Eosinophils circulate in the blood and are recruited in tissues during allergic inflammation. Gap junctions mediate direct communication between adjacent cells and may represent a new way of communication between immune cells distinct from communication through cytokines and chemokines. We characterized the expression of connexin (Cx)43 by eosinophils isolated from atopic individuals using RT-PCR, Western blotting, and confocal microscopy and studied the biological functions of gap junctions on eosinophils. The formation of functional gap junctions was evaluated measuring dye transfer using flow cytometry. The role of gap junctions on eosinophil transendothelial migration was studied using the inhibitor 18-a-glycyrrhetinic acid. Peripheral blood eosinophils express Cx43 mRNA and protein. Cx43 is localized not only in the cytoplasm but also on the plasma membrane. The membrane impermeable dye BCECF transferred from eosinophils to epithelial or endothelial cells following coculture in a dose and time dependent fashion. The gap junction inhibitors 18-a-glycyrrhetinic acid and octanol did not have a significant effect on dye transfer but reduced dye exit from eosinophils. The gap junction inhibitor 18-a-glycyrrhetinic acid inhibited eosinophil transendothelial migration in a dose dependent manner. Thus, eosinophils from atopic individuals express Cx43 constitutively and Cx43 may play an important role in eosinophil transendothelial migration and function in sites of inflammation.
Eosinophils circulate in the blood and are recruited to peripheral tissues during allergic inflammation [
Cells of the immune system have been shown to communicate through gap junctions [
In peripheral tissues, eosinophils come in close contact to various cell types such as endothelial and epithelial cells. Endothelial and epithelial cells express connexins and form functional gap junctions with each other. Formation of gap junctions between eosinophils and tissue resident cells may provide another mechanism of interaction during inflammatory reactions. In this way activation of eosinophils through eosinophil specific stimuli could be transferred directly and specifically to immediately neighboring cells through gap junctions. This mechanism could lead to limited and specific activation of communicating cells compared to the activation obtained through the paracrine effects of released cytokines and chemokines. Gap-junction mediated interactions between immune cells and endothelial cells have been shown before [
In this paper we characterized the expression of Cx43 by eosinophils and showed the formation of gap junctions between eosinophils and epithelial or endothelial cells. We also present evidence that gap junctions may play a role in eosinophil transendothelial migration.
Mouse monoclonal anti-Cx43 antibody (Zymed, South San Francisco, CA, USA); BCECF and BODIPY-conjugated goat anti-mouse IgG antibody (Molecular Probes, Eugene, OR, USA); DMEM, RPMI-1640, FBS, penicillin and streptomycin and L-glutamine (Biowhittaker, Walkersville, MD, USA); rabbit polyclonal anti-mouse HRP-conjugated antibody (Amersham Biosciences Corp., Baie D’Urfe, QC, Canada); nonfat dry milk and polyvinylidene difluoride (PVDF) membrane (Bio-Rad Laboratories Ltd., Mississauga, ON, Canada); MMLV reverse transcriptase, Taq polymerase, and oligo-dT primers (Gibco, Burlington, ON, Canada); anti-CD16 coated magnetic beads (Miltenyi Biotech Inc., Auburn, CA, USA); Costar clear Transwells for 12-well plates with 3 mm pores (Fisher Biosciences, Edmonton, AB, Canada); fibronectin coated Biocoat 12-well inserts with 3 mm pores (Becton Dickinson, Mississauga, ON, Canada); and octanol and 18-a-glycyrrhetinic acid and all other chemicals used (Sigma-Aldrich Canada Ltd., Oakville, ON, Canada) were purchased as shown.
Eosinophils were isolated from peripheral blood of atopic individuals as previously described [
RNA was isolated using Trizol. Two
Cells were lysed in borate buffered saline with 1% Triton-X 100, 1 mM PMSF, 10 mg/mL aprotinin, 4 mg/mL leupeptin, and 10 mg/mL pepstatin, and protein concentration was measured with a colorimetric assay. Forty
Cytospins of eosinophils were fixed in 2% paraformaldehyde in PBS for 10 min. Slides were washed five times in Tris-buffered saline, followed by incubation in a 2% human IgG blocking solution for 1 h. After a second wash step, mouse anti-Cx43 monoclonal antibody (20
The human type II pneumocytes-like epithelial cell line A549 was acquired from the American Type Culture Collection (Rockville, MD, USA) and primary cultures of human microvascular endothelial cells isolated from the lung (HMVEC-L) were purchased from Clonetics (Walkersville, MD, USA). A549 cells were cultured in DMEM supplemented with 10% FBS, 4 mmol/L L-glutamine, and 100 mg/mL penicillin/streptomycin. HMVEC-L cells were grown in EBM-2 medium supplemented with human FGF, VEGF, IGF-1 and EGF, ascorbic acid, hydrocortisone, and 5% FBS as per distributor’s instructions. These cells were used for up to four passages while they retained endothelial cell characteristics.
Cells were washed in serum free medium and resuspended at
To study the formation of functional gap junctions dye transfer experiments were performed as described, with minor modifications [
Following, coculture cells were detached with trypsin and analyzed on a FACScan flow cytometer (Becton Dickinson). Eosinophils were differentiated from epithelial or endothelial cells according to their forward and side scatter characteristics. To determine dye transfer, the fluorescent intensity of unlabeled cells following coculture with labeled cells was compared to the fluorescence intensity of cells incubated with the same number of unlabeled cells. Results are shown either as “% positive cells” or as “increase of relative fluorescence” between the two conditions.
In particular for Figure
We then calculated the MFI of the labeled eosinophils or HMVEC-L cells in the same experiments in the presence of 18-a-glycyrrhetinic acid or diluent. We then calculated the “% change” in MFI between labeled cells cocultured in the presence of 18-a-glycyrrhetinic acid and labeled cells cocultured in the presence of diluent and this is the number shown in Figure
HMVEC-L cells were grown on 12-well clear or fibronectin-coated inserts (3 mm pores) until confluent. Eosinophils or neutrophils (
Student’s
Peripheral blood eosinophils expressed Cx43 mRNA (Figure
Cx43 expression by peripheral blood eosinophils: (a) RT-PCR for Cx43 expression in peripheral blood eosinophils from atopic donors. (b) Western blot analysis of Cx43 expression by freshly isolated eosinophils and eosinophils cultured for 24 h in the presence or absence of IL-5 (10 ng/mL).
To localize Cx43 protein in eosinophils we performed CLSM. All freshly isolated peripheral blood eosinophils expressed Cx43 (Figure
CLSM for Cx43 expression on freshly isolated human peripheral blood eosinophils. Staining with isotype control antibody (top panels) and mouse anti-Cx43 antibody (lower panels) is shown. Left panels show antibody staining alone and middle panels Cx43 staining along with DAPI to visualize the nuclei. The right panels show transmitted light images from the same slides.
To study the ability of connexins to form functional gap junctions between eosinophils and other cell types we cocultured eosinophils with A549 human airway epithelial cells or HMVEC-L cells for 1–3 h. In different experiments eosinophils or epithelial/endothelial cells were loaded with the gap junction-permeable dye BCECF. Cells were cultured at a ratio of unlabeled to labeled cells of 1 : 2.
Dye transfer was observed between A549 and eosinophils in both directions (Figure
Dye transfer between eosinophils and epithelial/endothelial cells. Eosinophils were cocultured with A549 airway epithelial cells or HMVEC-L for 3 h. In each case one of the cell types was labeled and the other unlabeled. Unlabeled cells (eosinophils in (a) and (c), A549 cells in (b) and HMVEC-L in (d)) were gated and analyzed for evidence of dye transfer from the labeled cells. A representative experiment (from 4 to 6 experiments) is shown for each condition. Numbers in the graph indicate % of positive cells. Conditions: (a) transfer from labeled A549 cells to unlabeled eosinophils, (b) transfer from labeled eosinophils cells to unlabeled A549 cells, (c) transfer from labeled HMVEC-L cells to unlabeled eosinophils, and (d) transfer from labeled eosinophils to unlabeled HMVEC-L cells.
Effect of Cx inhibitors on dye transfer and transendothelial migration. (a) Unlabeled HMVEC-L cells were cocultured with labeled eosinophils (Eosin) or HMVEC-L (HMVEC-L). The graph shows % decrease in mean fluorescent intensity of unlabeled HMVEC-L in the presence of various concentrations of 18-a-glycyrrhetinic acid compared to mean fluorescent intensity in the absence of 18-a-glycyrrhetinic acid (
To study whether this dye transfer was the result of gap junction-mediated communication between eosinophils and endothelial cells we used the gap junction inhibitor 18-a-glycyrrhetinic acid. We compared the ability of 18-a-glycyrrhetinic acid to affect dye transfer between eosinophils and endothelial cells with its effect on dye transfer between labeled and unlabeled endothelial cells. In both cases there was significant transfer of dye between labeled and unlabeled cells. Unlabeled HMVEC-L increased their fluorescence by 30–50-fold when cocultured with labeled HMVEC-L and only 3–10-fold when cocultured with labeled eosinophils. Preincubation of the cells for 15 min with 18-a-glycyrrhetinic acid decreased the amount of the dye transferred from labeled HMVEC-L to unlabeled HMVEC-L but had no effect on the transfer from labeled eosinophils to unlabeled HMVEC-L (Figure
There is evidence that gap junctions are important in regulating the transmigration of malignant cells through endothelial monolayers [
Similar experiments were repeated using transwells coated with FN to grow the endothelial cells. In this case again 18-a-glycyrrhetinic acid inhibited the transmigration of eosinophils (Figure
We have shown that human peripheral blood eosinophils from atopic individuals express Cx43 but not Cx32. Cx43 was expressed on the cell membrane and also diffusely in the cytoplasm of freshly isolated peripheral blood eosinophils. Dye transfer experiments showed evidence of functional gap junction formation between eosinophils and epithelial or endothelial cells. Inhibitors of gap junction function showed a role for gap junctions in transmigration of eosinophils and neutrophils through endothelial monolayers.
Connexins are a family of more than 20 proteins [
Gap junction formation has been observed between immature eosinophils in the bone marrow [
The two inhibitors of gap junctions we used did not show clear evidence of inhibition of dye transfer between eosinophils and epithelial/endothelial cells. However, this is to our knowledge the main pathway that could mediate direct transfer of a membrane impermeable dye between eosinophils and epithelial/endothelial cells. Different connexin molecules can form gap junctions with varying conductance properties and differential sensitivity to gap junction inhibitors [
Connexins, in the form of open hemichannels, are also functional between the intracellular and extracellular environments [
It is important to define the cell types that communicate with eosinophils through gap junctions. Endothelial cells could be candidate partners for gap junction formation with eosinophils and also with other immune cells. Immune cell transmigration through the endothelium is probably a more complicated process than currently understood. Gap junctions between transmigrating immune cells and endothelial cells could facilitate transfer of important stimuli that initiate events leading to immune cell transmigration.
Our data suggest gap junction involvement in eosinophil transendothelial migration. Recent observations have also implicated gap junction functions in neutrophil transendothelial migration [
In conclusion we have presented biochemical and functional evidence that eosinophils express connexins, which may facilitate transfer of small molecules to epithelial and endothelial cells through gap junctions. The physiological relevance of gap junctions between eosinophils and epithelial or endothelial cells remains to be investigated. Our evidence suggests that gap junctions may play a role in eosinophil transendothelial migration. These heterologous gap junctions between immune and nonimmune cells might represent an underrecognized pathway for communication between eosinophils and cells in the local microenvironment in sites of inflammation.
Eosinophils express Cx43 that is functional and allows communication of eosinophils with other cells. This communication may be very important for eosinophil transmigration through the endothelial monolayer and tissue infiltration.
18-a-Glycyrrhetinic acid
2′,7′-Bis(2-carboxyethyl)-5(6)-carboxyfluorescein
Connexin
Connexin 43
Connexin 32
Confocal laser scanning microscopy
Human microvascular endothelial cells from lung
T regulatory cells.
The authors declare that there is no conflict of interests regarding the publication of this paper.
This study was supported by the Canadian Institutes for Health Research (CIHR) and Alberta Heritage Foundation of Medical Research (AHFMR).