Alarmin Function of Cathelicidin Antimicrobial Peptide LL37 through IL-36γ Induction in Human Epidermal Keratinocytes

Cathelicidin is a gene family that is preserved in vertebrates from fishes to mammals. Some mammals have multiple cathelicidin genes; in humans, CAMP is the only cathelicidin gene, and it encodes the 18-kDa proprotein hCAP18. LL37 is one form of the mature cathelicidin peptides that are derived from hCAP18 by enzymatic cleavage with kallikreins in human epidermis (1). LL37 forms an α helical structure and has broad antimicrobial properties against bacteria, virus, and fungus (2–5). LL37 kills microbes and exerts “alarmin” activity, which works as a multifunctional innate immune effector that recruits and activates inflammatory cells, including dendritic cells (DC) and monocytes (6). LL37 drives biological responses through epidermal growth factor receptor (EGFR), formyl peptide receptor-like 1 (FPRL1), TLR4, and TLR9 (7–11), resulting in the induction of proinflammatory cytokines and chemokines, such as IL-8 and IL-6, in monocytes, mast cells, and epithelial cells (5, 12–15). Thus, cathelicidin peptides modulate inflammatory cascades, and the effects of cathelicidin stimulation depend on the host cells and tissues that induce, and are affected by, cathelicidin.

Danger signals, such as infection and injury, exacerbate dermatoses, including psoriasis, and the induction of new lesions by injury is known as the Koebner phenomenon. Danger signals provoke the epidermis to produce cathelicidin through TLR2 activation in keratinocytes (16, 17). Aberrant LL37 expression is observed in psoriasis, rosacea, and other skin inflammatory disorders (18–21). In psoriasis, LL37 enables plasmacytoid DC to recognize self-DNA through TLR9 (8) and enables keratinocytes to induce TLR9 and to react against TLR9 ligands (7). Altered cathelicidin peptides are observed in rosacea, and they induce dermatoses resembling rosacea in mouse skin (19). Thus, cathelicidin has roles in the initiation of inflammatory cascades in some diseases, although the functions and roles of cathelicidin peptides differ depending on the type of dermatosis. Hence, it is still generally unknown how the aberrant cathelicidin causes skin diseases, and how cathelicidin acts as an alarmin in epidermal keratinocytes where innate immune responses are initiated has not systematically analyzed.

To understand how aberrant epidermal cathelicidin affects the behavior of keratinocytes and consequent inflammatory reactions by innate immune systems, we performed cDNA array analysis and revealed that LL37 induces molecules related to innate immunity, such as the IL-1 family and antimicrobial peptides, as well as chemokines in human keratinocytes. Among them, IL-36γ is of particular interest because IL-36γ and LL37 are abundant in psoriasis epidermis, and they synergistically increase chemokine production in human keratinocytes. Furthermore, silencing of IL-36γ and its receptors attenuated LL-37–dependent chemokine induction in keratinocytes, suggesting that LL37 initiates and exacerbates inflammation coordinately with IL-36γ and that IL-36γ enhances the alarmin functions of epidermis in dermatoses.

Human neonatal epidermal keratinocytes were grown in serum-free medium supplemented with HuMedia-KG2 Growth Supplements (KURABO INDUSTRIES, Tokyo, Japan). All cells were passed at 80% confluence under low-calcium conditions (0.05 mM), and keratinocytes in the third or fourth passage were used for the experiments. To induce keratinocyte differentiation, confluent keratinocytes were cultured in 1 mM calcium media for 72 h; subsequently, the calcium concentration was increased to 1.6 mM for an additional 48 h. Human cathelicidin antimicrobial peptide (hCAMP) LL37 (amino acid sequence of LLGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTES) was synthesized by Scrum (Tokyo, Japan). Recombinant human IL-36 (IL-36α, β, and γ), monoclonal rat anti-human IL-36γ/IL-1F9 Ab (MAB2320), and human IL-36γ/IL-1F9 biotinylated Ab (BAF2320) were purchased from R&D Systems (Minneapolis, MN). Polyclonal goat anti-human IL-36γ/IL-1F9 Ab was purchased from Santa Cruz Biotechnology (Dallas, TX). P38 MAPK inhibitor SB203580, MEK inhibitor PD98059, rhodamine-conjugated donkey anti-goat IgG were purchased from Chemicon International (Temecula, CA). FITC-conjugated donkey anti-rabbit IgG, and G_i protein inhibitor pertussis toxin (PTx) were purchased from Merck Millipore (Billerica, MA). JNK1/2 inhibitor SP600125, heparin-binding EGF-like growth factor (HB-EGF) inhibitor CRM197 (nontoxic mutant of diphtheria toxin solution), and EGFR tyrosine kinase inhibitor AG1478 were purchased from Wako Pure Chemical Industries (Osaka, Japan). The NF-κB inhibitor curcumin and the endosomal acidification inhibitors chloroquine diphosphate salt (ChQ) and bafilomycin A1 (BAF) were purchased from Sigma-Aldrich (St. Louis, MO). TLR2 and TLR4 inhibitor OxPAPC, TLR9 antagonist ODN TTAGGG (ODN A151), and ODN TTAGGG control were purchased from InvivoGen (San Diego, CA).

Keratinocytes were maintained in an undifferentiated condition in low-calcium media (0.05 mM) or in a differentiated condition in high-calcium media (1.6 mM) and stimulated with LL37 at 0, 2.56, or 7.68 μM for 12 or 24 h. Total RNA was isolated using an RNeasy Mini Kit and the RNase-Free DNase set (both from QIAGEN, Valencia, CA). Total RNA concentration was measured using a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA). The RNA quality was determined using the RNA 6000 Nano Kit and an Agilent 2100 Bioanalyzer (both from Agilent Technologies, Palo Alto, CA), and the RNA Integrity Number was confirmed to be ≥9. Total RNA was amplified, labeled, and analyzed as previously described (22). Normalization of the expression data was performed using GeneSpring GX 12.6 (Agilent Technologies) to query the expression of 41,000 genes with 44,000 distinct probes. After data transformation to GeneSpring, per-chip normalization to the 75th percentile and baseline to the median of all samples was performed. After this normalization, extremely low-intensity probes were excluded, leaving 32,940 probes for analysis. The cDNA microarray data were deposited in the Gene Expression Omnibus database under accession number GSE49472 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE49472).

Total RNA was extracted using ISOGEN (NIPPON GENE, Toyama, Japan). cDNA synthesis and quantitative PCR were performed as previously described (23). The primers and TaqMan probe sets used are listed in Supplemental Table I.

Cells were lysed in RIPA buffer (50 mM HEPES, 150 mM NaCl, 0.05% SDS, 0.25% deoxycholate, 0.5% Nonidet P-40 [pH 7.4]) with cOmplete protease inhibitors (F. Hoffmann-La Roche, Basel, Switzerland). The protein concentration was determined using the Pierce BCA kit (Thermo Fisher Scientific). A total of 20 μg total protein or recombinant human IL-36γ/IL-1F9 (aa 18–169) was loaded/well for electroporation and Western blotting. Membranes were incubated at 4°C overnight with monoclonal rat anti-human IL-36γ/IL-1F9 Ab (1 μg/ml), rinsed with TBS-T (25 mM Tris [pH 7.5] and 0.1% Tween 20), and incubated with HRP-labeled polyclonal goat anti-rat IgG Ab. Chemiluminescence data from Western blots were collected using an Image Quant LAS 4000 mini (GE Healthcare UK, Little Chalfont, U.K.).

IL-36γ protein in the supernatants of the keratinocytes was measured by sandwich ELISA. Microlon 96-well high-binding plates (Greiner Bio-One, Frickenhausen, Germany) were coated with 100 μl/well monoclonal rat anti-human IL-36γ/IL-1F9 Ab (1 μg/ml) overnight at room temperature (RT). After washing thrice with 0.05% Tween 20 in PBS (PBST), 300 μl blocking solution containing 5% BSA in PBST was added for 1 h at RT. After washing with PBST, 100 μl samples or rIL-36γ/IL-1F9 was added to each well in duplicate and incubated for 2 h at RT. After washing with PBST, human IL-36γ/IL-1F9 biotinylated Ab (0.5 μg/ml in PBST) was added for 2 h at RT. Streptavidin-HRP and substrate solution was used for colorimetric assay following the manufacturer’s instructions (R&D Systems).

To detect multiple cytokines and chemokines in the supernatants of keratinocytes stimulated with IL-36 in the presence or absence of LL37, Bio-Plex Pro Human Cytokine Assay 1 × 96-well 27-Plex Group I and Bio-Plex suspension array system (both from Bio-Rad, Hercules, CA) were used, following the manufacturer’s instructions. IL-8, CXCL1, and CCL20 protein in cultured media were measured using a DuoSet ELISA Development human kit (R&D Systems).

Paraffin sections (3 μm thickness) were treated with a microwave, blocked with rabbit serum (Sigma-Aldrich), and incubated with polyclonal goat anti-human IL-36γ/IL-1F9 Ab (20 μg/ml) at 4°C overnight. Subsequently, sections were visualized with a Histofine SAB-PO (G) kit (Nichirei Bioscience, Tokyo, Japan) and 3,3′-diaminobenzidine substrate (Wako Pure Chemical Industries) and counterstained with Mayer’s hematoxylin. Images were obtained using an Axio Imager M1 microscope (Carl-Zeiss, Oberkochen, Germany).

For immunofluorescence, paraffin sections were incubated with polyclonal goat anti-human IL-36γ/IL-1F9 Ab and polyclonal rabbit anti-human LL37 Ab (Phoenix Pharmaceuticals, Burlingame, CA) at 4°C overnight. Rhodamine-conjugated donkey anti-goat IgG and FITC-conjugated donkey anti-rabbit IgG were used as secondary Abs. Images were obtained using a Zeiss LSM700 laser scanning confocal microscope (Carl-Zeiss).

The small interfering RNA (siRNA) for IL36B, IL36G, IL36RN, IL1RL2 (IL36R), and IL1RAP (FlexiTube GeneSolution; GS27177 for IL36B, GS56300 for IL36G, GS26525 for IL36RN, GS8808 for IL1RL2, and GS3556 for IL1RAP), control siRNA (AllStars Negative Control siRNA), and HiPerFect Transfection Reagent (both from QIAGEN) were used for the siRNA experiments. Keratinocytes were processed with siRNA (10 nM) following the protocol in the HiPerFect Transfection Reagent handbook (QIAGEN). After 24 h for gene silencing, LL37 (0, 2.56, or 7.68 μM) was added to each well. After 24 h of stimulation, the supernatants were collected for protein analysis, and the total RNA was isolated as described above.

To determine the signaling pathway by which LL37 induces IL36G, the keratinocytes were pretreated with inhibitors for 1 h and stimulated with LL37 at 0, 2.56, or 7.68 μM for 24 h in both low- and high-calcium conditions. The inhibitors used for the assay were HB-EGF inhibitor CRM197 (10 μg/ml), EGFR tyrosine kinase inhibitor AG1478 (50 nM), Gi protein inhibitor PTx (200 ng/ml), endosomal acidification inhibitors ChQ (5 μM) and BAF (100 nM), p38 MAPK inhibitor SB203580 (10 μM), MEK inhibitor PD98059 (20 μM), JNK1/2 inhibitor SP600125 (10 μM), NF-κB inhibitor curcumin (10 μM), TLR2 and TLR4 inhibitor OxPAPC (30 μg/ml), and TLR9 antagonist ODN TTAGGG (ODN A151, 1 μM) and ODN TTAGGG control (1 μM). Total RNA was collected to measure IL36G using quantitative RT-PCR.

To determine the signaling pathway by which IL-36γ induces chemokine induction, keratinocytes were pretreated with p38 MAPK inhibitor SB203580 (10 μM), MEK inhibitor PD98059 (20 μM), JNK1/2 inhibitor SP600125 (10 μM), or NF-κB inhibitor curcumin (10 μM) for 1 h and stimulated with 100 ng/ml IL-36 (α, β, γ) in the presence or absence of LL37 (2.56 μM) for 24 h under low-calcium conditions. Total RNA was collected to measure IL8, CXCL10, CXCL1, and CCL20 by quantitative RT-PCR.

Data were analyzed by one-way ANOVA with the Tukey multiple-comparisons test or two-way ANOVA with the Sidak multiple-comparisons test using GraphPad Prism 6 (GraphPad, La Jolla, CA), unless otherwise stated. The p values < 0.05 were considered significant. All experiments were performed in triplicate and repeated at least three times to confirm the reproducibility.

Because the cathelicidin peptide LL37 modifies the host immune responses, cell growth, migration, and differentiation (24), we conducted cDNA microarray analysis to understand the consequences of aberrant cathelicidin antimicrobial peptide (CAMP) expression in epidermal keratinocytes of dermatoses. We stimulated human epidermal keratinocytes with LL37 (2.56 or 7.68 μM) for 12 or 24 h and identified genes with altered expression in both undifferentiated keratinocytes (low calcium) and differentiated keratinocytes (high calcium). These LL37 concentrations were the ones observed in skin diseases (19, 20). IL8 induction confirmed a proper stimulation by LL37 (13) (Supplemental Table II). From the microarray data, we identified several gene groups that affect inflammatory reactions in skin diseases (5, 25–28). LL37 increased genes of the IL-1 family, antimicrobial peptides, and chemokines in keratinocytes (Fig. 1). Increases in the IL-1 family and antimicrobial peptides were more obvious in differentiated keratinocytes cultured in high-calcium media. LL37 increased Th17/Th1-related genes IL6 and IL23A and increased CSF2 (GM-CSF) in undifferentiated keratinocytes. We did not observe significant expression of IL10 or Th2 cytokines IL4 and IL13 by keratinocytes (data not shown). Most of the TLRs were increased in differentiated keratinocytes, and LL37 increased TLR2. Thus, LL37 induced proinflammatory cytokines related to the IL-1 family, Th1/Th17 cascades, and innate immune molecules antimicrobial peptides and TLRs.

The IL-1 family contains 11 members: IL-1α, IL-1β, IL-1R antagonist (IL-1RN), IL-18, IL-33, IL36RN/IL-1F5, IL-36α/IL-1F6, IL-36β/IL-1F8, IL-36γ/IL-1F9, IL-37/IL-1F7, and IL-1F10/IL-38 (29). LL37 increased IL36G, IL36RN, IL1F10, IL1A, IL1B, and IL1RN in both undifferentiated and differentiated keratinocytes, and it increased IL36B and IL37 in differentiated keratinocytes. IL36γ was significantly increased (>4-fold) by LL37 in both undifferentiated and differentiated keratinocytes (Supplemental Table II). Quantitative RT-PCR confirmed the induction of IL-1 cluster genes by LL37 (Fig. 2). IL36G, IL36B, IL1F10, and IL36RN were increased by LL37, and the high-calcium condition augmented their expression. IL37, IL1A IL1B, and IL1RN were increased by LL37, and the calcium conditions had virtually no effect on their expression. We also observed that LL37 increased the IL-36R IL1RL2/IL36R and IL1RAP. Consistent with the very low level of IL36A in the microarray analysis, IL36A was not detectable by quantitative RT-PCR (data not shown).

Because LL37 stimulation induced IL-36γ mRNA most significantly in the IL-1 cluster genes, we further examined the dynamics of IL-36γ. We observed that LL37 increased IL-36γ protein in a time- and a dose-dependent manner (Fig. 3A). IL-36γ release in cultured media was confirmed by ELISA (Fig. 3B). In parallel with the calcium effects on IL36G mRNA, IL-36γ protein was increased in the high-calcium condition. Truncated forms of IL-36γ have greater immunological activity than do their full-length counterparts (30). Keratinocytes stimulated by LL37, especially differentiated keratinocytes, produced both the full-length and the truncated active forms of IL-36γ (Fig. 3D). These data showed that cathelicidin augmented IL-36γ expression in keratinocytes and that calcium modulated the expression. Again, IL-8 served as positive controls for LL37 stimuli (Fig. 3C).

LL37 acts on several receptors and signaling pathways, including EGFR, FPRL1, TLR4, and TLR9, to induce proinflammatory cytokines and chemokines (5, 7–15). We examined which signaling pathways are involved in IL36G induction by LL37. PTx, a reagent known to selectively block G_i protein–coupled signaling, significantly decreased IL36G induction by LL37 in both undifferentiated and differentiated keratinocytes (Fig. 4). EGFR inhibition by HB-EGF inhibitor CRM197 and EGFR tyrosine kinase inhibitor AG1478, as well as blockade of TLRs by TLR2 and TLR4 inhibitor OxPAPC and TLR9 antagonist ODN TTAGGG, did not affect IL36G induction by LL37. Among the intracellular signaling pathways, p38 MAPK–specific inhibitor SB203580 blocked IL36G induction by LL37. JNK inhibitor SP600125, MEK inhibitor PD98059, NF-κB inhibitor curcumin, and endosomal acidification inhibitors ChQ and BAF did not affect IL36G induction by LL37. These observations suggest that LL37 induces IL36G through a G_i protein–coupled receptor (GPCR) and p38 MAPK–signaling pathway in human epidermal keratinocytes.

Because hCAMP is abundantly expressed in psoriatic epidermis (7, 18, 20), we examined the expression of hCAMP and IL-36γ in psoriatic skin. As reported previously (31, 32), we observed high IL-36γ expression in the lesional epidermis of psoriasis (Fig. 5). We also confirmed that IL-36γ and LL37 coexisted in the lesional skin of psoriasis. It is noteworthy that IL-36γ was observed more in suprabasal cells than in basal cells, which is consistent with the in vitro data showing greater IL-36γ induction by LL37 in differentiated keratinocytes in high-calcium condition (Figs. 2, 3).

To elucidate how the coexistence of abundant IL-36γ and LL37 affects the epidermis, we stimulated keratinocytes with LL37 and IL-36 and examined the induction of chemokines and cytokines using cDNA array analysis (Supplemental Table III). Exogenous IL-36α, β, and γ increased IL-8, CXCL1, CXCL10, and CCL20 mRNA in a dose-dependent manner at concentrations of 10 ng/ml to 1 μg/ml (Fig. 6A–D). Chemokine release by IL-36 also was observed, and 100 ng/ml of IL-36 efficiently increased the release of most chemokines (Fig. 6E–H). Because up to 30 ng/ml of IL-36γ was detected from keratinocytes in vitro (Fig. 3B), it is suggested that a physiological concentration of IL-36 released from keratinocytes would result in skin inflammation through chemokine induction. The presence of LL37 further augmented the production of CXCL10, CCL20, and G-CSF (CSF3) but showed little effect on CXCL1, GM-CSF (CSF2), IL-6, and RANTES (CCL5).

To explore the intracellular signaling pathways that induce chemokines by IL-36 and LL37, we treated keratinocytes with inhibitors for MAPK or NF-κB. p38 MAPK inhibitor SB203580 suppressed IL-36–derived IL8, CXCL10, and CCL20 expression, regardless of the presence of LL37 (Supplemental Fig. 1). The NF-κB inhibitor curcumin significantly suppressed CXCL1 and CXCL10 expression by IL-36, regardless of the presence of LL37. The JNK1/2 inhibitor SP600125 reduced IL8 and CXCL1 expression induced by IL-36, but it augmented CXCL10 and CCL20 expression (Supplemental Fig. 1). The MEK1/2 inhibitor PD98059 showed no effect on IL8 and CXCL1 but augmented CXCL10 and CCL20 expression (Supplemental Fig. 1). Thus, LL37 and IL-36 coordinately augment the production of these chemokines, primarily through p38 and JNK MAPK and NF-κB pathways, and MEK1/2 signaling negatively regulates the expression of these chemokines in keratinocytes.

Associating the IL-36γ induction by LL37 with the IL-36γ–inducible chemokines from keratinocytes, we sought to determine whether LL37 induces chemokines that are dependent on IL-36γ induction. We knocked down IL36G/IL1F9 by siRNA (IL1F9_3, IL1F9_5, and IL1F9_Mix) for 24 h and then stimulated it with LL37 for another 24 h. IL36G suppression was confirmed in LL37-treated keratinocytes (Fig. 7A). The stable IL36B expression showed the specificity of IL36G siRNA (Fig. 7B). IL36G siRNA (IL1F9_3, IL1F9_5, and IL1F9_Mix) significantly suppressed IL8, CXCL1, CXCL10, and CCL20 induction by LL37 (Fig. 7E–H). Silencing of IL-36R IL36R/IL1RL2 and IL1RAP also significantly decreased the induction of chemokines by LL37 (Fig. 8). Expression of IL36RN and IL38/IL1F10, which are known antagonists of IL-36 ligands, was not altered when IL36G was silenced (Fig. 7C, 7D). IL36RN knock down did not significantly alter chemokine induction by LL37 (Fig. 8). These results suggested that LL37 induces chemokines in keratinocytes, at least in part, via an IL-36γ and IL-36R–mediated mechanism, and IL-36 antagonist IL-36RN has little effect on chemokine induction by LL37.

In this study, we systematically examined the LL37-inducible genes in human keratinocytes by cDNA array analysis and demonstrated that a representative hCAMP LL37 induces IL-1 cluster genes, antimicrobial peptides, and chemokines. Among IL-1 cluster genes, we showed that LL37 induces IL-36γ in both undifferentiated and differentiated keratinocytes and that IL-36γ and hCAMP are both abundant and coexisted in psoriatic epidermis. IL-36γ and LL37 coordinately induced chemokine production from human epidermal keratinocytes. Furthermore, the induction of proinflammatory cytokines and chemokines by LL37 is carried out, at least in part, through IL-36γ induction and the IL-36R pathway. Thus, this study revealed that alarmin functions of LL37 in human epidermis are enhanced by IL-36γ induction and its receptors in keratinocytes, as well as through the induction of other molecules related to the innate immune reaction.

The Koebner phenomenon is well known in psoriasis: danger signals, such as infection and injury, provoke new skin lesions in psoriasis patients. Because cathelicidin can be induced in human epidermis by environmental changes, including infections and injuries, the coexistence of LL37 and IL-36γ in psoriatic epidermis would initiate skin exacerbations by stimuli from the microenvironment. IL-36γ and IL-36β are increased in human psoriatic epidermis (32), whereas IL-36α–transgenic mice show an inflammatory skin condition that is similar to human psoriasis (31). In patients with a familial history of generalized pustular psoriasis, homozygous loss-of-function mutations were identified in IL-36R antagonist (IL-36RN), which results in constitutive IL-36 activation (33, 34). Thus, the possible involvement of IL-36 signaling has gained attention in the pathogenesis of psoriasis. Our data showed that LL37 induces IL-36γ and that IL-36 efficiently induces chemokines and cytokines in the presence of LL37 in human keratinocytes. IL-36–inducible factors from human keratinocytes include chemokines IL-8, CXCL1, CXCL10, RANTES, and CCL20, and cytokines IL-6, G-CSF, and GM-CSF, which can recruit and activate DC, macrophages, neutrophils, T cells, and NK cells. Studies in other epithelial tissues, such as human bronchial epithelial cells and mouse lungs, showed that IL-36, particularly IL-36γ, induces the neutrophil-attracting chemokines IL-8 and CXCL1, the T cell–attracting chemokines CXCL10, and the DC-attracting and Th17-attracting chemokine CCL20, as well as cytokines and growth factors, including IL-6, G-CSF, and GM-CSF (35, 36). Data from previous studies combined with our data indicate that IL-36γ would activate epithelial cells to induce chemokines and recruit inflammatory cells for the initial stages of innate immunity in skin inflammation. Thus, in human epidermal keratinocytes, the induction of CAMP by innate immune stimuli can be a trigger to activate the IL-36 axis that initiates skin inflammation and exacerbates chronic dermatoses.

The cathelicidin LL37 acts on several receptors and signaling pathways (5, 7–15). By blocking with PTx, we demonstrated that LL37 induced IL36G through G_i protein–coupled signaling. Among GPCR, LL37 is known to activate FPRL1 in endothelial cells (10). However, keratinocytes are suggested to have little FPRL1 and induce transactivation of EGFR by LL37, which is mediated through GPCR other than FPRL1 (13). However, we observed that EGFR inhibition by HB-EGF inhibitor CRM197 and EGFR tyrosine kinase inhibitor AG1478 did not affect IL36G induction by LL37. Because EGFR dominantly activates p42/p44 MEK in keratinocytes, inhibition of LL37-mediated IL36G induction by p38 inhibitor SB203580, but not by MEK inhibitor PD98059, also suggested little involvement of EGFR in IL36G induction by LL37 in keratinocytes. Although specific receptors inducing IL36G by LL37 were not identified in this study, multiple GPCR might be involved in the process. The ligand-dependent and -independent activation and oligomerization of GPCR are recognized as a cross-talk of GPCR, which alters intracellular signals making them different from the non-oligomerized GPCR signaling (37–40). Because LL37 is a cationic peptide and can bind directly to the cell membrane without receptors (8, 41), it may activate GPCR in a ligand-independent manner to induce IL-36γ. Occasionally, we observed increases in chemokines when we treated keratinocytes with mock (transfection reagent) and control siRNA along with LL37 (Figs. 7, 8). These phenomena also suggested the ligand-independent activation of GPCR, because the transfection reagent is also a cation that affects cell membrane components. The molecular mechanism of LL37-mediated GPCR activation should be explored further to define the cascade of inflammatory reactions in innate immunity of the skin.

We demonstrated that LL37 induces IL-36R IL-1RL2/IL-36R and IL-1RAP, as well as IL-36γ, IL-36β, and other IL-1 family genes. Therefore, LL37 would amplify the stimulation of IL-36 by inducing both ligands and their receptors in keratinocytes, which was demonstrated, in part, by the experiments with siRNA for IL36G and its receptors IL36R/IL1RL2 and IL1RAP. We also showed that p38 and JNK MAPK and NF-κB signaling affected the induction of chemokines in combination with IL-36 and LL37; these results are similar to those from a previous study that demonstrated that IL-36 phosphorylates NF-κB and JNK and p38 MAPK (30). IL-1β and IL-1R–associated kinase-1 are aberrantly expressed in psoriatic lesions, and the synergy between IL-1β and TNF-α leads to sustained inflammatory responses (42). IL-18 induces IFN-γ, and the increased IL-18 participates in the development of the Th1 response in lesional skin of psoriasis (43). IL-36α, IL-36β, and IL-36γ are members of IL-1 family of cytokines and use the same receptor IL-36R/IL-1RL2 coupled with IL-1RAP (30), and IL-36 induces TNF and IL-6 in human keratinocytes (44). Combined with observations of the functions of IL-1 family molecules in psoriasis, cathelicidin-dependent induction of IL-1 family molecules might indicate that the mechanism of altered innate immunity exacerbates and modulates inflammatory responses in psoriasis.

In summary, hCAMP LL37 induces IL-36γ production, as well as IL-36R. LL37 induces IL-8, CXCL1, CXCL10, and CCL20 through IL-36γ induction and IL-36R signaling, which would recruit neutrophils, T cells (including Th17 cells), and DC for the epidermis. Our findings provide evidence that some of the alarmin activities of LL37 occur via IL-36γ induction; thus, IL-36γ facilitates innate immune reactions by cathelicidin. Cathelicidin and IL-36 are engaged, therefore, in the pathogenesis of psoriasis and other skin diseases during the initiation or occasional exacerbation of dermatoses by innate immune stimuli, such as local and general infections and injuries.

We thank Yumiko Ito, Natsue Sawaya, and Yuko Yoshida for technical assistance and Momo Miura and Yuko Yanagawa for secretarial support.

The authors have no financial conflicts of interest.