Human airway epithelial cell determinants of survival and functional phenotype for primary human mast cells

Abstract

Mast cells (MCs) are found in increased numbers at airway mucosal surfaces in asthmatic patients. Because human airway epithelial cells (HAECs) actively participate in airway inflammatory responses and are in direct contact with MCs in the mucosa, we hypothesized that HAEC-MC interactions may contribute to the differentiation and survival of MCs in the airway mucosa. Here, we show that HAECs express mRNA and protein for soluble and membrane-bound stem cell factor, releasing soluble stem cell factor into the cell culture supernatant at a concentration of 5.9 +/- 0.1 ng per 10(6) HAEC. HAECs were able to support MC survival in coculture in the absence of any exogenous cytokines for at least 4 d. Before the initiation of coculture, MCs were uniformly tryptase and chymase (MC(TC)) double positive, but by 2 d of coculture the majority of MCs expressed tryptase (MC(T)) alone. MCs supported in coculture generated low amounts of cysteinyl-leukotrienes (cys-LT) after FcepsilonRI-dependent activation (0.2 +/- 0.1 ng of cys-LT per 10(6) cells) and required priming with IL-4 and IL-3 during coculture to achieve a quantity of cys-LT generation within the range expected for human lung mucosal MC (26.5 +/- 16 ng of cys-LT per 10(6) cells). In these culture conditions, HAECs were able to direct mucosal MC protease phenotype, but T cell-derived Th2 cytokines were required for the expression of a functional airway MC eicosanoid phenotype. Thus, distinct cell types may direct unique aspects of reactive mucosal MC phenotype in the airways.

Fig. 1.

Expression of SCF mRNA by HAECs. Soluble (572 bp) and membrane-bound (488 bp) SCF expression by HAECs was identified by RT-PCR analysis. SCF cDNA was amplified by 35 cycles of PCR. Lanes 1 and 2, HAEC donor 1; lanes 3 and 4, primary human MC; lanes 5 and 6, HAEC donor 2; lanes 7 and 8, HAEC donor 3. Expected size of β-actin: 508 bp.

Fig. 2.

SCF protein expression by HAECs. Release of soluble SCF protein into HAEC culture media was detected by SCF-ELISA. (A) All data are mean ± SEM (n = 4). Detection of membrane-bound SCF expression by immunohistochemical staining of HAECs by using goat-anti-human SCF Ab (B) or irrelevant IgG (C) demonstrates membrane-bound expression of SCF protein (n = 2). (Magnification: ×200.)

Fig. 3.

MC survival in coculture. (A) MCs were grown in monoculture with recombinant human SCF (100 ng/ml or 4 ng/ml), in coculture with HAECs by seeding MCs directly onto subconfluent HAECs (direct coculture), in coculture with HAECs by using a transwell insert (coculture-transwell), with HAEC-conditioned media in the absence of any HAECs (conditioned media alone), or with media alone (media). All data are shown as mean ± SEM. When compared with monoculture in SCF, only media alone was unable to support MC survival (n = 4; comparison made by one-way ANOVA; ^*, P < 0.002). All other conditions were not statistically different from culture with SCF 100 ng/ml. (B) Inhibitory effect of anti-SCF blocking Ab (5 μg/ml) on MC survival in coculture compared with coculture without blocking Ab, isotype-matched IgG (5 μg/ml), or media without SCF. Data presented are mean ± SEM (n = 3).

Fig. 4.

MC protease expression identified by immunostaining over time. MCs were recovered after 2 d of culture in liquid suspension culture with SCF alone, in direct coculture with HAECs, in HAEC coculture separated by a transwell membrane, with HAEC-conditioned media, or in suspension culture lacking any cytokines. MCs were fixed with Carnoy's fixative, and expression of tryptase and chymase granule proteases was detected by immunocytochemistry. Representative cells for each condition are shown. (Magnification: ×400.) Cells (100) were counted per condition for each replicate (n = 4).

Fig. 5.

Effect of coculture of MCs with HAECs on IgE-dependent histamine release and eicosanoid generation as measured by ELISA. MCs were maintained in monoculture with SCF or in coculture with HAECs for 4 d. Results depict percentage histamine release, cys-LT generation_, and PGD₂ production after 1 d of passive sensitization with IgE followed by activation with anti-IgE. Results are the mean ± SEM of four experiments. ^*, P = 0.05 when compared with PGD₂ generation of MCs grown in SCF monoculture.

Fig. 6.

Effect of coculture with IL-4 priming of MCs with HAECs on IgE-dependent histamine release and eicosanoid generation. MCs were maintained in monoculture with SCF or in coculture with HAECs for 4 d in the presence of recombinant human IL-4 (10 ng/ml). Results depict percentage histamine release, cys-LT generation_, and PGD₂ production after 1 d of passive sensitization with IgE followed by activation with anti-IgE. Results are the mean ± SEM of four experiments. ^*, P < 0.05 when compared with cys-LT generation of unprimed MCs grown in SCF monoculture; ^**, P < 0.05 when compared with PGD₂ generation of unprimed MCs grown in SCF monoculture; ^***, P < 0.05 when compared with cys-LT generation of unprimed MCs grown with HAEC coculture.

Fig. 7.

Effect of coculture with IL-4 and IL-3 priming of MCs with HAECs on IgE-dependent histamine release and eicosanoid generation. MCs were maintained in monoculture with SCF or in coculture with HAECs for 4 d in the presence of recombinant human IL-4 (10 ng/ml) and IL-3 (5 ng/ml). Results depict percentage histamine release, cys-LT generation_, and PGD₂ production after 1 d of passive sensitization with IgE followed by activation with anti-IgE. Results are the mean ± SEM of four experiments. ^*, P < 0.05 when compared with cys-LT generation of IL-4 primed MCs grown with SCF in monoculture; ^**, P < 0.05 when compared with cys-LT generation of IL-4 primed MCs grown in HAEC coculture.