Recent studies suggest a role for autophagy in the secretion of IL-1 cytokines regulating the development of inflammatory diseases. The antimalarial drug and autophagy/lysosome inhibitor chloroquine (CHQ) is considered as potential trigger of drug-induced or drug-aggravated psoriasis, in which Th17 cells sustain a persistent inflammation. In this study, we investigated the effect of CHQ on human monocyte-derived Langerhans-like cells (MoLC) and dendritic cells (MoDC) in response to IL-1β. The presence of CHQ reduced IL-12p70 release in both subsets, but surprisingly increased IL-6 production in MoDC and IL-23 in MoLC. Importantly, CHQ-treated MoLC promoted IL-17A secretion by CD4+ T cells and elevated RORC mRNA levels, whereas IFN-γ release was reduced. The dysregulation of IL-12 family cytokines in MoLC and MoDC occurred at the transcriptional level. Similar effects were obtained with other late autophagy inhibitors, whereas PI3K inhibitor 3-methyladenine failed to increase IL-23 secretion. The modulated cytokine release was dependent on IL-1 cytokine activation and abrogated by a specific IL-1R antagonist. CHQ elevated expression of TNFR-associated factor 6, a common intermediate in IL-1R and TLR-dependent signaling. Accordingly, treatment with Pam3CSK4 and CHQ enhanced IL-23 release in MoLC and MoDC. CHQ inhibited autophagic flux, confirmed by increased LC3-II and p62 expression, and activated ERK, p38, and JNK MAPK, but only inhibition of p38 abrogated IL-23 release by MoLC. Thus, our findings indicate that CHQ modulates cytokine release in a p38-dependent manner, suggesting an essential role of Langerhans cells and dendritic cells in CHQ-provoked psoriasis, possibly by promoting Th17 immunity.
Autophagy plays an emerging role in immune responses against intracellular pathogens and regulates distinct immunologic processes, important for the initiation of potent innate or adaptive immune responses (1, 2). A complex interplay between autophagy and cytokines is widely acknowledged. Although Th2-related IL-4 and IL-13 inhibit autophagy, IL-1 triggers autophagy (3, 4). In addition, autophagosomes regulate the processing and release of IL-1β, which in turn has previously been considered to induce IL-23 secretion (5, 6). Thus, the complex cross-regulation of autophagosome formation and cytokine release highlights the important regulatory function of autophagy in inflammatory diseases. Indeed, recent studies support a correlation between a malfunction of autophagy-related genes and increased susceptibility to Crohn’s disease (7).
At present, the mechanisms by which autophagy mediates the regulation of immune responses in vivo are not fully understood, but Th1 and Th17 cells are important mediators of the pathology of inflammatory bowel disorders. The release of IL-1 cytokines, regulated by autophagosomes, together with TGF-β and IL-6, drives the differentiation of Th17 cells in vivo (8, 9), suggesting that autophagy possibly modulates the Th1/Th17 balance in Crohn’s disease and other inflammatory disorders. Moreover, although recognized as a sterile inflammatory skin reaction (10), strong evidence exists that Th17 cells play an important role in the pathogenesis of psoriasis, in which increased levels of IL-1 cytokines within the hyperproliferative tissue have long been considered (11). Thus, a dysregulated processing and release of IL-1 cytokines may indicate an altered autophagic flux in diseases with a psoriatic phenotype.
Treatment with the antimalarial drugs chloroquine (CHQ) or hydroxychloroquine (hydroxy-CHQ) is associated with induction or aggravation of psoriasis or psoriatic arthritis (12, 13). CHQ prevents the acidification of lysosomes, necessary for the autolysosomal formation and subsequent enzymatic degradation of enclosed constituents. Considering the lysosomotropic character of CHQ, these data suggest a pivotal role for autophagy in CHQ-provoked psoriasis by maintaining a potent Th17 activity mediated by stimulated APC, including epidermal Langerhans cells (LC) and dermal dendritic cells (DC).
Previously, we showed that inflammatory cytokines differently modulate the activation and maturation of monocyte-derived LC (MoLC) and monocyte-derived DC (MoDC) (14). In this study, we demonstrate that inhibitors of late-stage autophagy potently trigger production and release of the Th17 cytokines IL-23 and IL-6 by activated MoLC and MoDC, thereby possibly shifting T cell response from Th1 to Th17.
Materials and Methods
In vitro generation of LC-like cells and DC from human monocytes (MoLC and MoDC)
PBMC were obtained from buffy-coat donations from anonymous healthy volunteers (DRK-Blutspendedienst Ost, Berlin, Germany) after informed consent. All studies have been approved by the ethics committee of the Charité-Universitätsmedizin (Berlin, Germany). MoLC and MoDC were generated from plastic-adherent human monocytes (14, 15) cultured in RPMI 1640 (Sigma-Aldrich, Taufkirchen, Germany), 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin (all from PAA Laboratories, Cölbe, Germany), and 10% heat-inactivated FCS (Biochrom, Cambridge, U.K.), supplemented with recombinant human (rh)-GM-CSF (100 ng/ml) and rh-IL-4 (20 ng/ml), and with or without rh-TGF-β1 (20 ng/ml; all from Miltenyi Biotec, Bergisch-Gladbach, Germany). At day 6, the nonadherent cells were collected and additionally sorted by CD1a MicroBeads (clone HI149; Miltenyi Biotec), leading to immature CD1a+/CD207+/CD324+/TROP-2+/Axl+ MoLC and CD1a+/CD209+/CD207− MoDC, respectively (14).
In vitro stimulation of MoLC and MoDC
At day 7, generated MoLC and MoDC were harvested, washed three times in PBS, and seeded in a 24-well cell culture plate (BD Biosciences) in complete medium without supplemented cytokines at a density of 106 cells/ml. For subsequent stimulation, the following agonists were applied to the cell culture medium for 24 h: Pam3CSK4 (1 μg/ml; InvivoGen, Toulouse, France), rh-IL-1α (5 ng/ml; BioLegend, London, U.K.), and rh-IL-1β (30 ng/ml; eBioscience, Frankfurt, Germany) alone, or combined with rh-TNF-α (20 ng/ml; eBioscience) and ultrapure LPS from Escherichia coli serotype 0111:B4 (1 μg/ml; InvivoGen).
For autophagy-blocking experiments, cells were preincubated with CHQ (0.2–200 μM), hydroxy-CHQ (20 μM; all from Sigma-Aldrich), bafilomycin A1 (1 μM; Tocris Bioscience, Bristol, U.K.), or 3-methyladenine (3-MA, 10 mM; Sigma-Aldrich) for 1 h in a 24-well cell culture plate (BD Biosciences) at a density of 106 cells/ml in complete medium without supplemented cytokines. Subsequently, MoLC and MoDC were stimulated for additional 24 h with different agonists, as described above, in the presence of CHQ, hydroxy-CHQ, bafilomycin A1, or 3-MA, respectively. For receptor-blocking experiments, MoLC and MoDC were preincubated with the TGF-βR1 antagonist LY364947 (5 μM; Tocris Bioscience) or hydrocinnamoyl-l-valyl pyrrolidine (10–100 μg/ml; Merck KGaA, Darmstadt, Germany), a specific IL-1R antagonist (IL-1Ra) in complete medium without supplemented cytokines for 1 h in presence or absence of CHQ (20 μM). Afterward, MoLC and MoDC were stimulated, as described above. To study MAPK signaling experiments were performed analogously in the presence of the selective MAPK inhibitors U0126 (MAPK kinase [MKK]1/2 inhibitor), SB 202190 (p38 MAPK inhibitor), and SP 600125 (JNK inhibitor) at 10 μM (all from Tocris Bioscience). After 24 h, cell culture medium was collected and IL-23 levels were measured by ELISA.
Cell viability was determined by trypan blue exclusion assay. Experiments were performed in quadruplicates. Additionally, cells were harvested and labeled with FITC-conjugated annexin-V Abs (Enzo Life Sciences, Lörrach, Germany). Cells were examined using a FACSCalibur flow cytometer (BD Biosciences, Heidelberg, Germany) collecting a total of 1–2 × 104 events.
Coculture of stimulated MoLC with CD4+ T cells
Naive human CD4+ T cells were obtained by negative isolation from nonadherent human PBMCs using Naive CD4+ T Cell Isolation Kit II (Miltenyi Biotec) (14). Isolated T cells were cultured in a 96-well cell culture plate with round bottom (Corning) at a density of 105 cells per 100 μl in RPMI 1640 containing 10% heat-inactivated FCS (Biochrom) and 2 mM l-glutamine (PAA Laboratories). A total of 104 immature or mature MoLC, activated with rh-IL-1β (30 ng/ml) for 24 h with or without CHQ, was added and cocultured with naive human CD4+ T cells. To trigger Th17 priming of CD4+ T cells, rh-IL-1β (10 ng/ml), rh-IL-6 (10 ng/ml), and rh-TGF-β (5 ng/ml; all from Miltenyi Biotec) were added to the medium. Samples analyzed in quantitative PCR were taken at day 3. At day 7, 50 ng/ml PMA and 1 μg/ml ionomycin (all from Sigma-Aldrich) were added to the culture medium for restimulation, and coculture was continued for 48 h. Subsequently, cell culture supernatant was collected, and cytokine levels were measured by ELISA.
Intracellular cytokine staining
At day 7, generated MoLC and MoDC were stimulated with different agonists in the presence or absence of CHQ (20 μM), as described above. After 6 h of incubation, brefeldin A solution (1×; BioLegend) was added to the medium to stop the vesicular transport. After additional 18 h of activation, cells were harvested and cytokine production was analyzed by flow cytometry (14). MoLC and MoDC were washed and fixed with 2% paraformaldehyde (Merck KGaA) in PBS at room temperature for 20 min and stained with FITC-conjugated rat anti–IL-6 (clone MQ2-13A5; BioLegend) and PE-conjugated mouse anti–IL-12p40 (clone HP40; eBioscience) in the presence of BD Perm/Wash (BD Biosciences). A total of 1–2.5 × 104 cells was counted on a FACSCalibur flow cytometer. Cell debris was excluded by scatter gates.
The cell culture supernatant was collected at the conclusion of the experiments and assayed for IL-1α, IL-1β, IL-6, IL-23, IL-12p70, IFN-γ, and IL-17A by using commercially available ELISA kits (DuoSet, R&D Systems, Wiesbaden, Germany; ELISA-Ready Set Go, eBioscience).
RNA isolation and quantitative RT-PCR
Total RNA isolation, cDNA synthesis, and quantitative RT-PCR were performed, as described (16). Primers (synthesized by TIB Molbiol, Berlin, Germany) with the following sequences were used: GAPDH, SDHA, and YWHAZ, as published previously (14, 17); IL-12p35, 5′-AGCCTCCTCCTTGTGGCTA-3′ and 5′-TGTGCTGGTTTTATCTTTTGTG-3′; IL-12p40, 5′-CATGGTGGATGCCGTTCA-3′ and 5′-ACCTCCACCTGCCGAGAAT-3′; IL-23p19, 5′-TTCCCCATATCCAGTGTGGAG-3′ and 5′-TCAGGGAGCAGAGAAGGCTC-3′; p62/SQSTM1, 5′-CTGGGACTGAGAAGGCTCAC-3′ and 5′-GCAGCTGATGGTTTGGAAAT-3′; RORC, 5′-CAATGGAAGTGGTGCTGGTTAG-3′ and 5′-GGGAGTGGGAGAAGTCAAAGAT-3′; TBX21, 5′-TTGAGGTGAACGACGGAGAG-3′ and 5′-CCAAGGAATTGACAGTTGGGT-3′. Fold difference in gene expression was normalized to the housekeeping gene SDHA or YWHAZ and GAPDH, showing the most constant expression levels.
MoLC and MoDC were harvested at day 7, washed in PBS, and incubated for 24 h in a 24-well cell culture plate at a density of 106 cells/ml in complete medium without supplemented cytokines in presence or absence of CHQ (20 μM) or 3-MA (10 mM), respectively. Subsequently, MoLC and MoDC were harvested, washed, and applied on a poly-lysine–coated slide (Thermo Scientific, Langenselbold, Germany) using cytospin procedure, as described (18). Afterward, the adherent monolayer was fixed with 4% paraformaldehyde (Carl Roth, Karlsruhe, Germany) for 10 min, washed with PBS, and permeabilized with 0.5% Triton X-100 (Carl Roth) for additional 10 min, followed by washing with 0.0025% BSA (Aurion, Wageningen, the Netherlands) and 0.025% Tween 20 (Carl Roth) in PBS three times for 5 min. After washing, slides were blocked with goat serum (1:20; Dianova, Hamburg, Germany) for 30 min and incubated with the following primary Abs at 4°C overnight: rabbit anti-LC3A (1:400), rabbit anti-LC3B (1:400; all from Cell Signaling Technology Europe, Leiden, the Netherlands), mouse anti–TNFR-associated factor (TRAF)3 (1:50; Santa Cruz Biotechnology), and mouse anti-TRAF6 (1:50; Santa Cruz Biotechnology, Heidelberg, Germany). Secondary DyLight488- and DyLight594-conjugated anti-rabbit or anti-mouse Abs (1:400; Dianova) were applied after washing for 1 h at room temperature. All washing and Ab addition steps were performed with a combination of PBS, BSA, and Tween 20. Cells were mounted in ImmunoSelect Antifading Mounting Medium with DAPI (Dianova). Images were obtained using a BZ-8000 fluorescence microscope (Keyence, Neu-Isenburg, Germany).
Quantification of LC3A punctae
Images of MoLC and MoDC, obtained for immunofluorescence staining with rabbit anti-LC3A and secondary DyLight594-conjugated anti-rabbit Abs, were subsequently analyzed using DotCount Software V1.2. The LC3A dots per cell were counted and quantified for a minimum of 40 cells per sample.
After stimulation, MoLC were washed with PBS and lysed, according to a previously described protocol (19). Samples containing 25 μg protein (Pierce BCA Protein Assay; Thermo Scientific) were boiled in standard SDS-PAGE sample buffer in the presence of DTT and separated by 10% SDS-PAGE (Bio-Rad, Munich, Germany). Gels were blotted overnight onto polyvinylidene difluoride membranes (Immobilon P; Carl Roth). After blocking with 5% skimmed-milk powder (Sucofin, Zeven, Germany) for 1 h at 37°C, membranes were incubated with primary rabbit Abs at 1:1000 (SQSTM1/p62, LC3B, phospho-ERK1/2, phospho-p38 MAPK, and phospho-JNK, all from NEB) overnight at 4°C, and incubated with anti-rabbit HRP-conjugated secondary Ab (NEB; 1:1000) for 1 h. Then blots were developed with SignalFire ECL reagent (NEB) and visualized by PXi/PXi Touch gel imaging system (Syngene, Cambridge, U.K.). The membranes were stripped with Reprore Western blot stripping buffer (Thermo Scientific) and further reprobed with anti–β-actin rabbit Ab (NEB; 1:1000). Values of protein expression were measured by densitometry and normalized to β-actin levels using ImageJ version 1.46r, verifying for nonsaturation and subtracting background.
Data are expressed as means ± SEM. Statistical significance of differences was determined by one-way ANOVA, followed by a Bonferroni post hoc test, and considered significant at p < 0.05. Statistical analysis was performed using GraphPad Prism software.
CHQ inhibits IL-12p70 release and reduces Th1-priming capacity of activated MoLC
CHQ has widely been used in therapies against inflammatory diseases, leading to a reduced secretion of proinflammatory cytokines and IFN-γ release by CD4+ and CD8+ T cells derived from patients with rheumatoid arthritis (20, 21). To determine whether the release of the Th1 cytokine IL-12p70 by activated MoLC or MoDC is regulated by CHQ treatment, we used the proinflammatory cytokines rh-IL-1β and rh-TNF-α in combination with LPS to induce maturation. Due to the low responsiveness of LC toward bacterial Ags (14, 22), mature MoLC were less effective in inducing IL-12p70 release, compared with MoDC (Fig. 1A). Despite the different capability to mature in the presence of bacterial Ags, treatment with CHQ abrogated IL-12p70 release in both subsets.
Th1-priming capacity was assessed by coculture of purified naive CD4+ T cells with MoLC, previously activated with rh-IL-1β. Cell culture supernatants revealed high levels of IFN-γ when restimulated with PMA and ionomycin, confirming an increased type 1 T cell response (data not shown). In contrast, the presence of CHQ during the period of MoLC activation almost completely inhibited T cell–induced IFN-γ production (Fig. 1B). Additionally, RORC, but not TBX21 mRNA levels were slightly upregulated when MoLC were pretreated with CHQ (Fig. 1C). This effect was amplified when a cytokine mixture containing rh-IL-6, rh-IL-1β, and rh-TGF-β1 was added to the coculture of activated MoLC and CD4+ T cells, promoting the development of Th17 cells. Again, the Th1-priming capacity of activated MoLC in the presence of CHQ led to a decreased type 1 T cell response, compared with the respective IFN-γ levels obtained without CHQ pretreatment (Fig. 1D).
CHQ enhances IL-1–induced IL-23 secretion in MoLC and subsequently increases IL-17A release by primed CD4+ T cells
To characterize the effect of CHQ on the development of Th17 cells in more detail, Th17 cytokines (rh-IL-6, rh-IL-1β, and rh-TGF-β1) were added during the coculture of activated MoLC and naive CD4+ T cells, as described before. IL-17A levels were increased compared with cultivation of T cells in the absence of Th17-priming cytokines (data not shown). We subsequently analyzed Th17-priming capacity of MoLC when pretreated with CHQ. Indeed, we observed an elevated secretion of IL-17A by respective CD4+ T cells upon restimulation, when CHQ was present during the period of MoLC activation, confirming a possible shift of Th cell responses from Th1 (IFN-γ) to Th17 (IL-17A) (Fig. 2A). Unstimulated MoLC, pretreated with CHQ, similarly induced IL-17A release by CD4+ T cells, assuming a pivotal role for IL-1β as a key cytokine for Th17 development, as well as a potent activator of MoLC maturation. More precisely, the presence of rh-IL-1β in the cytokine mixture used for Th17 priming might additionally result in a subsequent stimulation of previously untreated MoLC. Together with the sustained effect of CHQ, this might trigger the IL-17A secretion by CD4+ T cells, similarly to activated MoLC.
IL-1β induces IL-23 secretion in human PBMC, MoLC, and MoDC (5, 14), and additional costimulation with rh-TNF-α and TLR agonists further increased IL-23 production by MoLC and MoDC (14). Conversely to IL-12p70, the presence of CHQ during the period of IL-1β stimulation substantially elevated IL-23 release by MoLC, whereas MoDC stayed rather unaffected or slightly decreased cytokine secretion (Fig. 2B). Costimulation with rh-TNF-α and LPS was dispensable for CHQ-induced IL-23 production in activated MoLC (data not shown). Therefore, we subsequently focused on the effects in response to rh-IL-1β alone. IL-23 secretion by activated MoLC peaked at a concentration of 20 μM CHQ (Fig. 2B), whereas higher doses of CHQ markedly reduced cytokine release, assuming an enhanced cell death in both subsets due to elevated trypan blue incorporation and annexin V–positive cells (Fig. 2C). Notably, treatment with CHQ alone was insufficient to drive IL-23 release, indicating that IL-1β, as a distinct activator of MoLC, is strictly required. CHQ-treated MoLC similarly elevated IL-23 in response to IL-1α (Fig. 2D), suggesting a general role for IL-1 cytokines.
To determine whether the alteration in cytokine release can be described as a specific effect of autophagy, we activated MoLC and MoDC in the presence of early- and late-phase autophagy inhibitors. Hydroxy-CHQ and bafilomycin A1, a specific inhibitor of vacuolar-type H+-ATPase, induced similar levels of IL-23 in respective cell culture supernatants, compared with treatment with CHQ (Fig. 2E). Conversely, the early autophagy inhibitor 3-MA did not alter IL-23 secretion, indicating that the restriction of cytokine release might be crucially dependent on distinct processes of the late phase of autophagy or inhibition of lysosomal activity.
MoDC enhance IL-6 secretion in the presence of CHQ
Besides IL-23, which is important for the maintenance of a pathogenic Th17 phenotype, primarily IL-6 induces Th17 differentiation in vivo, together with TGF-β and IL-1 cytokines. Thus, we determined whether IL-6 is regulated by CHQ-treated MoLC and MoDC in response to rh-IL-1β. Intracellular FACS analysis revealed that mainly activated MoDC showed an IL-6+ population, which was further increased upon CHQ treatment (Fig. 3A). Moreover, costaining for p40 subunit revealed a decrease in p40+ cells in the presence of CHQ, confirming the reduced levels of IL-12p70 in both subsets (Fig. 1A) and IL-23 in MoDC. Increased IL-6 production was also detected in cell culture supernatants being significant for MoDC (Fig. 3B).
CHQ differentially regulates gene expression of IL-12 family cytokines
To evaluate the differential regulation of the release of IL-12 family cytokines, we additionally assessed these effects on the transcriptional level. Stimulation with rh-IL-1β with or without CHQ did not affect mRNA levels of p35 (Fig. 4A). This was in accordance with previous findings, showing that stimulation with rh-IL-1β alone was insufficient to induce IL-12p70 release in both subsets (14, 23). The strong upregulation of p19 and p40 mRNA levels in activated MoLC confirmed the enhanced IL-23 release in the presence of CHQ (Fig. 4B, 4C). Interestingly, activated MoDC solely increased p19, but failed to enhance p40 mRNA expression. Similarly, treatment with 3-MA did not enhance the secretion of IL-23, although both subsets moderately upregulated p19 mRNA. These data might indicate that the enhanced release of biologically active IL-23 is crucially associated with the p19- and p40-dependent transcriptional activity in stimulated MoLC and MoDC, respectively.
Inhibition of IL-1R abrogates IL-23 release in activated MoLC, whereas TGF-βR1 signaling appears dispensable
Our data indicate that IL-23 secretion is dependent on the activation of IL-1R signaling (Fig. 2B, 2D). To confirm the crucial role of IL-1R signaling pathways, we used hydrocinnamoyl-l-valyl pyrrolidine (IL-1Ra), a specific inhibitor of Toll/IL-1R domain-mediated MyD88/IL-1R1 interaction. IL-23 secretion, induced by either rh-IL-1α or rh-IL-1β, is markedly reduced in the presence of increasing concentrations of IL-1Ra in CHQ-treated MoLC (Fig. 5A).
Considering the striking differences in IL-23 secretion by CHQ-treated MoLC compared with MoDC, we tested whether TGF-βR1 is involved in this process. TGF-β1, the key regulator of MoLC differentiation (24), has recently been shown to potently alter the activity of autophagy in human hepatoma cells and mammary carcinoma (25, 26). Thus, we activated MoLC and MoDC in the presence of a specific TGF-βR1 inhibitor (LY364947). The rh-IL-1β–dependent release of IL-23 in MoLC treated with CHQ did not significantly alter in the presence of LY364947 (Fig. 5B), indicating that TGF-βR1 activity is dispensable.
CHQ elevates IL-23 and IL-6 secretion in the presence of TLR2/1 ligand
TLRs and IL-1R share significant homology in downstream signaling pathways (27). Similarly to IL-1β, previous studies revealed a potent IL-23, but low IL-12p70 release in human DC and macrophages, induced by TLR2 activation (28, 29). Thus, we used the specific TLR2/1 ligand Pam3CSK4 to activate MoLC and MoDC. Indeed, similarly to IL-1R activation, mainly MoDC highly upregulated IL-6 in the presence of CHQ (Fig. 6B). However, conversely to stimulation with IL-1 cytokines, activation of TLR2/1 in the presence of CHQ led to increased IL-23 secretion in MoLC, but also significantly in MoDC (Fig. 6A). To determine whether the upregulated release of IL-6 and IL-23 might be mediated by a concomitant production and secretion of IL-1 cytokines, as a consequence of TLR2/1 stimulation, we analyzed the release of IL-1α and IL-1β. The quantification of IL-1 cytokines revealed no detectable levels of IL-1α and IL-1β (data not shown), thus confirming a distinct mechanism of cell activation and subsequent modulation of cytokine release, in response to Pam3CSK4 in the presence of CHQ.
LC3-II, p62/SQSTM1, and TRAF6 expression is increased in the presence of CHQ
The mammalian autophagy marker L chain 3 (LC3) undergoes posttranslational modifications during the process of autophagy (30), allowing LC3 to associate with autophagosomes. Immunofluorescence analysis revealed reduced amounts of endogenous LC3A-positive vesicles in both subsets, when treated with the PI3K inhibitor 3-methyladenine (3-MA) (data not shown). Conversely, the presence of CHQ increased the levels of LC3A-positive vesicles, compared with untreated cells, according to its lysosomotropic character. Similar results were obtained for LC3B (data not shown). Quantification of LC3A punctae per cell confirmed the modulated level of autophagosomes in MoDC and MoLC, treated either with CHQ or 3-MA (Fig. 7A).
As a critical protein involved in the specific uptake of proteins into autophagosomes, p62/SQSTM1 also regulates important signaling pathways activated by IL-1β (31). Thus, we compared the specific expression of p62 in untreated and CHQ-treated MoLC (Fig. 7B). Together with increased LC3-II, the presence of CHQ additionally elevated p62 levels. Moreover, activation of MoLC with TLR2/1 ligand Pam3CSK4 significantly elevated p62 mRNA levels when CHQ was present (Fig. 7C).
Previous studies revealed that p62 can act as an important intermediary in IL-1 signaling through TRAF6 interaction (31, 32). An increased activity of TRAF proteins, due to the presence of CHQ, might regulate downstream signaling pathways. Therefore, we assessed intracellular levels of TRAF3 and TRAF6 in MoLC and MoDC, respectively, with or without CHQ. Indeed, immunofluorescence staining revealed a substantial increase of TRAF6, but not TRAF3, in both subsets in response to CHQ in the absence of IL-1 cytokines (Fig. 7D).
CHQ activates MAPK signaling pathways, but only p38 acts as critical regulator of IL-23 release by activated MoLC
TRAF6 is involved in the regulation of various signaling molecules, including MAPK. Thus, we assessed the activation of ERK, JNK, and p38 MAPK by CHQ. In accordance with the increased expression of TRAF6, CHQ increased phosphorylation of all three MAPK in MoLC, independent of a simultaneous activation by rh-IL-1β or Pam3CSK4 (Fig. 8A). To evaluate whether MAPK activation is required for IL-1β–induced IL-23 secretion in the presence of CHQ, specific MAPK inhibitors were used. Although CHQ increased expression of phosphorylated ERK and JNK, inhibition of the respective MAPK did not affect IL-23 release (Fig. 8B). Strikingly, specific blockade of p38 by SB 202190 almost completely inhibited IL-23 production in activated MoLC.
Recent insights on the interaction between autophagy and cytokines led to growing interest to unravel the complex mechanisms in the coordination of innate and adaptive immune responses. In this study, we show a potential pivotal immune regulatory function for late-stage autophagy inhibitors in DC. Our current study clearly demonstrated a tight regulation of Th1/Th17-related cytokines by autophagy/lysosome inhibitors in response to IL-1 cytokines. Although IL-12p70 secretion was clearly reduced, activated MoLC showed enhanced IL-23 release in the presence of CHQ, whereas MoDC highly upregulated the release of IL-6. This is in accordance with a recent study, showing reduced IL-12p70 levels in murine bone marrow–derived DC when stimulated with the TLR4 agonist LPS or whole bacteria and 3-MA (33). Conversely to our results, this group further described unaffected IL-6 secretion and enhanced IL-23 release by human THP-1 cells and murine bone marrow–derived DC, similarly to a previous study (34), in which cytokine release by human DC was controlled by wortmannin, another PI3K inhibitor. However, most of these studies used conserved microbial products, mimicking the presence of infection, whereas in the current study we mainly focused on IL-1–dependent activation, describing immunological and/or inflammatory conditions. Although IL-1 cytokines are closely linked to the innate immune response against various pathogens, IL-1α and IL-1β also trigger acute or chronic sterile inflammatory diseases, mediated by endogenous, metabolic, or exogenous danger signals (35). Thus, various outcomes in cytokine release might be explained by the different inflammatory settings. However, to our knowledge, no other study previously reported an upregulation of IL-23 release in CHQ-treated Langerhans-like cells in response to IL-1 family cytokines.
Stimulation of MoLC with CHQ markedly reduced the IFN-γ release by primed Th1 cells. In contrast, the enhanced IL-23 secretion increased IL-17A release by CD4+ T cells and RORC gene expression, indicating the development of Th17 cells. However, the use of rh-IL-1β, rh-IL-6, and rh-TGF-β1 was necessary to induce IL-17A secretion by CD4+ T cells, assuming that IL-23 alone is incapable of promoting an effective Th17 response. Indeed, previous studies reported that IL-23 is dispensable for the priming of Th17 cells, but mediates the maintenance and stabilization of an ongoing Th17 response, thereby promoting the pathogenic phenotype of IL-17A–secreting T cells (36).
Our data indicate that IL-1R and TLR2/1 activation might critically affect autophagy-related downstream processes. Previous studies highlighted the Toll/IL-1R domain-containing adapter-inducing IFN-β–dependent pathways as possible promoters for an enhanced IL-23 production in response to TLR3 and TLR4 agonists (1, 4). Using IL-1 cytokines and Pam3CSK4, we suggest a crucial role for MyD88-dependent signaling. The TLR2/1-induced upregulation of IL-23 release by activated MoLC, although not significantly, was unexpected, due to the lack of TLR2, 4, and 5 surface proteins (22) and the associated unresponsiveness against bacterial TLR ligands. Thus, the elevated cytokine secretion might indicate that CHQ possibly enhanced the immune regulatory function of activated MoLC against bacterial Ags similar to proinflammatory cytokines (14). We also assumed a critical modulatory function of TGF-β1, the key regulator of LC differentiation, due to the distinct gene profile of both subsets and differences in the capacity to induce IL-23 secretion in the presence of CHQ, but obtained no significant changes when TGF-βR1 was inhibited. As a result, we suggest a crucial disparity in IL-1R– and TLR2/1-dependent signaling pathways, which might be differently affected by CHQ.
Both IL-1R and TLR share significant homologies in downstream adapter molecules, leading to the activation of TRAF family proteins. TRAF6 acts as a critical adapter protein, regulating the activity of distinct protein kinases and determining the subsequent transcriptional activity. In general, receptor engagement leads to the binding of IL-1R–associated kinase 1 and IRAK4 to the common adaptor protein MyD88 and subsequently to the activation of TRAF6, which is thought to induce the recruitment of the MAPK TAK1, thus leading to the phosphorylation of the IκB kinase complex. The following degradation of IκB or p105 induces the release of NF-κB (p50/p65)– or MKK1/2-dependent ERK, respectively, to the nucleus. Alternatively, TAK1 is released to the cytosol, leading to the phosphorylation of MKK3/6 or MKK7, thereby activating JNK and p38 as potent inducers of transcription (37). Indeed, our data indicate that CHQ differently regulates the production of proinflammatory cytokines by modulating the specific transcriptional activity in MoLC and MoDC, respectively. The gene profile of IL-12/23 subunits in response to rh-IL-1β demonstrated a remarkable increase of IL-23p19 transcription in cells treated with CHQ. This occurs possibly due to an inhibition of enzymatic degradation of critical regulatory proteins in autolysosomes or an intracellular accumulation of autophagy-related substrates, thereby affecting important downstream signaling molecules.
As expected, p62/SQSTM1 expression was highly increased in the presence of CHQ. Hence, we assumed a pivotal role for TRAF6, which has previously been described to be selectively regulated by p62 (31). The selective inhibition of MAPK signaling pathways downstream of TRAF6 identified p38 MAPK as critical mediator of the CHQ-induced IL-23 release by MoLC in response to IL-1β. The exact mechanisms triggering the dysregulation of cytokine release remain unknown, although previous studies reported a crucial dependency on p38 pathways and the induction of IL-23 secretion (38, 39). Interestingly, besides its importance in the activation of various transcription factors, p38 MAPK has been characterized as a critical enhancer of the accessibility of NF-κB binding sites by phosphorylation and phosphoacetylation of histone H3 (40). Indeed, in response to IL-1β, the p38 MAPK pathway is activated concurrently with NF-κB, which can bind to promoter regions of p40, p19, and IL-6 genes. Thus, p38-induced histone modifications may control NF-κB target gene activation, which in turn resulted in the increased release of IL-23 or IL-6 and the reduced production of IL-12p70, respectively.
Interestingly, inhibition of late- but not early-phase autophagy is required for enhanced IL-23 production. However, the presence of lysosomal lumen alkalizer (CHQ and hydroxy-CHQ) or vacuolar-type H+-ATPase inhibitors (bafilomycin A1) for long incubation periods (>24 h) possibly disabled proteasomes, endocytic trafficking, and other cellular processes additionally (41). Consistent with this, our data do not provide direct evidence linking autophagy-related proteins and IL-1R– or TLR2/1-dependent downstream signaling molecules, including TRAF6, although an elevated expression has been observed. The lack of colocalization of LC3 and TRAF6 indicates that the results, at least in part, might be explained by the reduced activity of proteasomes, lysosomes, or endosomes, instead of autophagy-related processes. Similarly, PI3K inhibitors such as 3-MA target other kinases as well and potently alter different cellular metabolic processes (42), and thus are also not necessarily specific autophagy inhibitors.
In summary, our data support previous studies on the critical regulatory function of autophagy/lysosome inhibitors on immunologic processes and further indicate a new role for human LC in the regulation of IL-1–mediated immune responses during CHQ treatment, most likely through TRAF6-dependent modulation of p38 MAPK. Indeed, previous case reports described exacerbations of psoriatic skin lesions, a sterile antibacterial skin reaction, during the treatment with CHQ. Moreover, psoriatic skin shows increased activity of p38 MAPK, suggesting that respective signaling pathways may play a role in the pathogenesis of psoriasis (43, 44). Currently, the pathophysiology of drug-provoked psoriasis is poorly understood. It is hypothesized that the inhibition of transglutaminase in the skin by antimalarial drugs might influence the cellular proliferation (45). Therefore, the use of CHQ or hydroxy-CHQ in patients with psoriatic skin lesions is considered to be contraindicated. The differential regulation of cytokines in the presence of CHQ indicates a critical stimulatory function of LC and dermal DC in drug-provoked psoriasis by maintaining a pathological IL-23/IL-6/Th17 axis in inflammatory skin conditions with elevated levels of IL-1 cytokines, independent of microbial stimuli.
This work was supported by the Erasmus Mundus – Action 2 program (Ph.D. scholarship to T.L.).
The authors have no financial conflicts of interest.