Reduced Nicotinamide Adenine Dinucleotide

The Journal of Immunology

Reduced Nicotinamide Adenine Dinucleotide Phosphate Oxidase-Independent Resistance to Aspergillus fumigatus in Alveolar Macrophages1 E. Jean Cornish,* Brady J. Hurtgen,* Kate McInnerney,* Nancy L. Burritt,* Ross M. Taylor,* James N. Jarvis,† Shirley Y. Wang,† and James B. Burritt2* The fungal pathogen Aspergillus fumigatus is responsible for increasing numbers of fatal infections in immune-compromised humans. Alveolar macrophages (AM) are important in the innate defense against aspergillosis, but little is known about their molecular responses to fungal conidia in vivo. We examined transcriptional changes and superoxide release by AM from C57BL/6 and gp91phoxⴚ/ⴚ mice in response to conidia. Following introduction of conidia into the lung, microarray analysis of AM showed the transcripts most strongly up-regulated in vivo to encode chemokines and additional genes that play a critical role in neutrophil and monocyte recruitment, indicating that activation of phagocytes represents a critical early response of AM to fungal conidia. Of the 73 AM genes showing >2-fold changes, 8 were also increased in gp91phoxⴚ/ⴚ mice by conidia and in C57BL/6 mice by polystyrene beads, suggesting a common innate response to particulate matter. Ingenuity analysis of the microarray data from C57BL/6 mice revealed immune cell signaling and gene expression as primary mechanisms of this response. Despite the well-established importance of phagocyte NADPH oxidase in resisting aspergillosis, we found no evidence of this mechanism in AM following introduction of conidia into the mouse lung using transcriptional, luminometry, or NBT staining analysis. In support of these findings, we observed that AM from C57BL/6 and gp91phoxⴚ/ⴚ mice inhibit conidial germination equally in vitro. Our results indicate that early transcription in mouse AM exposed to conidia in vivo targets neutrophil recruitment, and that NADPH oxidase-independent mechanisms in AM contribute to inhibition of conidial germination. The Journal of Immunology, 2008, 180: 6854 – 6867.

A

spergillus fumigatus is the leading airborne fungal pathogen in immune-compromised individuals, where it can cause potentially fatal invasive pulmonary aspergillosis (IPA)3 (1– 4). However, IPA is rare in individuals with a normal inflammatory response, due primarily to innate immunity in which phagocytic leukocytes including alveolar macrophages (AM), polymorphonuclear neutrophils (PMN), and dendritic cells play an essential early role in the defense against aspergillosis (5). Therefore, infections caused by A. fumigatus, which originate in the airway of immune-compromised humans, arise because of a defect in one or more of the innate mechanisms of resistance that normally protect from IPA.

*Department of Microbiology, Montana State University, Bozeman, MT 59717; and † Oklahoma University Health Sciences Center, Oklahoma City, OK 73104 Received for publication December 21, 2007. Accepted for publication March 13, 2008. 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 National Institutes of Health (NIH) Award 1 R03 AI057931-01 (to J.B.B.), American Heart Association Scientist Development Grant 0630253N (to R.M.T.), and was also made possible by NIH Grant 1 P20 RR020185-01 from the National Center for Research Resources. 2 Address correspondence and reprint requests to Dr. James B. Burritt, Department of Microbiology, Montana State University, 109 Lewis Hall, Bozeman, MT 59717. Email address: [email protected] 3 Abbreviations used in this paper: IPA, invasive pulmonary aspergillosis; AM, alveolar macrophage; PMN, polymorphonuclear neutrophil; qRT-PCR, quantitative RT-PCR; BALF, bronchoalveolar lavage fluid; MCLA, 2-methyl-6-(4-methoxyphenyl)imidazo[1,2-a] pyrazin-3(7H)-one; SOD, superoxide dismutase; HO, heme oxygenase.

Copyright © 2008 by The American Association of Immunologists, Inc. 0022-1767/08/$2.00 www.jimmunol.org

AM are resident pulmonary phagocytes that respond early to inhaled A. fumigatus conidia. Numerous studies have examined the response of AM to A. fumigatus conidia, both in vivo and in vitro. Despite the key role of AM in aspergillosis immunity, there is little in vivo data that indicate how A. fumigatus conidia affect gene expression in AM, limiting our understanding about the molecular mechanisms by which AM respond to this fungus. In contrast to previous studies on in vitro transcriptional and functional changes of peritoneal macrophages and AM following exposure to conidia (6 –9), the present study was designed to include transcriptional responses of AM to conidia in mouse lungs, thus providing better insight into the overall mechanism by which AM help resist infections by A. fumigatus in vivo. The generation of reactive oxygen species represents a wellcharacterized antimicrobial mechanism used by phagocytic leukocytes, and extensive clinical evidence implicates the superoxide-producing phagocyte NADPH oxidase in the resistance to aspergillosis (10). Human patients with defects in this system suffer from chronic granulomatous disease characterized by recurrent infections, extensive granuloma formation, and frequently, IPA (11). Although it is generally accepted that NADPH oxidase is necessary for killing A. fumigatus conidia by PMN (12, 13), there is conflicting information on the role of this mechanism in AM-mediated resistance to IPA (12, 14 –17). Therefore, in addition to transcriptional studies on AM exposed to A. fumigatus conidia, we have examined functional aspects of AM for evidence of NADPH oxidase involvement following contact with conidia. DNA microarray analysis is a powerful method for monitoring the effect of microbes on the global transcription of cellular gene products. Relevant to the present study, microarray analysis has

The Journal of Immunology previously been used to characterize changes in gene expression associated with oxidative defense in phagocytes responding to stimuli in vitro. For example, increased transcription of some genes encoding oxidant scavengers was observed in human monocytes in response to A. fumigatus conidia (6), as was increased transcription of the gene encoding the gp91phox subunit of the NADPH oxidase in mouse macrophages when exposed to LPS (18). Additionally, the analysis of in vitro changes in transcription of human monocytes has provided valuable clues about innate responses to A. fumigatus (6, 8). In those studies, the exposure of monocytes to conidia was found to induce genes involved in a diverse array of cellular functions including leukocyte adhesion, cell recruitment, endocytosis, phagocytosis, and oxidant stress responses. However, transcriptional differences between AM and monocytes (19, 20), and the influences on AM by other resident and recruited cells in the lung, underscore the importance of analyzing the response of AM to conidia in vivo. There are some published in vivo microarray data related to the effects of inflammation in AM (21), but to our knowledge, none relates specifically to the transcriptional responses to A. fumigatus in the lung. The present study was undertaken to extend our previous observation that alveolar PMN both use and require NADPH oxidase for resistance to aspergillosis, and to more clearly define the early in vivo responses of AM to A. fumigatus conidia, before conidial germination. Following the instillation of conidia into the lungs of C57BL/6 and gp91phox⫺/⫺ mice, DNA microarray analysis was used to evaluate both dose- and time-dependent changes in AM gene transcription. Of key importance in the early innate AM responses to A. fumigatus conidia was up-regulation of genes for PMN recruitment, previously associated with conidial display of ␤-glucan (22, 23), augmenting the inflammatory response that protects normal animals from IPA. In our analyses, we observed a group of genes commonly up-regulated by administration of conidia to C57BL/6 and gp91phox⫺/⫺ mice, and by the administration of polystyrene beads to C57BL/6 mice, suggesting an innate immune response to particulate agents. The current study examining in vivo responses of AM identifies a focused and decisive immune response to A. fumigatus conidia involving increased expression of the genes for TNF-␣ and PMN-recruiting chemokines to provide a protective inflammatory response. Because the role of reactive oxygen species produced by AM in response to conidia remains a matter of debate, this study also examined AM for superoxide production following exposure to A. fumigatus conidia for evidence of involvement of NADPH oxidase in fungal killing. Our in vitro analysis of extracellular superoxide production by luminometry indicated exposure of conidia to AM resulted in a reduction in measurable NADPH oxidase-dependent superoxide generation, and that AM from C57BL/6 mice and those from gp91phox⫺/⫺ mice lacking the NADPH oxidase were equally capable of suppressing conidial germination. Furthermore, we observed that in vivo contact of conidia with AM did not lead to evidence of intracellular superoxide production in AM when examined by NBT staining as it did in PMN. Taken together, our results indicate the responses of AM to A. fumigatus conidia include the production of soluble factors that lead to recruitment and activation of additional cell types of the innate immune system, but not activation of superoxide production by the NADPH oxidase complex within AM themselves. The ability of AM to generate proteins that recruit PMN and prime their oxidative burst supports the hypothesis that superoxide production represents a crucial microbicidal mechanism used by PMN to prevent infections by A. fumigatus.

6855

Materials and Methods Preparation of A. fumigatus conidia A clinical isolate (no. 13073; American Type Culture Collection) of A. fumigatus was grown on Sabouraud dextrose agar slants in 75-cm2 culture flasks at 37°C for 5 days and conidia were collected in 0.1% Tween 20 in HBSS (no. 10-547F; Cambrex Bio Science) by gentle rocking as described (13). Conidia were then diluted in HBSS without Tween 20 to obtain either 106 or 107 conidia in 40 ␮l for intrapharyngeal administration. All in vivo inoculations were performed using conidia harvested immediately before use.

Animal handling and sample collection All protocols involving mice were approved by the Institutional Review Board of the Institutional Animal Care and Use Committee at Montana State University. C57BL/6 male mice were obtained at 9 –11 wk of age from Charles River Laboratories, kept in the Animal Resource Center at Montana State University in microisolator cages, and given food and water ad libitum. Breeder mice with a null allele corresponding to the X-linked gp91phox component of the NADPH oxidase (B6.129S6-Cybbtm1Din) were previously produced by backcrossing carrier females with C57BL/6 males for 13 generations. Breeding pairs of these mice were then obtained from The Jackson Laboratory and reared in specific pathogen-free housing in the Animal Resource Center. Only male gp91phox⫺/⫺ mice were used for microarray, quantitative RT-PCR (qRT-PCR), and ELISA studies. They were housed in microisolator cages in an environment of filtered air, given autoclaved food ad libitum, and prophylactically treated with sulfamethoxazole-trimethoprim in their sterile, acidified drinking water. Three days before use, their drinking water was changed to sterile acidified water without antibiotics. Egr1⫺/⫺ mice generated on a C57BL/6 background were obtained from Taconic Farms. Homozygous female Egr1⫺/⫺ mice 9 –11 wk of age were used (no. 002013-M-F, B6.129-Egr1tm1Jmi N12) and female Egr⫹/⫹ C57BL/6 mice were used for comparison in phagocyte responses to conidia. For all inoculation studies, mice were briefly anesthetized with isoflurane and inhaled 40 ␮l intrapharyngeally administered vehicle only (sterile HBSS) for mock inoculations, or vehicle containing either 106 or 107 freshly harvested conidia. In some experiments, 3-␮m polystyrene microspheres (no. 17134; Polysciences) were used instead of conidia to test for transcriptional changes to an alternative particle. At indicated time points following in vivo incubation, mice were euthanized with isoflurane and bronchoalveolar lavage fluid (BALF) was collected through an 18-gauge angiocatheter needle (BD Biosciences) inserted into an incision in the exposed trachea as described (13). Lungs were perfused with six 1-ml volumes of HBSS containing 3 mM EDTA, which were pooled to produce a 6-ml BALF sample from each mouse. Two hundred fifty microliters of BALF were used immediately for cytospin and Wright staining (to obtain leukocyte differentials and to verify pulmonary conidial delivery), and for cell counting by a hemocytometer. In certain experiments, the levels of specific proteins were examined by ELISA on the supernatants of BALF obtained from the first 2 ml of fluid following collection and storage at ⫺80°C before analysis.

Isolation of mRNA from AM Following the removal of samples of the BALF for cell count and differential, a pellet containing AM ⫾ conidia was obtained by centrifugation at 150 ⫻ g for 10 min at 4°C. The cell pellet was then resuspended in 450 ␮l of RNA lysis tissue buffer from the RNeasy mini kit (no. 74904; Qiagen) containing 1% 2-ME. The phagocytes were lysed by cycling the suspension through a pipette tip 10 times and vortexing for 20 s, then the sample was centrifuged at 20,800 ⫻ g for 10 min at 4°C to pellet conidia (if present). The supernatant samples without conidia were then processed according to the RNeasy mini kit protocol. We have verified this procedure does not extract fungal RNA as determined by RT-PCR with A. fumigatus-specific catalase A primers (forward primer: 5⬘-AAGACCTCCTCCAAGGGCATCATT; reverse primer: 5⬘-ACGGACTTTGTGGGCAAGTTCTTC). Genomic DNA was removed from a final 45-␮l RNA eluate using a TURBO DNA-free kit (no. 1907; Ambion) according to the manufacturer’s instructions. The resulting RNA was precipitated in 2.5 M lithium chloride according to the manufacturer’s protocol (no. 9480; Ambion) and stored overnight at ⫺20°C. Following this treatment, the RNA was pelleted by centrifugation and washed three times with 70% ethanol. Following air drying, the resulting RNA pellet was dissolved in 8 –14 ␮l of nuclease-free water at 37°C, quantified using a Nano-Drop 1000 spectrophotometer and its quality was checked using an Agilent 2100 bioanalyzer.

Analysis of transcriptional changes by microarray For microarray analysis, isolated RNA was amplified and biotin-labeled using an Affymetrix two-cycle labeling and control reagent kit (no.

6856

AM RESISTANCE TO ASPERGILLOSIS Table I. Primers for qRT-PCR Gene

Forward Primer

Reverse Primer

Amplicon bp

Tnf Egr1 Ereg Gapdh Hmox1

5⬘-CCAACGGCATGGATCTCAAAGACA-3⬘ 5⬘-TTCCACAACAACAGGGAGACCTGA-3⬘ 5⬘-TTCTGACATGGACGGCTACTGCTT-3⬘ 5⬘-TCAACAGCAACTCCCACTCTTCCA-3⬘ 5⬘-TAGCCCACTCCCTGTGTTTCCTTT-3⬘

5⬘-TGAGATAGCAAATCGGCTGACGGT-3⬘ 5⬘-TGGGTTTGATGAGCTGGGATTGGT-3⬘ 5⬘-CTTTGCTCAAGGGTTGGTGAACAG-3⬘ 5⬘-ACCCTGTTGCTGTAGCCGTATTCA-3⬘ 5⬘-TGCTGGTTTCAAAGTTCAGGGCAC-3⬘

143 186 143 115 107

900494). The mouse genome array 430A 2.0 (no. 90499; Affymetrix) was used to analyze AM gene transcription following hybridization (Affymetrix Gene chip hybridization oven 640) and processing through a Gene Chip Fluids Station 450. For these studies, each conidial and mock inoculation condition was replicated in quadruplicate for C57BL/6 mice, triplicate for gp91phox⫺/⫺ mice, and duplicate for C57BL/6 with polystyrene beads, generating 51 GeneChip experiments. Affymetrix GCOS software was used to convert raw scans to CEL and CHP data files; all of which were imported into GeneSpring GX7.3 (Agilent Technologies) for analysis. According to a recent examination of Affymetrix probe accuracy (24), sequence verified, mismatch permissive chip definition files were imported into GeneSpring (from http://gbic.biol.rug.nl//supplementary/2006/probeverification/), and present, marginal, and absent gene call information was extracted from the CHP files. Following robust multichip averaging normalization with median polishing (25), CEL data was filtered for present gene calls and baseline raw signal intensity of 100 in at least one replicate set equivalent (3 of 51 chips), an operation that trimmed the data being analyzed roughly in half (to 11,591 genes). The remaining genes were filtered by fold change (2 in any comparison) and were then subjected to ANOVA with a Welch t test and Benjamini and Hochberg false discovery rate of 0.01. A hierarchical tree was generated using a Pearson’s correlation and an average linkage clustering algorithm. Subsequent functional enrichment analysis was conducted using DAVID software (www.DAVID.niaid.nih.gov) (26). Microarray data were deposited with Gene Expression Omnibus at the National Center for Biotechnology Information, and can be accessed through accession number GSE8997.

Analysis of mRNA by qRT-PCR To confirm results obtained in microarray studies, qRT-PCR was used to quantify transcription of a representative subset of AM genes significantly altered following exposure to conidia and suspected to be involved in the resulting immune response. Thus, transcription of Ereg, Egr1, and Tnf was examined relative to Gapdh (a constitutively expressed reference gene (27– 30) at 2, 4, and 5 h after inoculation with 107 conidia or mock inoculation in male C57BL/6 and gp91phox⫺/⫺ mice. The Tnf transcript and protein levels were of particular interest in the present study because of the regulatory role of TNF-␣ with respect to NADPH oxidase (31, 32). Due to possible effect on NADPH oxidase activity, transcription of Hmox1 was also examined by qRT-PCR, but only at 4 h in C57BL/6 mice and at 5 h in gp91phox⫺/⫺ mice. The gene-specific primer sets used in these analyses are shown in Table I. A Quanti-Tect SYBR Green RT-PCR kit (Qiagen) was used in combination with a Corbett Rotor Gene 3000 (no. 204243; Qiagen) to analyze transcription using 6 ng of purified total RNA per sample. For these studies, the PCR amplification efficiencies were 1.01 ⫾ 0.04 for Gapdh, 1.07 ⫾ 0.07 for Ereg, 1.02 ⫾ 0.01 for Ereg1, 1.01 ⫾ 0.03 for Tnf, and 1.01 ⫾ 0.01 for Hmox1. In each experiment, controls using no template and no reverse transcriptase were included. In addition, melt curve analysis and gel electrophoresis were used to check product specificity. Analysis of qRT-PCR data was conducted by relative quantification using REST software, which normalizes data and takes amplification efficiencies into account (27, 28, 33).

Molecular pathway analysis From C57BL/6 mice inoculated with 107 conidia and AM collected at 4 h, networks of AM genes were constructed using Ingenuity analysis (www. ingenuity.com) as described (34). Briefly, microarray data (from Table II) containing gene identifiers and corresponding expression values were uploaded to the Ingenuity program to probe the knowledge-based databases for information to suggest molecular participants involved in the response of AM to conidia. Only genes showing fold change values ⱖ2 were evaluated, and only Ingenuity pathways with network scores of ⱖ15 were considered further. We did not examine gp91phox⫺/⫺ data by Ingenuity analysis, because higher constitutive expression of inflammatory genes in gp91phox⫺/⫺ mice resulted in lower corresponding fold change values which were needed to generate the networks.

ELISA Analysis of TNF-␣, IL-1␤, and CXCL2 was conducted on cell-free BALF using the first 2 ml collected, according to the manufacturer’s protocol (TNF-␣ 88-7324-22 and IL-1␤ 88-7013-22, eBioscience; CXCL2 MM200, R&D Systems).

Detection of pulmonary phagocyte superoxide production by luminometry Extracellular superoxide production by AM and LPS-elicited alveolar PMN was examined in white 96-well plates (no. EK-25075; E&K Scientific), using 2-methyl-6-(4-methoxyphenyl)-3,7-dihydroimidazol[1,2a]pyrazin-3-one (MCLA; no. M-23800, Molecular Probes-Invitrogen), a luminometry reagent that has optimal sensitivity for detecting low-level superoxide generation by phagocytes (35). For luminometry studies on AM, naive 10- to 12-wk-old C57BL/6 and gp91phox⫺/⫺ mice were sacrificed and 8 ml of BALF was collected as described above. AM in the BALF were counted, centrifuged at 300 ⫻ g for 10 min at 4°C, and then resuspended at 106 cells/ml in DMEM-10 (no. D5648; Sigma-Aldrich) containing 10% FCS and 4 mM L-glutamine. For superoxide production assays, 105 freshly harvested AM in 100 ␮l of DMEM-10 were adhered to wells of a 96-well microtiter plate in the absence or presence of conidia (5:1 ratio of conidia:AM). Following 90- to 180-min incubation (data is shown for 90-min incubation, but not different from 180-min incubation) in a 37°C incubator with 5% CO2, the DMEM-10 was removed with rare nonadherent PMN and replaced with 100 ␮l of 5 ␮M MCLA in 140 mM NaCl/well. In certain wells, phagocytes were also examined for oxidant production for 30 min after exposure to 910 nM PMA (no. P-8139; Sigma-Aldrich). For examination of oxidant production in PMN, mice were first exposed to aerosolized LPS 12 h before BALF collection as described (13), producing BALF that contained ⬎85% PMN (RBC were removed by hypotonic lysis). Following cell counting and Wright staining, PMN were resuspended at 107 cells/ml in HBSS, and then 10 ␮l containing 105 PMN was suspended in MCLA reagent described above, in the presence of 910 nM PMA. Data for superoxide liberation by both AM and PMN were collected using a Turner Biosystems GloRunner luminometer (1-s data points collected each minute over a 30- to 120-min interval) with wells blanked against representative wells containing 310 U/ml superoxide dismutase (SOD; no. S-2515, Sigma-Aldrich). The average SOD-inhibitable superoxide generation rate was then reported ⫾SEM for five mice (n ⫽ 5). The data were found to be reproducible with additional analyses performed on 3 separate days.

Detection AM superoxide production by formazan deposition Intracellular production of superoxide in AM and PMN was evaluated by exposure of cells to NBT (no. N-6876; Sigma-Aldrich), and microscopic examination for evidence of formazan deposition as previously described (13, 17). Briefly, C57BL/6 or gp91phox⫺/⫺ mice were inoculated with 107 conidia as described above, then BALF was collected 4 or 12 h later and exposed to 500 ␮M NBT for 30 min at 37°C. In certain experiments, conidia were first preswollen in RPMI 1640 for 8 h at 37°C in 5% CO2 then pelleted and resuspended in HBSS before inoculation of mice. For microscopy, cytospins were then prepared and counterstained with safranin as described (13). In some cases, BALF phagocytes in contact with conidia were exposed to 910 nM PMA concurrently with NBT.

In vitro inhibition of A. fumigatus conidial germination by AM To compare the ability of AM from C57BL/6 and gp91phox⫺/⫺ mice to inhibit conidial germination, BALF was collected from naive mice as described above. AM were then pelleted at 150 ⫻ g for 5 min at 4°C, resuspended at 106 AM/ml in DMEM-10, and then introduced to wells in a 96-well tissue-culture plate at 105 cells/well. In this system, AM from uninoculated mice remained ⬎98% viable for 7 days (based on trypan blue exclusion). Immediately after plating AM, conidia were then introduced

The Journal of Immunology

6857

Table II. AM genes with roles in the immune response showing ⱖ2-fold changes (bolded) in transcription following in vivo exposure to conidia gp91phox⫺/⫺ vs Mock

C57BL/6 vs Mock Common

2-h 106

2-h 107

4-h 106

4-h 107

2-h 106

2-h 107

4-h 106

4-h 107

Cxcl1 Cxcl2 Egr1 Ccl3 Il1b 1810011O10Rik Socs3a Nfkbiz Ereg Rgs1 Gdap10 Plk2 Clecsf9a Ccrl2 Gadd45ba Tnf Maff Atf3 Ifrd1 Hmox1 Il1a Cflara Ptp4a1 Ets2 Zfp36 Becn1 Dhx9 Pde4b Nfatc1 Ccl4 Gsr Bcl3 Tde1 Dnaja1 Nck1 Slc20a1 Tob2 Icam1 Nfkbiaa Txnrd1 Jun Slc2a1 Tcf7l2 Abcc5 Fkbp2 Flt1 Calm3

1.7 3.5 2.4 3.0 1.4 1.0 1.4 1.2 1.4 2.1 1.1 1.0 1.2 1.2 1.1 1.3 1.1 1.4 1.0 1.1 1.0 1.2 1.5 1.5 1.2 0.9 1.0 1.1 1.1 0.9 1.2 1.1 0.9 1.3 1.1 0.9 1.3 1.0 1.4 1.2 1.6 0.9 1.1 1.3 0.9 1.2 0.7

7.4 14.6 6.7 4.9 2.1 3.2 2.1 2.6 3.8 6.1 1.7 1.9 2.5 2.2 1.6 1.7 1.3 2.8 1.6 1.5 1.4 1.8 2.4 2.6 1.0 0.8 1.3 1.7 1.7 1.4 1.8 1.8 1.0 2.4 1.3 1.4 1.7 1.3 1.5 1.7 2.4 0.8 1.2 2.0 0.7 1.3 0.6

10.1 20.7 6.5 10.2 15.4 3.5 7.5 7.4 7.7 3.1 1.7 2.0 5.2 4.0 3.4 3.6 1.8 2.2 1.5 1.9 3.0 1.5 1.0 1.9 1.9 1.0 0.6 1.5 1.2 2.7 0.9 3.1 0.7 1.2 1.0 1.2 0.9 1.8 1.7 1.3 1.8 2.4 0.9 0.9 1.0 1.2 1.0

48.2 39.8 28.1 18.6 15.8 12.8 12.0 10.7 9.8 7.5 7.2 6.3 6.2 6.1 5.3 4.9 4.1 3.7 3.3 3.2 3.1 3.0 2.8 2.7 2.6 2.6 2.6 2.5 2.5 2.3 2.2 2.2 2.1 2.1 2.1 2.1 2.0 2.0 2.0 2.0 2.0 1.9 1.8 1.5 1.4 1.4 1.0

1.1 1.3 1.4 1.0 0.9 1.0 1.3 0.9 1.0 0.9 1.1 1.2 1.0 0.9 1.2 1.0 1.0 1.2 1.0 0.6 0.9 1.1 0.8 1.1 1.0 0.9 0.8 0.9 1.1 1.2 0.9 1.1 0.7 0.9 1.0 1.2 1.3 1.1 1.1 1.0 1.1 1.0 1.1 1.0 0.9 1.2 1.2

3.4 3.7 3.4 1.1 1.0 0.8 0.9 1.2 1.7 1.4 1.4 1.9 1.5 1.0 1.3 2.3 1.5 1.5 1.7 0.4 1.2 1.5 1.5 1.4 1.5 0.6 2.0 1.3 1.4 0.9 1.8 1.4 0.7 1.1 0.9 1.6 2.6 1.3 1.5 1.2 1.3 0.7 2.0 1.5 0.5 2.0 0.5

1.4 1.5 1.6 1.4 0.8 1.6 1.0 1.0 1.3 2.1 1.1 1.3 1.1 1.2 1.1 1.0 1.1 1.1 1.1 0.9 1.1 1.3 0.7 1.2 0.8 0.9 1.1 1.2 1.4 1.0 1.0 0.7 0.8 1.0 0.9 1.1 1.2 1.0 1.2 1.2 1.1 1.1 1.0 1.4 1.0 0.9 0.8

8.1 4.3 4.4 4.3 1.9 1.6 2.4 1.6 3.3 2.4 1.4 2.1 2.1 1.9 1.6 2.1 1.9 1.8 1.7 1.3 2.1 2.2 1.2 1.5 1.8 1.2 2.2 1.2 1.3 1.4 1.2 1.4 1.6 1.0 1.5 1.4 1.4 1.3 1.4 1.9 1.0 1.2 1.4 2.5 1.0 1.1 1.0

a

Average of two or more probe sets for gene on array.

and incubated for 8 h (3:1 ratio, conidia:AM), during which time phagocytosis of most conidia occurred. Cytospin mounts were then generated and Wright stained to determine the percentage of germinated intracellular and extracellular conidia by microscopy.

Results Characterization of phagocytic cells in the mouse lung in response to A. fumigatus To better understand the in vivo engagement of A. fumigatus conidia by AM in a normal and susceptible mouse model, the BALF was examined at specific time points following instillation of conidia in C57BL/6 and gp91phox⫺/⫺ mice, respectively. At 2 and 4 h after intrapharyngeal administration of conidia, PMN were largely absent in the BALF of both the C57BL/6 and gp91phox⫺/⫺ strains, and in the rare cases when they were present, did not exceed 4% of the total leukocyte number. At 5 h following administration of 107 conidia, PMN recruitment was still generally not

evident in C57BL/6 mice, but was more frequently seen in gp91phox⫺/⫺ mice. At 6 h, 39.7 ⫾ 7.8% (n ⫽ 3) of the cells in the BALF of C57BL/6 mice were PMN, so the transcription studies described below were not extended to samples collected beyond the 5 h time point. Of the total AM recovered in the BALF at 4 h after inoculation of C57BL/6 mice using 106 and 107 conidia, 9.1 ⫾ 1.1 (n ⫽ 8) and 21.1 ⫾ 2.2% (n ⫽ 7) of AM were associated with conidia, respectively (Fig. 1A). In similar studies, the corresponding values at 4 h for gp91phox⫺/⫺ mice were 7.5 ⫾ 2.6 and 21.3 ⫾ 1.5% (n ⫽ 4), while the introduction of 107 polystyrene beads (as a model particulate) resulted in 16.7 ⫾ 5% (n ⫽ 3) of AM associated with the beads. For each strain of mice, the percentage of phagocytosing AM did not increase significantly ( p ⬎ 0.3) between 2 and 5 h at either dose of conidia (106 and 107), suggesting few if any additional AM became involved in phagocytosis during that time. Further characterization of AM

%AM with conidia/beads

6858

AM RESISTANCE TO ASPERGILLOSIS

A

gp91phox-/-

C57Bl/6 6

10 conidia 107 conidia 107 beads

30 25

30 25

20

20

15

15

10

10

5

5

0

0

Conidia/beads per AM

2

4

5

2

4

5

2

4

5

B 8

10

6

8 6

4 4

2

2

0 2

4

5

0

Time (h)

FIGURE 1. In vivo AM-particulate engagement 2–5 h following intrapharyngeal administration. A, Average percentage of AM containing A. fumigatus conidia or polystyrene beads at the indicated time points in C57BL/6 (left panel) and gp91phox⫺/⫺ mice (right panel). B, Average number of conidia or polystyrene beads in AM involved in phagocytosis at time points shown in A for the two mouse strains. Not all treatments were evaluated at each time point, thus not all bars are shown. Replicates for conidia are n ⬎ 3, and for polystyrene beads n ⫽ 3, with error bars indicating SEM.

phagocytosis in the lung demonstrated that the average number of particles within AM at the 4 h time point (after receiving 107 conidia) reached 5.3 ⫾ 0.17 (n ⫽ 9) in C57BL/6 mice and 6.5 ⫾ 0.30 (n ⫽ 4) in gp91phox⫺/⫺ mice, while 4.7 ⫾ 0.88 (n ⫽ 3) were observed in AM collected from C57BL/6 mice after inoculation with 107 polystyrene beads (Fig. 1B). By comparison, an in vitro study using A. fumigatus conidia and human monocytes, observed: 1) that approximately three times as many cells had engulfed conidia at 4 h; 2) that the number of cells involved in phagocytosis was still increasing between 4 and 6 h; and 3) an average of only two conidia per phagocytosing AM at the 6-h time point (6). Such differences between those studies and our current data demonstrate the dependence of cell type and cellular environment on outcomes when analyzing the response of phagocytic cells to A. fumigatus conidia. Transcriptional analysis of AM in the presence and absence of A. fumigatus conidia To identify potential effector mechanisms used by AM to prevent infection by A. fumigatus, AM were isolated from the lungs of both C57BL/6 and gp91phox⫺/⫺ mice (with or without conidia) for microarray analysis. In microarray studies comparing these two mouse strains, the constitutive transcription in the absence of conidia (time 0) was significantly different for many genes ( p ⬍ 0.05), demonstrating that elimination of functional NADPH oxidase has a marked effect on the basal AM transcriptome (data not shown). This basic microarray characterization highlights the need to account for basal transcriptional changes in studies using geneknockout mice, and is consistent with a broad role for superoxide in signaling and regulation of gene transcription in AM (36, 37).

When the microarray studies described above were conducted with AM following exposure to particulates, significant transcriptional changes were observed for cells isolated from both mouse strains. ANOVA analysis of data for C57BL/6 and gp91phox⫺/⫺ mice after administration of A. fumigatus conidia or polystyrene beads indicated 73 gene probes showed transcriptional changes at least 2-fold ( p ⬍ 0.01) when comparing inoculated animals to their mock controls, in one or more of the experimental conditions (data not shown). When these data were examined for conidiaresponsive genes, mouse strains tended to cluster together in groups segregating by conidial dosage and timing (Fig. 2). Specifically, AM from gp91phox⫺/⫺ mice showed overall higher constitutive gene expression and responded minimally to the 106 dose of conidia at both 2 and 4 h, clustering with gp91phox⫺/⫺ time 0 and mock arrays. C57BL/6 mice also responded modestly to the 106 dose of conidia at 2 h; however, at 4 h, those from 106 dose arrays clustered between those from 2 and 4 h following the 107 dose. Interestingly, AM from C57BL/6 mice 4 h after administered 107 polystyrene beads clustered most closely with 107 conidia at 2 h. As administration of polystyrene beads to C57BL/6 mice elicited less change than did conidia, it is possible that the type of particle is important in the degree of up-regulation. A group of eight genes was up-regulated at least 2-fold in C57BL/6 mice exposed to 107 polystyrene beads, and in both strains of mice exposed to 107 conidia (Table III). The strongest transcriptional changes in response to both conidia and polystyrene beads were observed for genes encoding proteins important in PMN recruitment. Up-regulation of AM genes involved in PMN recruitment was dependent on both conidial dose and duration of exposure in vivo After administration of 106 conidia to C57BL/6 mice, there were ⱖ2-fold ( p ⬍ 0.01) increases in transcription of Cxcl2 (MIP-2) at 2 h, and in Cxcl1 (KC), Cxcl2, and Il1b at 4 h (Table II). By contrast in gp91phox⫺/⫺ mice receiving 106 conidia, no significant increases in the genes encoding PMN recruiting molecules were observed at 2 and 4 h. After administration of 107 conidia, Cxcl1 and Cxcl2 were up-regulated in both strains at 2 h, while there was up-regulation of Il1b in C57BL/6 mice and Tnf in gp91phox⫺/⫺ mice. At 4 h after inoculation with 107 conidia, PMN were being recruited by Cxcl1, Cxcl2, Il1a, Il1b, and TNF-␣ in both mouse strains (although Il1b in gp91phox⫺/⫺ mice minimally missed the cutoff at a fold change of 1.9). The largest fold changes in the transcripts outlined above were observed at 4 h after 107 conidia in C57BL/6 mice, when there were ⬃48- and 40-fold increases for Cxcl1 and Cxcl2, respectively. In addition to these five chemokine and cytokine genes involved in PMN recruitment, transcriptional changes in 42 additional genes with roles in the immune system were detected in our microarray analyses (Table II). Up-regulation of AM genes encoding oxidoreductases following exposure to conidia Due to the importance of a functional NADPH oxidase in resisting aspergillosis, the list of AM oxidoreductase genes altered by exposure to conidia was evaluated. There was no up-regulation of transcripts encoding cytosolic regulatory subunits of the NADPH oxidase complex (p40phox, p47phox, p67phox, or Rac1/2) for any condition, and similarly, there was no significant change in transcription of Cyba or Cybb (encoding the membrane-bound p22phox and gp91phox subunits of flavocytochrome b-558, respectively). We also observed no up-regulation of genes encoding the oxidant scavengers superoxide dismutase or catalase under any conditions following inoculation of either mouse strain with conidia. Genes

The Journal of Immunology

6859

gp 0hr gp 2h e6 gp 2h mock gp 4h e6 gp 4h mock bl 2h mock bl 4h mock bl 0h bl 2h e6 bl 2h e7 bl 4h bead bl 4h e6 bl 4h e7 gp 2h e7 gp 4h e7

CXCl1 CXCl2 Egr1 AA408868 Ereg Plk2 1frd1 Slc20a1 Il1b Socs3 Socs3 Ccl4 Bcl3 Tnf Clecsf9 Clecsf9 Gadd45b Lcam1 Gadd45b Maff 1810011010Rik Cflar Cflar Cflar Txnrd1 Ccl3 Atf3 Rgs1 Ccrl2 Il1a Ets2 Pde4b Nfkbia Gtpbp4 Nfkbia Zfp36 Jun Hmox1 Slc2a1 Psmd10 Becn1 Tde1 Nck1 Gdap10 Nfatc1 Gsr Tob2 Tcf712 Flt1 Ptp4a1 Gopc Dnaja1 Dhx9 Abcc5 Fkbp5 Fkbp2 Grcc10 Mpdu1 Calm3 1110025Lo5Rik Mrpl23 Hist2h2aa1 Ndufb7 Ndufa8 Rab4b Rpl39 0610010L17Rik 2510027N19Rik 2610208E05Rik Sdh1 Atp5k Bloc1s1 Fus

FIGURE 2. Heat map showing in vivo differences in transcription of AM genes in C57BL/6 and gp91phox⫺/⫺ mice. The hierarchical tree reflects 73 genes found significantly different by ANOVA. Data compare AM from mice exposed to conidia with mock controls exposed to vehicle

involved in cellular redox regulation that were altered included an increase in the transcript for heme oxygenase 1 (HO-1) HmoxI, marginal increases in the transcripts for glutathione reductase (Gsr) and thioredoxin reductase 1 (Txnrd1), and a down-regulation of the transcript for NADH dehydrogenase (ubiquinone) 1 ␤ subcomplex, Ndufb7 (Table IV). Ingenuity analysis of microarray data Ingenuity analysis was conducted on data from C57BL/6 mice to explore relationships between AM genes involved in the response to conidia. This analysis can identify relationships between genes revealed by microarray analysis and overlay their corresponding expression levels (fold change), where the degree of up-regulation is indicated by the intensity of red color. None of the genes found to be down-regulated by microarray analysis appeared in the networks. In the result of the network analysis, molecules or molecular complexes predicted by the program to participate in this response, but not found significantly altered in microarray data, are included as well (symbols without color). Thus, Ingenuity analysis is capable not only of constructing associations of genes identified by microarray (including relative expression levels), but also of predicting involvement of additional molecules not associated with significant transcriptional changes. Using the Ingenuity search criteria described in Materials and Methods, AM genes significantly altered in C57BL/6 mice 4 h after receiving 107 conidia segregated into three separate pathways (Fig. 3) that differed in regard to the type of molecular response predicted by the program. Network A is involved primarily in immune response and contained many of the proinflammatory molecules found up-regulated by microarray analysis. Network B included responses involved in cell-to-cell signaling and cell interaction. Molecules predicted to have the most complex neighborhoods (hubs) in networks A and B were TNF-␣ and IL-1␤, both showing relationships with at least 20 other molecules in each of the two networks. In network C, molecules (including NF-␬B, p38MAPK, and AP-1) contributing primarily to early changes in gene expression are shown to form complex hubs, though none of these three genes were identified by microarray data. Verification of DNA microarray results by qRT-PCR To better validate the microarray results reported in this study, transcriptional changes for several genes of particular interest (Egr1, Ereg, Tnf, and Hmox1) were evaluated by qRT-PCR. All qRT-PCR analysis was conducted on AM from mice treated as indicated, but separate from those used for microarray analyses. When the transcript levels of Egr1, Ereg, and Tnf were quantified using qRT-PCR in the absence of conidia as well as 2, 4, and 5 h after instillation of 107 conidia, increases measured for all three genes determined by REST software (33) were found to parallel those observed by microarray analysis (Fig. 4). Furthermore, C57BL/6 mice showed more marked up-regulation than gp91phox⫺/⫺ mice by qRT-PCR, which was consistent with our microarray data. qRT-PCR was also used to show a median fold increase of 5.4 for Hmox1 (SE 3.9 –7.2) in C57BL/6 mice 4 h after inoculation with 107 conidia, which was similar to the 3.2-fold increase found by microarray analysis. The Hmox1 transcript in AM from gp91phox⫺/⫺ mice showed a 1.3-fold increase at 4 h

alone (Welch t test, Benjamini & Hochberg test correction, p ⬍ 0.01, fold change ⱖ2). Replicates for conditions are: C57BL/6 mice receiving conidia (106 or 107) or mock, n ⫽ 4; all gp91phox⫺/⫺ mice, n ⫽ 3, C57BL/6 mice receiving polystyrene beads, n ⫽ 2. Abbreviations: gp, gp91phox⫺/⫺; bl, C57BL/6; e6, 106 conidia; e7, 107 conidia; bead, 107 polystyrene beads.

6860

AM RESISTANCE TO ASPERGILLOSIS Table III. Genes in AM up-regulated ⱖ2-fold at 4 h by 107 A. fumigatus conidia and polystyrene beads

Gene

Name

C57BL/6 (Conidia)

C57BL/6 (Beads)

gp91phox⫺/⫺ (Conidia)

Cxcl2 Cxcl1 Egr1 Ereg Clecsf9/Clec4e Ccl3 Plk2 Socs3

Chemokine (cxc motif) ligand 1 Chemokine (cxc motif) ligand 2 Early growth response 1 Epiregulin C-type lectin domain family 4, member e Chemokine (cc motif) ligand 3 Polo-like kinase 2 (Drosophila) Suppressor of cytokine signaling 3

39.8 48.2 28.1 9.8 6.2 18.6 6.3 12.0

13.4 4.8 4.3 3.5 2.9 2.5 2.2 2.3

4.3a 8.1 4.4a 3.3 2.1a 4.3a 2.1 2.4

a

Constitutive expression was significantly greater in gp91phox⫺/⫺ mice than in C57BL/6 mice.

Functional analysis of the Egr1 gene product in response to conidia

by microarray and a 1.8 (SE 1.8 –2.4) fold increase by qRTPCR by 5 h.

In light of the degree of up-regulation of Egr1 by 107 conidia in both C57BL/6 and gp91phox⫺/⫺ mice (28.1 and 4.4-fold increase, respectively), and its connection with several other genes as predicted by Ingenuity analysis, responses of Egr1⫺/⫺ mice (n ⫽ 4) and parental strain C57BL/6 mice (n ⫽ 4) to conidia were compared. Administration of 107 conidia to Egr1⫺/⫺ mice resulted in 17.5 ⫾ 2.4% AM in the BALF with internalized conidia at 6 h (compared with 22 ⫾ 0.6% in C57BL/6) ( p ⫽ 0.16) and 6.5 ⫾ 0.6 conidia/AM following phagocytosis (compared with 6.0 ⫾ 0.13% in wild type) ( p ⫽ 0.47). PMN recruitment was also not significantly different between Egr1⫺/⫺ (1.6 ⫻ 105 ⫾ 2.7 ⫻ 104 PMN/ lavage) and C57BL/6 mice (1.2 ⫻ 105 ⫾ 1.5 ⫻ 104 PMN/lavage), and comparable levels of TNF-␣ were observed in the BALF of both mouse strains (89 ⫾ 11.4 pg/ml for Egr1⫺/⫺ mice and 124 ⫾ 15.2 pg/ml for C57BL/6 mice, p ⫽ 0.12).

Validation of DNA microarray results by ELISA To examine the relationship between the observed transcription of AM genes and levels of the protein product, three genes predicted by Ingenuity analysis to play key roles and also known to encode proteins secreted into the BALF (TNF-␣, IL-1␤, and CXCL2) were chosen for analysis. CXCL2 protein concentrations in the BALF of C57BL/6 mice were found to increase 22-fold by 4 h in mice receiving 107 conidia compared with those receiving sterile vehicle alone. In identically treated gp91phox⫺/⫺ mice, an increase in CXCL2 protein of ⬃8-fold was found. Increases of CXCL2 were found to be significant in both animal strains receiving conidia relative to mock inoculated animals (n ⫽ 5 p ⬍ 0.005), and the increase in C57BL/6 mice was found to be significantly greater than in gp91phox⫺/⫺ mice ( p ⫽ 0.012). Following inoculation of mice with 107 conidia, no changes in the concentrations of TNF-␣ or IL-1␤ protein in the BALF of C57BL/6 mice could be detected at 4 h, despite the increases in transcript levels observed in this study. When these experiments were extended to 5 h, the concentration of TNF-␣ in the BALF of C57BL/6 mice was determined to be 7-fold higher in mice receiving conidia compared with those mock-inoculated (n ⫽ 4, p ⫽ 0.012, Fig. 5). The TNF-␣ response was therefore slower in C57BL/6 mice following a dose of 107 conidia (this study), relative to what we previously observed in BALB/c mice after receiving 3 ⫻ 107 conidia (13). This apparent delay in TNF-␣ production in C57BL/6 relative to BALB/c mice may be linked to the slower PMN recruitment in C57BL/6 mice following conidial challenge (13). The conidia-treated gp91phox⫺/⫺ mice showed levels of TNF-␣ in the BALF similar to those measured in inoculated C57BL/6 mice ( p ⫽ 0.3, Fig. 5). In contrast to results obtained with TNF-␣ and CXCL2, no measurable increase in the concentration of IL-1␤ was observed up to 5 h in the present study (data not shown).

Correlation of phagocyte superoxide production with the inhibition of conidial germination Our observation that TNF-␣ was increased at both the transcript and protein level indicated it could be up-regulating gp91phox transcription and priming the respiratory burst in macrophages, as suggested (31, 32). However, our microarray data did not identify increased transcription in components of NADPH oxidase following in vivo exposure to conidia, suggesting either that the NADPH oxidase components were preformed and activated following contact with conidia, or that they were not used in this response. Because the molecular mechanisms by which phagocytes block germination of A. fumigatus in the lung remain to be fully characterized, and to better describe innate response of AM to conidia, a functional role of superoxide production in this process was examined to complement transcriptional results of this study.

Table IV. Genes in AM which encode oxidoreductases showing ⱖ2-fold changes (bolded) in transcription following inhalation of A. fumigatus conidia Fold Change 6

Gene

2-h 10 Conidia

C57BL/6 4-h 10 Conidia

Hmox1 Gsr Txnrd1 Ndufb7

1.5 1.2 1.2 0.8

1.5 1.8 1.7 0.6

6

Mice 2-h 10 Conidia

1.9 0.9 1.3 1.0

7

4-h 107 Conidia

2-h 106 Conidia

gp91phox⫺/⫺ 4-h 106 Conidia

Mice 2-h 107 Conidia

4-h 107 Conidia

3.2 2.2 2.0 0.8

0.6 0.9 1.0 1.0

0.4 1.8 1.2 0.5

0.9 1.0 1.2 0.8

1.3 1.2 1.9 1.1

The Journal of Immunology

6861

A

B CXCL1

CXCL2

C

Legend: A A A

binding only inhibits acts on

cytokine

B B

enzyme

B

group or complex transcription regulator

A

inhibits/acts on

B

A

translocates to

B

kinase

direct interaction

other

indirect interaction transmembrane receptor

FIGURE 3. AM gene network analysis by Ingenuity. Analysis of microarray data by Ingenuity predicts three networks showing molecular interactions of AM genes from C57BL/6 mice 4 h after inoculation with 107 A. fumigatus conidia. Primary molecular responses identified by the analysis were: A, immune response; B, cell-to-cell signaling and interaction; and C, gene expression. Molecule symbols are marked with intensity of red color to indicate the relative degree of transcriptional up-regulation.

To measure extracellular superoxide production by AM from C57BL/6 and gp91phox⫺/⫺ mice following exposure to conidia, a sensitive MCLA-dependent luminometry method (35) was used to

detect low-level NADPH oxidase activity. When exposed to 910 nM PMA, AM from C57BL/6 but not gp91phox⫺/⫺ mice showed a small but significant early increase in superoxide release ( p ⫽

6862

AM RESISTANCE TO ASPERGILLOSIS

Microarray

qRT-PCR

A

Egr1 30

80

C 57B L gp91 phox-/-

25

120

60

20 15

40

100

20

80

10 5 0

0

4

Ereg

2

3

4

5

150

12 10

100

8 6

50

4 2

0

0 2

4

Tnf

5 4

2

3

4

5

40 30

3

20

2

60

Relative Light Units (RLU)

Fold change

2

C57Bl/6 AM C57Bl/6 AM +conidia C57Bl/6 AM +PMA gp91phox-/- AM

140

40 20 0 -20

0

10

0

0 2

4

2

3

4

5

Time (h) after administration of conidia FIGURE 4. Comparison of microarray and qRT-PCR transcriptional data for select AM genes following administration of 107 A. fumigatus conidia to C57BL/6 mice and gp91phox⫺/⫺ mice. Microarray data (left panels) show fold changes in transcription at 2 and 4 h, while qRT-PCR data (right panels) show median fold change examined at 2, 4, and 5 h. For qRT-PCR, error bars indicate SE of median, n ⫽ 4.

30

B 5000

C57Bl/6 PMN gp91phox-/- PMN

4000

10

1

20

3000

2000

1000

0 0

10

20

30

minutes

BALF protein (pg/ml)

0.014 at 3 min), which subsided within ⬃15 min and became indistinguishable from C57BL/6 AM without PMA stimulation (Fig. 6A). The relatively high concentration of 910 nM PMA was used to elicit maximal response from phagocytes, but did not produce measurable nonspecific effects in gp91phox⫺/⫺ cells that lack the functional NADPH oxidase. Surprisingly, when AM from C57BL/6 mice were exposed to a 5-fold excess of freshly harvested conidia for 90 min at 37°C, the signal from superoxide release was reduced, making it indistinguishable from that of AM isolated from gp91phox⫺/⫺ mice. In these superoxide-production

500 400 300

C57Bl/6-mock gp91phox-/--mock C57Bl/6-conidia gp91phox-/--conidia

200 100 0

TNF- α

CXCl2

FIGURE 5. Comparison of TNF-␣ and CXCL2 concentrations in BALF from C57BL/6 and gp91phox⫺/⫺ mice after intrapharyngeal administration of 107 A. fumigatus conidia. ELISA was used to measure the TNF-␣ and CXCL2 concentration in the first 2 ml of BALF collected from each mouse. For TNF-␣ measurements, BALF was collected 5 h after dose (conidia or mock treatment) and BALF was collected 4 h after dose for CXCL2 measurements. Error bars indicate SEM (p ⬍ 0.05 for each strain of inoculated mice compared with uninoculated controls).

FIGURE 6. Superoxide release from murine alveolar phagocytes detected by luminometry. All data represent the superoxide release from 105 phagocytes using MCLA-dependent luminometry, displayed in relative light units (RLU). A, AM collected from the BALF of naive C57BL/6 or gp91phox⫺/⫺ mice were incubated in DMEM-10 for 90 min at 37°C in 5% CO2 and then examined for 30 min by luminometry. Adherent C57BL/6 AM were examined alone, after stimulation by 910 nM PMA, and after incubation with a 5-fold excess (5 ⫻ 105) of A. fumigatus conidia. Signal of superoxide release from gp91phox⫺/⫺ AM alone is shown, and is not distinguishable from that of gp91phox⫺/⫺ AM exposed to conidia or 910 nM PMA (data not shown). B, PMN were examined by luminometry for comparison to results obtained with AM under similar conditions. PMA-dependent superoxide release from LPS-recruited murine PMN is shown for C57BL/6 and gp91phox⫺/⫺ mice. Error bars indicate SEM (n ⫽ 5) and similar results were obtained in three additional experiments.

assays, PMA-triggered alveolar PMN from C57BL/6 mice generated substantially higher levels of superoxide compared with C57BL/6-derived AM (Fig. 6B), while superoxide production was insignificant for PMN from gp91phox⫺/⫺ mice. Because functional and transcriptional analysis of AM responding to A. fumigatus conidia did not indicate obvious NADPH oxidase involvement (in contrast to a report for murine peritoneal macrophages responding to the Streptococcus pyogenes bacterium (38)), we next examined AM from mice exposed to conidia in vivo for evidence of intracellular superoxide production. To assess intracellular superoxide production, phagocytosing AM from C57BL/6 and gp91phox⫺/⫺ mice collected in BALF 12 h after instillation of 106 conidia were exposed to NBT and examined by microscopy. Despite the presence of functional NADPH oxidase in

The Journal of Immunology

6863 ment in aggregates (Fig. 7A, panel 6). AM in the vicinity of the PMN shown in panel 6 (arrows) uniformly failed to produce formazan when in the presence of both conidia and PMA, which was consistent with our luminometry data shown in Fig. 6A. The in vivo studies outlined above comparing C57BL/6 and gp91phox⫺/⫺ mice suggest the phagocyte NADPH oxidase complex does not play a major role in the innate response of AM to A. fumigatus conidia, extending information we reported previously (13). However, it is possible that AM-independent influences in the mouse lung contribute to suppression of conidial germination inside AM. We therefore used an in vitro assay to determine whether a difference exists in the ability of AM from gp91phox⫺/⫺ and C57BL/6 mice to inhibit germination of phagocytosed conidia (Fig. 7B). Our results indicate that 8 h following exposure of AM to conidia in a tissue culture system (at which time conidia outside AM show abundant germination) inhibition of germination inside AM appeared equally effective for the two mouse strains, relative to conidia germinating outside the AM ( p ⬍ 0.0002). These results demonstrate that functional NADPH oxidase is not required by AM to prevent conidial germination under these in vitro assay conditions. This information suggests AM neither use nor require superoxide production by the NADPH oxidase to protect from aspergillosis in the mouse lung.

Discussion

FIGURE 7. AM suppress conidial germination independently of NADPH oxidase activity. A, Following instillation of 107 A. fumigatus conidia and in vivo incubation, BALF AM from C57BL/6 (panels 1 and 3) or gp91phox⫺/⫺ (panels 2 and 4) mice were found to contain conidia but no evidence of NADPH oxidase-dependent formazan deposition. After instillation of resting conidia and 12 h in vivo incubation, conidia were found to be ungerminated in both mouse strains (panels 1 and 2). Panels 3 and 4 show AM collected from mice 4 h after instillation of preswollen and germinating conidia. In contrast, BALF PMN from C57BL/6 mice were found to engage conidia with evidence of formazan deposition (panel 5), as did all BALF PMN (but not AM) from C57BL/6 mice after stimulation with 910 nM PMA (panel 6, arrows show AM and arrowhead identifies AM containing conidia). Scale bars represent 10 ␮m. B, In vitro suppression of conidial germination following phagocytosis in AM from C57BL/6 and gp91phox⫺/⫺ mice. Resting conidia were combined with AM collected from naive mice and incubated in vitro for 8 h at 37°C in 5% CO2. Germination percentages both inside and outside the AM were then determined by microscopy following cytospin and Wright staining. Error bars indicate SEM (n ⱖ 5).

AM derived from C57BL/6 mice, no obvious difference could be found in formazan deposition within AM obtained from the two mouse strains engulfing resting conidia (Fig. 7A, panels 1 and 2). In a previous study, oxidant production was observed when A. fumigatus germlings were engaged by bone marrow derived macrophages (39). We therefore inoculated mice with swollen and germinating conidia and found AM from C57BL/6 and gp91phox⫺/⫺ mice 4 h after in vivo incubation to be indistinguishable with regard to formazan deposition in the vicinity of engulfed developing conidia (Fig. 7A, panels 3 and 4). For comparison, obvious formazan was found among PMN aggregates formed in vivo in C57BL/6 mice in response to conidia (13) (Fig. 7A, panel 5). Formazan deposition was also observed in PMN from C57BL/6 mice when exposed to 910 nM PMA, regardless of their involve-

The goal of this study was to examine the responses of AM following inoculation of mice with A. fumigatus conidia to better understand the innate resistance to IPA. To this end, we used a combination of in vivo and in vitro analyses to provide information that reflects the complexity of the innate response to A. fumigatus conidia in the context of the lung. In vivo transcriptional changes of AM responding to conidia were first determined by microarray and Ingenuity analysis, with qRT-PCR verification of changes for selected genes. TNF-␣ and CXCL2 were also correlated to protein levels in the BALF. Of the 73 gene probes showing ⱖ2-fold changes in transcription in AM from C57BL/6 and gp91phox⫺/⫺ mice responding to conidia (data not shown), 47 actual genes were classified by DAVID Bioinformatics Resources 2007 Functional Annotation as being involved in the immune response (Table II). In addition, all significantly altered genes in C57BL/6 mice were examined by Ingenuity analysis to help predict key biochemical networks involved in this early response phase. To our knowledge, the present study represents the first global analysis of transcriptional changes for AM in response to A. fumigatus conidia in the lung. Phagocytic leukocytes contain a multisubunit NADPH oxidase complex that generates superoxide which serves as a precursor for a set of microbicidal reactive oxygen and nitrogen species. Because superoxide has also been shown to regulate signal transduction cascades that alter gene transcription (37, 40), the responses of AM to conidia in C57BL/6 mice (which are generally resistant to IPA) were compared with those in susceptible gp91phox⫺/⫺ mice (12) by microarray analysis. In these studies, we observed strong differences in basal expression of numerous genes when comparing C57BL/6 and gp91phox⫺/⫺ mice, even before exposure to conidia. These differences could be due in part to compensatory responses in gp91phox⫺/⫺ mice, which lack metabolic regulatory mechanisms contributed by NADPH oxidase activity (41). In support of this hypothesis, differences in constitutive gene transcription have also been found in PMN from human patients with Xlinked chronic granulomatous disease compared with normal individuals (42). Recently, the altered inflammatory response of chronic granulomatous disease patients to pathogens including A.

6864 fumigatus has been linked to dysregulation of tryptophan metabolism and alteration of T cell subsets (43). Despite marked differences in constitutive expression of many genes in the C57BL/6 and gp91phox⫺/⫺ mouse strains used in the present study, the innate response of the parental C57BL/6 strain to conidia was conserved in a subset of genes in gp91phox⫺/⫺ mice. However, a stronger overall transcriptional response to conidia was observed in AM from C57BL/6 than in gp91phox⫺/⫺ mice. This difference does not appear to be due to greater phagocytosis of conidia by AM from C57BL/6 than gp91phox⫺/⫺ mice, as AM from the two strains were not significantly different with respect to the percentage of AM containing conidia. In fact, the average number of conidia per AM was greater in gp91phox⫺/⫺ than in C57BL/6 mice at both 2 and 4 h ( p ⫽ 0.0002 and 0.02, respectively). Approximately 40% of the genes up-regulated in response to conidia in both strains of mice were also increased in C57BL/6 mice after administration of 107 polystyrene beads, suggesting some of these changes in gene transcription occur as a general response to inhaled particulates. Previous experimental evidence has shown that phagocytic cells respond to A. fumigatus conidia by producing proinflammatory chemokines and cytokines involved in PMN recruitment (6, 13, 44 – 46). Our microarray data from C57BL/6 mice extend those studies by indicating the strongest increases in AM transcription are in the chemokine genes Cxcl1 and Cxcl2. In the present microarray study, increases in the transcripts for Tnf, Il1a, and Il1b (genes that have additional roles in PMN recruitment (47, 48)) were also observed. The above findings for AM exposed to conidia in the mouse lung support the notion that early transcriptional changes observed in mouse AM represent a general proinflammatory response of mononuclear phagocytes that triggers PMN recruitment. Exposure of AM to conidia in the lung also up-regulated genes involved in monocyte recruitment, including Ccl3, which encodes Ccl3 (MIP-1␣), a chemoattractant for circulating monocytes (49, 50), and the chemokine receptor-like 2 gene (Ccrl2), which encodes a receptor for chemokines such as Ccl2 (51). In light of these results it is interesting to note that in BALB/c mice, we previously found no evidence of a net increase in AM numbers in the BALF until 24 h after intratracheal administration of 3 ⫻ 107 conidia, though PMN recruitment was evident within 3 h (13). Ingenuity analysis is a tool for comparing novel molecular data sets to a library of knowledge-based cellular and molecular responses. Through our examination of data from AM of C57BL/6 mice, this method of data mining revealed particular complexity in network neighborhoods for TNF-␣ and IL-1␤ (Fig. 3, networks A and B), supporting their importance in the inflammatory response to fungal pathogens including Aspergillus (8, 52, 53). Our Ingenuity analysis also revealed molecular components underrepresented in microarray data, or possibly those that did not participate as expected. For example, we did not observe increased transcription for p38 MAPK and NF-␬B in our microarray data, though each was predicted by Ingenuity analysis to participate with several genes that were transcriptionally activated (Fig. 3, network C). Our Ingenuity analysis results support the importance of NF-␬B in aspergillosis immunity, as reduced nuclear NF-␬B translocation in macrophages was previously implicated in increased susceptibility to infection following dexamethazone treatment (54), and in sepsis (55). Our data are also consistent with a role for both NF-␬B and p38 MAPK in dendritic cells following contact with A. fumigatus (56). Other than for Egr-1, the extent to which other hubs identified by Ingenuity analysis are important to aspergillosis immunity were not examined in the present study. The increased Tnf and Cxcl2 transcription observed by microarray was confirmed using ELISA by increased secretion of TNF-␣

AM RESISTANCE TO ASPERGILLOSIS and CXCL2 protein in the BALF. Increased expression of Tnf was characterized in detail at both the transcript and protein level due to its predicted connection with other molecules in the response of AM exposed to conidia, as well as its role in both regulation of gp91phox gene expression and priming the respiratory burst in macrophages and PMN (31). Increases in TNF-␣ protein production in response to A. fumigatus conidia have been reported in mouse and rat AM in vitro (46, 47, 57), and in mouse lung tissue (58). Increased levels of TNF-␣ have also been found in the BALF after inhalation of A. fumigatus conidia (13, 44), but this TNF-␣ is probably not secreted only by AM (which represent the main source of this cytokine in the lungs) but also by epithelial, mesenchymal and dendritic cells (59). In the present study using C57BL/6 mice, a delay was observed between increased transcription of Tnf and a measurable increase in TNF-␣ protein, consistent with a similar study using C57BL/6 mice (22). ELISA also revealed 22- and 8-fold increases in CXCL2 protein in the BALF 4 h after exposure to 107 conidia, corresponding to 40- and 4-fold transcriptional upregulation of the Cxcl2 gene in C57BL/6 and gp91phox⫺/⫺ mice, respectively. In contrast to results with TNF-␣ and CXCL2, we were unable to detect significant changes in IL-1␤ concentration in the BALF up to 5 h after administration of conidia to C57BL/6 mice, although IL-1 protein secretion has been reported to increase following prolonged exposure of murine AM to conidia (22, 60, 61). These observations are in agreement with other reports showing difficulty correlating IL-1␤ mRNA to protein levels (8, 44). In this study, the lack of a corresponding increase in IL-1␤ protein concentration in the BALF at 5 h despite more marked up-regulation of this gene transcript relative to that observed for TNF-␣ at 4 h could indicate posttranscriptional regulation for expression of this gene. The difference in the results when comparing TNF-␣, IL-1␤, and CXCL2 (measured using comparable methods) emphasizes the fact that increased transcription may not necessarily be directly linked with translation (62). Previous studies have shown that some pathogens can overcome natural resistance to infection by suppression of the host immune response. The present study showed that exposure of both C57BL/6 and gp91phox⫺/⫺ mice to conidia caused up-regulation of two immune suppressive AM genes, Rgs1 and Socs3. The regulator of G-protein signaling 1 (Rgs1) protein product has previously been shown to play a role in deactivation of signaling through chemoattractant receptors (63), while the suppressor of cytokine signaling (Socs3) protein product has been shown to mediate negative feedback effects on IL-6 responses (64). It is important to emphasize that the C57BL/6 mice used in our studies were not specifically immune suppressed and have been shown to be generally resistant to aspergillosis. Therefore in animals with normal immune status, anti-inflammatory effects induced by A. fumigatus conidia resulting from increased expression of Socs3 or Rgs1 would be unlikely to lead to infection unless additional levels of immune dysfunction (such as neutropenia) also existed. For this reason, responses in AM leading to anti-inflammatory effects may be more likely to identify mechanisms by which the inflammatory response in the lung is attenuated to reduce tissue damage, rather than those triggered by the fungus to overcome the immune response. It remains to be determined whether any host-induced antiinflammatory responses triggered by conidia in AM are exploited by the organism to cause disease. Several lectin-like receptors on leukocytes have previously been reported to participate in binding of A. fumigatus conidia or hyphae (22, 39, 65). Our in vivo data from C57BL/6 mice indicated a 6.2-fold up-regulation of the transcript for Clecsf9 4 h after 107 conidia. Clecsf9 (Clec4e; macrophage-inducible C-type lectin

The Journal of Immunology (Mincle)) encodes a group-II Ca2⫹-dependent transmembrane lectin with a role in inflammation and immunity (66), but has not been previously implicated in responses to conidia. In addition to being up-regulated in AM by conidia, Clecsf9 was also up-regulated by polystyrene beads, though to a lesser extent than by conidia. Additional novel findings in the present study were up-regulation of Ereg and Egr1 in both strains of mice following exposure to conidia, and in C57BL/6 mice exposed to polystyrene beads. Ereg encodes epiregulin, a transmembrane protein that is cleaved by the metalloproteinase ADAM17, to form a soluble 5.5-kDa epidermal growth factor receptor agonist that can exert a proliferative or dedifferentiating autocrine effect in AM (67, 68). Previous studies have shown that Egr1, which encodes early growth response 1, is up-regulated in human monocytes exposed to conidia in vitro (6), but not as strongly as we observed in murine AM isolated from the lung. Early growth response-1 is a zinc-finger transcription factor that binds enhancer elements of Tnf, Cxcl2, Ccl3, Gadd45, IL-1␤, M-CSF, VEGF, NF-␬B, jun-D, c-myc, gene 475, c-myb, and Egr1 (69 –72), raising the possibility that some of the genes upregulated in AM by conidia may be responding to Egr1. Despite the marked up-regulation of Egr1 by conidia observed in this study and suggested involvement of this gene by Ingenuity analysis, our functional characterization of Egr1⫺/⫺ mice (6 h after administration of 107 conidia) indicated the lack of this gene product did not significantly: 1) affect phagocytosis of conidia by AM; 2) influence the levels of TNF-␣ secretion into the BALF; or 3) delay PMN recruitment into the lung. These results support the notion that multiple levels of immune suppression are needed to predispose to infections by A. fumigatus (73). Further studies are required to identify the potential role of Egr1 in the immune response to A. fumigatus. Relevant to our analysis of superoxide production by AM in the present study, it is of interest to emphasize that while there was a general lack of transcriptional change in genes involved in the generation and removal of oxidants (such as SOD, catalase, and NADPH oxidase subunits), we measured the up-regulation of Hmox1 (encoding HO-1) in C57BL/6 mice in response to A. fumigatus conidia. HO-1 uses NADPH to convert heme to carbon monoxide and biliverdin, with both products showing antiinflammatory effects in cells (74, 75). Because both heme and NADPH are also required for function of gp91phox, it is possible that HO-1 activity in AM competes with, and therefore suppresses NADPH oxidase activity. A previous study in RAW 264.7 mouse macrophages showed HO-1 indeed suppresses the activity of NADPH oxidase through heme degradation and proteasome-dependent degradation of gp91phox (76). In support of this view, we did not observe an increase of Hmox1 in AM from gp91phox⫺/⫺ mice responding to conidia by either microarray or qRT-PCR analysis. In fact, the strongest down-regulation of any immune function gene shown in Table II was for Hmox1, where it occurred in gp91phox⫺/⫺ mice 4 h following exposure to 106 conidia. These results raise the possibility that Hmox1 up-regulation is influenced by, and may in fact act to suppress a functional NADPH oxidase in AM. Further concerning the NADPH oxidase complex, the lack of up-regulation observed for Cybb (encoding the gp91phox subunit of flavocytochrome b558) following exposure to conidia is consistent with a report in human monocytes exposed to conidia (6), although an increase in expression of this same gene was previously observed in macrophages or macrophage cell lines in response to LPS (18), Mycobacterium tuberculosis (77), Chlamydia pneumoniae (78), and TNF-␣ (31). Our DNA microarray results differed in two respects when compared with an in vivo study on peritoneal macrophages in response to S. pyogenes. Specifically, the peritoneal macrophages did not show increased transcription of

6865 Hmox1 after contact with the bacterium, but showed both transcriptional up-regulation and functional evidence of p47phox of the NADPH oxidase (38). Despite the requirement of a functional NADPH oxidase complex in humans for resistance to A. fumigatus, conflicting reports exist concerning both the level of superoxide generated by AM and the role of superoxide in fungal killing by AM (12, 14 –17). It is possible that differences in phagocyte types, animal strains, and cell sources have contributed to contrasting interpretations in these studies. In the present study using a sensitive luminometry assay system, we identified low but significant NADPH oxidase activity in adherent AM from C57BL/6 mice. We also observed this adherence-dependent superoxide production in AM incubated in 96well tissue culture plates using NBT to detect oxidant release (data not shown). The luminometry signal showing superoxide produced by adherent AM was increased only slightly by exposure to 910 nM PMA (a concentration that did not elicit an increased signal in AM from gp91phox⫺/⫺ mice), which was strongly in contrast to the behavior of PMN where robust superoxide generation was observed in the presence of PMA. An almost 2-log difference between superoxide production by PMA-simulated AM relative to that in PMN suggests that even small numbers of PMN that contaminate AM samples could lead to a significant overestimation of superoxide generation in AM. Thus, care was taken in the present study to isolate AM samples before recruitment of PMN into the lung, and to remove nonadherent PMN from AM samples during a medium replacement before luminometry analysis. Using the luminometry assay system outlined above, we found no increase in superoxide generation in AM incubated 90 –180 min with A. fumigatus conidia in vitro. In fact, the presence of a five-fold excess of conidia abolished all SOD-inhibitable NADPH oxidase activity in adherent AM. It is not known if this result is related to the reduced oxidant production from PMN following exposure to A. fumigatus hyphal culture filtrates (79). Additional studies are therefore required to better understand the mechanism by which conidia influence the luminometry signal resulting from superoxide produced by these phagocytes. Importantly, we observed a lack of formazan deposition within AM following phagocytosis of conidia in vivo, which supports our hypothesis that AM do not use NADPH oxidase activity to neutralize A. fumigatus conidia. The lack of obvious superoxide production in AM containing conidia was in contrast to results obtained with PMN aggregates, where abundant formazan deposition is readily apparent in the vicinity of engaged conidia, and is of interest in light of increased production of TNF-␣ (a known priming agent for superoxide production) by AM. It is possible that mechanisms in AM have evolved to suppress NADPH oxidase activity in vivo, thus reducing tissue damage triggered by inhalation of airborne particulates. Because AM appear able to neutralize A. fumigatus conidia by NADPH oxidase-independent mechanisms, it is unclear why AM do not appear capable of neutralizing the conidial exposure that predisposes IPA in neutropenic or chronic granulomatous disease patients. Thus, additional studies are required to better determine which microbicidal mechanisms are overcome in AM, PMN, or possibly nonphagocytic cells (80), that enable conidia to germinate and lead to infections in immune suppressed individuals. In summary, our results indicate that in murine AM exposed to A. fumigatus conidia in vivo, the most marked early transcriptional changes occur in genes involved in PMN recruitment. This cooperation between AM and PMN in aspergillosis immunity appears to be vital because both PMN and NADPH oxidase activity are essential for preventing IPA (13, 81). As an apparent balance to

6866 this inflammatory response in AM, Socs3 and Rgs1 are also upregulated, presumably moderating the inflammatory response to reduce damage to the lung. Of note, was also an observed paucity of conidia-induced transcription of AM genes encoding proteins associated with phagocytosis, pattern recognition receptors, oxidant scavengers and specific microbicidal systems, which have sometimes been described in leukocytes following in vitro exposure to conidia. We observed conserved transcriptional responses related to innate immunity in AM of gp91phox⫺/⫺ mice in response to conidia, despite marked differences in constitutive transcription compared with control C57BL/6 mice. In addition, lack of functional NADPH oxidase in gp91phox⫺/⫺ mice exposed to conidia did not significantly affect the phagocytic ability of AM or their ability to inhibit germination of internalized conidia. The findings in this study support the concept that a lack of superoxide production by the NADPH oxidase in PMN, rather than AM, is the key defect that contributes susceptibility to IPA in chronic granulomatous disease patients (10).

AM RESISTANCE TO ASPERGILLOSIS

20.

21.

22.

23.

24.

25.

26.

Disclosures The authors have no financial conflict of interest.

27.

References

28.

1. Ampel, N. M. 1996. Emerging disease issues and fungal pathogens associated with HIV infection. Emerg. Infect. Dis. 2: 109 –116. 2. Maertens, J., M. Vrebos, and M. Boogaerts. 2001. Assessing risk factors for systemic fungal infections. Eur. J. Cancer Care 10: 56 – 62. 3. Dykewicz, C. A. 2001. Hospital infection control in hematopoietic stem cell transplant recipients. Emerg. Infect. Dis. 7: 263–267. 4. Walsh, T. J., and A. H. Groll. 1999. Emerging fungal pathogens: evolving challenges to immunocompromised patients for the twenty-first century. Transpl. Infect. Dis. 1: 247–261. 5. Latge, J. P. 2001. The pathobiology of Aspergillus fumigatus. Trends Microbiol. 9: 382–389. 6. Cortez, K. J., C. A. Lyman, S. Kottilil, H. S. Kim, E. Roilides, J. Yang, B. Fullmer, R. Lempicki, and T. J. Walsh. 2006. Functional genomics of innate host defense molecules in normal human monocytes in response to Aspergillus fumigatus. Infect. Immun. 74: 2353–2365. 7. Brummer, E., J. H. Choi, and D. A. Stevens. 2005. Interaction between conidia, lung macrophages, immunosuppressants, proinflammatory cytokines and transcriptional regulation. Med. Mycol. 43(Suppl. 1): S177–S179. 8. Simitsopoulou, M., E. Roilides, C. Likartsis, J. Ioannidis, A. Orfanou, F. Paliogianni, and T. J. Walsh. 2007. Expression of immunomodulatory genes in human monocytes induced by voriconazole in the presence of Aspergillus fumigatus. Antimicrob. Agents Chemother. 51: 1048 –1054. 9. Forman, H. J., H. Zhou, E. Gozal, and M. Torres. 1998. Modulation of the alveolar macrophage superoxide production by protein phosphorylation. Environ. Health Perspect. 106(Suppl. 5): 1185–1190. 10. Johnston, R. B., Jr. 2001. Clinical aspects of chronic granulomatous disease. Curr. Opin. Hematol. 8: 17–22. 11. Almyroudis, N. G., S. M. Holland, and B. H. Segal. 2005. Invasive aspergillosis in primary immunodeficiencies. Med. Mycol. 43(Suppl. 1): S247–S259. 12. Morgenstern, D. E., M. A. Gifford, L. L. Li, C. M. Doerschuk, and M. C. Dinauer. 1997. Absence of respiratory burst in X-linked chronic granulomatous disease mice leads to abnormalities in both host defense and inflammatory response to Aspergillus fumigatus. J. Exp. Med. 185: 207–218. 13. Bonnett, C. R., E. J. Cornish, A. G. Harmsen, and J. B. Burritt. 2006. Early neutrophil recruitment and aggregation in the murine lung inhibit germination of Aspergillus fumigatus Conidia. Infect. Immun. 74: 6528 – 6539. 14. Schaffner, A., H. Douglas, A. I. Braude, and C. E. Davis. 1983. Killing of Aspergillus spores depends on the anatomical source of the macrophage. Infect. Immun. 42: 1109 –1115. 15. Philippe, B., O. Ibrahim-Granet, M. C. Prevost, M. A. Gougerot-Pocidalo, P. M. Sanchez, M. A. Van Der, and J. P. Latge. 2003. Killing of Aspergillus fumigatus by alveolar macrophages is mediated by reactive oxidant intermediates. Infect. Immun. 71: 3034 –3042. 16. Roilides, E., T. Sein, A. Holmes, S. Chanock, C. Blake, P. A. Pizzo, and T. J. Walsh. 1995. Effects of macrophage colony-stimulating factor on antifungal activity of mononuclear phagocytes against Aspergillus fumigatus. J. Infect. Dis. 172: 1028 –1034. 17. Nessa, K., L. Palmberg, U. Johard, P. Malmberg, C. Jarstrand, and P. Camner. 1997. Reaction of human alveolar macrophages to exposure to Aspergillus fumigatus and inert particles. Environ. Res. 75: 141–148. 18. Wells, C. A., T. Ravasi, R. Sultana, K. Yagi, P. Carninci, H. Bono, G. Faulkner, Y. Okazaki, J. Quackenbush, D. A. Hume, and P. A. Lyons. 2003. Continued discovery of transcriptional units expressed in cells of the mouse mononuclear phagocyte lineage. Genome Res. 13: 1360 –1365. 19. Li, J., D. K. Pritchard, X. Wang, D. R. Park, R. E. Bumgarner, S. M. Schwartz, and W. C. Liles. 2007. cDNA microarray analysis reveals fundamental differ-

29.

30.

31.

32.

33.

34.

35.

36.

37. 38.

39.

40. 41.

42.

43.

44.

45.

ences in the expression profiles of primary human monocytes, monocyte-derived macrophages, and alveolar macrophages. J. Leukocyte Biol. 81: 328 –335. Srivastava, M., S. Jung, J. Wilhelm, L. Fink, F. Buhling, T. Welte, R. M. Bohle, W. Seeger, J. Lohmeyer, and U. A. Maus. 2005. The inflammatory versus constitutive trafficking of mononuclear phagocytes into the alveolar space of mice is associated with drastic changes in their gene expression profiles. J. Immunol. 175: 1884 –1893. Woodruff, P. G., L. L. Koth, Y. H. Yang, M. W. Rodriguez, S. Favoreto, G. M. Dolganov, A. C. Paquet, and D. J. Erle. 2005. A distinctive alveolar macrophage activation state induced by cigarette smoking. Am. J. Respir. Crit. Care Med. 172: 1383–1392. Steele, C., R. R. Rapaka, A. Metz, S. M. Pop, D. L. Williams, S. Gordon, J. K. Kolls, and G. D. Brown. 2005. The ␤-glucan receptor dectin-1 recognizes specific morphologies of Aspergillus fumigatus. PLoS. Pathog. 1: e42. Hohl, T. M., H. L. Van Epps, A. Rivera, L. A. Morgan, P. L. Chen, M. Feldmesser, and E. G. Pamer. 2005. Aspergillus fumigatus triggers inflammatory responses by stage-specific ␤-glucan display. PLoS. Pathog. 1: e30. Alberts, R., P. Terpstra, M. Hardonk, L. V. Bystrykh, H. G. de, R. Breitling, J. P. Nap, and R. C. Jansen. 2007. A verification protocol for the probe sequences of Affymetrix genome arrays reveals high probe accuracy for studies in mouse, human and rat. BMC Bioinformatics 8: 132–142. Bolstad, B. M., R. A. Irizarry, M. Astrand, and T. P. Speed. 2003. A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics 19: 185–193. Dennis, G., Jr., B. T. Sherman, D. A. Hosack, J. Yang, W. Gao, H. C. Lane, and R. A. Lempicki. 2003. DAVID: database for annotation, visualization, and integrated discovery. Genome Biol. 4: 3. Pfaffl, M. W. 2001. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 29: 2002–2007. Livak, K. J., and T. D. Schmittgen. 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2⫺⌬⌬C(T) method. Methods 25: 402– 408. Radonic, A., S. Thulke, I. M. Mackay, O. Landt, W. Siegert, and A. Nitsche. 2004. Guideline to reference gene selection for quantitative real-time PCR. Biochem. Biophys. Res. Commun. 313: 856 – 862. Thellin, O., W. Zorzi, B. Lakaye, B. B. De, B. Coumans, G. Hennen, T. Grisar, A. Igout, and E. Heinen. 1999. Housekeeping genes as internal standards: use and limits. J. Biotechnol. 75: 291–295. Gauss, K. A., L. K. Nelson-Overton, D. W. Siemsen, Y. Gao, F. R. DeLeo, and M. T. Quinn. 2007. Role of NF-␬B in transcriptional regulation of the phagocyte NADPH oxidase by tumor necrosis factor-␣. J. Leukocyte Biol. 82: 729 –741. Phillips, W. A., M. Croatto, and J. A. Hamilton. 1990. Priming the macrophage respiratory burst with IL-4: enhancement with TNF-␣ but inhibition by IFN-␥. Immunology 70: 498 –503. Pfaffl, M. W., G. W. Horgan, and L. Dempfle. 2002. Relative expression software tool (REST) for group-wise comparison and statistical analysis of relative expression results in real-time PCR. Nucleic Acids Res. 30: e36. Calvano, S. E., W. Xiao, D. R. Richards, R. M. Felciano, H. V. Baker, R. J. Cho, R. O. Chen, B. H. Brownstein, J. P. Cobb, S. K. Tschoeke, et al. 2005. A networkbased analysis of systemic inflammation in humans. Nature 437: 1032–1037. Rinaldi, M., P. Moroni, M. J. Paape, and D. D. Bannerman. 2007. Evaluation of assays for the measurement of bovine neutrophil reactive oxygen species. Vet. Immunol. Immunopathol. 115: 107–125. Gwinn, M. R., and V. Vallyathan. 2006. Respiratory burst: role in signal transduction in alveolar macrophages. J. Toxicol. Environ. Health B Crit. Rev. 9: 27–39. Buetler, T. M., A. Krauskopf, and U. T. Ruegg. 2004. Role of superoxide as a signaling molecule. News Physiol. Sci. 19: 120 –123. Goldmann, O., M. von Kockritz-Blickwede, C. Holtje, G. S. Chhatwal, R. Geffers, and E. Medina. 2007. Transcriptome analysis of murine macrophages in response to infection with Streptococcus pyogenes reveals an unusual activation program. Infect. Immun. 75: 4148 – 4157. Gersuk, G. M., D. M. Underhill, L. Zhu, and K. A. Marr. 2006. Dectin-1 and TLRs permit macrophages to distinguish between different Aspergillus fumigatus cellular states. J. Immunol. 176: 3717–3724. Finkel, T. 2000. Redox-dependent signal transduction. FEBS Lett. 476: 52–54. Quinn, M. T., M. C. Ammons, and F. R. DeLeo. 2006. The expanding role of NADPH oxidases in health and disease: no longer just agents of death and destruction. Clin. Sci. Lond. 111: 1–20. Kobayashi, S. D., J. M. Voyich, K. R. Braughton, A. R. Whitney, W. M. Nauseef, H. L. Malech, and F. R. DeLeo. 2004. Gene expression profiling provides insight into the pathophysiology of chronic granulomatous disease. J. Immunol. 172: 636 – 643. Romani, L., F. Fallarino, L. A. De, C. Montagnoli, C. D’Angelo, T. Zelante, C. Vacca, F. Bistoni, M. C. Fioretti, U. Grohmann, et al. 2008. Defective tryptophan catabolism underlies inflammation in mouse chronic granulomatous disease. Nature 451: 211–215. Duong, M., N. Ouellet, M. Simard, Y. Bergeron, M. Olivier, and M. G. Bergeron. 1998. Kinetic study of host defense and inflammatory response to Aspergillus fumigatus in steroid-induced immunosuppressed mice. J. Infect. Dis. 178: 1472–1482. Kim, H. S., E. H. Choi, J. Khan, E. Roilides, A. Francesconi, M. Kasai, T. Sein, R. L. Schaufele, K. Sakurai, C. G. Son, et al. 2005. Expression of genes encoding innate host defense molecules in normal human monocytes in response to Candida albicans. Infect. Immun. 73: 3714 –3724.

The Journal of Immunology 46. Mehrad, B., R. M. Strieter, T. A. Moore, W. C. Tsai, S. A. Lira, and T. J. Standiford. 1999. CXC chemokine receptor-2 ligands are necessary components of neutrophil-mediated host defense in invasive pulmonary aspergillosis. J. Immunol. 163: 6086 – 6094. 47. Mehrad, B., R. M. Strieter, and T. J. Standiford. 1999. Role of TNF-␣ in pulmonary host defense in murine invasive aspergillosis. J. Immunol. 162: 1633–1640. 48. Dinarello, C. A. 1996. Biologic basis for interleukin-1 in disease. Blood 87: 2095–2147. 49. Haelens, A., A. Wuyts, P. Proost, S. Struyf, G. Opdenakker, and J. van Damme. 1996. Leukocyte migration and activation by murine chemokines. Immunobiology 195: 499 –521. 50. Sozzani, S., W. Luini, A. Borsatti, N. Polentarutti, D. Zhou, L. Piemonti, G. D’Amico, C. A. Power, T. N. Wells, M. Gobbi, et al. 1997. Receptor expression and responsiveness of human dendritic cells to a defined set of CC and CXC chemokines. J. Immunol. 159: 1993–2000. 51. Shimada, T., M. Matsumoto, Y. Tatsumi, A. Kanamaru, and S. Akira. 1998. A novel lipopolysaccharide inducible C-C chemokine receptor related gene in murine macrophages. FEBS Lett. 425: 490 – 494. 52. Filler, S. G., M. R. Yeaman, and D. C. Sheppard. 2005. Tumor necrosis factor inhibition and invasive fungal infections. Clin. Infect. Dis. 41(Suppl. 3): S208 –S212. 53. Shalit, I., D. Halperin, D. Haite, A. Levitov, J. Romano, N. Osherov, and I. Fabian. 2006. Anti-inflammatory effects of moxifloxacin on IL-8: IL-1␤ and TNF-␣ secretion and NF␬B and MAP-kinase activation in human monocytes stimulated with Aspergillus fumigatus. J. Antimicrob. Chemother. 57: 230 –235. 54. Choi, J. H., E. Brummer, Y. J. Kang, P. P. Jones, and D. A. Stevens. 2006. Inhibitor ␬B and nuclear factor ␬B in granulocyte-macrophage colony-stimulating factor antagonism of dexamethasone suppression of the macrophage response to Aspergillus fumigatus conidia. J. Infect. Dis. 193: 1023–1028. 55. Benjamim, C. F., C. M. Hogaboam, N. W. Lukacs, and S. L. Kunkel. 2003. Septic mice are susceptible to pulmonary aspergillosis. Am. J. Pathol. 163: 2605–2617. 56. Romani, L., F. Bistoni, R. Gaziano, S. Bozza, C. Montagnoli, K. Perruccio, L. Pitzurra, S. Bellocchio, A. Velardi, G. Rasi, et al. 2004. Thymosin ␣1 activates dendritic cells for antifungal Th1 resistance through Toll-like receptor signaling. Blood 103: 4232– 4239. 57. Shahan, T. A., W. G. Sorenson, J. D. Paulauskis, R. Morey, and D. M. Lewis. 1998. Concentration- and time-dependent upregulation and release of the cytokines MIP-2. KC, TNF, and MIP-1␣ in rat alveolar macrophages by fungal spores implicated in airway inflammation. Am. J. Respir. Cell Mol. Biol. 18: 435– 440. 58. Jiang, Y., D. I. Beller, G. Frendl, and D. T. Graves. 1992. Monocyte chemoattractant protein-1 regulates adhesion molecule expression and cytokine production in human monocytes. J. Immunol. 148: 2423–2428. 59. Wasylnka, J. A., and M. M. Moore. 2002. Uptake of Aspergillus fumigatus conidia by phagocytic and nonphagocytic cells in vitro: quantitation using strains expressing green fluorescent protein. Infect. Immun. 70: 3156 –3163. 60. Taramelli, D., M. G. Malabarba, G. Sala, N. Basilico, and G. Cocuzza. 1996. Production of cytokines by alveolar and peritoneal macrophages stimulated by Aspergillus fumigatus conidia or hyphae. J. Med. Vet. Mycol. 34: 49 –56. 61. Schelenz, S., D. A. Smith, and G. J. Bancroft. 1999. Cytokine and chemokine responses following pulmonary challenge with Aspergillus fumigatus: obligatory role of TNF-␣ and GM-CSF in neutrophil recruitment. Med. Mycol. 37: 183–194. 62. Gygi, S. P., Y. Rochon, B. R. Franza, and R. Aebersold. 1999. Correlation between protein and mRNA abundance in yeast. Mol. Cell. Biol. 19: 1720 –1730. 63. Denecke, B., A. Meyerdierks, and E. C. Bottger. 1999. RGS1 is expressed in monocytes and acts as a GTPase-activating protein for G-protein-coupled chemoattractant receptors. J. Biol. Chem. 274: 26860 –26868.

6867 64. Johnston, J. A., and J. J. O’Shea. 2003. Matching SOCS with function. Nat. Immunol. 4: 507–509. 65. Kan, V. L., and J. E. Bennett. 1988. Lectin-like attachment sites on murine pulmonary alveolar macrophages bind Aspergillus fumigatus conidia. J. Infect. Dis. 158: 407– 414. 66. Marshall, A. S., and S. Gordon. 2004. Commentary: C-type lectins on the macrophage cell surface-recent findings. Eur. J. Immunol. 34: 18 –24. 67. Shirakata, Y., T. Komurasaki, H. Toyoda, Y. Hanakawa, K. Yamasaki, S. Tokumaru, K. Sayama, and K. Hashimoto. 2000. Epiregulin, a novel member of the epidermal growth factor family, is an autocrine growth factor in normal human keratinocytes. J. Biol. Chem. 275: 5748 –5753. 68. Takahashi, M., K. Hayashi, K. Yoshida, Y. Ohkawa, T. Komurasaki, A. Kitabatake, A. Ogawa, W. Nishida, M. Yano, M. Monden, and K. Sobue. 2003. Epiregulin as a major autocrine/paracrine factor released from ERK- and p38MAPK-activated vascular smooth muscle cells. Circulation 108: 2524 –2529. 69. Coleman, D. L., A. H. Bartiss, V. P. Sukhatme, J. Liu, and H. D. Rupprecht. 1992. Lipopolysaccharide induces Egr-1 mRNA and protein in murine peritoneal macrophages. J. Immunol. 149: 3045–3051. 70. Yan, S. F., J. Lu, L. Xu, Y. S. Zou, J. Tongers, W. Kisiel, N. Mackman, D. J. Pinsky, and D. M. Stern. 2000. Pulmonary expression of early growth response-1: biphasic time course and effect of oxygen concentration. J. Appl. Physiol. 88: 2303–2309. 71. Kamimura, M., C. Viedt, A. Dalpke, M. E. Rosenfeld, N. Mackman, D. M. Cohen, E. Blessing, M. Preusch, C. M. Weber, J. Kreuzer, et al. 2005. Interleukin-10 suppresses tissue factor expression in lipopolysaccharide-stimulated macrophages via inhibition of Egr-1 and a serum response element/MEKERK1/2 pathway. Circ. Res. 97: 305–313. 72. Thyss, R., V. Virolle, V. Imbert, J. F. Peyron, D. Aberdam, and T. Virolle. 2005. NF-␬B/Egr-1/Gadd45 are sequentially activated upon UVB irradiation to mediate epidermal cell death. EMBO J. 24: 128 –137. 73. Zaas, A. K. 2006. Host genetics affect susceptibility to invasive aspergillosis. Med. Mycol. 44(Suppl.): 55– 60. 74. Ryter, S. W., J. Alam, and A. M. Choi. 2006. Heme oxygenase-1/carbon monoxide: from basic science to therapeutic applications. Physiol. Rev. 86: 583– 650. 75. Fredenburgh, L. E., M. A. Perrella, and S. A. Mitsialis. 2007. The role of heme oxygenase-1 in pulmonary disease. Am. J. Respir. Cell Mol. Biol. 36: 158 –165. 76. Taille, C., J. El-Benna, S. Lanone, M. C. Dang, E. Ogier-Denis, M. Aubier, and J. Boczkowski. 2004. Induction of heme oxygenase-1 inhibits NAD(P)H oxidase activity by down-regulating cytochrome b558 expression via the reduction of heme availability. J. Biol. Chem. 279: 28681–28688. 77. Wang, J. P., S. E. Rought, J. Corbeil, and D. G. Guiney. 2003. Gene expression profiling detects patterns of human macrophage responses following Mycobacterium tuberculosis infection. FEMS Immunol. Med. Microbiol. 39: 163–172. 78. Azenabor, A. A., S. Yang, G. Job, and O. O. Adedokun. 2005. Elicitation of reactive oxygen species in Chlamydia pneumoniae-stimulated macrophages: a Ca2⫹-dependent process involving simultaneous activation of NADPH oxidase and cytochrome oxidase genes. Med. Microbiol. Immunol. 194: 91–103. 79. Sugui, J. A., J. Pardo, Y. C. Chang, A. Mullbacher, K. A. Zarember, E. M. Galvez, L. Brinster, P. Zerfas, J. I. Gallin, M. M. Simon, and K. J. Kwon-Chung. 2007. Role of laeA in the regulation of alb1, gliP, conidial morphology and virulence in Aspergillus fumigatus. Eukaryot. Cell 1552–1561. 80. Filler, S. G., and D. C. Sheppard. 2006. Fungal invasion of normally non-phagocytic host cells. PLoS. Pathog. 2: e129. 81. Stephens-Romero, S. D., A. J. Mednick, and M. Feldmesser. 2005. The pathogenesis of fatal outcome in murine pulmonary aspergillosis depends on the neutrophil depletion strategy. Infect. Immun. 73: 114 –125.