Cystic Fibrosis (CF)
CF is an autosomal recessive disorder caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene [
109]. The airways of CF patients are characterized by chronic bacterial colonization and associated neutrophilic inflammation.
P. aeruginosa infection is the major cause of morbidity and mortality among CF-affected individuals, producing acute pneumonia or chronic lung disease with periodic acute exacerbations [
3,
110,
111]. A predisposition to chronic and progressive
P. aeruginosa infection occurs despite the finding that both CF and non-CF lung epithelial cells express functional TLRs that can mediate inflammatory responses to microbes. For example, in one study comparing human CFTE29o (trachea; homozygous for the delta F508 CFTR mutation) and 16HBE14o (bronchial non-CF) cells, comparable mRNA and surface protein expression of TLR1-5 and TLR9 was observed [
112]. TLR6 mRNA, but not protein, expression was observed in both cell lines; however, for unclear reasons only the CF line responded to TLR2/TLR6 agonist MALP-2 [
112]. Despite this similar TLR expression pattern, a more recent study showed increased inflammatory responses following stimulation with clinical
Pseudomonas isolates in a CF airway epithelial cell line (IB3-1) compared to a "CF-corrected" line stably expressing wild type CFTR [
113]. A detailed analysis showed that these responses were dependent on bacterial flagellin and TLR5 expression. Peripheral blood mononuclear cells from CF patients also responded more vigorously to stimulation with
P. aeruginosa and TLR ligands compared to healthy controls and expressed higher levels of TLR5 mRNA, suggesting that CFTR mutations modulate the host inflammatory response through undetermined mechanisms [
113]. In another study, a selective increase in TLR5 expression was found on airway, but not circulating, neutrophils from CF patients compared to patients with bronchiectasis and healthy control subjects [
38]. The functional relevance of neutrophil TLR5 expression was reflected by its correlation with lung function values in
P. aeruginosa-infected CF patients. Neutrophils also had increased flagellin dependent IL-8 secretion, phagocytosis, and respiratory burst activity that were attributed to chronic infection rather than as a primary consequence of mutant CFTR [
38]. TLR5-deficient mice showed impaired bacterial clearance, reduced airway neutrophil recruitment and MCP-1 production after low dose challenge with flagellated
P. aeruginosa that was not observed after challenge with an isotypic non-flagellated strain, confirming a specific contribution of TLR5-dependent pathways to the host inflammatory response [
114].
In addition to TLR5-dependent recognition of flagellin,
P. aeruginosa LPS is detected by TLR4 and the
P. aeruginosa ExoS toxin is recognized by both TLR2 and TLR4 [
11,
115‐
117]. Loss of a single TLR does not confer susceptibility to
P. aeruginosa infection while deletion of the adaptor molecule MyD88 does confer hypersusceptibility, increased lung bacterial load, and deficient neutrophil recruitment [
114,
117‐
123]. Interestingly, MyD88 may play an essential role only during the early phase of infection (4-8 hours) as inflammation and control of bacterial load 48 hours after low dose infection occurred through an undetermined MyD88-independent mechanism [
119]. Both TLR2 and TLR4 signal through MyD88-dependent and -independent pathways while TLR5 signals exclusively through MyD88. Studies to determine the relative contribution of TLR2, TLR4, and TLR5 have had conflicting results, possibly due to the complex pathogenesis of pseudomonal infection [
123‐
125].
Staphylococcus aureus and
Burkholderia cenocepacia have been associated with early and advanced CF lung disease, respectively [
3].
B. cenocepacia provokes lung epithelial damage and TNF-α secretion that leads to severe pneumonia and sepsis in CF patients [
126,
127]. Excess inflammation and mortality in
B. cenocepacia infection occurred through flagellin-dependent activation of TLR5 and MyD88 [
128,
129]. Another study showed that, despite higher bacterial load, MyD88-deficient mice had less inflammation and decreased mortality compared to wild type mice infected with
B. cenocepacia[
130].
Chronic Obstructive Pulmonary Disease (COPD)
COPD includes disorders of the respiratory system that are characterized by abnormal inflammation as well as expiratory airflow limitation that is not fully reversible. In humans, the main risk factor for COPD is smoking and the disease prevalence rises with age [
131]. Although the pathogenesis of COPD is not well understood, various aspects of lung innate immunity are impaired including mucociliary clearance, AM function, and expression of airway antimicrobial polypeptides [
132]. As a result, microbial pathogens frequently establish residence in the lower respiratory tract and induce a vicious circle of inflammation and infection that may contribute to progressive loss of lung function [
133] (figure
1).
There is accumulating evidence that impaired innate immunity is likely to contribute to the pathogenesis of COPD [
134]. An essential role for TLRs in the maintenance of lung structural homeostasis under ambient conditions was recently described [
135]. In this study, TLR4- and MyD88-deficient mice developed spontaneous age-related emphysema that was associated with increased oxidant stress, cell death, and elastolytic activity. A detailed mechanistic analysis showed that TLR4 maintains a critical oxidant/antioxidant balance in the lung by modulating the expression and activity of NADPH oxidase 3 in structural cells. In light of this finding, the free radicals and oxidant properties of tobacco smoke have been hypothesized to subvert innate immunity and cause lung cell necrosis and tissue damage [
136,
137]. Indeed, mice with short-term cigarette smoke exposure develop neutrophilic airway inflammation that is dependent on TLR4, MyD88, and IL-1R1 signaling [
138]. Consistent with these findings, C3H/HeJ mice that have naturally defective TLR4 signaling develop less chronic inflammation after 5 weeks of cigarette smoke exposure [
139]. Finally, long-term cigarette smoke exposure induced strain-dependent emphysema in mice in one study, although no specific association to TLRs was described [
140].
Several studies have evaluated TLR expression and function in AMs from COPD patients, smokers, and non-smokers. Using flow cytometry, one group showed reduced TLR2 expression on AMs of COPD patients and smokers compared to non-smokers following
ex vivo ligand stimulation. Upregulation of TLR2 mRNA and protein expression was observed only in AMs from non-smokers while no significant differences in TLR4 expression were demonstrated among these three groups [
141]. Another report showed comparable AM expression of TLR2, TLR4 or the co-receptors MD-2 or CD14 between smokers and non-smokers [
142], yet AM stimulation with TLR2 or TLR4 ligands elicited lower mRNA and protein expression of inflammatory cytokines (TNF-α, IL-1β, IL-6) or chemokines (IL-8, RANTES) in smokers that was associated with suppressed IRAK-1 and p38 phosphorylation and impaired NF-κB activation [
142]. From this data, the authors concluded that chronic LPS exposure via cigarette smoking selectively reprograms AMs and alters the inflammatory response to TLR2 and TLR4 ligands [
142]. Finally, another study showed reduced TLR4 mRNA expression in nasal and tracheal epithelial cells of smokers compared to healthy non-smoking control subjects with no differences in TLR2 expression [
143]. Relative to non-smokers, patients with mild or moderate COPD showed increased expression of TLR4 and HBD-2, a TLR4 inducible antimicrobial peptide, while those with advanced COPD had a reduction in TLR4 and HBD-2 expression [
143]. Modulation of TLR4 expression by cigarette smoke extract was studied
in vitro and revealed a dose-dependent reduction in TLR4 mRNA and protein expression as well as reduced IL-8 secretion in the A549 alveolar epithelial cells [
143]. Taken together, these findings point to dynamic regulation of airway epithelial and AM TLRs in response to diverse environmental stimuli. The differences in TLR expression across studies could be related to variable LPS content in tobacco smoke, bacterial colonization, or a persistent inflammatory state. Increased TLR4 expression in mild or moderate COPD may reflect a robust host response, while the decreased TLR4 expression level in association with severe COPD may reflect a loss of innate immunity or an adaptive regulatory response.
The interaction of cigarette smoke and PRR activation has been studied using mouse models. In one study, AMs from mice that had been exposed to cigarette smoke for eight weeks showed decreased cytokine (TNF-α, IL-6) and chemokine (RANTES) production following
in vitro stimulation with double-stranded RNA, LPS, or NLR agonists [
144]. No alteration of TLR3 or TLR4 expression was observed; however, there was decreased nuclear translocation of the transcription factor NF-κB. The functional impairment of cytokine release was specific to AMs and reversible after cessation of smoke exposure [
144]. A subsequent report found a synergistic interaction of cigarette smoke and dsRNA or influenza virus that leads to emphysema in mice through epithelial and endothelial cell apoptosis as well as proteolysis [
145]. This process was mediated by IL-12, IL-18, and IFN-γ as well as activation of antiviral response pathways including the intracellular signaling adaptor protein IPS-1 and the kinase PKR.
Defective innate immunity may predispose to acute exacerbations of COPD that are characterized by acutely worsening dyspnea, cough, sputum production, and accelerated airflow obstruction [
146]. Bacterial colonization (
Streptococcus pneumoniae, Haemophilus influenzae) or viral infection (Influenza A and B, Respiratory Syncytial Virus) of the lower respiratory tract are primary causes of acute COPD exacerbations [
146‐
152]. Virulent pneumococci express the toxin pneumolysin that is able to physically interact with TLR4 [
153‐
159]. Consistent with this finding, nasopharyngeal infection of TLR4-deficient mice with
S. pneumoniae causes enhanced bacterial load, dissemination, and death compared to wild type mice [
158]. Interestingly, the role of TLR4 seems to be specific to the nasopharynx as TLR4-deficient mice exhibit only a modest impairment of host defense following direct pneumococcal infection of the lower respiratory tract [
160]. TLR2 is also upregulated following pneumococcal infection and enhances host inflammatory responses [
161,
162]. Despite a modest reduction of inflammatory mediator production, TLR2-deficient mice can still clear high and low infectious doses of
S. pneumoniae, suggesting that another PRR compensates for the loss of TLR2 in this model [
160,
163]. TLR9-deficient mice are slightly more susceptible to pneumococcal infection compared to wild type animals [
164]. Conversely, abrogation of MyD88 signaling leads to uncontrolled airway pneumococcal growth, systemic bacterial dissemination and decreased immune mediator (TNF-α and IL-6) expression [
158,
165,
166]. The severe susceptibility phenotype of MyD88-deficient mice compared to mice with a single deletion of TLR9 or combined deletion of TLR2 and TLR4 highlights the crucial role of this downstream adaptor in host defense against
S. pneumoniae[
158,
160,
163,
164,
167].
Non-typeable
H. influenzae (NTHi) is another bacterium that commonly colonizes the respiratory epithelium and causes COPD exacerbations [
168‐
171]. While NTHi produces both TLR4 and TLR2 ligands, TLR4/MyD88 is the dominant immune signaling pathway
in vitro and mediates bacterial clearance
in vivo[
172]. TLR4 signaling in response to NTHi is entirely MyD88 dependent as TRIF KO mice had an identical bacterial load compared to wild type mice [
172]. TLR3 may also play a role in inflammatory mediator production in the immune response to NTHi although its relative contribution to bacterial clearance is not clear [
173].
Asthma
Asthma is a potentially life-threatening chronic inflammatory airway disease that is characterized by episodic bronchoconstriction, mucus hypersecretion, goblet cell hyperplasia and tissue remodelling that may begin in childhood. The underlying immune response in asthma is targeted against environmental antigens including pollen or dust particles and is characterized by the presence of antigen-specific Th2 cells in the lung that facilitate production of antigen specific IgE [
174,
175]. Viral and bacterial infections have been associated with induction or protection against asthma, suggesting that innate immunity plays an important role in disease pathogenesis [
176]. On the basis of several epidemiologic, human, and animal studies, the timing and extent of LPS exposure, and presumably TLR4 activation, appears to determine whether a protective Th1 response or a permissive Th2 response develops in the lung [
177]. For example, it was demonstrated that low dose administration of intranasal LPS induces a Th2 biased immune response in the lung whereas elsewhere in the body LPS is a strong inducer of a Th1 immune response [
178]. Nevertheless, experimental treatment of mice with microbes [
179] or TLR agonists [
180,
181] inhibits allergic sensitization, eosinophilic inflammation, and airways hyperresponsiveness. Recently, experimental intranasal infection of pregnant mice with
Acinetobacter lwoffii F78 was shown to confer protection against ovalbumin-induced asthma in the progeny. Using knockout mice, the protective effect was shown to be dependent on maternal TLR expression and suggests that microbial recognition during pregnancy somehow primes the fetal lung environment for a Th1 response later in life [
182].
Lung resident cells that express TLR4 also play an important role in the induction of allergen specific Th2 cells via recognition of house dust mite (a ubiquitous indoor allergen) that leads to the production of thymic stromal lymphopoietin, granulocyte-macrophage colony-stimulating factor, IL-25 and IL-33. This cytokine milieu can bias lung DCs towards a Th2 activating phenotype that drives the polarization of naïve lymphocytes [
183]. In addition, eosinophil derived neurotoxin can induce TLR2-dependent DC maturation, leading to Th2 polarization by secretion of high amounts of IL-6 and IL-10 [
184] while basophils may also instruct T cells to become Th2 cells [
185].
TLRs have been shown via genetic association studies as well as single and multiple gene knockout studies to play a role in the development of allergic asthma. For example TLR7 and TLR8 are associated with human asthma [
186] while ligands of TLR7 and TLR8 can prevent airway remodeling caused by experimentally induced asthma [
187,
188]. TLR10 single nucleotide polymorphisms have also been associated with asthma in two independent samples [
189] although the ligand for TLR10 has not been defined. Finally, in a multi-centre asthma study, TLR4 and TLR9 were both associated with wheezing and TLR4 was also associated with allergen specific IgE secretion [
190]. Based on this observation, TLR9 ligands are currently in clinical trials for the treatment or prevention of asthma [
191].
Asthma can be further exacerbated by bacterial respiratory tract infection including
Mycoplasma pneumoniae or
Chlamydophila pneumoniae[
192]. In one study, 50% of patients suffering from their first asthmatic episode were infected with
M. pneumoniae while 10% were serologically positive for acute
C. pneumoniae infection [
193,
194]. MyD88-deficient mice infected with
C. pneumoniae failed to upregulate cytokine and chemokine expression, had delayed CD8
+ and CD4
+ T cell recruitment, and could not clear the bacterium from the lungs leading to severe chronic infection and significantly increased mortality [
195]. At a later stage of infection, IL-1β, IFN-γ and other inflammatory mediators may be upregulated via a MyD88-independent pathway but are not sufficient to prevent mortality from
C. pneumoniae[
195]. TLR2 and TLR4-deficient mice can recover from
C. pneumoniae infection with no impairment of bacterial clearance suggesting that other PRRs are also involved in host defense or that TLR2/TLR4 act in concert during
C. pneumoniae infection [
195,
196].
TLR2 is also upregulated in response to
M. pneumoniae infection, leading to increased expression of airway mucin [
197,
198]. Allergic inflammation along with the induction of Th2 cytokines (IL-4, IL-13) leads to TLR2 inhibition during
M. pneumoniae infection, thereby decreasing the production of IL-6 and other Th1 proinflammatory mediators that are required for bacterial clearance [
199]. Antibiotic treatment of asthmatic patients infected with
M. pneumoniae improves their pulmonary function and highlights the increasingly important role that bacterial colonization and interactions with the host innate immune response play in asthma exacerbations and mortality [
200,
201].
Viral infection of the lower respiratory tract can also contribute to asthma development and exacerbations. Respiratory Syncytial Virus (RSV) is a particularly important cause of acute bronchiolitis and wheezing in children that may lead to the subsequent development of asthma [
202‐
206]. Wheezing after the acquisition of severe RSV infection early in life has been associated with elevated Th2 responses, eosinophilia, and IL-10 production [
207‐
211]. During RSV infection, the viral G protein mediates attachment to lung epithelial cells and the F protein leads to the fusion of the viral envelope with the host cell plasma membrane [
212]. In response to RSV infection, TLRs are broadly upregulated in the human tracheal epithelial cell line 9HTEo [
213]. In mice, TLR4 has been shown to recognize the F protein and activate NF-κB during RSV infection [
203,
214]. Accordingly, TLR4-deficient animals exhibit impaired NK cell function and increased viral load [
205,
215]. Defective TLR4 signalling has also been linked to increased pathology in a study of pre-term infants [
216]. An essential role for IL-12, rather than TLR4, in susceptibility to RSV has also been proposed [
214]; however, significant differences in experimental design limit the comparison of these apparently discordant studies [
217].
In human lung fibroblasts and epithelial cells, the formation of dsRNA during RSV replication can activate TLR3-mediated immune signaling, leading to the upregulation of the chemokines RANTES and IP-10 [
218]. TLR3-deficient mice have a predominantly Th2 response to RSV characterized by increased airway eosinophilia, mucus hypersecretion and expression of IL-5 and IL-13 [
219]. RIG-I-induced IFN-β expression during RSV infection was recently shown to trigger TLR3 activation, suggesting that TLR3 mediates a secondary immune signaling pathway [
220]. Interestingly, while TLR3 is involved in chemokine expression it has no role in RSV viral clearance, which is primarily mediated by the TLR2/TLR6 heterodimer [
218,
219].
In summary, the emerging picture of allergic asthma suggests that the disease can be mediated or exacerbated in genetically predisposed individuals by infection. In some cases these infections may induce an inflammatory state that protects against asthma, while in others the infection may elicit an acute allergic response or bias the host towards a subsequent Th2 response (figure
1).