Role of Macrophages in Early Host Resistance to Respiratory Acinetobacter baumannii Infection

Acinetobacter baumannii is an emerging bacterial pathogen that causes nosocomial pneumonia and other infections. Although it is recognized as an increasing threat to immunocompromised patients, the mechanism of host defense against A. baumannii infection remains poorly understood. In this study, we examined the potential role of macrophages in host defense against A. baumannii infection using in vitro macrophage culture and the mouse model of intranasal (i.n.) infection. Large numbers of A. baumannii were taken up by alveolar macrophages in vivo as early as 4 h after i.n. inoculation. By 24 h, the infection induced significant recruitment and activation (enhanced expression of CD80, CD86 and MHC-II) of macrophages into bronchoalveolar spaces. In vitro cell culture studies showed that A. baumannii were phagocytosed by J774A.1 (J774) macrophage-like cells within 10 minutes of co-incubation, and this uptake was microfilament- and microtubule-dependent. Moreover, the viability of phagocytosed bacteria dropped significantly between 24 and 48 h after co-incubation. Infection of J774 cells by A. baumannii resulted in the production of large amounts of proinflammatory cytokines and chemokines, and moderate amounts of nitric oxide (NO). Prior treatment of J774 cells with NO inhibitors significantly suppressed their bactericidal efficacy (P<0.05). Most importantly, in vivo depletion of alveolar macrophages significantly enhanced the susceptibility of mice to i.n. A. baumannii challenge (P<0.01). These results indicate that macrophages may play an important role in early host defense against A. baumannii infection through the efficient phagocytosis and killing of A. baumannii to limit initial pathogen replication and the secretion of proinflammatory cytokines and chemokines for the rapid recruitment of other innate immune cells such as neutrophils.


Introduction
Acinetobacter baumannii is a ubiquitous, Gram-negative, opportunistic pathogen that frequently induces nosocomial and community-acquired pneumonia, skin and urinary tract infections, and bacteremia [1][2][3], especially in immunocompromised individuals [4]. Moreover, A. baumannii infections are becoming increasingly difficult to treat due to the rapid development of resistance to antibiotics [3,5]. Thus, A. baumannii infection can lead to significant morbidity and mortality, with an overall 30-day mortality rate as high as 49% for respiratory tract infections [6].
Similar to the neutrophil, the macrophage is another important phagocyte that is generally involved in host defense against pathogen invasion. Alveolar macrophages (AMs) are the first line of innate immune cells in the distal respiratory tract that are capable of detecting and eliminating invading pathogens as well as initiating the early host immune response. In this regard, AMs play a critical role in host resistance against both intracellular and extracellular bacterial pathogens [14][15][16][17], and are capable of clearing a low inoculum of bacteria without the recruitment of neutrophils [18]. However, to the best of our knowledge, there are no studies that have systemically evaluated the macrophage function during respiratory A. baumannii infection. In this study, we examined the relative contribution of macrophages in the host defense against A. baumannii infection using in vitro J774A.1 (J774) macrophage cell culture and the mouse model of intranasal (i.n.) A. baumannii infection. Our data suggest that macrophages may play an important role in the early host defense against respiratory A. baumannii infection.

Alveolar macrophage responses to intranasal A. baumannii infection in mice
Since AMs are the front line of innate immune cells that combat respiratory pathogens, we first determined the kinetics of AM recruitment in C57BL/6 mice in response to an i.n. A. baumannii infection. As shown in Fig. 1A, the total number of bronchoalveolar lavage (BAL) cells was moderately reduced at 2 hours post infection (hpi) with approximately 10 8 colony-forming units (CFU) A. baumannii, but was significantly increased by 24 hpi (P,0.01). Macrophages comprised nearly all of the cells recovered from the BAL fluid at 0 and 2 (100% and 98.7%, respectively) hpi. By 4 hpi, some neutrophils (25%) were present in the BAL and by 24 hpi, approximately 90% of the all BAL cells were neutrophils (Fig. 1B). Despite the decrease in the proportion of macrophages in the BAL fluid at 24 hpi, the absolute number of macrophages actually increased significantly at this time point (P,0.01) (Fig. 1B).
Flow cytometric analysis of macrophage surface markers showed no substantial changes in the activation markers CD40, CD80, CD86 or MHC-II (I-A b ) at 2 or 4 hpi, but the expression of CD80, CD86 and MHC-II was significantly increased at 24 hpi (P,0.001) (Fig. 2). More importantly, many AMs at 4 hpi contained high numbers of bacteria within their cytoplasm (Fig. 3) while only a few bacteria were visible in the cytoplasm of small numbers of neutrophils at this time point. These results suggest that A. baumannii induce moderate activation and recruitment of AMs into the lungs, and AMs are capable of taking up A. baumannii cells shortly after i.n. infection of the mice.

In vitro uptake of A. baumannii by macrophages
To further characterize the interaction between A. baumannii and macrophages, we examined the in vitro uptake of A. baumannii by the murine macrophage cell line J774A.1 (J774 cells). A. baumannii were incubated with J774 cells at a multiplicity of infection (MOI) dose of 100. After 4 h incubation, 1.3660.13610 6 CFU bacteria were detected inside J774 macrophages, representing about 3% of the total initial inoculated bacteria. Moreover, the uptake of A. Figure 1. Brochoalveolar macrophage responses to intranasal inoculation with A. baumannii. Groups of C57BL/6 mice (n = 5 per group) were i.n. inoculated with 10 8 CFU of freshly cultured A. baumannii. The mice were euthanized at the indicated times and the lungs were lavaged. The total number of bronchoalveolar lavage (BAL) fluid cells was determined by a hemocytometer (A) and the differential count of macrophages (B, left) and neutrophils (B, right) was determined by examining Hema3-stained cytospin slides. Data are representative of three independent experiments with similar results and presented as mean 6 SD. **P,0.01 vs 0 h. doi:10.1371/journal.pone.0040019.g001 baumannii by J774 macrophages was time-dependent. The bacteria were internalized by the macrophages as early as 10 min after inoculation, and the level of uptake continued to increase until the end of the treatment (4 h) (Fig. 4). These results support the finding of the above in vivo study and showed that macrophages can rapidly and efficiently phagocytose A. baumannii in vitro in a timedependent fashion without the presence of antibody or complement opsonization. Figure 2. Activation of CD11c+ alveolar macrophages following intranasal inoculation of A. baumannii. Groups of C57BL/6 mice were i.n. inoculated with 10 8 CFU of freshly cultured A. baumannii. The mice were euthanized at the indicated times and the lungs were lavaged. The expression of CD40, CD80, CD86 and MHCII by CD11c+ macrophages, measured as mean fluorescence intensity (MFI), was determined by FACS analysis. Data are compiled from two independent experiments with similar results, and are presented as mean 6 SD (n = 6 per time point). ***P,0.001 vs 0 h. doi:10.1371/journal.pone.0040019.g002  Cytokine/chemokine responses in A. baumannii-infected J774 macrophages To further understand the interaction between A. baumannii and macrophages, we determined the levels of interleukin (IL)-1b, IL-6, macrophage inflammatory protein (MIP)-2, tumor necrosis factor (TNF)-a, IL-10 and nitric oxide (NO) in the culture supernatant of J774 cells at 0 (no infection), 4, 24 or 48 h after inoculation with 100 MOI of A. baumannii (Fig. 5). This panel of cytokines and chemokines was selected for study because they have previously been implicated in the immunopathogenesis of A. baumannii infection in mouse models and in human cell culture studies [7,9,11,13,[19][20][21]. At 4 h, the levels of IL-6 and TNFa increased .100-fold (P,0.05) and .400-fold (P,0.001), respectively, as compared to 0 h. Similarly, the level of MIP-2 also increased about 100-fold (P,0.001) although there was no significant change in the levels of IL-1b or IL-10. At 24 and 48 h, levels of MIP-2 and TNF-a were maintained at the levels observed at 4 h while IL-6 and IL-1b concentrations increased significantly and peaked at 24 h (P,0.01) (Fig. 5). The IL-10 level did not change significantly in the first 24 h but then increased significantly at 48 h, in comparison with uninfected cultures (P,0.001). Although NO was undetectable (,2.6 nmol) at 0 or 4 h, significant amounts of NO were detected in infected culture supernatants collected 24 and 48 hpi (P,0.001) (Fig. 5).
Our previous studies have shown that respiratory A. baumannii infection in mice induces local production of moderate amounts of inflammatory cytokines TNF-a, IL-1, IL-6, MIP-2, and antiinflammatory cytokine IL-10 [7]. However, the cellular sources of these cytokines were not identified. Here we show that macrophages, after infection with A. baumannii, produced high levels of IL-6, TNF-a and MIP-2 as early as 4 hpi. MIP-2 and TNF-a levels peaked at 4 hpi, and IL-6 and IL-1b levels peaked at 24 hpi, respectively (Fig. 5). These results suggest that macrophages participate in the early inflammatory responses and host defense against A. baumannii infection. In this regard, it is interesting to note that the production of MIP-2 (a key neutrophil chemotactic chemokine in mice) by macrophages rose rapidly and reached a high level 4 h after infection. It is possible that macrophagederived MIP-2 may contribute significantly to the early neutrophil recruitment following A. baumannii infection in vivo but this can only be definitively concluded after conducting studies using MIP-2 knockout mice.

Mechanisms of A. baumannii entry into macrophages
It has been well established that entry of bacteria into macrophages depends on different receptors, and requires certain cell skeleton changes. To elucidate the potential molecular mechanisms involved in the uptake of A. baumannii by macrophages, we examined the effect of cytochalasin D, nocodazole, and tunicamycin on the uptake of A. baumannii by J774 cells. Cytochalasin D interferes with the microfilament system rearrangement of eukaryotic cells by blocking actin polymerization, which is necessary for macrophage phagocytosis [22][23]. Nocodazole specifically disrupts the cellular microtubulin system, which is essential for receptor-mediated endocytosis of small particles [23][24] while tunicamycin interferes with surface receptor N-glycosylation [20]. As shown in Fig. 6, pretreatment of J774 macrophages with cytochalasin D inhibited the uptake of A. baumannii in a dose-dependent manner, with .80% of inhibition at a concentration as low as 0.1 mg/ml. The uptake of A. baumannii by J774 macrophages was also inhibited by nocodazole and tunicamycin (Fig. 6). These results indicate that the phagocytosis of A. baumannii by J774 macrophages requires intact/functional microfilament and microtubulin systems.

Killing of phagocytosed A. baumannii by J774 macrophages
The killing of phagocytosed pathogens is an important function of macrophages, and is also a major host defense strategy in combating extracellular bacterial infections. We therefore determined the ability of J774 macrophages to kill phagocytosed A. baumannii. J774 cells were infected for 2 h with 100 MOI of A. baumannii, and the cells were lysed after 24 or 48 h to recover viable A. baumannii. As shown in Fig. 7, more than 80% of J774 phagocytosed A. baumannii cells were killed at 24 h post-phagocytosis, and 99% were killed by 48 h. However, .90% of A. baumannii phagocytosed by mouse neutrophils were killed after only one hour of incubation (Fig. 7). Thus, it would appear that in comparison to neutrophils, macrophages are less efficient in killing phagocytosed A. baumannii.

Potential mechanism of macrophage-mediated killing of A. baumannii
Macrophages employ a variety of mediators such as NO, reactive oxygen species and phagolysosome-associated enzymes to kill phagocytosed pathogens [18,25]. Since NO is one of the major effector molecules in bactericidal action of macrophages [26] and NO production is significantly increased in A. baumannii-infected macrophages (Fig. 5), we examined the potential significance of NO in macrophage killing of A. baumannii. J774 cells were treated with NO inhibitor L-NAME or its non-functional isomer D-NAME one hour before being infected with A. baumannii. As shown in Fig. 8, both L-NAME and D-NAME treated macrophages killed substantial numbers of internalized bacteria at 24 hpi. However, the inhibition of A. baumannii in L-NAME-treated J774 cells was ,20% greater than in the D-NAME-treated cells (P,0.01), indicating a role for NO in macrophage-mediated killing of A. baumannii. However, after 48 h incubation, both L-NAME and D-NAME treated cells showed similar level of bacterial inhibition (data not shown). Treatment of J774 cells with L-NMMA, another NO inhibitor, for 24 h also showed a similar reduction in the killing of phagocytosed A. baumannii as L-NAME (Fig. 8). These results indicate that NO plays a moderate role in the killing of phagocytosed A. baumannii by macrophages in the first 24 h after infection. Thus, it is likely that NO contributes to the bactericidal function of macrophages at an early stage of infection before sufficient numbers of neutrophils are recruited at the site of infection. At the later stage of infection, the contribution of NO could be largely compensated or replaced by other more efficient mechanisms (such as neutrophil-mediated killing). These results are in agreement with our previous in vivo work, which showed that iNOS knockout mice are only slightly more susceptible to i.n. A. baumannii infection than wild-type mice [9].
Since our previous studies have shown that NADPH oxidasedeficient (gp91 phox2/2 ) mice that lack the production of reactive oxygen species (ROS) are highly susceptible to A. baumannii infection [9], we also measured ROS production by J774 macrophages in response to A. baumannii infection, and its potential contribution in the subsequent killing of A. baumannii. In contrast to NO responses, we found that J774 cells, either at rest or infected with A. baumannii, produced virtually no ROS although its production increased substantially after 1 h phorbol 12-myristate 13-acetate (PMA) stimulation (Fig. 9). Although our in vitro study excluded any significant contribution of ROS in macrophagemediated killing of A. baumannii, it remains possible that in vivo, the macrophages may be activated by a matrix of cytokines and inflammatory mediators and produce ROS to eliminate phagocytosed A. baumannii. Further studies of the activation of macro-phages by cytokines and other mediators during A. baumannii infection are necessary to resolve this question.

Bactericidal effect of NO on A. baumannii
The above results (Fig. 8) suggest that NO only partially contributes to the killing of A. baumannii by J774 macrophages, and A. baumannii is only moderately susceptible to the bactericidal effect of NO. We next tested the susceptibility of A. baumannii to NO directly in bacterial culture broth supplemented with an NO donor, DETA/NO which induced NO production in the culture broth (Fig. 10A). The viability of A. baumannii in the broth supplemented with 1 or 5 mM NO donor was reduced by approximately 1 log after 2 h, and by approximately 2 logs and 3 logs respectively after 6 h incubation (P,0.001) (Fig. 10B). By 24 h after DETA/NO addition, A. baumannii viability is similar at all concentrations of DETA/NO (Fig. 10C) while viability is still significantly reduced in the more susceptible E. coli bacteria (P,0.01) at the 5 mM dose of DETA/NO (Fig. 10D). In addition, we also compared the susceptibility of five different clinical isolates of A. baumannii with strain ATCC 17961 to 5 mM DETA/NO and found that all of these isolates had similar susceptibilities (Fig. 10E). In contrast, more than 3 log reduction in A. baumannii was observed after only 1 h incubation in the presence of 5 mM H 2 O 2 (data not shown). These data confirm that A. baumannii is only moderately susceptible to the bactericidal effect of NO.

Alveolar macrophages function in host resistance to intranasal A. baumannii infection in mice
Results of the above in vitro experiments (Figs. 4,5,6,7,8,9,10) suggest that macrophages may play some role in the early stage of host defense against A. baumannii infection. We next examined the role of AMs in host defense against i.n. A. baumannii infection in AM-depleted mice. Groups of C57BL/6 mice were intranasally treated with clodronate-liposome to deplete AMs (AM-depleted mice) or phosphate buffered saline (PBS)-liposome as treatment control (control mice) [15,27]. Twenty-four hours later, the mice were i.n. inoculated with 1610 8 CFU A. baumannii, and bacterial burdens in the lung, BAL fluid and spleen were determined at 4, 24, 48 and 72 h after infection (Fig. 11). The bacterial burdens in the lung and BAL fluid were not significantly different between AM-depleted and control mice at 4 hpi, but at 24 hpi significantly higher numbers of bacteria were recovered in the BAL fluid of AM-depleted mice (P,0.01). At 48 and 72 hpi, significantly more bacteria were detected in both the lung and BAL fluid of AMdepleted mice (P,0.01). The overall bacterial burdens in the spleens of infected mice were relatively low, and there was no significant difference between AM-depleted and control mice at any time point. We also determined a panel of 6 proinflammatory cytokine and chemokine levels in the BAL fluid of these mice at 4 hpi and found that the AM-depleted mice had significantly lower levels of IL-6, MIP-2 and TNF-a (p,0.001) than control mice (Fig. 12), whereas the levels of IL-1b, KC, and RANTES were similar between the two groups of mice.
These results indicated that AMs do play a subtle, but important role in early host resistance against i.n. A. baumannii infection. However, in comparison with previous studies with neutrophil-depleted mice [7,11], the contribution of AMs to overall host defense in this model is relatively insignificant. Depletion of neutrophils converted a self-limiting infection into a lethal infection, whereas there was no death observed in AM-depleted mice in this study. In addition, the bacterial burdens in the lungs of neutropenic mice are about 100 times higher than control IgGtreated mice [7], while the difference between AM-depleted and control mice was only about 10 times (Fig. 11). Several factors may contribute to the different resistance observed between AMdepleted and neutropenic mice during i.n. A. baumannii infection. Firstly, AMs may be inherently less efficient in killing A. baumannii since our in vitro experiments showed that it required 24 h for AMs to kill 80-90% of phagocytosed A. baumannii while neutrophils can kill a similar percentage of bacteria in one hour (Fig. 7). Secondly, clodronate-liposome administration can only deplete about 80% of AMs and the depletion is largely restricted to terminal bronchioles and alveolar space [28]. On the other hand, RB6-8C5 monoclonal antibody treatment depletes .90% of pulmonary neutrophils and .95% of circulating neutrophils for at least 2 days [7]. Finally, although administration of clodronate-liposome depleted AMs, it   is likely that this also induced the recruitment of a small number of neutrophils, presumably in response to the cellular debris and soluble molecules released from the lysed macrophages. These neutrophils may partially compensate for the loss of AMs in control of bacterial replication. Nevertheless, the early phagocytosis and killing of bacteria by AMs may be important to contain the initial bacterial burden until neutrophils can be recruited to kill A. baumannii more efficiently and clear the infection.
In summary, the interaction between A. baumannii and the host innate immune system is likely to govern the extent of bacterial proliferation and local host inflammatory response following pulmonary bacterial infection. In this regard, AMs are the predominant phagocytes in the lung, and reside in the distal airway and alveolar spaces. They form the front line of phagocytes that initiate early innate immune responses against pulmonary pathogen invasion, as schematically summarized (Fig. 13). In this regard, macrophages express a wide array of pathogen pattern recognition molecules (such as TLRs, surface scavenger receptors and mannose receptors) to recognize and phagocytose bacterial and non-bacterial pathogens, and initiate inflammatory responses [29][30]. In this study, we used both an in vitro macrophage cell line and an AM depletion mouse model to examine the function of macrophages in the phagocytosis, killing, and regulation of inflammatory responses during A. baumannii infection. Our results showed, for the first time, that although macrophages play a relatively minor role in the overall host defense against A. baumannii infection, they play an important role in the initial stage of host defense against respiratory A. baumannii infection partially through an NO-dependent mechanism. Although at this stage it is not clear whether macrophage dysfunction directly contributes to any nosocomial or community-acquired A. baumannii infection, many studies have revealed that some risk factors (such as cigarette smoke or alcohol abuse) are positively associated with nosocomial and community-acquired infections including A. baumannii [25,[31][32][33]. Smoke or alcohol consumption is also correlated with impaired immune responses including AM dysfunction in phagocytosis, killing of bacteria, and cytokine secretion [25,[34][35][36][37]. Further study of these risk factors and the correlated macrophage dysfunction may provide more detailed information about nosocomial and community-acquired A. baumannii infection. The mouse macrophage-like cell line J774A.1 (ATCC TIB-67, J774) was obtained from American Type Culture Collection (Manassas, VA). J774 is a well-established mouse macrophage-like cell line that has been commonly used in functional studies of phagocytosis and killing because of its similarity to normal macrophages [23]. All cells were maintained in Dulbecco's Modified Eagle Medium (DMEM) (Sigma-Aldrich, St. Louis, MO) supplemented with 10% heat-inactivated fetal bovine serum and 10 mM HEPES (Invitrogen, Burlington, ON) at 37uC with 5% CO 2 .

Intranasal infection of mice with A. baumannii
For each in vivo experiment, fresh inocula were prepared as previously described [7]. Anesthetized mice were inoculated i.n. with ,10 8 CFU A. baumannii in 50 ml saline. This dosage induces a self-limiting bronchopneumonia in C57BL/6 mice without mortality [7]. Actual inoculum concentrations were determined by plating 10-fold serial dilutions on brain-heart infusion agar supplemented with 50 mg/ml streptomycin (BHI-S). Mice were euthanized at the indicated times and the lungs were lavaged five times with a total of 5.0 ml PBS supplemented with 3 mM EDTA as previously described [7,38]. The total number of BAL cells was determined with a hemacytometer and differential cell counts were determined by examining 200 cells on cytospin slides (Cytospin 3, Shandon, Pittsburgh, PA) stained with Hema-3H (Fisher Scientific, Kalamazoo, MI) [7].

FACS analysis of BAL cells
The percentage and activation of AMs in the BAL fluid were determined by FACS analysis as described previously [39]. Briefly, BAL cell samples were washed in PBS containing 1% BSA. The cells were incubated with unlabeled anti-CD16/CD32 (clone 2.4G2) monoclonal antibodies (BD Biosciences) for 15 min to block non-specific Fc receptor binding. Aliquots containing ,10 6 cells were stained with antibody cocktails containing appropriate fluorochrome-conjugated mAb for 30 min at 4uC. The anti-CD11c (clone HL3) mAb was used as a marker for AMs and the activation of CD11c+ AMs was further analyzed using anti-CD80 (clone 16-10A1), anti-CD86 (clone GL1), anti-CD40 (clone 3/23), and anti-MHCII (AF6-120.1) antibodies. After staining, the cells were washed twice with the above PBS/BSA solution, fixed with 200 ml of 1% paraformaldehyde (Polysciences Inc., Warrington, PA), and stored in the dark at 4uC until ready for quantitative analysis. The data were acquired using a FACS Canto flow cytometer (BD Biosciences, San Jose, CA) and analyzed using FlowJo software (Tree Star, Inc., Ashland, OR).

In vitro infection of J774 cells with A. baumannii
Freshly grown A. baumannii (ATCC 17961) were prepared from a frozen stock as previously described [7]. The bacterial cells were then concentrated in Tryptic Soy Broth (TSB) with 40% glycerol and aliquots were stored at 280 u C for use in infection studies. One ml aliquots of J774 (5610 5 cells/ml) cells were seeded into 24-well tissue culture plates (Becton Dickinson, Mississauga ON). After overnight culture, confluent cells were infected with 100 MOI of A. baumannii in 500 ml complete medium per well for 2 h as described by Choi et al. [40]. The cells were then washed 3x with Phagocytosis assay J774 cells that were incubated with A. baumannii for the indicated times were lysed with 0.1% Triton X-100 after a 2-h incubation with gentamicin. The number of viable bacteria recovered was determined by culturing serial dilutions of the lysates on BHI-S plates. The phagocytosis index was calculated by using the following formula: [bacteria recovered (CFU/ml)/ bacteria inoculated (CFU/ml)]6100 = % phagocytosed. The same phagocytosis assay was used to test time-dependent phagocytosis and the potential mechanism of A. baumannii phagocytosis. In these experiments, J774 cells were either co-incubated with A. baumannii for a series of time periods ranging from 10 min to 4 h as indicated, or were treated with cytochalasin D (0.1 to 1 mg/ml), nocodazole (0.1 to 5 mg/ml), or tunicamycin (2 to 5 mg/ml)(all from Sigma-Aldrich) for 30 (cytochalasin D) or 60 (nocodazole and tunicamycin) min before the addition of A. baumannii. The total number of viable A. baumannii recovered was similarly determined and used for the calculation of phagocytosis index. Pilot experiments confirmed that all the chemical reagents showed no direct effect on the replication of A. baumannii or the viability of J774 cells in the dose and time used (data not shown).

Bactericidal assay
To determine the killing of A. baumannii by J774 cells, the above phagocytosis assay was modified by further incubation of the infected monolayers for up to 48 h. Intracellular viable bacteria at 4, 24 and 48 h were determined as described above. For the 24 and 48 h cultures, cells were maintained in DMEM supplemented with only 10 mg/ml gentamicin to minimize the potential intracellular accumulation of antibiotics and subsequent killing of intracellular bacteria. Pilot experiments confirmed that this concentration of gentamicin completely killed the extracellular bacteria after overnight incubation (data not shown).
To study the role of nitric oxide (NO) in the killing of phagocytosed A. baumannii by macrophages, J774 cells were treated with L-NMMA (1 mM), or L-NAME (2 mM), or D-NAME (2 mM) 1 h before infection [41]. Cells were then infected with 100 MOI A. baumannii as described above. NOS inhibitors and controls were added again after infection. Intracellular viable bacteria were determined at 4, 24 and 48 h after the addition of A. baumannii.

Production of cytokines, chemokines and NO by A. baumannii-stimulated macrophages
In some experiments, culture supernatants were harvested at 0, 4, 24 and 48 h after A. baumannii infection. Supernatants from  sham-infected cells were also harvested at the same time points and served as baseline controls. The supernatants were filter sterilized and stored at 220uC until assay. The level of selected cytokines and chemokines were determined by using the mouse panel of Fluorokine MAP Multiplex Kits (R & D Systems, Inc. Minneapolis, MN) on a LuminexH 100 IS system (Luminex, Austin, TX). The analysis was done in duplicate, and cytokine/ chemokines concentrations were calculated against the standards using BeadviewH software (ver 1.03, Upstate) as described previously [7]. The NO 2 2 levels in the culture supernatants were determined by Griess assay [42]. Briefly, 100 ml culture supernatants were mixed with an equal volume of Griess reagent (0.5% sulfanilamide and 0.05% N-1-naphthylethylenediamide hydrochloride in 2.5% acetic acid). After 30 min incubation at room temperature, the mixture was measured spectrophotometrically at 550 nm. The NO 2 2 concentration was determined from a standard curve prepared with NaNO 2 .

DETA/NO and bacterial viability
To test the direct bactericidal effect of NO, NO was generated using diethylenetriamine/nitric oxide adduct (DETA/NO, Sigma-Aldrich, St. Louis), a compound that decomposes spontaneously with a half-life of approximately 20 h [43]. Bacteria were freshly grown in TSB broth and then diluted in tryptic soy broth to a concentration of approximately 10 5 CFU/ml and exposed to 0, 1 or 5 mM DETA/NO. The viable bacterial counts (CFU/ml) at various hours after incubation at 37 u C were determined by plating serial dilutions on BHI plates. Figure 13. Schematic depiction of the potential role of alveolar macrophages during early host defense against intranasal A. baumannii infection. Intranasally inoculated A. baumannii are rapidly taken up by resident alveolar macrophages (AM) shortly after infection. The infection stimulates the synthesis and production of bactericidal molecules such as nitric oxide (NO) which kills some A. baumannii inside AMs or in the alveolar space and partially controls the replication of A. baumannii. Some AMs containing viable A. baumannii migrate to the regional lymph nodes and initiate extrapulmonary dissemination. The proinflammatory cytokines/chemokines (such as MIP-2 and KC) produced by AMs and other pulmonary resident cells (such as dendritic cells and alveolar epithelial cells) in response to A. baumannii induce the recruitment of neutrophils from peripheral circulation into bronchoalveolar spaces where they release oxidative radicals (O 2 2 and H 2 O 2 ) to clear the infection. doi:10.1371/journal.pone.0040019.g013