Induction and activation of human Th₁₇ by targeting antigens to dendritic cells via Dectin-1

Abstract

Recent compelling evidence indicates that Th₁₇ confer host immunity against a variety of microbes, including extracellular and intracellular pathogens. Therefore, understanding mechanisms for the induction and activation of antigen-specific Th₁₇ is important for the rational design of vaccines against pathogens. To study this, we employed an in vitro system in which influenza HA1 was delivered to dendritic cells (DCs) via Dectin-1 using anti-hDectin-1-HA1 recombinant fusion proteins. We found that healthy individuals maintained broad ranges of HA1-specific memory Th₁₇ that were efficiently activated by DCs targeted with anti-hDectin-1-HA1. Nonetheless, these DCs were not able to induce significant level of HA1-specific Th₁₇ response even in the presence of Th₁₇-promoting cytokines, IL-1β and IL-6. We further found that the induction of surface IL-1R1 expression by signals via TCRs and common γ-chain receptors were essential for naïve CD4⁺ T cell differentiation into HA1-specific Th₁₇. This process was dependent on MyD88, but not IRAK1/4. Thus, interruptions in STAT3 or MyD88 signaling led to substantially diminished HA1-specific Th₁₇ induction. Taken together, the de novo generation of pathogen-specific human Th₁₇ requires complex but complementary actions of multiple signals. Data from this study will help us design new and effective vaccine strategy that can promote Th₁₇-mediated immunity against microbial pathogens.

Introduction

IL-17-producing CD4⁺ T cells (Th₁₇) have been broadly linked to inflammatory diseases (1). However, recent compelling evidence indicates that Th₁₇ also play an important role against both extracellular and intracellular microbial pathogens, including bacteria, fungi, parasites and viruses (2-13). Furthermore, the immunity conferred by Th₁₇ is associated with improved survival of cancer patients (14, 15). Accordingly, Th₁₇-mediated therapeutic immunity has also been demonstrated in murine cancer models (16, 17). Therefore, it is important to understand molecular and cellular mechanisms for the induction and activation of antigen-specific Th₁₇ in the context of T cell receptor (TCR) ligation by peptides and major histocompatibility complexes (MHCs).

The induction of Th₁₇ has been mainly studied in the context of inflammatory cytokine milieu. In mice, TGFβ (18, 19), IL-6 (19), IL-1β (20), IL-21 (21), IL-23 (22), and IL-9 (23) contribute to Th₁₇ induction. In humans, IL-1β with IL-6 was initially reported to induce Th₁₇ differentiation, and this was inhibited by TGFβ and IL-12 (9). TGFβ was also reported to be required for Th₁₇ development (24), but Yang et al. (25) demonstrated that human Th₁₇ could be developed in the presence of TGFβ and IL-21, but not TGFβ and IL-6. In contrast, Volpe et al. (26) showed that pro-inflammatory cytokines were all required and acted synergistically to generate human Th₁₇. These series of findings suggest that each of these cytokines might contribute to Th₁₇ development at certain stages of human T cell differentiation, although a recent finding has shown that IL-1β is essential in Candida albicans-induced human Th₁₇ differentiation (27). Unlike in mice, however, our understanding of the induction and activation of antigen-specific Th₁₇ in humans is still limited. This is mainly due to limitations of reliable experimental systems as well as difficulties in the assessment of antigen-specific T cell responses after in vitro priming of T cells particularly when the frequency of antigen-specific T cells is low. Thus, previous studies (9, 24-27) employed polyclonal T cell activators, such as anti-CD3/CD28 antibodies and phorbol 12-myristate 13-acetate (PMA), to prime and/or reactivate T cells to assess the magnitude and quality of T cell responses. Although these studies led to great progresses in our understanding of human Th₁₇ especially in the context of inflammatory diseases, biology of T cells primed and/or re-activated with such polyclonal activators may not always represent the biology of T cells primed and/or re-activated with MHC II/peptide complexes presented by antigen presenting cells (APCs). Therefore, it is valuable to study the induction and activation of antigen-specific human Th₁₇ in the context of T cell receptor (TCR) ligation by the complexes of MHC II and antigen-derived peptides presented by APCs.

DCs are major APCs that can induce and shape the types of T cell response during microbial infections. DCs express pattern-recognition receptors (PRRs), including toll-like receptors (TLRs) and C-type lectin receptors, which are linked to antimicrobial immunity through the sensing of pathogen-associated molecular patterns (28, 29). Of these PRRs, Dectin-1 is particularly relevant to the Th₁₇-mediated immunity in both mice and humans (3, 7, 30, 31). We and others have shown that DCs can take up protein antigens via Dectin-1 and present antigenic peptides to both CD4⁺ and CD8⁺ T cells (32-34). Thus, we established an in vitro system in which HA1 subunit from hemagglutinin (HA) of influenza virus (H1N1, PR8), as a model antigen, could be delivered to DCs via hDectin-1 using recombinant proteins of an agonistic anti-hDectin-1 fused to HA1. This system allowed us for the first time to dissect the complex and dynamic processes of the generation of HA1-specific human Th₁₇ in the context of TCR ligation with MHC II/peptide complexes presented by DCs. In addition, we demonstrated that antigen targeting to DCs via hDectin-1 along with TLR2 ligands could promote antigen-specific Th₁₇ responses in human.

Materials and methods

Cells and culture medium

Blood from healthy volunteers were acquired under a protocol approved by the Institutional Review Board (IRB) of Baylor Research Institute (BRI). Peripheral blood mononuclear cells (PBMCs) of healthy volunteers were isolated by density gradient centrifugation using Ficoll-Paque™ PLUS (GE Healthcare, Sweden). IFNDCs were generated by culturing monocytes from healthy donors in serum free media (Cellgenix, Germany) supplemented with GM-CSF (100 ng/ml) and IFNα (500 units/ml). The medium was replenished with cytokines on day 1. IFNα and GM-CSF were from the Pharmacy at the Baylor University Medical Center (Dallas, TX). Autologous CD4⁺ T cells were purified using EasySep Human CD4⁺ T Cell Enrichment Kit (StemCell Technologies, Canada). Naïve (CD45RA⁺CD45RO⁻CCR7⁺), memory (CD45RA⁻CD45RO⁺) CD4⁺ T cells, and mDCs (Lin⁻HLA-DR⁺CD11c⁺CD123⁻) were sorted by FACS Aria (BD Biosciences, CA) (purity>99.0%). Culture medium consisted of RPMI 1640 (GIBCO, NY) supplemented with HEPES buffer, 2 mM L-glutamine, 1% nonessential amino-acids, sodium pyvurate, 50 units/ml penicillin, 50 μg/ml streptomycin and 10% normal human serum AB (GemCell, TX).

Antibodies and reagents

Anti-CD4-APC Cy7, anti-IFNγ–PE Cy7, anti-CCR5-pacific blue and anti-CCR6-Alexa Fluor 488 were purchased from Biolegend (CA). Anti-IL-1R1-PE, anti-IL-6R-PerCP, anti-CCR9-PE, anti-CXCR3-FITC, anti-IL-23p19 and control IgG were from R&D Systems (MN). Neutralizing anti-IL-1β and anti-IL-6/IL-6R were made in house. Anti-CD45RA-FITC, anti-CD45RO-PE, anti-integrin β7-PE, and anti-CD161-PE were purchased from BD Biosciences. Anti-IL-17-APC (eBioscience, CA) and anti-human IgG-PE (Jackson ImmunoResearch Laboratories, PA) were used. GolgiPlug was purchased from BD Pharmingen (CA). CFSE (Molecular probes, Oregon) was used for measuring CD4⁺ T cell proliferation. TLR ligands, including Pam(3)CysSK(4) were purchased from Invivogen (Oregon). Jak2 inhibitor (1,2,3,4,5,6-Hexabromocyclohexane), Jak3 inhibitor (4-(4′-Hydroxyphenyl)amino-6,7-dimethoxyquinazoline), pan Jak inhibitor (2-(1,1-Dimethylethyl)-9-fluoro-3,6-dihydro-7H-benz[h]-imidaz[4,5-f]isoquinolin-7-one), STAT5 inhibitor (Pimozide) and IRAK-1/4 inhibitor (N-(2-Morpholinylethyl)-2-(3-nitrobenzoylamido)-benzimidazole) were purchased from EMD chemicals (PA). STAT3 inhibitor (Stattic) was from Sigma-Aldrich (MO). MyD88 homodimerization inhibitory peptide (Imgenex, CA) was used. Anti-CD3/CD28 microbeads were purchased from Miltenyi Biotec (Germany).

Cloning and purification of chimeric recombinant mAbs fused to HA1

The Flu HA1 fragment containing nt 52-993 of Influenza A virus (A/Puerto Rico/8/34/Mount Sinai(H1N1)) hemagglutinin (gi|21693168|gb|AF389118.1|) (with A619C and C959T changes) was engineered with a flanking NheI site followed by a flexible sequence (GATACAACAGAACCTGCAACACCTACAACACCTGTAACAACA) at the 5-prime end and 6 X His tag coding sequence followed by stop codon and NotI site at the 3-prime end (35). The NheI and NotI enzyme-digested fragment was fused downstream of and in-frame with the heavy chain coding region at the NheI site by NheI-NotI. Stable CHO-S cells were grown in GlutaMAX and HT media (Life Technologies, NY) and recombinant proteins were purified by protein A colum chromatography. Purified proteins were confirmed by reduced-SDS gel analysis.

Peptides

Overlapping 17-mer peptides (staggered by 11 amino acids) spanning the entire HA1 subunits of hemagglutinin (HA) (A/PR/8/34 H1N1) were purchased from Biosynthesis (TX).

Binding of recombinant fusion proteins to hDectin-1 and hDEC205

293F cells transfected with a full-length hDectin-1 or hDEC205 and IFNDCs were incubated with different concentrations of recombinant fusion proteins (anti-hDectin-1-HA1, anti-hDEC205-HA1, and IgG4-HA1) for 25 min on ice. Cells were then washed twice with 2% FCS in phosphate-buffered saline (PBS), and then stained with anti-human IgG-PE for 20 min. Cells were analyzed by flow cytometry.

DC activation

1×10⁵ DCs were incubated with 1 μg/ml anti-hDectin-1-HA1 or 1 μg/ml IgG4-HA1 in the presence or absence of 40 ng/ml Pam(3)CysSK(4) for 24h. Cytokines in the culture supernatants were assessed by the BeadLyte cytokine assay kit (Upstate, MA) as per the manufacturer’s protocol. IL-23 was measured using a human IL-23 ELISA kit (eBiosciences, CA).

DC/CD4⁺ T cell co-cultures and antigen-specific CD4⁺ T cell responses

1-2×10⁵ CFSE-labeled purified autologous CD4⁺ T cells were co-cultured with 5×10³ DCs loaded with indicated antibodies, in the absence or presence of different TLR ligands. After 7 days, CD4⁺ T cell proliferation was measured by CFSE-dilution. In some experiments, anti-IL-23p19, anti-IL-6 and anti-IL-6R, anti-IL-1β antibodies or control IgG (10 μg/ml) were added into the co-cultures of DCs and CD4⁺ T cells. In some experiments, CD4⁺ T cells were incubated overnight with 50 units/ml IL-2 (Hoffmann-LaRoche, NJ), 50 units/ml IL-7 (R&D Systems, MN), 50 ng/ml IL-15 (Peprotech, NY) and 50 ng/ml IL-21 (Invitrogen, NY) and then naïve CD4⁺ T cells were FACS-sorted to obtain IL-1R1⁻ and IL-1R1⁺ cells. For assessing antigen-specific CD4⁺ T cell responses, T cells were restimulated with 0.5-1 μM indicated peptides for 6h in the presence of Brefeldin A, and stained with 7-AAD, anti-CD4, anti-IFNγ and anti-IL-17 antibodies labeled with fluorescent dyes. CD4⁺ T cells expressing intracellular IFNγ and IL-17 were detected by flow cytometry (FACS Canto, BD). In separate experiments, CD4⁺ T cells were restimulated with indicated peptides for 48h, and cytokines in the supernatants were assessed by the BeadLyte cytokine assay kit (Upstate, MA) as per the manufacturer’s protocol.

Conventional and quantitative Real-time RT-PCRs

Total RNA was isolated from CD4⁺ T cells using Ambion’s RNAqueous kits (Life Technologies) and cDNA was synthesized with Reverse Transcription System (Promega, CA). Conventional RT-PCR was performed for TBX21, RORC, GATA3, IL-1R1, IL-1B and ACTB (β-actin) using the primers, TBX21 (forward, 5′-GAGGGGCGGGTCCTCGACGG-3′; reverse, 5′-TCGCGGCGGGTAGGCGTAGG-3′), GATA3 (forward, 5′-AACTGTGGGGCAACCTCGAC-3′; reverse, 5′-TTGCAGACAGGGTCCCCATT-3′), RORC (forward, 5′-TCTGGAGCTGGCCTTTCATCATCA-3′; reverse, 5′-TCTGCTCACTTCCAAAGAGCTGGT-3′), IL-1R1 (forward, 5′-CCACAGCCCAAGGGCGGGGCTA-3′; reverse, 5′-CTGCAGCACCTCTCAGGAGAGCCGC-3′), IL1B (forward, 5′-GCAAGGGCTTCAGGCAGGCCG-3′; reverse, 5′-GCTGTGAGTCCCGGAGCGTGC-3′) and ACTB (forward, 5′-CTCGCCTTTGCCGATCCGCCGC-3′; reverse, 5′-GCTCTGGGCCTCGTCGCCCACAT-3′). IL1B, RORC, IL-1R1 and ACTB were also quantified through real-time RT-PCR using the Lightcycler 480 machine (Roche Applied Bioscience, NJ) using SYBR Green master mix (Roche). Expression levels of individual molecules were normalized to the amount of ACTB mRNA.

Statistical Analysis

Statistical significance was determined using the Student’s t-test and significance was set at p<0.05.

Results

Generation and characterization of recombinant protein of HA1 fused to anti-hDectin-1 mAb

To deliver HA1 to DCs via hDectin-1, a recombinant fusion protein (anti-hDectin-1-HA1) of an agonistic anti-hDectin-1 (34) and HA1 was made on a mouse variable region-human IgG4κ chimera with two site mutations (S228P and L235E) that further diminishes antibody binding to the Fc receptor (36). IgG4-HA1 with the same mutations was made as a control. The putative structure of the recombinant fusion protein is presented in Supplemental Fig. 1. Recombinant proteins were expressed from stable CHO cell lines, purified by Protein A chromatography, and confirmed by the reduced SDS-gel analysis (Fig. 1A). Anti-hDectin-1-HA1, but not IgG4-HA1, specifically bound to 293F cells transfected with hDectin-1 (Fig. 1B). Neither anti-hDectin-1-HA1 nor IgG4-HA1 bound to mock transfectants (data not shown). Anti-hDectin-1-HA1 (upper panel in Fig. 1C) but not IgG4-HA1 (lower panel in Fig. 1C) also bound to the surface of monocyte-derived IFNDCs in a dose-dependent manner. Taken together, HA1 could be efficiently delivered to DCs via hDectin-1 using anti-hDectin-1-HA1 fusion proteins.

FIGURE 1.

DCs loaded with anti-hDectin-1-HA1 can elicit HA1-specific Th₁, Th₂, and Th₁₇ responses. (A) Reduced SDS-PAGE analysis of anti-hDectin-1 mAb (lane 1), anti-hDectin-1-HA1 (lane 2) and IgG4-HA1 (lane 3). (B and C) Binding of anti-hDectin-1-HA1 and IgG4-HA1 to 293F cells transfected with hDectin-1 (B) and IFNDCs (C). (D) Total CD4⁺ T cell proliferation induced by IFNDCs alone or IFNDCs loaded with 1 μg/ml anti-hDectin-1-HA1, IgG4-HA1, anti-hDectin-1, or combination of IgG4-HA1 (1 μg/ml) and anti-hDectin-1 (1 μg/ml). (E) CFSE-labeled total CD4⁺ T cells were co-cultured for 7 days with 1 μg/ml anti-hDectin-1-HA1-loaded IFNDCs or blood mDCs. T cells were restimulated with 1 μM of the clusters (upper panels, IFNDCs) or individual peptides from cluster 8 (middle panels, IFNDCs and lower panels, mDCs). (F) CD4⁺ T cells were restimulated for 48h in the presence or absence of 1 μM peptides indicated. Cytokines in the supernatants were assessed. In (B-F), three independent experiments showed similar results. (G) CFSE-labeled total CD4⁺ T cells were co-cultured for 7 days with 1 μg/ml anti-hDectin-1-HA1-loaded IFNDCs. T cells were restimulated with 1 μM of the indicated peptides and IL-17 and IFNγ expression were analyzed by flow cytometry. Two independent experiments using cells from donors #1 and #2 showed similar results. One representative data generated with donor #1 are presented. (H) 1×10⁵ IFNDCs were incubated overnight with 1 μg/ml HA1 fusion proteins. The amount of IL-23 in the supernatants was assessed. (I) Total CD4⁺ T cells were co-cultured with 1 μg/ml anti-hDectin-1-HA1-loaded IFNDCs for 7 days in the presence of 5 μg/ml anti-IL-23p19 or control antibodies. CD4⁺ T cells were then restimulated with 1 μM peptides for 48h. Cytokines in culture supernatants were assessed. (J and K) Experiments in (H and I) were performed with blood mDCs. In (H-K), Error bars indicate mean±SD of triplicate assay.

DCs loaded with anti-hDectin-1-HA1 can elicit HA1-specific Th₁₇ responses

Consistent with their bindings to DCs (Fig. 1C), anti-hDectin-1-HA1-loaded IFNDCs induced greater proliferation of CD4⁺ T cells than did IgG4-HA1-loaded IFNDCs (Fig. 1D, upper panel). Anti-hDectin-1 mAb alone or combination of anti-hDectin-1 and IgG4-HA1 induced slightly enhanced CD4⁺ T cell proliferation compared to DCs alone (Fig. 1D, lower panel). To test antigen specificity of the proliferating CD4⁺ T cells, T cells were restimulated with clusters of HA1 peptides (6 peptides in each cluster; 17-mers overlapping by 11 amino acids) (Fig. 1E, upper panels). Cluster 8 showed the greatest percentage of IFNγ⁺CD4⁺ T cells. Individual peptides in cluster 8 were further tested in separate experiments (Fig. 1E, middle panels), and pep 43 and pep 45 resulted in substantially increased percentage of IFNγ⁺CD4⁺ T cells. Pep 5 from cluster 1 and no peptide were negative controls. Experiments performed with blood myeloid DCs (mDCs) showed similar results (Fig. 1E, lower panels), although IFNDCs were more efficient than mDCs at eliciting CD4⁺ T cell responses. IFNDCs also resulted in a greater by-stander proliferation of CD4⁺ T cells than did mDCs. We (35) and others (37) previously showed that DCs induced by-stander T cell proliferation, as shown in Fig. 1D, and this by-stander proliferation is further enhanced when the antigen-specific T cells are activated (35, 37).

Total CD4⁺ T cells co-cultured with anti-hDectin-1-HA1-loaded IFNDCs (Fig. 1F, upper panels) or mDCs (Fig. 1F, lower panels) were restimulated with pep 43, pep 45, control pep 5 or no peptide for 48h and the amounts of IFNγ, IL-13, and IL-17 in the supernatants were assessed. They secreted increased amount of IFNγ and IL-13 in response to pep 43 and pep 45. In addition, they also secreted increased amount of IL-17 in response to both pep 43 and pep 45 compared to control pep 5 or no peptide. Although the level of IL-17 was lower than those of IFNγ and IL-13, our data suggested the presence of HA1-specific Th₁₇ in the cultures. In separate experiments using cells from the same donor (donor #1), we were also able to detect HA1-specific IL-17-producing CD4⁺ T cells (Fig. 1G, left panel), although the frequency of IL-17⁺CD4⁺ T cells were far less than that of IFNγ⁺CD4⁺ T cells (Fig. 1G, right panel) that are specific for HA1.

IL-23 secreted from IFNDCs (Fig. 1H) and mDCs (Fig. 1J) contributed to HA1-specific Th₁₇ and Th₁ responses (Fig. 1I, 1K), as blocking IL-23p19 during the co-cultures of DCs and CD4⁺ T cells resulted in decreased IL-17- and slightly decreased IFNγ-producing CD4⁺ T cell responses. DCs loaded with anti-hDectin-1 alone did not result in HA1-specific CD4⁺ T cell responses (data not shown).

Taken together, we concluded that anti-hDectin-1-HA1-loaded DCs can efficiently activate HA1-specific IFNγ-, IL-13-, and IL-17-producing CD4⁺ T cells. HA1-derived peptides used throughout this study were characterized by performing the experiments in Fig. 1E and their corresponding HLA class II types are summarized in Table 1.

Table 1.

MHC class II types of healthy subjects and peptide binding scores of individual peptides to corresponding MHC class II

Donors	HLA types	Peptides	Amino acid residues	Peptide sequence	ARB score (IC50 values)
#1 (Fig. 1E, 1F, 1G, 1H; Fig. 2A, 2B; Fig. 3A, 3B; Fig. 7B, 7C, 7D)	HLA-DRB1*01	pep 43	250-266	LEPGDTIIFEANGNLIA	10.7
		pep 45	262-278	GNLIAPWYAFALSRGFG	9.6
	HLA-DQB1*05:01	NA			NA
#2 (Fig. 2B, 2D; Fig.4A 4C, 4D; Fig. 6A, 6B, 6C; Fig. 7A, 7B, 7C)	HLA-DRB1*13:02	pep 43	250-266	LEPGDTIIFEANGNLIA	0
		pep 45	262-278	GNLIAPWYAFALSRGFG	10004
	HLA-DRB1*15:01	pep 43	250-266	LEPGDTIIFEANGNLIA	41.6
		pep 45	262-278	GNLIAPWYAFALSRGFG	54.6
	HLA-DRB3*03:01	NA			NA
	HLA-DQB1*06:04	NA			NA
#3 (Fig. 2B, 2C, 2D; Fig, 4D, 4E; Fig. 7C)	HLA-DRB1*01	pep 22	126-142	SSFERFEIFPKESSWPN	826
		pep54	316-332	IDECPKYVRSAKLRMVT	6.7
	HLA-DRB1*13	pep 22	126-142	SSFERFEIFPKESSWPN	25784.9
		pep54	316-332	IDECPKYVRSAKLRMVT	1354
	HLA-DQB1*03:01	NA			NA
	HLA-DQB1*05	NA			NA
#4 (Fig. 2B; Fig. 4D Fig. 7B)	HLA-DRB1*04:01	pep 43	250-266	LEPGDTIIFEANGNLIA	346.8
		pep 45	262-278	GNLIAPWYAFALSRGFG	10.4
	HLA-DRB1*09	pep 43	250-266	LEPGDTIIFEANGNLIA	378.9
		pep 45	262-278	GNLIAPWYAFALSRGFG	6.1
	HLA-DQB1*03:01	NA			NA
	HLA-DQB1*03:03	NA			NA
#5 (Fig. 1I, 1J; Fig. 2D; Fig. 4E)	HLA-DRB1*04:02	NA			NA
	HLA-DRB1*13	pep32	186-202	EKEVLVLWGVHHPPNIG	351.1
	HLA-DQB1*06:04	NA			NA
	HLA-DQB1*03:02	NA			NA
#6 (Fig. 7B)	HLA-DRB1*07	pep 43	250-266	LEPGDTIIFEANGNLIA	103.4
		pep 45	262-278	GNLIAPWYAFALSRGFG	1.6
	HLA-DRB1*08	pep 43	250-266	LEPGDTIIFEANGNLIA	79.5
		pep 45	262-278	GNLIAPWYAFALSRGFG	596.9
	HLA-DQB1*02	NA			NA
	HLA-DQB1*04:02	NA			NA
#7 (Fig. 7B)	HLA-DRB1*01	pep 22	126-142	SSFERFEIFPKESSWPN	826
		pep 54	316-332	IDECPKYVRSAKLRMVT	6.7
	HLA-DRB1*11	pep 22	126-142	SSFERFEIFPKESSWPN	50.6
		pep 54	316-332	IDECPKYVRSAKLRMVT	6.3
	HLA-DQB1*03	NA			NA
	HLA-DQB1*05	NA			NA
#8 (Fig. 7C)	HLA-DRB1*03	pep 22	126-142	SSFERFEIFPKESSWPN	16976.1
		pep 54	316-332	IDECPKYVRSAKLRMVT	726.5
	HLA-DRB1*11	pep 22	126-142	SSFERFEIFPKESSWPN	50.6
		pep 54	316-332	IDECPKYVRSAKLRMVT	6.3
	HLA-DQB1*02	NA			NA
	HLA-DQB1*06	NA			NA
# 9 (Fig. 7C)	HLA-DRB1*03	pep 7	37-53	LEKNVTVTHSVNLLEDS	478.4
		pep 45	262-278	GNLIAPWYAFALSRGFG	53285.8
		pep 46	268-284	WYAFALSRGFGSGIITS	61616.3
		pep 52	304-320	SSLPFQNVHPVTIGECP	7485.1
	HLA-DRB1*07	pep 7	37-53	LEKNVTVTHSVNLLEDS	13.6
		pep 45	262-278	GNLIAPWYAFALSRGFG	1.6
		pep 46	268-284	WYAFALSRGFGSGIITS	1.6
		pep 52	304-320	SSLPFQNVHPVTIGECP	13.6
	HLA-DQB1*02	NA			NA
# 10 (Fig. 7C)	HLA-DRB1*07	pep 22	126-142	SSFERFEIFPKESSWPN	2641.5
		pep 45	262-278	GNLIAPWYAFALSRGFG	1.6
		pep 46	268-284	WYAFALSRGFGSGIITS	1.6
		pep 52	304-320	SSLPFQNVHPVTIGECP	13.6
	HLA-DRB1*11	pep 22	126-142	SSFERFEIFPKESSWPN	50.6
		pep 45	262-278	GNLIAPWYAFALSRGFG	104.5
		pep 46	268-284	WYAFALSRGFGSGIITS	104.5
		pep 52	304-320	SSLPFQNVHPVTIGECP	2981.4
	HLA-DQB1*02	NA			NA
	HLA-DQB1*03	NA			NA
# 11 (Fig. 7D)	HLA-DRB1*03	pep 7	37-53	LEKNVTVTHSVNLLEDS	478.4
		pep 45	262-278	GNLIAPWYAFALSRGFG	53285.8
		pep 46	268-284	WYAFALSRGFGSGIITS	61616.3
		pep 52	304-320	SSLPFQNVHPVTIGECP	7485.1
	HLA-DRB1*11	pep 22	126-142	SSFERFEIFPKESSWPN	50.6
		pep 45	262-278	GNLIAPWYAFALSRGFG	104.5
		pep 46	268-284	WYAFALSRGFGSGIITS	104.5
		pep 52	304-320	SSLPFQNVHPVTIGECP	2981.4
	HLA-DQB1*02	NA			NA
	HLA-DQB1*03	NA			NA

Peptides were predicted by the algorithm in the web: (http://tools.immuneepitope.org/analyze/cgi-bin/mhc_II_binding.py).

NA: alleles are not available in the algorithm.

Anti-hDectin-1-HA1-loaded DCs expand memory Th₁₇ but cannot efficiently induce HA1-specific Th₁₇

To test whether anti-hDectin-1-HA1-loaded DCs could induce HA1-specific Th₁₇, FACS-sorted naïve and memory CD4⁺ T cells were co-cultured with 1 μg/ml anti-hDectin-1-HA1-loaded IFNDCs for 7 days. CD4⁺ T cells were then restimulated with indicated peptides for 48h and the amounts of IFNγ, IL-13, and IL-17 in the supernatants were assessed. Compared to unstimulated or control pep 5-stimulated CD4⁺ T cells, both naïve (Fig. 2A, upper left and middle) and memory CD4⁺ T cells (lower left and middle) stimulated with pep 22 secreted significantly increased amounts of both IFNγ and IL-13. Interestingly, however, increased amount of IL-17 by pep 22 was observed only in the culture supernatant of memory, but not naïve CD4⁺ T cells. Data from experiments using cells from five healthy donors further demonstrated that IFNDCs loaded with anti-hDectin-1-HA1 could efficiently activate HA1-specific memory Th₁₇ cells, but they could not efficiently prime HA1-specific Th₁₇ responses (Fig. 2B). Similar observations were made from experiments using blood mDCs (Fig. 2C, 2D). Naïve and memory CD4⁺ T cells were co-cultured with 1 μg/ml anti-hDectin-1-HA1-loaded mDCs for 7 days. CD4⁺ T cells were then restimulated for 48h with indicated peptides. Pep 45-stimulated naïve T cells secreted increased amounts of both IFNγ (Fig. 2C, upper left panel) and IL-13 (Fig. 2C, upper middle panel), but not IL-17 (Fig. 2C, upper right panel), compared to naïve T cells stimulated with control pep 5 or no peptide. Pep 45-stimulated memory T cells secreted increased amounts of the all three cytokines, including IL-17, compared to memory T cells stimulated with control pep 5 or no peptide (Fig. 2C, lower panel). Data from experiments using cells from 3 healthy donors (Fig. 2D) further showed that anti-hDectin-1-HA1-loaded mDCs efficiently activate HA1-specific memory Th₁₇, but they are not efficient to induce naïve CD4⁺ T cell differentiation into HA1-specific Th₁₇. Thus, we concluded that DCs loaded with anti-hDectin-1-HA1 can efficiently activate HA1-specific memory Th₁₇, but cannot efficiently prime HA1-specific Th₁₇ responses in vitro.

FIGURE 2.

DCs loaded with anti-hDectin-1-HA1 activate memory Th₁₇, but cannot efficiently induce HA1-specific Th_17. (A and C) FACS-sorted naïve and memory CD4⁺ T cells were co-cultured for 7 days with anti-hDectin-1-HA1-loaded IFNDCs (A) or blood mDCs (C). T cells were restimulated for 48h in the presence or absence of indicated peptides (1 μM). Pep 5 was used as a control peptide. Cytokines in the supernatants were assessed. Error bars indicate mean±SD of triplicate assay. (B and D) Summarized data generated with IFNDCs (B) and blood mDCs (D) from healthy donors. Background values acquired with pep 5 were subtracted. P values were acquired by the Student’s t-test. Indications of donors and peptides tested in (B and D) are summarized in the table.

Phenotypes of HA1-specific Th₁₇ are different from those of other Th₁₇

HA1 peptide-specific IFNγ⁺CD4⁺ and IL-17⁺CD4⁺ T cells were further characterized by assessing the expression levels of chemokine receptors, β7 integrin, and CD161 (Fig. 3A). Compared to IFNγ⁺CD4⁺ T cells (Fig. 3A, left panel) expressing CXCR3, IL-17⁺CD4⁺ T cells (Fig. 3A, right panel) expressed increased levels of CCR6 and CD161 (38, 39). Interestingly, fractions of the HA1-specific IL-17⁺CD4⁺ T cells also expressed a high level of CCR9, which could support the presence of Th₁₇ in the gut mucosa (40). Fractions of IFNγ⁺CD4⁺ and IL-17⁺CD4⁺ T cells expressed β7 integrin. CCR5 was similarly expressed on both IFNγ⁺CD4⁺ and IL-17⁺CD4⁺ T cells.

FIGURE 3.

Phenotypes of HA1-specific and PMA/ionomycin-activated Th₁₇. Total CD4⁺ T cells were co-cultured for 7 days with 1 μg/ml anti-hDectin-1-HA1-loaded IFNDCs. T cells were restimulated with 1 μM pep 43 (A) or PMA/ionomycin (B) for 6h. Cells were stained for intracellular IFNγ and IL-17 as well as surface receptors indicated. Four independent experiments using cells from two different healthy donors showed similar results.

We also assessed the expression levels of these receptors on IFNγ⁺CD4⁺ and IL-17⁺CD4⁺ T cells restimulated with PMA and ionomycin (Fig. 3B). Similar to the HA1-specific IFNγ⁺CD4⁺ T cells (Fig. 3A, left panel), PMA/ionomycin-induced IFNγ⁺CD4⁺T cells (Fig. 3B, left panel) expressed CCR6, CXCR3, β7-integrin, and CD161. Both pep 43- (Fig. 3A, right panel) and PMA/ionomycin-induced IL-17⁺CD4⁺ T cells (Fig. 3B, right panel) also expressed similar levels of these receptors. However, PMA/ionomycin-induced IL-17⁺CD4⁺ T cells did not express CCR9 that was expressed on pep 43-induced IL-17⁺CD4⁺ T cells. Pep 43-induced IFNγ⁺CD4⁺ and IL-17⁺CD4⁺ T cells also expressed increased levels of CCR5 compared to those induced with PMA/ionomycin.

In conclusion, these data demonstrated that phenotypes of HA1-specific human Th₁₇ cells are not the same to those of other Th₁₇ cells activated with polyclonal activators.

IL-1β and IL-6 are insufficient to induce naïve CD4⁺ T cell differentiation into HA1-specific Th₁₇

Both IL-1β with IL-6 have been known to contribute to the induction of human Th₁₇ differentiation (9, 27). We have shown that anti-hDectin-1 mAb and TLR2 ligand synergistically act on DCs to secrete IL-6 and IL-1β (34). Synergistic actions of signals via Dectin-1 and TLR2 can also increase IL-23, but decrease IL-12 secretion from DCs (41), thereby favoring Th₁₇ responses (9). Thus, we first tested whether DCs loaded with anti-hDectin-1-HA1 plus TLR2 ligand could enhance HA1-specific Th₁₇ responses (Fig. 4A). Total CD4⁺ T cells were co-cultured for 7 days with IFNDCs loaded with anti-hDectin-1-HA1 in the presence or absence of different concentrations of Pam(3)CysSK(4). CD4⁺ T cells were then restimulated for 48h with the indicated peptides and cytokines in the supernatants were assessed. HA1-specific Th₁₇ (Fig. 4A, right panel) and Th₁ responses (Fig. 4A, left panel) were enhanced by Pam(3)CysSK(4). Of the concentrations tested, 40 ng/ml Pam(3)CysSK(4) resulted in the highest levels of IL-17 production. Accordingly, DCs loaded with 1 μg/ml anti-hDectin-1-HA1 plus 40 ng/ml Pam(3)CysSK(4) secreted increased amounts of both IL-1β (Fig.4B, left panel) and IL-6 (Fig. 4B, right panel).

FIGURE 4.

IL-1β and IL-6 can promote HA1-specific memory Th₁₇ responses but fail to induce naïve CD4⁺ T cell differentiation into HA1-specific Th_17. (A) Total CD4⁺ T cells were cultured for 7 days with 1 μg/ml anti-hDectin-1-HA1-loaded IFNDCs in the presence of different concentrations of Pam(3)CysSK(4). T cells were then restimulated for 48h with 1 μM pep 43 or pep 5 (control peptide). Cytokines in the supernatants were assessed. Error bars indicate mean±SD of triplicate assay. Two independent experiments showed similar results. (B) IL-1β and IL-6 levels in the culture supernatants of 1×10⁵ IFNDCs incubated with 1 μg/ml anti-hDectin-1-HA1, 40 ng/ml Pam3CysSK4, or 1 μg/ml anti-hDectin-1-HA1 plus 40 ng/ml Pam(3)CysSK(4). Data are pooled from three independent experiments. (C and D) FACS-sorted naïve and memory CD4⁺ T cells were co-cultured for 7 days with IFNDCs (C) and blood mDCs (D) loaded with 1 μg/ml anti-hDectin-1-HA1 in the presence or absence of 40 ng/ml Pam(3)CysSK(4). T cells were restimulated for 48h with 1 μM peptides. The amount of IL-17 in the supernatants was assessed. In (C) and (D), background values generated with pep 5 were subtracted. P values were acquired by the Student’s t-test. Indications of donors and peptides tested in (C and D) are summarized in the table.

We next tested whether DCs loaded with anti-hDectin-1-HA1 plus 40 ng/ml Pam(3)CysSK(4) could induce HA1-specific Th₁₇ (Fig. 4C, 4D). FACS-sorted naïve (CD45RA⁺CD45RO⁻CCR7⁺) and memory CD4⁺ T cells (CD45RA⁻CD45RO⁺) were co-cultured with either IFNDCs or blood mDCs loaded with anti-Dectin-1-HA1 in the absence or presence of 40 ng/ml Pam(3)CysSK(4). Both IFNDCs (Fig. 4C) and mDCs (Fig. 4D) were able to activate HA1-specific memory Th₁₇, and these Th₁₇ responses were further enhanced by Pam(3)CysSK(4). However, neither IFNDCs nor blood mDCs induced HA1-specific Th₁₇ even in the presence of Pam(3)CysSK(4). Thus, the HA1-specific Th₁₇ responses observed in Fig. 4A were due to the activation of HA1-specific memory CD4⁺ T cells. TLR2 is known to be expressed in human T cells (42), but Pam(3)CysSK(4) alone did not induce CD4⁺ T cells to secrete significant amount of cytokines tested, although it slightly increased the amount of TNFα secreted by anti-CD3/CD28-activated CD4⁺ T cells (data not shown). DCs washed after Pam(3)CysSK(4) treatment were still able to enhance anti-hDectin-1-HA1-loaded DC-mediated HA1-specific Th₁₇ responses (data not shown). Therefore, the enhanced HA1-specific memory Th₁₇ responses by the combination of anti-hDectin-1-HA1 and Pam(3)CysSK(4) were due to the activation of DCs via Dectin-1 and TLR2.

Taken together, we concluded that DCs loaded with anti-hDectin-1-HA1 and TLR2 ligand could greatly promote HA1-specific memory Th₁₇ responses. However, they were not efficient to induce naïve CD4⁺ T cell differentiation into HA1-specific Th₁₇ even in the presence of IL-1β and IL-6. These results suggested that naïve CD4⁺ T cells require additional signals to differentiate into HA1-specific Th₁₇.

Naïve CD4⁺ T cells express IL-6R, but not IL-1R1 that is inducible by synergistic actions of signals via common γ-chain receptors and TCRs

Despite IL-1β and IL-6 secretion by DCs after stimulation with anti-hDectin-1-HA1 and TLR2 ligand, there was no significant enhancement of the induction of HA1-specific Th₁₇ (Fig. 4C, 4D). We therefore investigated the expression levels of cytokine receptors on the surface of naïve and memory CD4⁺ T cells in peripheral blood of healthy donors. Both naïve and memory CD4⁺ T cells expressed similar levels of surface IL-6R (Fig. 5A, left panel). However, surface IL-1R1 was detected only on memory CD4⁺ T cells (Fig. 5A, right panel). The presence of IL-6 in the culture medium (Fig. 4B) and the constitutive expression of IL-6R on naïve CD4⁺ T cell (Fig. 5A, top left panel), but no significant level of HA1-specific Th₁₇ response (Fig. 4C, 4D) suggested that IL-6 is not the key factor that turns on the program for the differentiation of naïve CD4⁺ T cells into HA1-derived peptide-specific Th₁₇. Thus, we hypothesized that the induction of surface IL-1R1 expression on naïve CD4⁺ T cells could be a prerequisite for the enhanced de novo induction of antigen-specific Th₁₇.

FIGURE 5.

Naïve CD4⁺ T cells do not express IL-1R1 that is inducible by synergistic actions of signals from TCRs and common-γ chain receptors. (A) IL-1R1 and IL-6R expression on naïve and memory CD4⁺ T cells in the peripheral blood. CD4⁺ T cells from 6 healthy donors showed similar results. (B) Real-time (lower panel) and conventional (upper panel) RT-PCR analysis of IL-1R1 expression in naïve CD4⁺ T cells treated for 18h with indicated cytokines. (C) Surface IL-1R1 expression on naïve CD4⁺ T cells treated for 36h with the combination of the cytokines in (B). (D and E) Real-time RT-PCR analysis of IL-1R1 expression in naïve CD4⁺ T cells treated with the combination of the cytokines in (B) in the presence or absence of indicated inhibitors. Data are pooled from three independent experiments. P values were acquired by the Student’s t-test. (F and G) Real-time (lower panel) and conventional (upper panel) RT-PCR analysis of IL-1R1 expression in naïve CD4⁺ T cells co-cultured overnight with IFNDCs alone, IFNDCs loaded with HA1 peptide (Pep 43, 1 μM), or IFNDCs loaded with 1 μg/ml anti-hDectin-1-HA1 in the presence or absence of the combination of common γ-chain cytokines (F) or different amounts of anti-CD3/CD28-coated microbeads (G). (F) Error bars indicate mean±SD of duplicate assay of two independent experiments. In (B), (C), and (G), Error bars indicate mean±SD of triplicate assay and three independent experiments showed similar results. P values were acquired by the Student’s t-test.

Induction of IL-1R1 expression on the surface of naïve CD4⁺ T cells is known to be very modest and takes 5-6 days after TCR stimulation in the presence of IL-2 (43). Naïve CD4⁺ T cells from cord blood started to express IL-1R1 on 6 days of culture in the presence of TGFβ, IL-7 and IL-15 (44). Therefore, we first sought cytokine signaling that could rapidly induce IL-1R1 expression on the surface of naïve CD4⁺ T cells. Fig. 5B showed that naïve CD4⁺ T cells from the peripheral blood of healthy donors expressed IL-1R1, although it was not detected on the cell surface (Fig. 5A). IL-1R1 expression was upregulated within 18h by treating naïve CD4⁺ T cells with the combination of IL-2, IL-7, IL-15, and IL-21 (Fig. 5B). In contrast to the previously published data (44), naïve CD4⁺ T cells treated with the combination of common γ-chain cytokines also expressed surface IL-1R1 within 36h (Fig. 5C) and this was consistent with the mRNA expression levels (Fig. 5B). The individual cytokines showed variable effects on IL-1R1 expression (Fig. 5B), but they were not able to induce surface IL-1R1 expression on naïve CD4⁺ T cells (data not shown). This was further confirmed by the data showing that Janus kinase 3 (Jak 3) and pan-Jak inhibitors but not Jak 2 inhibitor suppressed the common-γ chain cytokine-induced IL-1R1 upregulation (Fig. 5D). Accordingly, STAT3 inhibitor resulted in decreased expression of IL-1R1 (Fig. 5E). These data indicated that IL-2, IL-7, IL-15 and IL-21 act synergistically to induce surface IL-1R1 expression on naïve CD4⁺ T cells.

TCR ligation by peptide/MHC complexes is an inevitable process for the induction of antigenic peptide-specific T cell responses. Thus, we tested whether TCR signaling contributed to the IL-1R1 expression in naïve CD4⁺ T cells (Fig. 5F, 5G). FACS-sorted naïve CD4⁺ T cells were co-cultured for 18h with IFNDCs alone, anti-hDectin-1-HA1-loaded IFNDCs, and HA1 peptide-loaded IFNDCs in the presence or absence of the combination of common γ-chain cytokines (Fig. 5F). DCs loaded with either anti-hDectin-1-HA1 or HA1 peptide slightly, but significantly (p= 0.0107 and p= 0.0239 for anti-Dectin-HA1 and pep43 loaded IFNDCs, respectively) increased the expression of IL-1R1 in naïve CD4⁺ T cells. However, the combination of common γ-chain cytokines was more efficient than the antigen-loaded DCs at inducing IL-1R1 expression in naïve CD4⁺ T cells. This was not surprising because the frequency of HA1 antigen-specific CD4⁺ T cells in the naïve pool is low. In contrast, the combination of common γ-chain cytokines acts on majority of the naïve T cells in the cultures.

To further explore the roles of signaling via TCR in the induction of IL-1R1 expression, naïve CD4⁺ T cells were cultured with different ratios of anti-CD3/CD28-coated beads. Fig. 5G shows that anti-CD3/CD28 microbeads increased IL-1R1 expression and this increase was anti-CD3/CD28 dose-dependent. Thus, naïve CD4⁺ T cells treated with the beads at 1:2 ratio expressed higher levels of IL-1R1 than those treated with the beads at 1:8 ratio, demonstrating that IL-1R1 expression is positively correlated to the strength of TCR signaling. Furthermore, anti-CD3/CD28-induced upregulation of IL-1R1 expression was enhanced by the combination of common γ-chain cytokines, showing a synergistic effect of signals from common γ-chain receptors and TCRs on naïve CD4⁺ T cells to upregulate IL-1R1 expression.

Taken together, naïve CD4 T cells express IL-6R, but not IL-1R1 on their surface. Nevertheless, IL-1R1 expression can be rapidly induced by the synergistic actions of signals via common γ-chain receptors and TCRs.

IL-1R1⁺, but not IL-1R1⁻, naïve CD4⁺ T cells differentiate into HA1-specific Th₁₇

To test whether anti-hDectin-1-HA1-loaded DCs were able to induce IL-1R1⁺ naïve CD4⁺ T cell differentiation into HA1-specific Th₁₇, IL-1R1⁻ and IL-1R1⁺ naïve CD4⁺ T cells were co-cultured with anti-hDectin-1-HA1-loaded IFNDCs. After 7 days, both IL-1R1⁻ and IL-1R1⁺ naïve CD4⁺ T cells secreted IFNγ and IL-13 in response to pep 43 during 48h restimulation (Fig. 6A, left and middle panels). HA1-specific Th₁₇ responses were observed only in the cultures of IL-1R1⁺ naïve CD4⁺ T cells (Fig. 6A, right panel). We were not able to detect HA1-specific IL-17⁺CD4⁺ T cells by FACS. However, we observed that fraction of IL-1R1⁺ naïve CD4⁺ T cells co-cultured with DCs expressed intracellular IL-17 in response to PMA/ionomycin (Fig. 6A). Blocking IL-1β during the co-cultures of IL-1R1⁺ naïve CD4⁺ T cells and DCs substantially impaired HA1-specific Th₁₇ responses (Fig. 6B, right panel). Thus, the surface IL-1R1 induction on naïve CD4⁺ T cells by signals via both TCR and common γ-chain receptors is a crucial for the enhanced induction of HA1-specific human Th₁₇.

FIGURE 6.

Anti-hDectin-1-HA1-loaded DCs induce IL-1R1⁺ naïve CD4⁺ T cell differentiation into HA1-specific Th_17. (A) IL-1R1⁻ and IL-1R1⁺ naïve CD4⁺ T cells were co-cultured for 7 days with anti-hDectin-1-HA1-loaded IFNDCs. (left panels) T cells were then restimulated for 48h with 1 μM pep 43 or pep 5. Cytokines in the supernatants were assessed. (right panels) T cells were then restimulated for 6h with PMA and ionomycin. Cells were stained for intracellular IFNγ and IL-17 expression. (B) Experiments in (A) were performed with IL-1R1⁺ naïve CD4⁺ T cells in the presence or absence of indicated antibodies (10 μg/ml). T cells were restimulated for 48h with 1 μM pep 43 or pep 5. In (A) and (B), three independent experiments showed similar results. Error bars indicate mean±SD of triplicate assay. (C) Real time RT-PCR analysis of RORC expression in IL-1R1⁺ naïve CD4⁺ T cells co-cultured for 5 days with 1 μg/ml anti-hDectin-1-HA1-loaded IFNDCs in the absence or presence of indicated inhibitors. (D) IL-1R1⁻ and IL-1R1⁺ naïve CD4⁺ T cells were incubated with anti-CD3/CD28-coated microbeads (beads: cells=1:40) in the absence or presence of indicated cytokines for 5 days. mRNA was analyzed by conventional RT-PCR (left panel) and real time RT-PCR (right panel). Two independent experiments showed similar results. Error bars indicate mean±SD of triplicate assay. (E) IL-1R1⁻ and IL-1R1⁺ naïve CD4⁺ T cells were incubated for 4h with 50 ng/ml IL-6, anti-CD3/CD28-coated microbeads or combination of IL-6 and anti-CD3/CD28-coated microbeads. IL1B expression was assessed by real time RT-PCR. Error bars indicate mean±SD of triplicate assay. One representative of three experiment is shown. (F) IL-1R1⁺ naïve CD4⁺ T cells were incubated with anti-CD3/CD28-coated microbeads in the presence of indicated antibodies for 5 days. mRNA was analyzed by real time RT-PCR. In (C) and (F), Data are pooled from three independent experiments. P values were acquired by the Student’s t-test.

Blocking IL-6 also greatly decreased the induction of HA1-specific Th₁₇ from IL-1R1⁺ naïve CD4⁺ T cells (Fig. 6B, right panel). Thus, STAT3 inhibitors, which block IL-6-mediated signals, also decreased RORC expression in the IL-1R1⁺ naïve CD4⁺ T cells on day 5 of the co-cultures with anti-hDectin-1-HA1-loaded DCs (Fig. 6C). Fig. 6C also demonstrated that IL-1β-induced RORC expression is dependent on myeloid differentiation primary response gene 88 (MyD88), but not IL-1R-associated kinase 1 (IRAK1) or IRAK4.

Taken together, we concluded that both IL-1β and IL-6 contribute to the induction of HA1-specific Th₁₇. However, surface IL-1R1 expression on naïve CD4⁺ T cells is essential for the enhanced induction of HA1-specific Th₁₇. Surface IL-1R1⁻ naïve CD4⁺ T cells did not efficiently differentiate into HA1-specific Th₁₇ (Fig. 2 and Fig. 6A).

IL-6 contributes to the induction of HA1-specific Th₁₇ through IL-1β induction in TCR-activated T cells

IL-6 has been known to contribute to Th₁₇ induction in both mice (19) and humans (9). However, IL-6 can contribute to the induction of HA1-specific Th₁₇ only when naïve CD4⁺ T cells express IL-1R1 (Fig. 6B), although IL-6R is expressed on the surface of naïve CD4⁺ T cells (Fig. 5A, upper left panel). To further investigate the roles of IL-6 in the induction of Th₁₇, both IL-1R1⁻ and IL-1R1⁺ naïve CD4⁺ T cells were activated with anti-CD3/CD28-coated microbeads at a 1:40 ratio in the presence or absence of IL-1β, IL-6, or both IL-1β and IL-6 (Fig. 6D). Anti-CD3/CD28-coated microbeads at 1:40 did not upregulate IL-1R1 expression in naïve CD4⁺ T cells (data not shown). After 5 days, the expression levels of RORC, TBX21, and GATA3 in the two groups of naïve CD4⁺ T cells were assessed. Consistent with the data in Fig. 6A, IL-1R1⁻ naïve CD4⁺ T cells expressed both TBX21 and GATA3, but not RORC, in any tested condition. The expression levels of TBX21 and GATA3 in IL-1R1⁻ CD4⁺ T cells treated with either IL-6 alone or combination with IL-1β were variable in different experiments (data not shown). However, IL-1R1⁺ naïve CD4⁺ T cells expressed RORC in the presence of either IL-1β or IL-6 alone. IL-1β plus IL-6 induced the greatest level of RORC expression. In line with the data in Fig. 6B (right panel) and 6C, IL-6 alone could also induce RORC expression in IL-1R1⁺ naïve CD4⁺ T cells (Fig. 6D, right panel). Indeed, IL-6 was more efficient than IL-1β at inducing RORC expression in IL-1R1⁺ naïve CD4⁺ T cells. Therefore, we hypothesized that IL-6 could induce these naïve CD4⁺ T cells to express IL-1β followed by the induction of RORC expression, and thus could amplify IL-1β-induced RORC expression. As shown in Fig. 6E, exogenous IL-6 was able to induce IL-1R1⁻ and particularly IL-1R1⁺ naïve CD4⁺ T cells to upregulate IL1B expression. Furthermore, blocking IL-1β led to decreased IL-6-mediated RORC expression in IL-1R1⁺ naïve CD4⁺ T cells (Fig. 6F).

Taken together, IL-1R1⁺ naïve CD4⁺ T cells were able to differentiate into HA1-derived peptide-specific Th₁₇ when they were activated via both TCRs and IL-1R1. IL-6 could argument Th₁₇ responses by inducing IL-1β and thus enhanced RORC expression in the TCR-activated IL-1R1⁺ naïve CD4⁺ T cells.

Targeting antigens to DCs via Dectin-1 can elicit antigen-specific human Th₁₇ responses that can be further promoted by TLR2 and TLR4 ligands

Antigen targeting to DCs via Dectin-1 using recombinant fusion protein of an agonistic anti-hDectin-1 mAb and antigen could elicit antigen-specific Th₁₇ responses. IL-23 secreted from anti-hDectin-1-HA1-loaded DCs contributed to HA1-specific Th₁₇ responses (Fig. 1H, 1J). In addition, anti-hDectin-1-HA1 and TLR2 ligands synergized to activate DCs to secrete IL-1β and IL-6, resulting in the enhanced HA1-specific Th₁₇ responses (Fig. 4A, 4B).

We further tested the effects of other TLR ligands on the magnitude of HA1-specific Th17 responses elicited by DCs loaded with anti-hDectin-1-HA1 (Fig. 7A). Compared to other TLR ligands tested, Pam(3)CysSK(4) was the most efficient at enhancing the HA1-specific Th₁₇ responses, as it synergizes with anti-hDectin-1-HA1 to induce increased amounts of both IL-1β and IL-6 secretion from DCs (Fig. 4B). The enhanced HA1-specific Th₁₇ responses by the combination of anti-hDectin-1-HA1 and TLR2 ligands were further confirmed (Fig. 7B, 7C). Total CD4⁺ T cells from eight healthy donors were co-cultured for 7 days with DCs loaded with 1 μg/ml anti-hDectin-1-HA1 in the absence or presence of TLR2 ligands, Pam(3)CysSK(4) (Fig. 7B) and P. gingivalis lipopolysaccharide (LPS) (Fig. 7C). Consistent with the data in Fig. 4B and Fig. 7A, both Pam(3)CysSK(4) and P. gingivalis LPS promoted HA1-specific Th₁₇ responses elicited by anti-hDectin-1-HA1-loaded DCs. E. coli LPS was also able to enhance HA1-specific Th₁₇ responses elicited by anti-hDectin-1-HA1-loaded IFNDCs (Fig 7D, upper panels) and mDCs (Fig. 7D, lower panels). We previously reported that anti-hDectin-1 mAb and E. coli LPS synergize to activate DCs to secrete increased amount of both IL-1β and IL-6 (34). Taken together, healthy individuals tested in this study maintained HA1-specific memory Th₁₇, although the magnitudes of Th₁₇ responses in the healthy individuals were variable. In addition, a vaccine model made of an agonistic anti-hDectin-1 and antigens along with TLR2 or TLR4 ligands could offer great potential in promoting Th₁₇-mediated host immunity against infections.

FIGURE 7.

Healthy individuals maintain influenza HA1-specific Th₁₇ that can be promoted by the combination of anti-hDectin-1-HA1 and TLR2 ligands. (A) Total CD4⁺ T cells were co-cultured for 7 days with IFNDCs loaded with 1 μg/ml anti-hDectin-1-HA1 in the presence or absence of indicated TLR ligands (40 ng/ml Pam(3)CysSK(4), 1μg/ml poly IC, 50 ng/ml flagellin, 500 ng/ml R848). T cells were then restimulated for 48h with 1 μM of indicated peptides. Cytokines in the culture supernatants were assessed. Two independent experiments showed similar results. (B and C) Total CD4⁺ T cells from healthy donors were co-cultured for 7 days with 1 μg/ml anti-hDectin-1-HA1-loaded IFNDCs in the presence or absence of 40 ng/ml Pam(3)CysSK(4) (B) or P. gingivalis LPS (C). T cells were then restimulated with 1 μM indicated peptides. Cytokines in the culture supernatants were assessed. (D) Total CD4⁺ T cells were co-cultured for 7 days with IFNDCs (top panels) or mDCs (lower panels) loaded with 1 μg/ml anti-hDectin-1-HA1 in the presence or absence of P. gingivalis LPS (40 ng/ml) or E. Coli LPS (100 ng/ml). T cells were then restimulated for 48h with 1 μM of indicated peptides. Cytokines in the culture supernatants were assessed. Two independent experiments showed similar results. In (A-D), Error bars indicate mean±SD of triplicate assay.

Data in Fig. 8 further support our notion that targeting antigens to DCs via Dectin-1 using recombinant fusion protein of an agonistic anti-hDectin-1 mAb and antigen is an efficient strategy to elicit antigen-specific Th₁₇ responses. Anti-hDEC205-HA1 fusion protein was made (Fig. 8A). Anti-hDEC205-HA1 (Fig. 8B, lower panel) but not IgG4-HA1 (Fig. 8B, upper panel) bound to the surface of monocyte-derived IFNDCs in a dose-dependent manner. IFNDCs loaded with anti-hDEC205-HA1 resulted in HA1-specific Th₁, Th₂, and Th₁₇ responses (Fig. 8C). More importantly, HA1-specific Th₁₇ responses elicited with anti-DEC205-HA1 were further enhanced by anti-hDectin-1 mAb (Fig. 8C, right panel). Anti-hDectin-1 mAb did not increase HA1-specific IFNγ⁺CD4⁺ T cell responses (Fig. 8C, left panel), but significantly decreased IL-13⁺CD4⁺ T cell responses (Fig. 8C, middle panel). We next tested whether anti-hDEC205-HA1 and Pam(3)CysSK(4) synergized to promote HA1-specific CD4⁺ T cell responses (Fig. 8D). Only the higher concentration (200 ng/ml) of Pam(3)CysSK(4) could enhance HA1-specific Th₁, Th₂, and Th₁₇ responses. These data showed that TLR2 signaling does not synergize with DEC205 to promote HA1-specific Th₁₇ responses.

FIGURE 8.

HA1-specific Th₁₇ responses elicited by anti-hDEC205-HA1 can be enhanced by the activation of DCs via Dectin-1 but not TLR2. (A) Reduced SDS-PAGE analysis of anti-hDEC205 mAb (lane 1) and anti-hDEC205-HA1 (lane 2). (B) Binding of anti-hDEC205-1-HA1 and IgG4-HA1 to IFNDCs. (C) Total CD4⁺ T cells were co-cultured for 7 days with IFNDCs loaded with 1 μg/ml anti-hDEC205-HA1 in the presence 5 ug/ml control IgG or anti-hDectin-1 mAb. T cells were then restimulated for 48h with 1 μM of pep 43 or control pep 5. Cytokines in the culture supernatants were assessed. Values acquired with control pep5 were subtracted. Two independent experiments showed similar results. (D) Effects of different concentrations of Pam(3)CysSK(4) on HA1-specific CD4⁺ T cell responses were assessed.

Discussion

This study dissected the pathways for the de novo generation of protein antigen-derived peptide-specific human Th₁₇ in the context of TCR ligation by MHC II/peptide complexes presented by DCs. It appears that the induction of protein antigen-specific Th₁₇ requires more than the actions of previously known Th₁₇-promoting cytokines, but rather occurs through the complementary actions of signals delivered from both innate and adaptive immune cells. Notably, synergistic actions of signals via both TCR and common γ-chain receptors play pivotal roles in programming naïve CD4⁺ T cells to respond to IL-1β followed by the induction of RORC expression. It was also important to note that phenotypes of Th₁₇ activated with HA1 peptides and those activated with polyclonal activators are not the same. Finally, this study demonstrates that targeting antigens to DCs via Dectin-1 using an agonistic anti-hDectin-1 mAb, particularly with TLR2 ligands, can efficiently promote antigen-specific Th₁₇ where antigen-specific memory T cells have already been established.

It has been known that the immune system can favor Th₁₇ responses to certain microbial pathogens, including fungi and bacteria, and this relies mainly on recognition of such pathogens via PRRs, particularly Dectin-1, followed by the induction of Th₁₇-promoting cytokines from APCs (3, 45). However, Th₁₇ are now considered as important immune arms against both extracellular and intracellular pathogens (2-11, 13, 30) as well as cancers (14-17). The assumption that human Th₁₇ are mainly effector T cells with a short life span, as they are often found in peripheral tissues and organs (46), had raised a question as to the value of such vaccine-induced Th₁₇-mediated immunity. However, recent studies have shown that human Th₁₇ consist of long lived-effector memory cells (47, 48), and thus can contribute to long-lasting immunity. We also found that healthy individuals maintained influenza HA1-specific memory Th₁₇, although the levels of Th₁₇ among donors varied. Such memory Th₁₇ could contribute to the protective immunity against influenza infections, presumably by enhancing CD8⁺ T cell and antibody responses (6, 13, 16, 49). More importantly, those Th₁₇ responses can be greatly enhanced by the vaccine model, recombinant fusion protein of anti-hDectin-1 and antigens along with TLR2 and TLR4 ligands.

The mechanisms of human Th₁₇ differentiation remained obscure, but previous studies have revealed several key cytokines that promote Th₁₇ differentiation mainly in the context of inflammatory diseases. Data from this study are in agreement with the previous data, showing that IL-1β/IL-1R and IL-6 play key roles in human (9) and mouse Th₁₇ differentiation (19, 20). The major role of IL-23, expanding memory Th₁₇ (1, 50), is recapitulated by our data. We could detect active forms of TGFβ in the co-cultures of DCs and T cells (data not shown). This suggests that TGFβ may not be a key cytokine that determines the generation of antigen-specific human Th₁₇, while the induction of IL-1R1 expression and signals by IL-1β play key roles. It is important to note that experimental systems used in many of the previous studies were designed to test Th₁₇ differentiation mainly in the context of inflammatory diseases. Thus, T cells were activated with polyclonal stimuli in polarized conditions made with exogenous cytokines or neutralizing antibodies specific for targeted cytokines.

We have shown that the induction of antigen-specific Th₁₇ requires more than the actions of currently known Th₁₇-promoting cytokines. Foremost, the induction of surface IL-1R1 expression was the key step for naïve CD4⁺ T cell differentiation into pathogen-specific Th₁₇. A rapid induction of surface IL-1R1 expression on naïve CD4⁺ T cells is directed by the synergistic actions of signals via TCRs and common γ-chain receptors.

In support of the important roles of common γ-chain cytokines, inhibitors of STAT3 abolished the upregulation of IL-1R1 expression. Although IL-2 and STAT5 showed suppressive functions in Th₁₇ development (51), our data indicated that these could also contribute to the generation of antigen-specific Th₁₇ by enhancing the expression of IL-1R1. Furthermore, IL-2 supports T cell proliferation. The contribution of STAT5 in the upregulation of IL-1R1 expression could not be measured because STAT5 inhibitor (52) suppressed the phosphorylation of STAT1, 3, and 5, as measured by phospho-flow cytometry (data not shown). IL-21 also uses the common γ-chain receptors and thus can contribute to the generation of pathogen-specific Th₁₇ by promoting IL-1R1 expression. The roles of other common-γ chain cytokines, including IL-9 (23) as well as IL-7 and IL-15 (53) in Th₁₇ responses have also been reported.

Although IL-1β/IL-1R1 was the key axis for the induction of RORC, IL-6-mediated STAT3 activation was required for efficient generation of pathogen-specific human Th₁₇. This was further supported by the facts that mutations in STAT1 or STAT3 resulted in the deficiency of Th₁₇-mediated immunity in patients with fungal infections (12, 54). Interestingly, however, IL-6 promotes the induction Th₁₇ only when naïve CD4⁺ T cells express IL-1R1. Further experiments show that IL-6 contributes to the IL-1β-induced RORC expression by enhancing IL-1β expression in TCR-activated IL-1R1⁺ naïve CD4⁺ T cells. The key role of IL-1β in the induction of RORC expression through the action of MyD88 is supported by the previous studies, showing that MyD88 deficiency results in the lack of Th₁₇-mediated immunity against chlamydial infection (55, 56). Moreover, defective IL-1R/MyD88 signaling is associated with impaired Th₁₇ responses (57).

In conclusion, this study revealed the complex but complementary mechanisms for the induction of protein antigen-specific human Th₁₇ in the context of TCR ligation by MHC/peptide complexes. Although there could be alternative pathways for the generation of antigen-specific human Th₁₇, the pathway characterized with DCs, major immune inducers and modulators, could represent the major one. In addition, this study provides a rational strategy that can potentially enhance Th₁₇-mediated host immunity against infections and certain types of cancers in humans.