Human β-defensins (hBDs) are antimicrobial peptides that play important roles in host defense against infection, inflammation and immunity. Previous studies showed that micro-organisms and proinflammatory mediators regulate the expression of these peptides in airway epithelial cells. The aim of the present study was to investigate the modulation of expression of hBDs in cultured primary bronchial epithelial cells (PBEC) by rhinovirus-16 (RV16), a respiratory virus responsible for the common cold and associated with asthma exacerbations. RV16 was found to induce expression of hBD-2 and -3 mRNA in PBEC, but did not affect hBD-1 mRNA. Viral replication appeared essential for rhinovirus-induced β-defensin mRNA expression, since UV-inactivated rhinovirus did not increase expression of hBD-2 and hBD-3 mRNA. Exposure to synthetic double-stranded RNA (dsRNA) molecule polyinosinic:polycytidylic acid had a similar effect as RV16 on mRNA expression of these peptides in PBEC. In line with this, PBEC were found to express TLR3, a Toll-like receptor involved in recognition of dsRNA. This study shows that rhinovirus infection of PBEC leads to increased hBD-2 and hBD-3 mRNA expression, which may play a role in both the uncomplicated common cold and in virus-associated exacerbations of asthma.
The airway epithelium constitutes an effective barrier against infection and contributes to host defense by producing a variety of compounds. This includes the epithelial secretion of antimicrobial peptides that equip the host with effector molecules active against pathogens . One of the families of antimicrobial peptides expressed by human bronchial epithelial cells is that of the β-defensins. These peptides belong to the family of defensins, characterized by three disulfide linkages among six cysteine residues . The importance of β-defensins for host defense in the lung is supported by studies such as those using airway epithelial cell cultures from patients with cystic fibrosis  and in mice rendered β-defensin deficient by targeted recombination . The human β-defensin (hBD) family comprises multiple members. hBD-1, which is constitutively expressed , and hBD-2 and hBD-3, expression of which is inducible upon stimulation with bacteria and/or cytokines [6,7]. Expression of hBD-1 and hBD-2 mRNA has been described in epithelial cells of various organs [6–9] and in mononuclear phagocytes and dendritic cells . hBD-3 mRNA expression has been shown in primary bronchial epithelial cells (PBEC) and keratinocytes [11–13]. Recently, hBD-4 has been characterized by screening of genomic sequences . Expression of this 50 amino acid long peptide in airway epithelial cells is induced upon stimulation with bacteria but not with other inflammatory cytokines . Finally, using bioinformatics methods to screen several sequence databases, five conserved β-defensin gene clusters were identified encoding up to 28 β-defensins . In addition to antimicrobial activity, hBDs also display other activities which may enable these peptides to link innate and adaptive immunity. These activities include mast cell degranulation  and chemotactic activity for monocytes, immature dendritic cells and memory T cells [17,18]. Another host defense protein expressed by airway epithelial cells is secretory leukocyte protease inhibitor (SLPI). This low molecular mass, cationic protein is a mucosal serine protease inhibitor that blocks the activity of proteinases, including neutrophil elastase and cathepsin G, but also displays broad-spectrum antimicrobial activity . Expression of SLPI is constitutive in airway epithelial cell lines and primary cell cultures, and can be increased in, e.g., the epithelial cell line cells upon stimulation of these cells with cytokines .
Rhinovirus is the main cause of the common cold. The virus is a member of the family of picornaviridae, small non-enveloped viruses, with a genome that consists of a single RNA strand. The observation that rhinovirus infections are associated with exacerbations in asthma in time  suggests that rhinovirus infections have a more severe impact in patients with asthma than in otherwise healthy humans. Whereas the effect of rhinovirus on epithelial expression of proinflammatory cytokines has been well documented, little is known about the effect of this virus on expression of antimicrobial peptides in epithelial cells. Recent preliminary reports suggest that rhinovirus may modulate β-defensin and SLPI expression. In cells of the lung epithelial cell line A549 and PBEC hBD-2 mRNA expression is induced upon stimulation with rhinovirus [22,23] or respiratory syncytical virus , whereas the levels of SLPI in induced sputum of asthma patients is reduced during an experimental rhinovirus infection .
In the present study we investigated the effect of rhinovirus infection in PBEC on hBD-1, -2 and -3 and SLPI mRNA expression. Since double-stranded RNA (dsRNA), which is formed during viral replication, may trigger antiviral responses in infected cells , we also investigated mRNA expression of hBD-1, -2 and -3 and SLPI in PBEC after stimulation with the synthetic dsRNA molecule polyinosinic:polycytidylic acid (polyI:C) [26–28] and compared these results with the mRNA expression found during rhinovirus infection.
2Materials and methods
The rhinovirus-16 strain (RV16, a generous gift from Dr. E.C. Dick, Madison, WI, USA) was propagated in H1-Hela cells. Briefly, H1-Hela cells were grown till near-confluence in a T150 culture flask (Greiner GmbH, Frickenhausen, Germany) in modified Eagle's medium (Gibco, Grand Island, NY, USA) containing 10% (v/v) heat-inactivated fetal calf serum, 2 mM l-glutamine, 200 U ml−1 penicillin, and 200 U ml−1 streptomycin. Next, cells were infected with RV16 (multiplicity of infection of 0.1) for 90 min at 32°C. Thereafter, extra medium was added and the cells were maintained at 32°C until a cytopathic effect of 3+ (range: 0–4+) was observed. Intracellular RV16 was harvested by three cycles of freezing–thawing of the cells, cellular debris was removed by repeated centrifugation (10 min, 4500×g at 4°C), and the virus was further purified according to the methods described by Gern et al. . The virus stock was stored at −80°C. Where indicated, RV16 was rendered replication deficient by exposure to UV light (RV16-UV). This was achieved by exposition of 800 μl medium containing 5×107 RV16 to 400 μW cm−12 UV light for 4 min, while being kept on ice. These conditions were optimized in pilot experiments, thus preventing unwanted artifacts associated with prolonged exposure of the virus to high doses of UV light.
Bronchial tissue specimens were obtained from patients who underwent a lobectomy or pneumectomy for lung cancer at the Leiden University Medical Center, Leiden, The Netherlands. PBEC were isolated from the tissue and cultured as described . When cells reached approximately 80–90% confluence, the medium was replaced by keratinocyte medium (KSFM, Gibco) containing 1 mM calcium chloride, 5 nM retinoic acid (Sigma Chemical Company, St. Louis, MO, USA), 0.2 ng ml−1 epidermal growth factor (Gibco), 25 μg ml−1 bovine pituitary extract (Gibco), 200 U ml−1 penicillin and 200 μg ml−1 streptomycin to allow cellular differentiation. Thirty-six hours thereafter the cells were exposed for 18, 24 or 48 h at 32°C to 1.7×106 TCID50/ml RV16, 1.7×106 TCID50/ml RV16-UV or 0.37 μg ml−1 polyI:C (Sigma). Preliminary experiments revealed that 0.37 μg ml−1 polyI:C was the lowest concentration inducing hBD-2 and hBD-3 mRNA expression in PBEC (data not shown). As negative control PBEC were cultured at 32°C in KSFM medium without stimuli. As positive control PBEC were stimulated with a mixture of cytokines, i.e. 10 U ml−1 recombinant human IL-1β (Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, Amsterdam, The Netherlands; CLB), 500 U ml−1 recombinant human IFNγ (Boehringer Ingelheim, Ingelheim, Germany), and 100 ng ml−1 recombinant human TNFα (PeproTech, Rocky Hill, NJ, USA), further referred to as cytomix.
At the indicated intervals, total RNA was extracted from the cells and subjected to RT-PCR as described before  using primers for hBD-1 (5′-TGAGTGTTGCCTGCCAGT-3′ 5′-TCTTCTGGTCACTCCCAG-3′) , hBD-2 (5′-CATCAGCCATGAGGGTCT-3′ 5′-AGGCAGGTAACAGGATCG-3′) , hBD-3 (5′-TAGCAGCTATGAGGATCCA-3′ 5′-CTTCGGCAGCATTTTCGG-3′) , SLPI (5′-TCACTCCTGCCTTCACCA-3′ 5′-CTCCTCCATATGGCAGGA-3′) , and GAPDH (5′-CATCACCATCTTCCAGGAGC-3′ 5′-GGATGATGTTCTGGAGAGCC-3′) . To confirm infection of PBEC by RV16, RNA of these cells was subjected to RT-PCR with primers for RV16 (5′-GAAGCCAAGTATTGGACAAGG-3′ 5′-AAGTCCAATCGAGAAGCACC-3′) . Toll-like receptor (TLR) expression was analyzed using primers for TLR2 (5′-TTTATCGTCTTCCTGCTTCAAGCC-3′ 5′-TCTCGCAGTTCCAAACATTCCAC-3′), TLR3 (5′-GCAAACACAAGCATTCGGAATCT-3′ 5′-TTGAAGGCTTGGGACCAAGGCA-3′) or TLR4 (5′-TTTCTGCAATGGATCAAGGACCA-3′ 5′-GGACACCACAACAATCACCTTTC-3′) . To control for contamination with genomic DNA, PCR reactions were also performed using the total RNA instead of cDNA. The amplified products were analyzed on a 2% agarose gel (Promega, Madison, WI, USA) and visualized with ethidium–bromide. For semiquantitative RT-PCR, 8 μl of the PCR product were diluted with 80 μl of 1× working solution SYBR Green I (Molecular Probes, Eugene, OR, USA) and next fluorescence at 485/535 nm was measured. The hBD-1–3 and SLPI mRNA levels were calculated using the GAPDH mRNA levels as internal standards.
The IL-8 concentrations in cell supernatants were determined using a commercial enzyme-linked immunosorbent assay (ELISA; CLB, Amsterdam, The Netherlands).
Results, expressed as fold increase relative to values obtained with cells cultured in KSFM medium for 18 h, are means±S.E.M. of three to five experiments using cells of different donors. Data were analyzed for statistical difference by the Wilcoxon Signed Ranks test for related samples.
PBEC cultured in medium alone for 24 h showed a marked hBD-1 mRNA expression, which did not change after infection with RV16 (Fig. 1A). The mRNA expression of hBD-2 and hBD-3 was low or absent in unstimulated PBEC, but was clearly increased during infection with RV16 (Fig. 1A). The results of a semiquantitative analysis of hBD-2 and hBD-3 mRNA expression are summarized in Fig. 1B. Expression of SLPI mRNA was detectable in unstimulated cells and no change in expression occurred during infection with RV16 (Fig. 1A). Cells infected with RV16 showed a clear mRNA expression of RV16 as detected by RT-PCR (data not shown). In addition, UV-inactivated RV16 did not affect mRNA expression of hBD-1, hBD-2, hBD-3 and SLPI in PBEC (Fig. 1A), indicating that replication of the virus was required to induce hBD-2 and hBD-3 mRNA in PBEC. To investigate whether dsRNA is sufficient to stimulate the expression of the antimicrobial peptides, PBEC were exposed for 24 h to polyI:C. The results revealed that polyI:C stimulation did not affect hBD-1 mRNA expression in PBEC (Fig. 1A). In contrast, polyI:C markedly increased the expression of hBD-2 and hBD-3 mRNA in PBEC (Fig. 1A,B), while SLPI mRNA expression was not affected (Fig. 1A).
Cytomix, used as a positive control, induced hBD-2 and hBD-3 mRNA expression in PBEC (Fig. 1). The hBD-1 mRNA and SLPI mRNA expression in these cells was not affected by this stimulus. The mRNA expression pattern for hBD-1–2 and -3 and SLPI in PBEC after 24 h of stimulation was the same as that at 18 and 48 h of stimulation (data not shown). In agreement with the results for SLPI mRNA, we found no significant change in SLPI protein as detected by ELISA in the cell culture medium of the epithelial cells upon exposure to the various stimuli (data not shown).
The production of IL-8 by PBEC was markedly increased by the various stimuli. RV16 and polyI:C caused a 8±1.6- and 6±1.3-fold increase in IL-8 production by PBEC (n=3), respectively, whereas cytomix resulted in a much higher increase (83±51.3). UV-inactivated RV16 caused only a slight increase (2±0.7) in IL-8 production by these cells.
Because of the role of TLRs in recognizing conserved microbial molecular patterns, and because TLR3 has been shown to recognize dsRNA, we assessed the expression of TLRs in PBEC by RT-PCR. The results showed that PBEC not only express the TLR3, but also TLR2 and TLR4 that are involved in recognizing Gram-positive (TLR2) and Gram-negative (TLR4) bacteria (Fig. 2).
The results from the present study show that RV16 differentially affects antimicrobial peptide expression in subcultures of PBEC, as demonstrated by an increase in hBD-2 and hBD-3 mRNA, and unchanged hBD-1 and SLPI mRNA levels. This increase in hBD-2 and hBD-3 mRNA was paralleled by an increase in IL-8 release. The induction of hBD-2 and hBD-3 mRNA in PBEC appears to be dependent on viral replication, since UV-inactivated virus did not affect hBD-2 or hBD-3 mRNA expression in PBEC. Another observation in this study was that polyI:C also enhances the mRNA expression of hBD-2 and hBD-3 mRNA, but does not affect that of hBD-1 and SLPI. This indicates that dsRNA is sufficient to induce hBD-2 and hBD-3 expression. Since UV-inactivated RV16 does not form dsRNA in the infected epithelial cell, formation of dsRNA during viral replication  appears to be a prerequisite for the observed hBD-2 and hBD-3 induction. In support of these findings, earlier studies have reported that single-stranded RNA, in contrast to dsRNA, does not affect this expression of a variety of genes [36–39].
It is now clear that cells of the innate immune system employ pattern recognition receptors such as the TLRs to respond to micro-organisms or microbial products. Stimulation of epithelial cells through such receptors resulted in an increase in the expression of, e.g., cytokines and hBD-2 . A recent study revealed the involvement of TLR3 in response to dsRNA, using polyI:C as a model stimulus . These results suggest that the stimulatory effects of RV16 on β-defensin and IL-8 expression as observed in the present study, may be mediated not only by ICAM-1 , but may also involve TLR3. In agreement with this hypothesis, we noted the presence of TLR3 mRNA, in addition to TLR2 and TLR4 mRNA, in cultured PBEC using RT-PCR. Further functional studies are required to elucidate the contribution of TLR3 to RV16-induced β-defensin expression in airway epithelial cells.
Whereas we observed marked increases in the levels of hBD-2 and hBD-3 mRNA upon stimulation of bronchial epithelial cells with RV16 or polyI:C, we were not able to demonstrate an accompanying increase in the amount of β-defensin peptides. This may be explained by the lack of suitable immunoassays, which is clearly the case for the recently discovered hBD-3. However, a few reports demonstrated detectable hBD-2 peptide levels in conditioned media obtained from cultured epithelial cells. Therefore, we developed a sensitive ELISA for hBD-2 using a commercially available polyclonal antibody against hBD-2 (Peptide International, Osaka, Japan) reaching a detection limit of 2 ng ml−1. Using this assay, we failed to detect hBD-2 peptide in the cell culture supernatants, indicating that the peptide levels were below the detection limit of the assay. An alternative explanation is that the epitopes recognized by the anti-hBD-2 antibody are masked in hBD-2 present in conditioned medium from epithelial cell cultures due to, e.g., complex formation. In contrast, in a Western blot analysis of cell lysates several immunoreactive compounds (all differing from synthetic hBD-2) were detected using the same polyclonal antibody, suggesting either complex formation or lack of specificity of the antibody under these conditions. Other studies reported that cultured epithelial cells secrete small amounts of hBD-2 peptide upon stimulation, which are close to the limit of detection of our ELISA. This was shown, e.g., for colon epithelial cell lines (5 ng ml−1 hBD-2)  and the A549 lung epithelial cell line (7 ng ml−1) . It should be realized that in these studies as well as in our study submerged cell cultures were used, which are known to result in reduced mediator levels as compared to air–liquid interface epithelial cell cultures. Singh et al. reported concentrations of 8–10 μg ml−1 hBD-2 in the thin layer of airway surface fluid produced by air–liquid interface cultures of airway epithelial cells, whereas hBD-1 peptide was not detected . These levels of hBD2 appear to be sufficient to display antimicrobial activity, which was reported to require 0.1–10 μg ml−1[7,8,43]. Furthermore, it is likely that hBD-2 concentrations lower than 1 μg ml−1 contribute to the defense of the airways against microbes, since β-defensins act in concert with other antimicrobial peptides secreted by epithelial cells .
Our studies are in line with those from a recent study showing increased β-defensin mRNA in nasal brushes from individuals with an acute cold . What is the role of rhinovirus-enhanced expression of β-defensins in airway epithelium?α-Defensins inactivate various viruses, including herpes simplex virus, cytomegalovirus, influenza virus and HIV-1 [45–47]. Less is known about antiviral activities of β-defensins, except for a recent study showing that hBD-1 inactivates adenovirus . It is not clear whether β-defensins are effective against rhinoviruses. However, based on recent insight into the biology of β-defensins, properties other than antimicrobial activity may be relevant for understanding the putative role of increased epithelial β-defensin expression during rhinovirus infection. These peptides may be involved in orchestrating an immune response against the viral infection, as suggested by the finding that hBDs display chemotactic activity towards monocytes, immature dendritic cells and T cells [17,18]. Finally, the ability of β-defensins to stimulate mast cell exocytosis and prostaglandin synthesis  may be of importance to understand the role of β-defensins in rhinovirus-induced exacerbations of asthma.
This study was supported by grants from the Dutch Asthma Foundation (grant 96.27) and the Netherlands Organization for Scientific Research (NWO; grant 902-11–092).