State Key Laboratory of Respiratory Disease, The First Affiliated Hospital of Guangzhou Medical University, Guangzhou, China
Correspondence: Ai-huan Chen, State Key Laboratory of Respiratory Disease, The First Affiliated Hospital of Guangzhou Medical University, 151 Yanjiang Road, Guangzhou 510120, China. Email: email@example.com
Respiratory syncytial virus (RSV) results in acute wheezing in infants and is frequently associated with recurrent wheezing. Although RSV-induced wheezing clinically resembles that of asthma, corticosteroids are not equivalently effective in RSV-associated wheezing. The study sought to determine the mechanisms of RSV-induced wheezing by establishing an in vitro model of RSV-infected human bronchial epithelial cells (16-HBEC).
Leukotriene C4 synthase (LTC4S) messenger RNA (mRNA) expression in 16-HBEC was detected using fluorescence quantitative polymerase chain reaction, and the relative level of LTC4S mRNA was expressed as quotient cycle threshold (qCt) based on the threshold cycle number value compared with that of β-actin. Cysteinyl leukotrienes (CysLT) in culture supernatant were measured by enzyme-linked immunosorbent assay. RSV-infected 16-HBEC was incubated with gradient concentration of budesonide (BUD) to assess its effects on LTC4S expression and CysLT secretion.
RSV infection resulted in increased LTC4S mRNA expression between 48 and 96 h post-infection. High level of CysLT was detected in the supernatant of RSV-infected 16-HBEC. BUD at concentrations of 10−10 to 10−5 mol/L did not significantly alter LTC4S mRNA expression.
RSV infection upregulated LTC4S expression in HBEC leading to increased CysLT secretion. Such induction was not attenuated by BUD, suggesting that CysLT might contribute to the pathogenesis of RSV-induced wheezing.
Respiratory syncytial virus (RSV)-related lower respiratory tract infection is a common cause of acute wheezing in infants. Although RSV-induced wheezing clinically resembles asthma, glucocorticoids are not as equivalently effective as in asthma. The activation of a number of host genes during RSV infection results in the induction of pro-inflammatory cytokine and chemokine synthesis such as interleukin (IL)-6, IL-8 and 5-lipoxygenase.[2-4] Conceivably, RSV infection may induce inflammatory mediators in airway epithelial cells that might contribute to acute wheezing that presents different responses to glucocorticoids.
Cysteinyl leukotrienes (CysLT) may elicit bronchoconstriction, inflammatory cell infiltration, increased vascular permeability, mucus hypersecretion and airway mucosal oedema. High levels of leukotriene C4 (LTC4) have been documented in nasal lavage obtained from infants with bronchiolitis, and the level of LTC4 was positively correlated with disease severity. Furthermore, synthesis and secretion of leukotrienes were not effectively attenuated by glucocorticoids.[6, 7] It is likely that RSV-induced wheezing is associated with increased CysLT synthesis. In addition, first-line airway defences, airway epithelial cells, are able to generate inflammatory mediators and to express LTC4 synthase (LTC4S), which may be induced through 5-lipoxygenase signalling pathway leading to increased leukotrienes secretion.[8-10] To investigate the potential mechanisms of RSV-induced wheezing in infants, we sought to determine if LTC4S gene expression and leukotriene secretion in RSV-infected bronchial epithelial cell line, and if they could be suppressed by budesonide (BUD).
HEp-2 cells (human laryngeal carcinoma cell line; Shanghai Institute of Biochemistry and Cell Biology, Shanghai Institute of Biochemistry and Cell Biology, Shanghai, China) and human bronchial epithelial cell (HBEC) line (Southampton University, Southampton, UK) were incubated at 37°C with 5% CO2 in minimum essential medium supplemented with 10% foetal bovine serum (Hangzhou Sijiqing Biotechnology Co. Ltd, Hangzhou, China).
RSV preparation and quantification
RSV A2 strain (ATCC VR-1302, from American Type Culture Collection, Rockville, MD, USA) was inoculated to HEp-2 cells. Following sufficient propagation, the viral culture supernatant and cells were stored in a −80°C freezer. This was followed by repetitive freezing and thawing for thrice before centrifugation at 1000 g for 10 min. The supernatant was subsequently stored in 1.5-mL aliquots at −80°C.
A simplified plaque assay for RSV quantification was performed using Dulbecco's modified Eagle's medium supplemented with 1.5% carboxymethyl cellulose (Sigma, St. Louis, MO, USA) and 2% foetal bovine serum (carboxymethyl cellulose-Dulbecco's modified Eagle's medium) as the overlay. Virus sample was serially diluted with Dulbecco's modified Eagle's medium and was inoculated to HEp-2 cell monolayer. Five days later, the carboxymethyl cellulose-Dulbecco's modified Eagle's medium was aspirated, and the monolayer was fixed with phosphate buffer containing 0.1% crystal violet and 10% formalin for 2 h followed by plaque counting (Fig. S1 in the online supporting information). The plaque forming unit in per millilitre virus sample was calculated by using the formula plaque forming unit = a*b/v (a, the average number of plaques; b, reciprocal value of virus dilution; v, volume of the virus dilution). Stocks of virus contained 2.2 × 107 or 1.6 × 106 plaque forming unit/mL.
Infection of cells with RSV
Aliquots of 16-HBEC (1.0 × 104) were pipetted into a 96-well tissue culture plate and subjected to RSV infection at a serial multiplicity of infection (MOI) of 0.001, 0.01, 0.1, 1 and 10, respectively. The percentage of survival cells was measured by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay[12-14] every 24 h between 24 and 120 h post-infection. The plates were also examined under microscope to assess the cytopathic effect. Infection with MOI = 0.1 was found to induce apparent cytopathic effect, while the cell survival rate remains high at various time points (Fig. S2 in the online supporting information); therefore, MOI = 0.1 was employed to determine the kinetics of gene expression of LTC4S and CysLT in HBEC.
16-HBEC was cultured in a six-well tissue culture plate and was infected with RSV at MOI = 0.1 when cell confluence approached to 80%. The virus inoculum was replaced by minimum essential medium containing 2% foetal bovine serum after 2 h incubation, and then the cultures were maintained for up to 120 h. The LTC4S messenger RNA (mRNA) expression in 16-HBEC and CysLT in supernatant were measured every 24 h. The infection of RSV was examined at 72 h post-infection by fluorescence microscopy using fluorescein isocyanate-conjugated RSV monoclonal antibodies (Chemicon International, Inc., Temecula, CA, USA), and RSV fusion gene was targeted by reverse transcriptase-polymerase chain reaction (PCR).
Trizol reagent (1 mL, Life Technologies, Carlsbad, CA, USA) was added into each well of the six-well tissue culture plate of the 16-HBEC. Total RNA was extracted according to the manufacturer's instructions.
RNA (2 μg) was used as template for complimentary DNA synthesis. For fluorescent quantification polymerase chain reaction, the primers for LTC4S were 5′-TACCGAGCCCAGGTGAACTG-3′ (forward) and 5′-CGAACAGGTAGACCAGGCCG-3′ (reverse), and those for β-actin were 5′-GACTACCTCATGAAGATCCTCACC-3′ (forward) and 5′-TCTCCTTAATGTCACGCACGATT-3′ (reverse). The sequences of the probes were 5′-(FAM)-CTTCCCGCTGTTCCTCGCCACGC (Eclipse)-3′ for LTC4S and Taqman probe, and 5′-(FAM)-CGGCTACAGCTTCACCACCACGGC (Eclipse)-3′ for β-actin (Rui Zhen Biological Technology Co. Ltd, Guangzhou, China). PCR was performed using 1 μL complimentary DNA, 25 μL 2 × PCR buffer, 0.5 μL (5 U/μL) Taq polymerase (TaKaRa Co. Ltd, Tokyo, Japan), 2 μL of each primer, 0.4 μL of probe (25 mmol/mL), 8 μL of 10 mmol/L deoxyribonucleoside triphosphates (dNTP) Mix and 11.1 μL of deionized water to a total volume of 50 μL. Following denaturation for 5 min at 95°C, PCR amplification was performed by 40 cycles of 94°C for 1 s, 60°C for 20 s and 72°C for 40 s. Quantification of PCR products was achieved by plotting the fluorescence signals versus the cycle numbers at which the signals crossed the baseline.
The threshold cycle number was calculated by using Sequence Detection System software (iQ 5 Multicolor Real-Time PCR, Bio-Rad, Inc., Hercules, CA, USA) with an automatic setting of the baseline. Standard curves were derived by plotting the threshold cycle number values against the logarithm of the number of copies and that of unknown samples inferred from the regression line. The threshold cycle number value of each sample was adjusted by that of β-actin, and the ratio (qCt) was inversely proportional to both LTC4S mRNA levels and the gene copies.
Enzyme-linked immunosorbent assa
CysLT in the culture supernatant was measured using enzyme-linked immunosorbent assay kits (Cayman, Ann Arbor, MI, USA). Each sample was assayed in duplicate, and the experiment was performed with the same sample twice.
Infection of cells with inactivated RSV
For viral inactivation, aliquots of stock RSV virus were exposed to 1800 mJ/cm2 ultraviolet (UV) radiation for 1 h and referred to as UV-RSV. The 16-HBEC was infected with UV-RSV at MOI = 0.1. The LTC4S mRNA expression in 16-HBEC and total CysLT in culture supernatant were measured at 72 h after inoculation.
BUD (AstraZeneca Co. Ltd, Lund, Sweden) was dissolved in ethanol at 10−2 mol/L and stored at −20°C. 16-HBEC infected by RSV (MOI = 0.1) were incubated with different concentrations of BUD (10−10, 10−8, 10−6 and 10−4 mol/L) containing 2% foetal bovine serum medium for maintenance culture of 72 h followed by harvesting the cells and supernatant.
Statistical analysis was conducted using SPSS for Windows version 12.0 (SPSS Inc., Chicago, IL, USA). All variables were presented as mean ± standard deviation. Paired t-test was performed to compare between-group difference, and repeat-measure analysis of variance was employed for among-group comparison when appropriate. Statistical significance was deemed when P < 0.05.
In vitro model of RSV-infected 16-HBEC
The 16-HBEC did not show evidence of viral cytopathic effect under light microscope after RSV inoculation except for non-specific morphological changes including cell enlargement, increased opacity and clustered dead cells (Fig. 1a). In contrast, HEp-2 cells developed typical cytopathic effect following RSV inoculation (Fig. 1b). The green fluorescence emitted from the cytoplasm of RSV-infected 16-HBEC (Fig. 2a) showed that RSV-specific antigens were localized in the cytoplasm. To demonstrate successful RSV infection in 16-HBEC, the RSV fusion gene fragment (317 bp) was amplified by reverse transcriptase-PCR (Fig. 2b).
LTC4S mRNA expression
RSV infection yielded an increase of LTC4S mRNA expression in 16-HBEC at 48 h, and the level culminated at 72 h, corresponding to the most significantly reduced qCt (1.03 ± 0.12) following inoculation (P < 0.05). At 72 h post-infection, a difference in LTC4S mRNA expression between RSV-infected and non-RSV-infected 16-HBEC was observed (P < 0.05) (Fig. 3a).
At 72 h after inoculation with UV-RSV, LTC4S mRNA expression in 16-HBEC was not markedly different from that in control cells ((1.70 ± 0.10) vs (1.73 ± 0.21), P > 0.05). In contrast, there was a significant difference between RSV-infected and UV-RSV-inoculated 16-HBEC ((1.19 ± 0.06) vs (1.70 ± 0.10), P < 0.05, Fig. 3b).
CysLT secretion was shown by enzyme-linked immunosorbent assay to increase at 48 h following viral inoculation and culminated at 72 h, an effect that was well maintained until 96 h. An impressively higher level of CysLT secretion was noted at 72 h compared with 24, 48 and 120 h (all P < 0.01). RSV-infected 16-HBEC yielded considerably higher CysLT secretion at 72, 96 and 120 h compared with non-RSV-infected cells (all P < 0.01) (Fig. 3c).
In contrast, at 72 h after inoculation with UV-RSV, CysLT secretion did not differ markedly from that in control cells ((10.67 ± 1.97) vs (8.98 ± 1.51), P > 0.05). However, RSV-infected 16-HBEC showed increased secretion of CysLT compared with UV-RSV-inoculated 16-HBEC ((23.54 ± 2.48) vs (10.67 ± 1.97), P < 0.01, Fig. 3d).
Effects of BUD on LTC4S mRNA expression and CysLT secretion
Because BUD at 10−4 mol/L led to diffuse cell death under light microscope (Fig. S3 in the online supporting information), the concentration of 10−5 mol/L was employed as the upper limit in this study.[16, 17] Noticeably, high concentration of BUD at 10−5 mol/L resulted in neither significantly decreased LTC4S mRNA expression (1.38 ± 0.22) nor CysLT secretion (18.0 ± 2.42) when compared with those in RSV-infected cells without BUD (P > 0.05). BUD at concentrations of 10−10 to 10−5 mol/L did not effectively suppress LTC4S mRNA expression or CysLT secretion in RSV-infected 16-HBEC (Fig. 4).
RSV-lower respiratory tract infection is a common cause of wheezing in infancy and leads to high rate of hospitalization. Unfortunately, an effective treatment for RSV-induced wheezing is lacking. Despite this anti-asthmatic drugs are frequently prescribed to prevent and treat wheezing pertinent to RSV-lower respiratory tract infection.[18, 19] A systematic reviews and meta-analyses involving approximately 1200 children with viral bronchiolitis failed to warrant the use of corticosteroids in RSV-lower respiratory tract infection.[20-22] It has been therefore postulated that a few cytokines or inflammatory mediators could be responsible for the development of wheezing.[5, 23, 24]
A few studies have documented that secretion of CysLT might be triggered by RSV infection during acute wheezing. Volovitz et al. reported an approximately fivefold increase in the concentration of LTC4 in nasopharyngeal secretion obtained from children with RSV-induced bronchiolitis. In another study that determined the levels of CysLT in bronchoalveolar lavage fluid, Kim et al. reported significantly higher levels of CysLT in children with RSV-induced bronchiolitis than those in control group. HBEC, acknowledged as airway structural cells, are capable of releasing a spectrum of cytokines and chemokines. Furthermore, HBEC such as human bronchial epithelial (BEAS-2B) cells have been documented to express the 5-lipoxygenase pathway. It has been shown that primary HBEC may produce as much as 11 folds of LTC4 than that of 16-HBEC. This may support the hypothesis that the primary cells appear to be a more authentic host of RSV infection due to a significant increase of leukotrienes secretion.
In an attempt to investigate the impact of RSV infection on 5-lipoxygenase and endothelin-1 expression, Behera et al. established an in vitro model of BEAS-2B cells (MOI = 5) yet failed to delineate the cytopathic effect and the optimal condition for observation. In our study, with MOI = 0.1, the cytopathological changes in 16-HBEC were observed at 72 h, while most cells retained their viability. This suggested that at 72 h post-infection with an MOI of 0.1 could be deemed as the optimal condition for observation of LTC4S mRNA expression and CysLT secretion in HBEC.
A limited number of publications investigate CysLT synthesis in bronchial epithelial cells with RSV infection. Behera et al. showed that RSV infection upregulated 5-lipoxygenase expression in BEAS-2B cells and led to more intensive CysLT synthesis and secretion. Our study demonstrated that RSV infection significantly upregulates LTC4S mRNA expression and CysLT synthesis in epithelial cells. However, LTC4S expression and CysLT synthesis was only upregulated at 72 h or more following RSV infection. Villenave et al. and Yoon et al. reported that IL-6 and IL-8 was significantly upregulated at 24 h or less in HBEC with RSV infection. Furthermore, leukotrienes reportedly upregulated IL-6 expression possibly due to a positive feedback from the inflammatory responses. If there was a biphasic induction of LTC4S and other cytokines such as IL-6, then the early-phase increase in IL-6 and IL-8 may be responsible for late-phase augmented release of CysLT. Additionally, 16-HBEC inoculation with UV-RSV, did not show any increased LTC4S mRNA expression and CysLT secretion, suggesting that those increases in RSV infected 16-HBEC were likely correlated with the activity of RSV, and the release of inflammatory mediators in airway epithelial cells might contribute to the pathogenesis of RSV-induced acute wheezing.
The impact of corticosteroids on synthesis and release of leukotrienes in asthmatic patients has been inconclusive. Tanaka et al. documented considerably reduced urinary Leukotriene E4 (LTE4) in asthmatic patients who received inhaled fluticasone propionate (800 μg once daily) for 2 weeks. In other studies,[29, 30] however, neither urinary LTE4 nor sputum leukotrienes were altered by high-dose inhaled corticosteroids (>1000 μg daily). Our results suggested that inhaled corticosteroid might not effectively suppress the upregulation of LTC4S mRNA expression and CysLT secretion in RSV-infected 16-HBEC. This may be helpful to interpret the different effects of corticosteroids between asthmatic patients and children with RSV-induced bronchiolitis. The concentration of BUD (10−5 mol/L) used in this study was higher than that employed in literatures (10−10 to 10−6 mol/L).[31-33] The effect of BUD on cell apoptosis has been of interest, as suggested by inhibition of apoptosis in HBEC induced by microbial infection via Mitogen-activited protein kinase (MAPK) pathway, but corticosteroids was able to induce apoptosis of airway epithelial cells. Although it is reported that phospholipase in human lung mast cells is insensitive to corticosteroid and dexamethasone could not block leukotriene pathway enzymes including LTC4S, desensitization of glucocorticoids receptors may partly explain the no response to steroid.[37, 38] The precise mechanism why BUD is ineffective in inhibiting LTC4S expression has yet to be elucidated. It is unknown if BUD could effectively suppress the expression of LTC4S and CysLT in primary HBEC with higher levels. The lack of measurement of LTC4S at protein levels may therefore be a limitation of the study.
Our study revealed that the increase of CysLT synthesis in bronchial epithelial cells correlated with RSV infection and that BUD was unable to attenuate CysLT synthesis. Further studies of corticosteroids on LTC4S mRNA expression and CysLT secretion in children with acute RSV bronchiolitis will be undertaken to provide additional clinical evidence.
We thank Wei-jie Guan (First Affiliated Hospital of Guangzhou Medical University, State Key Laboratory of Respiratory Disease) for his linguistic consideration and suggestions.