Analysis of Human Bronchial Epithelial Cell Proinflammatory Response to Burkholderia cenocepacia Infection

Background: Airway epithelial cells form the first line of defense against B. cenocepacia infection. Results: Airway epithelial cells express low levels of caspase-1 and do not release processed IL-1β. Conclusion: Low level expression of caspase-1 results in the inability of airway epithelial cells to secrete IL-1β. Significance: This study provides novel insight into airway epithelial cell response in host defense. Burkholderia cenocepacia, the causative agent of cepacia syndrome, primarily affects cystic fibrosis patients, often leading to death. In the lung, epithelial cells serve as the initial barrier to airway infections, yet their responses to B. cenocepacia have not been fully investigated. Here, we examined the molecular responses of human airway epithelial cells to B. cenocepacia infection. Infection led to early signaling events such as activation of Erk, Akt, and NF-κB. Further, TNFα, IL-6, IL-8, and IL-1β were all significantly induced upon infection, but no IL-1β was detected in the supernatants. Because caspase-1 is required for IL-1β processing and release, we examined its expression in airway epithelial cells. Interestingly, little to no caspase-1 was detectable in airway epithelial cells. Transfection of caspase-1 into airway epithelial cells restored their ability to secrete IL-1β following B. cenocepacia infection, suggesting that a deficiency in caspase-1 is responsible, at least in part, for the attenuated IL-1β secretion.

Burkholderia cenocepacia, the causative agent of cepacia syndrome, primarily affects cystic fibrosis patients, often leading to death. In the lung, epithelial cells serve as the initial barrier to airway infections, yet their responses to B. cenocepacia have not been fully investigated. Here, we examined the molecular responses of human airway epithelial cells to B. cenocepacia infection. Infection led to early signaling events such as activation of Erk, Akt, and NF-B. Further, TNF␣, IL-6, IL-8, and IL-1␤ were all significantly induced upon infection, but no IL-1␤ was detected in the supernatants. Because caspase-1 is required for IL-1␤ processing and release, we examined its expression in airway epithelial cells. Interestingly, little to no caspase-1 was detectable in airway epithelial cells. Transfection of caspase-1 into airway epithelial cells restored their ability to secrete IL-1␤ following B. cenocepacia infection, suggesting that a deficiency in caspase-1 is responsible, at least in part, for the attenuated IL-1␤ secretion.
Burkholderia cenocepacia infects a relatively small proportion of cystic fibrosis patients, but the resulting cepacia syn-drome carries a high mortality rate. Burkholderia is a family of naturally multidrug-resistant bacteria that are routinely cleared by those with fully functional immune systems. However, immunocompromised individuals including cystic fibrosis patients have great difficulty in fighting B. cenocepacia infection. The lung epithelial barrier is typically the first line of defense against airway infection by B. cenocepacia and other pathogens, but the molecular responses of airway epithelial cells to B. cenocepacia infection have yet to be fully elucidated.
Here, we examined the responses of airway epithelial cells (primary and cell line) to infection with B. cenocepacia. Interestingly, we found that although infection elicited powerful immune responses such as NF-B activation and cytokine/ chemokine secretion, the secretion of IL-1␤ was compromised. Further examination showed that IL-1␤ was produced in significant quantity within the cell, but that levels of caspase-1 were insufficient to effect the cleavage and release of this cytokine. Transfection of caspase-1 abrogated this limitation, restoring the ability of airway epithelial cells to secrete IL-1␤. These results suggest that although airway epithelial cells are capable of producing IL-1␤, the inflammasome, namely caspase-1, is a significant rate-limiting step.
Bacterial Culture and Infections-B. cenocepacia strain K56-2 isolate (kindly provided by Dr. Miguel Valvano) was grown in LB broth (Sigma-Aldrich) overnight (12-14 h) until post-logarithmic phase. A 600 was taken and used to calculate the multiplicity of infection (MOI) for each experiment. Prior to infection, epithelial cells were washed twice with 1ϫ PBS, and culture medium was replaced with antibiotic-free medium. Airway epithelial cells were infected at MOI of 5 for 1 Both authors contributed equally to this work. 2  5, 15, 30, and 60 min for phosphorylation studies and for 24 h in the mRNA, cytokine production, and transfection studies. ELISA-Cell-free supernatants were collected from resting and infected cells and analyzed using sandwich ELISA kits specific for IL-1␤, TNF␣ (R&D Systems), and IL-6 and IL-8 (eBioscience, San Diego, CA). In addition, whole cell lysates were also used for the measurement of IL-1␤ in specified experiments.
Real-time RT-qPCR-RNA was extracted using TRIzol, reverse-transcribed to cDNA, and then run in triplicate on an Applied Biosystems StepOnePlus system, with automatically calculated thresholds. Relative expression was calculated as 2ˆϪ ⌬Ct , with ⌬C t calculated by subtracting the average C t of two housekeeping controls (CAP-1 and GAPDH) from the experimental sample C t (3).

RESULTS AND DISCUSSION
B. cenocepacia Drives Early Signaling Events and Cytokine/ Chemokine Production-We began by testing the signaling responses of 16HBE14o Ϫ bronchial epithelial cells to B. cenocepacia infection. Cells were infected with 5 MOI B. cenocepacia for 5, 15, or 30 min. Protein-matched lysates were analyzed by Western blotting. Results showed significant up-regulation of Erk and NF-B phosphorylation, in accordance with earlier findings (4,5), as well as increased phosphorylation of Akt as seen in monocytic cells (6) (Fig. 1A). We then infected cells overnight and examined the induction of TNF␣, IL8, IL6, and IL1␤ by RT-qPCR. Transcripts of all four cytokines were significantly up-regulated following infection (Fig. 1B). Next, to detect cytokine levels in the cell supernatants, ELISAs were performed. Results indicated significant increases in TNF␣, IL-8, and IL-6 levels in the supernatants after infection (Fig.  1C). It had previously been reported that primary airway epithelial cells did not secrete TNF␣ (nor IL-1␤) following 24 h of infection with B. cenocepacia (7). Our contrasting findings may be due to the lack of a mucosal layer in our experimental system, as well as differences in strain and MOI.
Despite significant increases in transcript, IL-1␤ protein was below the detection threshold in cell supernatants. To confirm that this was not a cell line-specific phenomenon, we repeated all of these experiments using Calu-3 cells and found similar results (data not shown).
Airway Epithelial Cells Do Not Release IL-1␤ following B. cenocepacia Infection-Because no mature IL-1␤ was detected in the supernatants of infected cells despite significant increases in IL1␤ transcript, we measured levels of both unsecreted and secreted IL-1␤, comparing levels between various human airway epithelial cells and monocytes (known to release IL-1␤). As shown in Fig. 2, human peripheral blood monocytes (PBM) secreted IL-1 ␤ in ng/ml quantities following infection ( Fig. 2A), and moderate levels were found in PBM lysates (data not shown). However, upon examination of cell lysates, significant quantities of IL-1␤ (likely pro-IL-1␤) were found in 16HBE14o Ϫ (Fig. 2B), Calu-3 (Fig. 2C), Beas-2B (Fig. 2D), and HBEC (Fig. 2E), but none was detectable in the supernatants. To confirm that this lack of secretion was not a phenomenon specific to B. cenocepacia infection, we also compared responses of Beas-2B cells to LPS treatment. Results (Fig. 2F) showed that LPS treatment also led to significant quantities of IL-1␤ within the cells, with little to none secreted. These results suggest that the processing and release of IL-1␤ may be defective in airway epithelial cells. To further confirm that IL-1␤ was neither processed nor secreted into the supernatants, we analyzed the presence of IL-1␤ in lysates and supernatants of Beas-2B cells infected with B. cenocepacia by Western blotting. The results confirmed that although pro-IL-1␤ was abundantly produced, there was virtually no detectable mature IL-1␤ (17.5 kDa) in the lysates nor in the supernatants (Fig.  2G). In contrast, mature IL-1␤ was readily detected in the lysates and supernatants of PBM infected with B. cenocepacia (data not shown).
Airway Epithelial Cells Express Minimal Caspase-1-The inflammasome is a multiprotein complex, the assembly of which is critical for the activation of caspase-1. Caspase-1 is required for cleavage and release of IL-1␤. Thus, we measured caspase-1 transcripts in human airway epithelial cells versus PBM. Real-time RT-PCR with human PBM, 16HBE14o Ϫ , Beas-2B, and HBECs showed significantly lower caspase-1 transcript in epithelial cells as compared with PBM (Fig. 3A). To verify these findings, we performed Western blot analyses using protein-matched lysates from these cells. Results showed that although caspase-1 expression was readily detected in PBM, there was little to no detection of caspase-1 protein in airway epithelial cells (Fig. 3, B and C). In contrast, the expression of ASC was abundant in both PBM and airway epithelial cells (Fig.  3, D-F).
Transfection of Caspase-1 Restores the Ability to Secrete IL-1␤-Next, we tested whether by adding caspase-1 to the epithelial cells we could overcome the defective secretion. We transiently transfected Beas-2B cells to express YFP-caspase-1 (Fig.  3G). We then infected the transfected cells with B. cenocepacia and measured IL-1␤ in both lysates and supernatants. Caspase-1-transfected cells readily secreted IL-␤ into the supernatants, whereas mock-transfected cells failed to do so despite the presence of significant amounts of IL-1␤ in the lysates (Fig. 3, H, left and right panels, respectively). Collectively, these results suggest that airway epithelial cells do produce substantial levels of IL-1␤ upon stimulation but that the vast majority of it remains within the cell due to deficient levels of caspase-1.
Previous studies have shown that airway epithelial cells could, at least to some degree, produce and secrete IL-1␤ in response to stimuli such as infection (8), cigarette smoke (9, 10), and particulate matter (11,12). However, with the exception of Mortaz et al. (10) at levels of secreted IL-1␤ were relatively low. In light of the present study, our explanation for this would be that there is a specific deficiency in caspase-1 expression within these cells, such that IL-1␤ accumulates inside and is secreted only at near undetectable levels. Transfection of caspase-1 permits secretion of IL-1␤ at near ng/ml levels, suggesting that low caspase-1 is a major factor limiting inflammasome function in epithelial cells. Further studies will be required to uncover the molecular mechanisms responsible for the low caspase-1 expression and to determine whether this inherent deficiency contributes to reduced bacterial clearance or serves to minimize collateral tissue damage within the context of B. cenocepacia infection.