. 6C and E). To determine if rCCL19 has a delayed effect on the expression of some Th2-associated genes, we analyzed their gene expression at a later time point. Five days after naïve CD4⁺ T cell activation in vitro, the Th2-associated genes, Stat5a, Stat5b, Stat6, and Il4ra, were all relatively increased in rCCL19-treated cells (Fig. 6F, Fig. E7B). The expression of Gata3, however, remained unaltered by rCCL19 treatment. Together, these results suggest that CCL19 skews cells toward Th2 differentiation in a manner independent of Gata3 expression, possibly by increasing the sensitivity of T cells to IL-2.

Fig. 6. Transcriptome analysis of CD4⁺ T cells stimulated by CCL19.

Gene expression in naïve CD4⁺ T cells cultured with immobilized anti-CD3 and -CD28 Abs. (A-E) Naïve CD4+ T cells were isolated from C57BL/6 male mice. RNA from cells cultured with or without rCCL19 (1 ng/mL) for 6 hours was analyzed by bulk RNA-Seq (n=3). (A) DEGs between rCCL19-treated and -untreated cells as determined by DESeq2 (FDR: 0.05) and displayed by a volcano plot. (B) Positive Z-score pathways (predicted upregulation) of DEGs as revealed by IPA. Pathways with log (p-value) > 1 are shown. (C) Heatmap displaying the expression level of genes in ‘Th2 Pathway’. (D) Heatmap displaying the expression of genes in ‘IL-2 Pathway’. (E) Compiled gene expression data (TMM-normalized CPM) of Gata3, Stat5a, Stat5b and Stat6 in bulk RNA-Seq analysis of CD4⁺ T cells after 6h of culture. Data were analyzed by unpaired t-test (n=3). (F) Expression of indicated genes in cultured CD4⁺ T cells with or without rCCL19 for 5 days was analyzed by qPCR. Data were analyzed by unpaired t-test (n=3). Representative results from 2 independent experiments are shown.

The RNA-Seq analysis also revealed that S1pr1 expression was reduced in CCL19-treated CD4⁺ T cells (Fig. 6C). Since Sphingosine-1-phosphate (S1P) and its receptor (S1PR1) mediate lymphocyte egress from the LNs⁴⁴, altered expression of S1pr1 might reduce retention of allergen-specific CD4⁺ T cells and thus alter their ultimate responses. However, as the number of OT-II donor CD45.1⁺ CD4⁺ T cells in mLNs were comparable between WT and Ccl19^–/– recipient mice (both CD45.2) after adoptive transfer and OVA/HDE inhalation (Fig. E7C), lacking CCL19 unlikely enhances the egress of allergen-specific CD4⁺ T cells.

CCL19 promotes Th2 differentiation by activating STAT5

Since our bulk RNA-Seq analysis of CD4⁺ T cells revealed that CCL19 promotes Il2 and IL-2 receptor gene expression, we investigated the role of CCL19 in IL-2 production and response. Addition of rCCL19 to cultures of naïve CD4⁺ T cells that were stimulated by immobilized anti-CD3 and -CD28 Abs increased IL-2 production (Fig. 7A). The role of CCL19 in enhancement of IL-2 production from CD4⁺ T cells was further supported by diminished IL-2 production by mLN cells isolated from Ccl19^–/– mice following HDE/OVA inhalation compared with that of WT mice (Fig. 7B). Addition of rCCL19 to cultures of stimulated naïve CD4⁺ T cells also increased the surface level of CD25/IL-2Rα (Fig. 7C), which is in agreement with our RNA-Seq analysis (Fig. 6D), and suggests that CCL19 potentially increases sensitivity of CD4⁺ T cells to IL-2.

Fig. 7. STAT5 is activated by CCL19.

(A) IL-2 in the supernatants of naïve CD4⁺ T cells cultured with indicated amount of rCCL19 for 3 days with immobilized anti-CD3 and -CD28 Abs was measured by ELISA. Combined results from two experiments are shown. Data were analyzed by one-way ANOVA (n=7). Statistical significance for each group against control cells not treated with rCCL19 is shown. (B) IL-2 in the supernatants of mLNs from WT and Ccl19^–/– mice harvested 24 hours after allergic sensitization with OVA/HDE was measured by ELISA. Combined results from three experiments are shown. Data were analyzed by unpaired t-test (n=17). (C) CD25 on CD4⁺ T cell surface was analyzed by flow cytometry. Naïve CD4⁺ T cells were cultured for 16 hours with indicated amount of rCCL19 and immobilized anti-CD3 and −CD28 Abs. Representative histograms and compiled data are shown. Data were analyzed by one-way ANOVA (n=4). (D) Naïve CD4⁺ T cells were cultured for 16 hours with or without rCCL19 and immobilized anti-CD3 and -CD28 Abs. pSTAT5⁺ cells in CD4⁺ T cells induced by IL-2 treatment were analyzed by flow cytometry. Data were analyzed by one-way ANOVA (n=8). Statistical significance against no CCL19 control with IL-2 is shown. Combined results from two experiments are shown. (E, F) pSTATs and pERK1/2 in CD4⁺ T cells. LN and spleen cells (E) or naïve CD4⁺ T cells (F) were activated with immobilized anti-CD3 and -CD28 Abs and indicated amount of rCCL19 for 1 hour. pSTAT⁺ or pERK1/2⁺ cells in CD4⁺ T cells were evaluated by flow cytometry. Data were analyzed by Kruskal-Wallis test (n=6) (E) or one-way ANOVA (n=4–7) (F). Combined results from two experiments are shown. Male and female mice were used.

IL-2 signals through its receptor to activate the transcription factor STAT5, which promotes Th2 differentiation^{3, 45}. Our findings prompted us to examine whether CCL19 promotes STAT5 activation through IL-2 receptor, and we therefore measured STAT5 phosphorylation (pSTAT5) in CD4⁺ T cells upon IL-2 treatment after activation of these cells with immobilized anti-CD3 and CD28 Abs together with rCCL19. Overnight incubation of naïve CD4⁺ T cells with rCCL19 increased pSTAT5 in response to rIL-2 compared with CD4⁺ T cells incubated without rCCL19 (Fig. 7D). This result suggests that CCL19 increases the sensitivity of CD4⁺ T cells to IL-2, probably through upregulation of IL-2 receptor production.

The above findings indicate that CCL19 indirectly activates STAT5 through enhancement of IL-2 and IL-2R production. We next questioned whether CCL19 directly activates signal transducers, which direct helper T cell differentiation, without exogenous IL-2. Another signal transducer STAT6 is activated by IL-4, and promotes Th2 differentiation, while other signal transducing factors, STAT1, STAT4, and extracellular regulated kinase 1/2 (ERK1/2) promote Th1 differentiation^{3, 9}. To determine whether CCL19 directly activates signal transducers, we measured phosphorylation of various STATs and ERK in CD4⁺ T cells treated with various amounts of rCCL19 together with activation of these cells with immobilized anti-CD3 and CD28 Abs. One hour after incubation of LN and spleen cells in vitro, pSTAT5 was significantly increased in CD4⁺ T cells in a CCL19 dose-dependent manner, while phosphorylation of other signal transducers, STAT1, STAT4, STAT6 and ERK1/2, were unaffected by CCL19 (Fig. 7E, E7D). We then tested whether this CCL19-mediated increase in pSTAT5 in CD4⁺ T cells is dependent on TCR signaling. To do this, we incubated naïve CD4⁺ T cells with various amounts of CCL19 in the presence or absence of anti-CD3 and anti-CD28 Abs. This experiment revealed that the CCL19-dependent increase in pSTAT5 does indeed require TCR activation (Fig. 7F). Interestingly, pSTAT5 was not increased by rCCL21 treatment at any dose we tested (Fig. E7E). Taken together, these results show that CCL19 selectively augments Th2 differentiation in TCR-stimulated CD4⁺ T cells by promoting the phosphorylation of STAT5.

DISCUSSION

The ability of cDCs to promote the differentiation of naïve CD4⁺ T cells to Th2 cells has been recognized for decades, but the underlying molecular mechanisms have remained elusive, in part because cDCs do not produce IL-4, the cytokine driving Th2 differentiation. Our current findings reveal a novel molecular mechanism, besides antigen presentation and co-stimulation, that contributes to Th2 differentiation. We demonstrate that the chemokine CCL19 can promote antigen-specific Th2 differentiation in vitro and is required for robust allergic airway inflammation in a mouse model of asthma. Eosinophilia, which is promoted by IL-5, was significantly diminished in Ccl19^–/– mice after allergen challenge. IL-5 production in mLNs was clearly diminished in Ccl19^–/– mice after allergic sensitization, whereas the production of this cytokine from lungs was not diminished in the deficient mice, suggesting that IL-5 production at early stage of allergic responses might be important for eosinophil expansion. Alternatively, as some eosinophils can develop without IL-5 signals in mouse and human⁴⁶, other mechanisms besides IL-5 may contribute to eosinophilia. cDCs lacking CCL19 have a reduced capability to induce Th2 differentiation in vitro, but it remains unclear whether CCL19 production by cDCs also drives Th2 development in vivo. The proximity of cDCs to naïve CD4⁺ T cells in regional LNs suggests that CCL19 production by cDCs might indeed contribute to Th2 development. We also found that CCL19 from FRCs contributes to Th2 development, as FRCs from CCL19-deficient mouse LNs have a lesser capability to promote Th2 differentiation compared with their counterparts from WT mice. FRCs in LNs have been previously linked to the recruitment of naïve CD4⁺ T cells and mature cDCs – both of which express CCR7 – through their secretion of CCL19 and CCL21¹⁰. However, our findings suggest an additional role for the CCL19 produced by FRCs, namely its contribution to Th2 differentiation. Our findings with bone marrow chimera mice suggested that contribution of non-hematopoietic cell-derived CCL19 is more important than cDC-derived CCL19 for allergic airway inflammation. Selective deletion of Ccl19 in different cell types, including cDCs and FRCs, will be necessary to disentangle the role of each of these cell sources in allergic airway inflammation. Although not trivial, such efforts are warranted by the importance demonstrated here of this chemokine to not only Th2 development and airway inflammation, but also AHR, an important clinical aspect of allergic asthma.

Mechanistically, IL-2 promotes Th2 differentiation through STAT5 phosphorylation^{3, 30}. We found that in vitro, CCL19 augments gene expression of Il2 and of IL-2 receptor subunits, Il2ra, Il2rb and Il2rg, and production of CD25/IL-2Rα on and IL-2 from CD4⁺ T cells. In agreement with the IL-2 receptor upregulation, CCL19-treated CD4⁺ T cells augmented pSTAT5 in response to IL-2. These findings therefore suggest that increased sensitivity to IL-2 responses is at least one of the mechanisms of CCL19-promoting Th2 differentiation. Interestingly, CCL19 augmented STAT5 phosphorylation in TCR-stimulated naïve CD4⁺ T cells even without exogenous IL-2, whereas it did not increase phosphorylation of another Th2-promoting transcription factor, STAT6, nor that of the Th1-promoting transcription factors, STAT1, STAT4 and ERK1/2. The selective induction of pSTAT5 is intriguing and suggests that CCL19 has a specific role in augmenting type 2 responses. This is consistent with a previous report showing that CCL19 is dispensable for cytotoxic T cell responses in skin-draining LNs¹⁵. Although we cannot completely rule out the contribution of IL-2 produced from naïve T cells, CCL19 might contribute to STAT5 activation in early stage of CD4⁺ T cell response. Noteworthy, CCL19 upregulated expression of a ‘Th2 pathway’ gene, Gfi1, which encodes growth factor independent-1 (GFI1), a transcriptional repressor promoting Th2 cell expansion, GATA3 stabilization and STAT5 activation^47–49. The contribution of GFI1 in CCL19-mediated Th2 responses, especially IL-2-independent STAT5 activation, are to be investigated in future studies. It is known that pSTAT5 induces expression of Il4, Il4ra and Gata3^{30, 50}, which encode factors driving Th2 differentiation. In in vitro cultured CD4⁺ T cells, Il4ra gene expression was upregulated by CCL19 treatment, suggesting that CCL19 promotes Th2 differentiation through enhancement of sensitivity to IL-4.

We found that another CCR7-ligand chemokine, CCL21, did not increase pSTAT5 in naïve CD4⁺ T cells, possibly because CCL21 has a lower CCR7-binding affinity than CCL19^{51, 52}. A recent study also demonstrated the difference between CCL19 and CCL21 in binding to the 2^nd extracellular loop of CCR7, and only CCL19 induces β-arrestin-2 recruitment⁵³. This unique feature of CCL19 might be relevant to Th2 development and airway inflammation because like CCL19-deficient mice, β-arrestin-2-deficient mice also display impaired Th2 cytokine production and allergic airway inflammation⁵⁴. Additional studies are warranted to more fully characterize how CCR7 and β-arrestin-2 cooperate to promote Th2 differentiation.

Although our findings suggest that CCL19 promotes Th2 responses, it is not the only factor because Ccl19^–/– cDCs still induce some Th2 cytokine production from CD4⁺ T cells, and Ccl19^–/– mice still developed mild allergic airway inflammation. Furthermore, Gata3 upregulation by CCL19 was not evident without cDCs on our experiments. These lines of evidence suggest that other factor(s) besides CCL19 produced by cDCs might be required to upregulate Gata3 expression, and contribute to Th2 differentiation.

The present study reveals a novel role for CCL19 produced by cDCs and LN FRCs in promoting Th2 differentiation and allergic airway inflammation. Our findings contribute to an improved mechanistic understanding of Th2 development, and might lead to novel strategies that employ CCL19 blockade to prevent or treat allergic asthma.

Key Message:

A novel function of CCL19 inducing Th2 cells in allergic airway disease suggests that strategies to block this pathway might help to reduce the incidence or severity of allergic asthma.

Abbreviations used:

AECs

alveolar epithelial cells

AHR

airway hyperresponsiveness

BALF

bronchoalveolar lavage fluid

BM

bone marrow

CCR7

CC-motif chemokine receptor 7

CCL19

CC-motif chemokine 19

DC

dendritic cells

DEG

differentially expressed gene

FBS

fetal bovine serum

FRCs

fibroblastic reticular cells

HDE

house dust extract

HDM

house dust mite

ILCs

innate lymphoid cells

LPS

lipopolysaccharide

LNs

lymph nodes

MACS

magnetic activated cell sorter

o.p.

oropharyngeal

OVA

ovalbumin

plt

paucity of lymph node T cells

qPCR

quantitative polymerase chain reaction

STAT

signal transducer and activation of transcription

TCR

T cell receptor

Th2

type 2 helper T cells

WT

wildtype

Data and materials availability:

Bulk RNA-seq data have been deposited at the National Center for Biotechnology Information in the Gene Expression Omnibus repository under accession number: GSE222700.