Download - Immune responses to Pneumocystis murina are robust in healthy mice but largely absent in CD40 ligand-deficient mice

Immune responses to Pneumocystis murina are robust inhealthy mice but largely absent in CD40 ligand-deficientmice

Beatriz Hernandez-Novoa,*,1 Lisa Bishop,* Carolea Logun,* Peter J. Munson,† Eldad Elnekave,‡

Zoila G. Rangel,† Jennifer Barb,† Robert L. Danner,* and Joseph A. Kovacs*,2

*Critical Care Medicine Department, National Institutes of Health Clinical Center, †Mathematical and StatisticalComputing Laboratory, Division of Computational Bioscience, Center for Information Technology, and ‡Laboratoryof Parasitic Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda,Maryland, USA

Abstract: Pneumocystis is a pathogen of immuno-compromised hosts but can also infect healthyhosts, in whom infection is rapidly controlled andcleared. Microarray methods were used to exam-ine differential gene expression in the lungs ofC57BL/6 and CD40 ligand knockout (CD40L-KO)mice over time following exposure to Pneumocystismurina. Immunocompetent C57BL/6 mice, whichcontrol and clear infection efficiently, showed arobust response to infection characterized by theup-regulation of 349 primarily immune response-associated genes. Temporal changes in the expres-sion of these genes identified an early (Week 2),primarily innate response, which waned before theinfection was controlled; this was followed by pri-marily adaptive immune responses that peaked atWeek 5, which coincided with clearance of theinfection. In conjunction with the latter, there wasan increased expression of B cell-associated (Ig)genes at Week 6 that persisted through 11 weeks.In contrast, CD40L-KO mice, which are highlysusceptible to developing severe Pneumocystispneumonia, showed essentially no up-regulation ofimmune response-associated genes at Days 35–75.Immunohistochemical staining supported these ob-servations by demonstrating an increase in CD4�,CD68�, and CD19� cells in C57BL/6 but notCD40L-KO mice. Thus, the healthy host demon-strates a robust, biphasic response to infection byPneumocystis; CD40L is an essential upstream reg-ulator of the adaptive immune responses that effi-ciently control infection and prevent developmentof progressive pneumonia. J. Leukoc. Biol. 84:420–430; 2008.

Key Words: immunodeficiency diseases � rodent � lung

INTRODUCTION

Pneumocystis jirovecii is a ubiquitous fungus that causes life-threatening pneumonia in immunodeficient patients, especially

those with HIV infection [1]. Although clinically significantpneumonia does not occur in immunocompetent individuals,serological studies have demonstrated that most humans havebeen infected by the age of 2, presumably as a result of aclinically inapparent infection [2, 3]. Studies of immune re-sponses to Pneumocystis have largely used immunodeficientanimal models or cell-depletion studies to identify importantimmunoregulatory mechanisms. Such studies have demon-strated a critical role for CD4 cells in immunity against Pneu-mocystis infection. However, only a limited number of studieshave examined immune responses in unmanipulated, immuno-competent hosts [4].

Pneumocystis can infect a variety of mammalian species.Each species is infected by a genetically distinct member ofthe genus: P. jirovecii infects humans, Pneumocystis carinii andPneumocystis wakefieldiae infect rats, and Pneumocystis mu-rina infects mice. Key features of P. murina infection haverendered it experimentally useful for studying immunity toPneumocystis, in part as a result of the availability of mice withdefined immune defects. Following exposure to the organism,immunocompromised murine hosts, such as scid mice, or thosedeficient in CD40 ligand (CD40L), develop a progressive pul-monary disease, which histologically and clinically resembleshuman Pneumocystis pneumonia (PcP) [5–7]. In contrast, as weand others have shown recently [8, 9], healthy mice becomeinfected following exposure but have a peak organism load �5weeks after exposure that is two to three logs below the peakseen in immunodeficient hosts and can clear the infection in amatter of weeks with little residual compromise. Little is knownabout early events, including innate immune responses, insuch healthy hosts.

1 Current address: Servicio de Enfermedades Infecciosas, 4ª Planta Centro.Control A, Hospital Ramon y Cajal, Ctra. de Colmenar Km 9.100, 28034Madrid, Spain.

2 Correspondence: Critical Care Medicine Department, National Institutesof Health Clinical Center, NIH, Building 10, Room 2C145, MSC 1662,Bethesda, MD 20892-1662. E-mail: [email protected]

Received December 7, 2007; revised April 9, 2008; accepted April 11,2008.

doi: 10.1189/jlb.1207816

420 Journal of Leukocyte Biology Volume 84, August 2008 0741-5400/08/0084-420 © Society for Leukocyte Biology

CD40L is a costimulatory molecule expressed on activated Tcells that plays a critical role in T cell- and B cell-mediatedresponses [10, 11]. CD40L is also expressed on a number ofadditional immune-related cells, including �� T, NK, NKT,and dendritic cells (DC) and macrophages, as well as nonim-mune cells. CD40L knockout (KO) mice have a specific im-mune defect that makes them highly susceptible to PcP, withlevels of infection similar to that seen in scid or Rag-1-deficient mice; humans with a homologous abnormality, result-ing in the hyper-IgM syndrome, are similarly highly suscepti-ble to Pneumocystis infection [12, 13]. However, the interactionof CD40L with other aspects of the immune response to Pneu-mocystis has not been defined. In the current study, we soughtto investigate the responses to Pneumocystis infection thatoccur in healthy hosts and to determine which of these re-sponses were deficient in CD40L-KO hosts. We used microar-ray techniques to examine changes in expression of over12,000 genes in the lungs of wild-type and CD40L-KO miceinfected with P. murina. Such a genome-wide approach pro-vides a means to simultaneously examine a broad range ofimmune responses and thereby yield biologically meaningfulresults, not only about individual genes but also about genenetworks and multi-network genetic programs of host defense.

MATERIALS AND METHODS

Animals

Healthy, �8-week-old, female C57BL/6 mice were obtained from the NationalCancer Institute (NCI; Frederick, MD, USA) or from The Jackson Laboratory(Bar Harbor, ME, USA). CD40L-KO (B6;129S2-Tnfsf5tm1Imx/J) and scid(CBySmn.CB17-Prkdcscid/J) mice were bred in-house, initially using breederspurchased from The Jackson Laboratory. All studies were carried out underprotocols approved by the National Institutes of Health (NIH) Clinical CenterAnimal Care and Use Committee (Bethesda, MD, USA).

Study design

As a primary goal was to characterize immune responses to naturally acquiredinfection (similar to what occurs in humans), infection was transmitted via therespiratory route by co-housing study animals with immunosuppressed seederanimals (scid or CD40L-KO) that had active PcP, which was subsequentlyverified by a quantitative real-time PCR (Q-PCR) assay [8]. This has beenpreviously shown to result in infection of co-housed animals with predictablekinetics and infection of healthy animals peaking at 35–42 days [8]. We usedthis approach rather than intratracheal inoculation of organisms, as the latter,although commonly used in studies of Pneumocystis, will deliver a substan-tially larger organism load as a bolus and may induce immune responses thatdiffer qualitatively or kinetically from those seen during naturally acquiredinfection [14]. In four separate experiments, C57BL/6 and/or CD40L-KO micewere co-housed with a P. murina-infected seeder or remained unexposed. Mice(three to 10 per time-point per group) were killed at varying time-pointsranging from 7 to 75 days.

Lungs and blood were obtained from each animal when they were killed.Lungs were split in two and stored at –20°C; one portion was placed inRNAlaterTM (Qiagen, Valencia, CA, USA) for microarray analysis and theother in PBS for quantitation of organism load by Q-PCR. Blood was obtainedby cardiac puncture, and serum was stored at –20°C for subsequent detectionof anti-P. murina antibodies by ELISA [8]. P. murina Q-PCR and ELISAresults for animals from the first experiment have been reported previously ina study describing the Q-PCR assay [8]. For immunohistochemistry studies,mice were exposed to P. murina, animals were killed at 14, 35, or 42 days afterexposure, and portions of the lung were stored at –20°C for Q-PCR or Westernblot analysis or frozen with O.C.T. compound (Sakura Finetek USA, Torrance,

CA, USA) in cryomolds on a slurry of dry ice with isopentane and stored at–20°C for subsequent immunohistochemical staining.

Q-PCR and ELISA

DNA extraction from lung homogenates and subsequent quantitation of P.murina dihydrofolate reductase (DHFR) gene copies per mg lung by means ofa real-time PCR were performed as described previously [1, 8]. As DHFR isa single-copy gene, the number of copies reflects the number of nuclei. Thepresence of anti-P. murina antibodies was detected using an ELISA techniqueas described previously [8].

Microarray hybridization and analysis

Total RNA was extracted from lung tissue using an RNeasy mini kit includingQiashredder columns and DNase treatment (Qiagen), following the manufac-turer’s instructions.RNA (4–10 �g) was reverse-transcribed using a Super-ScripTM ds cDNA synthesis kit (Invitrogen Life Technologies, Carlsbad, CA,USA) and the GeneChip� T7-Oligo(dT) promoter primer kit (Affymetrix, SantaClara, CA, USA). The cDNA was purified following the procedures describedby Affymetrix and was transcribed in vitro and biotin-labeled using theBioArray High Yield RNA transcript labeling kit (ENZO Life Science Inc.,Farmingdale, NY, USA). The resulting labeled cRNA was purified, and 20 �gwas fragmented using 5� fragmentation buffer and heated to 95°C for 35 min.

Hybridization cocktail (Affymetrix) containing 15 �g fragmented cRNA wasadded to an MG-U74Av2 array (Affymetrix) and hybridized for 16 h at 45°C ina Hybridization Oven 640 (Affymetrix). The arrays were stained and washedusing the GeneChip� Fluidics Station 400 using R-PE streptavidin (MolecularProbes, Eugene, OR, USA) and a biotinylated goat anti-streptavidin antibody(Vector Laboratories, Burlingame, CA, USA). Experiments 1 and 2 werescanned using the GeneArray� 2500 scanner. The GeneChip� Scanner 3000was used to scan Experiments 3 and 4. The signal intensity was quantitatedusing Microarray Suite Software, Version 4.0 (MAS 4.0), or GeneChip� Op-erating Software v1.2 (GCOS 1.2; both from Affymetrix). The data discussed inthis publication have been deposited in the National Center for BiotechnologyInformation (NCBI; U.S. National Library of Medicine, Bethesda, MD, USA)Gene Expression Omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/) and areaccessible through GEO Series Accession Number GSE11005.

RT-PCR analysis of gene expression

To validate the microarray gene expression results, expression of a panel ofdifferentially expressed genes, including genes that were up-regulated, un-changed, and down-regulated, was examined using a real-time, two-step,semi-QRT-PCR assay, based on TaqMan� technology (Applied Biosystems,Foster City, CA, USA). The oligonucleotide sequences for individual genes areprovided as supplemental data (Supplemental Table 1). For CCR2, CCL2, andCXCL10, reagents were purchased as Taqman� predeveloped assay reagentsfor gene expression. The cDNA was amplified in triplicate in an ABI PRISMTM

7900 HT sequence detection system instrument. Relative standard curves wereused to calculate the fold-change (FC) of the gene of interest after normalizingthe values to those of GAPDH (rodent GAPDH control reagents), which wererun in parallel. We used the same RT-PCR assay for determining the changein expression of TLR-2 and TLR-4 in the early exposure to P. murina, as theyare important, innate immunity effectors but are not included in the MG-U74Av2 array. The oligonuclotide sequences are provided as supplementaldata (Supplemental Table 1).

Statistical analysis

Affymetrix MAS 5 signal and present call values were stored in the NIHLIMS,a local database for storage and retrieval of Affymetrix GeneChip data, whichwere retrieved and analyzed using the Mathematical and Statistical ComputingLab Analyst’s Toolbox (P. J. Munson, J. Barb, 2004; http://abs.cit.nih.gov/MSCLtoolbox/) and the JMP statistical software package (SAS Inc., Cary, NC,USA). Data were first normalized to the median value of each chip and thenlogarithmically transformed after the addition of a small value (1% of themedian) to dampen the influence of extremely small data values. A one-wayANOVA was performed to determine the significance of changes within eachexperiment. Affymetrix genes were selected if the P value was less than 0.001,and the FC between the exposed compared with the unexposed animals of thesame strain was at least 1.5-fold or at most 1/1.5-fold. K-means and hierar-

Hernandez-Novoa et al. Immune responses to Pneumocystis infection 421

chical clustering were performed using JMP. Heat maps were generated usingGenesis [15].

Gene list analysis

NCBI databases, primarily LocusLink, its substitute Entrez Gene, andPubmed, as well as Mouse Genome Informatics 3.2 database (The JacksonLaboratory) were used to further explore the gene lists provided by thestatistical treatment of the data. Gene list subsets assigned to specific geneontology categories were annotated by means of the National Institute ofAllergy and Infectious Diseases-developed tool Database for Annotation, Vi-sualization and Integral Discovery (DAVID; http://david.abcc.ncifcrf.gov/home.jsp) and Expression Analysis Systematic Explorer (EASE; http://david.abcc.ncifcrf.gov/ease/ease.jsp) [16, 17]. To explore the relations of geneswithin the relevant gene lists and the possible existence of networks compris-ing them, we used EASE as well as Ingenuity Pathways analysis (Ingenuity�

Systems, Redwood City, CA, USA). The latter was also used for additional dataannotation and mining. Genes annotated by means of the Affymetrix websitereflect the available data as of March 2005.

Immunohistochemistry

To verify the results of the microarray analysis that suggested the presence ofdifferent cell populations during the course of Pneumocystis infection, in situimmunohistochemistry was performed using frozen lung sections from healthyC57BL/6 mice obtained at Weeks 2, 5, and 6 following exposure and fromCD40L-KO mice obtained at Weeks 2 and 5 following exposure. Unexposedmice were used as controls. The following primary antibodies were used:anti-Ly49s, a cocktail of three mAb (4D11, 4E5, and 1F8, kindly provided byDrs. John Ortaldo and Robin Winkler-Pickett, NCI-Frederick) to detect NKcells, anti-CD4 (clone H129.19), and anti-CD8 (clone 53-6.7) T cell markers(BD Biosciences PharMingen, San Diego, CA, USA); anti-CD19 (clone 1D3) Bcell marker (BD Biosciences PharMingen); and anti-CD68 (clone FA-11)macrophage marker (AbD Serotec, Raleigh, NC, USA). The secondary antibodywas an Alexa Fluor 488-conjugated donkey anti-rat IgG (H&L; 20 �g/ml,Molecular Probes). Staining was performed by Histoserv, Inc. (Germantown,MD, USA). Images were collected on a Leica TCS-NT/SP1 confocal microscope(Leica Microsystems, Exton, PA, USA) using a 40� oil-immersion objectivenumerical aperture (NA) 1.32. Images of CD19 staining were collected on aNikon Eclipse E800 using a 40� plan-fluor oil-immersion objective NA 1.30.

All regions of the specimen were examined to ensure that observed changeswere consistent throughout the sample. To quantitate cell populations, positivecells were visually counted in 10 random fields per section using an epifluo-rescence microscope and a 40� or 60� objective; the same objective was usedfor all samples of a given mouse strain and cell type. The average number ofpositive cells per field was compared. To quantitate reactivity with anti-mouseCD68, the mean fluorescence intensity (MFI) was calculated for 10 fields perlung-tissue section using Leica TCS-SP software. MFIs were compared usingStudent’s t-test.

Western blot

Lung tissue was homogenized at 100 mg tissue per ml homogenization buffer[0.05 M Tris, pH 8.0, 0.12 M NaCl, 1% deoxycholate, 1% Triton X-100, 0.1%SDS, 10 mM DTT, 1 mM EDTA, 0.2 mM PMSF, 1� protease inhibitor cocktail(Sigma Chemical Co., St. Louis, MO, USA)], separated on 4–20% Tris-glycineSDS-PAGE gels (Invitrogen Life Technologies), and transferred to nitrocellu-lose membranes (Invitrogen Life Technologies). After blocking, the membranewas incubated with anti-ClCa3 antibody (kindly provided by Dr. HirokiIwashita, Takeda Pharmaceutical Co. Ltd., Osaka, Japan), followed by perox-idase-conjugated anti-rabbit IgG antibody (Jackson ImmunoResearch Labora-tories, Inc., West Grove, PA, USA). Anti-mouse GAPDH mAb (Calbiochem, LaJolla, CA, USA) reactivity was analyzed simultaneously to demonstrate thatequal amounts of protein were loaded on the gel. Reactivity was detected usingAmersham ECLTM advance Western blotting detection kit (GE HealthcareBio-Sciences Corp., Piscataway, NJ, USA).

RESULTS

Microarray techniques identify substantialimmune responses to Pneumocystis

We initially sought to determine if microarray techniques coulddetect changes in gene expression in the lungs of animalsfollowing infection with P. murina. To address this, the lungsof C57BL/6 (n�3) and CD40L-KO (n�3) animals were exam-ined 32 days after exposure, which approximately correspondsto the peak of infection in the wild-type animals. At this time,the mean P. murina load (log organisms per mg lung tissue)was similar for the two groups of animals: 2.39 log copies/mg(�245 copies/mg) for the wild-type C57BL/6 mice and 3.05 logcopies/mg (�1100 copies/mg) for the CD40L-KO mice. Re-sults from co-housing experiments previously reported by ourgroup showed that from this time-point on, the organism loadincreases in the CD40L-KO mice, reaching 5.5 log copies/mg(�300,000 copies/mg) at Day 75 of exposure, and the infectionis cleared by 7–8 weeks in wild-type C57BL/6 mice [8].Unexposed, control cages were negative for P. murina byQ-PCR. No anti-P. murina antibodies were detected in anycage at this time-point [8].

Microarray results for Experiment 1 were analyzed initiallyby ANOVA, identifying 418 probe sets that were differentiallyexpressed (P�0.003) in exposed compared with unexposedanimals of the same strain. EASE analysis revealed that thislist was highly enriched in immune response-related genecategories, as all eight categories selected using a Bonferroni-corrected P value �0.05 were immune response-related.EASE analysis, restricted to up-regulated genes in the healthy,exposed group (n�152), found 18 over-represented categories,of which 17 (94.4%) were immune response-related (Table 1),including genes induced by IFN-�, macrophage, and T cell-related genes and genes for chemokine receptors and ligands.Up-regulated genes in the CD40L-KO-exposed group (n�70)

TABLE 1. EASE Analysis of Genes Significantly Up-Regulated inHealthy Mice at Day 32 of Exposure

Gene categoryBonferroni-

corrected P value

Defense response 1.67 � 10�23

Response to biotic stimulus 8.48 � 10�22

Immune Response 2.67 � 10�19

Response to external stimulus 3.20 � 10�16

Response to stimulus 3.97 � 10�15

Organismal physiological process 5.11 � 10�10

Response to pest/pathogen/parasite 1.99 � 10�07

Antigen presentation 1.09 � 10�06

Antigen processing 2.93 � 10�06

Inflammatory response 1.12 � 10�05

Innate immune response 1.75 � 10�05

Response to wounding 5.87 � 10�05

Response to stress 7.93 � 10�04

Chemokine receptor binding 9.87 � 10�04

Chemokine activity 9.87 � 10�04

G-protein-coupled receptor binding 1.55 � 10�03

Antigen processing/exogenous antigen via MHCclass II 8.00 � 10�03

Antigen presentation/exogenous antigen 3.62 � 10�02

422 Journal of Leukocyte Biology Volume 84, August 2008 http://www.jleukbio.org

and down-regulated genes in either group (n�132) were notover-represented significantly in any category. RT-PCR for sixgenes encompassing a broad range of levels of expressionshowed FC similar in magnitude to the microarray results forthis experiment (Table 2).

Biphasic immune response to Pneumocystis inimmunocompetent animals

This initial experiment demonstrated that microarray analysisof lung homogenates was able to identify robust immune re-

sponses in wild-type but not in CD40L-KO mice. We under-took additional experiments to better characterize the kineticsof immune responses to P. murina infection in wild-type mice.In all studies, exposed animals showed the expected increasein organism load and anti-P. murina antibodies with time, andunexposed animals were negative for both (data not shown).

ANOVA analysis was performed for each experiment sepa-rately to identify those genes that were differentially regulatedin exposed compared with unexposed animals, using a P value�0.001 and a FC �1.5 or �1/1.5 at any time-point as criteriafor selection, following which the selected genes from allexperiments were combined into a single list for further anal-ysis. A total of 448 genes fulfilled the selection criteria, ofwhich 349 were significantly up-regulated (FC�1.5) in at leastone experiment. To validate the reliability of our model, wecompared the log10 FC in all 448 genes in wild-type mice atDay 34, Experiment 2, to the log10 FC for the same genes atDay 32, Experiment 1. We found a highly significant correla-tion (R�0.85; P�0.0001) between the values for the twosamples (Fig. 1).

K-means clustering was used to categorize the up-regulatedgenes into four clusters (Fig. 2A), which segregated by timefollowing exposure. Complete lists of the genes that wereincluded in each cluster are included as supplemental material(Supplemental Tables 2–5). Cluster A represents 52 genes thatshare a similar expression profile, showing an early increasewith peak expression at Day 14, followed by a return tobaseline levels after Day 21. Cluster B represents 218 genesthat exhibit a mild peak at Day 34, whereas cluster C repre-sents a smaller group of 32 genes that exhibits a more dramaticpeak at the same time-point. Finally, cluster D represents agroup of 47 genes that rises dramatically after Day 34 (Fig. 2B).

Each of these gene clusters was analyzed individually byEASE as well as by a literature search for individual genes tocharacterize biological responses represented by that cluster.Cluster A has hallmarks of a primarily innate, NK cell-asso-ciated, immune response to Pneumocystis in wild-type mice,characterized by a marked increase in expression of granzymesB–G (Fig. 3A) [18–21], killer cell lectin-like receptors (Klrd1and Klra3) [22–25], and cathepsin W [26]. Two genes involvedin the signaling transduction of NK receptors are also ex-pressed {TYRO protein tyrosine kinase-binding protein (Ty-robp) and phospholipase C�2 [27–31]}. At this stage of P.murina infection, there was significant overexpression of threeIFN-type I (, )-related genes: IFN regulatory factor 7 (Irf7),which appears to be a critical regulatory factor for productionof Type 1 IFNs [32–34]; IFN--inducible protein (G1P2; cloneIFI15; and a similar expressed sequence tag), which amongother properties, plays a role in NK differentiation and prolif-eration and is involved in stimulating IFN-� production by Tcells [35–40]; and IFN-stimulated exonuclease gene 20 kDa(Isg20), which is induced by type-1 IFNs.

Up-regulation of cathepsin E [41], CCR2, Irf7, Isg20, andTyrobp suggests activation of DC, and increased expression ofthe �-chain of the TCR likely reflects the influx of �� T cells,a subset usually present in the airway mucosa with limitedantigen recognition that contributes to host defense early ininfection [42]. Additional up-regulated genes include thoseencoding acute-phase proteins such as calgranulin A (S100a8)

TABLE 2. Validation of FC in Gene Expression by Real-TimeRT-PCR

GeneExposure

dayFC

microarrayFC

RT-PCR

Experiment 1Saa3 32 45.9 65Clca3 32 33.8 65.7Ccl2 32 19.0 34.0Ligp2 32 3.5 3.5H2afx 32 2.0 0.9Psmb9 32 3.4 1.3

Experiment 2Saa3 34 (1) 17.7 30.8

34 (2) 18.1 28.741 (1) 2.6 3.741 (2) 6.1 6.6

Clca3 34 (1) 48.6 206.134 (2) 47.1 394.041 (1) 12.1 22.341 (2) 16.8 29.8

Ccl2 34 (1) 5.9 11.434 (2) 5.8 10.641 (1) 1.6 2.841 (2) 2.3 4.4

Ccl8 34 (1) 11.5 22.434 (2) 8.5 22.741 (1) 4.6 8.041 (2) 5.7 9.5

J558 34 (1) 1.5 4.834 (2) 1.2 0.941 (1) 10.6 11.841 (2) 6.1 4.1

Experiment 3Tlr2a 7 NA 1.3

14 NA 1.721 NA 1.6

Tlr4a 7 NA 1.114 NA 1.421 NA 1.3

Ccr2 7 1.6 1.314 2.1 1.021 1.4 1.8

GzmD 7 1.4 2.014 18.3 1143.221 1.4 9.3

Ccl8 7 1.7 1.514 1.3 1.121 1.8 2.2

Values represent the mean FC in expression in C57BL/6 mice for theindicated gene as measured by microarray or RT-PCR. In Experiment 2, thenumbers in parentheses represent the cage number, which is included todemonstrate the consistency of results among replicate cages. a Tlr2 and Tlr4were not included in the microarray chip. NA, Not applicable.


and calgranulin B (S100a9), which form a heterodimer and aresecreted by macrophages or neutrophils [43–45], and a subsetof chemokine-related genes, including CCR2, CCR5, andCCL5 (RANTES).

As the genes in cluster A waned, those in clusters B and C,which represent aspects of an adaptive immune response,increased. Both clusters represent common immunological pro-cesses taking place at Day 34 of exposure; the clusters differ,mainly in the intensity of the up-regulation. Genes in clustersB and C are expressed mainly in DC, macrophages/monocytes,and T cells. There are also some genes expressed in B cells andlung cells (alveolar macrophages and epithelial and endothelial

cells). We found a subset of genes coding for cell surfacemolecules corresponding to antigen receptors, mainly the TCR,and chemokine receptor CXCR3 (genes associated with sig-naling pathways of both receptors were also up-regulated atthis time-point), costimulatory molecules, FcRs, and CD andhistocompatibility antigens. The functional analysis of clustersB and C also identified a substantial number of IFN-�-inducedgenes [46] (Fig. 3A), including members of the CC and CXCchemokine families, complement components, histocompatibil-ity class II antigens, FcRs, GTPases, immunoproteasome-ac-tive sites, and signal transducer activators and inhibitors. Twoother gene subsets related to antigen presentation and involvedin the immunological synapse were also identified. Signaturesof adaptive immune response such as genes related to phago-cyte function and T cell activation, differentiation, and regu-lation were also categorized, together with genes involved in Bcell function regulation. A network created by means of Inge-nuity Pathways highlights the widely interconnected relationsamong some of the up-regulated key genes included in clustersB and C (Fig. 4).

Taken together, the pattern of up-regulated genes following34 days of exposure of healthy mice to P. murina suggests aTh1 response. In addition to evidence of IFN-� responses,there is up-regulation of three chemokine receptors expressedby Th1 cells: CCR2, CCR5, and CXCR3. CCR2 and CCR5peak at Day 14 and therefore, are included in cluster A butremain up-regulated at Day 34. Among the chemokines ex-pressed by Th1 cells, CCL2 (MCP-1), CCL4 (MIP-1), CCL5(RANTES at Day 14), CXCL9 (monokine induced by IFN-�),and CXCL10 (IFN-inducible protein 10) are all up-regulated.These in turn attract cells bearing the aforementioned chemo-kine receptors, colocalizing Th1 cells, NK cells, and macro-phages [47]. Also, four MCPs, CCL2, CCL7, CCL8, andCCL12, are up-regulated at this time-point.

Two of the most highly up-regulated genes at Day 34 ofexposure are ClCa3 (also know as Gob-5; FC�47.9) and Saa3(FC�17.9). ClCa3 is a member of the calcium-activated chlo-ride channel family and is selectively expressed in the airwaygoblet cells [48–50]. Western blot analysis confirmed the highClCa3 expression found by microarray and RT-PCR (Fig. 5).Although its function is not elucidated completely, it has beenshown to increase mucin (muc5ac) expression and mucusproduction and to have a role in diseases with secretory dys-functions and airway hyper-responsiveness (asthma and cysticfibrosis) [48–50]. Recently, its role as a chloride channel hasbeen questioned, as no transmembrane region seems to bepresent [51, 52]. Also, its role in mucus overproduction hasbeen found not to be essential in a ClCa3 KO mouse [53], buta role in regulation of tissue inflammation in the innate immuneresponse has been suggested recently [49]. Saa3 is a majoracute-phase protein that is expressed in macrophages, andalthough its role in immune defense is yet to be determined, ithas been described as a T cell chemoattractant [54, 55].

The fourth cluster (D) represents a group of genes that isup-regulated primarily at Day 41. This cluster is composedalmost exclusively of Ig genes (Fig. 3A) and signals B cellinfiltration and/or activation. It is noteworthy that up-regulationof these genes is maintained through Day 75 following expo-sure.

Fig. 1. Correlation between Experiments 1 (32 days of exposure to P. murina,C57BL/6, and CD40L-KO mice) and 2 (34 days of exposure to P. murina,C57BL/6 mice) for the 448 genes that were differentially expressed in thestudy. There was a highly significant correlation between the log10 FC in theexpression of these genes for the wild-type mice (P�0.0001, A) but not for theCD40L-KO mice (P�0.9, B), highlighting the tight correlation between the twoexperiments for the responses in C57BL/6 mice, and also the strong responseof C57BL/6 (A) in contrast with the absence of response of CD40L-KO mice(B). For each experiment, FC was determined by comparing expression of thegene set in exposed mice to expression in unexposed mice of the same strain.


Microarray gene expression results for a subset of geneswere confirmed by RT-PCR. Table 2 shows the FC in geneexpression, as determined by microarray and RT-PCR, forselected genes when compared with unexposed animals. Al-though the absolute FC varied between the two assays, the rankorder and relative magnitude of change were in general similar.By RT-PCR, TLR-2 was up-regulated significantly at Day 14 ofexposure (P�0.005), and TLR-4 was not up-regulated signif-icantly at any tested time-point.

Lack of immune response in CD40L-KO mice

In contrast to the robust immune response detected in C57BL/6mice, CD40L-KO mice exposed to Pneumocystis were lackingsuch immune responses at the later stages of infection thatwere examined by microarray analysis. At 35 days post-expo-sure, only 17 genes met the selection criteria for differentialexpression (P value �0.001 and a FC �1.5 or �1/1.5). Ninewere up-regulated, five of which were also up-regulated inC57BL/6 mice (Fig. 3B), and eight were down-regulated. At 39days post-exposure, no gene with a false discovery rate of�20% met the selection criteria. At 75 days post-exposure, ata time when the Pneumocystis organism load was �300,000copies/mg lung tissue, 96 genes were differentially expressed,of which 56 were up-regulated and 40 down-regulated. Only 11of these genes were also included in the 448 genes identified inimmunocompetent C57BL/6 mice, and none was immune-func-tion genes. EASE analysis of these 96 genes did not identify

enrichment of any immune response-related gene categories;identified categories were primarily related to intracellularorganelles and metabolic processes (Supplementary Table 4).

Immunohistochemical staining identifies immunecell influx into the lungs of immunocompetentbut not CD40L-KO mice

Frozen lung tissue sections were stained to examine the fre-quency of various cell populations at different time-points(Supplemental Fig. 1). In healthy C57BL/6 animals, an in-crease in CD4�, CD19�, and CD68� cells was seen whencompared with unexposed controls, most prominently at Week6, at which point, the changes were significant for all threepopulations (Table 3). No changes were seen at Week 2 in anyof these cell populations, and no increase in CD8� or Ly49�cells was seen at any time. Among the CD40L-KO animals, nochanges were seen at Weeks 2 or 5 compared with unexposedanimals.

DISCUSSION

Our studies have shown the evolution of the immune responseto naturally acquired Pneumocystis in the healthy host, whichprogresses in a bimodal manner. In contrast to this, CD40L-KOanimals, which are highly susceptible to Pneumocystis infec-tion, showed little evidence by microarray or immunohisto-

Fig. 2. (A) Heat map showing the log10 FC over time for the 448 genes identified in the study.Days following exposure to Pneumocystis are shown at the top. The values on the left (Days 7–75)are for C57BL/6 mice (57B); those on the right are for CD40L-KO mice (40L; Days 35–75). Eachrow represents a single gene. The brackets on the right indicate the five clusters into which thegenes were grouped. Clusters A–D show the genes that were up-regulated significantly at one ormore time-points, and the fifth cluster comprises significantly down-regulated genes. Colorsrepresent the log10 FC, using the scale at the top of figure (B). The graphs (A–D) show for C57BL/6mice the log10 FC over time for the individual genes in each of the four clusters, into whichsignificantly up-regulated genes were segregated using K-means clustering.


Fig. 3. Heat map for select subgroups of genes. Colors represent the log10 FC, using the scale at the top of the figure. Each row represents a unique gene andis labeled with the Affymetrix probe set ID and gene title. (A) Genes that were up-regulated in C57BL/6 mice following Pneumocystis infection included granzymes,which peaked at Day 14 and were subsequently down-regulated; IFN-�-induced genes (including chemokine ligands and receptors), the majority of which peakedat Day 34, remained elevated but dampened at Day 41, and declined to base-line by Day 75; and Ig-related genes, the majority of which increased only at Day41 but remained elevated through Day 75. (B) The 17 genes that were differentially expressed at Day 35 in CD40L-KO mice are shown together with results forthe same genes in C57BL/6 mice at Days 7–75. The first five were also up-regulated significantly in C57BL/6 mice.


chemical analysis of an immune response to infection at thelater time-points that appear critical to clearance in the healthyhost.

Our animal model mimics infection of humans by Pneumo-cystis, as mice get the infection through inhalation of presum-ably a small number of organisms rather than by direct inoc-ulation into the respiratory tract. Inoculation of a bolus (106–108) of Pneumocystis organisms in healthy animals results inmore rapid clearance of the organisms in a response that ischaracterized by a profound eosinophilic pulmonary infiltra-tion, which is different from what was observed in the currentstudy [14].

We used microarray analysis to obtain a broad survey of hostresponses following Pneumocystis infection. The advantage ofthis approach is that it permits the simultaneous evaluation ofa large number of genes; thus, the role of different cell typesand pathways can be explored at the same time. The repro-ducibility of this approach was validated by the high level ofcorrelation between the results of the first two experiments(R�0.845; P�0.0001).

Initially there is an up-regulation of genes in the healthyhost that reflects an innate immune response (cluster A), whichpeaks at Day 14 and wanes through Day 21. This initialresponse is ineffective in clearing Pneumocystis, as the organ-

Fig. 4. Network created by means of Ingenuity Pathways using genes up-regulated in clusters B and C in C57BL/6 mice. The network was one of many generatedfollowing input of the genes in clusters B and C and was selected to illustrate the complex interactions of the genes in these clusters. All the genes in the networkare included in the two clusters. The intensity of the color is proportional to the FC seen at Day 34. The abbreviations and full gene names are included inSupplementary Tables 3 and 4.

Fig. 5. Western blot showing the differential expression of ClCa3 in C57BL/6mice unexposed and exposed to P. murina for 2, 5, and 6 weeks. Increasedlevels of protein can be seen in four of the six exposed animals at Weeks 5 and6. No expression is detected by Western blot in the unexposed animals or theexposed animals at 2 weeks. GAPDH is included as a control to demonstratethat approximately equal amounts of protein were loaded into each lane.


ism load continues to rise until it peaks at �5 weeks. The maineffector cells of this first phase are likely NK or NKT cells andDC, as reflected by the gene networks and families that wereup-regulated, such as granzymes. Although the immunohisto-chemical staining did not identify an increase in NK cells, thismay reflect activation in situ without recruitment of such cellsor recruitment of a small subset that was not apparent by thestaining used. The influx may also be short-lived and notdetected by the single, early time-point that was stained.

Also, during this first phase, type I IFN activity was in-creased, which in other models, down-regulates Th2 responsesin Pneumocystis-infected mice, lowering pulmonary complica-tions associated with infection [14]. There is also evidence ofan influx of �� T cells, which have previously been reported tobe associated with slower clearance in intratracheally inocu-lated models of Pneumocystis infection [4, 56]. Although TLRsare not interrogated by the array, we studied the gene expres-sion of TLR-2 and -4 and found that only TLR-2 was up-regulated significantly at Day 14 of infection. In accordancewith this, a recent study [57] has shown that normal C57BL/6mouse macrophages respond to P. murina through TLR-2, withsubsequent production of TNF- and CXCL1, and TLR-4transcription was not increased. Interestingly, CXCL1 is one ofthe chemokines highly up-regulated in our study at Day 34.

As the innate response declines, robust, adaptive immuneresponses are apparent at Days 34 and 41. Coincident withthis, there is a stabilization and subsequent decline in organ-ism load, with clearance of Pneumocystis by Day 75 or earlier.At Day 34, genes included in clusters B and C are up-regulated; these genes largely reflect IFN-� induction andinflux or activation of T cells and macrophages. There isevidence of a Th1 response, characterized by the up-regulationof a large set of IFN-�-induced genes and chemokine signalinggenes (Th1-related receptors and ligands). Although previousstudies have shown that IFN-� is not essential for clearance ofPneumocystis, the current study demonstrates a prominent rolefor IFN-�-induced genes in the immunocompetent host re-sponse [58, 59]. Also, CXCR3 and its ligands CXCL9 andCXCL10 were greatly up-regulated. CXCR3 is expressed by

CD4 (preferentially Th1), CD8, NK cells, and B cells. A recentstudy by McAllister et al. [60] found that clearance of Pneu-mocystis was delayed, but not impaired, in otherwise immuno-competent, CXCR3-deficient mice, demonstrating that CXCR3plays an important but not essential role in clearance of Pneu-mocystis. Our data suggest that the increased expression of abroad range of chemokines and chemokine receptors providesa redundancy to the clearance mechanisms.

Our studies are consistent with the previously established,critical role of CD4 cells and macrophages in clearance ofPneumocystis [61–63]. Both cell populations increased atWeeks 5 and 6, and multiple genes related to their function areup-regulated. There was no increase in CD8 cell number,suggesting that this population does not play an important rolein clearance of Pneumocystis in healthy animals. Althoughmost studies in immunodeficient models also support thisconclusion, CD8 cells can be manipulated to facilitate clear-ance of the organism in CD4-depleted models [64, 65].

A surprising and previously unreported finding was theprominent influx of B cells at Day 41, which was initiallysuggested by up-regulation of Ig-related genes in cluster D andconfirmed by immunohistochemistry. The role of these B cellsin anti-Pneumocystis immunity is unclear. It is noteworthy thatgenes in this cluster remained up-regulated at Day 75, sug-gesting that B cells may play a role in long-term immunity toPneumocystis. Consistent with our data, two studies have re-cently highlighted an important role for B cells in clearance ofPneumocystis that is independent of antibody production andmay be related to the antigen-presenting function of B cells[66, 67].

One of the most striking findings of this study was themarked absence of late immune responses in CD40L-KO mice.Among the immune-related genes up-regulated in C57BL/6mice, only CCL9, CXCR1, and integrin-2 were up-regulatedin these animals. In two separate experiments looking at anearlier time-point (Day 14) in CD40L-KO animals, inconsistentresults were obtained, and no significant up-regulation of in-nate responses was noted for the group as a whole (data notshown). However, in individual animals, there was up-regula-

TABLE 3. Quantitation of Different Cell Populations Over Time Following Exposure to P. murina

CD4 CD8 CD19 Ly49 CD68

57BL, 40L-KO 57BL, 40L-KO 57BL, 40L-KO 57BL, 40L-KO C57BL CD40L-KO

Unexposed Mean 36.6 20.9 18.8 11.8 0.2 0.0 28.3 9.3 121,743 82,651SD 12.9 4.5 8.9 0.4 0.3 0.0 7.6 3.7 24,403 16,201

Week 2 Mean 34.9 22.3 19.8 8.7 0.0 0.1 23.5 11.5 155,074 85,265SD 5.1 3.9 12.7 4.7 0.0 0.1 4.9 6.0 50,344 7141

Week 5 Mean 71.0 23.7 18.3 8.8 6.9 0.1 26.1 8.4 216,794 127,033SD 65.1 2.3 4.1 1.0 11.5 0.1 4.1 0.6 88,217 14,666

Week 6 Mean 90.1 21.4 15.6 17.3 378,937SD 16.3 4.6 1.3 2.0 89,283

P versus UnexposedWeek 2 0.85 0.77 0.92 0.52 0.42 0.50 0.41 0.72 0.38 0.86Week 5 0.46 0.54 0.94 0.11 0.42 0.50 0.68 0.78 0.20 0.10Week 6 0.013 0.68 0.001 0.12 0.03

Values represent the mean for three animals per time-point and strain. For all cell populations except CD68� cells, cell numbers were quantitated by manuallycounting 10 fields using a 40-� or 60-� objective. For CD68, MFI for 10 fields was determined as described in Materials and Methods. P values were determinedusing Student’s t-test.


tion of granzymes as well as calgranulin A and B, suggestingthat innate responses do occur in these animals (Supplemen-tary Fig. 2). Thus, CD40L is a critical molecule for inductionof adaptive immune responses to Pneumocystis and is upstreamof essentially all of these adaptive responses. This is consistentwith reports that CD40L is essential in certain models to thelink between innate and adaptive immune responses [68].Additional studies are needed to better understand the mech-anisms by which CD40L orchestrates anti-Pneumocystis re-sponses.

Our study highlights the intricate mechanisms involvingmultiple cell types that lead to the eradication of Pneumocystisin the healthy host and the critical role that CD40L plays in thegeneration of these responses. These studies have identified anumber of potentially important cellular responses that cannow be explored in greater detail to determine the role that theyplay in Pneumocystis infection in healthy and immunocompro-mised hosts.

ACKNOWLEDGMENTS

This research was supported by the Intramural Research Pro-gram of the NIH Clinical Center. The authors have no conflict-ing financial interests. This paper was presented in part at VIIIInternational Workshops on Opportunistic Protists (IWOP-8)and International Conference on Anaerobic Protists (ICAP;Hilo, HI, USA), July 25–29, 2003, Abstract B29; 42nd AnnualMeeting of the Infectious Diseases Society of America (IDSA;Boston, MA, USA), September 30–October 3, 2004, Abstract1078; and 44th Annual Interscience Conference on Antimicro-bial Agents and Chemotherapy (ICAAC; Washington, DC,USA), October 30–November 2, 2004, Abstract B-144. Wethank Drs. Meggan Czapiga and Owen Schwartz for performingthe confocal imaging and quantitation of fluorescence; Drs.John Orlando and Robin Winkler-Pickett for providing anti-Ly49s antibodies; Dr. Hiroki Iwashita for providing anti-ClCa3antibody; Drs. James Shelhamer and Anthony Suffredini forproviding support and advice; and Rene Costello and HowardMostowski for providing animal care.

REFERENCES

1. Kovacs, J. A., Gill, V. J., Meshnick, S., Masur, H. (2001) New insights intotransmission, diagnosis, and drug treatment of Pneumocystis carinii pneu-monia. JAMA 286, 2450–2460.

2. Bishop, L. R., Kovacs, J. A. (2003) Quantitation of anti-Pneumocystisjiroveci antibodies in healthy persons and immunocompromised patients.J. Infect. Dis. 187, 1844–1848.

3. Meuwissen, J. H., Tauber, I., Leeuwenberg, A. D., Beckers, P. J., Sieben,M. (1977) Parasitologic and serologic observations of infection with Pneu-mocystis in humans. J. Infect. Dis. 136, 43–49.

4. Steele, C., Shellito, J. E., Kolls, J. K. (2005) Immunity against theopportunistic fungal pathogen Pneumocystis. Med. Mycol. 43, 1–19.

5. Wiley, J. A., Harmsen, A. G. (1995) CD40 ligand is required for resolutionof Pneumocystis carinii pneumonia in mice. J. Immunol. 155, 3525–3529.

6. Roths, J. B., Marshall, J. D., Allen, R. D., Carlson, G. A., Sidman, C. L.(1990) Spontaneous Pneumocystis carinii pneumonia in immunodeficientmutant scid mice. Natural history and pathobiology. Am. J. Pathol. 136,1173–1186.

7. Shultz, L. D., Schweitzer, P. A., Hall, E. J., Sundberg, J. P., Taylor, S.,Walzer, P. D. (1989) Pneumocystis carinii pneumonia in scid/scid mice.Curr. Top. Microbiol. Immunol. 152, 243–249.

8. Vestereng, V. H., Bishop, L. R., Hernandez, B., Kutty, G., Larsen, H. H.,Kovacs, J. A. (2004) Quantitative real-time polymerase chain-reactionassay allows characterization of Pneumocystis infection in immunocompe-tent mice. J. Infect. Dis. 189, 1540–1544.

9. An, C. L., Gigliotti, F., Harmsen, A. G. (2003) Exposure of immunocom-petent adult mice to Pneumocystis carinii f. sp. muris by cohousing: growthof P. carinii f. sp. muris and host immune response. Infect. Immun. 71,2065–2070.

10. van Kooten, C., Banchereau, J. (2000) CD40-CD40 ligand. J. Leukoc. Biol.67, 2–17.

11. Quezada, S. A., Jarvinen, L. Z., Lind, E. F., Noelle, R. J. (2004) CD40/CD154 interactions at the interface of tolerance and immunity. Annu. Rev.Immunol. 22, 307–328.

12. Etzioni, A., Ochs, H. D. (2004) The hyper IgM syndrome—an evolvingstory. Pediatr. Res. 56, 519–525.

13. Winkelstein, J. A., Marino, M. C., Ochs, H., Fuleihan, R., Scholl, P. R.,Geha, R., Stiehm, E. R., Conley, M. E. (2003) The X-linked hyper-IgMsyndrome: clinical and immunologic features of 79 patients. Medicine(Baltimore) 82, 373–384.

14. Meissner, N. N., Swain, S., Tighe, M., Harmsen, A., Harmsen, A. (2005)Role of type I IFNs in pulmonary complications of Pneumocystis murinainfection. J. Immunol. 174, 5462–5471.

15. Sturn, A., Quackenbush, J., Trajanoski, Z. (2002) Genesis: cluster analysisof microarray data. Bioinformatics 18, 207–208.

16. Dennis Jr., G., Sherman, B. T., Hosack, D. A., Yang, J., Gao, W., Lane,H. C., Lempicki, R. A. (2003) DAVID: Database for Annotation, Visual-ization, and Integrated Discovery. Genome Biol. 4, P3.

17. Hosack, D. A., Dennis Jr., G., Sherman, B. T., Lane, H. C., Lempicki,R. A. (2003) Identifying biological themes within lists of genes with EASE.Genome Biol. 4, R70.

18. Grossman, W. J., Revell, P. A., Lu, Z. H., Johnson, H., Bredemeyer, A. J.,Ley, T. J. (2003) The orphan granzymes of humans and mice. Curr. Opin.Immunol. 15, 544–552.

19. Sedelies, K. A., Sayers, T. J., Edwards, K. M., Chen, W., Pellicci, D. G.,Godfrey, D. I., Trapani, J. A. (2004) Discordant regulation of granzyme Hand granzyme B expression in human lymphocytes. J. Biol. Chem. 279,26581–26587.

20. Lieberman, J. (2003) The ABCs of granule-mediated cytotoxicity: newweapons in the arsenal. Nat. Rev. Immunol. 3, 361–370.

21. Russell, J. H., Ley, T. J. (2002) Lymphocyte-mediated cytotoxicity. Annu.Rev. Immunol. 20, 323–370.

22. Lowin-Kropf, B., Kunz, B., Schneider, P., Held, W. (2002) A role for thesrc family kinase Fyn in NK cell activation and the formation of therepertoire of Ly49 receptors. Eur. J. Immunol. 32, 773–782.

23. Gosselin, P., Lusignan, Y., Brennan, J., Takei, F., Lemieux, S. (1997) TheNK2.1 receptor is encoded by Ly-49C and its expression is regulated byMHC class I alleles. Int. Immunol. 9, 533–540.

24. Sundback, J., Achour, A., Michaelsson, J., Lindstrom, H., Karre, K. (2002)NK cell inhibitory receptor Ly-49C residues involved in MHC class Ibinding. J. Immunol. 168, 793–800.

25. Spaggiari, G. M., Contini, P., Dondero, A., Carosio, R., Puppo, F., Indiveri,F., Zocchi, M. R., Poggi, A. (2002) Soluble HLA class I induces NK cellapoptosis upon the engagement of killer-activating HLA class I receptorsthrough FasL-Fas interaction. Blood 100, 4098–4107.

26. Ondr, J. K., Pham, C. T. (2004) Characterization of murine cathepsin Wand its role in cell-mediated cytotoxicity. J. Biol. Chem. 279, 27525–27533.

27. Gilfillan, S., Ho, E. L., Cella, M., Yokoyama, W. M., Colonna, M. (2002)NKG2D recruits two distinct adapters to trigger NK cell activation andcostimulation. Nat. Immunol. 3, 1150–1155.

28. Lanier, L. L., Corliss, B., Wu, J., Phillips, J. H. (1998) Association ofDAP12 with activating CD94/NKG2C NK cell receptors. Immunity 8,693–701.

29. Kaifu, T., Nakahara, J., Inui, M., Mishima, K., Momiyama, T., Kaji, M.,Sugahara, A., Koito, H., Ujike-Asai, A., Nakamura, A., Kanazawa, K.,Tan-Takeuchi, K., Iwasaki, K., Yokoyama, W. M., Kudo, A., Fujiwara, M.,Asou, H., Takai, T. (2003) Osteopetrosis and thalamic hypomyelinosiswith synaptic degeneration in DAP12-deficient mice. J. Clin. Invest. 111,323–332.

30. Tassi, I., Presti, R., Kim, S., Yokoyama, W. M., Gilfillan, S., Colonna, M.(2005) Phospholipase C-� 2 is a critical signaling mediator for murine NKcell activating receptors. J. Immunol. 175, 749–754.

31. Long, E. O. (2002) Versatile signaling through NKG2D. Nat. Immunol. 3,1119–1120.


32. Honda, K., Yanai, H., Negishi, H., Asagiri, M., Sato, M., Mizutani, T.,Shimada, N., Ohba, Y., Takaoka, A., Yoshida, N., Taniguchi, T. (2005)IRF-7 is the master regulator of type-I interferon-dependent immuneresponses. Nature 434, 772–777.

33. Kawai, T., Sato, S., Ishii, K. J., Coban, C., Hemmi, H., Yamamoto, M.,Terai, K., Matsuda, M., Inoue, J., Uematsu, S., Takeuchi, O., Akira, S..(2004) Interferon- induction through Toll-like receptors involves a directinteraction of IRF7 with MyD88 and TRAF6. Nat. Immunol. 5, 1061–1068.

34. Lu, R., Pitha, P. M. (2001) Monocyte differentiation to macrophagerequires interferon regulatory factor 7. J. Biol. Chem. 276, 45491–45496.

35. Kim, K. I., Giannakopoulos, N. V., Virgin, H. W., Zhang, D. E. (2004)Interferon-inducible ubiquitin E2, Ubc8, is a conjugating enzyme forprotein ISGylation. Mol. Cell. Biol. 24, 9592–9600.

36. D’Cunha, J., Knight Jr., E., Haas, A. L., Truitt, R. L., Borden, E. C. (1996)Immunoregulatory properties of ISG15, an interferon-induced cytokine.Proc. Natl. Acad. Sci. USA 93, 211–215.

37. D’Cunha, J., Ramanujam, S., Wagner, R. J., Witt, P. L., Knight Jr., E.,Borden, E. C. (1996) In vitro and in vivo secretion of human ISG15, anIFN-induced immunomodulatory cytokine. J. Immunol. 157, 4100–4108.

38. Giannakopoulos, N. V., Luo, J. K., Papov, V., Zou, W., Lenschow, D. J.,Jacobs, B. S., Borden, E. C., Li, J., Virgin, H. W., Zhang, D. E. (2005)Proteomic identification of proteins conjugated to ISG15 in mouse andhuman cells. Biochem. Biophys. Res. Commun. 336, 496–506.

39. Recht, M., Borden, E. C., Knight Jr., E. (1991) A human 15-kDa IFN-induced protein induces the secretion of IFN-�. J. Immunol. 147, 2617–2623.

40. Reich, N., Evans, B., Levy, D., Fahey, D., Knight Jr., E., Darnell Jr., J. E.(1987) Interferon-induced transcription of a gene encoding a 15-kDaprotein depends on an upstream enhancer element. Proc. Natl. Acad. Sci.USA 84, 6394–6398.

41. Chain, B. M., Free, P., Medd, P., Swetman, C., Tabor, A. B., Terrazzini, N.(2005) The expression and function of cathepsin E in dendritic cells.J. Immunol. 174, 1791–1800.

42. Lahn, M., Kanehiro, A., Takeda, K., Terry, J., Hahn, Y. S., Aydintug,M. K., Konowal, A., Ikuta, K., O’Brien, R. L., Gelfand, E. W., Born, W. K.(2002) MHC class I-dependent V�4� pulmonary T cells regulate Tcell-independent airway responsiveness. Proc. Natl. Acad. Sci. USA 99,8850–8855.

43. Hobbs, J. A., May, R., Tanousis, K., McNeill, E., Mathies, M., Gebhardt,C., Henderson, R., Robinson, M. J., Hogg, N. (2003) Myeloid cell functionin MRP-14 (S100A9) null mice. Mol. Cell. Biol. 23, 2564–2576.

44. Vandal, K., Rouleau, P., Boivin, A., Ryckman, C., Talbot, M., Tessier,P. A. (2003) Blockade of S100A8 and S100A9 suppresses neutrophilmigration in response to lipopolysaccharide. J. Immunol. 171, 2602–2609.

45. Xu, K., Yen, T., Geczy, C. L. (2001) IL-10 up-regulates macrophageexpression of the S100 protein S100A8. J. Immunol. 166, 6358–6366.

46. Boehm, U., Klamp, T., Groot, M., Howard, J. C. (1997) Cellular responsesto interferon-�. Annu. Rev. Immunol. 15, 749–795.

47. Curtis, J. L. (2005) Cell-mediated adaptive immune defense of the lungs.Proc. Am. Thorac. Soc. 2, 412–416.

48. Leverkoehne, I., Gruber, A. D. (2002) The murine mCLCA3 (alias gob-5)protein is located in the mucin granule membranes of intestinal, respira-tory, and uterine goblet cells. J. Histochem. Cytochem. 50, 829–838.

49. Long, A. J., Sypek, J. P., Askew, R., Fish, S. C., Mason, L. E., Williams,C. M., Goldman, S. J. (2006) Gob-5 contributes to goblet cell hyperplasiaand modulates pulmonary tissue inflammation. Am. J. Respir. Cell Mol.Biol. 35, 357–365.

50. Nakanishi, A., Morita, S., Iwashita, H., Sagiya, Y., Ashida, Y., Shirafuji,H., Fujisawa, Y., Nishimura, O., Fujino, M. (2001) Role of gob-5 in mucusoverproduction and airway hyperresponsiveness in asthma. Proc. Natl.Acad. Sci. USA 98, 5175–5180.

51. Mundhenk, L., Alfalah, M., Elble, R. C., Pauli, B. U., Naim, H. Y.,Gruber, A. D. (2006) Both cleavage products of the mCLCA3 protein aresecreted soluble proteins. J. Biol. Chem. 281, 30072–30080.

52. Gibson, A., Lewis, A. P., Affleck, K., Aitken, A. J., Meldrum, E., Thomp-son, N. (2005) hCLCA1 and mCLCA3 are secreted non-integral membraneproteins and therefore are not ion channels. J. Biol. Chem. 280, 27205–27212.

53. Robichaud, A., Tuck, S. A., Kargman, S., Tam, J., Wong, E., Abramovitz,M., Mortimer, J. R., Burston, H. E., Masson, P., Hirota, J., Sliptez, D.,Kennedy, B., O’Neill, G., Xanthoudakis, S. (2005) Gob-5 is not essentialfor mucus overproduction in preclinical murine models of allergic asthma.Am. J. Respir. Cell Mol. Biol. 33, 303–314.

54. Ferreira, L. R., Silva, A. M., Michailowsky, V., Reis, L. F., Gazzinelli,R. T. (1999) Expression of serum amyloid A3 mRNA by inflammatorymacrophages exposed to membrane glycoconjugates from Trypanosomacruzi. J. Leukoc. Biol. 66, 593–600.

55. Meek, R. L., Eriksen, N., Benditt, E. P. (1992) Murine serum amyloid A3is a high density apolipoprotein and is secreted by macrophages. Proc.Natl. Acad. Sci. USA 89, 7949–7952.

56. Steele, C., Zheng, M., Young, E., Marrero, L., Shellito, J. E., Kolls, J. K.(2002) Increased host resistance against Pneumocystis carinii pneumoniain �� T-cell-deficient mice: protective role of � interferon and CD8(�) Tcells. Infect. Immun. 70, 5208–5215.

57. Zhang, C., Wang, S. H., Lasbury, M. E., Tschang, D., Liao, C. P., Durant,P. J., Lee, C. H. (2006) Toll-like receptor 2 mediates alveolar macrophageresponse to Pneumocystis murina. Infect. Immun. 74, 1857–1864.

58. Rudmann, D. G., Preston, A. M., Moore, M. W., Beck, J. M. (1998)Susceptibility to Pneumocystis carinii in mice is dependent on simulta-neous deletion of IFN-� and type 1 and 2 TNF receptor genes. J. Immunol.161, 360–366.

59. Garvy, B. A., Ezekowitz, R. A., Harmsen, A. G. (1997) Role of � interferonin the host immune and inflammatory responses to Pneumocystis cariniiinfection. Infect. Immun. 65, 373–379.

60. McAllister, F., Ruan, S., Steele, C., Zheng, M., McKinley, L., Ulrich, L.,Marrero, L., Shellito, J. E., Kolls, J. K. (2006) CXCR3 and IFN protein-10in Pneumocystis pneumonia. J. Immunol. 177, 1846–1854.

61. Roths, J. B., Sidman, C. L. (1992) Both immunity and hyperresponsivenessto Pneumocystis carinii result from transfer of CD4� but not CD8� T cellsinto severe combined immunodeficiency mice. J. Clin. Invest. 90, 673–678.

62. Harmsen, A. G., Stankiewicz, M. (1990) Requirement for CD4� cells inresistance to Pneumocystis carinii pneumonia in mice. J. Exp. Med. 172,937–945.

63. Limper, A. H., Hoyte, J. S., Standing, J. E. (1997) The role of alveolarmacrophages in Pneumocystis carinii degradation and clearance from thelung. J. Clin. Invest. 99, 2110–2117.

64. McAllister, F., Steele, C., Zheng, M., Young, E., Shellito, J. E., Marrero,L., Kolls, J. K. (2004) T cytotoxic-1 CD8� T cells are effector cells againstPneumocystis in mice. J. Immunol. 172, 1132–1138.

65. Beck, J. M., Newbury, R. L., Palmer, B. E., Warnock, M. L., Byrd, P. K.,Kaltreider, H. B. (1996) Role of CD8� lymphocytes in host defenseagainst Pneumocystis carinii in mice. J. Lab. Clin. Med. 128, 477–487.

66. Lund, F. E., Hollifield, M., Schuer, K., Lines, J. L., Randall, T. D., Garvy,B. A. (2006) B cells are required for generation of protective effector andmemory CD4 cells in response to Pneumocystis lung infection. J. Immunol.176, 6147–6154.

67. Lund, F. E., Schuer, K., Hollifield, M., Randall, T. D., Garvy, B. A. (2003)Clearance of Pneumocystis carinii in mice is dependent on B cells but noton P. carinii-specific antibody. J. Immunol. 171, 1423–1430.

68. Fujii, S., Liu, K., Smith, C., Bonito, A. J., Steinman, R. M. (2004) Thelinkage of innate to adaptive immunity via maturing dendritic cells in vivorequires CD40 ligation in addition to antigen presentation and CD80/86costimulation. J. Exp. Med. 199, 1607–1618.


Download - Immune responses to Pneumocystis murina are robust in healthy mice but largely absent in CD40 ligand-deficient mice

Top Related