Murine Invasive Aspergillosis in Pulmonary Host Defense in Α Role

Home , Pidotimod

Role of TNF-α in Pulmonary Host Defense in Murine Invasive Aspergillosis Borna Mehrad, Robert M. Strieter and Theodore J. Standiford This information is current as J Immunol 1999; 162:1633-1640; ; of September 24, 2021. http://www.jimmunol.org/content/162/3/1633 Downloaded from References This article cites 29 articles, 8 of which you can access for free at: http://www.jimmunol.org/content/162/3/1633.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/ • Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 24, 2021 Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1999 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Role of TNF-␣ in Pulmonary Host Defense in Murine Invasive Aspergillosis1

Borna Mehrad, Robert M. Strieter, and Theodore J. Standiford2

Invasive pulmonary aspergillosis is a common and devastating complication of immunosuppression, whose incidence has increased dramatically in tandem with the increase in the number of immunocompromised patients. Given the role of TNF-␣ in other pulmonary infections, we hypothesized that TNF-␣ is an important proximal signal in murine invasive pulmonary aspergillosis. Intratracheal challenge with Aspergillus fumigatus conidia in both neutropenic (cyclophosphamide-treated) and nonneutropenic BALB/c mice resulted in the time-dependent increase in lung TNF-␣ levels, which correlated with the histologic development of a patchy, peribronchial infiltration of mononuclear and polymorphonuclear cells. Ab-mediated neutralization of TNF-␣ resulted in an increase in mortality in both normal and cyclophosphamide-treated animals, which was associated with increased lung fungal burden as determined by histology and as quantified by chitin content. Depletion of TNF-␣ resulted in a reduced lung Downloaded from neutrophil influx in both normal and cyclophosphamide-treated animals, which occurred in association with a decrease in lung levels of the C-X-C chemokine, macrophage inflammatory protein-2 and the C-C chemokines macrophage inflammatory pro- ␣ ␣ tein-1 and JE. In cyclophosphamide-treated animals, intratracheal administration of a TNF- agonist peptide (TNF70–80) 3 days before, but not concomitant with, the administration of Aspergillus conidia resulted in improved survival from 9% in control mice ␣ to 55% in TNF70–80-treated animals. These studies indicate that TNF- is a critical component of innate immunity in both immunocompromised and immunocompetent hosts, and that pretreatment with a TNF-␣ agonist peptide in a compartmentalized http://www.jimmunol.org/ fashion can significantly enhance resistance to A. fumigatus in neutropenic animals. The Journal of Immunology, 1999, 162: 1633–1640.

nvasive aspergillosis is a feared complication of immunode- and IL-1 (13, 14), as well as cytokines that promote Th-1 pheno- ficiency. Its incidence has increased dramatically over the type immune responses (15). I past 2 decades, in tandem with an increase in the number of TNF-␣ is a 17-kDa cytokine secreted predominantly by various immunocompromised hosts (1, 2). Invasive pulmonary aspergillo- macrophage populations, including alveolar macrophages. It has sis, the commonest form of the disease (3), carries a crude mor- shown to be a critical proximal signal in the initiation and main- tality rate of Ͼ80% with current therapy (4). Aspergillus species tenance of innate pulmonary immunity in animal models of pneu- by guest on September 24, 2021 are the second commonest fungal pathogen in immunocompro- monia (16, 17) and human pneumonia (18). In the single study mised hosts, and A. fumigatus is responsible for 90% of cases of examining the role of TNF-␣ in invasive aspergillosis in vivo, invasive aspergillosis. immunocompetent animals administered A. fumigatus conidia i.v. The respiratory tract is the portal of entry in the majority of had improved survival when treated with systemic murine rTNF-␣ cases of human invasive aspergillosis (3). It is postulated that in- (19). The role of TNF-␣ in pulmonary host defense against A. nate immunity is the principal pathway by which A. fumigatus is fumigatus has not been defined. cleared from the lung. This innate response consists of alveolar Given that TNF-␣ has been shown to be a critical mediator of macrophages, which represent the first line of defense against innate immunity against several respiratory pathogens, attempts conidia entering the alveolus, and recruited neutrophils. Neutro- have been made to administer TNF-␣ to augment innate and ac- phils are believed to kill conidia that have survived to form hyphae quired host responses. However, the therapeutic administration of and cause invasive infection (5). Various cells of macrophages TNF-␣ is limited by substantial dose-related toxicity, particularly lineage can engulf and kill Aspergillus conidia in vitro (6–8) and when this cytokine is administered systemically. A TNF-␣ agonist in vivo (9, 10). Both macrophages (6, 11) and polymorphonuclear peptide composed of the 11 amino acids that constitute the binding cells (5, 12) have been shown to damage hyphae in vitro. When site of native human TNF-␣ to its receptors (referred to as 3 exposed to conidia in vitro, cells of macrophage lineage secrete TNF70–80) has recently been characterized (20, 21). Binding of ␣ ␣ inflammatory mediators, including the proximal cytokines TNF- TNF70–80 to TNF- receptors (both p55 and p75) has been shown to mediate many leukocyte-activating effects of native TNF-␣, including stimulation of neutrophils for enhanced pro- Department of Medicine, Division of Pulmonary and Critical Care Medicine, Uni- versity of Michigan Medical School, Ann Arbor, MI 48109 tease release and respiratory burst, and enhancement of neutrophil phagocytic activity and killing. When administered sys- Received for publication July 6, 1998. Accepted for publication October 8, 1998. temically, TNF70–80 was associated with less toxicity than that 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 observed with native TNF- , due in part to the fact that this with 18 U.S.C. Section 1734 solely to indicate this fact. peptide did not alter adhesive properties of the endothelium 1 This work was supported in part by National Institutes of Health Grants HL57243, (20). This peptide has been successfully administered directly HL58200, and P50HL60289. 2 Address correspondence and reprint requests to Dr. Theodore J. Standiford, Division 3 ␣ ␣ of Pulmonary and Critical Care Medicine, University of Michigan Medical Center, Abbreviations used in this paper: TNF70–80, TNF- agonist peptide; MIP-1 ; mac- 6301 MSRB III, 1150 West Medical Center Drive, Ann Arbor, MI 48109-0360. rophage inflammatory protein-1␣; MIP-2, macrophage inflammatory protein-2; i.t., E-mail address: [email protected] intratracheal; MPO, myeloperoxidase; CCR-1, C-C chemokine receptor-1.

into the lungs of mice with Gram-negative bacterial pneumonia, was perfused with PBS via the right ventricle. For histologic examination, lungs were perfused with 1 ml of 4% paraformaldehyde in PBS, inflated although the beneficial effects of TNF70–80 on bacterial clearance and survival were only observed when the peptide was with 1 ml of 4% paraformaldehyde in PBS via the trachea, and then excised en bloc. Lungs for various assays were perfused with 1 ml of PBS con- administered 3 or 7 days before, but not concomitant with, bac- taining 5 mM EDTA, removed, frozen in liquid nitrogen, and stored at terial challenge (22). Ϫ20°C until the day of the assay. Lungs for cytokine and myeloperoxidase In the current study we hypothesized that TNF-␣ is an important (MPO) assays were homogenized in 1 ml of 2ϫ complete protease inhib- proximal signal in murine invasive pulmonary aspergillosis. We itor mixture buffer (Boehringer Mannheim, Mannheim, Germany) in PBS ␣ using a tissue homogenizer (Biospec Products, Bartlesville, OK). A 900-␮l tested this hypothesis by assessing the production of TNF- in the aliquot of PBS was added to 900 ␮l from each sample, sonicated for 10 s, lungs of normal and immunocompromised mice challenged with and centrifuged at 500 ϫ g for 10 min. Supernatants were passed through A. fumigatus conidia. We next examined the effect of TNF-␣ de- a 0.45-␮m pore size filter (Gelman, Ann Arbor, MI) and stored at 4°C for pletion on survival, the burden of fungal organisms, and the host cytokine ELISA. inflammatory response. Finally, the effects of intrapulmonary de- Lung chitin assay livery of native TNF-␣ or TNF on survival in cyclophos- 70–80 Given that molds (including the Aspergillus species) do not reliably form phamide-treated animals were determined. reproductive units in tissue, we employed an assay for chitin to measure the burden of organisms in lungs. Chitin, a component of the hyphal wall, is Materials and Methods absent from mammalian cells and conidia. The assay was adapted from a Reagents previously described method (24). Lungs were homogenized in 5 ml of distilled water and centrifuged (1500 ϫ g, 5 min, 20°C). The supernatants ␣ ␣ Polyclonal anti-murine TNF- , MIP-1 , JE, MIP-2, and KC Abs used in were discarded, and pellets were resuspended in sodium lauryl sulfate (3%, Downloaded from the ELISAs were produced by immunization of rabbits with murine re- w/v) and heated at 100°C for 15 min. Samples were then centrifuged combinant cytokines in multiple intradermal sites with CFA. Carrier-free (1500 ϫ g, 5 min, 20°C), and pellets were washed with distilled water and ␣ murine rTNF, MIP-1 , JE, MIP-2, and KC were purchased from R&D resuspended in 3 ml of KOH (120%, w/v). Samples were heated at 130°C ␣ Systems (Minneapolis, MN). In TNF- neutralization experiments, 0.5 ml for 60 min. After cooling, 8 ml of ice-cold ethanol (75%, v/v) was added ␣ of rabbit anti-murine TNF- serum or control rabbit serum was adminis- to each sample, and tubes were shaken until ethanol and KOH made one tered i.p. 1 day before A. fumigatus administration. The i.p. injections of phase. Samples were incubated on ice for 15 min, and 0.3 ml of Celite 0.25 ml of antiserum or control rabbit serum were repeated every 48 h for suspension (supernatant of1gofCelite 545 (Fisher Scientific, Pittsburgh,

two additional doses. TNF70–80 and control peptides were synthesized at PA) added to 75% ethanol and allowed to stand for 2 min) was added to http://www.jimmunol.org/ the University of Michigan Protein and Carbohydrate Structure Facility each. Samples were centrifuged (1500 ϫ g, 5 min, 4°C), and supernatants and were composed of the amino acid sequences H-Pro-Ser-Thr-His-Val- were discarded. Pellets were washed once with ethanol (40%, v/v) and Leu-Ileu-Thr-His-Thr-Ileu-OH and H-Gly-Gly-Asp-Pro-Gly-Ileu-Val-Thr- twice with distilled water, and resuspended in 0.5 ml of distilled water. His-Ser-OH, respectively (20). Both peptides were purified by HPLC and Standards, consisting of 0.2 ml of distilled water and 0.2 ml of glucosamine mass spectrometry, and contained no detectable LPS, as determined by (10 ␮g/ml), were made up. NaNO (0.2 ml; 5%, w/v) and KHSO (0.2 ml; Limulus lysate assay (ICN Biomedicals, Costa Mesa, CA). 2 4 5%, w/v) were added to each standard, and NaNO2 (0.5 ml; 5%, w/v) and Animals KHSO4 (0.5 ml; 5%, w/v) were added to each tissue prep; all samples were mixed gently for 15 min and then centrifuged (1500 ϫ g, 2 min, 4°C). Two Specific pathogen-free BALB/c mice (8-wk-old females; Charles River 0.6-ml aliquots of supernatant from each tissue prep were transferred to Breeding Laboratories, Wilmington, MA) were used in all experiments. All separate tubes. Ammonium sulfamate (0.2 ml) was added to each tube, and by guest on September 24, 2021 mice were housed in specific pathogen-free conditions within the animal all tubes were shaken vigorously for 5 min. A fresh solution of 3-methyl- care facility at the University of Michigan until the day of sacrifice. Im- 2-thiazolone hydrazone HCl monohydrate (50 mg in 10 ml of distilled munocompromised animals were treated with 200 mg/kg of cyclophos- water) was made, and 0.2 ml was added to each tube. Samples were then ⅐ phamide (Sigma, St. Louis, MO), administered i.p., 4 days before intra- heated to 100°C for 3 min and cooled. FeCl3 6H2O (0.2 ml; 0.83%, w/v) tracheal (i.t.) inoculation. Treatment with cyclophosphamide resulted in was added to each, and OD was measured at 650 nm after 25 min. Chitin peripheral blood neutropenia (absolute neutrophil count, Ͻ200 cells/␮l) by content, measured in glucosamine equivalents, was measured by the fol- days 4 and 6 after treatment, with a return of peripheral counts to pretreat- lowing formula: chitin content ϭ [5 ϫ (OD of organ Ϫ OD of control ment levels by day 10. organ)]/(OD of glucosamine Ϫ OD of water). Preparation and administration of A. fumigatus conidia Lung MPO activity We chose to use A. fumigatus strain 13073 (American Type Culture Col- Lung MPO activity was measured as a marker of neutrophil sequestration, lection, Manassas, VA) in our studies, as this strain has previously been as described previously (25). Briefly, a 100-␮l aliquot of each lung ho- shown to induce invasive aspergillosis in immunocompromised mice (23). mogenate was added to 100 ␮l of a buffer containing 50 mM potassium The organism was grown on Sabouraud dextrose agar plates (Becton Dick- phosphate (pH 6.0), 5% hexadecyltrimthylammonium bromide, and 5 mM inson, Cockeysville, MD) for 7–10 days at 37°C. The surface of each plate EDTA. Samples were sonicated for 10 s and centrifuged at 3000 ϫg for 15 was then washed with 100 ml of sterile 0.1% Tween-80 (SigmaUltra, St. min. The supernatant was mixed 1/15 with assay buffer and read at 490 nm. Louis, MO) in normal saline. The resulting suspension of conidia was MPO units were calculated as the change in absorbance over time. filtered through sterile gauze to remove clumps and hyphal debris, and then washed once and resuspended in 4 ml of 0.1% Tween-80. The concentra- Cytokine ELISA tion of Aspergillus conidia in the suspension was determined by a particle Murine TNF-␣, MIP-1␣, JE, MIP-2, and KC were quantified using a mod- counter (Z2 particle analyzer, Coulter, Hialeah, FL). The suspension was ification of a double ligand method, as described previously (26). Briefly, then diluted to the desired concentration, and the concentration was again flat-bottom 96-well microtiter plates (Immuno-Plate I 96-F, Nunc, Copen- measured before administration. In preliminary experiments, the number of hagen, Denmark) were coated with 50 ␮l/well of rabbit anti-cytokine Ab (1 particles determined by the particle counter was in close agreement with ␮ g/ml in 0.6 M NaCl, 0.26 M H3BO4, and 0.08 N NaOH, pH 9.6) for 16 h the number of viable colony-forming units found by serial dilution and at 4°C and then washed with PBS (pH 7.5) and 0.05% Tween-20 (wash plating of the suspension. On the day of inoculation, each animal was buffer). Microtiter plate nonspecific binding sites were blocked with 2% anesthetized with 1.8–2 mg of pentobarbital i.p. Using standard aseptic BSA in PBS and incubated for 90 min at 37°C. Plates were rinsed four ␮ technique, the trachea was exposed and a 30- l inoculum (A. fumigatus times with wash buffer, and undiluted and diluted (1/10) cell-free super- suspension or 0.1% Tween-80) was administered via a sterile 26-gauge natants (50 ␮l) in duplicate were added followed by incubation for1hat needle. The skin incision was closed with surgical staples. In all experi- 37°C. Plates were washed four times, followed by the addition of 50 ␮l/ ϫ 7 ments, immunocompetent animals were challenged with 1–2 10 well biotinylated rabbit anti-cytokine Abs (3.5 ␮g/ml in PBS (pH 7.5), conidia, and cyclophosphamide-treated animals were challenged with 0.05% Tween-20, and 2% FCS), and plates were incubated for 30 min at ϫ 6 1–2 10 conidia. 37°C. Plates were washed four times, streptavidin-peroxidase conjugate Lung harvest (Bio-Rad, Richmond, CA) was added, and plates were incubated for 30 min at 37°C. Plates were washed again four times, and chromogen sub-

At designated time points, the mice were sacriﬁced by CO2 asphyxiation. strate (Bio-Rad, Richmond, CA) was added. The plates were incubated at The chest cavity was opened aseptically, and the pulmonary vasculature room temperature to the desired extinction, and the reaction was terminated The Journal of Immunology 1635 Downloaded from http://www.jimmunol.org/

FIGURE 1. Effect of i.t. administration of A. fumigatus conidia on histopathology. A and B depict representative lung hematoxylin and eosin (H&E) and Gomori methanamine silver (GMS) stains in immunocompetent mice 2 days after inoculation with 1–2 ϫ 107 A. fumigatus conidia (magnification, ϫ400). Inflammatory cellular infiltrate occurred in peribronchial areas, corresponding to sites of conidial deposition. Hyphal forms were not present. C and D depict representative lung H&E and GMS stains in cyclophosphamide-treated 2 days after inoculation with 1–2 ϫ 106 A. fumigatus conidia (magnification, ϫ100). Inflammatory cellular infiltrate occurred peribronchially, in areas where branching hyphae were present. The data shown are representative of three experiments.

␮ by guest on September 24, 2021 with 50 l/well of3MH2SO4 solution. Plates were read at 490 nm in an matory response was noted in cyclophosphamide-treated animals ELISA reader. Standards were 1/2 log dilutions of murine rTNF, MIP-1␣, up to 7 days after challenge (data not shown). Conidia were visible JE, MIP-2, or KC, from 1 pg/ml to 100 ng/ml. This ELISA method con- in normal mice 2 days after fungal inoculation and were nearly sistently detected the relevant cytokine at concentrations Ͼ25 pg/ml. The ELISAs did not cross-react with IL-1, IL-2, IL-4, IL-6, IL-10, or IFN-␥.In completely cleared by 5 days. In contrast, hyphae (the invasive addition, each ELISA did not cross-react with other members of the murine form of A. fumigatus) were present only in cyclophosphamide- chemokine family. treated mice and were most numerous by 2–3 days after A. fumigatus administration. The inflammatory infiltrates observed cor- Statistical analysis responded to areas of conidial deposition in normal mice and to Data were analyzed by a Power Macintosh 8600/300 computer using the hyphae in cyclophosphamide-treated mice. InStat version 2.01 statistical package (GraphPad, San Diego, CA). Sur- vival data were compared using the Fisher’s exact test. All other data were Lung TNF-␣ expression after administration of A. fumigatus expressed as the mean Ϯ SEM and compared using unpaired two-tailed Mann-Whitney (nonparametric) test. The p values were considered statis- conidia tically significant if they were Ͻ0.05. We next correlated changes in lung TNF-␣ levels with the development of pulmonary inflammation in normal and cyclophospha- Results mide-treated mice after administration of A. fumigatus conidia. As Development of pulmonary inflammation in normal and shown in Fig. 2A, a rapid increase in lung TNF-␣ levels was noted cyclophosphamide-treated mice after A. fumigatus in immunocompetent mice when they were challenged with 1–2 ϫ administration 107 conidia, which reached a plateau by 24 h. Administration of To characterize the host response to Aspergillus in immunocom- 1–2 ϫ 106 A. fumigatus conidia to cyclophosphamide-treated mice petent and immunocompromised animals, normal and cyclophos- resulted in a marked elevation in lung TNF-␣ levels by 1 day after phamide-treated BALB/c mice were challenged i.t. with 1–2 ϫ 107 administration, with maximal TNF-␣ levels at 2 days, and a return and 1–2 ϫ 106 A. fumigatus conidia, respectively. Lungs were to baseline levels by 4 days after A. fumigatus administration (Fig. examined histologically at various time points after challenge. By 2B). No significant induction of TNF-␣ was observed in the lungs 2 days after A. fumigatus administration in both normal and cy- of immunocompetent animals challenged with 1–2 ϫ 106 A. fu- clophosphamide-treated mice, infiltration of inflammatory cells migatus conidia or in cyclophosphamide-treated animals chal- was noted within the alveolar and interstitial compartments in a lenged with 0.1% Tween-80 (vehicle) i.t. patchy, peribronchial pattern (Fig. 1). The inflammatory cell infil- ␣ trate consisted predominantly of neutrophils and, to a lesser extent, Effect of TNF- neutralization on survival in normal and mononuclear cells. By 4 days after A. fumigatus administration, immunocompromised mice inoculated with A. fumigatus conidia cellular infiltrates were primarily mononuclear, with resolution oc- Additional studies were undertaken to evaluate the contribution of curring in immunocompetent mice, whereas a continued inflam- TNF-␣ to survival after inoculation of conidia. To ascertain the 1636 TNF-␣ IN MURINE INVASIVE ASPERGILLOSIS

FIGURE 3. Effect of TNF-␣ neutralization on survival after A. fumiga- Downloaded from tus challenge in immunocompetent and cyclophosphamide-treated mice. A depicts survival in immunocompetent mice after A. fumigatus conidia administration (1.0 ϫ 107 conidia). B depicts survival in cyclophosphamide- treated mice after A. fumigatus challenge (9.3 ϫ 105 conidia). Cy, cyclo- p Ͻ 0.01 compared with animals ,ء .phosphamide. Experimental n ϭ 15 receiving control rabbit serum. http://www.jimmunol.org/

mal and cyclophosphamide-treated mice receiving control serum. All animals that survived beyond 7 days were considered long term survivors (Ͼ4 wk; data not shown).

Effect of TNF-␣ neutralization on lung fungal burden in normal FIGURE 2. Time-dependent production of TNF-␣ protein in lungs from and cyclophosphamide-treated mice

BALB/c mice after the i.t. administration of vehicle or A. fumigatus To determine the mechanism of increased lethality in mice pas- by guest on September 24, 2021 conidia. A depicts lung TNF-␣ levels in immunocompetent mice after A. sively immunized with anti-TNF-␣ Ab, we examined fungal bur- ϫ 7 ␣ fumigatus (1–2 10 conidia) or vehicle challenge. B depicts lung TNF- den in lungs of infected animals by histology and assessment of levels in cyclophosphamide-treated mice after A. fumigatus challenge (1– lung chitin content. Immunocompetent BALB/c mice or mice 2 ϫ 106 conidia), immunocompetent mice after A. fumigatus challenge treated with cyclophosphamide were given either anti-TNF-␣ Ab (1–2 ϫ 106 conidia), and cyclophosphamide-treated mice after vehicle ϫ 7 -p Ͻ 0.05 com- or rabbit serum 1 day before challenge with 1–2 10 (immuno ,ء .challenge. Cy, cyclophosphamide. Experimental n ϭ 6 ϫ 6 pared with animals receiving vehicle i.t. competent) or 1–2 10 (cyclophosphamide-treated) A. fumigatus conidia. Animals were then sacrificed after 3 days for histology and lung chitin measurement. In immunocompetent mice, treatment with anti-TNF-␣ Ab resulted in the appearance of both effectiveness of the polyclonal anti-murine TNF-␣ Ab in depleting conidial and hyphal forms at 3 days after the administration of TNF-␣ in vivo, we measured lung TNF-␣ levels in normal and conidia, whereas only conidial forms of A. fumigatus were seen in cyclophosphamide-treated animals passively immunized with anti- animals receiving preimmune serum (data not shown). In cyclo- TNF-␣ Ab or control serum 1 day before inoculation with A. fu- phosphamide-treated animals, a marked increase in the number of migatus conidia. On the basis of earlier studies showing time of hyphae was noted in animal treated with anti-TNF-␣ Ab, as evi- maximum expression of TNF, lungs were harvested 1 day after dent on histology, compared with that in control animals (Fig. 4, B inoculation in normal mice and 2 days after inoculation in cyclo- and D). To quantitate differences in fungal burden, lung chitin phosphamide-treated mice. Lung homogenate TNF-␣ levels in levels were determined in cyclophosphamide-treated animals. normal mice were 3.53 Ϯ 0.54 in the control group and 0.13 Ϯ Lungs of TNF-depleted animals had a greater than twofold in- 0.051 in the anti-TNF-␣ group (mean Ϯ SEM; n ϭ 6 mice/group; crease in chitin content compared with those of control animals p Ͻ 0.01). In cyclophosphamide-treated animals, levels were (Fig. 5). 6.16 Ϯ 0.70 in the control group and 0.080 Ϯ 0.007 in anti-TNF-␣ group (mean Ϯ SEM; n ϭ 6 mice/group; p Ͻ 0.01). Effect of TNF-␣ neutralization on recruitment of inflammatory Having demonstrated significant in vivo neutralization of TNF, cells normal and cyclophosphamide-treated animals were treated with To determine whether impaired fungal clearance in TNF-depleted anti-murine TNF-␣ or control rabbit serum 1 day before challenge animals occurred as a result of changes in lung neutrophil influx, with conidia. Normal and cyclophosphamide-treated mice were lungs from cyclophosphamide-treated and normal mice given anti- then inoculated with 1.0 ϫ 107 and 9.3 ϫ 105 conidia, respec- TNF-␣ Ab or control serum were harvested for estimation of MPO tively. As shown in Fig. 3, treatment with anti-murine TNF-␣ Ab activity, a surrogate measure of neutrophil presence. On the basis resulted in 14% mortality in normal mice and 47% mortality in of studies showing the time of maximum expression of MPO in cyclophosphamide-treated mice, compared with no deaths in nor- normal and cyclophosphamide-treated animals (not shown), lung The Journal of Immunology 1637 Downloaded from http://www.jimmunol.org/

FIGURE 4. Effect of TNF-␣ neutralization on lung histopathology after A. fumigatus challenge in cyclophosphamide-treated mice. Cyclophosphamide- treated animals were given rabbit anti-TNF-␣ serum or rabbit control serum 1 day before challenge with A. fumigatus. A and B depict representative lung hematoxylin and eosin (H&E) and Gomori methanamine silver (GMS) stains in cyclophosphamide-treated mice given control serum 3 days after inoculation of 1–2 ϫ 106 A. fumigatus conidia (magniﬁcation, ϫ100). C and D depict representative lung H&E and GMS stains in cyclophosphamide-treated mice given anti-TNF-␣ serum 3 days after inoculation of 1–2 ϫ 106 A. fumigatus conidia (magniﬁcation, ϫ100). Greater number of branching hyphae are visible in animals treated with anti-TNF. The data shown are representative of three experiments. by guest on September 24, 2021 MPO levels were evaluated 1 day after inoculation in normal an- Effect of TNF-␣ neutralization on expression of C-X-C and C-C imals and 2 days after inoculation in cyclophosphamide-treated chemokines animals. In normal animals treated with anti-TNF, there was a 79% decrease in lung MPO levels compared with control values at 1 Given that TNF-␣ has been shown to be a potent inducer of both day, indicating reduced neutrophil recruitment in the absence of C-X-C and C-C chemokines, levels of the C-X-C chemokines TNF. In cyclophosphamide-treated mice, the lung MPO level was MIP-2 and KC, and of the C-C chemokines, MIP-1␣ and JE, were reduced by 45% in TNF-depleted animals compared with that in measured in lung homogenates from normal and cyclophospha- controls at 2 days (Fig. 6). mide-treated mice. On the basis of preliminary studies showing the

FIGURE 6. Effect of TNF-␣ neutralization on lung MPO activity after FIGURE 5. Effect of TNF-␣ neutralization on lung chitin after A. fu- A. fumigatus challenge in immunocompetent and cyclophosphamide- migatus challenge in cyclophosphamide-treated mice. Lung chitin levels in treated mice. Lung MPO activity was measured 1 day after A. fumigatus cyclophosphamide-treated mice given rabbit anti-murine TNF-␣ serum or challenge in normal mice (1–2 ϫ 107 conidia) and 2 days after A. fumigatus control serum were determined 3 days after A. fumigatus challenge (1–2 ϫ challenge in cyclophosphamide-treated mice (1–2 ϫ 106 conidia). Asp, A. -p Ͻ 0.05 com ,ء .p Ͻ 0.05 compared with fumigatus; Cy, cyclophosphamide. Experimental n ϭ 6 ,ء .Asp, A. fumigatus. Experimental n ϭ 6 .(106 animals receiving control rabbit serum. pared with animals receiving control rabbit serum. 1638 TNF-␣ IN MURINE INVASIVE ASPERGILLOSIS Downloaded from http://www.jimmunol.org/

FIGURE 7. Effect of TNF-␣ neutralization on lung MIP-2, KC, MIP- 1␣, and JE levels after A. fumigatus administration in immunocompetent mice. A and B depict C-X-C and C-C chemokine levels in immunocom- FIGURE 8. Effect of TNF-␣ neutralization on lung MIP-2, KC, MIP- ϫ 7 by guest on September 24, 2021 petent mice 1 day after A. fumigatus challenge (1–2 10 ), respectively. 1␣, and JE levels after A. fumigatus administration in cyclophosphamide- Ͻ ء ϭ Experimental n 6. , p 0.05 compared with animals receiving control treated mice. A and B depict C-X-C and C-C chemokine levels in cyclo- rabbit serum. phosphamide-treated mice 3 days after A. fumigatus challenge (1–2 ϫ 106), p Ͻ 0.05 compared with animals ,ء .respectively. Experimental n ϭ 6 receiving control rabbit serum. time of maximum expression of these chemokines, levels were measured 1 day after Aspergillus inoculation in normal mice, and 3 days after Aspergillus inoculation in cyclophosphamide-treated pretreated with TNF70–80 i.t. 3 days before the administration of A. mice. In both normal and cyclophosphamide-treated animals, there fumigatus compared with that in animals pretreated with either was a significant decrease in MIP-2 levels in anti-TNF-treated an- control peptide or vehicle 3 days before inoculation with A. fu- imals challenged with conidia compared with that in control ani- migatus conidia (Fig. 9). mals receiving serum. No significant change in KC levels was detected in either group. Treatment of normal or cyclophospha- Discussion mide-treated animals with anti-TNF-␣ serum also resulted in sig- Invasive pulmonary aspergillosis is a common and devastating nificant reduction in the C-C chemokines MIP-1␣ and JE/mono- complication of immunosuppression. Given its poor outcome with cyte chemoattractant protein-1 (Figs. 7 and 8). current therapy, the precise mechanism of the immune response to A. fumigatus is of interest, since immunomodulation may be ben- ␣ Effect of intrapulmonary delivery of rmTNF- or TNF70–80 on eficial as adjunctive therapy. survival in invasive pulmonary aspergillosis We studied the pulmonary host response to A. fumigatus in nor- Having demonstrated that TNF-␣ depletion increases fungal bur- mal and immunocompromised animal models. Despite the 10-fold den and mortality, we next examined the effect of compartmental- greater concentration of conidia used in all experiments involving ized TNF-␣ augmentation on survival in cyclophosphamide- immunocompetent animals, normal hosts readily cleared these treated animals challenged with A. fumigatus. To avoid systemic large inocula without developing invasive disease. Cyclophos- effects, we administered either rmTNF-␣ (1 ␮g/animal) or phamide, an alkylating agent in common clinical use as a chemo- ␮ TNF70–80 (10 g/animal) i.t. concomitant with the administration therapeutic and immunosuppressive agent, was used to render an- of A. fumigatus conidia (1.8 ϫ 106). Neither rmTNF-␣ nor imals susceptible to invasive disease. Cyclophosphamide induces a

TNF70–80 administration altered survival in cyclophosphamide- variety of effects on immunologically active cells; these include a treated animals when given at the time of Aspergillus challenge dose-dependent depletion of circulating neutrophils as well as al- compared with that in animals receiving a control peptide of ve- terations in the number and function of circulating and pulmonary hicle control (data not shown). However, signiﬁcantly enhanced monocyte/macrophages and various lymphocyte populations (27, survival was observed in cyclophosphamide-treated mice that were 28). Although peripheral blood neutropenia is an easily measurable The Journal of Immunology 1639

a receptor for MIP-1␣ and RANTES, develop disseminated infection when administered A. fumigatus i.v. (31). In this context, our study provides evidence that TNF-␣ drives the expression of both C-X-C and C-C chemokines, including MIP-1␣, JE, and MIP-2, which may in part mediate the inﬂux and activation of neutrophils and mononuclear cells in the lungs. Reduced amounts of MIP-2 and MIP-1␣ may explain the reduction in early neutrophil inﬂux (and possibly activation), increased fungal burden, and increased mortality in TNF-depleted animals. In addition, MIP-1␣ and JE/ monocyte chemoattractant protein-1 may play important roles in mononuclear cell recruitment and activation in the setting of invasive pulmonary aspergillosis. Studies establishing causal rela- tionships are in progress.

The intratracheal administration of TNF70–80 provided survival beneﬁts when given 3 days before the i.t. inoculation of A. fumigatus conidia, but not when given concomitantly with conidia. These results parallel our earlier observations in murine bacterial pneu- FIGURE 9. Effect of TNF70–80 on survival in cyclophosphamide- monia (22). The lack of a beneﬁcial effect in animals treated with treated mice pretreated with TNF70–80. Mice were treated with vehicle, ␮ ␮ Downloaded from control peptide (10 g), or TNF70–80 (10 g) i.t. 3 days before inoculation TNF70–80 concomitant with conidial administration may be due to Ͻ ء ϫ 6 with A. fumigatus (1.8 10 conidia). , p 0.05 compared with animals several possible mechanisms. Firstly, TNF70–80 is more potent as receiving vehicle or control peptide. p-control, control peptide; p-TNF, a priming agent than as a direct activator of leukocyte respiratory ϭ TNF70–80 peptide. Experimental n 15. burst and degranulation in vitro (20). Therefore, pretreatment with

TNF70–80 may provide sufficient macrophage priming, whereas concomitant treatment may result in insufficient leukocyte activa- clinical risk factor for the development of invasive aspergillosis in tion. Alternatively, leukocyte release of proteolytic enzymes in http://www.jimmunol.org/ humans, this susceptibility may also be related to the numerous response to infectious organisms within the airspace may lead to other effects of cyclophosphamide on other immune effector cells. cytokine degradation and loss of biologic effects. Lastly, the ad- ␣ In our model, animals pretreated with cyclophosphamide before ministration of native TNF- or TNF70–80 given at the time of inoculation with Aspergillus conidia developed an acute, severe fungal challenge may induce lung injury in the setting of pneu- pneumonia that clinically and histologically resembled invasive monia, thereby negating any beneficial effect on mortality. pulmonary aspergillosis complicating severe immunosuppression Given that immunocompromised patients at risk of developing in humans. Therefore, this model has the advantage of notable pulmonary invasive aspergillosis are easily identifiable, early in- clinical relevance, but is complicated by the diverse effects of cy- stitution of prophylactic immunomodulatory treatment is a poten- clophosphamide on various immune effector cells. tial application of this study. Such clinical applications require by guest on September 24, 2021 In this study we demonstrated that in both normal and cyclo- further study. phosphamide-treated mice, TNF-␣ is produced in the lungs in response to the i.t. administration of A. fumigatus. In immunocom- Acknowledgments petent animals, the TNF-␣ level is significantly elevated by 8 h after A. fumigatus inoculation. Resident alveolar macrophages are We are indebted to Dr. Thomas Moore for numerous instructive discus- most likely to be the source of this early and localized release of sions, to Dr. Steven Kunkel for generously providing us with anti-TNF-␣ TNF, which occurs before significant recruitment of other immune serum, and to Mrs. Mary Glass for technical assistance in performing effector cells occurs (data not shown). This view is supported by ELISAs. the fact that TNF-␣ is released from alveolar macrophages coin- cubated with A. fumigatus conidia in vitro (13). Elevated TNF-␣ at References later time points probably represents release from various recruited 1. Fraser, D. W., J. I. Ward, L. Ajello, and B. D. Plikaytis. 1979. Aspergillosis and cells, including neutrophils, and recruited monocyte/macrophages. other systemic mycoses: the growing problem. JAMA 242:1631. Notably, neutrophils were present in the lungs of cyclophospha- 2. Groll, A. H., P. M. Shah, C. Mentzel, M. Schneider, G. Just-Nuebling, and K. Huebner. 1996. Trends in the postmortem epidemiology of invasive fungal mide-treated mice infected with A. fumigatus while the animals infections at a university hospital. J. Infect. 33:23. had peripheral blood neutropenia. This is presumably due to pref- 3. Bardana, E. J., Jr. 1981. The clinical spectrum of aspergillosis. II. Classification erential accumulation of neutrophils at the site of inflammation and description of saprophytic, allergic, and invasive variants of human disease. Crit. Rev. Clin. Lab. Sci. 13:85. despite a reduced pool of available neutrophils. 4. Denning, D. W. 1996. Therapeutic outcome in invasive aspergillosis. Clin. Infect. In vivo depletion of TNF-␣ resulted in increased fungal burden Dis. 23:608. and mortality, which occurred in association with a reduction in 5. Schaffner, A., H. Douglas, and A. Braude. 1982. Selective protection against ␣ conidia by mononuclear and against mycelia by polymorphonuclear phagocytes lung neutrophil infiltration. TNF- is likely to contribute to host in resistance to Aspergillus: observations on these two lines of defense in vivo defense against A. fumigatus by several mechanisms. Although not and in vitro with human and mouse phagocytes. J. Clin. Invest. 69:617. directly chemotactic for neutrophils, TNF-␣ induces leukocyte and 6. Levitz, S. M., M. E. Selsted, T. Ganz, R. I. Lehrer, and R. D. Diamond. 1986. In vitro killing of spores and hyphae of Aspergillus fumigatus and Rhizopus oryzae endothelial cell expression of adhesion molecules, thus influencing by rabbit neutrophil cationic peptides and bronchoalveolar macrophages. J. In- neutrophil trafficking in the lungs (29). In addition, our studies fect. Dis. 154:483. indicate that the neutralization of TNF-␣ resulted in significant 7. Nessa, K., L. Palmberg, U. Johard, P. Malmberg, C. Jarstrand, and P. Camner. 1997. Reaction of human alveolar macrophages to exposure to Aspergillus fu- reductions in both C-X-C and especially C-C chemokines. Previ- migatus and inert particles. Environ. Res. 75:141. ous studies have suggested a role for chemokines in host defense 8. Nessa, K., C. Jarstrand, A. Johansson, and P. Camner. 1997. In vitro interaction against A. fumigatus: isolated rat alveolar macrophages produce of alveolar macrophages and Aspergillus fumigatus. Environ. Res. 74:54. ␣ ␣ 9. Chilvers, E. R., C. L. Spreadbury, and J. Cohen. 1989. Bronchoalveolar lavage in MIP-1 , MIP-2, and KC, as well as TNF- in response to A. an immunosuppressed rabbit model of invasive pulmonary aspergillosis. Myco- fumigatus conidia in vitro (30), and knockout mice lacking CCR1, pathologia 108:163. 1640 TNF-␣ IN MURINE INVASIVE ASPERGILLOSIS

10. Waldorf, A. R., S. M. Levitz, and R. D. Diamond. 1984. In vivo bronchoalveolar 22. Laichalk, L. L., K. A. Bucknell, G. B. Huffnagle, J. M. Wilkowski, T. A. Moore, macrophage defense against Rhizopus oryzae and Aspergillus fumigatus. J. Infect. R. J. Romanelli, and T. J. Standiford. 1998. Intrapulmonary delivery of tumor Dis. 150:752. necrosis factor agonist peptide augments host defense in murine Gram-negative 11. Roilides, E., A. Holmes, C. Blake, D. Venzon, P. A. Pizzo, and T. J. Walsh. 1994. bacterial pneumonia. Infect. Immun. 66:2822. Antifungal activity of elutriated human monocytes against Aspergillus fumigatus 23. Allen, S. D., K. N. Sorensen, M. J. Nejdl, C. Durrant, and R. T. Proffit. 1994. hyphae: enhancement by granulocyte-macrophage colony-stimulating factor and Prophylactic efficacy of aerosolized liposomal (AmBisome) and non-liposomal interferon-␥. J. Infect. Dis. 170:894. 12. Rex, J. H., J. E. Bennett, J. I. Gallin, H. L. Malech, and D. A. Melnick. 1990. (fungizone) amphotericin B in murine pulmonary aspergillosis. J. Antimicrob. Normal and deficient neutrophils can cooperate to damage Aspergillus fumigatus Chemother. 34:1001. hyphae. J. Infect. Dis. 162:523. 24. Lehmann, P. F., and L. O. White. 1975. Chitin assay used to demonstrate renal 13. Taramelli, D., M. G. Malabarba, G. Sala, N. Basilico, and G. Cocuzza. 1996. localization and cortisone-enhanced growth of Aspergillus fumigatus mycelium Production of cytokines by alveolar and peritoneal macrophages stimulated by in mice. Infect. Immun. 12:987. Aspergillus fumigatus conidia or hyphae. J. Med. Vet. Mycol. 34:49. 25. Goldblum, S. E., K. M. Wu, and M. Jay. 1985. Lung myeloperoxidase as a 14. Taramelli, D., M. G. Malabarba, N. Basilico, G. Sala, R. Mattina, and measure of pulmonary leukostasis in rabbits. J. Appl. Physiol. 59:1978. C. Cocuzza. 1994. Effects of pidotimod on macrophage functions in methylpred- nisolone-treated mice. Arzneimittelforschung 44:1425. 26. Greenberger, M. J., R. M. Strieter, S. L. Kunkel, J. M. Danforth, R. E. Goodman, 15. Grazziutti, M. L., J. H. Rex, R. E. Cowart, E. J. Anaissie, A. Ford, and and T. J. Standiford. 1995. Neutralization of IL-10 increases survival in a murine C. A. Savary. 1997. Aspergillus fumigatus conidia induce a Th1-type cytokine model of Klebsiella pneumonia. J. Immunol. 155:722. response. J. Infect. Dis. 176:1579. 27. Hunninghake, G. W., and A. S. Fauci. 1977. Immunological reactivity of the 16. Gosselin, D., J. DeSanctis, M. Boule, E. Skamene, C. Matouk, and D. Radzioch. lung. IV. Effect of cyclophosphamide on alveolar macrophage cytotoxic effector 1995. Role of tumor necrosis factor ␣ in innate resistance to mouse pulmonary function. Clin. Exp. Immunol. 27:555. infection with Pseudomonas aeruginosa. Infect. Immun. 63:3272. 28. Cooper, J. A., Jr., W. W. Merrill, and H. Y. Reynolds. 1986. Cyclophosphamide 17. Laichalk, L. L., S. L. Kunkel, R. M. Strieter, J. M. Danforth, M. B. Bailie, and T. J. Standiford. 1996. Tumor necrosis factor mediates lung antibacterial host modulation of bronchoalveolar cellular populations and macrophage oxidative defense in murine Klebsiella pneumonia. Infect. Immun. 64:5211. metabolism: possible mechanisms of pulmonary pharmacotoxicity. Am. Rev. Re- 18. Moussa, K., H. J. Michie, I. A. Cree, A. C. McCafferty, J. H. Winter, spir. Dis. 134:108. Downloaded from D. P. Dhillon, S. Stephens, and R. A. Brown. 1994. Phagocyte function and 29. Gamble, J. R., J. M. Harlan, S. J. Klebanoff, and M. A. Vadas. 1985. Stimulation cytokine production in community acquired pneumonia. Thorax 49:107. of the adherence of neutrophils to umbilical vein endothelium by human recom- 19. Nagai, H., J. Guo, H. Choi, and V. Kurup. 1995. Interferon-␥ and tumor necrosis binant tumor necrosis factor. Proc. Natl. Acad. Sci. USA 82:8667. ␣ factor- protect mice from invasive aspergillosis. J. Infect. Dis. 172:1554. 30. Shahan, T. A., W. G. Sorenson, J. D. Paulauskis, R. Morey, and D. M. Lewis. 20. Kumaratilake, L. M., D. A. Rathjen, P. Mack, F. Widmer, V. Prasertsiriroj, and 1998. Concentration- and time-dependent up regulation and release of the cyto- A. Ferrante. 1995. A synthetic tumor necrosis factor-␣ agonist peptide enhances kines MIP-2, KC, TNF, and MIP-1␣ in rat alveolar macrophages by fungal spores human polymorphonuclear leukocyte-mediated killing of Plasmodium falcipa- implicated in airway inflammation. Am. J. Respir. Cell. Mol. Biol. 18:435. rum in vitro and suppresses Plasmodium chabaudi infection in mice. J. Clin. http://www.jimmunol.org/ Invest. 95:2315. 31. Gao, J. L., T. A. Wynn, Y. Chang, E. J. Lee, H. E. Broxmeyer, S. Cooper, 21. Rathjen, D. A., A. Ferrante, and R. Aston. 1993. Differential effects of small H. L. Tiffany, H. Westphal, J. Kwon-Chung, and P. M. Murphy. 1997. Impaired tumour necrosis factor-␣ peptides on tumour cell cytotoxicity, neutrophil activa- host defense, hematopoiesis, granulomatous inflammation and type 1-type 2 cy- tion and endothelial cell procoagulant activity. Immunology 80:293. tokine balance in mice lacking CC chemokine receptor 1. J. Exp. Med. 185:1959. by guest on September 24, 2021