Abstract
Nucleotide-binding domain leucine-rich repeat (NLR) proteins are regulators of inflammation and immunity. Although first described 8 y ago, a physiologic role for NLRP12 has remained elusive until now. We find that murine Nlrp12, an NLR linked to atopic dermatitis and hereditary periodic fever in humans, is prominently expressed in dendritic cells (DCs) and neutrophils. Nlrp12-deficient mice exhibit attenuated inflammatory responses in two models of contact hypersensitivity that exhibit features of allergic dermatitis. This cannot be attributed to defective Ag processing/presentation, inflammasome activation, or measurable changes in other inflammatory cytokines. Rather, Nlrp12−/− DCs display a significantly reduced capacity to migrate to draining lymph nodes. Both DCs and neutrophils fail to respond to chemokines in vitro. These findings indicate that NLRP12 is important in maintaining neutrophils and peripheral DCs in a migration-competent state.
Nucleotide-binding domain leucine-rich repeats (NLRs) constitute a large family of mammalian genes that are homologous to innate immune defense genes extending back to the plant kingdom. In animals, NLRs function as important components of inflammation and immunity. To date, at least five NLR proteins are predicted to form a large complex termed the inflammasome, which serves as a scaffold for caspase 1 activation and subsequent release of bioactive IL-1β and related cytokines. Other NLR proteins, such as NOD1, NOD2, and NLRX1, have been shown to function as regulators of inflammatory cytokines, chemokines, and antimicrobial peptides (1, 2); however, the in vivo functions of the majority of NLRPs remain to be elucidated.
We have previously shown that human NLRP12 is expressed by monocytes/macrophages and granulocytes and inhibits the activation of noncanonical NF-κB by associating with and inducing proteasome-mediated degradation of NF-κB–inducing kinase (3, 4). More recently, mutations that result in a truncated form of NLRP12 are linked to hereditary periodic fevers that manifest with recurrent fevers, joint pain, and skin urticaria (5). In addition, a single nucleotide polymorphism in intron 9 of NLRP12 is loosely associated with atopic dermatitis (6). In this study, we present, to our knowledge, the first identified in vivo role for NLRP12 by demonstrating its role in the migration of dendritic cells (DCs) and neutrophils. This impacts contact hypersensitivity (CHS), but cannot be attributed to impaired IL-1β production and hence is distinct from the inflammasome function.
Materials and Methods
Mice
Nlrp12−/− and Nlrp3−/− mice were generated by homologous recombination and backcrossed for nine generations to C57BL/6 (The Jackson Laboratory, Bar Harbor, ME) and maintained in specific pathogen-free housing. OT-II mice, which express the OVA 323–339-specific TCR transgene on the C57BL/6 background, were kindly provided by M. Croft (La Jolla Institute of Allergy and Immunology, La Jolla, CA). Experiments were performed with 6–12-wk-old age- and sex-matched mice in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and the Institutional Animal Care and Use Committee guidelines of University of North Carolina at Chapel Hill (Chapel Hill, NC).
Nlrp12−/− genotyping
Genomic tail DNA was amplified with F1 5′-CCCACAAAGTGATGTTGGACTG-3′, F2 5′-GCAGCGCATCGCCTTCTATC-3′, and R1 5′-GAAGCAACCTCCGAATCAGAC-3′.
Expression analysis of Nlrp12
cDNA was synthesized from total RNA using Moloney murine leukemia virus reverse transcriptase (Invitrogen, Carlsbad, CA). Nlrp12 intron-spanning primers were: forward, 5′-GTCCAGACTCAGTCCACATA-3′, reverse 5′-GTATAAGGCCAGCTCGATCA-3′; and GAPDH: forward, 5′-TGAAGCAGGCATCTGAGGG-3′, reverse 5′-CGAAGGTGGAAGAGTGGGAG-3′. Cell populations were isolated as described: T cells and B cells by negative selection; neutrophils from bone marrow (BM) (7); DCs (8), macrophages (9), mast cells (10), and osteoclasts (11) from BM precursors; Raw264.7 macrophages from American Type Culture Collection (Manassas, VA); and resident peritoneal macrophages by lavage with PBS and overnight adherence.
Generation of DCs and isolation of neutrophils
Contact hypersensitivity
Mice were sensitized by topical application of hapten to the footpads and depilated abdomen (12): either 200 μl 3% oxazolone (Sigma-Aldrich) in ethanol or 200 μl 0.5% FITC (Sigma-Aldrich) in 1:1 acetone/dibutyl phthalate (Sigma-Aldrich). Five days later, 20 μl 1% oxazolone in ethanol or 20 μl 0.5% FITC in acetone/dibutyl phthalate was topically applied to one ear. The opposite ear was mock-treated with solvent, and control mice were treated with solvent on both ears. After 24 h, mice were euthanized, and 8-mm circular samples of ear tissue were excised and weighed, then the weight of the carrier-treated ear was subtracted from that of the hapten-treated ear. Ear tissue was frozen on dry ice (see below) or fixed in formalin, paraffin embedded, sectioned, and stained with H&E. Immune cell infiltration was quantified as average pixel density (× 104) using ImageJ software (National Institutes of Health, Bethesda, MD) from four fields per ear.
Ear tissue homogenates
Individual ears were manually homogenized in T-PER reagent (Thermo Fisher Scientific, Waltham, MA) using RNase/DNase-free plastic pestles (Kimble Kontes, Vineland, NJ) then sonicated. Total protein concentration was determined by Bradford assay (Bio-Rad, Hercules, CA). IL-1β, TNF-α, and myeloperoxidase (MPO) were measured by ELISA (IL-1β and TNF-α, BD Biosciences, San Jose, CA; MPO, Hycult Biotechnology, Uden, The Netherlands).
FITC-induced in vivo migration
Twenty microliters 0.5% FITC in 1:1 acetone/dibutyl phthalate was applied topically to both ears (13). After 24 or 48 h, draining (auricular and cervical) and nondraining lymph nodes (LNs) were removed. Cells were stained with fluorochrome-conjugated Abs: anti-CD11c (HL3, BD Biosciences), anti-CD19 (eBioscience, San Diego, CA), and anti-langerin/CD207 (eBioL31, eBioscience) and analyzed by flow cytometry using a CyAn ADP flow cytometer (Beckman Coulter, Fullerton, CA) and FlowJo software (Tree Star, Ashland, OR). Ear epidermal sheets (14) were stained with biotin-labeled anti–I-Ab (AF6-120.1, BD Biosciences) plus streptavidin-Alexa Fluor 595 (Invitrogen), then visualized by fluorescent microscopy. DCs per ×400 field are the mean of four fields per sample counted by a blinded reader.
OVA-induced in vivo migration
Mice were injected s.c. into one footpad with 20 μl Alexa Fluor 647-labeled OVA (2 mg/ml in PBS; Invitrogen) emulsified in CFA (15). After 24 h, draining (popliteal) and nondraining LNs were removed and analyzed by flow cytometry as above.
In vitro migration
BM-derived DCs (BMDCs) were seeded at 2 × 105 per upper well of 96-well transwell plates with 5-μm pores (ChemoTx System, NeuroProbe, Gaithersburg, MD), over chemokines (PeproTech, Rocky Hill, NJ) in serum-free RPMI and incubated at 37°C for 3 h. Migrated cells were quantified using XTT (Sigma-Aldrich). For neutrophil migration: 3-μm pores, cultured 40 min at 37°C, migrated cells quantified using ToxiLight Bioassay Kit (Lonza, Basel, Switzerland).
Statistics
Central tendencies are presented as mean ± SEM. Pairwise comparisons were made using two-tailed tests, all α = 0.05: Mann–Whitney U test (in vivo, Figs. 1B–E, 3B–D); Student t test (in vitro, Figs. 2F, 2G, 4I); Wilcoxon matched-pairs (paired data, Fig. 3E); one-sample Wilcoxon signed-rank, hypothesized median 100% (Fig. 3F). Outliers were by Grubb’s test and statistics computed with Prism 4 (GraphPad, San Diego, CA).
Results and Discussion
Nlrp12−/− mice were generated by replacing a region of exon 3 containing the Walker A and Walker B sequences with the neomycin resistance gene (Fig. 1A, 1B). Nlrp12−/− mice displayed no gross abnormalities, and there were no identifiable abnormalities in the cellularity of the peripheral blood, BM, spleen, or LNs (Supplemental Fig. 1). Similar to human NLRP12, murine Nlrp12 was expressed in the BM and spleen and at the cellular level in neutrophils and DCs (Fig. 1C). Unlike human NLRP12, however, murine Nlrp12 was not highly expressed in transformed Raw264.7 macrophages, macrophages differentiated from BM precursors, or macrophages resident within the peritoneum.
NLRP12 is not required for IL-1β secretion in response to TLR agonist including LPS + ATP stimulation
Given the critical role of other NLR proteins in IL-1β processing, we tested the ability of Nlrp12−/− cells to produce IL-1β poststimulation with a variety of elicitors. No significant difference in IL-1β production was detected in Nlrp12−/− versus wild-type (WT) BM cells or BMDCs stimulated with LPS, LPS + ATP, other TLR ligands, or TNF-α (Fig. 1D). As a positive control, Nlrp3−/− cells failed to produce IL-1β when properly stimulated (Fig. 1D). Nlrp12 did not affect the production of IL-12p40, IL-6, and TNF-α (Supplemental Fig. 2A–E). In vivo, Nlrp12−/− mice treated with two doses of LPS trended toward reduced survival, but the difference was not statistically different, and there was no difference in liver and kidney function (Supplemental Fig. 2F–H). Thus, more extensive studies are required to delineate the role of Nlrp12 in LPS-mediated inflammatory responses.
NLRP12 promotes hapten-induced CHS
The importance of NLRP12 in cutaneous inflammation has been suggested through its recent linkage to human hereditary periodic fevers with skin urticaria (5). We evaluated the role of NLRP12 in cutaneous inflammation by assessing CHS, a mouse model of allergic dermatitis. WT and Nlrp12−/− mice were sensitized topically on the abdomen with hapten and either oxazolone or FITC and elicited 5 d later on the ear. Compared to controls, Nlrp12−/− mice displayed a weaker response to both haptens, as indicated by significantly reduced swelling (Fig. 2A–C) and reduced cellular accumulation at the site of elicitation (Fig. 2D). The partial effect of Nlrp12 deletion is typical of other NLR responses in mice (16). During a CHS response, neutrophils migrate out of the bloodstream and into the affected skin tissue (17). In Nlrp12−/− mice, the accumulation of MPO-positive neutrophils in the skin was significantly reduced in hapten-treated ears compared with WT mice (Fig. 2E).
NLRP12 potentiates DC migration from the periphery to the draining LNs
Given the critical role of DCs in the CHS model (19) coupled with the high expression of Nlrp12 in these cells, we determined if DC function is dependent upon NLRP12. Expression of MHC class II (MHC-II) and costimulatory factors CD80, CD86, and CD40 was not affected by Nlrp12 during BMDC maturation (Supplemental Fig. 3A–C). To determine if Nlrp12 affected Ag processing and presentation, WT and Nlrp12−/− BMDCs were pulsed with OVA and then cocultured with CFSE-labeled OT-II splenocytes specific for the OVA peptide 323–339. WT and Nlrp12−/− BMDCs induced equivalent Ag-dependent T cell proliferation, indicating that NLRP12 was not required for Ag presentation by DCs (Supplemental Fig. 3D, 3E).
A key function of DCs in the CHS model is to collect peripheral Ags and migrate to draining LNs. To evaluate DC migration in vivo, FITC was applied topically to the ears of WT and Nlrp12−/−mice. Draining LNs were removed 24 and 48 h later, and the presence of FITC+ CD11c+ DCs was assessed by flow cytometry. FITC+ DCs in the draining LNs of Nlrp12−/− mice were significantly reduced compared with WT mice 24 and 48 h after FITC application (Fig. 3A–C). This was specific to Nlrp12−/− mice, as DC migration was comparable to WT in mice deficient in a related NLR, Nlrp3−/− (Fig. 3D). Migrated DCs in the draining LNs of WT mice were MHC-II+ (mean ± SEM, 96.5 ± 0.86%) and immunostained for the dermal/epidermal DC marker langerin/CD207 (85.0 ± 2.30%). Similar results were obtained with Nlrp12−/− mice (MHC-II+, 96.8 ± 0.65%; langerin/CD207+, 88.1 ± 1.91%). In agreement with earlier results (Fig. 3A–C), the number of FITC+ langerin/CD207+ cells in the draining LNs of Nlrp12−/− mice was less than half that of WT mice (Fig. 3E), suggesting a defect in DC migration from the skin to the draining LNs. Nlrp12−/− DCs did not traffic to other immune organs, as FITC+ CD11c+ DCs were not detected in nondraining LNs, BM, or spleen (Fig. 3A and data not shown). Quantification of DCs in untreated skin revealed that Nlrp12 did not affect DC numbers in untreated mice (Fig. 3F). However, following FITC treatment, the number of WT skin DCs decreased by >40%, whereas the number of Nlrp12−/− skin DCs changed by <10% (Fig. 3F), indicating that Nlrp12 is required for DC egress from the skin.
To further examine the migratory capacity of Nlrp12−/− DCs, we measured migration of DCs to the draining LNs in response to s.c. OVA Ag. The absence of Nlrp12 reduced OVA+ DCs in the draining LNs by nearly 75% (Fig. 3G). These results support migratory defects in Nlrp12−/− DCs.
Nlrp12−/− DCs and neutrophils fail to respond to chemokines in vitro
DCs migrate to draining LNs in response to chemokine engagement of CCR7 and CXCR4 on DCs (20, 21). Nlrp12 did not affect the expression of CCR7 and CXCR4 on BMDCs (Supplemental Table I). However Nlrp12−/− BMDCs demonstrated significantly reduced migration toward CCR7 and CXCR4 ligands CCL19, CCL21, and CXCL12 in an in vitro transwell assay (Fig. 4A–C, Supplemental Table II). In contrast, Nlrp3−/− BMDCs exhibited normal migration toward these chemokines (Fig. 4D–F). As a negative control, BMDCs from all genotypes failed to migrate toward CCL5 (Fig. 4G, 4H), a chemokine for immature DCs (22). The migration of Nlrp12−/− neutrophils to the neutrophil-attracting chemokine CXCL1 was also reduced by ~50% when compared with WT neutrophils (Fig. 4I).
NLR proteins intersect various pathways that are integral to inflammation and immunity (1). Based upon our previous observation of inappropriate noncanonical NF-κB activation in human monocytic cell lines with reduced NLRP12, we predicted a proinflammatory phenotype for Nlrp12-deficient mice. Surprisingly, proinflammatory cytokine production was unaffected by the absence of NLRP12, and the CHS response mounted in these mice was significantly attenuated relative to WT mice. Nlrp12 did not affect BMDC maturation or Ag-presenting functions in vitro. Instead, Nlrp12 affected DC migration from the periphery to the draining LNs during CHS and s.c. immunization and migration to LN-homing chemokines. Nlrp12 also affected neutrophil response to chemotactic stimuli, indicating a central role for NLRP12 in licensing cellular migration. Hereditary periodic fever is associated with two mutations in human NLRP12 (5), and it is tantalizing to speculate that altered neutrophil migration or retention of activated DCs in the periphery may lead to the recurring cutaneous inflammation experienced by these patients. The current study does not address if DC progenitors can properly migrate from the BM to the skin in the absence of Nlrp12. Also, the mechanisms by which DCs and neutrophils rely on Nlrp12 for optimal migratory capacity remain unclear.
We did not find an in vivo effect of NLRP12 on proinflammatory cytokine production. In contrast to these results, a previous study showed that silencing of NLRP12 with small hairpin RNA leads to increased production of IL-6 and TNF-α in human cells of monocytic lineage (4). This implies that NLRP12 may exhibit distinct functions in DCs and monocytes/macrophages; however, because mouse macrophages do not express detectable Nlrp12, its deletion is not predicted to affect macrophage function in mice. Nonetheless, it will be interesting to investigate if the absence of Nlrp12 affects monocyte egress from the BM to sites of inflammation. We also were unable to find an effect of Nlrp12 deletion on IL-1β production, which is a well-established role for other NLRs. Human NLRP12 can colocalize with inflammasome components and promote IL-1β secretion when overexpressed (23). It is possible that the proper agonist/ligand, once identified, can activate NLRP12 inflammasome function. In agreement with previous studies from our group and others (4, 24), we found that NLRP12 was not required for IL-1β secretion in response to TNF-α or TLR ligands ± ATP.
In summary, we reveal a regulatory role for NLRP12 in licensing cellular migration that affects a DC-dependent model of cutaneous inflammation. Unlike Nlrp3−/− mice, this is not correlated with a defect in inflammasome activation and IL-1β production. Considering that NLRs appear to cluster by their functional properties, it is likely that other NLRs will be found to similarly affect physiological processes by controlling cellular migration.
Acknowledgements
Disclosures J.D.L. and J.B. are current employees and shareholders at GlaxoSmithKline. The other authors have no financial conflicts of interest.
Footnotes
This work was supported by National Institutes of Health Grants A1063031, AI067798, DK38108, DE016326, and HL068141, a National Institutes of Health Research Service Trainee Award (to J.C.A., J.E.W.), and the American Cancer Society (to J.D.L.). J.P.-Y.T is a Sandler Program Awardee.
The online version of this article contains supplemental material.
Abbreviations used in this paper:
- BM
bone marrow
- BMD
bone marrow-derived
- BMDC
bone marrow-derived dendritic cell
- CHS
contact hypersensitivity
- cLPS
commercial LPS
- DC
dendritic cell
- iDCs
immatureDCs before maturation stimulus
- LN
lymph node
- mac
macrophage
- mDCs
DCs after TNF-α maturation
- MHC-II
MHC class II
- NLR
nucleotide-binding domain leucine-rich repeat
- no tx
no treatment
- pIC
polyinosinic-polycytidylic acid
- pLPS
phenol-purified LPS
- UnTx
untreated
- WT
wild-type.