Chemoattraction of Macrophages, T Lymphocytes, and Mast Cells Is Evolutionarily Conserved within the Human α-Defensin Family¹

Human defensins are short cationic β-sheet peptides with molecular masses ranging from 3 to 5 kDa. They are divided into two families, α and β, based on the differences in the positions and three disulfide bonds of their six cysteine residues (1). α- and β-defensins are products of distinct gene families that are thought to have evolved from an ancestral β-defensin gene (2). Despite only modest amino acid sequence identity among α-defensins, a striking similarity on the level of tertiary structures was found (3, 4).

α-Defensins are produced primarily by neutrophils (human neutrophil peptides (HNP))⁵ or intestinal Paneth cells (human defensins (HD)) (5). The best known property of defensins is their ability to kill microbial pathogens. Their spectrum of activity includes bacteria, fungi, and viruses (6). An important role of α-defensins in the maintenance of intact mucosal barriers is suggested by the finding that mucosal peptide antibiotics are involved in the pathogenesis of inflammatory bowel disease (7). Notably, a deficiency of HD5 in Paneth cells is associated with Crohn′s disease of the ileum (8).

Recently, α-defensins have been shown to play significant roles in adaptive as well as innate immunity. They modulate the cytokine response of monocytes and lymphocytes (9, 10) and promote systemic Ag- or tumor-specific immune responses (9, 11, 12). In keeping with these findings, HNP-1 and -2 have been found to mobilize immature dendritic cells (DC) and naive but not memory T lymphocytes via as yet unidentified receptors (13). However, controversial results exist with regard to HNP-1/-2-induced migration of monocytes (13, 14, 15), whereas HNP-3 and HD5 are supposed to completely lack chemotactic activity (4, 14, 15). Clearly, studies of the immunomodulatory properties of α-defensins are still in their infancy despite the well-appreciated importance of these peptides for the host defense (16).

The purpose of the present study was to investigate the impact of human α-defensins on cells of the innate and adaptive immune system to 1) identify those immune cells that are primary targets for α-defensins, 2) define receptor usage by α-defensins, and 3) uncover cross-regulation between α- and β-defensins.

Pertussis toxin, MAPK ERK inhibitor PD098059, JNK inhibitor SP600125, p38 inhibitor SB203580, the general protein kinase C (PKC) inhibitor GF109, Gö6976, which inhibits calcium-dependent PKC isotypes, and PMA were all purchased from Calbiochem. Recombinant (human or murine) CCL3, CCL19, and CCL20 were obtained from PeproTech. Recombinant human anaphylatoxin C5a was described elsewhere (17). Commercial sources for HNP-1 were The Peptide Institute, Bachem, and PeproTech; HNP-3, HD-5, and human β-defensin (HBD) 2 were obtained from The Peptide Institute.

Leukocytes were obtained by leukapheresis from volunteer blood donors in the Department of Transfusion Medicine, University Clinic Göttingen (Göttingen, Germany). PBMC were isolated as described previously (17) and cultured for 1 h at 1 × 10⁷ cells/ml in endotoxin-free RPMI 1640 (Biochrom) supplemented with 5% heat-inactivated autologous serum in flat-bottom plates. After washing off nonadherent cells, adherent mononuclear cells (>90% CD14⁺ monocytes) were used to generate macrophages or DC. DC were cultured in RPMI 1640 supplemented with 10% FCS (PAN Biotech), 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, GM-CSF (300 U/ml), and IL-4 (300 U/ml) (both from R&D Systems). After 7 days, cultured DC expressed high levels of HLA-DR but no CD14 and were characterized as immature due to their moderate expression of CD86 and low expression of CD83 (17).

To obtain macrophages, adherent PBMC were cultured in RPMI 1640 supplemented with 5% FCS, 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 100 U/ml GM-CSF. After 1–3 days, cultured macrophages expressed low to moderate levels of HLA-DR but high levels of CD14.

The human mast cell line HMC-1 was cultured as described elsewhere (18).

Human peripheral blood CD3⁺CD4⁺CD45RA⁺ naive or CD3⁺CD4⁺CD45RO⁺ memory T cells were purified from PBMC (see above) by magnetic cell sorting using a negative selection technique according to the manufacturer’s recommendations (Miltenyi Biotec). The purity of T cell subsets was verified by FACS analysis and was always >95%.

Generation of murine immature DC derived from BALB/c bone marrow precursor cells has been described previously (17). Nonadherent bone marrow cells were cultured in RPMI 1640 containing 5% FCS, 0.1 mM nonessential amino acids, 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, 50 μM 2-ME, along with 100 U/ml murine GM-CSF (PeproTech) for a total of 6–7 days.

Murine bone marrow-derived macrophages were generated by culturing nonadherent bone marrow cells from BALB/c mice in RPMI 1640 containing 5% FCS, 0.1 mM nonessential amino acids, 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, 50 μM 2-ME, and murine M-CSF (20 ng/ml; PeproTech) for 6–7 days. Adherent cells were >95% macrophages as judged by morphology and homogenous surface expression of C5a and C3a receptors and absence of CD11c.

The murine macrophage cell line J774A.1 was cultured as described elsewhere (19).

In vitro chemotaxis was assayed using the HTS Transwell-24 system from Corning. Cells diluted at 1 × 10⁶/ml in migration buffer (RPMI 1640 with 1% BSA) were placed in the upper wells, whereas the various chemoattractants diluted in migration buffer as indicated were added to the lower wells. For chemotaxis of human and murine macrophages and DC, polycarbonate membranes with a pore size of 8 μm were used and incubation was performed at 37°C in a 5% CO₂ atmosphere for 90 min; if human T lymphocytes were used, pore size was 5 μm and incubation time was 4 h and for HMC-1 cells pore size was 8 μm and incubation was 3 h. Migration was stopped and migrated cells were detached by placing Transwell chambers for 15 min on ice. Thereafter, migrated cells in the lower chambers were counted using a hemocytometer. All determinations were performed in duplicate.

To investigate the role of PKC in receptor desensitization, cells (1 × 10⁶ in 1 ml of medium) were preincubated in the presence or absence of the PKC inhibitors GF109 (4 μM) or Gö6976 (2 μM) at 37°C for 30 min and centrifuged.

For desensitization studies, cells (1 × 10⁶ in 1 ml of medium) were preincubated at 37°C for 30 min with selected chemotaxins (200 ng/ml) or PMA (2 μM) and centrifuged.

To study the role of Gα_i in signal transduction, cells were incubated for 2 h in the presence of different concentrations of pertussis toxin (Calbiochem) before their use in chemotaxis experiments. Treatment with pertussis toxin did not affect cell viability, as assessed by trypan blue exclusion.

To study the role of MAPK in signal transduction, cells were preincubated at 37°C for 30 min with PD098059, SP600125, or SB203580 (20 μM each) and centrifuged. Treatment with MAPK inhibitors did not affect cell viability, as assessed by trypan blue exclusion.

All animal work was conducted in accordance with guidelines for animal welfare and was approved by the government of Lower Saxony, Germany. In vivo migration studies were performed as described elsewhere (17). SCID mice (strain CB-17 SCID of both sexes; weight 19–24 g) were obtained from Charles River Breeding Laboratories. Cells (1 × 10⁷ in 200 μl of PBS) were injected into the tail vein of SCID mice along with a chemotaxin (10 or 20 μg in 200 μl of PBS) that was injected into the peritoneal cavity. After 4 or 15–18 h, mice were sacrificed and peritoneal cells were harvested by lavage. Subsequently, cells were stained with anti-HLA-DR-PE in combination with FITC-anti-CD14 or FITC-anti-CD86 to identify migrated human macrophages and DC, respectively. Absolute numbers of migrated human cells were calculated from the percentage of HLA-DR⁺ cells as determined by FACS analysis and the total peritoneal cell count.

For desensitization studies, human cells (1 × 10⁷ in 1 ml of PBS) were preincubated at 37°C for 1 h with selected chemotaxins (2 μg), washed twice in PBS, and resuspended in 200 μl of PBS for injection. Pretreatment of cells did not affect cell viability, as assessed by trypan blue exclusion.

In vivo migration studies were performed as described elsewhere (17). Murine bone marrow-derived cells were labeled with the red fluorescent dye PKH-26 (Sigma-Aldrich) according to the manufacturer’s instructions. Cells (1 × 10⁷ in 200 μl of PBS) were injected into the tail vein of syngeneic BALB/c mice (weight 20–24 g; age 8–20 wk) along with a chemotaxin (10 or 20 μg in 200 μl, as indicated) that was injected into the peritoneal cavity. Approximately 15–18 h later, mice were sacrificed and peritoneal lavage performed. Subsequently, peritoneal cells were counted and analyzed by FACS. Absolute numbers of migrated labeled cells were calculated from the percentage of red fluorescent cells and the total peritoneal cell count.

For desensitization studies, labeled cells (1 × 10⁷ in 1 ml of PBS) were preincubated at 37°C for 1 h with selected chemotaxins (2 μg), washed twice in PBS, and resuspended in 200 μl of PBS for injection.

A total of 10⁶ cells was loaded in 700 μl of RPMI containing 5% FCS, 1 μM Fluo3-AM, and 0.02% Pluronic F127 (both from Molecular Probes). Subsequently, the cell suspension was diluted 2-fold with RPMI 1640/10% FCS and was incubated for 10 min at 37°C. Cells were washed twice with Krebs-Ringer solution composed of 10 mM HEPES (pH 7.0), 140 mM NaCl, 4 mM KCl, 1 mM MgCl₂, 1 mM CaCl₂, and 10 mM glucose. The changes in fluorescence intensity of Fluo3 were monitored on a LSRII cytometer (BD Biosciences). Equal loading of the samples was controlled by treatment with 100 nM ionomycin (Sigma-Aldrich).

For the calculation of p values, the paired Student’s t test was used. Differences with p < 0.05 were considered to be significant.

We observed human α-defensins HNP-1, HNP-3, and HD-5 to induce concentration-dependent, bell-shaped chemotactic responses in human monocyte-derived macrophages in vitro (Fig. 1 A).

To address the issue of signaling via one or more chemotactic receptors, functional desensitization between human α-defensins was investigated. Preincubation of macrophages with HNP-1 or HD5 is shown to inhibit subsequent migratory responses to HD5 (Fig. 1,B). Although desensitization can be of homologous or heterologous nature, only the latter is PKC-dependent (20, 21). We observed that the general PKC inhibitor GF109 was unable to block desensitization between HNP-1 and HD5 (Fig. 1,B). This finding indicates that α-defensins signal through a common receptor in macrophages resulting in PKC-independent homologous receptor desensitization. Receptors for α-defensins were subject to heterologous desensitization as shown by the ability of the PKC activator PMA to desensitize HD5-induced chemotaxis (Fig. 1,B). Human macrophages have been found to migrate in response to different human β-defensins (22). We observed that PKC was instrumental in the cross-desensitization between α- and β-defensins because PKC blockade by GF109 abrogated the desensitizing effects of HNP-1 and HD5 on HBD-2-induced migration of macrophages (Fig. 1,C). Desensitization by α-defensins did not extend to other chemotactic receptors as shown in the case of C5a (Fig. 1 D) or CCL3 (data not shown).

α-Defensin-mediated migration of human macrophages was further studied using a SCID mouse model. We have recently shown that accumulation of human macrophages in the peritoneal cavity of SCID mice as a result of chemotactic cell migration was inducible by injection of anaphylatoxins and chemokines (17). We now observed that human macrophages could be mobilized into the peritoneal cavity of SCID mice by i.p. injections of α-defensins (Fig. 1, E and F). Similar to in vitro chemotaxis experiments, in vivo migration of macrophages was abrogated if α-defensins were used for both preincubation and chemoattraction. Cross-desensitization between both defensin families was also observed in vivo (Fig. 1, E–G). No inhibitory effect by α-defensin preincubation was observed when C5a was used as a chemotaxin in vivo (Fig. 1,H). C5a was also unable to desensitize HNP-1-induced chemotaxis (Fig. 1 E). These experiments confirm the in vivo relevance of α-defensin-induced chemoattraction and desensitization.

IL-4 which is commonly used for the generation of DC in vitro was recently observed to inhibit migration of monocyte-derived macrophages and DC in response to anaphylatoxins (17). In line with these findings, IL-4 treatment also inhibited migration of human monocyte-derived DC in response to HNP-1 in vitro and in vivo (Fig. 2,A). Thus, macrophages but not DC appear to constitute primary target cells for α-defensins. To find further support for this hypothesis, we investigated α-defensin-induced migration of murine macrophages and DC which had been generated from bone marrow precursors in the presence of M-CSF and GM-CSF, respectively (Fig. 2,B). It was found that HNP-1 preferentially mobilized macrophages but not DC in vitro and in vivo. Of note, murine DC retained the ability to migrate as shown by their potent chemotactic response to CCL20 (Fig. 2,B). In support of its role as a chemoattractant for macrophages, HNP-1 recruited the murine macrophage cell line J774A.1 (Fig. 2 C).

HNP-1 and HD5 were observed to be potent chemotaxins for both naive as well as memory T lymphocytes (Fig. 3, A and B). Of note, we also identified HNP-3 to recruit memory T cells (Fig. 3 B).

In analogy to human macrophages, the general PKC inhibitor GF109 was unable to block functional desensitization between human α-defensins (Fig. 3,C). This finding suggests that HNP-1, HNP-3, and HD5 signal through a common receptor also in T lymphocytes, resulting in PKC-independent homologous receptor desensitization. The α-defensin receptor on T cells was subject to heterologous desensitization following PKC stimulation by PMA (Fig. 3 C).

The cell line HMC-1 is widely used as a model for immature human mast cells. We observed human α-defensins HNP-1, HNP-3, and HD-5 to induce concentration-dependent, bell-shaped chemotactic responses in HMC-1 cells (Fig. 4 A).

In analogy to human macrophages and T cells, the general PKC inhibitor GF109 was unable to block the functional desensitization between HNP-1 and HD5 (Fig. 4,B). Thus, human α-defensins signal through a common receptor also in mast cells. Mast cells have been shown to migrate in response to HBD-2 (23). We observed that the α-defensin receptor on HMC-1 mast cells was subject to heterologous desensitization by the PKC activator PMA and to PKC-dependent cross-desensitization by HBD-2 (Fig. 4,B). Conversely, α-defensins desensitized HBD-2-mediated migration of mast cells in a PKC-dependent manner, suggesting unique receptors for both defensin families (Fig. 4,C). Usage of the PKC inhibitor Gö6976 allowed us to investigate the involvement of calcium-dependent PKC isotypes (Fig. 4, D and E). Gö6976 abrogated the desensitizing impact of HBD-2 on HNP-1-induced migration but, vice versa, failed to inhibit the desensitizing influence of HNP-1 on HBD-2. Thus, HBD-2 but not HNP-1 or HD5 signal through calcium-dependent PKC isotypes, a finding which is consistent with the inability of HNP-1 or HD5 to induce calcium fluxes in HMC-1 in contrast to HBD-2 (Fig. 4, F and G).

To determine the role of MAPK in α-defensin-induced signaling pathways, we used a pharmacological approach. Human macrophages (Fig. 5 A) were pretreated with PD098059, SP600125, or SB203580, which inhibit MAPK ERK, JNK, and p38 phosphorylation, respectively. We found that predominantly ERK and p38 contributed to α-defensin-induced chemotaxis.

Furthermore, pretreatment of macrophages (Fig. 5,B) and HMC-1 mast cells (Fig. 5 C) with pertussis toxin for 2 h inhibited HNP-1-induced chemotaxis in a dose-dependent manner. The 50% inhibitory concentration of pertussis toxin was ∼50 ng/ml in macrophages and 12.5 ng/ml in HMC-1 cells. These data indicate that Gα_i proteins are instrumental in α-defensin-induced signal transduction.

Defensins function as important effectors of innate immunity. Their main function appears to be the killing or inactivation of microorganisms (24). Recent evidence suggests that defensins may be involved in adaptive immunity by attracting immunocompetent cells to the sites of inflammation and infection (22).

Investigating chemotaxis in vitro and in vivo, we found that HNP-1, HNP-3, and HD5 are chemoattractive for human monocyte-derived macrophages. HNP-1 and -2 have been shown in the past to recruit monocytes (14) or to be devoid of this activity (13, 15). HNP-3, in contrast, was described as chemotactically inactive, a finding which was surprising because HNP-1 to -3 differ at only the N-terminal amino acid (25). The intestinal α-defensin HD5 shows only moderate sequence similarity to neutrophil α-defensins (26) and was also observed to lack chemotactic activity (4). According to our data, however, recruitment of human macrophages is a function common to neutrophil as well as intestinal α-defensins including HNP-1, -3, and HD5. α-Defensin-induced migration was also detected using the murine macrophage cell line J774A.1 or murine macrophages differentiated from bone marrow precursors in the presence of M-CSF. Thus, mobilization of macrophages represents a common, evolutionarily conserved property of the human α-defensin family. The discrepancies between our data and those shown by others could originate from the different defensin preparations applied. We obtained synthetic HNP-3 from a commercial source whereas HNP-3 found to be chemotactically inactive was purified from neutrophils (14, 15). Other possible reasons may be the different methods for investigating in vitro chemotaxis (Transwell system vs microchemotaxis chambers). Furthermore, we used cultured macrophages which may express higher numbers of defensin receptors than freshly isolated monocytes.

DC have been demonstrated to migrate in response to gradients of HNP-1 via an unidentified receptor (13). According to our results, however, human and murine DC generated from monocytes and bone marrow precursor cells, respectively, were only marginally mobilized by α-defensins in vitro and no migration was observed using an in vivo migration assay. Therefore, we conclude that macrophages but not DC represent primary target cells for α-defensins. In keeping with this reasoning, IL-4 which is commonly used for the generation of human monocyte-derived DC was identified as a suppressor of α-defensin-induced chemotaxis. This finding is reminiscent of the inhibitory effect of IL-4 on anaphylatoxin-induced DC migration as a consequence of receptor down-regulation (17).

Chertov et al. (15) identified HNP-1 and HNP-2 as chemoattractants for CD3⁺ T cells whereas HNP-3 was inactive. Later, Yang et al. (13) reported that naive but not memory T cells migrated in response to HNP-1. In contrast, we found neutrophil HNP-1 and -3 to be comparably chemotactic for naive and memory T lymphocytes. Furthermore, our data prove that intestinal HD5 has the capacity to recruit naive and memory T cells, a finding which contrasts with a previous report on the chemotactic inactivity of intestinal α-defensins (4). It is tempting to speculate that the protective role of HD5 in host defense, specifically in the pathogenesis of ileal Crohn′s disease (7, 8), may in part result from its chemotactic effects on immune cells, including T lymphocytes and macrophages which may supplement or even exceed its antimicrobial activity. Indeed, the antimicrobial effects of many host defense peptides are sensitive to physiological salt concentrations in vitro which may favor the view that immunomodulation is the predominant function of defensins in vivo (27).

Rat peritoneal mast cells respond to human neutrophil α-defensins with histamine secretion (28). The underlying mechanism of action was found to be receptor-independent similar to other polycationic secretagogues such as substance P or polylysine. In contrast, the cell line HMC-1 which is widely used as a surrogate for human mast cells responded with chemotaxis to gradients of HNP-1, -3, and HD5 as shown in the present study. The chemotactic impact of α-defensins on mast cells is obviously receptor-mediated because chemoattraction was subject to homologous as well as heterologous desensitization.

Receptor desensitization is a regulatory mechanism to prevent the damaging effects of prolonged or excessive activation of G protein-coupled receptors (21). Heterologous or cross-desensitization involving different G protein-coupled receptors is commonly mediated by second messenger-regulated kinases such as PKC and does not require agonist occupancy (20). In contrast, homologous desensitization affects receptors in the agonist-occupied state and involves phosphorylation by G protein-coupled receptor kinases (21). We found that the functional desensitization between α-defensins occurred without participation of PKC, indicating that a single receptor is engaged. The receptor for α-defensins was also subject to heterologous desensitization because the PKC activator PMA abrogated defensin-induced chemotaxis.

Intriguingly, the ability of α-defensins to desensitize the chemotactic activity of β-defensins and vice versa was also PKC-dependent, indicating cross-desensitization to occur between distinct defensin receptors. In support of our findings, HBD-2 has been shown to signal through as yet unidentified G protein-coupled receptors on various immune cell subtypes such as macrophages and mast cells (22, 23). In HMC-1 mast cells, we made the observation that calcium-dependent PKC isotypes were involved in the HBD-2-mediated cross-desensitization of the α-defensin receptor which was consistent with the capability of HBD-2 to induce cellular calcium fluxes. Conversely, HNP-1 and HD5 did not engage calcium-dependent PKC isotypes when cross-desensitizing the HBD-2 receptor which was in line with their inability to induce calcium fluxes. It is noteworthy that calcium-independent PKC isotypes have been demonstrated to mediate cross-desensitization of chemokine receptors by opioid receptors (29). In contrast to their impact on defensin-induced cross-desensitization, PKC inhibitors were unable to block migration in response to α-defensins and HBD-2. These results are in line with published evidence that the monocyte chemotactic response to chemoattractants such as fMLP and chemokines is independent of PKC activation (30).

α-Defensin peptide sequences exhibit a large degree of sequence variability and divergent antimicrobial properties (1, 6). The molecular basis for the evolutionarily conserved binding of neutrophil as well as intestinal α-defensins to a common chemotactic receptor on immune cells may reside in the similarity of their tertiary structures which are stabilized through three conserved disulfide bridges (3, 4).

We found α-defensin receptors to be coupled to Gα_i proteins due to their sensitivity to pertussis toxin, which confirms a previously published observation (13). G protein-coupled receptors are known to be connected to the MAPK signaling pathways by classical G protein-stimulated synthesis of second messengers (31). Mammalian cells contain three major classes of MAPK: ERK, JNK, and p38. All three kinases have been shown to play crucial roles in cell migration by distinct mechanisms (32). Our results confirm the importance of MAPK also for α-defensin-induced chemotaxis with p38 and ERK being predominantly required for HNP-1-induced migration of macrophages. In line with these data, p38 and ERK have been observed to participate in the degranulation of mast cells by HBD-3 and -4 (33).

In summary, human α-defensins are characterized by evolutionarily conserved receptor usage in macrophages, mast cells, and lymphocytes. The existence of a common receptor for α-defensins strengthens the view that immunomodulation may represent a principal function of the defensin system.

We thank Olga Walter and Kerstin Eckelmann for excellent technical assistance.

The authors have no financial conflict of interest.