calcineurin target crza regulates conidial germination ...calcineurin effectors have not been...

EUKARYOTIC CELL, July 2008, p. 1085–1097 Vol. 7, No. 71535-9778/08/$08.00�0 doi:10.1128/EC.00086-08Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Calcineurin Target CrzA Regulates Conidial Germination, HyphalGrowth, and Pathogenesis of Aspergillus fumigatus�†

Robert A. Cramer, Jr.,1‡ B. Zachary Perfect,2 Nadthanan Pinchai,2 Steven Park,3 David S. Perlin,3Yohannes G. Asfaw,4 Joseph Heitman,1,5,6 John R. Perfect,1,6 and William J. Steinbach1,2*

Department of Molecular Genetics and Microbiology, Duke University Medical Center, Durham, North Carolina1; Department ofPediatrics, Duke University Medical Center, Durham, North Carolina2; Public Health Research Institute, International Center forPublic Health, UMDNJ-New Jersey Medical School, University of Medicine and Dentistry of New Jersey, Newark, New Jersey3;

Department of Laboratory Animal Resources, Duke University Medical Center, Durham, North Carolina4; Department ofPharmacology and Cancer Biology, Duke University Medical Center, Durham, North Carolina5; and Department of

Medicine, Duke University Medical Center, Durham, North Carolina6

Received 9 March 2008/Accepted 17 April 2008

The calcineurin pathway is a critical signal transduction pathway in fungi that mediates growth, morphol-ogy, stress responses, and pathogenicity. The importance of the calcineurin pathway in fungal physiologycreates an opportunity for the development of new antifungal therapies that target this critical signalingpathway. In this study, we examined the role of the zinc finger transcription factor Crz1 homolog (CrzA) in thephysiology and pathogenicity of the opportunistic human fungal pathogen Aspergillus fumigatus. Geneticreplacement of the crzA locus in A. fumigatus resulted in a strain with significant defects in conidial germi-nation, polarized hyphal growth, cell wall structure, and asexual development that are similar to but withdifferences from defects seen in the A. fumigatus �cnaA (calcineurin A) strain. Like the �cnaA strain, the �crzAstrain was incapable of causing disease in an experimental persistently neutropenic inhalational murine modelof invasive pulmonary aspergillosis. Our results suggest that CrzA is an important downstream effector ofcalcineurin that controls morphology in A. fumigatus, but additional downstream effectors that mediatecalcineurin signal transduction are likely present in this opportunistic fungal pathogen. In addition, theimportance of CrzA to the production of disease is critical, and thus CrzA is an attractive fungus-specificantifungal target for the treatment of invasive aspergillosis.

Invasive pulmonary aspergillosis (IPA) is a devastating dis-ease characterized by extensive invasion of lung tissue in im-munocompromised patients by hyphae of saprophytic molds inthe genus Aspergillus (26, 48, 61). As the number of immuno-compromised patients continues to rise with advances in med-ical technology, there has been a concomitant increase in thenumber of patients with IPA (29). Despite expanding treat-ment options for IPA, current therapy still has a dismal 40 to50% success rate (19, 28). Thus, the development of new an-tifungal therapies to augment existing treatment strategies isurgently needed.

One potential antifungal drug target in Aspergillus species isthe calcineurin pathway (55). The importance of the cal-cineurin signaling pathway in fungal physiology is vividly dem-onstrated by gene expression analyses in Saccharomyces cerevi-siae that identified 163 genes regulated by the calcineurin/Crz1signaling pathway. These regulated genes were involved inion/small-molecule transport, signal transduction, cell wallmaintenance, and vesicular transport, indicating the importantrole that this pathway plays in fungal physiology (66).

Unlike the case for many new candidate antifungal drugtargets, drugs that target the calcineurin pathway already exist.The calcineurin inhibitors FK506 and cyclosporine, which areused routinely to inhibit human calcineurin and have revolu-tionized modern organ transplantation, exhibit antifungal ac-tivity in vitro, indicating their potential as therapeutic agentsfor IPA (3, 20, 54, 56, 57). In addition, our group and anotherindependent research group have shown that deletion of thegene encoding the calcineurin A catalytic subunit in Aspergillusfumigatus resulted in a strain with significant defects in hyphalgrowth. Furthermore, these �cnaA strains are no longer capa-ble of causing IPA in distinct experimental murine models(14, 53).

The importance of the calcineurin pathway in fungal physi-ology and the dramatic phenotypes associated with pharmaco-logic inhibition and genetic deletions of this pathway make itan attractive antifungal drug target. However, the critical roleof calcineurin in mammalian immune system responses bringsinto question the feasibility of directly targeting fungal cal-cineurin pathways in vivo with existing calcineurin inhibitors(8, 27, 34, 36). The impact of calcineurin inhibition on the host,via immunosuppression, may be stronger than its antifungalactivity, with the net result being exacerbation of the disease.Thus, further studies are needed to identify fungus-specificcomponents of the calcineurin signaling pathway that are notpresent in mammals. In this study, we characterize a cal-cineurin effector protein, CrzA, which is not present in mam-mals but is integrally linked with calcineurin-mediated signaltransduction in many fungi (12, 21, 23, 43, 50).

* Corresponding author. Mailing address: Duke University MedicalCenter, Box 3499, Pediatric Infectious Diseases, Durham NC 27710.Phone: (919) 681-1504. Fax: (919) 684-8902. E-mail: [email protected].

† Supplemental material for this article may be found at http://ec.asm.org/.

‡ Present address: Department of Veterinary Molecular Biology,Montana State University, Bozeman, MT.

� Published ahead of print on 2 May 2008.

1085

on February 1, 2020 by guest

http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

In S. cerevisiae, calcineurin-dependent transcription is regu-lated in part by the zinc finger transcription factor Crz1 (30,50). Calcineurin regulates Crz1 activity via dephosporylation ofCrz1, which results in nuclear localization of cytoplasmic Crz1and requires the karyopherin Nmd5 (39, 51). S. cerevisiae �crz1mutants display many phenotypes, including growth and via-bility defects, that are similar to those of S. cerevisiae cal-cineurin mutants (30, 50). Additional downstream effectors ofcalcineurin have also been identified. In S. cerevisiae, Hph1 andHph2 have been shown to be novel calcineurin-regulated re-sponse components and Crz1-independent downstream effec-tors of calcineurin (18).

Additional downstream effectors of calcineurin are alsolikely to exist in the opportunistic human fungal pathogenCandida albicans. C. albicans crz1 mutant strains display apathobiologic phenotype different from that of C. albicanscalcineurin mutants (23, 37, 43). For example, C. albicans�crz1/�crz1 mutant strains, unlike C. albicans calcineurin mu-tant strains, are not sensitive to serum and do not display adecrease in virulence in murine models of candidiasis. Theseresults suggest the presence of additional calcineurin down-stream effectors in this pathogenic yeast.

However, in the pathogenic mold A. fumigatus downstreamcalcineurin effectors have not been identified or functionallycharacterized. In this study, we identified the A. fumigatus Crz1homolog, CrzA, and tested the hypothesis that CrzA is a pri-mary downstream effector of the calcineurin pathway in A.fumigatus. Our results suggest that CrzA is an important down-stream effector for calcineurin in A. fumigatus but that addi-tional calcineurin downstream effectors also exist in this mold.We observed that CrzA is critical for in vivo fungal growth,development, and pathogenicity and thus is an excellent anti-fungal drug target for treatment of invasive aspergillosis due toits role in pathogenesis and lack of a conserved mammalianhomolog.

MATERIALS AND METHODS

Strains, media, and growth conditions. Aspergillus fumigatus strain AF293.1was used to generate the crzA replacement (�crzA) strain (�crzA::Aspergillusparasiticus pyrG). AF293.1 is a uracil/uridine-auxotrophic (pyrG�) mutant of A.fumigatus strain AF293 (38), and AF293 was used as the wild-type strain for allexperiments. All A. fumigatus cultures were grown on glucose minimal mediumas previously described (9, 47, 53) at 37°C. The Escherichia coli One-Shot TOP10strain (Invitrogen, Carlsbad, CA) was used for routine cloning and grown inLuria broth (Fisher Chemicals, Fair Lawn, NJ) supplemented with appropriateantibiotics at 37°C.

Generation of �crzA and �crzA � crzA reconstituted strains of A. fumigatus.The 2.3-kb A. fumigatus crzA gene (Afu1g06900 [www.cadre.man.ac.uk]) wasreplaced with the 3.1-kb A. parasiticus pyrG gene to generate �crzA (Fig. 1A).Approximately 1 kb of upstream flanking and 1.2 kb of downstream flankingsequence of crzA was cloned to flank the 3.1-kb A. parasiticus pyrG in plasmidpJW24 (a gift from Nancy Keller, University of Wisconsin, Madison) to generatethe replacement construct. This resulting replacement construct plasmid wasused as a template to generate the approximately 4.5-kb PCR amplicon for usein transformation.

Isolation of fungal protoplasts and polyethylene glycol-mediated transforma-tion of A. fumigatus were performed as previously described (5, 9, 53). Trans-formants were initially screened by PCR with primers designed to amplify crzA,which is predicted to be absent in �crzA strains. Transformants were additionallyscreened by PCR for the junctions of the crzA::pyrG allele to indicate replace-ment and homologous recombination. Strains with both absence of crzA andevidence of homologous recombination, as determined by PCR, then underwentsingle-spore isolation. Confirmation of gene replacement was performed viaSouthern analysis using the digoxigenin labeling system (Roche Molecular Bio-chemicals, Mannheim, Germany) and three separate probes, including a 500-bpprobe for crzA, a 1.2-kb probe for the 3� crzA flanking sequence, and a 600-bpprobe for the A. parasiticus pyrG gene.

�crzA was complemented to generate the �crzA � crzA complemented strainby transformation with native crzA, 550 bp of flanking upstream sequence, and1.7 kb of flanking downstream sequence. The native crzA and flanking sequenceswere cloned into the TOPO TA pCR 2.1 system (Invitrogen, Carlsbad, CA),followed by EcoRV blunt cloning of a 1.4-kb segment of the E. coli hygromycinB phosphotransferase gene (hph), conferring hygromycin B resistance, fromplasmid pCB1636 (59). The approximately 4.5-kb complementation cassette wasamplified from this new TOPO vector (named pZP5) and the PCR producttransformed as detailed above into the recipient �crzA strain. Putative comple-mented transformants, with ectopic integrations of the complementation con-struct, were selected on medium supplemented with hygromycin B (150 �g/ml).

FIG. 1. Construction of A. fumigatus �crzA and �crzA � crzA complemented strains. (A) Schematic of crzA gene replacement in A. fumigatusstrain 293.1. The wild-type crzA locus (2.3 kb) was replaced with the 3.1-kb A. parasiticus pyrG gene from plasmid pJW24. (B) Southern blot analysisof SmaI-digested genomic DNA with a 1.2-kb digoxigenin-labeled probe revealed successful gene replacement of crzA with a single copy of pyrGin a transformant (KO) and successful ectopic reintroduction of the wild-type (WT) crzA gene (2.7-kb band) in the crzA mutant background(Comp). Lane L, ladder.

1086 CRAMER ET AL. EUKARYOT. CELL


http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

PCR analyses were performed to screen for crzA as described above, and strainswith amplicons underwent single-spore isolation followed by Southern analysesto confirm that isolates contained both crzA and hphB.

Radial growth, conidiation, and germination. Radial growth on solid mediumwas obtained by measuring colony diameters of the wild-type, �crzA, and �crzA �crzA complemented strains once every 24 h over a period of 5 days as previouslydescribed (53). Experiments were performed in triplicate, and the mean andstandard error of the colony diameter at each 24-hour period are reported.

Conidiation was quantified by harvesting conidia from the wild-type, �crzA,and �crzA � crzA complemented strains as previously described (53). Harvestsfrom each strain were performed in triplicate, and the mean and standard errorof the total number of conidia per ml collected from each strain are reported.

The germination percentage and rate were determined with the wild-type,�crzA, �crzA � crzA complemented, and �cnaA strains. Conidia (1 � 106

conidia/ml) of the respective strains were inoculated into RPMI 1640 brothsupplemented with 0.15% (wt/vol) Junlon to promote dispersal as previouslydescribed (7) and were incubated over a time course at 37°C. Sampling wasperformed at 4, 6, 8, 10, 12, 16, 24, and 48 h, and samples were then fixed in 10%neutral buffered formalin and analyzed by light microscopy. Germination rateswere calculated in triplicate by scoring 100 conidia for germination to calculatea germination percentage. Hyphal morphology was also assessed at each timepoint. Conidia (1 � 105/ml) were incubated in RPMI for 48 h as described above,and hyphae and ungerminated conidia were examined for viability using thefluorescent dye 5,(6)-carboxyfluorescein diacetate, which stains live cells. (7, 54).

Quantification of cell wall �-1,3-glucan. The aniline blue fluorescence assay ishighly specific for quantifying �-1,3-glucan (46). Using a previously reported (54)slight modification of an earlier study (22), flasks containing 25 ml of RPMI wereinoculated with 1 � 107 conidia/ml of the A. fumigatus strains with and withoutantifungals and grown for 48 h with shaking at 200 rpm and 37°C. The myceliawere vortexed, harvested, washed three times in 0.1 M NaOH, and lyophilized.Mycelial tissue was weighed and normalized to place 5 mg of tissue from eachsample in a new tube containing 0.25 ml of 1 M NaOH. This tube was thensonicated with a microprobe. Tubes were incubated at 52°C for 30 min, andaliquots of 50 �l were treated with aniline blue reagent (22, 46) for 30 min at52°C followed by 30 min at 25°C. Relative fluorescence units were measured ona SpectraMax M2 fluorimeter (Molecular Devices, Sunnyvale, CA) at 405-nmexcitation and 460-nm emission wavelengths. A standard curve was created usingcurdlan, a �-1,3-glucan analog (Sigma, St. Louis, MO), to normalize values.Values are expressed as relative fluorescence units per mg of mycelial tissue; thevalues were analyzed using an unpaired t test and statistical significance isreported as a two-tailed P value, with significance defined as a P value of �0.05.

Cell wall inhibitor effects on radial growth. The activities of a selective �-1,3-glucan synthesis inhibitor (caspofungin at 1 �g/ml), a chitin synthesis inhibitor(nikkomycin Z at 1 �g/ml), and a calcineurin inhibitor (FK506 at 20 ng/ml) weremeasured using minimal medium impregnated with clinically relevant concen-trations. The antifungal activity of each was measured by radial growth followinginoculation of 1 � 104 conidia (10 �l of a solution of 1 � 106 conidia/ml) of theA. fumigatus strain. Plates were incubated at 37°C, and images were obtainedafter 4 days of growth. Each experiment was performed in triplicate.

Transcriptional profiling of calcineurin and cell wall-related genes. One-stepSybr green I quantitative reverse transcriptase PCR (qRT-PCR) was performedfor transcriptional profiling of calcineurin and cell wall-related genes. Primerswere designed to the calcineurin, cell wall-related genes, and the housekeepingbeta-tubulin gene using Beacon Designer software version 5.0 (Premier BiosoftInternational, Palo Alto, CA). All primers used in this study are listed in TableS1 in the supplemental material. All qRT-PCRs were performed using a Strat-agene Mx3005p quantitative PCR system instrument (Stratagene, La Jolla, CA).Each primer set was optimized by using serially diluted total RNA extracted fromwild-type A. fumigatus and the Brilliant II Sybr green qRT-PCR Master Mix Kit,1-Step (Stratagene), in a 25-�l reaction volume according to the manufacturer’sassay optimization instructions. The optimized qRT-PCR assay for each RNAsample was carried out in 25-�l reaction volumes that consisted of 1� Sybr greenreaction mix, 50 nM of each primer, 100 ng of total RNA, and 1 �l of theRT-RNase block enzyme mixture. No-template and no-RT controls for eachprimer set were also assayed to confirm that no primer-dimers and no DNAcontamination were present, respectively. The thermal cycling parameters con-sisted of a 30-min RT step at 48°C and a 10-min Taq polymerase hot start at 95°C,followed by template amplification of 45 cycles of 95°C for 30 s, 55°C for 30 s, and72°C for 30 s. Fluorescence was measured during the extension step (72°C).Following amplification, a disassociation analysis (melt curve) was performed toconfirm that a single amplified product was present. The qRT-PCR assay wasperformed in duplicate. The threshold cycle (CT) for each reaction was deter-mined using the Stratagene MxPro v.4 software’s “adaptive baseline” algorithm

enhancements. For each RNA sample, expression levels of all genes of interestwere normalized to beta-tubulin. The 2��CT method of analysis was used todetermine fold changes of gene expression in the mutants relative to the wild-type AF293 A. fumigatus strain.

Light and scanning electron microscopy. All light microscopy was performedon a Zeiss microscope using Nomarski optics (differential interference contrast).All scanning electron microscopy images were obtained using an environmentalscanning electron microscope (Philips XL30 ESEM TMP; FEI Company, Hill-sboro, OR) as previously described (53).

Murine inhalational model of invasive aspergillosis. Six-week-old CD1 malemice (Charles River Laboratories) (mean weight, 22.5 g) were immunosup-pressed with both cyclophosphamide (Cytoxan; Bristol-Myers Squibb, Princeton,NJ) at 150 mg/kg intraperitoneally on days �2 and �3 of infection and triam-cinolone acetonide (Kenalog-40; Bristol-Myers Squibb, Princeton, NJ) at 40mg/kg subcutaneously on days �1 and �6 of infection as previously described(53). Mice were housed under sterile conditions and supplied sterile drinkingwater supplemented with vancomycin (1 mg/ml), gentamicin (0.2 mg/ml), andclindamycin (1 mg/ml). Four groups of 20 immunosuppressed, unanesthetizedmice each inhaled 40 ml of an aerosolized suspension of the 1 � 109 conidia/mlof AF293 wild-type, �crzA, or �crzA � crzA complemented strain or a diluentcontrol (0.05% Tween 80) in a Hinners inhalational chamber for 30 min aspreviously described (53). Mice were evaluated for morbidity and mortality in ablinded fashion twice daily. Survival was plotted on a Kaplan-Meier curve foreach Aspergillus strain, and the log rank test was used for pairwise comparison ofsurvival (GraphPad Prism 5.0; GraphPad, San Diego, CA). Statistical signifi-cance was defined as a two-tailed P value of �0.05.

Histopathologic evaluation of fungal burden. To evaluate the in vivo his-topathologic progression of disease, four groups of four additional mice weresimilarly infected in the murine inhalational model with each strain and eutha-nized at defined time points (days 4 and 7 after infection). Lungs were harvestedand prepared for histologic examination as previously described (53).

RESULTS

Identification of the crzA homolog in A. fumigatus. Saccha-romyces cerevisiae Crz1 (S. cerevisiae genome database acces-sion no. YNL027W) was used with the BLASTP algorithm tosearch the A. fumigatus genome database. The search returneda subject, Afu1g06900, with an expected value of 1.2e�46 andan identity of 65%. The predicted protein sequence forAfu1g06900 was then used in a reciprocal BLASTP searchagainst the S. cerevisiae genome database. This search re-turned a yeast protein, YNL027W (Crz1), as the best match.This result suggested that Afu1g06900 was homologous to S.cerevisiae Crz1, YNL027W. Additional reciprocal BLASTanalyses were conducted with additional Crz1 homologs inNeurospora crassa (NCU07952.3) and Candida albicans(orf19.7359) and confirmed the result that Afu1g06900 is mostlikely the A. fumigatus Crz1 homolog (CrzA).

The predicted calcineurin interaction site in S. cerevisiaeCrz1, PIISIQ, was not identified in the amino acid sequence ofA. fumigatus CrzA, nor was the known docking site for cal-cineurin on the mammalian transcription factor NFAT, PxIxITidentified (Fig. 2) (6). However, significant amino acid identitywas shared between Crz1 and CrzA in the serine-rich regionknown to be involved in calcineurin-dependent regulation ofCrz1 localization, activity, and phosphorylation (51). Severalserine residues are conserved between Crz1 and CrzA in thisregion (amino acids 186 to 232) (Fig. 2).

The Crz1 nuclear localization signal (NLS) (residues 394 to422) does not seem to be conserved in CrzA in the samelocation (Fig. 2) (39). A putative NLS may exist in CrzA down-stream of the Crz1 NLS site, amino acids 460 to 490. Thisregion contains the basic amino acids arginine and lysine,which are typically found in NLS regions. In addition, this

VOL. 7, 2008 CrzA AND A. FUMIGATUS PATHOGENICITY 1087


http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

putative NLS site is serine rich, similar to Crz1’s NLS region(Fig. 2). The highest degree of sequence similarity betweenCrz1 and CrzA occurs in the C terminus, where both proteinshave two conserved zinc finger domains (residues 576 to 672)of the C2H2 class (Fig. 2). Taken together, these in silicoresults strongly suggest that the protein Afu1g06900 (CrzA) isa Crz1 homolog.

Generation of A. fumigatus crzA strains. Transformation ofthe uracil/uridine-auxotrophic A. fumigatus AF293.1 with thecrzA replacement construct yielded 26 transformants, including7 which had PCR products with primers designed to amplifythe predicted �crzA replacement locus after homologous re-combination, and no detectable PCR amplicon with primersdesigned to amplify the wild-type crzA locus (data not shown).These seven strains underwent Southern analysis with SacI/XbaI-digested genomic DNA and a 1.2-kb probe for the 3�segment of crzA flanking sequence, which revealed that two ofthe seven transformants had successful homologous recombi-nation events at the crzA locus without presence of the wild-type allele (data not shown). An additional Southern analysisusing SmaI-digested genomic DNA (Fig. 1B) and a 500-bpprobe from a section of the 3� flank of the crzA replacementconstruct was performed. This hybridization yielded two bands(4.7 kb and 2.7 kb) in the wild-type strain and the two predictedbands for a gene replacement (4.7 kb and 6.2 kb) in the �crzAstrain due to successful homologous recombination at the crzAlocus. A final Southern analysis with XhoI-digested genomicDNA and a 550-bp probe for the A. parasiticus pyrG geneyielded a single 3.9-kb band in both �crzA strains and theabsence of a band in the wild type, confirming the homologous

replacement of crzA with a single insertion of the gene replace-ment cassette. The two �crzA strains were phenotypically iden-tical, and one was selected to continue as the �crzA strain forfurther experimentation.

To complement the �crzA defect, a 4.5-kb complementationcassette (from pZP5) was transformed into the recipient �crzAstrain for ectopic integration. A total of 85 transformantswhich grew on the hygromycin B selection medium were ob-tained. PCR screening of transformants with the same 500-bpsection of crzA revealed two colonies with the expected crzAproduct. Southern analysis with SmaI-digested genomic DNA(Fig. 1B) revealed the persistence of the �crzA locus (4.7-kband 6.2-kb bands) and an additional 2.7-kb band correspond-ing to the wild-type crzA allele integrated in an ectopic locationin the genome. Additionally, the complemented strain wasverified by probing for the hygromycin B resistance (hph) gene(data not shown). One strain was selected as the �crzA � crzAcomplemented strain for further experimentation.

Confirmation of crzA gene replacement and complementa-tion was also done using real-time RT-PCR, which revealedsimilar crzA mRNA abundances in both the wild-type and�crzA � crzA complemented strains but no expression in the�crzA strain (data not shown).

crzA is necessary for germination, radial growth, and asex-ual development. Conidia of the �crzA strain displayed a sig-nificant delay and defect in germination at 37°C in RPMImedium (Fig. 3). At 8 h postinoculation, 48% and 59% of the�crzA � crzA complemented and wild-type strains, respec-tively, had germinated, while only 18% and 6% of the �cnaAand the �crzA strain conidia, respectively, germinated. Up

FIG. 2. Multiple alignment of Saccharomyces cerevisiae Crz1p (YNL027W) and putative Aspergillus fumigatus CrzA (Afu1g06900), created withCLUSTALW using the BLOSUM62 scoring matrix and default parameters. Conserved residues are in black. *, serine-rich region; ˆ, NLS site inCrz1p; $, calcineurin interaction site present in Crz1p but absent in CrzA; #, C2H2 zinc finger domain region.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

until 24 h postinoculation, the �crzA strain displayed an evengreater defect in conidial germination than the �cnaA strain.By 48 h postinoculation only 37% of the �crzA conidia hadgerminated, compared with 81% and 95% of conidia from the�crzA � crzA complemented and wild-type strains, respec-tively. Importantly, 5,(6)-carboxyfluorescein diacetate viabilitystaining revealed that the ungerminated �crzA conidia after

48 h of incubation were still viable (data not shown). Theseresults demonstrate the importance of calcineurin signaling, asprimarily directed through CrzA, for A. fumigatus conidiumgermination.

In addition to significant defects in germination, the �crzAstrain demonstrated a severe defect in polarized hyphal growth(Fig. 4A and B). Radial growth assays demonstrated a 68%decrease in hyphal growth compared with the wild-type and�crzA � crzA complemented strain (P � 0.0001). Further-more, the �crzA strain was also defective in conidium produc-tion (P � 0.0001) (Fig. 4C); when normalized to the decreasedcolony area, the �crzA strain still possessed a 73% decrease inconidia produced per mm2 compared to the wild-type strain.Scanning electron microscopy of the hyphae of the strains alsorevealed that �crzA hyphae were slightly misshapen withblunted hyphal tips compared with the wild-type and �crzA �crzA complemented strains (Fig. 5). However, the hyphal mor-phology defect was not as severe as previously seen in the�cnaA strain (53). Similar to the case for the �cnaA strain, theconidial surfaces of the �crzA strain were relatively smoothcompared to those of the wild-type and �crzA � crzA comple-mented strains (Fig. 6) yet not as smooth as those of the �cnaAstrain. This may be due to a lack of rodlet formation; however,this was not definitively examined. These results indicate thatthe �crzA strain has a phenotype remarkably similar to that ofthe �cnaA strain, with a slightly more severe defect in germi-nation but less of a defect in hyphal growth and conidial mor-phology. These results also suggest the likely existence of ad-

FIG. 3. The calcineurin pathway is required for in vitro conidiumgermination. Conidium germination of the wild-type (AF293), �cnaA,�crzA, and �crzA � crzA strains was examined at 37°C in RPMI 1640liquid medium over a 48-h time course. Significant differences in boththe rate and quantity of conidia germination are seen among thewild-type, the crzA complemented strain, and the cnaA and crzA mu-tant strains. Percent germination is based on analysis of 100 conidiaunder light microscopy (magnification, �400), repeated in triplicate.

FIG. 4. (A and B) CrzA is necessary for in vitro hyphal growth. (A) In vitro culture morphology of the wild-type (AF293), �crzA, and �crzA � crzAstrains at 37°C on glucose minimal medium after 4 days of growth. (B) The �crzA strain displayed a 68% decrease in radial growth over a 5-daytime course compared with the wild-type and complemented strains (P � 0.0001). (C) CrzA is required for the production of asexual conidia inA. fumigatus. Conidia were harvested from glucose minimal medium plates grown at 37°C as described in Materials and Methods. Conidia werequantified with a hemacytometer. The �crzA strain had a 73% decrease in conidia produced per mm2 of colony hyphal growth compared to thewild-type strain (P � 0.0001). Error bars indicate standard errors.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

ditional calcineurin effectors involved in polarized hyphalgrowth and morphology.

Cell wall �-1,3-glucan levels are reduced in the �cnaA and�crzA strains. Defects in hyphal growth may be due to defectsin the synthesis of important cell wall components such aschitin and �-1,3-glucan. We utilized an aniline blue fluores-cence assay to measure the �-1,3-glucan content in hyphae ofthe wild-type, �cnaA, and �crzA strains. Figure 7 reveals thatthe �cnaA strain had a statistically significantly reduced �-1,3-glucan content in hyphae compared to the wild-type strain (P 0.03). While a difference in fluorescence was consistently seenbetween the �crzA and wild-type strains, this was not statisti-cally significant (P 0.09), and it may reflect the greater abilityof the �crzA strain to form hyphae in vitro due to an unchar-acterized crzA-independent pathway in A. fumigatus.

When treated with the �-1,3-glucan synthesis inhibitorcaspofungin, all three strains showed statistically significantdifferences in �-1,3-glucan content compared to the untreatedsamples (wild type compared to wild type with caspofungin, P 0.03; �cnaA strain compared to �cnaA strain with caspofungin,P � 0.0001; �crzA strain compared to �crzA strain with caspo-fungin, P 0.01). However, when the strains were treated withthe chitin synthesis inhibitor nikkomycin Z, a statistically sig-nificant compensatory increase in �-1,3-glucan content wasseen in the �cnaA strain (P � 0.0001) but not in the �crzA(P 0.68) and wild-type (P 0.12) strains. These resultssuggest that the calcineurin pathway is involved in mediating�-1,3-glucan biosynthesis in A. fumigatus but that CrzA plays alimited role in this regulation. In addition, it seems apparentthat calcineurin-independent pathways for �-1,3-glucan bio-synthesis also exist in A. fumigatus and may be activated in cellsconfronted with chitin synthesis inhibitors.

We also transcriptionally profiled genes predicted to be in-

volved in the calcineurin pathway and cell wall biosynthesis inA. fumigatus to help clarify the role of CnaA and CrzA in A.fumigatus physiology (Table 1). With the exception of gelA, themRNA abundance of genes involved in �-1,3-glucan biosyn-thesis was consistently decreased in the �cnaA strain whileremaining at virtually wild-type levels in the absence of CrzA.These data correlate with the decrease in the �-1,3-glucancontent observed in the �cnaA strain and suggest that, unlikein other fungi, CrzA does not appear to play a significant rolein the regulation of �-1,3-glucan biosynthesis in A. fumigatus.A gene replacement mutation of gelA (gel1) was shown to haveno effect on the cell wall composition of A. fumigatus. Thus, itis not surprising that expression of gelA in the �cnaA strain wasnot able to rescue the defect in �-1,3-glucan biosynthesis (32).It therefore appears that calcineurin A is involved in regulating�-1,3-glucan biosynthesis in A. fumigatus but that this regula-tion is not mediated solely by CrzA.

Other significant results from the mRNA abundance exper-iments were also observed. The large increase, 52-fold, incalmodulin mRNA in the �cnaA strain highlights the impor-tance of calcineurin in mediating responses to calcium (Table1). Interestingly, a similar compensatory increase in calmodu-lin mRNA was not observed in the �crzA strain. With theexception of a fourfold increase of calcineurin mRNA inthe �crzA strain, elimination of CrzA had minor effects on themRNA abundance of the genes examined in this study (Table1). These results strongly suggest that CrzA-independent cal-cineurin effectors controlling cell wall biosynthesis and re-sponses to calcium exist in A. fumigatus.

Cell wall inhibitors inhibit hyphal growth in calcineurinpathway mutants. Our data suggest that the calcineurin path-way plays an important role in cell wall biosynthesis in A.fumigatus but that this may be CrzA independent. Evidence

FIG. 5. The �crzA strain has malformed hyphae. Scanning electron microscopy of conidia from the three A. fumigatus strains grown at 37°Cin glucose minimal medium broth for 48 h is shown. The crzA mutant strain shows abnormal, malformed hyphal tips that form abnormal mycelia.Levels of magnification are as shown. Bars, 100 �m (top row) and 50 �m (bottom row).



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

that additional calcineurin-independent mechanisms of cellwall biosynthesis and hyphal growth exist in A. fumigatus isobserved by the enhanced effects of calcineurin and/or CrzAinhibition and specific cell wall inhibitors (Fig. 8). Inhibition ofthe calcineurin pathway via FK506, of �-1,3-glucan biosynthe-sis via caspofungin, or of chitin biosynthesis via nikkomycin Zhas clear effects on hyphal growth in the wild-type strain (Fig.8). Additional evidence for the existence of calcineurin-inde-pendent regulation of �-1,3-glucan biosynthesis is the in-creased inhibition of hyphal growth with caspofungin treat-ment of the �cnaA strain. Treatment of the �crzA strain withFK506 resulted in a phenotype virtually identical to that of the�cnaA strain (Fig. 8). This result may suggest that CrzA isdependent upon the presence of calcineurin in the cell for itsactivity. Additional support of this conclusion is the 4-folddecrease in crzA mRNA in the fungal strain lacking calcineurin(Table 1).

CrzA is necessary for pathogenicity in an experimental mu-rine model of IPA. Given the significant growth and morpho-logical defects in the �crzA strain, we hypothesized that thisstrain would be unable to establish invasive disease in an im-munosuppressed murine model of IPA, similar to our resultswith the �cnaA strain (53). To test the virulence of �crzA, a

FIG. 6. Conidia of the �crzA strain show abnormal conidial surface morphology. Scanning electron microscopy of sputter-coated A. fumigatusconidia from glucose minimal medium plates is shown. Levels of magnification are as shown. Bars, 10 �m (top row) and 2 �M (middle and bottomrows). �crzA strain conidia lack the bumpy and echinulate surface observed on wild-type and reconstituted strains.

FIG. 7. Cell wall �-1,3-glucan is decreased in calcineurin mutants. Ani-line blue fluorescence, measuring �-1,3-glucan in the cell wall, shows statis-tically significant decreases in �-1,3-glucan content in the �cnaA strain com-pared with the wild-type strain AF293 (P 0.03). In addition, treatment withthe �-1,3-glucan synthase inhibitor caspofungin results in a further decreaseof �-1,3-glucan content in A. fumigatus strains in the wild-type strain (P 0.03), in the absence of calcineurin A (P � 0.0001), and in the absence ofCrzA (P 0.011). Decreases in �-1,3-glucan content in the �crzA strain werenot statistically significant compared to the wild-type strain (P 0.09). Treat-ment of the strains with the chitin inhibitor nikkomycin Z resulted in acompensatory increase in the �cnaA strain (P � 0.0001), indicating thatcalcineurin-independent mechanisms of �-1,3-glucan biosynthesis exist in A.fumigatus. Asterisks indicate statistical significance (P � 0.05). Error barsindicate standard errors.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

persistently neutropenic inhalational murine model of IPA wasutilized (52). Infection with the wild-type strain yielded 85%mortality by 14 days after infection (Fig. 9A), while infectionwith the �crzA strain or inoculation of mice with 0.05% Tween80 did not lead to any mortality (P 0.0011). Mice infectedwith the �crzA � crzA complemented strain had 65% mortalityby 14 days after infection, which was not statistically different

from the mortality caused by the wild-type infection. Animalsinfected with the �crzA strain were symptomatically indistin-guishable from uninfected control animals during the entireexperimental course. Animals infected with the wild-type and�crzA � crzA complemented strains displayed progressive andsevere signs of invasive disease that have been previously welldescribed, including ruffled fur, hunched posture, weight loss,and an increased respiratory rate.

Histopathologic analysis revealed extensive invasive hyphaethroughout lung sections from additional animals infected inthe murine inhalational model with both the wild-type and�crzA � crzA complemented strains (Fig. 9B). In mice infectedwith the �crzA strain, hyphal growth was occasionally visible,consistent with the in vitro growth phenotype of this mutantstrain. However, compared to that in mice infected with thewild-type and �crzA � crzA complemented strains, hyphalgrowth was significantly decreased in �crzA strain-infectedmice. In addition, there was minimal lung inflammation inanimals infected with the �crzA strain and a greater number ofopen alveoli. In animals infected with the wild-type and �crzA �crzA complemented strains, infiltration of the alveoli withphagocytic cells and debris and alveolar destruction were ex-tensively observed. Uninfected control mice showed no archi-tectural or inflammatory changes in the lungs.

Pulmonary infarct scores at each time point during infectionwere greater for the wild-type and �crzA � crzA comple-mented strains than for the �crzA strain. The mean pulmonaryinfarct score at 7 days after infection for the �crzA strain was3, compared to mean scores of 5.125 for the wild-type strainand 4.75 for the �crzA � crzA complemented strain (P � 0.05).

FIG. 8. Enhanced effects of �-1,3-glucan and chitin synthesis inhi-bition on the wild-type, �cnaA, and �crzA strains. Strains were grownon glucose minimal medium for 96 h with the indicated treatments.The effects of the antifungal agents are enhanced with genetic replace-ment of calcineurin or crzA and are most evident with combinationtreatment with �-1,3-glucan (caspofungin) and chitin synthesis(nikkomycin Z) inhibitors. Importantly, treatment of the �crzA strainwith the calcineurin inhibitor FK506 resulted in a phenotype virtuallyidentical to that of the �cnaA strain, indicating that CrzA has little orno function in a cell lacking calcineurin activity.

TABLE 1. Transcriptional profiles of two calcineurin pathway mutants

Gene Predicted function A. fumigatusgenome locus

Fold expression changea in:

�cnaAmutant

�crzAmutant

agsA �-1,3-Glucan synthase Afu3g00910 �1.5 1.1agsB �-1,3-Glucan synthase Afu2g11270 �1.3 �1.5agsC �-1,3-Glucan synthase Afu1g15440 �1.7 �1.1cmdA Calmodulin Afu4g10050 52.7 2.7cbpA Calcineurin binding protein Afu2g13060 �1.4 1.1chsA Chitin synthase A Afu2g01870 �1.1 1.3chsB Chitin synthase B Afu4g04180 1.2 1.1chsC Chitin synthase C Afu5g00760 �1.2 1.6chsD Chitin synthase D Afu1g12600 1.7 1.4chsE Chitin synthase E Afu2g13440 �2.0 1.3chsF Chitin synthase F Afu8g05630 �1.7 1.2chsG Chitin synthase G Afu3g14420 �1.0 2.6cnaA Calcineurin catalytic subunit Afu5g09360 NDb 4.2cnaB Calcineurin regulatory subunit Afu6g04540 �1.0 2.2crzA C2H2 transcription factor Afu1g06900 �3.6 NDfksA �-1,3-Glucan synthase catalytic subunit Afu6g12400 �2.1 �1.3gelA 1,3-�-Glucanosyltransferase Afu2g01170 3.4 �1.3gelB 1,3-�-Glucanosyltransferase Afu6g11390 �6.0 �1.6gelC 1,3-�-Glucanosyltransferase Afu2g12850 �3.5 �1.2rhoA Rho GTPase Afu6g06900 10.9 1.0rhoB Rho GTPase Afu3g10340 2.9 1.7rhoC Rho GTPase Afu3g06300 3.2 �1.5rhoD Rho GTPase Afu2g05740 3.7 �2.5pmrA Secretory Ca2�-ATPase pump Afu2g05860 �4.4 �1.9pmcA Vacuolar Ca2�-ATPase pump Afu7g01030 �7.9 �1.9

a Compared to expression in wild-type strain AF293.b ND, not detected.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

In the immunosuppressed, uninfected control mice there wereno findings of necrosis, hemorrhage, or edema at any timepoint. These results from our animal model strongly suggestthat CrzA is required for in vivo fungal growth and diseasedevelopment.

DISCUSSION

Calcium is one of the most important secondary messengersin microbial responses to the environment. Ca2� functions asan essential cofactor for many proteins, and thus microorgan-isms have evolved several mechanisms to regulate calcium lev-els in cells and thereby control the transduction of importantintra- and intercellular Ca2�-mediated signals. One highly con-served mechanism of eukaryotic calcium-mediated signaltransduction involves the Ca2�-calmodulin-dependent serine/threonine phosphatase calcineurin (24, 27, 41).

The role of calcineurin-mediated signaling in fungal physi-ology has been extensively explored in S. cerevisiae and a hostof other important fungi, including the medically importantopportunistic human fungal pathogens C. albicans, Cryptococ-cus neoformans, and, recently, A. fumigatus (55). In S. cerevi-siae, calcineurin mediates resistance to various cation stresses,including Na�, Mn2�, and Li�, and regulates Ca2� homeosta-

sis, Ca2�-mediated G2 arrest, and the onset of mitosis (12, 13,16, 31, 63–65). In the human pathogenic yeast C. neoformans,calcineurin mediates pH and CO2 homeostasis, temperature-sensitive growth, antifungal drug tolerance, mating/haploidfruiting, resistance to Li�, and fungal pathogenicity (10, 15, 35,60). Fungal pathogenicity and growth in human serum are alsomediated in part by calcineurin in C. albicans, along withresistance to both cation stress and antifungal drugs (1, 2, 4,11, 42).

While these studies have clearly illustrated the importanceof calcineurin in the physiology of fungi, they also demonstrateunique attributes of calcineurin-mediated signal transductionin different fungal species. Recently, we began an explorationof calcineurin-mediated signal transduction in the opportunis-tic human mold, A. fumigatus (53, 54). Previous studies with A.nidulans have suggested that calcineurin is essential and re-quired for cell cycle progression through G1/S, nuclear divi-sion, and polarized hyphal growth in this species (33, 40). Ourresults and the results of a complementary study, however,suggested that while calcineurin was not essential for survivalin A. fumigatus, calcineurin mutants displayed severe defects inconidial germination, polarized hyphal growth, and conidiumdevelopment. Importantly, A. fumigatus calcineurin mutantswere incapable of causing disease in several experimental mu-rine models of IPA (14, 53).

In this study, we sought to identify the homolog of S. cerevi-siae Crz1 in A. fumigatus and examined the hypothesis thatCrzA was a downstream effector of calcineurin signaling in A.fumigatus as in other fungi. Bioinformatic analyses revealed thepresence of a single putative homolog of Crz1 in the A. fu-migatus genome sequence, which we named CrzA. Replace-ment of the crzA gene in A. fumigatus resulted in a strain withphenotypes similar to those of the calcineurin A mutant butdiffering in their severity. For example, while the calcineurin Amutant was virtually incapable of polarized hyphal growth, the�crzA strain was capable of limited polarized hyphal growth,with a greatly decreased growth rate after 3 days. Furthermore,similar to the �cnaA strain, the �crzA strain possessed signif-icant defects in conidium germination and conidium morphol-ogy, with an apparent absence of the bumpy and echinulatesurface typically found on A. fumigatus conidia. In addition,the �crzA strain did not display as significant a defect in�-1,3-glucan content as the �cnaA strain, indicating that inA. fumigatus, �-1,3-glucan biosynthesis may largely be CrzAindependent.

Treatment with nikkomycin Z led to a compensatory in-crease in �-1,3-glucan content in the �cnaA strain, indicatingthat �-1,3-glucan synthesis can occur in the absence of cal-cineurin-mediated signaling. While we did observe an increasein mRNA abundance of the �-1,3-glucosyltransferase in the�cnaA strain in liquid culture, this did not prevent the loss of�-1,3-glucan. However, it may be possible that treatment witha cell wall stress agent induces higher levels of gelA mRNA thataccount for the compensatory increase in �-1,3-glucan content.Alternatively, nikkomycin Z may induce calcineurin-indepen-dent pathways for �-1,3-glucan biosynthesis. A similar com-pensatory effect on �-1,3-glucan content, however, was notseen in the �crzA strain treated with nikkomycin Z, so it ispossible that this pathway is activated primarily in the absenceof calcineurin in the cell and is dependent on a functional

FIG. 9. CrzA is required for A. fumigatus pathogenesis in a persis-tently neutropenic inhalational murine model of IPA. (A) Kaplan-Meier survival curve. Mice inoculated with the wild-type (AF293) andcrzA complemented strains displayed significant mortality by 14 daysfollowing infection, while the mice infected with the �crzA strain allsurvived (P 0.0011). Each experimental arm consisted of 20 mice.(B) Lung histopathology. Top row, Gomori’s methenamine silverstaining demonstrated extensive hyphal proliferation in the lung tissueof mice infected with the wild-type or crzA complemented strain, whilemice infected with the �crzA strain displayed a limited amount ofhyphal proliferation. Magnification, �400. Bottom row, hematoxylinand eosin staining shows significant necrosis and inflammation in miceinfected with the wild-type and crzA complemented strains comparedwith mice infected with the �crzA strain. Magnification, �400.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

CrzA. Thus, it appears that calcineurin A is involved in regu-lating �-1,3-glucan biosynthesis in A. fumigatus, as previouslypostulated (54), but that this regulation is not solely dependentupon CrzA.

One area of clinical and biochemical exploration is the con-cept of the “paradoxical effect” seen in fungi following treat-ment with a �-1,3-glucan synthesis inhibitor (62), where in C.albicans there is a reported compensatory increase in chitincontent, therefore linking the two cell wall components in astress response fashion (58). Our study is the first to demon-strate a compensatory increase in �-1,3-glucan following chitinsynthesis inhibition in A. fumigatus. It appears that this glucan-chitin interaction is controlled or affected by calcineurin sig-naling. If we eliminate calcineurin, there does appear to be anincrease in �-1,3-glucan biosynthesis during chitin biosynthesisinhibition. This apparent cell wall stress response seems idealfor combination therapy in which calcineurin, �-1,3-glucan,and chitin targets can all be blocked by drugs and certainlysupports a multiple-pronged attack on cell wall targets andhyphal growth in A. fumigatus.

While the molecular mechanisms of calcineurin signaling inA. fumigatus are just beginning to be explored, recent experi-ments with other fungi have started to further elucidate thecomplex calcineurin signaling pathway and may provide cluesto the interplay between calcineurin and other signaling path-ways in A. fumigatus. A calcineurin-independent Crz1 ho-molog, Crz2, was identified in C. albicans and observed to workin concert with the Rim101/PacC pathway to mediate adapta-tion to alkaline pH environments (25). A putative homolog forCrz2 in A. fumigatus has not been identified, and it may be thatCrzA in A. fumigatus performs the roles of both Crz1 and Crz2.In the filamentous mold Botrytis cinerea, a subset of genes werefound to be coregulated by the G alpha subunit gene (bcg1),the phospholipase C gene (bcplc1), and the calcineurin path-way, suggesting an interconnection of these important signal-ing pathways in filamentous molds (45). Future studies willuncover the mechanism of calcineurin-dependent regulation ofCrzA and the likely identity of other signaling pathways thatmay intersect with calcineurin-mediated signal transductionto control hyphal growth and cell wall homeostasis in A.fumigatus.

Similar to the �cnaA strain, the �crzA strain was incapableof establishing extensive in vivo growth and disease in a per-sistently neutropenic murine model of IPA. Several explana-tions may account for the �crzA strain’s inability to causedisease. First, as in vitro germination assays demonstrated,fewer than 50% of the �crzA strain conidia were able to ger-minate at 37°C in a relatively nutrient-rich medium by 48 h.Thus, during infection, when initial nutrient availability is likelylow, the significant delay in germination may allow those in-nate immune cells that are functional in neutropenic animalsto clear the majority of the �crzA strain conidia.

On the other hand, histopathological analyses did demon-strate the presence of sparse fungal hyphae in mice infectedwith the �crzA strain, in contrast to our previous study of�cnaA strain-infected mice, which contained no histologicalevidence of hyphal growth. Given the low growth rate observedin vitro with the �crzA strain and the abnormal hyphal mor-phology observed in electron micrographs, it is possible thatthe �crzA strain hyphae are not capable of substantial invasive

growth in vivo and production of disease. However, it may alsobe possible that the low growth rate allows those immuneeffector cells that remain in the mice to clear the infection andprevent disease development. Furthermore, the �crzA strainmay be more susceptible to killing by immune effector cells dueto its altered cell wall morphology.

Given the close similarity in phenotypes observed in the�cnaA and the �crzA strains, two conclusions about the role ofCrzA in calcineurin signaling in A. fumigatus are possible fromour studies. First, like in other fungi, CrzA is an importantdownstream effector of calcineurin-mediated signal transduc-tion in A. fumigatus for morphology. The significant defectsobserved in hyphal growth and asexual development in the�crzA strain strongly suggest the importance of CrzA-medi-ated signaling in fungal morphology. Importantly, it appearsthat CrzA is largely dependent on calcineurin for its activity inA. fumigatus. Treatment of the �crzA strain with the cal-cineurin inhibitor FK506 resulted in a phenotype virtuallyidentical to that of the �cnaA strain, with almost completecessation of hyphal growth (Fig. 8). This result suggests thatCrzA has little or no function in a cell lacking calcineurinactivity and suggests that CrzA is an important downstreameffector of calcineurin in A. fumigatus. However, the moresevere conidium germination defect observed in the �crzAstrain may suggest that additional factors regulate CrzA in A.fumigatus. Interestingly, no obvious calcineurin docking site,similar to the one found in Crz1, was found in CrzA. However,Crz1 and CrzA do share a serine-rich region known to beinvolved in calcineurin-dependent regulation of Crz1 localiza-tion, activity, and phosphorylation in S. cerevisiae (Fig. 2) (51).Future studies will seek to discover the mechanism of cal-cineurin-dependent regulation of CrzA in A. fumigatus.

Second, our data suggest that additional CrzA-independentdownstream effectors of calcineurin signaling exist in A. fu-migatus. Given the intermediate phenotypes in growth andmorphology observed with the �crzA and �cnaA strains, wehypothesize that additional calcineurin-dependent down-stream effectors exist that partially maintain, albeit in a limitedfashion, hyphal growth in the �crzA strain. Data to support thishypothesis are observed in the �-1,3-glucan contents of the�cnaA and �crzA strains and in the mRNA abundance assays,which demonstrated little or no change in the abundance ofmRNAs from genes involved in cell wall biosynthesis in the�crzA strain (Fig. 7 and Table 1). Importantly, our results areconcordant with those reported for �crz1/�crz1 mutants of C.albicans, which also display a growth and morphology pheno-type intermediate to that of the �cna1/�cna1 mutants, suggest-ing the presence of additional downstream calcineurin effec-tors (23, 37, 43). In addition, our �crzA strain displays aphenotype similar to that of a crz1 mutant, �bccrz1, of theplant pathogenic mold Botrytis cinerea (44). The B. cinereamutant displayed significant in vitro growth defects and hadimpaired hyphal morphology, conidiation, and sclerotium for-mation. Unlike the A. fumigatus �crzA strain, the �bccrz1strain displayed significant defects in cell wall and membraneintegrity. Thus, like with calcineurin mutants, fungi lacking aCrz1 homolog display similar phenotypes with subtle differ-ences among genera.

During the writing of this paper, a complementary studyreporting the creation of a �crzA strain from A. fumigatus



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

strain CEA17 was published (49). The phenotype of the �crzAstrain in the CEA17 background is virtually identical to theresults we report here with the �crzA strain in the AF293background. Those authors observed localization of CrzA tothe nucleus in response to calcium chloride treatment. Ouranalysis of the CrzA amino acid sequence compared to theamino acid sequence of Crz1 in S. cerevisiae revealed a putativeNLS. However, this site was downstream of the Crz1 NLS sitein CrzA (amino acids 460 to 490) (Fig. 2). This region containsthe basic amino acids arginine and lysine, which are typicallyfound in NLS regions, and is serine rich, similar to Crz1’s NLSregion (Fig. 2). Further experiments utilizing site-directed mu-tagenesis will confirm whether this site is indeed the NLS sitein A. fumigatus.

In addition to the nuclear localization study, Soriani et al.(49) observed that the �crzA strain in the CEA17 backgroundwas more sensitive to calcium and manganese chloride. Whilewe did not directly measure sensitivity to these chemicals, weobserved that germination of �crzA strain conidia in theAF293 background was severely inhibited in liquid culture inthe presence of 200 mM CaCl2 (data not shown). We alsoobserved a significant decrease in conidium germination in the�crzA strain in the AF293 background, while Soriani et al. (49)reported an increase in polar germ tube emergence in the�crzA strain in the CEA17 background. The explanation forthis significant difference is not readily apparent; however, thedifferences in methodology may be an important factor. Whilewe measured germ tube emergence over a defined time periodof 48 h, Soriani et al. (49) treated germlings with hydroxyureaand, following release from cell cycle arrest over 150 min,measured germling polarization. Interestingly, Soriani et al.were unable to identify germinating conidia or hyphae in theirhistopathological analyses of mice infected with the �crzAstrain. However, unlike our animal model results, Soriani et al.did observe 30% mortality in mice infected with the �crzAstrain. Viable �crzA colonies were recovered from mice in-fected with the �crzA strain that perished.

Soriani et al. (49) also examined the mRNA abundances ofvarious ion pump homologs in response to a short pulse of 200mM CaCl2, while we focused our mRNA abundance assays oncomponents of the cell wall biosynthesis machinery (Table 1).However, we did examine mRNA abundances of pmrA andpmcA in the �cnaA and �crzA strains and found results similarto those of Soriani et al., i.e., that both pmrA and pmcA appearto be regulated by the calcineurin pathway in A. fumigatus as inS. cerevisiae. These results likely explain the calcium sensitivityof the �crzA strain reported by Soriani et al. (49) and observedby us in our experiments.

These results continue to confirm the great potential thatinhibition of the calcineurin pathway in pathogenic fungi mayhave for the treatment of invasive mycoses. There is a highamino acid identity between the A. fumigatus CnaA and humancalcineurin A proteins (58.4% and 63.1%, respectively) (17).However, the closest possible homolog to A. fumigatus CrzA ishuman NFAT, and the amino acid identity between A. fumiga-tus CrzA and human NFAT is only 14.9% (34). This informa-tion suggests that inhibition of A. fumigatus CrzA, which blocksA. fumigatus hyphal growth in vitro and in vivo, would have farless cross-reactivity with potential similar human targets and

could more selectively damage A. fumigatus with targetedantifungal therapy.

In conclusion, our results strongly suggest that the A. fumiga-tus Crz1 homolog, CrzA, is a downstream effector of cal-cineurin signaling in this important opportunistic human fun-gal pathogen. A. fumigatus CrzA has a role in polarized hyphalgrowth, asexual development, and fungal pathogenicity muchlike that of CnaA. Future studies will seek to discover themechanism of CrzA-mediated signal transduction and to un-cover additional pathways that intersect with the calcineurinsignaling pathway that may account for some of the differencesobserved between the �crzA and �cnaA strains. Finally, giventhat CrzA seems to be fungus specific and strongly linked todisease development, this particular component of the cal-cineurin pathway is an ideal antifungal drug target for invasiveaspergillosis and possibly other molds.

ACKNOWLEDGMENTS

We thank Leslie Eibest for assistance with the scanning electronmicroscopy work.

R.A.C. was supported by the NIH/NIAID Molecular Mycology andPathogenesis Training Program (5 T32 AI052080) at Duke UniversityMedical Center and is currently supported by NIH/COBRE grantRR020185 and the Montana State Agricultural Experiment Station.W.J.S. is supported by NIH/NIAID award K08 A1061149, a BasicScience Faculty Development grant from the American Society forTransplantation, and a Children’s Miracle Network grant.

REFERENCES

1. Bader, T., B. Bodendorfer, K. Schroppel, and J. Morschhauser. 2003. Cal-cineurin is essential for virulence in Candida albicans. Infect. Immun. 71:5344–5354.

2. Bader, T., K. Schroppel, S. Bentink, N. Agabian, G. Kohler, and J. Mor-schhauser. 2006. Role of calcineurin in stress resistance, morphogenesis, andvirulence of a Candida albicans wild-type strain. Infect. Immun. 74:4366–4369.

3. Bell, N. P., C. L. Karp, E. C. Alfonso, J. Schiffman, and D. Miller. 1999.Effects of methylprednisolone and cyclosporine A on fungal growth in vitro.Cornea 18:306–313.

4. Blankenship, J. R., F. L. Wormley, M. K. Boyce, W. A. Schell, S. G. Filler,J. R. Perfect, and J. Heitman. 2003. Calcineurin is essential for Candidaalbicans survival in serum and virulence. Eukaryot. Cell 2:422–430.

5. Bok, J. W., and N. P. Keller. 2004. laeA, a regulator of secondary metabolismin Aspergillus spp. Eukaryot. Cell 3:527–535.

6. Boustany, L. M., and M. S. Cyert. 2002. Calcineurin-dependent regulation ofCrz1p nuclear export requires Msn5p and a conserved calcineurin dockingsite. Genes Dev. 16:608–619.

7. Bowman, J. C., S. Hicks, M. B. Kurtz, H. Rosen, D. M. Schmatz, P. A.Liberator, and C. M. Douglas. 2002. The antifungal echinocandin caspofun-gin acetate kills growing cells of Aspergillus fumigatus in vitro. Antimi-crob.Agents Chemother. 46:3001–3012.

8. Clipstone, N. A., and G. R. Crabtree. 1992. Identification of calcineurin as akey signalling enzyme in T-lymphocyte activation. Nature 357:695–697.

9. Cramer, R. A. J., M. P. Gamcsik, R. M. Brooking, L. K. Najvar, W. R.Kirkpatrick, T. F. Patterson, C. J. Balibar, J. R. Graybill, J. R. Perfect, S. N.Abraham, and W. J. Steinbach. 2006. Disruption of a nonribosomal peptidesynthetase in Aspergillus fumigatus eliminates gliotoxin production. Eukaryot.Cell 5:972–980.

10. Cruz, M. C., D. S. Fox, and J. Heitman. 2001. Calcineurin is required forhyphal elongation during mating and haploid fruiting in Cryptococcus neo-formans. EMBO J. 20:1020–1032.

11. Cruz, M. C., A. L. Goldstein, J. R. Blankenship, M. Del Poeta, D. Davis,M. E. Cardenas, J. R. Perfect, J. H. McCusker, and J. Heitman. 2002.Calcineurin is essential for survival during membrane stress in Candidaalbicans. EMBO J. 21:546–559.

12. Cyert, M. S. 2003. Calcineurin signaling in Saccharomyces cerevisiae: howyeast go crazy in response to stress. Biochem. Biophys. Res. Commun.311:1143–1150.

13. Cyert, M. S., R. Kunisawa, D. Kaim, and J. Thorner. 1991. Yeast hashomologs (CNA1 and CNA2 gene products) of mammalian calcineurin, acalmodulin-regulated phosphoprotein phosphatase. Proc. Natl. Acad. Sci.USA 88:7376–7380.

14. da Silva Ferreira, M. E., T. Heinekamp, A. Hartl, A. A. Brakhage, C. P.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

Semighini, S. D. Harris, M. Savoldi, P. F. de Gouvea, M. H. de SouzaGoldman, and G. H. Goldman. 2007. Functional characterization of theAspergillus fumigatus calcineurin. Fungal Genet. Biol. 44:219–230.

15. Fox, D. S., M. C. Cruz, R. A. Sia, H. Ke, G. M. Cox, M. E. Cardenas, andJ. Heitman. 2001. Calcineurin regulatory subunit is essential for virulenceand mediates interactions with FKBP12-FK506 in Cryptococcus neoformans.Mol. Microbiol. 39:835–849.

16. Garrett-Engele, P., B. Moilanen, and M. S. Cyert. 1995. Calcineurin, theCa2�/calmodulin-dependent protein phosphatase, is essential in yeast mu-tants with cell integrity defects and in mutants that lack a functional vacuolarH(�)-ATPase. Mol. Cell. Biol. 15:4103–4114.

17. Guerini, D., and C. B. Klee. 1989. Cloning of human calcineurin A: evidencefor two isozymes and identification of a polyproline structural domain. Proc.Natl. Acad. Sci. USA 86:9183–9187.

18. Heath, V. L., S. L. Shaw, S. Roy, and M. S. Cyert. 2004. Hph1p and Hph2p,novel components of calcineurin-mediated stress responses in Saccharomy-ces cerevisiae. Eukaryot. Cell 3:695–704.

19. Herbrecht, R., D. W. Denning, T. F. Patterson, J. E. Bennett, R. E. Greene,J. W. Oestmann, W. V. Kern, K. A. Marr, P. Ribaud, O. Lortholary, R.Sylvester, R. H. Rubin, J. R. Wingard, P. Stark, C. Durand, D. Caillot, E.Thiel, P. H. Chandrasekar, M. R. Hodges, H. T. Schlamm, P. F. Troke, andB. de Pauw. 2002. Voriconazole versus amphotericin B for primary therapyof invasive aspergillosis. N. Engl. J. Med. 347:408–415.

20. High, K. P. 1994. The antimicrobial activities of cyclosporine, FK506, andrapamycin. Transplantation 57:1689–1700.

21. Hirayama, S., R. Sugiura, Y. Lu, T. Maeda, K. Kawagishi, M. Yokoyama, H.Tohda, Y. Giga-Hama, H. Shuntoh, and T. Kuno. 2003. Zinc finger proteinPrz1 regulates Ca2� but not Cl� homeostasis in fission yeast. Identificationof distinct branches of calcineurin signaling pathway in fission yeast. J. Biol.Chem. 278:18078–18084.

22. Kahn, J. N., M. J. Hsu, F. Racine, R. Giacobbe, and M. Motyl. 2006.Caspofungin susceptibility in Aspergillus and non-Aspergillus molds: inhibi-tion of glucan synthase and reduction of beta-D-1,3 glucan levels in culture.Antimicrob Agents Chemother. 50:2214–2216.

23. Karababa, M., E. Valentino, G. Pardini, A. T. Coste, J. Bille, and D. Sang-lard. 2006. CRZ1, a target of the calcineurin pathway in Candida albicans.Mol. Microbiol. 59:1429–1451.

24. Klee, C. B., and M. H. Krinks. 1978. Purification of cyclic 3�,5�-nucleotidephosphodiesterase inhibitory protein by affinity chromatography on activatorprotein coupled to Sepharose. Biochemistry 17:120–126.

25. Kullas, A. L., S. J. Martin, and D. Davis. 2007. Adaptation to environmentalpH: integrating the Rim101 and calcineurin signal transduction pathways.Mol. Microbiol. 66:858–871.

26. Latge, J. P. 1999. Aspergillus fumigatus and aspergillosis. Clin. Microbiol.Rev. 12:310–350.

27. Liu, J., J. D. Farmer, Jr., W. S. Lane, J. Friedman, I. Weissman, and S. L.Schreiber. 1991. Calcineurin is a common target of cyclophilin-cyclosporin Aand FKBP-FK506 complexes. Cell 66:807–815.

28. Maertens, J., I. Raad, G. Petrikkos, M. Boogaerts, D. Selleslag, F. B.Petersen, C. A. Sable, N. A. Kartsonis, A. Ngai, A. Taylor, T. F. Patterson,D. W. Denning, and T. J. Walsh. 2004. Efficacy and safety of caspofungin fortreatment of invasive aspergillosis in patients refractory to or intolerant ofconventional antifungal therapy. Clin. Infect. Dis. 39:1563–1571.

29. Marr, K. A., R. A. Carter, F. Crippa, A. Wald, and L. Corey. 2002. Epide-miology and outcome of mould infections in hematopoietic stem cell trans-plant recipients. Clin. Infect. Dis. 34:909–917.

30. Matheos, D. P., T. J. Kingsbury, U. S. Ahsan, and K. W. Cunningham. 1997.Tcn1p/Crz1p, a calcineurin-dependent transcription factor that differentiallyregulates gene expression in Saccharomyces cerevisiae. Genes Dev. 11:3445–3458.

31. Mizunuma, M., D. Hirata, K. Miyahara, E. Tsuchiya, and T. Miyakawa.1998. Role of calcineurin and Mpk1 in regulating the onset of mitosis inbudding yeast. Nature 392:303–306.

32. Mouyna, I., W. Morelle, M. Vai, M. Monod, B. Lechenne, T. Fontaine, A.Beauvais, J. Sarfati, M. C. Prevost, C. Henry, and J. P. Latge. 2005. Deletion ofGEL2 encoding for a �(1–3)glucanosyltransferase affects morphogenesis andvirulence in Aspergillus fumigatus. Mol. Microbiol. 56:1675–1688.

33. Nanthakumar, N. N., J. S. Dayton, and A. R. Means. 1996. Role of Ca��/calmodulin binding proteins in Aspergillus nidulans cell cycle regulation.Prog. Cell Cycle Res. 2:217–228.

34. Northrop, J. P., S. N. Ho, L. Chen, D. J. Thomas, L. A. Timmerman, G. P.Nolan, A. Admon, and G. R. Crabtree. 1994. NF-AT components define a familyof transcription factors targeted in T-cell activation. Nature 369:497–502.

35. Odom, A., S. Muir, E. Lim, D. L. Toffaletti, J. Perfect, and J. Heitman. 1997.Calcineurin is required for virulence of Cryptococcus neoformans. EMBO J.16:2576–2589.

36. O’Keefe, S. J., J. Tamura, R. L. Kincaid, M. J. Tocci, and E. A. O’Neill. 1992.FK-506- and CsA-sensitive activation of the interleukin-2 promoter by cal-cineurin. Nature 357:692–694.

37. Onyewu, C., F. L. Wormley, Jr., J. R. Perfect, and J. Heitman. 2004. Thecalcineurin target, Crz1, functions in azole tolerance but is not required forvirulence of Candida albicans. Infect. Immun. 72:7330–7333.

38. Osherov, N., D. P. Kontoyiannis, A. Romans, and G. S. May. 2001. Resis-tance to itraconazole in Aspergillus nidulans and Aspergillus fumigatus isconferred by extra copies of the A. nidulans P-450 14�-demethylase gene,pdmA. J. Antimicrob. Chemother. 48:75–81.

39. Polizotto, R. S., and M. S. Cyert. 2001. Calcineurin-dependent nuclear import ofthe transcription factor Crz1p requires Nmd5p. J. Cell Biol. 154:951–960.

40. Rasmussen, C., C. Garen, S. Brining, R. L. Kincaid, R. L. Means, and A. R.Means. 1994. The calmodulin-dependent protein phosphatase catalytic sub-unit (calcineurin A) is an essential gene in Aspergillus nidulans. EMBO J.13:2545–2552.

41. Rusnak, F., and P. Mertz. 2000. Calcineurin: form and function. Physiol.Rev. 80:1483–1521.

42. Sanglard, D., F. Ischer, O. Marchetti, J. Entenza, and J. Bille. 2003. Cal-cineurin A of Candida albicans: involvement in antifungal tolerance, cellmorphogenesis and virulence. Mol. Microbiol. 48:959–976.

43. Santos, M., and I. F. de Larrinoa. 2005. Functional characterization of theCandida albicans CRZ1 gene encoding a calcineurin-regulated transcriptionfactor. Curr. Genet. 48:88–100.

44. Schumacher, J., I. F. de Larrinoa, and B. Tudzynski. 2008. Calcineurin-responsive zinc finger transcription factor CRZ1 of Botrytis cinerea is re-quired for growth, development, and full virulence on bean plants. Eukaryot.Cell 7:584–601.

45. Schumacher, J., M. Viaud, A. Simon, and B. Tudzynski. 2008. The Galphasubunit BCG1, the phospholipase C (BcPLC1) and the calcineurin phos-phatase co-ordinately regulate gene expression in the grey mould fungusBotrytis cinerea. Mol. Microbiol. 67:1027–1050.

46. Shedletzky, E., C. Unger, and D. P. Delmer. 1997. A microtiter-based fluo-rescence assay for (1,3)-beta-glucan synthases. Anal. Biochem. 249:88–93.

47. Shimizu, K., and N. P. Keller. 2001. Genetic involvement of a cAMP-dependent protein kinase in a G protein signaling pathway regulating mor-phological and chemical transitions in Aspergillus nidulans. Genetics 157:591–600.

48. Singh, N., and D. L. Paterson. 2005. Aspergillus infections in transplantrecipients. Clin. Microbiol. Rev. 18:44–69.

49. Soriani, F. M., I. Malavazi, M. E. da Silva Ferreira, M. Savoldi, M. R. VonZeska Kress, M. H. de Souza Goldman, O. Loss, E. Bignell, and G. H.Goldman. 2008. Functional characterization of the Aspergillus fumigatusCRZ1 homologue, CrzA. Mol. Microbiol. 67:1274–1291.

50. Stathopoulos, A. M., and M. S. Cyert. 1997. Calcineurin acts through theCRZ1/TCN1-encoded transcription factor to regulate gene expression inyeast. Genes Dev. 11:3432–3444.

51. Stathopoulos-Gerontides, A., J. J. Guo, and M. S. Cyert. 1999. Yeast cal-cineurin regulates nuclear localization of the Crz1p transcription factorthrough dephosphorylation. Genes Dev. 13:798–803.

52. Steinbach, W. J., D. K. J. Benjamin, S. A. Trasti, J. L. Miller, W. A. Schell,A. K. Zaas, W. M. Foster, and J. R. Perfect. 2004. Value of an inhalationalmodel of invasive aspergillosis. Med. Mycol. 42:417–425.

53. Steinbach, W. J., R. A. Cramer, Jr., B. Z. Perfect, Y. G. Asfaw, T. C. Sauer,L. K. Najvar, W. R. Kirkpatrick, T. F. Patterson, D. K. Benjamin, Jr.,J. Heitman, and J. R. Perfect. 2006. Calcineurin controls growth, morphol-ogy, and pathogenicity in Aspergillus fumigatus. Eukaryot. Cell 5:1091–1103.

54. Steinbach, W. J., R. A. Cramer, Jr., B. Z. Perfect, C. Henn, K. Nielsen,J. Heitman, and J. R. Perfect. 2007. Calcineurin inhibition or mutationenhances cell wall inhibitors against Aspergillus fumigatus. Antimicrob.Agents Chemother. 51:2979–2981.

55. Steinbach, W. J., J. L. Reedy, R. A. Cramer, Jr., J. R. Perfect, and J. Heit-man. 2007. Harnessing calcineurin as a novel anti-infective agent againstinvasive fungal infections. Nat. Rev. Microbiol. 5:418–430.

56. Steinbach, W. J., W. A. Schell, J. R. Blankenship, C. Onyewu, J. Heitman,and J. R. Perfect. 2004. In vitro interactions between antifungals and immu-nosuppressants against Aspergillus fumigatus. Antimicrob. Agents Che-mother. 48:1664–1669.

57. Steinbach, W. J., N. Singh, J. L. Miller, D. K. Benjamin, Jr., W. A. Schell,J. Heitman, and J. R. Perfect. 2004. In vitro interactions between antifungalsand immunosuppressants against Aspergillus fumigatus isolates from trans-plant and nontransplant patients. Antimicrob Agents Chemother. 48:4922–4925.

58. Stevens, D. A., M. Ichinomiya, Y. Koshi, and H. Horiuchi. 2006. Escape ofCandida from caspofungin inhibition at concentrations above the MIC (par-adoxical effect) accomplished by increased cell wall chitin; evidence forbeta-1,6-glucan synthesis inhibition by caspofungin. Antimicrob. AgentsChemother. 50:3160–3161.

59. Sweigard, J., F. Chumley, A. Carroll, L. Farrall, and B. Valent. 1997. A seriesof vectors for fungal transformation. Fungal Genet. Newsl. 44:52–53.

60. Wang, P., M. E. Cardenas, G. M. Cox, J. R. Perfect, and J. Heitman. 2001.Two cyclophilin A homologs with shared and distinct functions important forgrowth and virulence of Cryptococcus neoformans. EMBO Rep. 2:511–518.

61. White, D. A. 2005. Aspergillus pulmonary infections in transplant recipients.Clin. Chest Med. 26:661–674, vii.

62. Wiederhold, N. P., D. P. Kontoyiannis, J. Chi, R. A. Prince, V. H. Tam, andR. E. Lewis. 2004. Pharmacodynamics of caspofungin in a murine model of



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

invasive pulmonary aspergillosis: evidence of concentration-dependent ac-tivity. J. Infect. Dis. 190:1464–1471.

63. Withee, J. L., J. Mulholland, R. Jeng, and M. S. Cyert. 1997. An essentialrole of the yeast pheromone-induced Ca2� signal is to activate calcineurin.Mol. Biol. Cell 8:263–277.

64. Withee, J. L., R. Sen, and M. S. Cyert. 1998. Ion tolerance of Saccharomycescerevisiae lacking the Ca2�/CaM-dependent phosphatase (calcineurin) isimproved by mutations in URE2 or PMA1. Genetics 149:865–878.

65. Ye, R. R., and A. Bretscher. 1992. Identification and molecular character-ization of the calmodulin-binding subunit gene (CMP1) of protein phos-phatase 2B from Saccharomyces cerevisiae. An alpha-factor inducible gene.Eur. J. Biochem. 204:713–723.

66. Yoshimoto, H., K. Saltsman, A. P. Gasch, H. X. Li, N. Ogawa, D. Botstein,P. O. Brown, and M. S. Cyert. 2002. Genome-wide analysis of gene expres-sion regulated by the calcineurin/Crz1p signaling pathway in Saccharomycescerevisiae. J. Biol. Chem. 277:31079–31088.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

calcineurin target crza regulates conidial germination ...calcineurin effectors have not been...

Documents