SEARCH

SEARCH BY CITATION

Keywords:

  • allergic inflammation;
  • cell–cell contact;
  • eosinophils;
  • immunologic synapse;
  • mast cells

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Author contributions
  8. Disclosure
  9. References
  10. Supporting Information

To cite this article: Elishmereni M, Alenius HT, Bradding P, Mizrahi S, Shikotra A, Minai-Fleminger Y, Mankuta D, Eliashar R, Zabucchi G, Levi-Schaffer F. Physical interactions between mast cells and eosinophils: a novel mechanism enhancing eosinophil survival in vitro. Allergy 2011; 66: 376–385.

Abstract

Background:  Mast cells (MCs) and eosinophils (Eos) are the key effector cells of the allergic reaction. Although classically associated with different stages of the response, the cells co-exist in the inflamed tissue in the late and chronic phases in high numbers and are likely to cross-talk. While some mediators of MCs are known to affect Eos biology and vice versa, paracrine and physical interplay between the two cells has not been described yet. We aimed to investigate whether intercellular MC–Eos communication could take place in the allergic response and exert functional bidirectional changes on the cells.

Methods:  Tissue sections from various allergic disorders were specifically stained for both cells. Human cord blood-derived MCs and peripheral blood Eos, co-cultured under different conditions, were studied by advanced microscopy and flow cytometry.

Results:  Several co-localized MC–Eos pairs were detected in human nasal polyps and asthmatic bronchi, as well in mouse atopic dermatitis. In vitro, MCs and Eos formed stable conjugates at high rates, with clear membrane contact. In the presence of MCs, Eos were significantly more viable under several co-culture conditions and at both IgE-activated and steroid-inhibited settings. MC regulation of Eos survival required communication through soluble mediators but was even more dependent on physical cell–cell contact.

Conclusions:  Our findings provide the first evidence for a complex network of paracrine and membrane interactions between MCs and Eos. The prosurvival phenotype induced by this MC–Eos interplay may be critical for sustaining chronic allergic inflammation.

Allergic inflammation (AI) is a multiphase process involved in several pathological disorders such as asthma and atopic dermatitis (AD). The early phase of this response is typically initiated upon allergen-mediated activation of mast cells (MCs) (1, 2). Later recruitment of circulating inflammatory cells, and particularly of eosinophils (Eos), results in a late-phase response that in many respects mimics the inflammatory features present in chronic allergic diseases such as atopic eczema, rhinitis, and asthma (1, 3).

The role of MCs in AI has been classically confined to its early/acute phase, whereas the late/chronic phase of inflammation is usually attributed to the effects of infiltrating Eos and leukocytes. Yet in reality, both MCs and Eos co-exist in high numbers in the inflamed tissue during the late/chronic stages of AI and are highly likely to interact with each other by numerous and complex pathways (4, 5). It is probable that bidirectional interactions between MCs and Eos serve to stimulate each other’s functions, thereby perpetuating AI (4, 5). Basic proteins and lipid mediators of Eos have been found to activate MCs, and MC mediators and metabolites were shown as Eos stimulators (6–8). In addition, growth factors of one cell can be a product of the other: Eos may produce stem cell factor (SCF) and nerve growth factor (NGF), both of which up-regulate MC survival, differentiation, and function (5, 9). Reciprocally, granulocyte macrophage colony-stimulating factor (GM-CSF), interleukin (IL)-5 and IL-3, and tumor necrosis factor α (TNF-α), which can be released by MCs, are survival cytokines for Eos (10).

Beyond the possible interactions mediated by soluble factors, physical interfaces could also occur between the many MCs and Eos present in AI. Such direct contact between MCs and Eos may result in the exchange of critical regulatory information and transfer of signals, which further augment functional attributes of these effector cells. Indeed, cell–cell communication is an established cornerstone of regulation of cell development, survival, and activation. Intercellular binding has been widely documented in the context of the immunologic synapse, processes of chemotaxis, and cell adhesion (11, 12). Similarly, MCs and Eos might interact via surface molecules expressed on both cells (5), or in a less specific manner through membrane nanotube contact or exosomal transfer (13, 14). Still, evidence for direct MC–Eos contact in allergy is lacking, and its potential effects on the functions of either cell in late/chronic AI are yet to be deciphered.

In this study, we have investigated MC–Eos interactions in several in vitro and AI conditions. Our results provide the first evidence for a functional MC–Eos effector unit, which depends on both paracrine cross-talk and cell–cell physical contact. This ‘vicious cycle’ between MCs and Eos could significantly contribute to the persistence of AI and thus represents a novel therapeutic target in these diseases.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Author contributions
  8. Disclosure
  9. References
  10. Supporting Information

Tissue samples

Human nasal polyp (NP) and bronchial biopsies, and murine AD skin samples, were obtained and processed as detailed in the Supporting Information file. Analysis of tissues by histological staining, immunohistochemistry, and immunofluorescence is described in the Supporting Information.

Cells and co-cultures

MCs and Eos were obtained from cord blood (CB) and purified from peripheral blood (PB), respectively (see Supporting Information). In conjugate visualization and imaging experiments, MCs and Eos were co-cultured for up to 1 h at various conditions (see Supporting Information). Selected samples were pretreated 20 min on ice with mAbs against CD48 (clone MEM108, 1 μg/ml; Biolegend, San Diego, CA, USA) and/or 2B4 (clone C1.7, 1 μg/ml; eBioscience, San Diego, CA, USA), or isotype-matched Abs (IgG1; Dakocytomation, Glostrup, Denmark), and washed twice prior to co-culturing. For adequate comparison, cell amounts and volume in control MC/Eos monocultures were set exactly as in co-culture. Detailed sample preparation and evaluation by advanced microscopy and multispectral imaging flow cytometry (MIFC) is provided in the Supporting Information.

For cell survival experiments, MCs and Eos were co-cultured for different times (1 h to 7 days) and ratios (1 : 1–10), in media with SCF (100 ng/ml) and/or GM-CSF (20 ng/ml). Partial co-cultures were conducted using 0.4-μm transwell inserts (Greiner Bio-One, Frickenhausen, Germany). For activation, MCs presensitized for 3 days with human myeloma IgE (1 μg/ml; Calbiochem-Merck, Schwalbach, Germany) were washed extensively and stimulated by anti-IgE Ab (5 μg/ml; Dakocytomation) at co-culture initiation. In other samples, dexamethasone (1 μM; Holland Moran, Yehud, Israel) was added at co-culture initiation. In some experiments, the following reagents were added 20 min prior to co-culture: GM-CSF, TNFα, IL-5-, and IL-3-blocking Abs (1 μg/ml; R&D Systems, Minneapolis, MN, USA); mouse mAbs against CD48 (MEM108, 1 μg/ml), 2B4 (C1.7, 1 μg/ml), intercellular adhesion molecule (ICAM)-1 (1 μg/ml; Biolegend), and Nectin-2 (mouse hybridoma, a gift of Dr A. Moretta, Genoa, Italy; a 1 : 10 dilution); 2B4 Fab′ fragments (prepared as described in the Supporting Information; 1 μg/ml); recombinant human CD48 (rhCD48, 0.1 μg/ml; R&D systems). Control samples included matched isotype Abs (mouse IgG1 and IgG2a, 1 μg/ml; Dakocytomation and R&D Systems) or Fab′ fragments (see Supporting Information). In all experiments, control monocultures were set exactly as in co-culture. Viability was assessed by propidium iodide (PI) exclusion (see Supporting Information).

Additional methods are detailed in the Supporting Information file.

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Author contributions
  8. Disclosure
  9. References
  10. Supporting Information

MC–Eos contact in AI

We first investigated whether MC–Eos pairs could be found in inflamed tissues. Congo-red and Toluidine-blue (CTB)-stained NP tissues of patients with allergic rhinitis revealed several MCs and Eos in close proximity (Fig. 1A). MCs appeared partially degranulated at times, with released granules near Eos (Fig. 1A; images 2–3). Sections of asthmatic bronchi showed similar co-localization of MCs and Eos, identified, respectively, by immunohistochemical staining of tryptase and major basic protein (MBP) (Fig. 1B). This was independent of asthma severity, as neighboring MCs and Eos were found in severe (Fig. 1B; image 1), moderate (Fig. 1B; image 2–3), and mild (Fig. 1B; images 4–6) asthmatic bronchi. We also examined CTB-stained skin of mice with chronic AD, induced by epicutaneous sensitization via ovalbumin (OVA) with or without staphylococcal enterotoxin B (SEB) for up to three sensitization weeks (15). In OVA-sensitized mice, MC–Eos contacts were abundant especially after the full 3-week sensitization period (Fig. S1A). This was confirmed by immunolabeling of OVA/SEB skin sections: MCs, evident by their granules stained by the cationic compound avidin-sulforhodamine (AS), were near Eos, stained for MBP (Fig. S1B). In OVA-sensitized mice, pairing was apparent already from the 2-week sensitization stage, coinciding with MC/Eos accumulation in the inflamed tissue (Fig. S1C). Maximal couple counts (2.5 ± 1.5 pairs per high-power field) were observed at the most severe 3-week stage, with 15.8% of MCs and 12.8% of Eos in contact.

image

Figure 1.  Mast cell–eosinophil (MC–Eos) contact in human allergic inflammation disorders. (A) Nasal polyp tissue sections stained by Congo-red and Toluidine-blue for co-localized MCs and Eos (×100 magnification). (B) Bronchi sections from a severe asthmatic patient (1), moderate asthmatic patient (2, 3), and mild asthmatic patient (4–6), dual-stained by anti-tryptase and anti-major basic protein Abs for MCs and Eos, respectively (×40). Bold arrows-MCs; Thin arrows-Eos.

Download figure to PowerPoint

Physical interactions between human MC–Eos in vitro

To examine the binding kinetics, human CB-derived MCs were co-cultured with PB-Eos (1 : 1) for 1 h. The cells interacted to form a well-defined interface within 5 min, and contacts lasted for ∼3–4 min (Fig. 2A), in contrast to the transient (<10 s) random homotypic binding in monocultured MCs or Eos. The stability of MC–Eos pairs was confirmed by measuring membrane distances (Fig. 2B). Pseudopodia-like membrane structures of MCs appeared often approaching the Eos (Fig. 2C). Tight interactions were seen also in transmission electron microscopy (TEM) analysis (Fig. 2D), as further characterized in our recent ultrastructural investigation of MC–Eos interactions (16). To quantify conjugation rates, co-cultures containing carboxyfluorescein diacetate succinimidyl ester (CFSE)-labeled MCs and unlabeled Eos were analyzed by MIFC. Monocultures contained homotypic MC–MC or Eos–Eos pairs, yet in co-culture the formation of heterotypic MC–Eos couples was clearly preferred over such conjugates (Fig. 2E,F). Binding rates were highest in 5- and 15-min co-cultures but still present following 1 h (Fig. 2F). Notably, complex multicellular MC–Eos aggregates were detected especially in early co-culture phases (Fig. S2). These consisted mostly of single MCs surrounded by several Eos (Fig. S2A).

image

Figure 2.  Physical interactions between mast cells (MCs) and eosinophils (Eos) in vitro. (A) CB-MCs (bold arrows) and peripheral blood Eos (thin arrows) interact in 1-h co-cultures (time-lapse photomicrographs, ×40). 00 : 00-arbitrary time within the 1-h period. (B) Binding dynamics given by quantifying MC–Eos membrane distances (mean ± SEM, several pairs analyzed in n = 3 experiments). MC pseudopodia-like membrane structures approaching Eos (C), transmission electron microscopy-observed couples (D), and multispectral imaging flow cytometry-analyzed Eos interactions with carboxyfluorescein diacetate succinimidyl ester-labeled MCs (E). (F) Homotypic/heterotypic conjugation rates in co-cultures vs. monocultures (representative example of n = 3 experiments).

Download figure to PowerPoint

Influence of MC–Eos interactions on cell viability

As MCs can produce Eos growth factors and vice versa (6–10), we examined whether cell survival would be influenced by the communication between the cells. MCs and Eos from several donors (n = 12) were mixed (1 : 1) with or without their respective growth factors SCF and GM-CSF and incubated for up to 1 week. Following a 72-h co-culture with MCs (with/without SCF), Eos showed significantly higher survival rates (a ∼2-fold increase), comparable to GM-CSF-monocultured Eos (Fig. 3A). Interestingly, the MC-induced survival effect on Eos was strongest in the presence of SCF. In SCF-containing media, the MC prosurvival signal was evident as soon as 48 h following the start of co-culture, lasting for at least 7 days (Fig. 3B). High Eos viability levels were observed also under various MC:Eos ratios (between 1 : 1 and 1 : 10), with maximal rates in the 1 : 1 and 1 : 5 ratios (Fig. 3C). In contrast, in GM-CSF media, the viability of co-cultured Eos was lower than in monoculture. Notably, MC viability rates were only slightly, but not significantly, higher following co-culture in all conditions (Fig. 3A). However, in co-cultures with the high Eos numbers (1 : 10), MC were slightly less viable, suggesting a possibly toxic effect of Eos on MCs in such settings (Fig. 3C).

image

Figure 3.  Cell survival is modulated in mast cell–eosinophils (MC–Eos) co-cultures. (A) Viability of CB-MCs and peripheral blood Eos, co-cultured (1 : 1) with/without stem cell factor (SCF) and/or granulocyte macrophage colony-stimulating factor, analyzed at 72 h by PI exclusion (n = 12, *P < 0.05, **P < 0.001). (B) Survival levels of co-cultured MCs/Eos in SCF-containing media, for different durations (n = 4, *P < 0.01, **P < 0.001). (C) Viability in SCF co-cultures (72 h) with various MC:Eos ratios (n = 6, *P < 0.05). Values in control monocultures (cells alone) are indicated. Data are mean ± SEM.

Download figure to PowerPoint

MC-induced Eos survival following IgE activation or dexamethasone treatment

In AI, Eos infiltration follows IgE activation and degranulation of MCs (1). To mimic this pathophysiologic scenario, co-cultures were carried out concomitantly with IgE-activated MCs. Activated MCs increased Eos survival after 72 h, yet to a slightly lower extent when compared with co-cultures containing resting MCs (Fig. 4A). We next tested whether MC-induced Eos survival could occur in the presence of dexamethasone, which alters cytokine secretion patterns (10, 17–19). Monocultured Eos exposed to dexamethasone have displayed poor survival rates (20), yet here dexamethasone-treated co-cultured Eos were protected from apoptosis: viability rates (51%) were only slightly lower than those in a untreated MC co-culture (72%) (Fig. 4B).

image

Figure 4.  Mast cell (MC)-mediated augmentation of eosinophils (Eos) survival occurs in IgE activation or dexamethasone treatment. (A) Eos viability in 72-h stem cell factor (SCF) co-cultures with IgE-presensitized MCs (1 : 1), at resting (−) or anti-IgE Ab-activated (+) conditions (n = 3, *P < 0.01). (B) Survival levels of Eos in 72-h SCF co-cultures, at untreated or dexamethasone (1 μM)-treated conditions (n = 3, *P < 0.05, ns – not statistically significant). Viability rates of control monocultures are indicated. All data are mean ± SEM.

Download figure to PowerPoint

Enhanced Eos viability owing to both soluble and physical communication with MCs

Eos growth factors (GM-CSF, TNFα, IL-3, and IL-5) can be produced by MCs (1, 5), suggesting a prominent role of these cytokines in MC-induced Eos survival. Indeed, in SCF-supplemented 24-h co-cultures, the release of GM-CSF was synergistically increased and TNFα was additively increased, when compared with MCs or Eos cultured alone under similar conditions (Fig. 5A). Yet when blocking these soluble communication pathways by cytokine-neutralizing mAbs added to co-culture, the survival effect was still apparent (Fig. 5B); viability rates in GM-CSF/TNFα-blocked co-cultures at 72 h were only slightly lower (∼63% and 67%) and in IL-3/IL-5-blocked co-cultures were unaltered, when compared to the high Eos survival in normal (isotype-treated) co-cultures (73%). We hypothesized that MC–Eos physical contact could compensate for the blocked soluble signaling, inducing Eos survival by direct signaling. To clarify the regulating role of soluble vs. physical interactions in this effect, transwell cell-separating co-cultures enabling only paracrine interactions were studied. In partial co-cultures, Eos viability was elevated to 44%, yet normal (full-contact) co-cultures further augmented survival rates to 68% (Fig. 5C), implicating both communication modes as important. In contrast to the outcome in full co-cultures (Fig. 5B), blockade of GM-CSF (but not of TNFα) in transwell co-cultures abrogated most of the MC-induced Eos survival effect (Fig. 5D), suggesting that GM-CSF is the primary cytokine involved in the soluble MC–Eos interplay that sustains Eos.

image

Figure 5.  Mast cell (MC)-induced eosinophils (Eos) survival requires both soluble and physical interactions. (A) Tumor necrosis factor α/granulocyte macrophage colony-stimulating factor (TNFα/GM-CSF) levels in 24-h MC–Eos co-culture supernatants (n = 3, *P < 0.05). Samples contain equal cell concentrations. (B) Eos viability in 72-h full-contact co-cultures containing various cytokine-blocking Abs (n = 7, *P < 0.01). Eos survival in transwell (tw) vs. full-contact co-cultures, in normal conditions (C, n = 9, *P < 0.01), or under TNFα/GM-CSF-neutralized Abs (D, n = 3, *P < 0.01). Viability in control monocultures and isotype-treated settings are shown. Data are mean ± SEM.

Download figure to PowerPoint

CD48 and 2B4 receptors mediate the MC–Eos physical interface and survival effects

Because MC–Eos direct interactions exist and affect Eos survival (Fig. 5C), we conjectured that this contact could be facilitated in part by binding of the surface molecule CD48 [on both MCs and Eos (21)] to its ligand 2B4 [on Eos (22)]. Indeed, in several MC–Eos pairs after a 15-min co-culture, CD48 is found at the contact site (Fig. 6A), indicating that it could interact with Eos 2B4. To clarify the role of the CD48-2B4 axis in MC–Eos conjugation, co-cultures were held after Ab blockade of CD48 on MCs and/or 2B4 on Eos. MC–Eos coupling was less abundant in 2B4-neutralized or CD48/2B4-neutralized settings (Fig. 6B), when compared with control isotype co-cultures. Neutralization of other MC/Eos receptors (i.e., c-kit) did not reduce couple formation (Fig. 6B), supporting a specific role of CD48 and 2B4 in the contact.

image

Figure 6.  CD48 and 2B4 mediate mast cell–eosinophil (MC–Eos) interactions and increased Eos survival. (A) Conjugates of CB-MC (bold arrows) and peripheral blood Eos (thin arrows) immunostained for CD48 (red) and nuclei (Hoechst, blue). Merged images are displayed, and white arrows mark the interface between the cells. (B) Conjugation following receptor neutralization on MC and/or Eos (mean ± SEM, 20 high-power fields in n = 3 experiments, *P < 0.01). (C) Seventy-two hours viability of Eos in monoculture and co-culture in the presence or absence of whole Abs, Fab′ fragments, or ligands of indicated cell-surface molecules (mean ± SEM, n = 3, *P < 0.01). Values in isotype-treated samples are indicated. (D) Murine atopic dermatitis tissues immunolabeled for CD48 (green), MC granules (red), and Eos major basic protein (blue). Merged images and surface-rendered projections (SRP) are displayed. Arrows mark the MC–Eos interface.

Download figure to PowerPoint

We next tested whether CD48-2B4 signaling could also affect Eos viability. Following Ab ligation of CD48 or 2B4, monocultured Eos (72 h) exhibited significantly higher survival (66% and 69%, respectively) compared to isotype-treated controls (Fig. 6C). Abs for other receptors (i.e., Nectin-2 or ICAM-1) had no such impact on Eos survival. Interestingly, Eos incubated with specific anti-2B4 Fab′ fragments or soluble rhCD48 were also unaffected; this suggests that crosslinking of Eos 2B4 (enabled by whole agonistic bivalent Abs) was critical for the viability effect (Fig. 6C). Notably, in co-cultured Eos, the whole Abs did not contribute to a further rise in the viability rates already enhanced by MCs (88%, Fig. 6C), suggesting that Eos had reached maximal survival rates.

These data led us to hypothesize that the direct MC-induced Eos survival (Fig. 5C) is mediated, at least partially, by CD48-2B4 binding. Evidence for this was provided by co-cultures pretreated with anti-2B4 Fab′s or rhCD48: in these 2B4-blocked settings, the MC-induced effect on Eos viability was reduced (74%, Fig. 6C). The physiologic relevance of CD48 in MC–Eos interactions was also supported in vivo. Inflamed skin sections of OVA/SEB-induced AD mice were stained for MCs/Eos (by AS and anti-MBP Ab, respectively), as well as for CD48 expression. Both cells expressed CD48, particularly located at the MC–Eos interface in several cases (Fig. 6D). These findings collectively propose the CD48-2B4 axis as a mechanism underlying MC–Eos direct contact and induced effects on Eos viability.

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Author contributions
  8. Disclosure
  9. References
  10. Supporting Information

Despite advances in studying AI, the cellular and molecular events that induce and support its late/chronic outcomes remain elusive. It is clear however that MCs and Eos are highly instrumental in these responses (1–3, 5). We have postulated that both physical and paracrine MC–Eos interactions exist and are a functionally essential component of chronic AI. In fact, as the symptoms of allergy are mostly a direct consequence of the mediators of MCs and Eos, it is highly likely that this soluble and/or physical MC–Eos communication is particularly dominant in its effects, when compared to other possible cell–cell interactions in AI. Our work provides the first documentation and characterization of this novel MC–Eos interplay.

The high MC–Eos binding rates that we detected in situ and in vitro suggest that this cell–cell contact may indeed be influential during chronic states. Interestingly, MC–Eos physical contact was found in nonallergic syndromes [i.e., Ascaris infection, gastric carcinoma (5)]. Contact-dependent communication has several advantages over paracrine interplay: it is highly precise, efficiently relaying the signal selectively to the target cell. In light of this, direct and continuous MC–Eos contact may give rise to significant signaling events, especially because similar stable cell–cell interactions in the immunologic synapse are functional (11, 12). Eos-bound MCs in the examined tissues often appeared degranulated, supporting the possibility that direct contact mediates such effects.

Several MC mediators are known as Eos survival factors (6–8), but the physiologic relevance in co-cultures has not been explored. Here, MCs enhanced and sustained Eos survival, particularly in optimal MC growth conditions (i.e., in SCF). This was seen even in low MC/Eos ratios, an important finding as such tissue ratios characterize chronic AI (1). This alteration of the life span of Eos (normally short-lived cells) could strongly perpetuate chronic AI features critically affected by Eos [i.e., inflammation, tissue remodeling, angiogenesis (1, 3)]. The unchanged MC viability rates in co-culture suggest that these cells are more robust and less dependent on late-phase cells (23), in keeping with their upstream role in AI.

MCs elevated Eos survival in both resting and stimulated settings, further supporting the relevance of this effect to all AI stages involving normal, desensitized, or activated MCs. This was apparent even with dexamethasone, which directly inhibits Eos survival [an effect partially reversed by GM-CSF and IL-3 (20)]. Co-cultured Eos were protected from dexamethasone-induced apoptosis, likely due to cytokines such as GM-CSF (which we found to be partially responsible for the MC-mediated effects on Eos). Of clinical relevance, MC-induced Eos survival may explain why anti-IL-5 therapy fails to eliminate Eos from asthmatic airways (1) and might play a significant role in corticosteroid-resistant patients. Targeting this effect may thus provide a novel approach to the treatment of corticosteroid-resistant asthma.

Our findings offer a novel distinction between soluble and membrane interaction pathways. Both communication modes contribute to the MC-mediated Eos survival effect. The soluble route likely involves Eos survival factors GM-CSF and TNFα, as both were elevated in co-culture. However, Eos survival in transwell co-cultures depended solely on GM-CSF, coinciding with the notion that TNFα signaling precedes GM-CSF signaling: MC-secreted TNFα can ignite an autocrine loop of GM-CSF production in Eos (10, 24), implying GM-CSF as the key factor. MCs themselves may also release GM-CSF (25).

Interestingly, however, Eos were more viable in co-cultures enabling direct MC contact. In fact, the failure of GM-CSF-neutralizing Abs to block MC-mediated Eos survival in full-contact co-cultures shows that direct MC–Eos binding is functional in itself and can compensate for insufficient paracrine signaling. Indeed, MCs and Eos can physically engage other cells (i.e., fibroblasts, endothelial cells, smooth muscle cells, lymphocytes) and undergo functional changes (5, 26, 27). In this context, some surface molecules on MCs/Eos are potential candidates for mediating the AEU physical contact; these include stimulatory receptors, i.e., DNAX accessory molecule-1 (DNAM-1) and its ligand Nectin-2, and adhesion molecules, i.e., lymphocyte function-associated antigen (LFA)-1 and its ligand ICAM-1 (5, 28, 29). Our results, however, propose a particular role of the CD2-like receptors CD48 and 2B4 (21, 22) in mediating MC–Eos conjugation and MC-sustained survival of Eos: Specific ligation of both receptors increases Eos viability in monocultures, and the MC-induced rise in Eos viability is partly reduced by blockade of Eos 2B4. Thus, it is likely that co-cultured Eos are more viable due to activation of 2B4 (by CD48 on MCs) and/or of the CD48 receptor (by a yet unidentified MC ligand). Signaling through both molecules is known to increase lymphocyte survival and induce stimulatory/co-stimulatory effector signals in several cells (30, 31). In fact, CD48 is elevated on Eos of allergic patients (21), supporting a correlation between this molecule, increased MC–Eos binding affinities, and AI. Our finding that CD48 is located at the MC–Eos interface, in both human co-cultures and mouse AD, shows that its regulation of MC–Eos interactions is highly feasible. Additional studies should corroborate the importance of the CD48-2B4 axis in the contact between these effector cells.

The MC-mediated rise in Eos survival could serve as a positive feedback effect, sustaining Eos in chronic AI. Yet this effect may not be straightforward; in GM-CSF-containing co-cultures, Eos viability was not higher, but rather lower (Fig. 3). MCs express the GM-CSF receptor (32) and are receptive to the cytokine, although inhibited Eos survival was probably not because of MC consumption of GM-CSF (as ruled out by high levels of the cytokine in co-culture, data not shown). GM-CSF inhibits MC maturation and alters mediator secretion in immature MCs (33). In our cultures, GM-CSF slightly increased the viability of MCs alone, yet it may have negated their ability to sustain Eos in co-culture. Indeed, activation of some MC phenotypes could counteract Eos viability: Histamine from stimulated MCs attenuates IL-5-induced survival of Eos (34), and MC-derived prostaglandin D2 and J2 induce Eos apoptosis (35). The MC influence on Eos survival is likely a net effect of several signals reflecting the cell activation state and cytokine/mediator milieu and could therefore be altered under diverse pathophysiologic settings.

Beyond cell homeostasis and survival phenomena, bidirectional MC–Eos interactions might also modulate effector functions: Eos-derived factors (i.e., MBP, leukotrienes) activate MCs, and MC mediators and metabolites (i.e., tryptase, histamine, arachidonic acid derivatives) stimulate Eos (5). Indeed, Eos-paired MCs in the inflamed tissues analyzed herein appeared mostly activated. These findings warrant deeper investigation of how MC–Eos activation profiles change in co-culture and in vivo.

In conclusion, our work serves as a basis for the novel concept that soluble and physical interplay between MC–Eos can create favorable conditions for persistence of inflammation. Moreover, as therapeutic intervention is still largely unsatisfactory in the context of chronic allergy, this MC–Eos cross-talk may present a basis for the future development of new therapeutics targets in AI.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Author contributions
  8. Disclosure
  9. References
  10. Supporting Information

We thank Dr Ido Bachelet, Alon Nissim Ben-Efraim and Dr Micha Ben-Zimra for their help in experimental setup, Levi-Schaffer laboratory members for discussions, Terhi Savinko and Dr Nanna Fyhrquist for processing mouse samples, Dr Yael Feinstein-Rotkopf for valuable assistance with confocal microscopy, Ariel Roytman for help in MIFC, and Dr Maria Rosa Soranzo and Dr Francesca Vita for TEM analysis. This work was supported by the Aimwell Charitable Trust (London, UK), the Israel Science Foundation (grant 213/05), and the Adolph and Klara Brettler Center for Research in Molecular Pharmacology and Therapeutics at The Hebrew University of Jerusalem.

Author contributions

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Author contributions
  8. Disclosure
  9. References
  10. Supporting Information

M.E. designed the research, performed experiments, analyzed data, and wrote the manuscript. H.T.A. provided mice samples and edited the manuscript, P.B. provided stained human asthma sections and edited the manuscript, S.M. prepared Ab fragments and helped to design the blocking experiments, A.S. stained bronchial sections, Y.M.F. helped to set experiments for TEM, G.Z. performed and analyzed TEM, D.M. provided CB samples, R.E. provided NP samples. F.L.S. designed the research, provided overall supervision, analyzed the data and edited the manuscript.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Author contributions
  8. Disclosure
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Author contributions
  8. Disclosure
  9. References
  10. Supporting Information

Data S1. Methods.

Figure S1. MC –Eos pairs detected in a mouse AD model.

Figure S2. MC and Eos form stable aggregates in short-term co-cultures.

FilenameFormatSizeDescription
ALL_2494_sm_FigureS1.tif10300KSupporting info item
ALL_2494_sm_FigureS2.tif2037KSupporting info item
ALL_2494_sm_DataS1.doc88KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.