TLRs are important for the recognition of conserved motifs expressed by invading bacteria. TLR4 is the signaling receptor for LPS, the major proinflammatory component of the Gram-negative cell wall, whereas CD14 serves as the ligand-binding part of the LPS receptor complex. Triggering of TLR4 results in the activation of two distinct intracellular pathways, one that relies on the common TLR adaptor MyD88 and one that is mediated by Toll/IL-1R domain-containing adaptor-inducing IFN-β (TRIF). Nontypeable Haemophilus influenzae (NTHi) is a common Gram-negative respiratory pathogen that expresses both TLR4 (LPS and lipooligosaccharide) and TLR2 (lipoproteins) ligands. To determine the roles of CD14, TLR4, and TLR2 during NTHi pneumonia, the following studies were performed: 1) Alveolar macrophages from CD14 and TLR4 knockout (KO) mice were virtually unresponsive to NTHi in vitro, whereas TLR2 KO macrophages displayed a reduced NTHi responsiveness. 2) After intranasal infection with NTHi, CD14 and TLR4 KO mice showed an attenuated early inflammatory response in their lungs, which was associated with a strongly reduced clearance of NTHi from the respiratory tract; in contrast, in TLR2 KO mice, lung inflammation was unchanged, and the number of NTHi CFU was only modestly increased at the end of the 10-day observation period. 3) MyD88 KO, but not TRIF mutant mice showed an increased bacterial load in their lungs upon infection with NTHi. These data suggest that the MyD88-dependent pathway of TLR4 is important for an effective innate immune response to respiratory tract infection caused by NTHi.

The innate immune system is of crucial importance for the initiation of an efficient immune response against invading pathogens. To start an appropriate inflammatory response leading to the elimination of microorganisms, pathogens have to be distinguished from self. In recent years, a novel family of receptors has been identified as the key receptors recognizing pathogens: TLRs (for review, see Refs.1 and 2).

TLRs are pattern recognition receptors that can sense conserved microbial patterns that are shared by groups of pathogens. TLR4, the receptor for LPS, is important for the recognition of Gram-negative bacteria, whereas cell wall components of Gram-positive bacteria (such as lipoteichoic acid and lipoproteins) are primarily recognized by TLR2. After recognition of the bacterial pathogen-associated molecular pattern, distinctive intracellular pathways are activated, which include several different adaptor molecules, such as MyD88, Mal (MyD88 adaptor-like, also known as TIRAP) (3, 4), Toll/IL-1R domain-containing adaptor-inducing IFN-β (TRIF;3 also known as TICAM-1) (5, 6) and TRIF adaptor-related adaptor molecule (TRAM, also known as TICAM-2) (7, 8, 9). TLR4 uses two different pathways. Until recently, it was believed that TLR4 only signals via the adaptor proteins MyD88 and Mal/TIRAP, leading directly to NF-κB activation (MyD88-dependent pathway). Recently, a second, MyD88-independent, signaling route for TLR4 was discovered involving the adaptor proteins TRIF and TRAM. This MyD88-independent pathway initiates the type 1 IFN response as well as late NF-κB activation (10, 11, 12, 13).

Nontypeable Haemophilus influenzae (NTHi) is a Gram-negative bacterium that is a commensal organism in the human respiratory tract and an important cause of localized infections such as middle ear infection, sinusitis, and conjunctivitis. Furthermore, this bacterium is a common cause of exacerbations in asthma or chronic obstructive pulmonary disease, and NTHi has been implicated as a frequent cause of community-acquired pneumonia (14, 15, 16, 17). NTHi abundantly expresses lipooligosaccharides (LOS) as well as LPS on the bacterial cell wall. Like LPS, LOS contains a lipid A portion, and this is the portion recognized by TLR4. In line, Wang et al. (18) recently reported that C3H/HeJ mice, which have a mutated nonfunctional TLR4, are less able to mount an inflammatory response to an encapsulated type b H. influenzae in their lungs accompanied by impaired bacterial clearance. Furthermore, recent research demonstrated that capsule porins and bacterial lipoproteins from H. influenzae type B signal via TLR2 (19). In contrast to other H. influenzae isolates, NTHi lacks a polysaccharide capsule (nontypeable). Nevertheless, NTHi has several TLR2 ligands on its surface, including the bacterial lipoproteins protein D, P2, or P6, a 16-kDa lipoprotein highly conserved in the outer membrane of all NTHi strains (20).

In the present study we sought to determine the contributions of CD14, TLR2, and TLR4 to the innate immune response to NTHi in the lung. In addition, after we had identified TLR4 as the main receptor contributing to an effective clearance of NTHi, we examined the relative roles of MyD88-dependent and MyD88-independent TLR signaling by making use of MyD88 knockout (KO) and TRIF mutant mice.

Pathogen-free, 8- to 10-wk-old, wild-type (WT) C57BL/6 mice were purchased from Harlan Sprague Dawley. TLR4, MyD88, and TLR2 KO mice, backcrossed six times to a C57BL/6 background, were generated as described previously (21, 22, 23). CD14 KO mice, backcrossed six times to a C57BL/6 background, were obtained from The Jackson Laboratory. Mice deficient in TRIF were obtained by inducing random germline mutations in C57BL/6 mice using by N-ethyl-N-nitrosourea (13). Age- and sex-matched animals were used in all experiments. The animal care and use of committee of University of Amsterdam approved all experiments.

AM were harvested from CD14, TLR4, TLR2, and MyD88 KO, TRIF mutant, and WT mice by bronchoalveolar lavage (n = 3–7/strain). The trachea was exposed through a midline incision and cannulated with a sterile 22-gauge Abbocath-T catheter (Abbott Laboratories). Bronchoalveolar lavage was performed by instilling three 0.5-ml aliquots of sterile saline. Total cell numbers were counted from each sample using a hemocytometer. Cells were resuspended in RPMI 1640 containing 1 mM pyruvate, 2 mM l-glutamine, penicillin, streptomycin, and 10% FCS in a final concentration of 1 × 104 cells/100 μl. Cells were then cultured in 96-well microtiter plates (Greiner) for 2 h and washed with RPMI 1640 to remove nonadherent cells. Adherent monolayer cells were stimulated with LPS (from Escherichia coli O55:B5; 200 ng/ml; Sigma-Aldrich), heat-killed NTHi strain 12 (1 × 105 CFU/ml), or RPMI 1640 for 16 h. Supernatants were collected and stored at −20°C until assayed for TNF and KC.

H. influenzae strain 12 (donated by S. J. Barenkamp, St. Louis University School of Medicine, St. Louis, MO) is a clinical isolate that has been used by our and other laboratories in investigations on murine pneumonia (24, 25, 26). The strain was classified as nontypeable based on the absence of agglutination with typing antisera for H. influenzae types a–f (Burroughs Wellcome) and the failure to hybridize with pUO38, a plasmid that contains the entire cap b locus. The NTHi strain was stored at −80°C in brain heart infusion broth with 20% glycerol. For preparation of the inoculum, bacteria were streaked from frozen aliquots onto a chocolate agar plate and incubated overnight at 37°C in a 5% CO2 incubator. Next, bacteria obtained from the chocolate agar plate were grown for ∼3 h to midlogarithmic phase in brain heart infusion broth supplemented with 10 μg/ml hemin and 3.5 μg/ml NAD at 37°C (all reagents were from Difco). Bacteria were harvested by centrifugation at 1500 × g for 15 min, washed, and resuspended in sterile isotonic saline at a concentration of 1 × 107 CFU/50 μl, as determined by plating serial 10-fold dilutions on chocolate agar plates. Pneumonia was induced by intranasal (i.n.) inoculation of 50 μl (107 CFU) of the bacterial suspension as described previously (24). For this procedure mice were lightly anesthetized by inhalation of isofluorane (Upjohn).

Six hours, 24 h, 44 h, and 10 days after infection, mice were anesthetized with Hypnorm (Janssen Pharmaceutica; active ingredients: fentanyl citrate and fluanisone) and midazolam (Roche) and were killed by bleeding from the vena cava inferior. The lungs were harvested and homogenized at 4°C in 4 vol of sterile saline using a tissue homogenizer (BioSpec Products). CFUs were determined from serial dilutions of lung homogenates and blood, plated on chocolate agar plates, and incubated at 37°C at 5% CO2 for 20 h before colonies were counted.

For cytokine measurements, lung homogenates were diluted 1/2 in lysis buffer containing 300 mM NaCl, 30 mM Tris, 2 mM MgCl2, 2 mM CaCl2, 1% Triton X-100, and pepstatin A, leupeptin and aprotinin (all 20 ng/ml (pH 7.4)) and incubated at 4°C for 30 min. Homogenates were centrifuged at 1500 × g at 4°C for 15 min, and supernatants were stored at −20°C until assays were performed.

For detection of IL-1α, IL-1β, IL-6, KC, and MIP-1α, a Bio-Plex Cytokine Array was used (27) (Bio-Rad; detection limit, 7.8 pg/ml). TNF and MPO were measured using specific ELISAs (for TNF from R&D Systems; for MPO from HyCult Biotechnology). The detection limits were 31.25 pg/ml for TNF and 7.8 pg/ml for MPO.

Lungs for histology were harvested after infection, fixed in 10% formalin, and embedded in paraffin. Four-micron sections were stained with H&E and analyzed by a pathologist who was blinded for groups. To score lung inflammation and damage, the entire lung surface was analyzed with respect to the following parameters (24, 28, 29): interstitial inflammation, edema, endothelialitis, bronchitis, and pleuritis. Each parameter was graded on a scale of 0–4: 0, absent; 1, mild; 2, moderate; 3, severe; and 4, very severe. The total lung inflammation score was expressed as the sum of the scores for each parameter; the maximum was 20.

All data are expressed as the mean ± SEM. Serial data were analyzed by one-way ANOVA using GraphPad PRISM version 4.00. When comparing two groups at multiple time points, two-way ANOVA was used. If appropriate, ANOVAs were followed by Bonferroni post test. Statistical analyses of bacterial counts were performed after log transformation. A value of p < 0.05 was considered statistically significant.

To gain insight into the requirement of TLR signaling upon the first encounter between bacterium and AM, we tested the capacity of isolated AM to release TNF and KC upon stimulation with LPS (positive control) or heat-killed NTHi (E:T cell ratio, 1:10). As expected, CD14 and TLR4 KO AM failed to respond to LPS, whereas the LPS responsiveness of TLR2 KO AM was indistinguishable from that of WT AM (data not shown). AM from CD14 and TLR4 KO mice also produced significantly less TNF and KC in response to NTHi, whereas WT AM released considerable amounts of TNF and KC (Fig. 1). In addition, TLR2 KO AM released significantly less TNF and KC upon stimulation with NTHi than WT AM. These data suggest that the responsiveness of AM to NTHi depends on CD14, TLR4, and, to a lesser extent, TLR2.

FIGURE 1.

Membrane CD14, TLR4 and TLR2 are required for responsiveness of AM to NTHi. Freshly isolated AM of WT, CD14, TLR4, and TLR2 KO mice (n = 7/group) were incubated with RPMI 1640 (control) or heat-killed NTHi (equivalent 105 CFU; T:E ratio, 1:10) for 16 h before TNF and KC production was measured. ∗∗∗, p < 0.01; ∗∗, p < 0.01 (KO vs WT).

FIGURE 1.

Membrane CD14, TLR4 and TLR2 are required for responsiveness of AM to NTHi. Freshly isolated AM of WT, CD14, TLR4, and TLR2 KO mice (n = 7/group) were incubated with RPMI 1640 (control) or heat-killed NTHi (equivalent 105 CFU; T:E ratio, 1:10) for 16 h before TNF and KC production was measured. ∗∗∗, p < 0.01; ∗∗, p < 0.01 (KO vs WT).

Close modal

Having demonstrated that CD14, TLR4 and TLR2 mediate the responsiveness of AM to NTHi in vitro, we determined whether the lack of these receptors alters host defense after infection with live NTHi in vivo. Therefore, we infected WT, CD14 KO, TLR4 KO, and TLR2 KO mice with NTHi and killed the mice after 6, 24, and 44 h as well as after 10 days. Bacterial counts were similar in the different mouse strains after 6 h of infection (Fig. 2). Although all strains were able to partially clear the infection during the following days, CD14 and TLR4 KO mice in particular had substantially higher bacterial loads from 24 h onward. By 24 h after infection, bacterial counts were >600-fold higher in CD14 and TLR4 KO mice compared with WT mice or TLR2 KO mice. Remarkably, 10 days after infection, TLR2 KO mice also displayed higher NTHi loads in their lungs than WT mice. At this time point, bacterial clearance was so efficient in WT mice that only two of seven animals still had detectable bacteria in their lungs (detection limit, 100 CFU/g lung). In contrast, all TLR4 and CD14 KO mice and six of seven TLR2 KO mice had detectable numbers of bacteria in their lungs. Hence, these data suggest that CD14, TLR4, and, to a lesser extent, TLR2 contribute to effective clearance of NTHi from the respiratory tract.

FIGURE 2.

Bacterial clearance: WT (♦, •, and ▴) and TLR4, CD14, and TLR2 KO mice (⋄, ○, and ▵) were infected i.n. with 1 × 107 CFU of NTHi. Six, 24, and 44 h as well as 10 days after infection, mice were killed, and bacterial loads were determined in lung homogenates. Data are the mean ± SEM of five to eight mice per group. The p value is derived from curve comparison: ∗∗, p < 0.01; ∗∗∗, p < 0.001 (KO vs WT at the indicated time point).

FIGURE 2.

Bacterial clearance: WT (♦, •, and ▴) and TLR4, CD14, and TLR2 KO mice (⋄, ○, and ▵) were infected i.n. with 1 × 107 CFU of NTHi. Six, 24, and 44 h as well as 10 days after infection, mice were killed, and bacterial loads were determined in lung homogenates. Data are the mean ± SEM of five to eight mice per group. The p value is derived from curve comparison: ∗∗, p < 0.01; ∗∗∗, p < 0.001 (KO vs WT at the indicated time point).

Close modal

The success of combating pulmonary infections strongly depends on the efficacy of the local inflammatory response elicited (30, 31). To study the extent and kinetics of the inflammatory response, we killed WT, CD14, TLR4, and TLR2 KO mice at multiple time points after infection and measured the concentrations of proinflammatory cytokines (IL-1α, IL-1β, TNF, and IL-6) and chemokines (KC and MIP-1α) in lung homogenates (Fig. 3). After 6 h of infection, when the bacterial load was still similar in all mouse strains, the pulmonary levels of all inflammatory mediators were strongly reduced, especially in TLR4 and to a lesser extent in CD14 KO mice. Interestingly, from 24 h onward, CD14 KO mice showed much higher levels of IL-1α, IL-6, and KC in their lungs than WT mice. This response in CD14 KO mice contrasted with the response observed in TLR4 KO mice, in which lung cytokine and chemokine concentrations remained low throughout the entire observation period (with the exception of IL-1β, which showed a late increase in TLR4 KO mice). In sharp contrast with the altered inflammatory response in lungs of CD14 and TLR4 KO mice, the pulmonary levels of cytokines and chemokines in TLR2 KO mice were largely similar to those in WT mice; only KC levels were modestly reduced in TLR2 KO mice 6 and 48 h after infection.

FIGURE 3.

Cytokine and chemokine responses: WT mice (A–R; •), CD14 KO mice (A–F; ○), TLR4 KO mice (G–L; ○), and TLR2 KO mice (M–R) were infected i.n. with 1 × 107 CFU of NTHi. Six, 24, and 44 h after infection, mice were killed, lungs were removed, and IL-1β, IL-1α, IL-6, TNF, KC, and MIP-1α were measured in lung homogenates. Data are the mean ± SEM of five to eight mice per group per time point. The p value is derived from curve comparison: ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001 (KO vs WT at the indicated time point).

FIGURE 3.

Cytokine and chemokine responses: WT mice (A–R; •), CD14 KO mice (A–F; ○), TLR4 KO mice (G–L; ○), and TLR2 KO mice (M–R) were infected i.n. with 1 × 107 CFU of NTHi. Six, 24, and 44 h after infection, mice were killed, lungs were removed, and IL-1β, IL-1α, IL-6, TNF, KC, and MIP-1α were measured in lung homogenates. Data are the mean ± SEM of five to eight mice per group per time point. The p value is derived from curve comparison: ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001 (KO vs WT at the indicated time point).

Close modal

To obtain additional insight into the involvement of CD14, TLR4, and TLR2 in the inflammatory response in the lung upon infection of the airways with NTHi, we semiquantitatively scored lung histology slides generated from WT, CD14, TLR4, and TLR2 KO mice at various time points after infection (Table I and Fig. 4). In line with their attenuated initial cytokine/chemokine response, both TLR4 (Fig. 4,B) and CD14 KO (Fig. 4,C) mice displayed reduced lung inflammation 6 h after infection compared with WT (Fig. 4,A) mice. At later time points, CD14 KO mice in particular (Fig. 4,F) demonstrated more lung inflammation than WT mice (Fig. 4,D). In accordance, lung MPO levels, reflecting the number of neutrophils present in this organ, were lower in CD14 and TLR4 KO mice than in WT mice 6 h after infection, whereas they were higher in these genetically modified animals at 44 h (Table II). In contrast with the altered lung inflammation in CD14 and TLR4 KO mice, pathology scores (Table I) and lung MPO levels (Table II) in TLR2 KO mice were not different from those in WT mice, although in line with a reduction in KC levels at 6 h, MPO content was lower in TLR2 KO mice early in infection.

Table I.

Inflammation scoresa

WTCD14 KOTLR4 KOTLR2 KO
6 h 9.6 ± 0.6 5.5 ± 0.9c 4.9 ± 0.3c 7.1 ± 0.7 
24 h 9.8 ± 0.6 7.2 ± 0.4 9.9 ± 0.9 8.1 ± 0.7 
44 h 8.9 ± 0.7 12.8 ± 1.3c 10.6 ± 0.7 9.6 ± 1.1 
10 days 1.5 ± 0.3 8.1 ± 0.5c 4.4 ± 0.5c 3.1 ± 0.3b 
WTCD14 KOTLR4 KOTLR2 KO
6 h 9.6 ± 0.6 5.5 ± 0.9c 4.9 ± 0.3c 7.1 ± 0.7 
24 h 9.8 ± 0.6 7.2 ± 0.4 9.9 ± 0.9 8.1 ± 0.7 
44 h 8.9 ± 0.7 12.8 ± 1.3c 10.6 ± 0.7 9.6 ± 1.1 
10 days 1.5 ± 0.3 8.1 ± 0.5c 4.4 ± 0.5c 3.1 ± 0.3b 
a

Groups of five to nine mice were i.n. inoculated with 1 × 107 CFU of NTHi. After 6, 24, and 44 h as well as 10 days, inflammation was scored on H&E-stained slides (for details, see Materials and Methods). Data are expressed as mean ± SEM.

b

, p < 0.05;

c

, p < 0.01 vs WT control.

FIGURE 4.

Histopathology: representative lung histology of WT (A and D), TLR4 KO (B and E), and CD14 KO (C and F) mice 6 h (A–C) and 44 h (D–F) after i.n. infection with 1 × 107 CFU of NTHi, showing reduced inflammation, pleuritis, bronchitis, and endothelialitis in TLR4 and CD14 KO mice compared with WT mice early in infection. Nevertheless, although not very pronounced, the same parameters were increased in TLR4 and CD14 KO mice 44 h after infection. The lung sections are representative of at least five animals per group per time point. G, Histology of an uninfected mouse lung. H&E staining was used. Magnification, ×10; inset, ×40.

FIGURE 4.

Histopathology: representative lung histology of WT (A and D), TLR4 KO (B and E), and CD14 KO (C and F) mice 6 h (A–C) and 44 h (D–F) after i.n. infection with 1 × 107 CFU of NTHi, showing reduced inflammation, pleuritis, bronchitis, and endothelialitis in TLR4 and CD14 KO mice compared with WT mice early in infection. Nevertheless, although not very pronounced, the same parameters were increased in TLR4 and CD14 KO mice 44 h after infection. The lung sections are representative of at least five animals per group per time point. G, Histology of an uninfected mouse lung. H&E staining was used. Magnification, ×10; inset, ×40.

Close modal
Table II.

MPO contenta

WTCD14 KOTLR4 KOTLR2 KO
6 h 6.9 ± 1.1 3.4 ± 0.2b 4.9 ± 0.3c 3.1 ± 0.3c 
44 h 7.7 ± 0.9 11.3 ± 1.2b 10.1 ± 1.2b 6.8 ± 0.9 
WTCD14 KOTLR4 KOTLR2 KO
6 h 6.9 ± 1.1 3.4 ± 0.2b 4.9 ± 0.3c 3.1 ± 0.3c 
44 h 7.7 ± 0.9 11.3 ± 1.2b 10.1 ± 1.2b 6.8 ± 0.9 
a

Groups of five to nine mice were i.n. inoculated with 1 × 107 CFU of NTHi. After 6 and 44 h, mice were sacrificed, lungs were removed, homogenized and MPO content (μg/ml) was measured. Data are expressed as mean ± SEM.

b

, p < 0.05;

c

, p < 0.01 vs WT control.

After having established that a functional CD14/TLR4 complex is important for the initiation of a local inflammatory response and subsequent clearance of NTHi, we were interested in the relative importance of MyD88 and TRIF in host defense against NTHi. At first we studied whether AM from TRIF mutant and MyD88 KO mice were able to respond to NTHi in vitro (Fig. 5). AM from MyD88 KO failed to respond to NTHi and did not produce significant amounts of TNF and KC. Although AM from TRIF mutant mice produced significantly reduced levels of KC after stimulation with NTHi, no significant difference was detected in TNF production. We then inoculated WT, TLR4 KO, TRIF mutant, and MyD88 KO mice with NTHi and determined pulmonary bacterial loads 44 h after infection, i.e., the time point at which the difference between WT and TLR4 KO mice was most pronounced in our first experiments. Of considerable interest, the lungs of TLR4 and MyD88 KO mice contained much higher NTHi numbers than the lungs of WT mice, differences that were 3–4 logs, on the average (Fig. 6). In contrast, the bacterial loads harvested from the lungs of TRIF mutant mice were similar to those obtained from WT mice. In line with the studies described above, TLR4 and MyD88 KO, but not TRIF mutant, mice displayed higher cytokine, chemokine, and MPO levels in lung homogenates obtained 44 h after infection compared with WT mice (Table III). Moreover, semiquantitative inflammation scores were higher in MyD88 KO mice, but the difference for WT mice did not reach statistical significance (Table III and Fig. 7).

FIGURE 5.

TRIF and MyD88 are required for responsiveness of AM to NTHi. Freshly isolated AM of WT, TRIF mutant, and MyD88 KO mice (n = 3–6/group) were incubated with RPMI 1640 (control) or heat-killed NTHi (equivalent 105 CFU; T:E ratio, 1:10) for 16 h before TNF and KC production was measured. ∗∗, p < 0.01 vs WT.

FIGURE 5.

TRIF and MyD88 are required for responsiveness of AM to NTHi. Freshly isolated AM of WT, TRIF mutant, and MyD88 KO mice (n = 3–6/group) were incubated with RPMI 1640 (control) or heat-killed NTHi (equivalent 105 CFU; T:E ratio, 1:10) for 16 h before TNF and KC production was measured. ∗∗, p < 0.01 vs WT.

Close modal
FIGURE 6.

Roles of MyD88 and TRIF in bacterial clearance. WT and TLR4-, TRIF-, and MyD88-deficient mice were infected i.n. with 1 × 107 CFU of NTHi. Forty-four hours after infection, mice were killed, and bacterial loads were determined in lung homogenates. Data are the mean ± SEM of five or six mice per group. ∗∗, p < 0.01 (KO vs WT).

FIGURE 6.

Roles of MyD88 and TRIF in bacterial clearance. WT and TLR4-, TRIF-, and MyD88-deficient mice were infected i.n. with 1 × 107 CFU of NTHi. Forty-four hours after infection, mice were killed, and bacterial loads were determined in lung homogenates. Data are the mean ± SEM of five or six mice per group. ∗∗, p < 0.01 (KO vs WT).

Close modal
Table III.

Role of TRIF and MyD88 in NTHi induced lung inflammationa

WTTLR4 KOTRIF MutantMyD88 KO
IL-1β 0.5 ± 0.1 2.5 ± 0.9c 0.8 ± 0.4 2.8 ± 0.4c 
IL-1α 0.3 ± 0.1 1.4 ± 0.5b 0.5 ± 0.1 1.6 ± 0.9b 
IL-6 0.4 ± 0.1 0.5 ± 0.1 0.1 ± 0.1 0.9 ± 0.1b 
TNF 0.2 ± 0.05 0.3 ± 0.1 0.3 ± 0.05 0.9 ± 0.1c 
KC 0.07 ± 0.03 0.1 ± 0.02 0.1 ± 0.02 0.6 ± 0.2 
MIP-1α 1.4 ± 0.2 4.2 ± 0.7 2.3 ± 0.4 5.6 ± 1.0b 
MPO 4.3 ± 1.1 9.4 ± 1.2b 8.0 ± 1.1 10.2 ± 1.1b 
Inflammation scores 9.5 ± 1.2 9.8 ± 1.2 9.4 ± 0.8 11.1 ± 0.7 
WTTLR4 KOTRIF MutantMyD88 KO
IL-1β 0.5 ± 0.1 2.5 ± 0.9c 0.8 ± 0.4 2.8 ± 0.4c 
IL-1α 0.3 ± 0.1 1.4 ± 0.5b 0.5 ± 0.1 1.6 ± 0.9b 
IL-6 0.4 ± 0.1 0.5 ± 0.1 0.1 ± 0.1 0.9 ± 0.1b 
TNF 0.2 ± 0.05 0.3 ± 0.1 0.3 ± 0.05 0.9 ± 0.1c 
KC 0.07 ± 0.03 0.1 ± 0.02 0.1 ± 0.02 0.6 ± 0.2 
MIP-1α 1.4 ± 0.2 4.2 ± 0.7 2.3 ± 0.4 5.6 ± 1.0b 
MPO 4.3 ± 1.1 9.4 ± 1.2b 8.0 ± 1.1 10.2 ± 1.1b 
Inflammation scores 9.5 ± 1.2 9.8 ± 1.2 9.4 ± 0.8 11.1 ± 0.7 
a

Groups of five to seven mice were i.n. inoculated with 1 × 107 CFU of NTHi. After 44 h, mice were sacrificed, lungs were removed, homogenized and cytokines, chemokines (both in ng/ml) and MPO content (μg/ml) were measured. Inflammation was scored from H&E-stained slides (for details, see Materials and Methods). Data are expressed as mean ± SEM.

b

, p < 0.05;

c

, p < 0.01 vs WT control.

FIGURE 7.

Histopathology: representative lung histology of WT (A), TRIF mutant (B), and MyD88 KO (C) mice 44 h after i.n. infection with 1 × 107 CFU of NTHi. The lung sections are representative of at least five animals per group per time point. H&E staining was used. Magnification, ×10; inset, ×40.

FIGURE 7.

Histopathology: representative lung histology of WT (A), TRIF mutant (B), and MyD88 KO (C) mice 44 h after i.n. infection with 1 × 107 CFU of NTHi. The lung sections are representative of at least five animals per group per time point. H&E staining was used. Magnification, ×10; inset, ×40.

Close modal

NTHi is a commensal Gram-negative bacterium of the human nasopharynx that is able to colonize the epithelium of the upper and lower respiratory tract in patients with underlying lung diseases such as chronic obstructive pulmonary disease, chronic bronchitis, or cystic fibrosis (14, 15, 16, 17). NTHi is recognized as an important cause of infection of the lower airways and one of the most common causes of community-acquired bacterial pneumonia in adult populations (32, 33). Therefore, it is of importance to obtain insight into the innate immune response to NTHi respiratory tract infection.

An earlier study by Wang et al. (18) reported that TLR4 mutant C3H/HeJ and C57BL/10ScSn mice were more susceptible to a type b strain of H. influenzae. These TLR4-deficient mice demonstrated reduced pulmonary cytokine and chemokine levels as well as reduced infiltration of neutrophils into the lungs early in infection, which was associated with a delay in clearance of bacteria. We extend these data using a different (nonencapsulated) strain of H. influenzae relevant for human disease and using a genetically “cleaner” mouse model (i.e., TLR4 KO mice backcrossed to a C57BL/6 background). To substantiate and further extend the role of different TLRs during NTHi infection and to elucidate the relative contributions of the MyD88-dependent and MyD88-independent signaling pathways, we included CD14, TLR2, and MyD88 KO as well as TRIF mutant mice in our investigations.

Abundant evidence indicates that TLR4 is the signal-transducing element of the LPS receptor complex (1, 2). In this complex, CD14 functions as the ligand-binding component that facilitates LPS-induced TLR4 activation. NTHi expresses both LPS and LOS, which, like LPS, has been found to activate cells via TLR4 (34, 35). In our initial in vitro studies using AM harvested from CD14 and TLR4 KO mice, we established that both components of the LPS receptor complex are indispensable for activation of AM by NTHi. These in vitro data were in part confirmed in the in vivo model of NTHi pneumonia. In particular, TLR4 KO mice displayed a strongly reduced inflammatory response in the lung, especially early after infection, indicating that TLR4 is of crucial importance for mounting an innate immune response to NTHi in the airways. CD14 KO mice also showed less inflammation than WT mice early after infection, albeit to a lesser extent than TLR4 KO mice. A remarkable difference between CD14 KO and TLR4 KO mice became apparent at later time points, when CD14 KO mice demonstrated increased cytokine and chemokine levels in their lungs, whereas in TLR4 KO mice, the pulmonary concentrations of these mediators remained low. It is likely that CD14 KO mice were able to generate an enhanced cytokine response late during the infection via TLR4, TLR2, or other patter recognition receptors due to the higher bacterial load (providing a more potent inflammatory stimulus). The fact that such a late rise in cytokine levels was not observed in TLR4 KO mice suggests that TLR4 is pivotal for this response. In any case, the results obtained in CD14 KO and TLR4 KO mice suggest that the early inflammatory response to NTHi infection is crucial for the subsequent clearance of these bacteria from the respiratory tract; in both mouse strains, NTHi pulmonary loads were several logs higher than in WT mice from 24 h onward.

In vitro studies have identified several TLR2 ligands expressed by NTHi, including lipoproteins D, P2, and P6 (20). In this study we demonstrate that AM respond to NTHi in part via TLR2, although clearly CD14 and TLR4 play a greater role. It is likely that the response of WT AM is mostly dependent on LPS/LOS present in NTHi, and TLR2 ligands enhance this response. LPS/LOS was able to activate TLR2 KO AM, but without the additional effect of TLR2 ligands, such as lipoprotein D, P2, and P6, TLR2 KO AM responded less well in vitro compared with WT AM. In contrast to our in vitro findings using purified AM, TLR2 KO mice displayed an unaltered inflammatory response to NTHi infection in vivo. A possible explanation for this discrepancy is that in intact animals a variety of factors influence the host response, e.g., natural Abs or complement, which might compensate for the lack of TLR2. In addition, the presence of TLR4 may be able to compensate TLR2 deficiency in vivo, and respiratory epithelial cells also produce proinflammatory mediators that may compensate for a reduced response of AM. Interestingly, 10 days after infection, the lungs of TLR2 KO mice contained ∼1 log more NTHi than those of WT mice, suggesting that TLR2 deficiency does have a modest impact on bacterial clearance in this model despite the fact that the innate immune response was not impaired to an extent that could be detected by our methods. Additional research is warranted about how TLR2 influences the late clearance of NTHi from the airways after intranasal infection.

TLR4 triggers at least two intracellular signaling pathways, one mediated by the common TLR adaptor MyD88 and one that relies on TRIF. The different intracellular adaptors of TLR4 lead to different ways of cell activation; MyD88 activation leads to early phase NF-κB activation using, among others, IL-1R-associated kinase and TNFR-associated factor-6 (1), whereas activation of TRIF also leads to NF-κB activation, but with delayed kinetics (11). The MyD88-independent pathway activates IFN regulatory factor, resulting in the expression of IFN-β (11, 36). Considering that TLR4 especially appeared to be of importance for the clearance of NTHi from the lungs in our first experiments, we determined the relative contributions of MyD88 and TRIF in antibacterial defense in this model. First, we demonstrated that AM from MyD88 KO mice failed to respond to NTHi during 16 h of stimulation. This finding points out that the MyD88-dependent pathway is crucial in the in vitro response to NTHi. Interestingly, TRIF mutant AM produced less KC than WT AM in response to NTHi despite a functional MyD88-dependent signaling pathway. These data suggest that TRIF contributes to the response of AM to NTHi in vitro. Knowledge of the relative roles of the MyD88-dependent and MyD88-independent pathways of TLR4 signaling in host defense against pulmonary pathogens in vivo is highly limited. Two recent studies investigated the role of MyD88 in Pseudomonas aeruginosa pneumonia (37, 38). Both studies focused on the early host response (up to 24 h), showing that in the absence of MyD88, the clearance of Pseudomonas is impaired. In the present study we evaluated the roles of MyD88 and TRIF in the clearance of NTHi by determining the bacterial load 44 h after infection, which is the time point at which our first studies had established an important role for TLR4. Interestingly, TRIF mutant mice were indistinguishable from WT mice with respect to bacterial load and lung inflammation. In contrast, MyD88 KO mice demonstrated an NTHi load in their lungs that was >4 logs higher than that measured in WT lungs, which was associated with more pronounced pulmonary inflammation. In this regard, MyD88 KO mice resembled TLR4 KO mice. Because MyD88 also serves as the essential adaptor for other IL-1R/TLR family members, including the type I IL-1R and IL-18R, and all TLRs except for TLR3, it is intriguing to document that even MyD88 KO mice are apparently able to mount an inflammatory response in the lung in the presence of a very high bacterial burden. It will be of interest to study which part of the immune system is responsible for the increased inflammation in the MyD88 KO mice. Possible MyD88-independent mediators include complement and Ab responses and the TRIF pathway of TLR4 signaling.

Our study demonstrates that TLR4 is of utmost importance for a successful host defense in pulmonary infection with NTHi, and that for an effective function of TLR4, CD14 (in the extracellular compartment) and MyD88 (intracellular) are crucial. Finally, our study reveals that the role of the MyD88-independent, TRIF-dependent pathway of TLR4 signaling is of minor importance for the clearance of NTHi from the respiratory tract.

We thank Marieke ten Brink, Ingvild Kop, and Joost Daalhuisen for expert technical assistance.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by grants from Mr. Willem Bakhuys Roozeboom Foundation (to C.W.W.), the Netherlands Organization of Scientific Research (to S.F.), and the Dutch Asthma Foundation (to N.A.M.).

3

Abbreviations used in this paper: TRIF, Toll/IL-1R domain-containing adaptor-inducing IFN-β; TRAM, TRIF adaptor-related adaptor molecule; AM, alveolar macrophage; i.n., intranasal; KO, knockout; LOS, lipooligosaccharide; MPO, myeloperoxidase; NTHi, nontypeable Haemophilus influenzae; WT, wild type.

1
Akira, S., K. Takeda.
2004
. Toll-like receptor signalling.
Nat. Rev. Immunol.
4
:
499
.-511.
2
Takeda, K., T. Kaisho, S. Akira.
2003
. Toll-like receptors.
Annu. Rev. Immunol.
21
:
335
.-376.
3
Fitzgerald, K. A., E. M. Palsson-McDermott, A. G. Bowie, C. A. Jefferies, A. S. Mansell, G. Brady, E. Brint, A. Dunne, P. Gray, M. T. Harte, et al
2001
. Mal (MyD88-adapter-like) is required for Toll-like receptor-4 signal transduction.
Nature
413
:
78
.-83.
4
Horng, T., G. M. Barton, R. Medzhitov.
2001
. TIRAP: an adapter molecule in the Toll signaling pathway.
Nat. Immunol.
2
:
835
.-841.
5
Yamamoto, M., S. Sato, K. Mori, K. Hoshino, O. Takeuchi, K. Takeda, S. Akira.
2002
. Cutting edge: a novel Toll/IL-1 receptor domain-containing adapter that preferentially activates the IFN-β promoter in the Toll-like receptor signaling.
J. Immunol.
169
:
6668
.-6672.
6
Oshiumi, H., M. Matsumoto, K. Funami, T. Akazawa, T. Seya.
2003
. TICAM-1, an adaptor molecule that participates in Toll-like receptor 3-mediated interferon-β induction.
Nat. Immunol.
4
:
161
.-167.
7
Oshiumi, H., M. Sasai, K. Shida, T. Fujita, M. Matsumoto, T. Seya.
2003
. TIR-containing adapter molecule (TICAM)-2, a bridging adapter recruiting to toll-like receptor 4 TICAM-1 that induces interferon-β.
J. Biol. Chem.
278
:
49751
.-49762.
8
Fitzgerald, K. A., D. C. Rowe, B. J. Barnes, D. R. Caffrey, A. Visintin, E. Latz, B. Monks, P. M. Pitha, D. T. Golenbock.
2003
. LPS-TLR4 signaling to IRF-3/7 and NF-κB involves the Toll adapters TRAM and TRIF.
J. Exp. Med.
198
:
1043
.-1055.
9
Yamamoto, M., S. Sato, H. Hemmi, S. Uematsu, K. Hoshino, T. Kaisho, O. Takeuchi, K. Takeda, S. Akira.
2003
. TRAM is specifically involved in the Toll-like receptor 4-mediated MyD88-independent signaling pathway.
Nat. Immunol.
4
:
1144
.-1150.
10
Yamamoto, M., S. Sato, H. Hemmi, K. Hoshino, T. Kaisho, H. Sanjo, O. Takeuchi, M. Sugiyama, M. Okabe, K. Takeda, et al
2003
. Role of adaptor TRIF in the MyD88-independent Toll-like receptor signaling pathway.
Science
301
:
640
.-643.
11
Kawai, T., O. Takeuchi, T. Fujita, J. Inoue, P. F. Muhlradt, S. Sato, K. Hoshino, S. Akira.
2001
. Lipopolysaccharide stimulates the MyD88-independent pathway and results in activation of IFN-regulatory factor 3 and the expression of a subset of lipopolysaccharide-inducible genes.
J. Immunol.
167
:
5887
.-5894.
12
Doyle, S., S. Vaidya, R. O’Connell, H. Dadgostar, P. Dempsey, T. Wu, G. Rao, R. Sun, M. Haberland, R. Modlin, et al
2002
. IRF3 mediates a TLR3/TLR4-specific antiviral gene program.
Immunity
17
:
251
.-263.
13
Hoebe, K., X. Du, P. Georgel, E. Janssen, K. Tabeta, S. O. Kim, J. Goode, P. Lin, N. Mann, S. Mudd, et al
2003
. Identification of Lps2 as a key transducer of MyD88-independent TIR signalling.
Nature
424
:
743
.-748.
14
Bandi, V., M. A. Apicella, E. Mason, T. F. Murphy, A. Siddiqi, R. L. Atmar, S. B. Greenberg.
2001
. Nontypeable Haemophilus influenzae in the lower respiratory tract of patients with chronic bronchitis.
Am. J. Respir. Crit. Care Med.
164
:
2114
.-2119.
15
Rao, V. K., G. P. Krasan, D. R. Hendrixson, S. Dawid, J. W. St. Geme, III.
1999
. Molecular determinants of the pathogenesis of disease due to non-typeable Haemophilus influenzae.
FEMS Microbiol. Rev.
23
:
99
.-129.
16
Turk, D. C..
1984
. The pathogenicity of Haemophilus influenzae.
J. Med. Microbiol.
18
:
1
.-16.
17
Murphy, T. F., M. A. Apicella.
1987
. Nontypeable Haemophilus influenzae: a review of clinical aspects, surface antigens, and the human immune response to infection.
Rev. Infect. Dis.
9
:
1
.-15.
18
Wang, X., C. Moser, J. P. Louboutin, E. S. Lysenko, D. J. Weiner, J. N. Weiser, J. M. Wilson.
2002
. Toll-like receptor 4 mediates innate immune responses to Haemophilus influenzae infection in mouse lung.
J. Immunol.
168
:
810
.-815.
19
Galdiero, M., E. Finamore, F. Rossano, M. Gambuzza, M. R. Catania, G. Teti, A. Midiri, G. Mancuso.
2004
. Haemophilus influenzae porin induces Toll-like receptor 2-mediated cytokine production in human monocytes and mouse macrophages.
Infect. Immun.
72
:
1204
.-1209.
20
Shuto, T., H. Xu, B. Wang, J. Han, H. Kai, X. X. Gu, T. F. Murphy, D. J. Lim, J. D. Li.
2001
. Activation of NF-κB by nontypeable Hemophilus influenzae is mediated by Toll-like receptor 2-TAK1-dependent NIK-IKK α/β-IκBα and MKK3/6-p38 MAP kinase signaling pathways in epithelial cells.
Proc. Natl. Acad. Sci. USA
98
:
8774
.-8779.
21
Adachi, O., T. Kawai, K. Takeda, M. Matsumoto, H. Tsutsui, M. Sakagami, K. Nakanishi, S. Akira.
1998
. Targeted disruption of the MyD88 gene results in loss of IL-1- and IL-18-mediated function.
Immunity
9
:
143
.-150.
22
Hoshino, K., O. Takeuchi, T. Kawai, H. Sanjo, T. Ogawa, Y. Takeda, K. Takeda, S. Akira.
1999
. Cutting edge: Toll-like receptor 4 (TLR4)-deficient mice are hyporesponsive to lipopolysaccharide: evidence for TLR4 as the Lps gene product.
J. Immunol.
162
:
3749
.-3752.
23
Takeuchi, O., K. Hoshino, T. Kawai, H. Sanjo, H. Takada, T. Ogawa, K. Takeda, S. Akira.
1999
. Differential roles of TLR2 and TLR4 in recognition of Gram-negative and Gram-positive bacterial cell wall components.
Immunity
11
:
443
.-451.
24
Branger, J., C. W. Wieland, S. Florquin, N. A. Maris, J. M. Pater, P. Speelman, T. Shimizu, S. Ishii, T. van der Poll.
2004
. Platelet-activating factor receptor-deficient mice show an unaltered clearance of nontypeable Haemophilus influenzae from their respiratory tract.
Shock
22
:
543
.-547.
25
Frick, A. G., T. D. Joseph, L. Pang, A. M. Rabe, J. W. St. Geme, III, D. C. Look.
2000
. Haemophilus influenzae stimulates ICAM-1 expression on respiratory epithelial cells.
J. Immunol.
164
:
4185
.-4196.
26
Hu, W. G., J. Chen, J. F. Battey, X. X. Gu.
2000
. Enhancement of clearance of bacteria from murine lungs by immunization with detoxified lipooligosaccharide from Moraxella catarrhalis conjugated to proteins.
Infect. Immun.
68
:
4980
.-4985.
27
Carson, R. T., D. A. Vignali.
1999
. Simultaneous quantitation of 15 cytokines using a multiplexed flow cytometric assay.
J. Immunol. Methods
227
:
41
.-52.
28
Wieland, C. W., S. Knapp, S. Florquin, A. F. de Vos, K. Takeda, S. Akira, D. T. Golenbock, A. Verbon, T. van der Poll.
2004
. Non-mannose-capped lipoarabinomannan induces lung inflammation via Toll-like receptor 2.
Am. J. Respir. Crit. Care Med.
170
:
1367
.-1374.
29
Knapp, S., C. W. Wieland, C. van’t Veer, O. Takeuchi, S. Akira, S. Florquin, T. van der Poll.
2004
. Toll-like receptor 2 plays a role in the early immune response to murine pneumococcal pneumonia but does not contribute to antibacterial host defense.
J. Immunol.
172
:
3132
.-3138.
30
Welsh, D. A., C. M. Mason.
2001
. Host defense in respiratory infections.
Med. Clin. North Am.
85
:
1329
.-1347.
31
Strieter, R. M., J. A. Belperio, M. P. Keane.
2002
. Cytokines in innate host defense in the lung.
J. Clin. Invest.
109
:
699
.-705.
32
Campbell, G. D., Jr.
1999
. Commentary on the 1993 American Thoracic Society guidelines for the treatment of community-acquired pneumonia.
Chest
115
:
14S
.-18S.
33
Brown, P. D., S. A. Lerner.
1998
. Community-acquired pneumonia.
Lancet
352
:
1295
.-1302.
34
Gioannini, T. L., A. Teghanemt, K. A. Zarember, J. P. Weiss.
2003
. Regulation of interactions of endotoxin with host cells.
J. Endotoxin Res.
9
:
401
.-408.
35
Zughaier, S. M., Y. L. Tzeng, S. M. Zimmer, A. Datta, R. W. Carlson, D. S. Stephens.
2004
. Neisseria meningitidis lipooligosaccharide structure-dependent activation of the macrophage CD14/Toll-like receptor 4 pathway.
Infect. Immun.
72
:
371
.-380.
36
Hoshino, K., T. Kaisho, T. Iwabe, O. Takeuchi, S. Akira.
2002
. Differential involvement of IFN-β in Toll-like receptor-stimulated dendritic cell activation.
Int. Immunol.
14
:
1225
.-1231.
37
Power, M. R., Y. Peng, E. Maydanski, J. S. Marshall, T. J. Lin.
2004
. The development of early host response to Pseudomonas aeruginosa lung infection is critically dependent on myeloid differentiation factor 88 in mice.
J. Biol. Chem.
279
:
49315
.-49322.
38
Skerrett, S. J., H. D. Liggitt, A. M. Hajjar, C. B. Wilson.
2004
. Cutting edge: myeloid differentiation factor 88 is essential for pulmonary host defense against Pseudomonas aeruginosa but not Staphylococcus aureus.
J. Immunol.
172
:
3377
.-3381.