INFECTION AND IMMUNITY, Jan. 2006, p. 504–515 Vol. 74, No. 10019-9567/06/$08.00�0 doi:10.1128/IAI.74.1.504–515.2006Copyright © 2006, American Society for Microbiology. All Rights Reserved.

Urease Produced by Coccidioides posadasii Contributes to theVirulence of This Respiratory Pathogen

Fariba Mirbod-Donovan, Ruth Schaller, Chiung-Yu Hung, Jianmin Xue,Utz Reichard, and Garry T. Cole*

Department of Medical Microbiology and Immunology, Medical College of Ohio, Toledo, Ohio 43614

Received 19 May 2005/Returned for modification 17 June 2005/Accepted 12 October 2005

Urease activity during in vitro growth in the saprobic and parasitic phases of Coccidioides spp. is partlyresponsible for production of intracellular ammonia released into the culture media and contributes toalkalinity of the external microenvironment. Although the amino acid sequence of the urease of Coccidioidesposadasii lacks a predicted signal peptide, the protein is transported from the cytosol into vesicles and thecentral vacuole of parasitic cells (spherules). Enzymatically active urease is released from the contents ofmature spherules during the parasitic cycle endosporulation stage. The endospores, together with the ureaseand additional material which escape from the ruptured parasitic cells, elicit an intense host inflammatoryresponse. Ammonia production by the spherules of C. posadasii is markedly increased by the availability ofexogenous urea found in relatively high concentrations at sites of coccidioidal infection in the lungs of mice.Direct measurement of the pH at these infection sites revealed an alkaline microenvironment. Disruption of theurease gene of C. posadasii resulted in a marked reduction in the amount of ammonia secreted in vitro by thefungal cells. BALB/c mice challenged intranasally with the mutant strain showed increased survival, a well-organized granulomatous response to infection, and better clearance of the pathogen than animals challengedwith either the parental or the reconstituted (revertant) strain. We conclude that ammonia and enzymaticallyactive urease released from spherules during the parasitic cycle of C. posadasii contribute to host tissuedamage, which exacerbates the severity of coccidioidal infection and enhances the virulence of this humanrespiratory pathogen.

Coccidioides spp. are airborne molds and respiratory patho-gens of humans which grow as a filamentous saprobe in desertand semiarid soils. They are the causative agents of coccidioid-omycosis, which is found in areas of endemicity such as thesouthwestern United States, northern Mexico, and Central andSouth America, including Guatemala, Honduras, northern Co-lumbia, Venezuela, Paraguay, and Argentina (29). Two specieshave been identified: Coccidioides immitis, which appears to bebiogeographically restricted to the San Joaquin Valley of Cal-ifornia, and C. posadasii, which is widespread throughout theregions of endemicity of the Americas (16). The saprobic formof Coccidioides survives under harsh environmental conditions,including meager rainfall, hot summers, and sandy, alkalinesoil with high salinity (33). The mycelial phase has been re-ported to tolerate a broad pH range of pH 2 to 12 (4). Inresponse to incubation in an acidic, sugar-free medium, thesaprobe is induced to release ammonium ions and ammonia inquantities which increase the pH of its extracellular environ-ment (8). Parasite-phase cultures similarly respond to acidifi-cation of their external environment by secretion of NH4

� andNH3. A source of intracellular ammonia is urea hydrolysis, andCoccidioides has been reported to produce an enzymaticallyactive urease (25, 39). This cytosolic enzyme catalyzes thehydrolysis of urea to yield ammonia and carbamate. The latter

spontaneously hydrolyzes to form carbonic acid and a secondmolecule of ammonia. At physiological pH, the carbonic acidproton dissociates and ammonia molecules equilibrate withwater to become protonated, with a resultant increase in pH(26).

Urease has been shown to be an important virulence factorin several bacterial pathogens (5). In Helicobacter pylori, thecausative agent of chronic gastritis and duodenal ulcers, ureaseactivity is induced both by acidic conditions of the host micro-environment and the availability of gastric urea (26). Thisresults in production of high amounts of intracellular ammo-nia, some of which is released and neutralizes the acid in theimmediate vicinity of the pathogen. High concentrations ofbacterially derived ammonia at sites of infection have beensuggested to injure the gastric mucosa as a result of tissueexposure to ammonium hydroxide. Evidence has also beenpresented that urease stimulates the host gastric epithelial cellsto produce proinflammatory cytokines (36). The persistent in-flammatory response to H. pylori infection ultimately damagesmucosal tissue and exacerbates the ulcerated condition. Re-cent studies have suggested that engulfment of urease-produc-ing bacteria by macrophages in the presence of exogenous urearesults in intraphagocytic release of ammonia, which has aninhibitory effect on macrophage surface expression of majorhistocompatibility complex class II molecules (34). This, inturn, could have a suppressive effect on cell-mediated immuneresponse to the microbial pathogen. Although urease activityhas been reported for several genera of medically importantfungi (e.g., Aspergillus, Candida, Cryptococcus, Rhodotorula,and Trichosporon spp.), this enzyme has been suggested to play

* Corresponding author. Present address: Department of Biology,Margaret Batts Tobin Building, Room 1.308E, University of Texas atSan Antonio, 6900 North Loop 1604 West, San Antonio, TX 78249.Phone: (210) 458-7017. Fax: (210) 458-5658. E-mail: garry.cole@utsa.edu.

504

on March 31, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

a role in pathogenesis only in Coccidioides and Cryptococcusspp. (8, 11, 27). In this report we show that disruption of the C.posadasii urease gene results in a major reduction in the abilityof the pathogen to kill intranasally challenged BALB/c mice.Survival appears to be related to the host’s ability to mount aneffective granulomatous response to the pathogen, which leadsto clearance of the fungal cells from infected lung tissue. Wepresent evidence that urease activity is an important virulencedeterminant of Coccidioides during interaction of the pathogenwith murine lung tissue.

MATERIALS AND METHODS

Reagents. Nutrients used in the fungal culture medium were obtained fromDifco Laboratories (Detroit, Mich.). All other chemicals and solutions werepurchased from Sigma Chemical Company (St. Louis, Mo.), unless otherwisespecified.

Fungal strain, media, and growth conditions. For production of arthroconidia,the saprobic (mycelial) phase of C. posadasii strain C735 was grown on GYE agar(1% glucose, 0.5% yeast extract, 1.5% agar) at 30°C for 3 to 4 weeks. For DNAextraction, parental and transformed strains of the pathogen were grown as thesaprobic phase in GYE liquid medium at 30°C. For parasite-phase growth, thesesame strains were cultured in defined liquid glucose-salts medium at 39°C in thepresence of 20% CO2 as previously described (17). Ammonium acetate (16.0mM) is the sole nitrogen source provided in the glucose-salts medium. Alterna-tively, arthroconidia were converted to spherules by growth in RPMI 1640medium supplemented with 10% heat-inactivated calf serum (catalog no. 30-2001 and 30-2020; American Type Culture Collection, Manassas, Va.) plusNa2HPO4, Tamol, and antibiotics as previously reported (30). The supplementedRPMI 1640 growth medium was used to test the effect of the addition ofexogenous urea on ammonia production by C. posadasii. For these studies, theculture medium was contained in 24-well, flat-bottom, polystyrene tissue cultureplates (Becton Dickinson and Co., Franklin Lakes, N.J.) (1.0 ml/well). Each wellwas inoculated with 106 arthroconidia. The parasitic cells were incubated with orwithout the addition of urea (10 mM) at 39°C in the presence of 20% CO2 for 7days.

Immunolocalization. In vitro-grown parasite-phase cells were cryofixed, sub-jected to freeze substitution, and embedded in L.R. White resin at �10°C inpreparation for immunoelectron microscopy as previously described (19). Anti-serum was raised in BALB/c mice against the purified recombinant ureaseprotein as previously reported (39). The immunoglobulin G (IgG) fraction wasisolated from immune sera by protein A affinity chromatography. Thin sectionsof cultured parasite-phase cells (24-h [first-generation] spherule initials, 72-hsegmentation-phase spherules, and 132-h endosporulation-phase spherules; Fig.1) were reacted with the specific antiurease antibody followed by goat anti-mouse

secondary antibody conjugated to colloidal gold (Ted Pella Inc., Redding, Calif.)(20-nm particles) as previously reported (19). Thin sections of resin-embeddedcells were examined by transmission electron microscopy. Incubation conditions,poststaining methods, and controls employed were the same as previously re-ported (19). Thick sections of these same cryofixed and resin-embedded spher-ules were reacted with antiurease IgG followed by secondary antibody conju-gated with fluorescein isothiocyanate (FITC) as previously described (17). Thestained sections were examined by fluorescence microscopy.

A secreted, spherule outer wall (SOW) fraction, which has previously beendescribed (19), together with intact endospores released from ruptured spher-ules, was isolated from 132-h parasite-phase cultures. The freshly collected SOWfraction and endospores were reacted with either antiurease serum or preim-mune murine serum at a 1:200 dilution in phosphate-buffered saline (PBS). Theintact cells and SOW preparation were then washed with PBS, incubated withgoat anti-mouse immunoglobulin conjugated with FITC, and examined by fluo-rescence microscopy as previously described (17).

Immunoblot analysis. Detection of the C. posadasii urease protein in cellhomogenates and culture filtrates of parasitic cells obtained from 24-, 72-, and132-h cultures was conducted using aliquots of equilibrated total protein of eachpreparation separated by sodium dodecyl sulfate-polyacrylamide gel electro-phoresis (SDS-PAGE). Protein extraction was conducted as previously described(18). Immunoblot analysis with urease-specific murine antiserum was performedas previously reported (19). The SOW fraction, which was isolated by differentialcentrifugation as previously described (17), was obtained from both 72-h and132-h parasite-phase cultures and subjected to SDS-PAGE and immunoblotanalysis as described above.

Urease activity assay. Determination of urease activity in parasitic cell ho-mogenate, wall, and culture filtrate preparations was conducted as previouslyreported (25). One unit of urease activity is defined as the amount of enzymerequired to hydrolyze 1 �mol of urea per min at 37°C. Hydrolysis was measuredby monitoring the rate of release of ammonium ions (NH4

�) plus ammonia(NH3) from the urea substrate as determined by the Bertholet reaction (25).

Isolation of the URE gene. The original genomic clone of the C. posadasiiurease gene (URE) was sequenced by the dideoxy chain termination method(39). The reported genomic sequence was compared to the nucleotide sequenceof the same strain now available in the C. posadasii genome database (www.tigr.org), and a suspected error was detected in the former. The original genomicclone was resequenced using an ABI Prism 310 genetic analyzer (Perkin-Elmer,Foster City, Calif.) and examined with MacDNASIS sequence analysis software(version 3.5) (Hitachi, San Bruno, Calif.). The revised genomic sequence of theURE gene matches the sequence in the genome database. The genomic fragmentsequenced in this study includes the full-length URE gene plus fragments of the5� and 3� untranslated regions. To clone this genomic fragment, a PCR was firstconducted with template genomic DNA of C. posadasii and the following senseand antisense primers: 5�-CTGAGAAAAAGAAAAGGGGAA-3� and 5�-CTCCACAAGCACTTGAGTA-3�. The primer sequences were based on the C.posadasii genome database. The 5.6-kb PCR product was cloned into the pCR2.1TOPO plasmid vector (Invitrogen, Carlsbad, Calif.) and sequenced as previouslyreported (12). The nucleotide sequence of the full-length clone was determinedand shown to be identical to the sequence reported in the C. posadasii genomedatabase.

Disruption of the URE gene. Transformation of C. posadasii for generation ofa mutant strain that lacks a functional URE gene was conducted as previouslydescribed (19). The p�ure construct was designed as a gene disruption vector(13) and consisted of the pAN7.1 plasmid plus a 1.6-kb genomic fragment of theURE gene. The latter was obtained by digestion of the 5.6-kb genomic clone withBglII and SnaBI (bp 1884 to 3446) and was ligated with BglII and StuI restrictionsites of the pAN7.1 plasmid, respectively. Prior to transformation, the p�ureconstruct was linearized with SplI and then purified by ethanol precipitation. Thetransformation procedure and method employed to isolate homokaryotic trans-formants were the same as previously described (19). Hygromycin (Hph)-resis-tant colonies were selected for examination by PCR, Southern hybridization, andimmunoblot analysis as previously reported (19).

Disruption of the URE gene was first confirmed by PCR using the primer pairslisted in Table 1. The HPH gene probe (Pb1), which was used for Southernhybridization to confirm single-site integration of the plasmid construct, con-sisted of a 354-bp PCR product generated by use of the primer pair e and f (listedin Table 1). A second 524-bp probe, generated with primer pair g and h, con-sisted of a 5� fragment of the URE gene (Pb2) and was used for Southernhybridization to confirm that homologous recombination had occurred in themutant strain and that the transformant was homokaryotic. A restriction map of

FIG. 1. Diagrammatic representation of the in vitro parasitic cycleof Coccidioides posadasii strain C735. The sequence of morphogeneticevents is indicated in incubation times (in hours) after arthroconidiuminoculation of the parasite-phase culture. SOW, spherule outer wallfraction.

VOL. 74, 2006 COCCIDIOIDES UREASE CONTRIBUTES TO VIRULENCE 505

on March 31, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

the genomic locus containing the URE gene and a hypothetical restriction mapof the URE gene disruption locus (�ure) were constructed to predict the size ofthe hybridized fragments in the Southern blots.

Generation of the revertant strain. To restore expression of the urease gene inthe �ure host, the mutant strain was transformed with a plasmid (pUREr), whichcontained the parental strain-derived 5.6-kb genomic clone of the URE gene inthe pCR2.1 TOPO vector. No drug-resistant selection marker was used to screenfor transformants in this case. When the parental strain of C. posadasii was grownon Difco Bacto Urea agar base medium containing a phenol red pH indicator, ayellow-to-red color change was clearly visible at the margin and under the surfaceof the fungal colony within 4 days of incubation. The �ure mutant, on the otherhand, failed to show a color change when grown on the same medium. Thisphenotypic difference was used to select putative transformants which containedthe restored wild-type URE gene. In order to distinguish the revertant strainfrom the parental strain, the 5.6-kb genomic clone used to transform the �urehost was subjected to site-directed mutagenesis using a QuickChange kit (Strat-agene, La Jolla, Calif.) according to the manufacturer’s protocol. An adenine wasadded to the genomic sequence between nucleotide positions 2299 and 2300 tocreate an AflII restriction site (2297CTTAAG2302) within an intron of the UREgene. This introduced restriction site was used for PCR analysis and Southernhybridization to confirm restoration of the wild-type URE gene in the mutantstrain and to establish that the 5.6-kb clone had integrated into chromosomalDNA of the �ure host by homologous recombination. Immunoblot analysis andthe urease activity assay were used to confirm that the revertant strain (UreR)produced amounts of enzymatically active urease which were comparable to thatof the parental strain grown under identical conditions.

Comparison of in vitro growth of the parental, mutant, and revertant strains.The growth rates of the saprobic phase of the parental, �ure, and revertant(UreR) strains were compared on the basis of dry weight during a 4-day periodof incubation in GYE liquid medium. Each culture contained 100 ml of mediumand was inoculated with 103 arthroconidia. The results of parasitic cell develop-ment of each strain cultured in defined glucose-salts medium (17) were com-pared by light microscopy to determine the percentages of endosporulatingspherules produced after a 6-day period of incubation. Each parasite-phaseculture contained 100 ml of medium and was inoculated with 107 arthroconidia.Culture aliquots of each strain were examined by light microscopy.

In an attempt to more closely simulate physiological conditions, we culturedthe parasite phase of each strain in supplemented RPMI 1640 medium andexamined the change in the extracellular ammonium-ammonia concentrationover a 7-day period of incubation. On the basis of our detection of relatively highconcentrations of urea at sites of coccidioidal infection reported in this study (anapproximately fourfold-higher concentration in Coccidioides-infected mice thanin healthy mice), we compared the amounts of NH4

�/NH3 released from para-sitic cells into the RPMI 1640 medium in the absence or presence of exogenousurea (10 mM). The range of physiological concentration of urea in blood ofnormal mice is 1 to 3 mM. We also compared the consumption of exogenousurea by the parental versus the �ure mutant strain during the same 7-day periodof incubation. Aliquots (50 �l) of the supplemented RPMI 1640 medium plusurea which supported growth of each strain were sampled daily. Each aliquot wasdiluted 10� with urease assay buffer (25) and incubated with 10 units of purified,commercially available, recombinant jack bean urease for 30 min at 37°C. Theammonium-ammonia content of each aliquot was determined by the Bertholet

reaction (25). The concentration of ammonium plus ammonia in each aliquotprior to the addition of urease was subtracted from the estimate of total NH4

�/NH3 concentration in the sample, and the final determination of ammoniaconcentration was divided by a factor of 2 for calculation of the urea content (2mol of ammonia are produced per mole of urea). All analyses were conducted intriplicate, and standard deviations were determined.

Virulence studies. To compare the virulence of the parental, mutant, andrevertant strains, arthroconidia (50 viable cells) were first obtained from GYEplate cultures of each strain grown for 30 days at 30°C. The conidia weresuspended in PBS and used to separately inoculate BALB/c mice (purchasedfrom the National Cancer Institute, Bethesda, Md.) (12 8-week-old female miceper group) by the intranasal (i.n.) route as previously reported (19). The survivalplots were subjected to Kaplan-Meier statistical analysis as previously described (19).

Three separate groups of BALB/c mice were challenged intranasally with theparental, mutant, or revertant strain as described above and then sacrificed at 12,25, or 50 days postinfection to determine the residual CFU of C. posadasii inhomogenates of the lungs and spleen as previously reported (12). Groups of fivemice per time point were examined. The CFU counts per organ were expressedon a log scale, and the Mann-Whitney U test was used to compare mediannumbers as previously described (12). The detection limit of the CFU assay is 100colonies per organ (log10 CFU � 2.00).

In situ pH measurements. For this in vivo assay, BALB/c mice were infectedeither by the intraperitoneal route with 103 viable arthroconidia as previouslyreported (12) or by the i.n. route as described above. Four groups of mice (fivemice per group) were separately challenged via one of these two routes witheither the parental or �ure mutant strain. Two weeks after challenge, each mousewas subjected to necropsy; abscesses associated with the mesentery, spleen, andlungs were immediately probed with a MI-408 needle electrode (Microelec-trodes, Londonderry, N.H.) to determine the internal pH of the infected tissue.

Estimates of urea concentration at sites of lung infection. BALB/c mice wereinfected by the i.n. route with arthroconidia of either the parental or �ure mutantstrain as described above. At 12 days postchallenge, the mice were sacrificed,lung abscesses were excised and quick-frozen, and the infected tissue was sub-jected to cryomicrotomy as previously reported (37). The 8-�m-thick sectionsfrom each abscess (10 sections per abscess) were pooled. Three mice infectedwith the parental or mutant strain were examined in this study. The pooledsections of each abscess were extracted separately with three aliquots of 150 �lof distilled water, and the total extract was assayed for urea concentration by useof commercially available, purified, jack bean urease as described above. Threeseparate lung abscesses obtained from each mouse were examined. This sameprocedure was conducted to determine the urea concentration in lung sectionsobtained from noninfected control mice.

Histopathology. Separate groups of BALB/c mice challenged intranasally witheither the parental or �ure mutant strain of C. posadasii as described above wereused for comparative histopathology. Nine mice infected with the �ure mutantstrain were sacrificed at 12, 20, or 25 days postchallenge. Their lungs wereexcised, formalin fixed, and paraffin embedded in preparation for histologicalexamination using standard procedures. Three mice infected with the parentalstrain were examined only at 12 days postchallenge.

Nucleotide sequence accession number. The revised genomic sequence of theURE gene has been deposited in GenBank under accession number U81509.3.

TABLE 1. PCR primers used to screen for disruptants and amplify selected probes for Southern hybridization

Primer Nucleotide sequence Sequencederivationa Size and use of PCR product

a 5�-GGCTGCTGCAAGTAGATCG-3� URE 1.9 kb; to screen for wild-type URE geneb 5�-GTTTGGCTTCTCGACGACG-3� URE

c 5�-AGGTTGAATCATGCAGAAGCG-3� URE 1.9 kb; to screen for disrupted URE gened 5�-CCAAATGAAGCTATGCTACCTCC-3� pAN7.1

e 5�-TGCTTTGCCCGGTGTATGAAACC-3� pAN7.1 354 bp; probe 1 (Pb1) for Southern hybridationf 5�-AAGGGATGGGAAGGATGGAGTATGG-3� pAN7.1

g 5�-TGGACACTCCTTGACCTAGCG-3� URE 524 bp; probe 2 (Pb2) for Southern hybridizationh 5�-TCCGTAGACAATAGGGCGATC-3� URE

a Nucleotide sequences of primers are based on reported DNA sequence of pAN7.1 (GenBank accession no. Z32698) and URE gene (GenBank accession no.U81509.3) (39).

506 MIRBOD-DONOVAN ET AL. INFECT. IMMUN.

on March 31, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

RESULTS

Urease is localized in the cytoplasm, vesicles, and centralvacuole of spherules and is released during endosporulation.C. posadasii strain C735 typically completes its first generationof the parasitic cycle in vitro within 5 to 6 days, and thesequence of morphogenetic events appears to be identical tothat observed in vivo (35). In addition, first-generation para-sitic cells of strain C735 grown in vitro have been shown to benear-synchronous in their development (18). This has permit-ted isolation of different cell types (e.g., spherule initials,spherules in early stages of segmentation, and mature spher-ules in the process of endosporulation) for biochemical andmolecular studies using cultures incubated for 24 h, 72 h, and132 h, respectively (Fig. 1). Arthroconidia first undergo isotro-pic growth to form spherule initials. On the basis of ultrastruc-tural studies of this early stage of parasitic cell differentiation,it appears that a multitude of tiny vesicles first visible in spher-ule initials later coalesce to form a large central vacuole (9).Segmentation of the spherule cytoplasm occurs by ingrowth ofthe parasitic cell wall and results in the formation of compart-ments which surround the still-intact, central vacuole (9). It isevident from examinations of thin sections of this stage ofdevelopment that all of the cytoplasm of the presegmentedspherule is not incorporated into the compartments that laterdifferentiate into endospores. The cytosolic debris, combinedwith the newly formed endospores and residual contents of thecentral vacuole, is released upon rupture of the mature spher-ules (Fig. 1). Another peculiar morphogenetic feature of Coc-cidioides spp. is that a lipid-rich, membranous SOW layer isproduced and shed from the spherule surface during the par-asitic cycle (Fig. 1) (17). The SOW material, which can beeasily isolated by differential centrifugation from parasite-phase cultures as a cell-free fraction, has been shown to behighly immunogenic (17). In this study we provide evidencethat enzymatically active urease, derived from the vacuolarcontents and cytosolic debris, is released from ruptured spher-ules at the time of endosporulation and is transiently associ-ated with the SOW and surface of endospores.

Immunolocalization of urease protein in organelles of dif-ferent parasitic cell types was conducted by transmission elec-tron microscopy using gold particles conjugated to C. posadasiiurease-specific IgG. Thin sections of germinated arthroconidia(24-h spherule initials) which were incubated with antibody-coated gold particles revealed small amounts of gold label inthe cytoplasm and within vesicles (not shown). Sections of invitro-grown spherules which had initiated segmentation butnot endospore differentiation (72- to 84-h cultures; Fig. 1)showed labeling in both the cytoplasmic vesicles and centralvacuole (Fig. 2A). Gold particles were also observed associatedwith cytosolic debris trapped between the central vacuole andthe newly formed segmentation wall (Fig. 1 and 2A). Thicksections of these same segmented spherules showed concen-trated urease protein-specific fluorescent labeling in both thevesicles and the central vacuole (Fig. 2B), while control sec-tions of these cells labeled with secondary antibody-FITC con-jugate alone were devoid of labeling (Fig. 2C). As segmenta-tion proceeds, the central vacuole ruptures and the preexistingvacuolar space accommodates the newly differentiated endos-pores (Fig. 1). Thin sections of endospore initials located

within the mature, unruptured spherule (120-h culture)showed antiurease-labeled gold particles concentrated in ves-icles (Fig. 2D). Endospores incubated with gold particles whichhad been conjugated to preimmune mouse serum showed nolabeling (Fig. 2E). When the spherule eventually ruptures andreleases its endospores (132- to 144-h parasite-phase culture;Fig. 1), the pool of urease protein which had presumably ac-cumulated both in the central vacuole and regions of cytosolicdebris is also discharged. The released urease protein associ-ates with both the surface of intact endospores and the mem-branous sheets of lipid-rich SOW, as revealed by immunoflu-orescence (Fig. 2F). Control endospore and SOW preparationsincubated with preimmune serum followed by secondary anti-body-FITC conjugate showed no labeling (Fig. 2G).

Results of immunoblot-urease activity assays correlate withobservations of immunolocalization of urease protein. Immu-noblot analyses of total protein preparations obtained fromparasitic cell homogenates, SOW, and culture filtrates wereconducted using urease-specific antibody (Fig. 3A). The esti-mated size of the denatured urease protein in SDS-PAGE gelsis 101 kDa (25). Aliquots of equal amounts of total proteinisolated from parasitic cell homogenates were first separatedby electrophoresis and then immunoblotted. A single, faint101-kDa band was revealed in the 24-h spherule initial prep-aration. Single bands of the same size but of much higherintensity were detected in preparations obtained from seg-mented spherules (72-h culture) and endosporulating spher-ules (132-h culture). The total protein fraction of SOW iso-lated from pre-endosporulating spherules (72 h) revealed nourease in the immunoblot, while the SOW derived from en-dosporulating cultures showed a distinct 101-kDa band. Thetotal protein fractions of concentrated culture filtrates ob-tained from 24-h, 72-h and 132-h parasite-phase cultures hadno detectable urease protein in the respective immunoblots.

Total equilibrated protein aliquots isolated from cell homog-enates, SOW fractions, and culture filtrates examined (Fig. 3A)were also tested by the urease assay for the presence of activeenzyme. The data obtained from these assays correlated wellwith results of the immunoblot analyses (Fig. 3B). A compar-ison of units of urease activity in each of the preparationsrevealed that the highest amounts of active enzyme werepresent in homogenates of the 72-h and 132-h spherules (Fig.3B). Relatively low amounts of urease activity were detected inhomogenates of the spherule initials (24-h parasite-phase cul-tures) and the SOW fraction isolated from 132-h cultures ofendosporulating spherules. No urease activity was detected inthe SOW fraction isolated from pre-endosporulating (72-h)cultures, and only minor activity was observed in the concen-trated filtrate obtained from the 132-h culture. The other cul-ture filtrates tested were devoid of urease activity.

Transformation and generation of the URE-deficient mu-tant (�ure) and URE-reconstituted (revertant) strain (UreR).The plasmid construct p�ure (not shown) which was used fortransformation of C. posadasii was designed to integrate intochromosomal DNA by a single crossover event (13) that re-sulted in disruption of the URE gene. This latter process ischaracterized by generation of two mutated copies of the tar-get gene separated by the HPH gene cassette (Fig. 4A). Trans-formation of approximately 107 germ tube-derived protoplastswith 3 �g of SplI-linearized plasmid DNA yielded 30 to 50

VOL. 74, 2006 COCCIDIOIDES UREASE CONTRIBUTES TO VIRULENCE 507

on March 31, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

hygromycin-resistant colonies. To obtain homokaryons, indi-vidual colonies were serial transferred to fresh GYE platescontaining 75 �g/ml of hygromycin. Previous C. posadasii genedisruption experiments have shown that at least three suchtransfers of putative transformants, each followed by 10 days ofincubation, are necessary to obtain homokaryotic mutants (19,32). One of the candidate transformants, referred to as the�ure strain, was selected for further analysis by Southern hy-bridization. The �ure transformant was used as the host forgeneration of a revertant strain. A transformation construct,pUREr (not shown), was designed to integrate into chromo-somal DNA of the �ure strain by a double crossover event

(12), resulting in replacement of the entire disruption locus ofthe host strain with the wild-type URE gene. The 5.6-kb trans-formation construct contained an introduced AflII restrictionsite within intron 4 of the URE gene downstream from its startcodon (Fig. 4B). Putative transformants were initially screenedby growth on Bacto Urea agar containing phenol red as a pHindicator. Transformation of 5 � 106 protoplasts with 3 �g ofthe 5.6-kb plasmid DNA yielded 15 colonies which generated acolor change (yellow to red) after 4 days of incubation at 30°Con the agar plates. DNA was separately isolated from 4 of the15 putative transformants and used for PCR amplification ofthe 1.9-kb URE gene fragment by use of primer pair a and b

FIG. 2. (A) Thin section of segmented spherule (72- to 84-h culture; see Fig. 1) labeled with antiurease and colloidal gold conjugate, whichreveals sites of intracellular localization of urease protein in the cytoplasm, vesicles (Vs), central vacuole (C.V.), and cytosolic debris trappedbetween spherule compartments and central vacuole. (B and C) Thick sections of segmented spherules stained with antiurease and FITC conjugate(B) or FITC-conjugated secondary antibody alone (C) are shown. Note the comparable results for localization of label in panels A and B. (D) Thinsection of antiurease and colloidal gold-labeled endospore (120- to 132-h culture; see Fig. 1), which shows localization of urease protein in vesicles(arrow). Control section (E) of endospore was reacted with preimmune antibody and colloidal gold conjugate. (F) Intact endospores and SOWisolated from endosporulation-phase cultures (132 h; Fig. 1), both of which are labeled with antiurease and FITC conjugate. The control cells andSOW fraction in panel G were incubated with preimmune antibody and FITC conjugate. Mt, mitochondria; Seg. wall, segmentation wall; SW,spherule wall. Bars in panels A, B, D, and F represent 1.5 �m, 60 �m, 0.5 �m, and 5 �m, respectively.

508 MIRBOD-DONOVAN ET AL. INFECT. IMMUN.

on March 31, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

(Fig. 4B; Table 1). Template genomic DNA from the parentalstrain was subjected to the same PCR amplification. Each1.9-kb amplicon derived from the putative transformants andparent strain was separately digested with AflII, and the prod-ucts were subjected by agarose gel electrophoresis. The AflII-digested amplicons of the four transformants selected from theGYE-phenol red plates each revealed 1.1-kb and 0.8-kb bandsin the ethidium bromide-stained agarose gels, as predicted(Fig. 4B). The 1.9-kb amplicon derived from the parentalstrain was not digested by AflII. One transformant, referred toas the UREr (revertant) strain, was selected from this PCR-restriction digest screening for further evaluation by Southernhybridization.

A restriction map was constructed for a 7.0-kb nucleotidesequence identified in the C. posadasii genome database (Fig.4C), which included the 5.6-kb genomic clone of the URE genedeposited in GenBank (accession no. U81509.3). Restrictionmaps were also constructed for the hypothetical structures ofthe chromosomal fragments which contained the integratedp�ure disruption plasmid (Fig. 4D) and the pUREr plasmidused to generate the revertant strain (Fig. 4E). On the basis ofthese maps, two Southern hybridization probes (Pb1 and Pb2)were generated by PCR using the primer pairs listed in Table1. Comparison of results of hybridization of Pb1 with SpeI- and

XbaI-restricted genomic DNA and ClaI- and SacII-restrictedgenomic DNA from the parental versus the �ure strain (Fig.5A) provided evidence for single-site integration of the p�ureconstruct into the host chromosomal DNA. As indicated, thePb1 probe failed to hybridize with DNA from the parentalstrain but did hybridize with bands of the predicted size in thecase of the �ure strain. The Pb2 probe hybridized with SpeI-

FIG. 3. (A) Antiurease immunoblot and (B) corresponding assayof urease activity of samples obtained from parasitic cell homogenates,isolated SOW fraction, and culture (Cult.) filtrates. The isolates werederived from first-generation cultures in near-synchronous isotropicgrowth phase (24 h), segmentation phase (72 h), and endosporulationphase (132 h). Std., standard.

FIG. 4. (A) Urease gene (URE) structure and integrated disruptant(�ure). Disruption of the URE gene resulted from the chromosomalintegration of p�ure by a single crossover event. The 1.6-kb internalfragment of URE was cloned into the pAN7.1 plasmid (construct notshown). Homologous recombination between the plasmid constructand chromosomal DNA generated two incomplete copies of the UREgene, separated by the linearized pAN7.1 sequence. Primers used toscreen by PCR for the wild-type gene or disruptant are a and b or c andd, respectively. Primers used for PCR amplification of probes (Pb1 andPb2) employed in Southern hybridization are labeled e and f andlabeled g and h, respectively. (B) Structure of disrupted urease gene(�ure) which was replaced by a double crossover event with the mod-ified wild-type gene (UREr) that contained an engineered AflII restric-tion site in an intron. Primers a and b were used for PCR amplificationof the reconstituted URE gene fragment (1.9 kb) that contained theAflII restriction site. (C) Restriction map of the 7.0-kb chromosomalfragment containing the wild-type URE gene. (D and E) Hypotheticalrestriction maps of chromosomal fragments containing the urease dis-ruption construct (�ure) and reconstituted urease gene (UREr), re-spectively.

VOL. 74, 2006 COCCIDIOIDES UREASE CONTRIBUTES TO VIRULENCE 509

on March 31, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

and StuI-restricted genomic DNA, SpeI- and SnaBI-restrictedgenomic DNA, SpeI- and NheI-restricted genomic DNA, andBstEII- and XhoI-restricted genomic DNA fragments from theparental and putative transformant (Fig. 5B). The sizes of thehybridization fragments of genomic DNA from the parentaland mutant strains matched the sizes predicted using theirrespective restriction maps (Fig. 4C and D). These data pro-vide evidence that homologous recombination occurred at asingle chromosomal locus and argue that the �ure strain ishomokaryotic. The Pb2 hybridization probe shown in Fig. 4Ewas predicted to identify a 1.5-kb SpeI- and AflII-restrictedgenomic DNA fragment of the revertant strain, which is absentfrom the restricted DNA of both the parental and �ure strains.The results of Southern hybridization in Fig. 5C provide evi-dence that homologous recombination occurred between thepUREr transformation plasmid and host chromosomal DNAby a double crossover event at a single locus.

Phenotypic comparison of the parental, mutant, and rever-tant strains. The saprobic phase of the three strains grown inGYE liquid medium revealed identical growth curves (notshown). Comparison of parasitic cell development in definedglucose-salts medium demonstrated that the three strains eachgave rise to endosporulating spherules at 5 to 6 days after thecultures were inoculated with equal numbers of arthroconidia.The percentages of parasitic cells which had ruptured andreleased their endospores in cultures of the parental, �uremutant, and revertant strains after 6 days of incubation as

revealed by microscopic examination were 89.2% � 4.8%,87.3% � 2.9%, and 83.5% � 6.3%, respectively. No significantdifference in growth and development results between thethree strains was observed.

Equal aliquots of total protein obtained from mycelial ho-mogenates of 4-day-old cultures of the three strains were equil-ibrated and separated by SDS-PAGE (8% gel) in preparationfor immunoblot analysis using antiserum raised against therecombinant urease (Fig. 6A and B). The results showed thatthe parental and revertant strains each contained a prominent101-kDa band, which correlates with the molecular size of thepurified urease protein (25). The immunoblot of the �urehomogenate, on the other hand, showed no seroreactive pro-tein band. The same equilibrated total protein samples wereused to conduct separate urease activity assays. The reactionmixtures (50 �l total volume), which consisted of 50 mMHEPES buffer (pH 7.75), 50 mM urea, and 0.5 mM EDTA plusan equal amount of total protein from each homogenate (10�g), were incubated at 37°C. Urease activity was measured bymonitoring the rate of NH4

�/NH3 release per minute in thepresence of the urea substrate over a 20-min incubation period(data not shown). The parental and revertant strains revealedcomparable levels of urease activity, while the �ure mutantstrain showed no detectable ammonia released into the reac-tion mixture. Control assays were conducted with boiled ho-mogenates (100°C, 10 min) as well as with reaction mixtureswithout the urea substrate. In both cases, the control reactionsshowed no ammonia released during the period of incubation.

The amount of ammonium and ammonia released into theculture medium during growth of the parasitic phase of eachstrain was also measured over a 7-day period of incubation.Equal numbers of arthroconidia (106 cells) from agar plates ofthe three strains were used to separately inoculate RPMI 1640medium with or without the addition of 10 mM urea (Fig. 6C).The parental and revertant strains showed comparableamounts of ammonia released into the culture filtrate whenincubated in the presence of exogenous urea, achieving NH4

�/NH3 concentrations of approximately 9.5 to 10 mM. In con-

FIG. 5. Results of Southern hybridization of restricted genomicDNA derived from the parental, �ure mutant (�), and revertant(R) strains with the Pb1 or Pb2 probe. The sizes of the restrictedfragments correlated with the predicted sizes (see Fig. 4C to E). Std.,standard.

510 MIRBOD-DONOVAN ET AL. INFECT. IMMUN.

on March 31, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

trast, when the same strains were incubated in RPMI mediumwithout exogenous urea, the NH4

�/NH3 concentrationspeaked at about 1.0 to 1.1 mM (Fig. 6C, inset). Ammonialevels detected in the culture filtrate of the �ure mutant in theabsence of exogenous urea showed a sharp increase over thefirst 2 days of incubation but then began to decrease. With theaddition of 10 mM urea to the culture medium, levels ofammonia detected in the culture filtrate of the mutant strainrevealed only a slight increase over the 7-day period of incu-bation. Aliquots of the culture medium supplemented with 10mM urea were also sampled daily during incubation of theparental, revertant, and mutant strain to test for consumptionof urea as described above in the Materials and Methodssection. The urea concentration in the filtrate obtained fromcultures of the parental and revertant strains showed a near-linear decrease from 10 mM to 4.2 mM over the 7-day periodof incubation. The amount of urea consumed correlated ap-proximately with the amount of NH4

�/NH3 released into themedium supporting growth of the two strains. The culturefiltrate of the mutant strain, on the other hand, showed a slightincrease in urea content from 10 mM to 12 mM between days5 and 7 of incubation.

Comparison of the virulence of the parental, mutant, andrevertant strains. Equal numbers of arthroconidia (50 viablecells) isolated from the parental, mutant (�ure), and revertant(UreR) strains were used to separately inoculate three groupsof BALB/c mice by the intranasal route (Fig. 7A). Mice chal-

lenged with arthroconidia derived from the parental and re-vertant strains showed comparable high rates of mortality. Allthe animals had died by 22 days postchallenge. On the otherhand, animals infected with the mutant strain showed a slowerrate of mortality over the first 3 weeks after intranasal inocu-lation, and approximately 55% of the mice survived beyond 50days postchallenge. The difference between the survival plotsfor mice infected with the mutant versus the parental or re-vertant strain was highly significant (P � 0.001).

The CFU in homogenates of the lungs and spleen ofBALB/c mice infected intranasally with the parental or mutantstrain of C. posadasii were determined at 12, 25, and 50 dayspostchallenge. Comparable numbers of CFU were detectedafter 12 days in the lungs of mice infected with the parentalstrain (median of 6.41 [log10]; range of 5.85 to 6.87) or mutant(5.83; 5.70 to 5.94). On the other hand, the spleens of micechallenged with the parental strain were infected (2.85; 2.10 to3.55), but no CFU were detected in spleens of mice challengedwith the �ure strain. The lungs of mice challenged with themutant strain showed a significant reduction in CFU at 25 dayspostchallenge (2.78; 2.70 to 3.78) compared to the results seenwith mice infected with the parental strain at 12 days postchal-lenge (P � 0.002). The spleen was still uninfected. None of themice challenged with the parental strain survived to 25 days. At50 days postchallenge, no CFU were detected on culture platesinoculated with undiluted homogenates of the lungs or spleenof mice which survived challenge with the �ure mutant strain.

FIG. 6. (A) SDS-PAGE separation of equal amounts of total protein obtained from mycelial homogenates (4-day cultures) of the parental (P),�ure mutant (�), and revertant (R) strains. (B) Immunoblot of protein separation shown in panel A produced using antiurease antibody. Notethe absence of the 101-kDa band in the preparation of the mutant strain (�). (C) Comparative analysis of amounts of NH4

�/NH3 released intothe culture medium during incubation of each strain in the presence or absence of exogenous urea. The concentrations of extracellular ammoniawere determined during 7 days of incubation of the parental (�), �ure mutant (F), and revertant (E) strains in supplemented RPMI medium plus10 mM exogenous urea. The changes in concentration are plotted as first-order linear regression lines. (Inset) The change in concentration ofammonia in the absence of exogenous urea is plotted point to point. Std., standard.

VOL. 74, 2006 COCCIDIOIDES UREASE CONTRIBUTES TO VIRULENCE 511

on March 31, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

In situ determinations of pH at infection sites. BALB/c micechallenged via the intraperitoneal route with 103 arthroconidiaof either the parental or mutant strain, and sacrificed 2 weekslater, responded by production of visible abscesses in theirlungs and spleen, on the surface of the diaphragm, and inassociation with the mesenteric tissue. Immediately after themice were sacrificed and internal organs were exposed, pHmeasurements of abscesses were performed in situ by penetra-tion of the host tissue surface layer with a needle electrode.The results in Fig. 7B show significantly lower pH values forabscesses produced in response to the �ure mutant straincompared to the host response to the parental strain (P 0.02). BALB/c mice were also inoculated via the intranasalroute with 50 arthroconidia of either the parental or mutantstrain and sacrificed at 2 weeks postchallenge, and their lungswere examined as described above. The average pH of lungabscesses in mice infected with the parental strain was approx-imately 7.7. In contrast, the average pH of lung abscesses inmice infected intranasally with the �ure strain was approxi-mately 7.2 (Fig. 7C). This difference in pH of lung abscesses inthe intranasally challenged mice is also statistically significant(P � 0.001).

Detection of high concentrations of urea at sites of infectionwith the parental strain. On the basis of the substantial in-crease in the in vitro production of ammonia by the parentalstrain in the presence of 10 mM urea, we speculated that ifexogenous urea is present at sites of lung infection it may alsoinfluence the levels of NH4

�/NH3 production and the pH of

the microenvironment. Figure 7D shows the results of a com-parison of the amounts of urea detected in aqueous extracts offrozen sections of normal, noninfected murine lung tissue ver-sus sectioned abscesses from the lungs of mice at 14 days afterintranasal challenge with either the parental or �ure mutantstrain of C. posadasii. The approximate fourfold increase inurea concentration at sites of infection with the parental strainmay at least partly account for the elevated pH values recordedfor lungs of mice infected with this strain (Fig. 7B and C). Onthe other hand, a significantly lower concentration of urea wasdetected at sites of infection with the mutant strain comparedto the parental strain results (P � 0.002).

Comparative histopathology. Two groups of BALB/c micewere challenged via the intranasal route with equal numbers ofarthroconidia produced by either the parental or �ure mutantstrain. Mice infected with the parental strain at 12 days post-challenge showed multiple abscesses in their lungs. Histologi-cal sections of these abscesses suggested that an intense in-flammatory response had occurred without evidence ofgranuloma differentiation. Large numbers of spherules in pre-and post-endosporulation stages of development were visible,and apparent host tissue damage had occurred at sites of in-fection (Fig. 8A). Inflammatory cells (see “Inf. cells” in Fig.8A) appeared to have responded in a chemotactic manner tothe released contents of ruptured spherules. Host phagocyteswere frequently visible within the lumen of the parasitic cells(arrow at right in Fig. 8A). Sections of murine lung tissueinfected with the mutant strain after 12 days showed compa-rable high numbers of the pathogen but also an indication of amore organized host inflammatory response at sites of lunginfection than observed in lung tissue infected with the paren-tal strain (not shown). At 20 days postchallenge, mice infectedwith the parental strain were moribund or dead, while thesurviving animals which had been challenged with the �uremutant showed visible signs of recovery (activity, weight gain,smooth rather than ruffled fur) and histological evidence ofearly differentiation of granulomas which had largely con-tained the pathogen (arrows in Fig. 8B). At 25 days afterchallenge, mice infected with the mutant strain showed devel-opment of well-differentiated granulomas (Fig. 8C). Spherulescontained within the granuloma were either destroyed or sur-rounded by host inflammatory cells, and few endospores werevisible. Results of these studies correlate with our observationsof a marked reduction in CFU in the lungs of BALB/c mice at25 days after intranasal challenge with the �ure mutant strain.

DISCUSSION

In vitro growth of the saprobic phase of Coccidioides spp. for8 days at 30°C in liquid medium containing only glucose andyeast extract resulted in a final extracellular NH4

�/NH3 con-centration of 7.5 mM and an increase in pH from 6.7 to 7.8(25). Growth of the parasitic phase in defined glucose-saltsmedium has also been reported to result in release of ammoniafrom the fungal cells, with a concomitant increase in the ex-tracellular pH (8). Urease activity is a major source of ammo-nia and an important virulence determinant for microbialpathogens (6). However, little is known about the function offungal ureases compared to the wealth of information on theirhomologs in plants and bacteria. The most extensively studied

FIG. 7. (A) Survival plots of BALB/c mice challenged intranasallywith equal numbers of arthroconidia isolated from the parental, �uremutant, or revertant strain. (B and C) pH of infected host tissue(abscesses) at 2 weeks after intraperitoneal challenge with 103 arthro-conidia (B) or after intranasal challenge with 50 arthroconidia (C) ofeither the parental or �ure mutant strain. Mesen., mesentery; Spl.spleen. (C) Urea concentration in extracts of frozen sections of lungtissue obtained from normal, noninfected BALB/c mice (Cont.), miceinfected with the parental strain (P), or mice infected with the mutantstrain (�ure) of C. posadasii and sacrificed at 12 days postchallenge.Equal numbers of sections (10 per abscess) were collected, pooled, andextracted with the same total volume (450 �l) of distilled water.

512 MIRBOD-DONOVAN ET AL. INFECT. IMMUN.

on March 31, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

urease is that of jack bean (Canavalia ensiformis); it was thefirst enzyme to be crystallized and is the first example of anickel [Ni (II)] metalloproteinase (2). The native plant ureaseis a hexamer composed of identical polypeptide subunits, eachwith an estimated molecular size of 91 kDa. The molecular sizeof the monomer of the C. posadasii urease, whose amino acidsequence is 76% homologous to that of the jack bean urease,has been estimated to be 101 kDa (25). The approximate sizeof the native fungal enzyme is 450 kDa, and it is apparentlycomposed of four subunits. The C. posadasii urease demon-strates biochemical similarities to other fungal ureases, includinga pH optimum of 8.0 and thermostability up to 60 to 65°C (39).

Although urease is a cytosolic enzyme that lacks a signalpeptide, evidence derived from studies of both plant and bac-terial ureases suggests that the active enzyme may also be cellsurface associated (14, 24). In the case of H. pylori, the extra-cellular urease has been suggested to be essential for the bac-teria to resist exposure to acid (22). Host exposure to bacterialurease has been reported to elicit an inflammatory response(36), which may lead to tissue damage and exacerbation of the

microbial infection when this response is persistent and intense(26). Recombinant urease of Coccidioides has been shown tobe highly immunogenic in BALB/c mice (23). Both the recom-binant fungal protein and a mammalian plasmid vector con-taining the URE gene have been used to vaccinate BALB/cmice and demonstrated significant protective efficacy againstcoccidioidal infection. These latter results further argue thatthe fungal urease is presented to the host during the course ofa natural infection (23).

The mechanism by which urease becomes associated withthe surface of H. pylori involves initial release of the enzymefrom cells that undergo autolysis, followed by adsorption of theenzymatically active protein to intact, viable bacteria (15). Acomparable sequence of events is suggested to occur duringthe parasitic cycle of C. posadasii. The urease protein has beenimmunolocalized to the spherule cytoplasm and vesicles andthe large central vacuole. Active urease escapes from spheruleswhen the parasitic cells rupture and release their contents. Theenzyme subsequently associates both with the surface of intactendospores and the membranous, SOW fraction which is pro-

FIG. 8. (A) Histopathology of lung tissue from BALB/c mouse infected intranasally with the parental strain of C. posadasii at 12 dayspostchallenge. Note the high concentration of host inflammatory cells (Inf. cells), some of which surrounded and entered a ruptured spherule(arrow at right). (B) Sectioned lobe of lung tissue from a BALB/c mouse infected intranasally with the �ure mutant strain of C. posadasii andsacrificed at 20 days postchallenge. Note early development of granulomas (arrows). (C) Sectioned granuloma in lung of mouse infected with the�ure mutant strain and sacrificed 25 days postchallenge. Partially decomposed spherules are surrounded by inflammatory cells. The bar in panelA represents 40 �m, and the bars in panels B and C represent 2 mm.

VOL. 74, 2006 COCCIDIOIDES UREASE CONTRIBUTES TO VIRULENCE 513

on March 31, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

duced in abundance during in vitro growth of the parasiticphase (17, 19). Positive correlation of data obtained from im-munoblot analyses and enzymatic assays of parasitic cell ho-mogenates suggested that most of the active urease is presentin the spherule cytosol during the pre-endosporulation (72 h)and endospore differentiation (132 h) stages of development.However, a significant amount of active enzyme was also de-tected in association with the cell-free, SOW fraction. Prelim-inary studies have demonstrated that the urease located byimmunofluorescence at the surface of endospores is also en-zymatically active (not shown).

The results of these initial investigations suggested that ure-ase activity of C. posadasii in vivo may contribute to the gen-eration of an alkaline microenvironment at the surface of thefungal pathogen as well as stimulate host inflammatory re-sponse to the extracellular protein. Together, these conditionscould lead to host tissue damage and exacerbation of the se-verity of coccidioidal infections. On the basis of this hypothesis,we designed a strategy to disrupt the URE gene. Targeteddisruption of Coccidioides genes by homologous recombina-tion (13) is now well established in our laboratory (19, 32).Application of the protoplast method to transform C. posadasii(32) has proved to be superior to alternative approaches, suchas electroporation, biolistic transformation (38), and Agrobac-terium-facilitated DNA transfer (1). Results of PCR analysisand Southern hybridization have demonstrated that chromo-somal integration of the gene disruption construct occurred ata single site by homologous recombination. In order to satisfy“Koch’s postulates,” we have successfully developed a genereconstitution strategy for Coccidioides spp. (19). We haveshown that this process of gene-targeted replacement also oc-curred by single-site integration of the wild-type gene intochromosomal DNA. We have confirmed that the normal func-tion of the targeted URE gene was restored and that the pa-rental and revertant strains have identical phenotypes.

Acetohydroxamic acid is a reversible urease inhibitor andwas previously shown to completely block in vitro activity ofthe purified Coccidioides enzyme when added to mycelial ho-mogenates at a concentration of 10 mM (25). The addition ofacetohydroxamic acid at the same concentration to live myce-lial cultures coincubated for 12 h resulted in a dramatic reduc-tion in urease activity compared to control culture results, butNH4

�/NH3 continued to be released into the growth medium(25). Metabolic sources of intracellular ammonia synthesisother than urease activity have been previously proposed (6)but have not yet been explored for Coccidioides spp. Never-theless, disruption of the URE gene of C. posadasii resulted ina significant difference between the amounts of extracellularammonia detected in cultures of the parental and revertantstrains versus the �ure mutant strain cultures. This differencewas most clearly demonstrated upon addition of 10 mM ureato the culture medium. The pH change over the 7-day periodof incubation of the parental strain in RPMI medium plusexogenous urea was not dramatic (6.6 to 7.2), presumablybecause of the buffering capacity of the medium. Generation ofan alkaline, extracellular microenvironment by bacteria thatrelease ammonia has been suggested to occur by diffusionrather than active transport of NH4

�/NH3 (3). This wouldrequire that the intracellular concentration of ammonia ex-ceeds that of extracellular ammonia. Although we have not

measured intracellular NH4�/NH3 concentrations of the par-

asitic cells of C. posadasii, indirect evidence suggests that theyare high. We have shown that the concentration of urea in theculture medium which supported growth of the parental straindecreased (10 mM to 4.2 mM) over 7 days of incubation,suggesting that the parasitic cells are capable of uptake andconsumption of the exogenous substrate. Intracellular ureamost likely undergoes hydrolysis immediately after it enters theparasitic cells of C. posadasii because of the toxicity of thissubstrate. Assuming that the intracellular pH of spherules andendospores is controlled, it is likely that high ammonium-ammonia concentrations resulting from urea uptake and ure-ase activity generate a diffusion gradient for release of NH4

�

and NH3 from the cells. However, it is also possible that morethan one NH4

�/NH3 transport system exists in C. posadasii, ashas been proposed for certain ammonia-releasing bacteria (3).

Loss of urease activity from the �ure strain of C. posadasiiresulted in a marked reduction in the pathogenicity of theorganism in the lungs of BALB/c mice. Approximately 55% ofthe animals survived an arthroconidial challenge which is typ-ically lethal to this strain of mice, and all the survivors clearedthe infection between 25 and 50 days postchallenge. The sig-nificance of this partial loss of virulence is reinforced by thefact that BALB/c mice represent a host which is exceptionallysusceptible to disseminated coccidioidomycosis following in-tranasal challenge with arthroconidia (21). Inoculation ofBALB/c mice with the same number of spores of the parentalor revertant strain resulted in 100% mortality after 22 days. Weused both intraperitoneally and intranasally infected mice togenerate coccidioidal tissue abscesses which were large enoughto allow measurement of the internal pH. We recorded asignificantly higher pH at sites of infection with the parentalstrain (7.4 to 7.7) than with the urease knockout strain (6.8 to7.1). We also showed that abscesses from BALB/c mice result-ing from infection with the parental strain contained higherconcentrations of urea (approximately 630 �M) than abscessesproduced in response to infection with the mutant strain. Ourestimation of the concentration of urea detected in sections oflung tissue obtained from normal, noninfected mice was 160 �M.

We have observed that murine macrophages respond invitro to infection with the wild-type strain of C. posadasii by anapproximately threefold increase in the level of expression ofthe host arginase I gene compared to that in noninfected mac-rophages (unpublished data). A sharp rise in arginase I pro-duction in vivo during Leishmania infection is indicative of aTh2 pathway of immune response (20), which in both leish-maniasis and coccidioidomycosis is an indicator of a poor dis-ease outcome (10). Arginase I competes with inducible nitricoxide synthase in macrophages for the common substrate, L-arginine, which results in reduction of the basal level of nitricoxide and increased production of ornithine and urea (20).Expression of the arginase gene of C. posadasii (28) during invitro growth of the parasitic phase in the presence of exoge-nous urea is constitutive (unpublished data). On this basis, wesuggest that the high amounts of urea detected at sites ofcoccidioidal infection are host derived. The availability of ex-ogenous urea would provide substrate for both extracellularand intracellular urease activity, result in elevated productionand accumulation of NH4

�/NH3 at infection sites, and cause apH increase that could be damaging to host tissue (31). Results

514 MIRBOD-DONOVAN ET AL. INFECT. IMMUN.

on March 31, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

of histopathological studies of lung tissue infected with theparental strain suggested that extensive tissue damage occursduring the course of disease; this damage may be mediatedboth by the pathogen and the host (7). Persistently high levelsof proinflammatory cytokine production (e.g., interleukin-6)during coccidioidal lung infection may result in additional hosttissue damage due to localized and intense inflammatory cellresponse (10). On the other hand, BALB/c mice infected withthe urease knockout strain showed a robust, protective re-sponse to intranasal challenge characterized by early formationof granulomas and clearance of the pathogen from animalswhich survived to 50 days postchallenge. In summary, thesedata suggest that elevated levels of intracellular and extracel-lular urease activity occur at sites of coccidioidal infection withvirulent strains of C. posadasii and result in localized highconcentrations of NH3/NH4

�, together with persistent and in-tense inflammatory response to the pathogen. These combinedfeatures of coccidioidomycosis, which are attributed at least inpart to urease production, contribute significantly to the patho-genesis of Coccidioides spp.

ACKNOWLEDGMENTS

Support for this study was provided by Public Health Service grantsAI19149 and AI37232 from the National Institute of Allergy andInfectious Diseases, National Institutes of Health.

We are grateful to Kalpathi Seshan for his assistance in conductingthe ultrastructural studies.

REFERENCES

1. Abuodeh, R. O., M. J. Orbach, M. A. Mandel, A. Das, and J. N. Galgiani.2000. Genetic transformation of Coccidioides immitis facilitated by Agrobac-terium tumefaciens. J. Infect. Dis. 181:2106–2110.

2. Blakeley, R. L., and B. Zerner. 1984. Jack bean urease: the first nickelenzyme. J. Mol. Catal. 23:263–292.

3. Brewin, B., P. Woodley, and M. Drummond. 1999. The basis of ammoniumrelease in nifL mutants of Azotobacter vinelandii. J. Bacteriol. 181:7356–7362.

4. Bump, W. A. 1925. Observations on growth of Coccidioides immitis. J. Infect.Dis. 36:561–565.

5. Burall, L. S., J. M. Harro, X. Li, C. V. Lockatell, S. D. Himpsl, J. R. Hebel,D. E. Johnson, and H. L. T. Mobley. 2004. Proteus mirabilis genes thatcontribute to pathogenesis of urinary tract infection: identification of 25signature-tagged mutants attenuated at least 100-fold. Infect. Immun. 72:2922–2938.

6. Bury-Mone, S., J.-M. Thiberge, M. Contreras, A. Maitournam, A. Labigne,and H. D. Reuse. 2004. Responsiveness to acidity via metal ion regulatorsmediates virulence in the gastric pathogen Helicobacter pylori. Mol. Micro-biol. 53:623–638.

7. Casadevall, A., and L.-A. Pirofski. 2003. The damage-response framework ofmicrobial pathogenesis. Nat. Rev. Microbiol. 1:17–24.

8. Cole, G. T. 1997. Ammonia production by Coccidioides immitis and its pos-sible significance to the host-fungus interplay, p. 247–263. In H. vandenBossche, D. A. Stevens, and F. C. Odds (ed.), Host-fungus interplay. PlenumPress, New York, N.Y.

9. Cole, G. T., D. Kruse, K. R. Seshan, S. Pan, S. J. Szaniszlo, J. Richardson,and B. Bian. 1993. Factors regulating morphogenesis in Coccidioides immitis,p. 191–212. In H. Van den Bossche, F. C. Odds, and D. Kerridge (ed.),Dimorphic fungi in biology and medicine. Plenum Press, New York, N.Y.

10. Cole, G. T., J.-M. Xue, C. N. Okeke, E. J. Tarcha, V. Basrur, R. A. Schaller,R. A. Herr, J.-J. Yu, and C.-Y. Hung. 2004. A vaccine against coccidioid-omycosis is justified and attainable. Med. Mycol. 42:189–216.

11. Cox, G. M., J. Mukherjee, G. T. Cole, A. Casadevall, and J. R. Perfect. 2000.Urease as a virulence factor in experimental cryptococcosis. Infect. Immun.68:443–448.

12. Delgado, N., J. Xue, J.-J. Yu, C.-Y. Hung, and G. T. Cole. 2003. A recom-binant -1,3-glucanosyltransferase homolog of Coccidioides posadasii pro-tects mice against coccidioidomycosis. Infect. Immun. 71:3010–3019.

13. d’Enfert, C., G. Weidner, P. C. Mol, and A. A. Brakhage. 1999. Transforma-tion systems of Aspergillus fumigatus. New tools to investigate fungal viru-lence. Contrib. Microbiol. 2:149–166.

14. Dunn, B. E., and M. G. Grutter. 2001. Helicobacter pylori springs anothersurprise. Nat. Struct. Biol. 8:480–482.

15. Dunn, B. E., and S. H. Phadnis. 1998. Structure, function and localization ofHelicobacter pylori urease. Yale J. Biol. Med. 71:63–73.

16. Fisher, M. C., G. L. Koenig, T. J. White, and J. W. Taylor. 2002. Molecularand phenotypic description of Coccidioides posadasii sp. nov., previouslyrecognized as the non-California population of Coccidioides immitis. Myco-logia 94:73–84.

17. Hung, C. Y., N. M. Ampel, L. Christian, K. R. Seshan, and G. T. Cole. 2000.A major cell surface antigen of Coccidioides immitis which elicits both hu-moral and cellular immune responses. Infect. Immun. 68:584–593.

18. Hung, C.-Y., J.-J. Yu, P. F. Lehmann, and G. T. Cole. 2001. Cloning andexpression of the gene which encodes a tube precipitin antigen and wall-associated -glucosidase of Coccidioides immitis. Infect. Immun. 69:2211–2222.

19. Hung, C.-Y., J.-J. Yu, K. R. Seshan, U. Reichard, and G. T. Cole. 2002. Aparasitic phase-specific adhesin of Coccidioides immitis contributes to thevirulence of this respiratory fungal pathogen. Infect. Immun. 70:3343–3456.

20. Iniesta, V., L. C. Gomez-Nieto, I. Molano, A. Mohedano, J. Carcelen, C.Miron, C. Alonso, and I. Corraliza. 2002. Arginase I induction in macro-phages, triggered by Th2-type cytokines, supports the growth of intracellularLeishmania parasites. Parasite Immunol. 24:113–118.

21. Kirkland, T. N., and J. Fierer. 1983. Inbred mouse strains differ in resistanceto lethal Coccidioides immitis infection. Infect. Immun. 40:912–916.

22. Krishnamurthy, P., M. Parlow, J. B. Zitzer, N. B. Vakil, H. L. Mobley, M.Levy, S. H. Phadnis, and B. E. Dunn. 1998. Helicobacter pylori containingonly cytoplasmic urease is susceptible to acid. Infect. Immun. 66:5060–5066.

23. Li, K., J.-J. Yu, C.-Y. Hung, P. F. Lehmann, and G. T. Cole. 2001. Recom-binant urease and urease DNA of Coccidioides immitis elicit an immuno-protective response against coccidioidomycosis in mice. Infect. Immun. 69:2878–2887.

24. Millanes, A. M., B. Fontaniella, M. L. Garcia, M. T. Solas, C. Vicente, andM. E. Legaz. 2004. Cytochemical location of urease in the cell wall of twodifferent lichen phycobionts. Tissue Cell 36:373–377.

25. Mirbod, F., and R. A. Schaller. 2002. Purification and characterization ofurease isolated from the pathogenic fungus Coccidioides immitis. Med. My-col. 40:35–44.

26. Mobley, H. L., M. D. Island, and R. P. Hausinger. 1995. Molecular biologyof microbial ureases. Microbiol. Rev. 59:451–480.

27. Olszewski, M. A., M. C. Noverr, G.-H. Chen, G. B. Toews, G. M. Cox, J. R.Perfect, and G. B. Huffnagle. 2004. Urease expression by Cryptococcus neo-formans promotes microvascular sequestration, thereby enhancing centralnervous system invasion. Am. J. Pathol. 164:1761–1771.

28. Pan, S., M. Zhang, and G. T. Cole. 1995. Isolation and characterization of thearginase-encoding gene (arg) from Coccidioides immitis. Gene 154:115–118.

29. Pappagianis, D. 1980. Epidemiology of coccidioidomycosis, p. 68–85. InD. A. Stevens (ed.), Coccidioidomycosis: a text. Plenum Press, New York, N.Y.

30. Petkus, A. F., L. L. Baum, R. B. Ellis, M. Stern, and D. L. Danley. 1985. Purespherules of Coccidioides immitis in continuous culture. J. Clin. Microbiol.22:165–167.

31. Prusky, D., J. L. McEvoy, B. Leverentz, and W. S. Conway. 2001. Localmodulation of host pH by Colletotrichum species as a mechanism to increasevirulence. Mol. Plant-Microbe Interact. 14:1105–1113.

32. Reichard, U., C.-Y. Hung, P. W. Thomas, and G. T. Cole. 2000. Disruptionof the gene which encodes a serodiagnostic antigen and chitinase of thehuman fungal pathogen Coccidioides immitis. Infect. Immun. 68:5830–5838.

33. Saubolle, M. A. 1996. Life cycle and epidemiology of Coccidioides immitis, p.1–8. In H. Einstein and C. Catanzaro (ed.), Coccidioidomycosis. NationalFoundation for Infectious Diseases, Washington, D.C.

34. Sendide, K., A.-E. Deghmane, J.-M. Reyrat, A. Talal, and Z. Hmama. 2004.Mycobacterium bovis BCG urease attenuates major histocompatibility com-plex class II trafficking to the macrophage cell surface. Infect. Immun. 72:4200–4209.

35. Sun, S. H., G. T. Cole, D. J. Drutz, and J. L. Harrison. 1986. Electron-microscopic observations of the Coccidioides immitis parasitic cycle in vivo.J. Med. Vet. Mycol. 24:183–192.

36. Tanahashi, T., M. Kita, T. Kodama, Y. Yamaoka, N. Sawai, T. Ohno, S.Mitsufuji, Y. P. Wei, K. Kashima, and J. Imanishi. 2000. Cytokine expres-sion and production by purified Helicobacter pylori urease in human gastricepithelial cells. Infect. Immun. 68:664–671.

37. Xue, J., C.-Y. Hung, J.-J. Yu, and G. T. Cole. 2005. Immune response ofvaccinated and non-vaccinated mice to Coccidioides posadasii infection. Vac-cine 23:3535–3544.

38. Yu, J.-J., and G. T. Cole. 1998. Biolistic transformation of the human patho-gen Coccidioides immitis. Microbiol. Methods 33:129–141.

39. Yu, J.-J., S. L. Smithson, P. W. Thomas, T. N. Kirkland, and G. T. Cole.1997. Isolation and characterization of the urease gene (URE) from thepathogenic fungus Coccidioides immitis. Gene 198:387–391.

Editor: T. R. Kozel

VOL. 74, 2006 COCCIDIOIDES UREASE CONTRIBUTES TO VIRULENCE 515

on March 31, 2020 by guest

http://iai.asm.org/

Dow

nloaded from