Cutting Edge: LL-37–Mediated Formyl Peptide Receptor-2 Signaling in Follicular Dendritic Cells Contributes to B Cell Activation in Peyer’s Patch Germinal Centers

The gut mucosa is exposed to various microorganisms, and secretory IgA plays a major regulatory role in mucosal immune homeostasis through immune exclusion and neutralization by binding to pathogenic components (1). Class-switch recombination to produce IgA occurs in germinal centers (GCs) of Peyer’s patches (PPs), sites of mucosal immune induction (2). Specifically, GCs are compartments within a secondary lymphoid organ and are closely associated with B cell clonal expansion and affinity maturation, leading to the production of high-affinity Abs, a hallmark of adaptive immunity, against specific Ags (3). In peripheral lymph nodes (pLNs), GC formation is induced by pathogen infection. However, GCs in PPs are continuously active as a result of stimulation by commensal bacteria in the gut; mice treated with antibiotics (ABXs) or raised under germ-free conditions contain few PP GCs (4). Although the formation and maintenance of GC follicles in PPs are poorly defined, pre-existing GCs are closely associated with the induction of strong, highly synchronized, and oligoclonal IgA responses dominated by affinity-matured cells against T-dependent Ags following oral immunization (5).

During GC formation and maintenance, follicular dendritic cells (FDCs) derived from perivascular precursors of stromal cells have long been regarded as Ag-retaining and Ag-presenting reticular cells. However, recent reports demonstrate that FDCs are directly associated with the induction of humoral immunity through GC formation and maintenance (6). For example, FDCs express CXCL13, which is involved in the maintenance of an organized follicular structure in GCs by attracting CXCR5-expressing B or T cells, IL-6, and B cell activating factor (BAFF), which enhances B cell survival (7). Additionally, FDCs matured by lymphotoxin and TNF signaling derived from B cells express surface receptors, such as VCAM-1, ICAM-1, MadCAM-1, low-affinity Fc receptor, and TLRs (8). In pLNs, TLR4 signaling in FDCs enhances FDC activation and contributes to Ig class-switched high-affinity plasma and memory B cell formation (9). Although the expression of TLR transcripts in gut FDCs is lower than that of pLN FDCs, expression of the retinoic acid receptor and cosignaling of this receptor with TLRs directly regulate the increased IgA isotype switching through secretion of TGF-β and BAFF (8). Consequently, we expect that specific receptor-mediated signaling in gut FDCs is closely associated with GC maintenance.

The formyl peptide receptors (FPRs) are classic chemoattractant G protein–coupled receptors that are associated with leukocyte trafficking, cell differentiation, and wound healing (10, 11). FPRs are also considered innate immune sensors because various bacteria-derived chemotactic peptides, such as N-formyl peptide derived from Escherichia coli, Listeria peptides, and phenol-soluble modulin peptide from Staphylococcus aureus, are cognate ligands for these receptors (12–14). Additionally, fpr1- and fpr2-deficient mice show increased susceptibility to Listeria monocytogenes and Pneumococcal meningitis infection (15, 16). The human FPR family consists of FPR1, FPR2, and FPR3, and the mouse FPR (Fpr) family is composed of eight members. Among them, Fpr2 is a low-affinity receptor for N-formyl peptide and has several endogenous ligands, such as phosphoenolpyruvate, LL-37, serum amyloid A, and prion protein 106–126 peptide (11). Notably, some of these endogenous ligands, such as LL-37 and serum amyloid A, are expressed in the mucosal compartment as a result of exposure to specific microbes and their metabolites (17, 18). Thus, it is conceivable that Fpr2 is involved in mucosal immune regulation, although its precise role is unclear. In this study, we identified the expression of Fpr2 in PP FDCs. We also found that Fpr2 signaling via interaction with LL-37 ligand enhances the levels of Cxcl13 and Tnfsf13b (BAFF gene) transcripts, as well as in vitro B cell proliferation and activation.

Syngeneic specific pathogen–free (SPF) BALB/c mice were purchased from Charles River Technology through Orient Bio (Sungnam, Korea). ABX-treated mice were subjected to oral administration of an ABX mixture (1 g/l ampicillin, 0.5 g/l vancomycin, and 0.1 g/l polymyxin) in drinking water for 30 d. All chemicals were purchased from Sigma Chemical (St. Louis, MO), unless otherwise specified. Synthetic LL-37 (LLGDFERKSKEKIGKEFKRIVQRIKDFLRNLVPRTES) and Biotin-LC–LL-37 peptides and WRWWWW [WRW(4)]-NH₂ were purchased from ANYGEN (Jangseong, Korea) and AnaSpec (Fremont, CA), respectively. rTNF-α was purchased from R&D Systems (Minneapolis, MN). Collagenase D and DNase I were purchased from Roche Applied Science (Mannheim, Germany).

Anti–FDC-M1, anti-B220 (biotin- or allophycocyanin-Cy7 conjugated), and anti–VCAM-1 (FITC-conjugated) Abs, purified mouse IgA, purified rat IgG2c (biotin-conjugated), streptavidin-FITC, streptavidin-allophycocyanin, streptavidin-PE, streptavidin–PE–CF594, and 7-aminoactinomycin D (7-AAD) were purchased from BD Biosciences (Franklin Lakes, NJ). Anti–FDC-M2 Ab was purchased from Amsbio (Cambridge, MA). Anti-TLR2 and Cy3-conjugated anti-CD45 Abs were purchased from Abcam (Cambridge, MA). FITC- or PE-conjugated secondary Abs were purchased from Santa Cruz Biotechnology (Dallas, TX). Anti-Fpr2, Alexa Fluor 594–, Alexa Fluor 680– or Alexa Fluor 700–conjugated secondary Abs, and DAPI were purchased from Invitrogen (Grand Island, NY). Biotin-conjugated anti-FDC–M2 Ab was purchased from eBioscience (San Diego, CA). FITC-conjugated GL7 and anti-IgM Abs were purchased from Bethyl Laboratories (Montgomery, TX).

PP FDCs were prepared according to a procedure reported previously with minor modifications (8, 19). Briefly, ∼50 PPs were collected from 10 mice, and PP lymphocytes (PPLs) were prepared after tissue digestion with collagenase. B220⁻ PPLs were collected using an AutoMACS Separator (Miltenyi Biotec, Bergisch Gladbach, Germany) and stained with anti-CD3 and anti–FDC-M1 Abs and 7-AAD. Data were acquired from the collected live B220⁻CD3⁻FDC-M1⁺ (FDC-M1⁺) cells using a FACSAria III (BD Biosciences) and analyzed using FlowJo software (TreeStar, Ashland, OR).

After stimulating the isolated FDCs with 2 μg/ml LPS or 1 μM synthetic LL-37 peptide for 2 h, 100 FDCs were cocultured with splenic B cells (5 × 10³) tagged with CFSE (eBioscience). After 3 d, cells were stained with 7-AAD and analyzed by flow cytometry. The absolute cell number was calculated using 123count eBeads (eBioscience), according to the manufacturer’s instructions. To analyze the level of IgA in the culture supernatant, 100 FDCs, which had been stimulated with the indicated molecules, were cocultured with splenic or PP B cells. After 5 d, the amount of IgA in the culture supernatant was measured by sandwich ELISA.

SPF BALB/c mice were orally administered synthetic LL-37 peptide (50 μg) or rEGFP conjugated or not with LL-37 (10 μg). After 7 d, B220⁺IgA⁺, B220⁺IgM⁺, and B220⁺GL7⁺CD95⁺ PPLs were analyzed by flow cytometry. The numbers of EGFP-specific IgA⁺ cells in PP and lamina propria (LP) from some experiments were determined after 14 d by ELISPOT analysis.

The PP specimens prepared from SPF BALB/c mice were stained with a biotin-conjugated anti-FDC–M2 Ab, anti-Fpr2 Ab, and anti-CD45 Ab and then counterstained with DAPI. Sorted FDC-M1⁺ cells were cultured on fibronectin-coated dishes with TNF-α (5 ng/ml) and anti-LTβR Ab (1 μg/ml) for 3 d. After treatment with biotin-conjugated LL-37 peptide (10 μM) for 5 or 15 min, cells were fixed with 4% paraformaldehyde, permeabilized with 0.2% Triton X-100, stained with an anti-Fpr2 Ab, fluorescent dye–conjugated streptavidin, and anti-CD21/35 Ab, and counterstained with DAPI. Specimens were analyzed by confocal laser scanning microscopy (CLSM) (LSM 510 META; Carl Zeiss, Thornwood, NY).

Sorted FDC-M1⁺ cells were cultured for 3 d, treated with 10 μM Fpr2 antagonistic peptide [WRW(4)] for 1 h, and loaded with Fluo-4 AM using a Fluo-4 Calcium Imaging Kit (Molecular Probes), according to the manufacturer’s recommendations. Briefly, cells were loaded with Fluo-4 am at 37°C for 30 min, incubated for 30 min at room temperature, washed, and suspended in live cell imaging solution containing 2 mM glucose. Cells were subjected to CLSM, and calcium influx was recorded by fluorescence emission at 506 nm after stimulation with synthetic LL-37 peptide for 200 s. Data were analyzed using LSM 501 software.

Sorted FDC-M1⁺ cells were treated or not with synthetic LL-37 peptide (10 μM) for 2 h. Total RNA was prepared using an RNeasy Plus Micro Kit, and cDNA was prepared with a QuantiTect Reverse Transcription kit (both from QIAGEN, Hilden, Germany). Quantitative real-time PCR (qRT-PCR) reactions were run on an ABI 7500 (Applied Biosystems, Waltham, MA) with a QuantiTect SYBR Green PCR Kit using the following specific primer sets: Cxcl13-F, 5′-TGT CCA AAG CAA AAG TCT GTC T-3′ and Cxcl13-R, 5′-ATA GTG GCT TCA GGC AGC TC-3′; Tnfsf13b-F, 5′-TGG GAG CAG AGT CCT GAT GT-3′ and Tnfsf13b-R, 5′-GCT TCT GGG TGA GTA CTG CT-3′; and Fpr2-F, 5′-GAG ACC TCA GCT GGT TGT GC-3′ and Fpr2-R, 5′-ACC ACC ACT TCT GAT CCA TTC A-3′. The relative quantity of cDNA was calculated by the ΔΔ threshold cycle method using 18S rRNA gene as a reference. Experimental results were acquired from three independent experiments, in which each gene was assayed in triplicate.

Statistical analyses were performed using Prism 6 software (GraphPad, La Jolla, CA). Data are presented as mean ± SD of triplicates. Differences among the means of multiple independent variables were compared between the control and each group using one-way ANOVA, followed by the Tukey post hoc test. Differences in mean values were considered significant at p < 0.05.

Regulation between innate and adaptive immunity in PPs is pivotal in host defense, because PP cells are directly or indirectly affected by microorganisms introduced from the intestinal lumen (20). We assumed that Fpr2 contributes to immune regulation in PPs to maintain GCs, because it may play a role as an innate receptor linking innate and adaptive immunity. To detect the expression of Fpr2 in PP cells, we investigated PP tissue slices containing GCs prepared from SPF mice. GCs primarily consisted of leukocyte lineage cells, such as B and T cells, and stroma-derived cells, such as FDCs; anti-CD45 and anti–FDC-M2 Abs can distinguish these cells. We initially detected the expression of Fpr2 in CD45⁻FDC-M2⁺ cells (Fig. 1A). FDCs can be further identified using anti–FDC-M1 or anti–FDC-M2 Ab; anti–FDC-M1 Ab interacts with fat globule epidermal growth factor VIII expression, which increases during the GC reaction, whereas anti–FDC-M2 Ab recognizes complement component C4 in naive FDCs (19). For further confirmation of the expression of Fpr2 in PP FDCs, B220⁻ PP cells were enriched, FDCs were sorted using anti–FDC-M1 Ab, and PP FDC-M1⁺ cells were analyzed. As reported previously, PP FDC-M1⁺ cells displayed the highest expression of molecules related to the GC reaction, such as VCAM-1 and Cxcl13 and Tnfsf13b transcripts, together with low-level TLR2 expression (Fig. 1B, 1C) (8). Notably, TLR transcript expression was reported to be downregulated in PP FDCs compared with that in pLNs (8), and we found that Fpr2 was expressed in PP FDCs. Collectively, these results show that Fpr2 is expressed in PP FDCs and suggest its role as a mucosal innate immune sensor.

We next analyzed whether signals are transduced from Fpr2 into PP FDCs through the interaction with a cognate ligand of Fpr2. To confirm the activation of Fpr2 on FDCs, FDC-M1⁺ cells sorted from PPs were stimulated with an endogenous Fpr2 ligand, cathelin-related antimicrobial peptide (CRAMP; a mouse ortholog of human cathelicidin LL-37), and the expression of Tnfsf13b and Cxcl13 transcripts was enhanced by CRAMP treatment (Supplemental Fig. 1) (21). In addition, LL-37 treatment of FDC-M1⁺ cells resulted in significantly (p < 0.05) increased expression of the Cxcl13 and Tnfsf13b transcripts. Additionally, the Fpr2 transcript level increased significantly (p < 0.05) by ∼60-fold in response to LL-37, suggesting positive-feedback regulation of Fpr2 in FDCs by an endogenous ligand of Fpr2 (Fig. 2A). We further investigated the direct interaction between LL-37 and Fpr2 in FDC-M1⁺ cells and found that Fpr2 was distributed in the subcellular compartment of purified B220⁻FDC-M1⁺ cells that were not stimulated with LL-37, whereas Fpr2 was localized at the cell periphery, together with LL-37, as long as 5 min after stimulation with LL-37 (Fig. 2C).

We next determined whether LL-37–mediated Fpr2 signaling triggers intracellular Ca²⁺ mobilization, because activation of Fpr2 triggers intracellular Ca²⁺ mobilization by G protein–dependent phospholipase C stimulation, which, in turn, induces chemotaxis (22). When we monitored Ca²⁺ influx by CLSM during incubation of isolated PP B220⁻FDC-M1⁺ cells that were pretreated or not with the Fpr2 antagonist peptide WRW(4) (10 μM), we found that LL-37 treatment of FDC-M1⁺ cells resulted in an increased level of intercellular Ca²⁺ that was abrogated in WRW(4)-pretreated cells (Fig. 2B, 2C). These results suggest that the interaction between LL-37 and Fpr2 in FDC-M1⁺ cells transduces the positive signal into the cells. In fact, LL-37 is continuously expressed in the ileum, and the finding that its expression is regulated negatively by Shigella infection and positively by microbiota-derived metabolites, such as short-chain fatty acids, suggests that LL-37–mediated FPR-2 signaling in PP FDCs likely plays a role in GC formation and the maintenance of PPs (18, 23).

To confirm the biological function of LL-37–mediated Fpr2 signaling in PP FDCs, we first monitored B cell proliferation, because the Tnfsf13b transcript, whose expression was enhanced by LL-37 treatment in PP FDCs, encodes the B-lymphocyte stimulator BAFF (24). When CFSE-stained splenic B cells were cultured for 3 d with isolated PP FDC-M1⁺ cells that were stimulated or not with LPS or LL-37, B cells cultured with PP FDCs activated by LPS or LL-37 exhibited significantly (p < 0.01) enhanced proliferation compared with B cells cultured in the absence of FDCs or cultured with FDCs without additional stimulation with LPS or LL-37 (Fig. 3A). We next measured the amount of IgA secreted from B cells that had been cocultured in vitro with FDCs that were stimulated or not with LPS or LL-37, because active PP FDCs express a high level of TGF-β1, which is involved in IgA isotype switching (8). The FDCs activated by LL-37 promoted efficient expression of IgA from B cells derived from PPs, as well as from the spleen (Fig. 3B). Additionally, enhanced IgA class switching was observed in PPs of mice orally administered rEGFP conjugated with LL-37 (Fig. 3C).

For further investigation of the role of Fpr2 in PP FDCs, GC B cells were analyzed from PPs of SPF mice orally administered the synthetic LL-37 peptide (50 μg). The recruitment of GC B cells expressing CXCR5, GL7, and CD95, which is dependent on CXCL13 secreted from FDCs and follicular Th cells, increased in SPF mice treated with LL-37 (Fig. 3D); this accords well with the increase in CXCL13 transcript expression after LL-37 treatment (Fig. 2A). This LL-37–mediated enhancement of GC B cell recruitment in PPs was more prominent in ABX-treated mice (Supplemental Fig. 2A). It is conceivable that the results from ABX-treated mice are closely associated with the level of Fpr2 expression in PP FDCs, because the level of Fpr2 was downregulated in FDC-M1⁺ cells prepared from ABX-treated mice compared with that in SPF mice and was increased by LL-37 treatment in vitro (Supplemental Fig. 2B). Additionally, LL-37 treatment increased the expression of Cxcl13 and Tnfsf13b transcripts in FDC-M1⁺ cells prepared from ABX-treated mice, and the increase was abrogated by WRW(4) pretreatment (Supplemental Fig. 2C). Consequently, these results support the possibility that the increase in the GC B cell population was dependent on Fpr2-mediated signaling in FDCs. Given that the frequency of GC B cells increased (Fig. 3D), and the interaction between FDCs and GC B cells triggers the generation of plasma cells producing Ag-specific Abs (25), it is conceivable that the increase in GC B cells positively regulated the plasma cells producing Ag-specific Abs. Finally, we confirmed this speculation using EGFP as an Ag and found that the oral administration of LL-37–conjugated EGFP evoked an increase in EGFP-specific IgA⁺ cells in PPs, as well as in LP, compared with those of EGFP alone (Fig. 3E). Taking these findings together, Fpr2 signaling potentiated FDC function in PPs via cognate ligand LL-37 binding.

Under steady-state conditions, gut epithelial cells continuously express cathelicidin LL-37 as a result of stimulation by butyrate, a microbial metabolite. Importantly, LL-37 expression is low in Crohn’s disease patients (18). We believe that the decreased LL-37 expression causes low-level production of IgA as a result of reduced FDC activation and is closely associated with disease susceptibility, because IgA binds preferentially to the colitogenic microbiota (26). In addition, CRAMP-knockout mice exhibited increased colonization of pathogenic intestinal bacteria (27). Therefore, we believe that Fpr2 signaling via LL-37 in FDCs facilitates mucosal homeostasis by contributing to maintenance of GCs in PPs. Collectively, these data confirmed that Fpr2 stimulation by LL-37 in PP FDCs may support the recruitment of germinal B and T cells, as well as B cell proliferation and activation. We believe that this interaction between LL-37 and FDCs supports the prolonged induction of PP GCs. Additionally, because reutilization of PP GCs promotes highly synchronized and affinity-matured gut IgA responses against the orally introduced Ags (28), we suggest that a GC-stimulation strategy using Fpr2 ligand LL-37 can be applied for the development of an effective mucosal vaccine adjuvant. Indeed, we reported previously that oral administration of an LL-37–conjugated Ag enhanced the Ag-specific immune responses by modulating the mucosal environment (29).

CLSM and flow cytometry were performed at the Center for University-Wide Research Facilities of Chonbuk National University.

The authors have no financial conflicts of interest.