INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Lipoarabinomannan (LAM) is a cell wall component of Mycobacterium tuberculosis, which shares many physicochemical properties with lipopolysaccharide (LPS) from gram-negative bacteria (1). LAM can induce the release of cytokines and chemokines by whole blood and mononuclear cells in vitro (2- 5) and by mechanisms that strongly resemble LPS (1, 6, 7). Considering that LPS is an important factor in lung inflammation during gram-negative pneumonia (8), it is conceivable that LAM contributes to the initiation of a proinflammatory response after infection with M. tuberculosis. However, knowledge of the in vivo effect of LAM within the pulmonary compartment is highly limited.

LPS-induced pulmonary inflammation is characterized by enhanced production of cytokines and chemokines within the lung, and the emigration of granulocytes into the bronchoalveolar airspace (9, 12). Several lines of evidence indicate that the proinflammatory cytokine interleukin-1 (IL-1) is an important mediator of LPS-induced inflammation in the lung. IL-1 is produced in the pulmonary compartment after intratracheal administration of LPS, and inhibition of IL-1 activity attenuates lung inflammation elicited by LPS (10, 13). Moreover, recombinant IL-1 causes neutrophilic infiltration in the lung comparable to LPS (9, 13). In tuberculosis (TB), IL-1 contributes to formation and maintenance of granulomas during mycobacterial infection (14). Elevated levels of IL-1 are found in bronchoalveolar lavage fluid (BALF) and lungs of patients with TB (15, 16), suggesting that IL-1 plays a role in the immune response to TB in the pulmonary compartment. IL-1 binds to IL-1 receptor (IL-1R) type I, which is responsible for the biological effects of IL-1, whereas IL-1R type II is a decoy receptor (17).

In this study, we evaluated the in vivo effect of LAM in the pulmonary compartment in mice. Furthermore, the role of IL-1 in LAM-induced lung inflammation was determined by comparison of LAM effects in IL-1R type I deficient (IL-1R^/) and wild-type mice.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals

IL-1R^/ mice back-crossed 6 times to a C57Bl/6 background were kindly provided by Immunex Corporation (Seattle, Horst, WA) (18). C57Bl/6 (IL-1R^+/+) mice (Harlan Sprague Dawley Inc., The Netherlands) were used as wild-type controls. Each experimental group consisted of 6 mice (sex- and age-matched; 7 to 8 wk old) per time point. IL-1R^/ mice were normal in size, weight, and fertility, and displayed no abnormalities in leukocyte subsets (18).

LAM-induced Pulmonary Inflammation

Mannose-capped LAM was isolated and prepared from the rapidly growing M. tuberculosis strain H37Rv (kindly provided by Colorado State University, Fort Collins, CO, under National Institutes of Health Contract NO1-A1-75320). A pilot study was performed to establish an immune activity-inducing dose. A volume of 1 µg of LAM corresponds with 10⁴ colony-forming units (cfu). The LAM preparation contained 21.6 ng/mg LPS as determined by the Limulus test. Mice were anesthesized briefly by inhalation of isoflurane (Upjohn, Ede, The Netherlands), and 50 µg LAM in 50 µl phosphate-buffered saline (PBS) was inoculated intranasally. Each experimental group consisted of six mice per strain (IL-1R^+/+ and IL-1R^/) per time point. Control mice received 50 µl PBS or 1.38 ng LPS dissolved in 50 µl PBS, i.e., the amount of LPS contamination of the LAM preparation. After 4, 8, or 24 h, mice were anesthesized with FFM (fentanyl citrate 0.079 mg/ml, fluanisone 2.5 mg/ml, midazolam 1.25 mg/ml in H₂O; of this mixture 7.0 ml/kg, intraperitoneally). Blood was collected from the vena cava inferior, and BALF was obtained by flushing the lungs 2 times with 500 µl of sterile saline using an endotracheal cannula (Abbocath-T catheter; Abbott, Sligo, Ireland). The lungs were harvested and homogenized immediately in 9 volumes of lysis buffer (0.5% Triton X-100, 150 mM NaCl, 15 mM Tris, 1 mM CaCl, and 1 mM MgCl, pH 7.40) at 4° C using a tissue homogenizer (Biospec Products, Bartlesville, OK). Homogenates were centrifuged at 14,000 rpm for 10 min to remove cell debris, after which the supernatants were stored at 20° C.

Cell Counts

Total cells present in BALF were counted in a hematometer, and leukocyte differentiation was determined on cytospin preparations stained with modified Giemsa stain (Diff-Quick products; Dade, Düdingen, Switzerland).

Assays

All assays were performed in duplicate. IL-1, IL-1, IL-1 receptor antagonist (IL-1RA), macrophage inflammatory protein 2 (MIP-2), keratinocyte (KC), and interferon- (IFN-) were measured by ELISA according to the instructions of the manufacturer (R&D Systems, Abingdon, UK). In addition, tumor necrosis factor- (TNF-; Genzyme, Leuven, Belgium) was measured. Detection limits of assays were 82 (IL-1), 156 (IL-1), 281 (IL-1RA), 111 (MIP-2), 93.8 (KC), and 31.2 (TNF-) pg/ml.

Statistical Analysis

Differences between IL-1R^/, IL-1R^+/+, and saline controls were assessed using the Mann-Whitney U test for unmatched samples. A value of p < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Dose-finding Pilot Study

To establish the dose of LAM that induced an immune response, four mice were intranasally inoculated with 25 or 50 µg of LAM, after which bronchoalveolar lavage (BAL) was performed after 4 and 8 h. Control mice received PBS. The low dose did not induce cell influx into the pulmonary compartment compared with the PBS controls (Table 1). A volume of 50 µg of LAM resulted in elevated numbers of leukocytes in the BALF compared with controls (Table 1). Leukocyte differentiation suggests that the increase in cells in BALF of mice inoculated with LAM results from a granulocyte influx (Table 1). Additional experiments using IL-1R^/ and IL-1R^+/+ mice were performed with 50 µg of LAM. Because the immune response was not diminished after 8 h, an additional time point of 24 h was included in these experiments.

Induction of IL-1

Intranasal administration of LAM to normal C57Bl/6 mice induced an increase in the lung levels of both IL-1 and IL-1 compared with saline controls, peaking at 8 h (2.6 ± 1.2 ng/mg tissue, and 27.4 ± 8.4 ng/mg tissue, respectively, both p < 0.05 versus controls; Figure 1). LPS controls (i.e., mice inoculated with the amount of LPS that contaminated 50 µg LAM, as determined by Limulus amebocyte lysate [LAL] assay) did not differ from saline controls with respect to lung cytokine concentrations (data not shown). To determine whether the absence of the functional IL-1R influences IL-1 production after LAM administration, IL-1 and IL-1 concentrations were also measured in lung homogenates of IL-1R^/ mice intranasally exposed to LAM. Both IL-1 and IL-1 tended to be lower in IL-1R^/ mice than in IL-1R^+/+ mice, although the differences did not reach statistical significance (Figure 1). In plasma, IL-1 and IL-1 remained undetectable in all mice at all time points. Although IL-1RA could be readily detected in lung homogenates of all mice evaluated, LAM administration did not result in an increase in IL-1RA concentrations in lungs, and there were no differences between groups at any time point (data not shown).

View larger version (25K): [in this window] [in a new window]   Figure 1.   Concentrations (mean ± SE) of IL-1 and IL-1 in lungs of IL-1R/ and IL-1R+/+ mice inoculated with 50 µg of mycobacterial lipoarabinomannan. Controls received sterile PBS. n = 6 mice per group per time point.

Recruitment of Leukocytes

Compared with PBS (control mice), LAM induced a significant increase in cell influx in BALF in both mouse strains, peaking at 8 h and lasting for at least 24 h (Table 2). Cells recruited to lungs after LAM administration mainly were granulocytes. Compared with IL-1R^+/+ mice, IL-1R^/ mice had less granulocytes in BALF at 4, 8, and 24 h, the difference only being significant at the later time point (Table 2). LAM effects on cell influx in BALF were not caused by LPS contamination, because mice challenged with the amount of LPS with which the LAM preparation was contaminated did not show changes in the cellular content of BALF (data not shown).

Induction of Chemokines and Cytokines

Having established that LAM induced an influx of granulocytes into lungs, we were interested in the concentrations of CXC chemokines in lung homogenates. LAM administration was associated with elevated lung levels of MIP-2 and KC when compared with PBS controls (p < 0.05, Figure 2). The concentrations of these chemokines did not differ between IL-1R^/ and IL-1R^+/+ mice. To determine the ability of both mouse strains to mount a proinflammatory cytokine response, TNF- and IFN- were measured. LAM induced production of lung TNF- in both strains (p < 0.05 versus controls, Figure 3). Notably, the TNF- response was sustained in IL-1R^+/+ mice and transient in IL-1R^/ mice, resulting in lower concentrations of TNF- in lungs of IL-1R^/ mice at later time points (p = 0.09). IFN- was not induced in the lung by LAM (data not shown). In plasma, TNF- and IFN- were not detectable.

View larger version (29K): [in this window] [in a new window]   Figure 2.   Concentrations (mean ± SE) of chemokines KC and MIP-2 in lungs of IL-1R/ and IL-1R+/+ mice inoculated with 50 µg of mycobacterial lipoarabinomannan. Controls received sterile PBS. n = 6 mice per group per time point.

View larger version (18K): [in this window] [in a new window]   Figure 3.   Concentrations (mean ± SE) of TNF- in lungs of IL-1R/ and IL-1R+/+ mice inoculated with 50 µg of mycobacterial lipoarabinomannan. Controls received sterile PBS. n = 6 mice per group per time point.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The capacity of LAM to induce cytokines and chemokines in cell cultures in vitro has been documented in a number of studies (4, 5, 7). This study is the first to report on the in vivo effects of LAM in the pulmonary compartment. Intranasal administration of LAM resulted in a neutrophilic cell influx in BALF, which was associated with activation of granulocytes and elevated lung levels of the CXC chemokines MIP-2 and KC, and of the proinflammatory cytokines IL-1 and TNF-. The local production of IL-1 contributed to LAM-induced inflammation in the lung, because mice lacking the IL-1R type I, in which IL-1 can not exert biological effects, demonstrated a relatively diminished influx of granulocytes in their BALF, and a lower concentration of TNF-. Together, these data suggest that LAM is an important stimulator of the host response to mycobacterial infection, and that IL-1 is necessary in mounting a full inflammatory response to a pulmonary challenge with LAM.

In vitro, LAM activates cells by mechanisms that are also used by LPS (1, 6, 7). The results of the present study suggest that the intrapulmonary effects of LAM in vivo, resemble previously described effects of LPS in lungs. Intratracheal administration of LPS caused a neutrophilic cell influx in BALF of mice, peaking at 6 h (9). Similarly, in this study, LAM installed intranasally resulted in an increase in neutrophils in BALF, with a peak at 8 h. Whereas LPS induced influx of lymphocytes and monocytes, LAM did not.

Several lines of evidence indicate that IL-1 is important for neutrophil accumulation during lung inflammation. First, both IL-1 and IL-1 can induce neutrophil influx in lungs after intratracheal administration to rodents (13, 19, 20). Second, IL-1 is found in high concentrations in lungs after a pulmonary challenge with LPS, and inhibition of this endogenous IL-1 activity inhibits LPS-induced neutrophil influx in BALF (9, 11, 21). We found that IL-1R^/ mice had less granulocyte influx into BALF than IL-1R^+/+ mice after administration of LAM, suggesting that also during LAM-induced inflammation, IL-1 produced in the lung is an important mediator, either directly or indirectly, of neutrophil migration. Further studies are warranted to support this hypothesis, for example, using anti IL-1 antibodies. Also, because IL-1R^/ and IL-1R^+/+ mice came from a different source, it is possible that the differences observed might have resulted from genetic differences of the mice unrelated to IL-1 receptor expression.

Other mediators that could be involved in LAM-induced neutrophil recruitment include MIP-2, KC, and TNF-. Intratracheal LPS administration induces MIP-2 and KC production by alveolar macrophages, resulting in elevated concentrations in BALF (22). MIP-2 and KC correlate with influx of neutrophils in LPS-induced inflammation (22), and inhibition of either MIP-2 or KC results in a reduced migration of neutrophils into BALF after intratracheal injection of LPS in rats (23, 24). In this study, mycobacterial LAM induced elevated concentrations of MIP-2 and KC in the pulmonary compartment. However, no clear correlation of MIP-2 or KC with the number of granulocytes was seen (data not shown), consistent with the fact that MIP-2 and KC are not exclusively responsible for granulocyte influx (23, 25). In addition, MIP-2 and KC concentrations were similar in IL-1R^/ and IL-1R^+/+ mice, making it unlikely that these CXC chemokines were responsible for the differences in LAM-induced granulocyte influx in BALF in the two mouse strains.

As in LPS-induced pulmonary inflammation (9), TNF- was produced in the lung after LAM administration. Lung TNF- concentrations in this study were lower in IL-1R^/ than in IL-1R^+/+ mice, suggesting that endogenous IL-1 induces TNF- production in the pulmonary compartment after stimulation with LAM. Furthermore, considering that inhibition of TNF- activity reduces LPS-induced neutrophil recruitment to BALF (11, 21), these data indicate that a reduction in TNF- production in the lungs of IL-1R^/ mice may have contributed to the diminished neutrophil accumulation in BALF of these mice. Hence, lower TNF- concentrations resulting from lack of an endogenous IL-1 response may result in a suboptimal induction of other mediators of the immune response. The precise role of IL-1 herein needs further studies.

The concentrations of IL-1 and IL-1 tended to be lower in IL-1R^/ mice, suggesting that IL-1 can induce its own production in vivo. In accordance, production of IL-1 can be induced by stimulation of the type I IL-1R in vitro (17). In addition, after LPS administration, granulocytes are an important source of IL-1 (11). Hence, the lower number of granulocytes in lungs of IL-1R^/ mice may also play a role in their relatively reduced ability to produce IL-1.

LAM is a cell wall component of M. tuberculosis with potent proinflammatory activities in vitro. We demonstrate here for the first time that LAM given intranasally induces an inflammatory response in the pulmonary compartment of mice in vivo, characterized by granulocyte recruitment in BALF, and enhanced local production of cytokines and chemokines. Granulocytes form a first line of defense during TB. However, lung tissue destruction has been associated with high numbers of granulocytes. Therefore, whether LAM contributes to host defense or rather facilitates mycobacterial invasion remains to be determined. LAM-induced lung inflammation may be in part mediated by endogenous IL-1 activity. LAM from M. tuberculosis may be a principal stimulator of innate immunity during TB.