INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Repetitive airway antigen challenges yield pulmonary T-cell- mediated immune responses in rats (1). Dendritic cells (DC) are effective antigen-presenting cells (APC) (4) and function as sentinel cells for the recognition of antigen in the airways. After recognition of inhaled soluble antigens, DC in the airways traffic to regional lymph nodes (LN), where they may present inhaled antigens directly to T lymphocytes (8).

Resident alveolar macrophages (AM) are poor APC and suppress the activities of both DC and T lymphocytes (9) through the release of nitric oxide (NO) and other soluble inhibitors (10). In the normal rat lung, AM are located less than 0.2 µM from DC in the alveolar interstitium, so that suppressive factors released by AM can potentially diffuse across short distances to downregulate the activities of pulmonary DC in vivo (11). The selective elimination of AM in vivo, via the intratracheal administration of dichloromethylene diphosphonate (Cl₂-MDP, or clodronate) encapsulated in liposomes (LIP-CLOD), enhances the presentation of soluble antigen by pulmonary DC (12) and promotes humoral responses to inhaled antigen (13). However, the interaction of phagocytic AM and DC in the response to particulate antigen in vivo has not been fully investigated.

In a previous study, we found that AM participate in the suppression of cellular pulmonary immune responses to heat-killed Listeria (HKL) by ingesting bacteria and sequestering them away from DC in the pulmonary interstitium (14). In the present study, we report that eliminating AM in vivo decreases NO production by pulmonary macrophages, specifically promotes the APC activities of pulmonary DC for HKL, and augments normally suppressed pulmonary T-cell-mediated immune responses in vivo.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals

Inbred, pathogen-free, 6- to 8-wk-old female Lewis rats (150 to 250 g) were obtained from Charles River Laboratories (Kingston, MA). Rats were housed in a restricted-access animal care facility, and were permitted access to food and water ad libitum.

Liposomes

Liposomes were composed of phosphatidylcholine (PC) and cholesterol (molar ratio 6:1), and contained either phosphate-buffered saline (PBS) or dichloromethylene diphosphonate (Cl₂-MDP, or clodronate, a generous gift of Boehringer Mannheim, Mannheim, Germany) dissolved in PBS. The liposomes were prepared as previously described by van Rooijen and coworkers (15); briefly, 86 mg of PC and 8 mg of cholesterol were dissolved in 10 ml chloroform, and, by low vacuum rotary evaporation, were used to produce a lipid film that was dissolved in either 10 ml PBS (LIP-SAL) or in a solution of 2.5 g Cl₂-MDP in 10 ml PBS (LIP-CLOD). The suspension was kept at room temperature. The LIP-CLOD suspension was then diluted in 100 ml PBS and centrifuged at 100,000 × g for 30 min to remove free Cl₂-MDP, after which the liposomes were resuspended in 4 ml PBS.

Heat-killed Listeria monocytogenes

Heat-killed Listeria monocytogenes (HKL) were obtained from the Massachusetts General Hospital bacteriology laboratory. The concentration of organisms was determined in a McFarland nephelometer, and bacteria were heat inactivated in a 63° C water bath for 90 min; viability was assessed by failure of the bacteria to grow on blood agar plates. Aliquots (10⁹ bacteria/ml in saline) were stored at 20° C.

Complete Medium and Culture Conditions

Cells were cultured in RPMI 1640 medium (JRH Biosciences, Lenexa, KS) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Sigma Chemical Co., St. Louis, MO), 50 µg/ml gentamycin (GIBCO BRL, Gaithersburg, MD), 0.5% 1 M 4-(2-hydroxyethyl)-1-piperazine-N'-2-ethanesulfonic acid (Hepes) buffer (GIBCO), and 2-mercaptoethanol (5 × 10⁵ M) (Sigma), and were incubated at 37° C in 95% air and 5% CO₂ in a humidified chamber.

Monoclonal Mouse Antirat Antibodies

Monoclonal antirat antibodies were used to purify and characterize cells in the studies. These included W3/25 (CD4), OX-8 (CD8), OX-1 (CD45R), OX-6 (Ia) (all from Accurate Chemical & Scientific Co., Westbury, NY); RMA-1 (antirat macrophage); anti-IgM, OX-12 (anti- light chain); OX-33 (CD45RA) (all from Pharmingen, San Diego, CA); and antirat inducible nitric oxide synthase (iNOS) (a gift of K. Bloch, Massachusetts General Hospital). The monoclonal antibodies (mAbs) were prepared as ascites or supernatants, and were used at predetermined optimal concentrations. For in situ localization, DC or lung tissues were stained according to an indirect avidin-biotin immunoperoxidase technique (7). Surface immune phenotype was examined in a FACS 440 cytofluorimeter (Becton Dickinson, Mountain View, CA) after direct staining with fluorescein-conjugated antirat mAbs or indirect staining with goat F(ab')₂ antimouse IgG-fluorescein isothiocyanate (IgG-FITC) (Tago Immunologicals, Camarillo, CA) (16).

Generation of Immune T Cells

Rats were immunized with an emulsion of 1 × 10⁷ HKL or 100 µg hen-egg lysozyme (HEL) in complete Freund's adjuvant (Difco Laboratories, Detroit, MI). The emulsion (0.1 ml) was injected bilaterally at the base of the tail; inguinal LN were harvested at 10 to 14 d, mechanically dispersed, and separated on Isolymph (Gallard Schlesinger Industries Inc., Carle Place, NY). LN mononuclear cells (2 × 10⁶) were incubated in 24-well culture plates with HEL (100 µg/ml) or HKL (10⁷/ml) antigen in a chamber containing humidified 95% air and 5% CO₂. Recombinant interleukin (IL)-2 (100 U/ml) was added to the medium at Day 5 and then 2 to 3 times weekly. Every 3 to 4 wk the cultures were restimulated with HKL or HEL, respectively, in the presence of normal irradiated (3,000 cGY) syngeneic spleen cells at a ratio of 10:1 of spleen cells to T-cell blasts. Immune phenotyping by cytofluorimetry showed that the antigen-specific lymphoblasts were > 95% W3/25⁺ (CD4) and OX22 (CD45RC). The specificity of the response was judged by a difference of more than threefold in the magnitude of (³H)-thymidine incorporation by T-cell blasts in response to HKL than in response to an irrelevant antigen (HEL).

Intratracheal Instillation

Rats were lightly anesthetized with chloral hydrate (400 mg/kg), and the trachea was surgically exposed. Equal volumes (0.1 ml) of LIP-CLOD, LIP-SAL (100 µg/rat), or normal saline were introduced into the trachea via a 25-gauge needle; indocyanine green (2.5 mg/ml; Becton Dickinson Microbiology Systems, Mountain View, CA) was used as a marker of distribution. The rats were subsequently killed at 24 to 72 h and RMA-positive macrophages were enumerated by immunohistochemical staining in situ. In some experiments, rats were rechallenged intratracheally with 1 × 10⁷ HKL on Day 3 and killed at 24 h, and DC were purified in order to assess their APC activities for HKL.

Dendritic Cell Purification

Dendritic cells were purified from lung, as previously described (17). Briefly, lungs were perfused with saline via the pulmonary artery in order to diminish blood contaminants. Excised lung tissue was digested with collagenase and deoxyribonuclease (DNAse) and fractionated by density gradient sedimentation in bovine serum albumin (BSA), and the selected mononuclear cell fraction was adhered on tissue culture dishes (No. 3003/Falcon; Fisher Scientific, Pittsburgh, PA) with complete medium (CM) for 2 h at 37° C. Nonadherent cells were discarded; the adherent fraction was cultured overnight at 37 ° C. Loosely adherent cells enriched for DC were further subjected to immunopanning with OX-6. The selected cells were retrieved by gentle scraping with a rubber policeman, and were applied to plastic tissue culture dishes for 1 h at 37° C to remove adherent macrophages. Nonadherent cells were judged to be highly enriched (> 90%) for DC, on the basis of their morphology, failure to stain for nonspecific esterase, immunostaining characteristics, and ability to stimulate a primary allogeneic mixed lymphocyte reaction (MLR), as previously described (7).

Accessory Cell Activities

DC from the lung, and LN, were irradiated (3,000 cGY), suspended in CM, and plated (1 × 10⁴ cells/well) into 96-well flat-bottom culture plates (Falcon; Fisher Scientific). Nylon-wool-treated and OX-6⁺ cell-depleted splenic lymphocytes or HKL-immune T cells (5 × 10⁴) were added to DC with or without HKL (10⁷ bacteria/ml). The culture plates were incubated for 72 h at 37° C and the wells were pulsed with (³H)- thymidine (1 µCi/well, 80 Ci/mmol; Dupont/NEN, Billerica, MA) for 6 h before harvesting. The wells were harvested in a semiautomatic cell harvester (Skatron Inc., Sterling, VA) and counted in a Tri-Carb liquid -scintillation spectrometer (Packard Instrument Co., Sterling, VA).

Generation of Pulmonary Cell-mediated Immunity

Equal volumes (0.1 ml) of saline or HKL (1 × 10⁷) were introduced into the tracheas of rats via a 25-gauge needle. At Day 14, rats received a second challenge of 1 × 10⁷ HKL or an irrelevant soluble antigen (HEL, 100 µg) intratracheally. The rats were killed at 48 to 72 h and the lungs excised. One lung was frozen in ornithyl carbamyltransferase (OCT) and the other was formalin fixed; 5-µm sections of lung were stained with toluidine blue or hematoxylin and eosin (H&E), respectively, and examined light microscopically for the development of a cell-mediated response.

Adoptive Transfer of AM

AM were retrieved via saline bronchoalveolar lavage (BAL) from the lungs of normal syngeneic Lewis rats, were labeled in vitro with the Hoechst lipophilic dye PKH-2 (Sigma), as previously described (14), and were reinstilled (5 × 10⁶ AM) into the lungs of LIP-CLOD- treated rats. The labeled AM were subsequently localized at 24 h in frozen sections of lung, through epifluorescence microscopy.

Nitric Oxide Production

RNA was extracted from whole lung in 4 M guanidine isothiocyanate/ 25 mM sodium acetate (BRL, Gaithersburg, MD) with 0.1 M -mercaptoethanol (Sigma), and was separated by ultracentrifugation through 5.7 M cesium chloride/25 mM sodium acetate in an SW55 rotor at 42,000 rpm for 8 h at 18° C. The RNA pellet was dissolved in sterile diethylprocarbonate (DEPC)-treated water, transferred to a new sterile tube, and reprecipitated in sodium acetate and 95% ethanol. The pellet was washed in 70% ethanol, dried, and reconstituted in water. The concentration of RNA was determined at OD₂₆₀, and 10 µg was fractionated for 6 to 8 h at 100 mV in 1.3% agarose formaldehyde gels containing ethidium bromide; the integrity of RNA was confirmed by examining 28S and 18S bands in UV light. The RNA was then transferred to a nylon filter, crosslinked, and hybridized overnight with a ³²P-labeled, isoform-specific iNOS complementary DNA (cDNA) probe (a generous gift from K. Bloch). The membrane was then washed at high stringency and exposed to X-ray film. After initial probing, the membrane was stripped and reprobed with ³²P-labeled oligonucleotide complementary to 18S RNA, to reaffirm equal loading of RNA for each sample. Autoradiograms were scanned with a color image scanner, and analyzed with NIH Image 1.44 software (National Institutes of Health, Bethesda, MD) to determine the NOS mRNA isoform: 18S ratio.

BAL was done on the lungs of rats that had received 1 × 10⁷ HKL and liposomes intratracheally. The cell pellet was centrifuged at 400 × g and resuspended in CM; cells were plated at 1 × 10⁶/dish and allowed to adhere overnight at 37° C in a 5% CO₂ incubator. Adherent cells were retrieved via gentle scraping with a rubber policeman. Conditioned supernatants were prepared from rat AM (2 × 10⁵/well) cultured for 48 h in CM with or without recombinant murine interferon- (IFN-, 500 U/ml; Genentech Inc., South San Francisco, CA). In some wells, macrophages were treated with an NOS inhibitor N^G-monomethyl-L-arginine [L-NMMA], 10 mM; Sigma). Nitrite in the conditioned supernatants was measured according to the method of Ding and coworkers (18). Briefly, 50 µl of conditioned supernatant was incubated in an equal volume of Griess reagent at 25° C for 20 min. Chromophore absorbance at 562 nm was determined in a microtiter plate reader (Biotek Instruments, Inc., Winooski, VT). Nitrite concentration was determined by using graded concentrations of sodium nitrite (Sigma) as a standard.

Statistics

Each experiment was repeated at least twice. Data are expressed as mean ± SD. Student's paired t test was used to assess differences between groups.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Elimination of AM In Vivo

Following the introduction of LIP-CLOD, as compared with LIP-SAL or saline controls, the lungs showed a progressive reduction (Figure 1) in the representation of RMA-positive AM in situ. At 72 h, few alveolar septa in the LIP-CLOD-treated lungs showed RMA-positive AM (Figure 2). Immunostaining revealed no difference in the numbers of pulmonary interstitial mononuclear cells staining weakly positive for RMA (9 ± 3 versus 7 ± 3 per hpf), or of OX-6-positive mononuclear cells (11 ± 2 versus 13 ± 4; both p > 0.05) in LIP-CLOD- versus LIP-SAL-treated lungs, respectively. Following pretreatment with LIP-CLOD and subsequent intrathecal challenge with 1 × 10⁷ HKL, the number of macrophages retrieved from the BAL fluid (BALF) was modestly reduced from that of controls (Figure 3). The majority of macrophages (> 85%) identified in the BALF of LIP-CLOD + HKL-treated rats were weakly RMA positive, a surface immune phenotype that is characteristic of exudate macrophages (manuscript in preparation).

View larger version (13K): [in this window] [in a new window]   Figure 1.   LIP-CLOD yielded a progressive decrease in AM in vivo. Rats received intratracheal injections of liposomes that contained Cl2-MDP (LIP-CLOD), saline (LIP-SAL), or saline alone. The rats were killed at 24 to 72 h and their lungs were inflated with cryopreservative, frozen, sectioned at 5 µm, and immunostained with anti-RMA antibody in an indirect immunoperoxidase technique. The number of RMA-positive macrophages lining the alveolar spaces was determined by direct visualization with a light microscope. Data represent mean ± SD of three separate experiments.

View larger version (100K): [in this window] [in a new window]   View larger version (99K): [in this window] [in a new window]   Figure 2.   RMA-positive AM after treatment with LIP-CLOD. Lewis rats were treated as described in the legend to Figure 1. (A) LIP-SAL-treated lung at 72 h. (B) LIP-CLOD-treated lung at 72h. The LIP-CLOD- treated lung shows virtually no RMA-postive AM, and minimal interstitial changes. (Peroxidase and 1% hematoxylin; original magnification ×150.)

View larger version (30K): [in this window] [in a new window]   Figure 3.   Accumulation at 24 h of macrophages in BALF of lungs treated with LIP-CLOD and HKL. Rats pretreated intratracheally with either LIP-CLOD (LIP) or saline subsequently received 1 × 107 HKL intratracheally at 72 h. The number of macrophages retrieved 24 h later from the BALF of control rats that received saline alone or HKL was substantially greater than from those that received LIP-CLOD alone (LIP) or LIP-CLOD + HKL (LIP + HKL). The data are from one of two representative experiments.

APC Activities of Antigen-pulsed Pulmonary DC

As previously reported, the immune phenotype of pulmonary DC was determined by immunostaining. Pulmonary DC were virtually all CD45R⁺, OX-6-positive, and OX-62-positive; fewer than 10% of DC stained for RMA antigen, which may be weakly expressed by a subset of OX-6-positive pulmonary DC (17); DC did not stain for T-cell antigens (CD4, CD8) or B-cell (CD45RA, surface immunoglobulin M [slgM],  light chain) markers.

Pulmonary DC purified from lungs pretreated with LIP-CLOD and subsequently challenged with 1 × 10⁷ HKL given intratracheally were examined for their ability to promote (³H)-thymidine incorporation specifically by HKL-immune T cells in vitro both with and without the addition of HKL to the culture wells. At 24 h, pulmonary DC from the LIP-CLOD- treated rats yielded an ~ 10-fold increase in (³H)-thymidine incorporation by HKL-immune T-cells, as compared with pulmonary DC purified from LIP-SAL controls (Table 1). The APC activities of DC were judged to be specific, based on their inability to serve as APC for "irrelevant" (HEL-immune) T lymphocytes or naive splenic lymphocytes (Table 1). The addition of HKL to the culture wells showed that DC purified from LIP-CLOD- and LIP-SAL-treated lungs had comparable APC capacities.

Generation of Pulmonary Cell-mediated Immunity

A repeat challenge with 1 × 10⁷ HKL given intratracheally was administered 2 wk after intratracheal sensitization. At 72 h, the lungs of the LIP-CLOD-treated rats showed pulmonary vessels cuffed by mononuclear leukocytes (Figure 4), whereas controls showed minimal pulmonary perivascular inflammation (45 ± 5 versus 7 ± 2 perivascular cells/hpf; p = 0.01). An intratracheal challenge with an irrelevant antigen (HEL, 100 µg/rat, given intratracheally) yielded no lymphocytic response (not shown).

View larger version (82K): [in this window] [in a new window]   View larger version (100K): [in this window] [in a new window]   Figure 4.   Intratracheal HKL challenge on Day 14 in sensitized lungs yielded a perivascular lymphoid infiltrate in lungs of rats pretreated with LIP-CLOD. Rats were pretreated intratracheally with LIP-CLOD or LIP-SAL at Day 3, and were then sensitized intratracheally with 1 × 107 HKL. At Day 14, the rats received a second intratracheal injection of 1 × 107 HKL. At 72 h the lungs were removed, inflated with 5% formalin, and examined light microscopically. Rats that received LIP-CLOD (A) showed dense perivascular cuffs of lymphoid cells, whereas the lungs of the LIP-SAL- treated rats (B) showed minimal perivascular inflammation. (H&E; original magnification: ×150.)

Adoptive Transfer of AM

At 72 h after pretreatment with LIP-CLOD given intratracheally, rats received 5 × 10⁶ PKH-2-labeled normal syngeneic AM, or nylon wool-treated splenocytes as controls, both of which were also given intratracheally. Direct visualization of frozen lung tissue by epifluorescence microscopy at 24 h after adoptive transfer revealed AM with fluorescence staining for PKH-2 lining most alveolar septa (Figure 5).

View larger version (93K): [in this window] [in a new window]   View larger version (131K): [in this window] [in a new window]   Figure 5.   Adoptive intratracheal transfer of AM into LIP-CLOD- treated rats reconstituted a population of AM in vivo. AM were retrieved from normal rats by BAL and stained ex vivo with the Hoechst lipophilic dye PKH-2. Rats pretreated intratracheally with LIP-CLOD at Day 1 received 5 × 106 PKH-2-stained AM intratracheally at Day 3 and were killed 24 h later. The lungs were frozen and examined with a Zeiss (Jena, Germany) epifluourescence microscope. The dark field (A) shows PKH-2-labeled AM distributed along alveolar septa in a distribution best appreciated by phase- contrast microscopy (B). (Magnification: ×150.)

Pulmonary DC purified from LIP-CLOD-treated rats that were reconstituted with AM prior to receiving HKL intratracheally were unable to support the proliferation of HKL- immune T cells in vitro (Figure 6); the adoptive transfer of splenocytes did not diminish specific APC activities of pulmonary DC. Histologic evidence of pulmonary cell-mediated responses, following the second challenge with HKL, did not develop in the lungs of LIP-CLOD-treated rats whose AM had been reconstituted by adoptive transfer (not shown).

View larger version (18K): [in this window] [in a new window]   Figure 6.   APC activities of purified pulmonary DC after an airway challenge with HKL. Rats pretreated intratracheally with LIP-CLOD or saline received adoptive transfers of either 5 × 106 normal AM or spleen cells at Day 3. Next, rats were treated intratracheally with 1 × 107 HKL and killed 24 h later. Dendritic cells (1 × 104/ well) were purified from the lung and examined for their ability to serve as APC for HKL-immune T lymphocytes (5 × 104/well) in the absence of added HKL. Data represent the incorporation of (3H)- thymidine at 72 h. (*p < 0.001 compared with DC purified from nontreated rats and rats that received LIP-CLOD and intratracheal adoptive transfer [AT] of AM.)

NO Secretion

Because NO is a potent suppressive factor for the immune activities of T lymphocytes and DC (9), we examined changes in NO production in this model. At 72 h after the instillation of LIP-CLOD or saline, rats were treated intratracheally with 1 × 10⁷ HKL. Lungs were examined at 24 h for expression of iNOS mRNA and for iNOS antigen expression by AM in situ. In parallel experiments, pulmonary macrophages were purified from BALF by adherence on plastic, and were cultured with IFN- (500 U/ml) and assayed for nitrite release at 48 h.

Lungs pretreated with LIP-CLOD and HKL showed substantially less expression of iNOS mRNA than those of saline controls (Figure 7). The expression of iNOS protein in situ in controls that received saline + HKL intratracheally was characterized by heterogeneous and intense focal intracytoplasmic staining of AM (not shown). Rats that received LIP-CLOD + HKL showed a substantial reduction in the number of iNOS- positive AM in situ (Table 2). Functionally, macrophages purified from the BALF of rats challenged with LIP-CLOD + HKL showed reduced nitrite secretion in response to IFN- as compared with those of controls. The adoptive transfer of AM prior to the intratracheal HKL challenge effectively reconstituted the number of iNOS-positve AM in situ, and IFN- stimulated NO secretion by BALF macrophages (Table 2).

View larger version (22K): [in this window] [in a new window]   Figure 7.   Expression of iNOS mRNA in lung extracts. Rats were pretreated with LIP-CLOD or saline. At 72 h they received HKL (107/rat) intratracheally. RNA was extracted and ~ 10 µg was fractionated on agarose gels, transferred onto nylon filters, and probed with 32P-isoform specific probes for iNOS. The figure shows that iNOS mRNA message is decreased in the LIP-CLOD- treated lungs as compared with the saline control. The autoradiogram is one of two comparable experiments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have previously reported that pulmonary DC exposed to HKL in vivo are unable to present antigen to HKL-immune T cells in vitro unless dosages of HKL in excess of 10⁹ rat are given intratracheally (14). It was suggested that AM suppress the activities of DC in vivo by ingesting HKL within the alveolar space, thereby diverting antigen away from DC in the lung interstitium and effectively abrogating the afferent limb of the immune response. However, because AM also suppress the APC activities of DC (12), it was hypothesized that release of NO and other suppressive factors by AM, after their ingestion of HKL might also contribute to downregulation of the cellular immune response in vivo.

In order to ascertain the role of AM in the immune response to HKL, an earlier study adopted a liposome-induced "suicide" technique that specifically targets resident AM (19). In the present study, the delivery of LIP-CLOD intratracheally effectively reduced the number of AM, so that only rare alveoli showed RMA-positive AM in situ at 72 h. Whereas LIP-CLOD eliminated AM, it had little effect on the representation of either weakly staining RMA-positive interstitial macrophages or OX-6-positive DC in situ. Because liposomes have limited capacities to penetrate the alveolar wall, their cytocidal effects were largely limited to superficially located AM. However, the number of BALF macrophages counted in response to an intratracheal challenge with 1 × 10⁷ HKL was minimally reduced, indicating that LIP-CLOD did not impede exudate macrophages from accumulating in the lung.

After the intratracheal injection of a limited number of HKL (10⁷/rat), only pulmonary DC from LIP-CLOD-treated rats were able to serve directly as APC for HKL-immune T lymphocytes in vitro. The activities of pulmonary DC pulsed with HKL in vivo were judged to be specific, as they failed to support (³H)-thymidine incorporation by "irrelevant" (HEL-immune) T lymphocytes or naive splenocytes. Following a second intratracheal challenge with HKL, cell-mediated immune responses developed only in the lungs of LIP-CLOD-treated rats.

Normal AM, adoptively transferred intratracheally, were able to assume their normal location and morphology along alveolar septa. Reconstituting the lung with AM abrogated the direct APC activities of DC for HKL-immune lymphocytes, and antagonized the development of a pulmonary cell- mediated response to HKL in situ. The critical suppressive role of AM in these responses in vivo was further supported by the inability of adoptively transferred naive splenocytes to inhibit immune responses in LIP-CLOD-treated lungs.

Production of NO was decreased in the lungs of rats treated with LIP-CLOD, as judged by diminished expression of iNOS message in whole-lung extracts after HKL challenge, diminished numbers of iNOS-positive macrophages in situ, and decreased levels of IFN--stimulated nitrite production by macrophages in BALF. These findings suggest that AM are the major source of NO production in the lung during inflammation, and confirm previous observations in vitro made in our laboratory (20).

Previous studies have shown that LIP-CLOD-mediated elimination of AM in vivo promotes the immune response to soluble antigen. Thepen and coworkers found that eliminating AM yielded a marked increase in local antibody production in preprimed mice (15). In a subsequent study, they demonstrated that eliminating AM in vivo yielded enhanced local and systemic humoral responses to aerosolized antigen in sensitized rats, but did not affect protective tolerance to inhaled antigen in immunologically naive rats (13). Thepen and coworkers concluded that the immune response to inhaled antigens was differentially regulated in the upper and lower respiratory tract, and suggested that the suppressive effects of AM were mediated by the release of monokines. The present study extends these observations to the response to inhaled particulate antigens, in which the question arises of how phagocytosis and the release of suppressive soluble factors by AM are coordinated in the local immune response.

A large body of evidence demonstrates that AM suppress the accessory activities of pulmonary DC and the mitogenic responses of T lymphocytes (10, 11, 21). The mechanisms mediating this suppression are complex, and include the release of monokines, including transforming growth factor- (TGF-) and IL-10, prostaglandins, and NO (9). Upham and colleagues have reported that AM selectively inhibit T-cell proliferation by antagonizing tyrosine phosphorylation, without substantially altering other aspects of T-cell activation, including Ca⁺² fluxes and lymphokine production (22).

The role of NO in immune responses is complex. The production of iNOS by macrophages is stimulated by certain microbes, lipopolysaccharide, IFN-, and TNF-, whereas TGF-, IL-4, and IL-10 all downregulate iNOS expression (23). NO inhibits the accessory activities of pulmonary DC and also specifically antagonizes the production of T-helper 1 (Th1) cytokines (12, 24). As a result, decreasing the levels of local NO production by AM in the alveolar space could promote Th 1 cell-mediated immune responses. Although the present study confirms that resident AM are a source of NO, and that their elimination is associated with increased APC activities by pulmonary DC, it does not prove that NO is the prime factor responsible for enhanced immune activities in vivo. Future studies that specifically target macrophage iNOS activity will be required to demonstrate the mechanisms for these activities.

How the pulmonary immune milieu is transformed in vivo to support the cellular immune response to HKL is uncertain. Thepen and coworkers have suggested that liposomes may contribute to enhanced immunity through their adjuvant effects for antigen (15). However, such effects are unlikely to account for the rapid increase in APC activities observed for pulmonary DC in our study, and they do not explain why increased cell-mediated responses to HKL developed in LIP-CLOD- but not in LIP-SAL-treated rats.

Resident AM play a critical role in scavenging particulates and in limiting the activation of immune cells in the alveolar wall; however, the macrophage populations participating in the response to inhaled particulates are not limited to AM. At least two other macrophage populations are likely to participate in the immune response to inhaled antigens. In our study, exudate macrophages, recruited from the circulation, represented the majority of BALF macrophages responding to HKL in LIP-CLOD-treated lungs. In addition, a subset of pulmonary interstitial macrophages that show low-level basal expression of iNOS (20) may contribute to the response to antigen. Currently, the way in which these macrophage subsets interact in the pulmonary response to antigen in vivo is uncertain.

We conclude that AM are critical elements in promoting a milieu of tolerance to inhaled particulate antigens. This is achieved by a complex response that includes the ingestion and sequestration of antigen away from interstitial DC, as well as the production of suppressive factors, including NO, that downregulate the activities of interstitial DC and T lymphocytes. Although the current study does not distinguish the relative importance of these events, it does suggest that antagonizing the activities of AM may allow specific T-cell-mediated immune responses to develop in vivo.