Response of Differentiated Human Airway Epithelia to Alcohol Exposure and Klebsiella Pneumoniae Challenge

Abstract

Alcohol abuse has been associated with increased susceptibility to pulmonary infection. It is not fully defined how alcohol contributes to the host defense compromise. Here primary human airway epithelial cells were cultured at an air-liquid interface to form a differentiated and polarized epithelium. This unique culture model allowed us to closely mimic lung infection in the context of alcohol abuse by basolateral alcohol exposure and apical live bacterial challenge. Application of clinically relevant concentrations of alcohol for 24 hours did not significantly alter epithelial integrity or barrier function. When apically challenged with viable Klebsiella pneumoniae, the cultured epithelia had an enhanced tightness which was unaffected by alcohol. Further, alcohol enhanced apical bacterial growth, but not bacterial binding to the cells. The cultured epithelium in the absence of any treatment or stimulation had a base-level IL-6 and IL-8 secretion. Apical bacterial challenge significantly elevated the basolateral secretion of inflammatory cytokines including IL-2, IL-4, IL-6, IL-8, IFN-γ, GM-CSF, and TNF-α. However, alcohol suppressed the observed cytokine burst in response to infection. Addition of adenosine receptor agonists negated the suppression of IL-6 and TNF-α. Thus, acute alcohol alters the epithelial cytokine response to infection, which can be partially mitigated by adenosine receptor agonists.

Keywords: Adenosine receptor; Air-liquid interface culture; Alcohol; Cytokines; Epithelial barrier function; Human airway epithelia; Klebsiella pneumoniae bacteria.

Fig. 1. Air-liquid interface culture of airway epithelial cells

A) A drawing depicting the air-liquid interface culture. NHBE cells are seeded on a semi-permeable filter and grown at the air-liquid interface. Nutritional support comes from the basolateral side and the apical side is exposed to the air. Such a setting facilitates epithelial polarization and differentiation. B & C) Confocal micrographs demonstrate that the cultured airway epithelia form a tight epithelial sheet as seen under horizontal optical scanning with apical ciliation discernable by vertical scanning. D) The electron micrograph shows the typical characteristics of polarized epithelium with apical cilia and microvilli, and pseudopodia extending into the basal filter membrane.

Fig. 2. Effect of alcohol exposure on epithelial integrity and transepithelial permeability

NHBE cells were grown at the air-liquid interface and were exposed to different concentrations of alcohol for 24 hours. A) The effect of alcohol on epithelial cell monolayer integrity was investigated by measuring the transepithelial electrical resistance (TEER). As displayed, alcohol did not induce any significant changes in TEER. B) The effect of alcohol on the transepithelial permeability was estimated by measuring the apical to basolateral diffusion of FITC-Inulin. The acute alcohol exposure did not markedly affect the transepithelial permeability. Data expressed is representative of 2 independent experiments carried out in quadruplicates.

Fig. 3. Effect of alcohol on epithelial integrity and transepithelial permeability under apical K. pneumoniae challenge

Polarized NHBE cells were challenged with ~10 MOIs of apical K. pneumoniae for 24 hours in the absence (0mM) or presence (100mM) of basolateral alcohol for 24 hours. A) TEER measurements of NHBE cells showed a marked TEER increase with K. pneumoniae challenge as compared to the no alcohol treatment and no bacterial challenge control. Further, alcohol did not have any negative effect on such bacterium-induced TEER increase. B) Similar to the TEER data, transepithelial FITC-inulin diffusion measurements indicated that K. pneumoniae challenge enhanced the epithelial barrier function by significant reduction in FITC-inulin transepithelial diffusion. Under this condition, alcohol exposure did not influence the K. pneumoniae effects on epithelial integrity. Asterisks denote statistically significant difference (p ≤0.05, n=5).

Fig. 4. Alcohol effect on apical bacterial viability and binding

A) Polarized NHBEs were treated with 0 and 50 mM of basolateral alcohol for 24 hours and apically challenged with K. pneumoniae (MOI= ~10). The epithelia were lysed and bacterial supernatants were diluted and plated on agar plates for 24 hours. The colonies were counted to estimate the viable colonies. Alcohol presence promoted the number of colonies on airway epithelia. B) NHBEs were exposed to 0 and 50 mM of basolateral alcohol for 24 hours and then apically applied with ¹⁴C-labelled K. pneumoniae for an hour. The cells were washed and lysed to count radioactivity. Bound bacterial numbers were determined by comparing the cell-retaining radioactivity with a pre-established standard curve with various bacterial numbers. As demonstrated, alcohol did not alter the number of K. pneumoniae bound to NHBEs. Asterisk denotes statistically significant difference (p≤0.05, n=4–6).

Fig. 5. Alcohol effect on epithelial basolateral cytokine release under apical bacterial challenge

NHBE cells were grown at the air-liquid interface and challenged with live K. pneumonia apically in the presence (50 mM) or absence (0 mM) of basolateral alcohol for 24 hours. Eight different cytokines in the basolateral media were estimated by cytokine Bio-plex assay. Data are expressed as fold changes as compared to the base levels of cytokines secreted by the no treatment control. The absolute amount of the base level of each cytokine is presented in Table 1.

Fig. 6. Adenosine receptor agonists partially restore alcohol-suppressed cytokines

Polarized epithelia were exposed to 0 or 50 mM of alcohol basolaterally and challenged with K. pneumoniae apically in presence of adenosine (100 μM) or NECA (10 μM). Bacterial challenge significantly elevated IL-6, TNF-α and MCP-1 release. Alcohol blunted such an increase. However, adenosine and NECA partially restored the alcohol-suppressed IL-6 (A) and almost completely restored the alcohol-suppressed TNF-α (B). In contrast, the adenosine receptor agonists had no effect on the alcohol-suppressed MCP-1. Asterisks denote statistically significant difference (p≤0.05, n=4–6).