Deletion of TLR3 Alters the Pulmonary Immune Environment and Mucus Production during Respiratory Syncytial Virus Infection

Severe Respiratory Syncytial virus (RSV)² infections in infants and high risk adults have been recognized to have two primary components that relate to the severity of the pathophysiology: the intensity of the inflammation and the overproduction of mucus (1, 2, 3, 4, 5). Because airway epithelium is the primary site of RSV replication, the immune response that develops within the lungs of infected individuals dictates the subsequent immune responses (4, 6, 7, 8). The immune environment will dictate several aspects, including the clearance of the virus itself, as well as the severity of the pathophysiologic responses that can accompany RSV infection. Thus, the ability to respond effectively to the viral response early in the process may be a pivotal aspect in defining the severity of the pathologic responses. In children with severe RSV infections, a number of leukocyte populations can be identified from airway samples, including neutrophils, macrophages, lymphocytes, and eosinophils (9, 10). Using animal models of disease to examine temporal expression of leukocyte accumulation, the initial cellular response within the first 4 days appears to consist of macrophages, neutrophils, and NK cell migration (11, 12). One of the more pathological mechanisms in severely infected infants during RSV infection is the dysregulation of mucus production within the airway. The overproduction of mucus is especially detrimental in infants with severe RSV responses, whose airways are small and can be easily obstructed (6, 13, 14, 15).

TLR provide a recognition system to monitor pathogen invasion and activate the immune response for expulsion of infectious agents. These important innate immune molecules are not only necessary for the initiation of effective immunity, but have also been implicated in the early activation pathways necessary for cellular recruitment through the generation of chemokines at the site of infection. A number of TLRs have been linked to viral infection, including TLR3 (dsRNA), TLR4 (binding to f-protein of RSV) (16, 17), TLR7/8 (ssRNA), and TLR9 (unmethylated CpG). To transcribe and translate the protein components of RSV for assembly, the virus must go through a dsRNA step. It is recognition of this stage by TLR3 that probably leads to recognition of the infection by the host immune response. The activation of TLR3 in dendritic cells using polyinosinic-polycytidylic acid leads to significant IL-12 and type I IFN production, which is required for initiation of a cytotoxic IFN-γ-mediated response (18, 19). Recent studies have indicated that although TLR3^−/− mice do not have defects in viral clearance, TLR7^−/− mice appear to have altered viral clearance (20, 21). In other studies examining mouse CMV infection, both TLR3 and TLR9 are essential for an optimal immune response for clearance of the virus (22). In addition, other important dsRNA-responding molecules may have an activating role during viral infections, including RNA-dependent protein kinase and retinoic acid-inducible gene-I (RIG-I) (23, 24). Our results indicate that RSV does not require TLR3 for effective clearance but suggest an important role for TLR3 in regulation of the pulmonary immune environment and subsequent mucus hypersecretion. Thus, although TLR3 may not alter viral clearance responses, the appropriate activation of TLR3 may be important for the regulation of pathogenic responses in the lung environment.

Specific pathogen-free C57BL6 wild-type mice were purchased from The Jackson Laboratory and housed in University of Michigan animal facilities under pathogen-free conditions. TLR3^−/− mice were originally generated in Dr. Flavell’s laboratory and were backcrossed 10 generations onto a C57BL/6 background; a breeding colony was established at the University of Michigan. Age- and sex-matched mice were used in these studies.

Human RSV A strain, originally isolated at University of Michigan Hospitals, was propagated in Hep2 cells as previously described (25, 26, 27). To determine viral titers, a plaque assay was performed in Vero cells that were grown until they were semiconfluent. The viral stock was serially diluted and added to each well in duplicate. The plates were incubated for 4–5 days at 37°C while syncytia formed. The methylcellulose solution was removed from each of the wells, and the cells were fixed with cold 80% methanol for 60 min at −80°C. The methanol was removed, and the plates were frozen at −80°C for 1 h to lyse the cells before the plates were stained for RSV proteins. Blocking was performed using 5% dry milk/PBS. The primary Ab used was goat anti-RSV polyclonal Ab (Chemicon International). This Ab was used at a dilution of 1/500. The secondary Ab used was rabbit anti-goat Ab conjugated to HRP (Serotec). It was diluted 1/100 for use. Ab incubations were conducted for 1 h at 37°C. One-Step chloronapthol (Pierce) was added to each well, and the cells were incubated for 10 min at room temperature. Cells were washed with PBS, and plaques were counted.

Mice were infected intratracheally with ∼1 × 10⁵ PFU of human RSV A strain virus isolated from University of Michigan Hospitals and used as previously described (25, 26, 27). Animals were then maintained under biohazard containment in University of Michigan animal facilities.

Murine cytokine concentrations were determined using a standardized sandwich ELISA technique. Briefly, Maxisorp ELISA plates (Fisher Scientific) were coated with 1–5 μg/ml polyclonal capture Ab (R&D Systems) in coating buffer (0.6 M NaCl, 0.26 M H₃BO₄, and 0.08 M NaOH (pH 9.6)) overnight at 4°C and washed with PBS containing 0.05% Tween 20. Nonspecific binding sites were blocked with 2% BSA in PBS for 1 h at 37°C. After washing, cell-free lung homogenate or standard was added to each well and incubated for 1 h at 37°C. Plates were subsequently washed and incubated with biotinylated, affinity-purified, polyclonal detection Ab (0.1–3.5 μg/ml) for 45 min at 37°C. After thorough washing, streptavidin-HRP (BD Pharmingen) was added to each well for 30 min at 37°C. Plates were washed again, and after the addition of chromogen substrate, OD readings were measured at 492 nm. Recombinant cytokines and chemokines (R&D Systems) were used to generate standard curves, and the limit of detection for the assays was 50 pg/ml.

RNA was isolated from lung tissue after homogenization in TRIzol (Invitrogen Life Technologies) according to the manufacturer’s protocol. Total RNA (0.5 μg) was reverse transcribed in a 25-μl volume. mRNA expression was determined in 1 μl of cDNA by TaqMan real-time PCR with a PRISM 7700 sequence detection system (Applied Biosystems) using gene-specific primers and probes labeled with 5′-6-FAM and 3′-TAMRA. The primers and probes used for the detection of mRNA expression were determined using predeveloped primer/probe sets (Applied Biosystems). Reactions were incubated for 2 min at 50°C, denatured for 10 min at 95°C, and subjected to 40 two-step amplification cycles with annealing/extension at 60°C for 1 min, followed by denaturation at 95°C for 15 s. A predeveloped primer/probe set for murine GAPDH (Applied Biosystems) was used as an internal control for quantification of the total amount of cDNA used in the reaction. Results are normalized to GAPDH expression and are presented as the fold increase in mRNA expression, compared with the level detected on day 0. The gob5 primer/probe set was used as previously described (25).

One milliliter of PBS bronchoalveolar lavage (BAL) was used to assess leukocyte influx into the infected animals. BAL samples were centrifuged, and the cellular pellets were resuspended in 200 μl of PBS; 50 μl of suspension was cytospun onto glass slides and differentially stained (Diff-Quick; BD Biosciences). Cellular contents were determined by counting 200 total cells/slide and determining the percentage of each population for each mouse. The total number of cells in BAL ranged from 2.5 × 10⁴ to 3.5 × 10⁴ and was not significantly different between wild-type and TLR3^−/− mice.

All results are expressed as the mean ± SE. Statistical significance was calculated by ANOVA, followed by Student-Newman-Keuls post test to calculate the p value. Significance was determined as a value of p < 0.05.

Our initial studies examined the pulmonary expression patterns of relevant virus associated TLRs during RSV infection in C57B/6 mice. The data in Fig. 1 indicate that during the infection, TLR3 and TLR7 mRNA were significantly up-regulated on day 8 of infection, whereas TLR9 was not. Previous studies have demonstrated that TLR3 has an integral role for responding to dsRNA (28, 29, 30, 31), a required intermediate for RSV that is a negative strand virus. It is important to note that previous studies have shown that this strain of mouse readily clears RSV and has little associated pathogenic response (27, 32, 33). The present studies found no significant difference in the TLR3^−/− animals’ ability to clear RSV, compared with wild-type control mice, examining RSV-specific pulmonary titers during infection (data not shown). Thus, in agreement with previous studies that examined several other RNA viruses in TLR3^−/− mice, our studies demonstrate that no alteration in the clearance response to RSV was evident (21).

One of the more insidious disease-associated responses to RSV infection in infants is the overproduction of mucus within the small airways, which often causes obstructive problems. The qualitative assessment of RSV infection was initially examined in TLR3^−/− mice by staining histologic sections of lungs with periodic acid-Schiff (PAS) stain to determine the degree of mucus expression. The images in Fig. 2,A illustrate that day 6 RSV-infected TLR3^−/− mice display significant expression of mucus staining, whereas wild-type control mice had none. A protein that has been associated with goblet cell hyperplasia/metaplasia and mucus overproduction, gob5, was examined by immunohistochemistry and was correspondingly expressed in the airways of TLR3^−/− mice (Fig. 2,B). In addition, we examined the temporal expression of gob5 mRNA by quantitative PCR analysis in the lungs during RSV infection and demonstrated a significant increase in gob5 mRNA expression in TLR3^−/− mice during RSV infection (Fig. 3). Thus, TLR3^−/− mice displayed an altered pathophysiology response to the virus related to the overexpression of mucus in the airways.

Regulation of the local immune environment is central to regulating the pathophysiologic responses of the lung. The data in Fig. 4 illustrate that IL-13 mRNA is not up-regulated in the lungs of wild-type mice, but the lungs of TLR3^−/− mice demonstrated significant increases in IL-13 throughout the infectious process. This cytokine has been closely linked to mucus overexpression in airways in RSV-induced disease as well as in asthmatic responses. Another Th2-type cytokine that has been linked to detrimental responses in RSV is IL-5, which demonstrated a significant increase in TLR3^−/− mice at later time points (Fig. 4). No difference in IL-4 was observed in the lungs of these animals.

The recruitment of particular leukocyte populations may be related to disease progression and severity. Examination of leukocytes in the airway by BAL analysis showed that the TLR3^−/− mice demonstrated a significant increase in eosinophils on day 8, whereas none were found in wild-type mice infected by RSV (Fig. 5). Although the presence of eosinophils in the airway diminished by day 12 of infection in TLR3^−/− mice, there was a significant increase in lymphocytes in TLR3^−/− animals.

Isolated lymph nodes were dispersed into single-cell suspensions and stimulated by anti-CD3 to assess the overall TCR-stimulated responses. The cells from infected wild-type mice displayed a predominant Th1-type response, with high levels of IFN-γ expression and little Th2 cytokine expression (Fig. 6). Although anti-CD3-stimulated lymph node cells from TLR3^−/− mice produced similar levels of IFN-γ, there was increased production of IL-5 and IL-13 in TLR3^−/− mice corresponding to the pulmonary responses. Uninfected mice did not display a divergent cytokine expression pattern upon activation, suggesting that RSV infection altered the immune environment.

To examine the mechanistic relationship of the pathophysiology in TLR3^−/− mice, we examined the role of IL-13 using neutralizing Abs to attempt to modify the immune environment. In TLR3^−/− mice given anti-IL-13, there was a diminished mucus response, and although not all airways in TLR3^−/− mice displayed plugging of mucus, none of the anti-IL-13-treated animals had any sign of mucus plugging (Fig. 7,A). Likewise, when we examined gob5 mRNA expression in lung samples from these same mice, a significant reduction of gob5 expression was clearly evident, even below that observed in wild-type, RSV-infected mice (Fig. 7 B). Thus, detrimental pathophysiology was significantly altered in TLR3^−/− mice given anti-IL-13, suggesting a causative link between appropriate activation via TLR3 and IL-13.

The severity and intensity of pulmonary viral disease are a combination of virus-induced damage and responses associated with the host’s immune response. The success of clearance of the viral infection depends upon the ability of the host to detect and properly respond to the infectious stimuli. At the same time, the ability to properly mount a relatively nonpathogenic reaction relies on the host to elicit the appropriate immune response. In these studies we have examined whether a pathogen pattern recognition receptor, TLR3, has a role in determining the outcome of the RSV infection. Although the deletion of TLR3 had little effect on clearance of the viral infection, it did have a significant function in determining the intensity of mucus overproduction. Although mucus production at the interface of the environment can be viewed as one mechanism of antipathogen response, clearly it can contribute to a pathogenic situation leading to long-term complications (34, 35, 36, 37, 38). These changes appeared to be linked directly to the cytokine phenotype within the lung and lymph nodes during RSV-induced responses. It is clear from a number of studies that the generation of a Th2-type response in the lungs can result in a severe pathogenic response, including mucus overproduction (37, 39, 40, 41, 42, 43, 44). In particular, the overproduction of IL-13 can be specifically linked to the intensity of the goblet cell hyper/metaplasia that often accompanies severe pulmonary diseases, highlighted in the above-mentioned reports using both overexpression systems and depletion strategies. More specifically, previous studies have identified a link between IL-13 and mucus overproduction in the pathogenesis of RSV-induced disease (26, 45, 46). Likewise, the influx of eosinophils that correlated with increased IL-5 expression may also contribute to the altered pathophysiologic responses. Interestingly, this disease phenotype appears to continue in the presence of substantial levels of IFN-γ production of TLR3^−/− mice. The continued production of IFN-γ may account for the efficient clearance of RSV and the lack of AHR (data not shown) observed in TLR3^−/− mice, but apparently is not sufficient to regulate the mucus overproduction. Especially convincing is the association of mucus/goblet cell staining and gob5 protein staining, indicating that TLR3^−/− mice have increased goblet cell differentiation and expansion. Because TLR3 has been implicated in efficient cross-presentation of soluble Ag to CD8 T cells, it may be that there is an altered CD4 and CD8 T cell response in TLR3^−/− mice (ongoing studies), because both CD4 and CD8 T cells have been identified as sources of IL-13 (47, 48, 49). The depletion of IL-13 in TLR3^−/− mice significantly reduced mucus expression in the lungs, as indicated by PAS/Alcian blue stain and gob5 mRNA expression. Together, these data suggest that TLR3 has a role in shaping the immune response related to the cytokine environment of the lung.

Previous studies have demonstrated that deletion of TLR3 in viral responses differentially altered viral clearance depending upon the virus used to examine the responses (21, 22). The data generated in these studies concur with those original findings that indicated little role for TLR3 in clearance of the virus. However, the role of TLR3 in dendritic cells has been linked to the generation of the Th1-type responses (50, 51). An important issue that should be considered is differential responses dependent upon the type of virus used to study the role of TLRs. For example, although TLR7 is important during influenza virus infection (20, 52), another negative strand virus, Sendai virus, is completely independent of TLR3 and -9 and relies on RIG-I for DC maturation (53). In addition, evidence indicates that vaccinia virus contains Toll/IL-1 receptor domain-containing proteins that block Toll/IL-1 receptor domain-containing adaptor-inducing IFN-β activation and therefore inhibit type I IFN activation (54, 55). Thus, the type of virus, how it infects cells (endosomal vs cytoplasmic), and whether it expresses inhibitor proteins may all become factors for determining the mode of cellular activation. Therefore, generalizations with regard to activation of immune responses by different classes of virus may be inappropriate for individual viral infections.

The findings in the present study have investigated the viral clearance and pathogenesis of disease in animals deficient in TLR3 during RSV infection. The data demonstrate that the activation of TLR3 during the progression of immune responses shape the pulmonary immune environment to promote a predominant Th1 type response that is overall less pathogenic. In contrast, in the absence of TLR3 the disease is skewed toward a Th2 like response with increases in more pathogenic cytokine phenotypes, such as IL-13 and IL-5. It is tempting to speculate about the inverse correlation of peak TLR3 expression on day 8 in wild-type mice with the peak gob5 expression on day 8 in TLR3^−/− mice. Perhaps TLR3 expression allows the host to properly maintain the optimal, least pathogenic response at this point. It will be important to examine how other innate molecules, such as TLR7, RIG-I, RNA-dependent protein kinase, etc., interact with TLR3 activation to allow the development of immune responses to RSV. These latter investigations will be important, because recent data have indicated the importance of TLR7 for clearance of RNA viruses (20, 54), and our TLR7 expression data indicated that in addition to TLR3, TLR7 was significantly up-regulated. There is probably a coordinated activation of multiple innate immune molecules that need to be activated for an appropriate and nonpathogenic immune response.

The authors have no financial conflict of interest.