of January 28, 2018.This information is current as

InfectionsCandida glabrataDoes Not Clear Toll Pathway Controls butDrosophilaThe

Marie-Céline Lafarge and Dominique FerrandonJessica Quintin, Joelle Asmar, Alexey A. Matskevich,

ol.1201861http://www.jimmunol.org/content/early/2013/02/10/jimmun

published online 11 February 2013J Immunol 

MaterialSupplementary

1.DC1http://www.jimmunol.org/content/suppl/2013/02/11/jimmunol.120186

        average*  

4 weeks from acceptance to publicationSpeedy Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright © 2013 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on January 28, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

by guest on January 28, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

The Journal of Immunology

The Drosophila Toll Pathway Controls but Does Not ClearCandida glabrata Infections

Jessica Quintin,1 Joelle Asmar,2 Alexey A. Matskevich,3 Marie-Celine Lafarge, and

Dominique Ferrandon

The pathogenicity of Candida glabrata to patients remains poorly understood for lack of convenient animal models to screen large

numbers of mutants for altered virulence. In this study, we explore the minihost model Drosophila melanogaster from the dual

perspective of host and pathogen. As in vertebrates, wild-type flies contain C. glabrata systemic infections yet are unable to kill the

injected yeasts. As for other fungal infections in Drosophila, the Toll pathway restrains C. glabrata proliferation. Persistent C.

glabrata yeasts in wild-type flies do not appear to be able to take shelter in hemocytes from the action of the Toll pathway, the

effectors of which remain to be identified. Toll pathway mutant flies succumb to injected C. glabrata. In this immunosuppressed

background, cellular defenses provide a residual level of protection. Although both the Gram-negative binding protein 3 pattern

recognition receptor and the Persephone protease-dependent detection pathway are required for Toll pathway activation by C.

glabrata, only GNBP3, and not psh mutants, are susceptible to the infection. Both Candida albicans and C. glabrata are restrained

by the Toll pathway, yet the comparative study of phenoloxidase activation reveals a differential activity of the Toll pathway

against these two fungal pathogens. Finally, we establish that the high-osmolarity glycerol pathway and yapsins are required for

virulence of C. glabrata in this model. Unexpectedly, yapsins do not appear to be required to counteract the cellular immune

response but are needed for the colonization of the wild-type host. The Journal of Immunology, 2013, 190: 000–000.

Candida species are the most common cause of invasivefungal disease in humans. Morbidity caused by dissem-inated candidiasis continues to increase despite the

availability of new antifungal drugs and has not changed much inthe past two decades (1, 2). Candida albicans remains the pre-dominant cause of disseminated candidiasis, accounting for morethan half of all cases. However, Candida glabrata, which is in-volved in mucosal and bloodstream infections, has emerged as thesecond most common cause of invasive candidiasis in multiple

countries (3, 4). C. glabrata systemic infections are difficult totreat because of their natural resistance to echinocandins and azolederivatives, especially to fluconazole, which is the classical anti-fungal treatment effective on C. albicans (3–5). Despite its name,C. glabrata is phylogenetically more closely related to Saccha-romyces cerevisiae, a nonpathogenic haploid yeast, than to C.albicans (6). Indeed, it is mostly a monomorphic yeast, a majordifference with the dimorphic C. albicans. Thus, a current chal-lenge is to understand what makes C. glabrata virulent whencompared with the nonvirulent yeast S. cerevisiae and the op-portunistic pathogen C. albicans. Mammalian models are notoptimal to identify virulence factors on a large scale because oflow throughput, cost and labor-intensive procedures, and ethicalissues. Furthermore, the study of the innate immune mechanisms,which are crucial for the outcome of systemic antifungal defense(7), are hampered in mammals by the adaptive immune responsethat is superimposed onto the innate immune response. Never-theless, several C. glabrata–specific virulence factors have beenidentified such as adhesins and cell-bound proteases (8–14).Invertebrates provide interesting alternative models to study

microbial virulence factors and host immune responses (15, 16).Drosophila melanogaster, a small fly that is easy to rear, benefitsfrom a century of genetic investigations. Its innate immunity hasbeen widely studied during the past two decades (for a review, seeRef. 17). D. melanogaster is an excellent tool for the identificationof novel virulence genes, for medium-throughput screening ofantifungal molecules against Candida species, and for investi-gating pathogenicity of C. albicans (18–20).Flies deficient for the evolutionarily conserved Toll signaling

pathway are sensitive to fungal infections (18, 21, 22). The Tollpathway of D. melanogaster controls the systemic antifungal andthe Gram-positive antibacterial host response, whereas immunedeficiency (IMD), a second NF-kB–like pathway, is involved inthe host response against Gram-negative bacteria infection. TheToll receptor is activated by the Spatzle cytokine, following its

Unite Propre de Recherche 9022 du Centre National de la Recherche Scientifique,Universite de Strasbourg, Institut de Biologie Moleculaire et Cellulaire, F67084Strasbourg Cedex, France

1Current address: Department of Experimental Internal Medicine, Radboud Univer-sity Nijmegen Medical Centre, Nijmegen, The Netherlands.

2Current address: Cegedim Strategic Data, Medical Research, Boulogne BillancourtCedex, France.

3Current address: Theramab, Moscow, Russia.

Received for publication July 6, 2012. Accepted for publication January 8, 2013.

This work was supported by Centre National de la Recherche Scientifique, Fondationpour la Recherche Medicale, and the ERAnet Pathogenomics FUNPATH program.J.Q. was supported by a fellowship from the French Ministry of Research. TheStrasbourg team is an “Equipe Fondation pour la Recherche Medicale,” a label awardedby the Fondation pour la Recherche Medicale.

J.Q. and D.F. designed most experiments; J.Q., J.A., A.A.M., and M.-C.L. performedthe experiments; J.Q. and D.F. analyzed the experiments; and J.Q. and D.F. wrote themanuscript.

Address correspondence and reprint requests to Dr. Dominique Ferrandon, UnitePropre de Recherche 9022 du Centre National de la Recherche Scientifique, Uni-versite de Strasbourg, Institut de Biologie Moleculaire et Cellulaire, 15 Rue ReneDescartes, F67084 Strasbourg Cedex, France. E-mail address: D.Ferrandon@ibmc-cnrs.unistra.fr

The online version of this article contains supplemental material.

Abbreviations used in this article: DIF, Dorsal-related immunity factor; GNBP3,Gram-negative binding protein 3; HOG, high-osmolarity glycerol; IMD, immunedeficiency; NR, nicotinamide riboside; PO, phenoloxidase; PRR, pattern recognitionreceptor; PSH, Persephone.

Copyright� 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1201861

Published February 11, 2013, doi:10.4049/jimmunol.1201861 by guest on January 28, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

proteolytic maturation into an active ligand by the Spatzle-processing enzyme (23). Gram-negative binding protein 3(GNBP3) binds to fungal b-(1,3)-glucan and activates a proteo-lytic cascade that ultimately matures Spatzle through the Spatzle-processing enzyme (22, 24). However, D. melanogaster does notsolely rely on pattern recognition receptors (PRRs) to detectinfections. The host protease Persephone (PSH) is able to sense theenzymatic activity of microbial virulence factors, namely, someproteases, and subsequently triggers the Toll pathway through thetranscription factor Dorsal-related immunity factor (DIF) (22, 25).The systemic immune response of D. melanogaster consists of thetranscriptional induction of tens of genes, antimicrobial peptidegenes such as Drosomycin, Diptericin, and Cecropins being highlyexpressed. These peptides are produced by the fat body, a func-tional equivalent of the mammalian liver, and subsequently se-creted into the hemolymph. A complementary facet of theDrosophilahost defense is the cellular immune response. Sessile and circu-lating hemocytes phagocytose microorganisms present in the he-molymph. Finally, a septic wound triggers immediate reactionsthat depend on proteolytic cascades, including the deposition ofmelanin at the wound site, which is thought to constitute anotherarm of host defense. Several steps of the melanization reactionsare catalyzed by phenoloxidase (PO), which becomes active afterthe cleavage of its prodomain after the activation of dedicatedproteolytic cascades (26–29). Of note, independently of its Toll-dependent function, the GNBP3 PRR is also required for the ac-tivation of PO after a fungal challenge and is required for theagglutination of C. albicans cells in the hemolymph (30). Basedon its ability to form complexes that include PO and a serpin, ithas been proposed that GNBP3 may muster the formation of anattack complex upon the detection of b-(1,3)-glucans (30).In this study, we develop D. melanogaster as a model to inves-

tigate C. glabrata pathogenesis. We find that C. glabrata kills onlyToll pathway immunocompromised flies. Like in mammals, wild-type Drosophila contain but are unable to clear C. glabrata infec-tions (10). Although C. glabrata infections are detected by thefungal PRR of Drosophila, the known Toll-independent functions ofGNBP3 do not appear to be required for host defense. C. glabratacells are ingested by hemocytes, yet phagocytosis appears to be anessential host mechanism against fungi only in Toll pathway mutantflies. Finally, we demonstrate that C. glabrata strains mutated forthe yapsin virulence factors or the high-osmolarity glycerol (HOG)pathway are hypovirulent, thus establishing Drosophila as a suitablemodel to identify and study novel C. glabrata virulence factors.

Materials and MethodsMicrobial strains

Gram-positive bacteria used in this study includedMicrococcus luteus (CIPA270; kind gift from H. Monteil, University Louis Pasteur, Strasbourg,France). Gram-negative bacteria used in this study included Escherichiacoli. Yeasts strains are listed in Supplemental Table I. Bacteria and yeastswere grown overnight in liquid cultures of Luria-Bertani broth (37˚C) andyeast extract/peptone/dextrose broth (29˚C), respectively, and collected bycentrifugation.

Fly strains

Stocks were raised on standard cornmeal-agar medium at 25˚C. w A5001and DD1 cn bw flies were used as wild-type throughout the experimentsbecause the majority of mutants were generated in these backgrounds.GNBP3hades, psh4-GNBP3hades, psh4, GNBP1osi, Dif1, spz, MyD88, key1,pll78/pll21 (pelle), tub238/tub118 (tube), dl1/J4 (Dorsal), PAE1, and PAEex1

stocks have been described previously (21, 22, 31–36).

Survival experiments

Batches of 20 to 25 wild-type and mutant strains were challenged by septicinjury. The flies were either infected using a needle previously dipped into

a concentrated solution of bacteria or yeasts, or using a thin capillary. Thecapillary was filled with a solution of yeast cells (OD = 20) resuspended infresh PBS containing Tween 0.01% to avoid agglutination and capillaryblockage. The injections of 9.2 nl in the thorax were performed usinga Nanoject II apparatus (Drummond Scientific). The vials containing theinfected flies were then put in an incubator at the desired temperature (29˚Cfor fungal infections, 25˚C for bacterial challenges), and the surviving flieswere counted as required. Flies were usually put into new vials every otherday. Results are expressed as a percentage of surviving flies at differenttime points postinfection. Each graph is representative of at least threeindependent experiments.

Quantitative RT-PCR

Quantitative RT-PCR was performed as described previously (22, 33).Primers for Cecropin were as follows: forward, 59-ACGCGTTGGTCAG-CACACT-39; reverse, 59-ACATTGGCGGCTTGTTGAG-39. The Droso-mycin and Cecropin mRNA expression level of wild-type flies challengedwith M. luteus was used as 100% reference. Diptericin mRNA expressionlevel of wild-type flies challenged with E. coli was used as 100% refer-ence.

Proliferation assay

Yeast titers were obtained by determining the CFUs. Infected flies werehomogenized in PBS supplemented with 0.1% Tween and serially diluted.Hemolymph of infected flies were collected in PBS and treated the sameway. Dilutions were plated on yeast extract/peptone/dextrose broth agarplates containing either hygromycin or streptomycin for selection andincubated overnight.

In vivo phagocytosis assays

Four thousand cells of C. glabrata were injected in flies. In vivo phago-cytosis assays were then performed as previously described (24).

Cell culture

Drosophila S2 cells were cultured in Schneider’s medium (Invitrogen,Carlsbad, CA) supplemented with 10% FBS, penicillin, and streptomycin.

In vitro phagocytosis assays

Drosophila S2 cells (1 3 105) were plated in eight-well labtech slides in150 ml Schneider’s medium with 10% FBS and penicillin/streptomycin.Fifty microliters C. glabrata preparation at 5 3 105 cells/ml was added toeach well containing S2 cells and incubated for 2 h at 25˚C. After coin-cubation, the cell culture medium was aspirated and cells were fixed with2% formaldehyde in PBS, washed with 13 PBS, and blocked with 2%BSA in PBS for 2 h. The cell surfaces of C. glabrata were detected withprimary Abs raised against whole Candida (a kind gift from Karl Kuchlerin Vienna). Primary Abs were visualized with Cy3-conjugated goat anti-rabbit (Zymed Laboratories). In a second step, to visualize engulfed yeaststhat were protected from binding the Ab by phagocytes, we permeabilizedcells in BSA and Triton X-100 (0.1%), and then incubated them again withthe same primary Ab. The subsequent procedure was identical exceptthat primary Abs were visualized with Alexa 488 goat anti-rabbit Ab(Invitrogen). In each case, C. glabrata cells were considered phagocytosedif they were visualized only with the Alexa 488 goat anti-rabbit secondaryAbs (green). DNA was visualized with DAPI. Slides were mounted inVectashield medium (Vector Laboratories) and were examined with afluorescent Zeiss axioscope microscope (Carl Zeiss, Gottingen, Germany).Slides were kept at 4˚C for storage. Images generated were processedusing Adobe PhotoShop CS (Adobe Systems) and analyzed using ImageJplug-in RVB profiler.

Hemolymph extraction

Hemolymph was collected from the flies using thin glass capillariesmounted on the Nanoject apparatus (Drummond Scientific) and added toPBS supplemented with protease inhibitors (Complete mixture tablets;Roche). Protein concentration was determined using a Bradford assay.

Saturation of phagocytosis

For experiments in which phagocyte function is ablated, surfactant-free red,0.3-mm diameter carboxylate-modified latex beads (Interfacial Dynamics)were washed in PBS and used 43 concentrated in PBS (corresponding to5–10% solids), and 69 nl was injected 18–24 h before septic injury. Theefficiency of the procedure in blocking phagocytosis was assessed usingfluorescent-labeled bacteria or yeast and Trypan blue (data not shown), aspreviously described (32, 37).

2 C. GLABRATA PERSISTENCE IN DROSOPHILA

by guest on January 28, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

Western blot

All hemolymph extracts, containing 10 mg protein, were equilibrated inLaemmli solution and denatured at 95˚C for 5 min before loading on a 12or 8% SDS-PAGE. After SDS-PAGE, proteins were blotted to HybondECL nitrocellulose membranes (Amersham) and blocked for 2 h at roomtemperature in 5% fat dry milk (Bio-Rad) in TBST (0.1%). Blots wereincubated overnight at 4˚C with the anti-PO Abs in 0.5% fat dry milk inTBST. After washing with TBST, the blots were incubated for 2 h at roomtemperature with the HRP-conjugated donkey anti-rabbit secondary Ab(Amersham). After washing with TBST, blots were detected by ECLaccording to the manufacturer’s instructions.

Agglutination and measurement of PO activity

Agglutination and PO activity assays were performed as described previ-ously (30).

Killing of microbes and immunolocalization

Killing of microbes and immunolocalization were performed as previouslydescribed (24).

Data processing and statistical analysis

Data are representative of at least three independent experiments (n $ 3 ineach experiment) that yielded similar results unless stated otherwise in thisarticle. Results from quantitative analyses are expressed as mean 6 SD ofthe data from at least three independent experiments, unless stated otherwisein this article. Statistical analyses were performed using Student t test, andp , 0.05 was considered significant. Statistical significance of survival ex-periment was calculated using the product limit method of Kaplan and Meierusing the log-rank test (GraphPad PRISM4 software), and p , 0.05 wasconsidered significant. All data significantly different from control values aremarked with asterisks, and p values are given in the figure legends.

ResultsThe Toll pathway plays an essential role in host defense againstinjected C. glabrata

We first injected C. glabrata, C. albicans, and S. cerevisiae inwild-type flies. Three reference wild-type strains of C. glabrataand C. albicans hardly killed flies, and there was no significantdifference in survival rates with flies infected with three strains ofS. cerevisiae (Fig. 1A and data not shown). The few flies that weredying apparently did not succumb to the fungal infection, becausethe microbial titer had not significantly increased in those de-ceased flies (data not shown). Next, we determined that flies de-ficient for Toll pathway activation, namely, MyD88 mutants, weresensitive to a C. glabrata challenge but not to S. cerevisiaeinfections (Fig. 1B). Note, however, that C. albicans appears to bemore virulent than C. glabrata. A similar sensitivity to C. glabratainfection was found for other mutants affecting the core Tollpathway, namely, spz, pll, and tub (Supplemental Fig. 1A). Ge-netically null Dif1 mutants displayed a somewhat variable, lowersensitivity to this challenge that was also observed in dose–re-sponse experiments (Fig. 1G; compare Supplemental Fig. 1A and1B, Supplemental Fig. 1C and 1D). The variable Dif1 phenotypehad, however, not been observed after C. albicans infection, whichkilled rapidly these mutants (22, 25). This lower sensitivity of theDif mutant may be accounted by a partial functional redundancywith dorsal (31, 38, 39), even though dorsal hemizygous mutantswere not at all susceptible to C. glabrata (Supplemental Fig. 1A).As expected, flies mutant for the IMD pathway gene kenny (key)(40) survived the infection, as well as wild-type flies (Supple-mental Fig. 1B). The C. glabrata titer in both whole flies (Fig. 1C)and extracted hemolymph (Fig. 1D) rapidly increased in MyD88flies, which suggests that mutant flies succumb to fungemia.Strikingly, we found that C. glabrata was not cleared in wild-typeflies because the C. glabrata titer actually slowly increased overa period of 2 wk. We conclude that the Toll pathway is a majordefense against injected C. glabrata because it prevents the

growth of the fungus. Importantly, our data are also compatiblewith a dynamic equilibrium between proliferation and the Toll-mediated elimination of the invading yeasts.Next, we monitored the induction of the humoral systemic im-

mune response using Drosomycin as readout of Toll pathway ac-tivation and Diptericin as that of the IMD pathway. As expected,we observed an induction of Drosomycin expression after a Can-dida or a S. cerevisiae challenge, which, however, was significantlylower than that induced byM. luteus, used as a reference in most ofour experiments (Fig. 1E). Surprisingly, S. cerevisiae, as comparedwith Candida, elicited the weakest response, whereas in mammalsit does robustly trigger the innate immune response (41). In keepingwith the persistence of C. glabrata in the course of the infection,we did detect a significant expression of Drosomycin up to 12 dafter the challenge (Supplemental Fig. 1E). In contrast, M. luteus,which is rapidly cleared by the fly immune response, triggered onlya 2-d-long expression of this antimicrobial peptide gene. We did notdetect any induction of Diptericin expression, even though severalfilamentous fungi induce a short-lived response (42) (SupplementalFig. 2A). CecropinA, which is regulated by both the IMD and Tollpathways in mixed bacterial infections (21), was induced 3 h afterchallenge to levels that were close to those induced by M. luteus(Supplemental Fig. 2B). As expected, the induction of Drosomycinand Cecropin was blocked in mutants that affect the core Tollpathway (Fig. 1F, Supplemental Fig. 2B). Taken together, theseexperiments indicate that C. glabrata is efficiently detected by theinnate immune system of the fly.To assess how the fly detects C. glabrata, we first checked that

GNBP1, which is required for sensing Gram-positive bacteria, wasnot involved in host defense against this fungus (survival andDrosomycin expression data not shown). Next, we monitored thesurvival of psh, GNBP3hades, and psh-GNBP3hades mutant flies(Fig. 1G). As previously reported for C. albicans, we found thatGNBP3hades and psh-GNBP3hades mutant flies succumbed likeMyD88 flies, whereas psh displayed only a weak susceptibilityphenotype (22). Indeed, Drosomycin induction was impaired inGNBP3hades and psh-GNBP3hades mutant flies (Fig. 1H). However,in contrast with infections with C. albicans, Drosomycin expres-sion was also significantly decreased in psh mutant flies. Theobservation that GNBP3hades, and not psh, mutants succumb rap-idly to C. glabrata, even though Drosomycin expression is af-fected to similar levels in both mutants, suggests that GNBP3 mayplay other roles in the immune defense that, as reported previously(30), are independent of its function in Toll pathway activation.

The known Toll-independent functions of GNBP3 may not beinvolved in the host defense against C. glabrata

We have previously reported that C. albicans are agglutinated ina GNBP3-dependent manner (30). We therefore tested whether C.glabrata cells are agglutinated after incubation with hemolymphcollected from wild-type or GNBP3hades mutant flies. As shown inFig. 2A, we did not detect any significant aggregates as had beenobserved previously with C. albicans (30).GNBP3 is also required for the activation of the proteolytic

cascade that turns pro-PO into an active PO enzyme, which cat-alyzes the melanization reactions. We found that PO activity wasinduced by C. glabrata, but to a lower extent, as compared with C.albicans (Fig. 2B). In Dif mutants, PO was unexpectedly activatedto similar levels as wild-type flies by C. glabrata, in contrast withchallenges with other microbes (28, 30). We also found that pro-PO was poorly cleaved into active PO after a C. glabrata chal-lenge, as compared with M. luteus and C. albicans (Fig. 2C). Asexpected, pro-PO cleavage was inhibited in MyD88 mutants. Fi-nally, we also failed to detect any sensitivity of CG3066 mutants

The Journal of Immunology 3

by guest on January 28, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

to a C. glabrata challenge, in which the PO activating enzyme ismutated (data not shown).

The cellular arm of the immune response provides anadditional layer of defense against C. glabrata

We determined that hemocytes phagocytose living C. glabratain vivo (Fig. 3A), as previously shown for C. albicans (30). Similar

findings were established with S2 cultured cells, which originatefrom embryonic macrophages (data not shown). Interestingly, westill did find ingested C. glabrata in fly hemocytes 7 d after thebeginning of the infection (Supplemental Fig. 3A).It has been reported that C. glabrata is not killed by mammalian

macrophages and actually grow inside them (8, 10). To determinewhether hemocytes may constitute an intracellular niche in which

FIGURE 1. C. glabrata and the systemic humoral response of Drosophila. (A and B) Survival experiments of wild-type (wt) (A) and MyD88 (B) mutant

flies to C. glabrata (C. g), S. cerevisiae (S. cere), and C. albicans (C. a) strains. Flies were challenged with a thin needle previously dipped in a pellet of C.

glabrata. All yeast pellets were prepared similarly for pricking. Survival was monitored at 29˚C. The survival rate is expressed as percentage. ***p ,0.0001. (C and D) In vivo proliferation assay in wt and MyD88 mutant flies (MyD88). C. glabrata (BG14 [HYG-R]; C. g) were retrieved from crushed

whole flies (C) or from the hemolymph (D). Graphs show the log CFUs per single fly. A total of 104 yeast cells were injected (C. g input) and indeed

recovered. PBST 0.1% was used for plating. Yeast cells were diluted in PBST 0.01% for injection. AllMyD88 mutant flies were dead after day 2, hence the

lack of data for subsequent time points. Data are representative of at least three independent experiments (pool of n $ 15 in each experiment) that yielded

similar results. (E and F) Quantitative analysis of Drosomycin mRNA induction after Candida infection in wt flies (E) and Toll mutant flies (F). Drosomycin

mRNA expression after C. glabrata or C. albicans infection is significantly lower than after a challenge withM. luteus at 24 h postinfection (E). Drosomycin

induction is blocked in Toll mutant flies (F). *p , 0.05, **p , 0.01, ***p , 0.0001 when compared with the corresponding values of wt challenged flies,

respectively, by C. glabrata and by M. luteus. (G) Survival experiments of Dif, GNBP3hades, Persephone (psh), double GNBP3hades-psh, andMyD88 mutant

flies to C. glabrata infection. For GNBP3hades, double GNBP3hades-psh and MyD88 mutant flies, ***p , 0.0001 when compared with the survival curve of

wt flies. (H) Drosomycin mRNA quantification 24 h after Candida challenge. C. glabrata detection and subsequent Toll activation pass through PSH and

GNBP3. *p, 0.05, **p, 0.01 for comparison with the corresponding values of wt C. glabrata challenged flies. Survival data are representative of at least

three independent experiments (n $ 20 in each experiment) that yielded similar results. Results from quantitative analyses are expressed as mean 6 SD of

the data from at least three independent experiments. NC, Nonchallenged; NI, noninduced; ns, not significant.

4 C. GLABRATA PERSISTENCE IN DROSOPHILA

by guest on January 28, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

C. glabrata can proliferate while being protected from the sys-temic immune response, we impaired the phagocytic activity ofhemocytes by the prior injection of nondigestible latex beads. Ourprediction was that the survival of wild-type flies would not be

affected and that the C. glabrata titer would decrease. Indeed, wedid not observe a significant change in the survival curves of C.glabrata–challenged wild-type flies preinjected with latex beadsas compared with PBS preinjected controls (Fig. 3B). However,

FIGURE 2. Agglutination and melanization of C. glabrata. (A) Agglutination of yeast cells in Drosophila hemolymph. Hemolymph was extracted from

wild-type (wt) or GNBP3hades mutant (hades) flies and incubated with C. glabrata. BSAwas used as a control. The GNBP3-dependent agglutination of C.

albicans has been previously published (30). Data are representative of at least three independent experiments. (B and C) C. glabrata and melanization. PO

enzymatic activity was measured after C. albicans and C. glabrata challenge (B). C. glabrata induced lower PO activity than C. albicans in wt flies. This

activity is not blocked in Dif mutants after C. glabrata challenge. Enzymatic activity is expressed as mean 6 SD of the data from three independent

experiments (n $ 20 in each experiment, *p , 0.05, **p , 0.01). Immunoblotting experiment showing PPO cleavage in its active PO form (C). PPO

cleavage is induced at 4 h after challenge with M. luteus, C. albicans, and to a lesser extent with C. glabrata. PPO cleavage is blocked in MyD88 mutant

flies. Data are representative of three independent experiments. ns, Not significant.

FIGURE 3. Phagocytosis of C. glabrata (C. g). (A) In vivo phagocytosis of C. glabrata by Drosophila hemocytes. Adult Drosophila carcasses were

dissected and fixed 45 min after C. glabrata injection. Hemocytes are shown in yellow. In each experiment, yeasts cells were first stained in red before

permeabilization and then in green. Ingested cells can only be stained in green, whereas yeasts on the outside are stained by both Abs. DNA was stained

with Hoechst 33258 and hemocytes with Cy5 (yellow). Data are representative of three independent experiments (n = 15). Original magnification 32000.

(B) Phagocytosis in adult flies plays an essential role in host defense of Toll mutant flies. Flies were either preinjected with latex beads (LXB) or PBS and

subjected to a septic injury with C. glabrata. Data are expressed as mean of the data from three independent experiments (n$ 20 in each experiment). LXB

preinjected MyD88 mutant flies were significantly more susceptible to infection than PBS-injected flies (**p , 0.01). (C) In vivo proliferation assay of

wild-type (wt) and MyD88 mutant flies (MyD88). C. glabrata cells were retrieved from crushed whole flies. Before the injection of 104 yeast cells, flies

were either preinjected with LXB or PBS. Graph shows the log CFUs of C. glabrata per single fly, and data are representative of three independent

experiments (pool of n $ 15 in each experiment) that yielded similar results. All MyD88 mutant flies were dead after day 7, and all MyD88 mutant flies

pretreated with latex beads succumbed after day 2, hence the absence of data for subsequent time points. (D) Inactivation of phagocytosis by injections of

LXB leads to increased Toll pathway activation after C. glabrata infection. Flies were either preinjected with LXB or PBS and subjected to a septic injury

with C. glabrata. Drosomycin mRNA rate were then quantified 24 h postinfection. **p , 0.005. Data are expressed as mean of the data from four in-

dependent experiments (n = 15 in each experiment). CI, Clean injury; M. l, M. luteus.

The Journal of Immunology 5

by guest on January 28, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

we found that C. glabrata was still able to persist in phagocytosis-impaired wild-type flies (Fig. 3C). The somewhat decreased titeras compared with controls might be accounted for by a strongerinduction of the Toll pathway (Fig. 3D), suggesting that the hostmore easily detects C. glabrata yeasts when hemocytes canno longer ingest them. Actually, the finding that MyD88 mutantsin which phagocytosis is blocked succumb more rapidly to C.glabrata than phagocytosis-competent MyD88 flies (Fig. 3B)strongly suggests that hemocytes form a remaining defense in Tollpathway-deficient flies. In keeping with this observation, the C.glabrata titer reached a lethal threshold around 2 d postinfection(Fig. 3B, 3C).

Drosophila as a model to study C. glabrata interactions with itshost

C. glabrata is pathogenic to immunodeficient patients; yet, onlya few animal models that lead to C. glabrata–induced death areavailable to date (e.g., see Ref. 11). In this study, we askedwhether flies might be useful to identify virulence factors of thispathogenic yeast. C. glabrata is auxotrophic for NAD+. C. glabratais able to grow in the bladder and urinary tract of patients becauseof its ability to synthesize NAD+ from external metabolites such asnicotinamide ribosides (NRs) through the use of either the NR ki-nase (Nrk1) or the nicotinamidase Pnc1 (43, 44). We found that nrk1and pnc1 mutants are as virulent as control wild-type C. glabratawhen introduced into MyD88 flies (Fig. 4A), which suggests thatNAD+ may be available in the hemocoel of the fly. Next, we askedwhether yapsins might be required for virulence in our model. The

11 C. glabrata yapsins are secreted GPI-anchored aspartyl pro-teases. Yapsins, especially YPS1 and YPS7, are required for theintegrity of the cell wall, control of the level of adherence, andsurvival within macrophages (8). We found that both a yps1-yps7double-mutant strain and a yps1-11D strain, in which all 11 yapsingenes have been deleted, were less virulent in MyD88-immunode-ficient Drosophila (Fig. 4A). The yps1-yps7 phenotype was, how-ever, variable in some experiments when compared with that ofyps1-11D. We next assessed whether the C. glabrata yapsins mightbe required to counteract the cellular response mediated by hosthemocytes that forms an essential defense in Toll pathway–deficientflies (Fig. 3B). Following this hypothesis, the yps1-11D C. glabratastrain should become as virulent as the wild-type C. glabrata strainin Toll mutant flies that are deprived of a functional phagocyticsystem. The virulence of the yps1-11D strain was enhanced inMyD88 mutants that were preinjected with latex beads, but, how-ever, not to the level observed with wild-type C. glabrata in sim-ilarly pretreated MyD88 flies (Fig. 4B). Thus, yapsins are stillrequired as virulence factors inMyD88-immunosuppressed flies thatalso lack a cellular immune response, implying that yapsin functionis not required to counteract phagocytosis. We next investigatedfurther the pathogenesis of the less virulent yps1-11D C. glabratastrain by assessing the hemolymph titer in MyD88 flies. Interest-ingly, the yapsin mutant strain initially proliferated in MyD88 mu-tant flies but was subsequently controlled (Fig. 4C). Importantly,unlike the wild-type strain, the persistence of the yps1-11D C.glabrata strain was abolished in wild-type flies, suggesting that theyps1-11D strain is more sensitive to the action of the Toll pathway.

A B C

D E

FIGURE 4. C. glabrata (C. g) mutants for virulence factors. (A) Survival of Drosophila to wild-type, yapsin, and NR salvage pathway mutant C. glabrata

infection.MyD88 mutant flies succumbed to nrk1D or pnc1D at a similar rate than to wild-type C. glabrata infection.MyD88 mutant flies succumbed more

slowly to a challenge with either yps1Dyps7D or yps1-11D than with wild-type C. glabrata (*p , 0.05 for comparison with the corresponding MyD88 flies

infected with the wild-type C. glabrata strain). (B) C. glabrata yapsins are not required to counteract phagocytosis (*p , 0.01). (C) The proliferation of the

yapsin mutant strain is controlled by both wild-type and MyD88 mutant flies. C. glabrata were retrieved from the hemolymph. Graphs show the log CFUs

per single fly. All MyD88 mutant flies infected with the wild-type C. glabrata strain or the yps1-11D C. glabrata strain were dead after days 2 and 7,

respectively; hence the lack of data for subsequent time points. Data are representative of three independent experiments (pool of n = 15 in each ex-

periment) that yielded similar results. (D) C. glabrata yapsins do not induce the PSH pathway. Drosomycin mRNA quantification 24 h after C. glabrata

challenge. Results from quantitative analyses are expressed as mean 6 SD of the data from 20 flies. (E) Survival of Drosophila to wild-type ATCC2001 or

HOG mutant C. glabrata infection. MyD88 mutant flies succumbed less rapidly to a challenge with either sho1D or pbs2D than with wild-type C. glabrata.

*p , 0.05 when compared with MyD88 mutant flies infected with the wild-type C. glabrata strain. (A, B, E) Flies were challenged with a thin needle

previously dipped in a pellet of C. glabrata. Survival was monitored at 29˚C. The survival rate is expressed in percentage. The survival results are the mean

of three to four independent experiments for each strain (n $ 20 in each experiment).

6 C. GLABRATA PERSISTENCE IN DROSOPHILA

by guest on January 28, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

We have previously reported that the Drosophila PSH proteaseactivates the Toll pathway when proteolytically matured by secretedfungal virulence factor (22) (Fig. 5). We therefore next assessedwhether the C. glabrata yapsins were required to elicit the PSHpathway. However, we did not detect any decrease in the Droso-mycin expression elicited by the yps1-11D strain as compared withwild-type C. glabrata in GNBP3hades flies (Fig. 4D).The HOG pathway triggers adaptation of yeasts through intra-

cellular accumulation of osmolytes. It has been reported that C.glabrata strains mutated for the HOG pathway showed severedefects under high-osmolarity conditions (45). Sho1 and Sln1 arecell-surface membrane proteins that redundantly detect and activatetwo independent signaling branches that converge on the Pbs2 ki-nase, which subsequently phosphorylates the downstream MAPKHOG1 (Supplemental Fig. 3B). Gregori et al. (45) have demon-strated that the ATCC2001 strain lacks the Ssk2 kinase gene, whichrenders the Sln1 branch nonfunctional. Therefore, the mutation ofthe cell-surface sensor of the unique functional branch, Sho1(sho1D), causes osmosensitivity similar to that obtained with themutation of the downstream MAPKK Pbs2 (pbs2D). Interestingly,MyD88 mutant flies infected with ATCC2001-sho1D or with theATCC2001-pbs2D strains were killed at the same slow rate ascompared with MyD88 flies infected with the wild-type ATCC2001strain (Fig. 4E). The ssk2 gene is functional in the BG2 strain, an-other wild-type strain, resulting in a functional redundancy betweenthe Sho1 and Sln1 activating branches of the HOG pathway (45).Interestingly, we found that only pbs2 mutant C. glabrata, and notsho1 or ssk2 mutants, were less virulent when injected into MyD88flies (Supplemental Fig. 3C), a situation that mirrors that describedfor osmosensitivity in the BG2 strain. These findings in two inde-pendent strains make it unlikely that the effects we observe are dueto second-site mutations, and thus allow us to conclude that the HOGpathway is required for C. glabrata virulence in our Drosophilainfection model. Altogether, these results suggest that Drosophila isa suitable model to identify some virulence factors of C. glabrata.

DiscussionIn this study, we deciphered the pathogenesis of C. glabrata in-fection, an increasingly common cause of morbidity and mortality

in humans, in the model host Drosophila. As compared with otherfungi, C. glabrata presents some specific features in this model.We discuss the unique and common attributes of C. glabrata, andexamine the value and limitations of D. melanogaster as a modelto study C. glabrata pathogenesis.

C. glabrata persist even in wild-type flies

Flies deprived of a functional Toll pathway succumb to C. glabratainfection, thus highlighting again the primary role of the Tollpathway in systemic antifungal host defense (18, 21, 22, 46). Theinoculum size of C. glabrata we used did not induce a highmortality in wild-type flies in this systemic infection model for atleast a period of 2 wk (data not shown). Remarkably, although theToll pathway is an essential antifungal defense mechanism, wild-type flies are not able to clear C. glabrata infections within afortnight. This persistence is reminiscent of the situation observedin mice in which C. glabrata persists up to 4 wk without killingthe immunocompetent animal (47). These observations suggesteither that C. glabrata is able to evade or withstand the host im-mune response or, alternatively, that a dynamic equilibrium isestablished between C. glabrata proliferation and the host-inducedantifungal response. It has been shown that C. glabrata survivesand multiplies in mammalian macrophages (8, 10). In Drosophila,C. glabrata is detected within hemocytes up to 7 d postinfection.It remains to be determined whether these yeasts had been freshlyengulfed or had been phagocytosed early on during the infectionand had been able to persist. It is, however, unlikely that they areable to proliferate in hemocytes because we did not detect aqualitatively distinct number of ingested C. glabrata after 1 or7 d (Supplemental Fig. 3A). Furthermore, it is unlikely thathemocytes constitute a major niche in which yeasts proliferate orat least persist because we have shown that phagocytosis is indeedan efficient host defense mechanism against C. glabrata infec-tions, at least in a Toll pathway mutant background. In addi-tion, C. glabrata still persists, although at decreased levels, inphagocytosis-impaired wild-type flies (Fig. 3C). Macrophagesare ultimately damaged by ingested C. glabrata (10). WhetherDrosophila hemocytes are also finally lysed at later stages of theC. glabrata infection remains unknown. We note, however, that no

FIGURE 5. PO is differentially activated through the Toll pathway by C. glabrata and C. albicans. Pro-PO is cleaved into active PO by the MP1 protease,

which is itself activated by the MP2 protease, an activation that is negatively regulated by the 27A serpin (SPN27A) (27–29). Pro-PO activation is directly

triggered by GNBP3, which actually binds pro-PO and PO (30). PO activation also requires the activation of the Toll pathway, which regulates the ex-

pression of unknown factors required for pro-PO cleavage (X, Y) (28). PO activation by C. albicans requires DIF, through the expression of X, whereas PO

activation by C. glabrata is not dependent on DIF. We propose in this study that in this latter case, a DIF/Dorsal heterodimer controls the expression of

a distinct factor (Y).

The Journal of Immunology 7

by guest on January 28, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

hematopoiesis has been described in adult Drosophila. Thus, theobservation of phagocytosed yeasts inside hemocytes at least up to7 d after an immune challenge demonstrates that not all hemocytesare lysed at this stage.The persistence of C. glabrata in immunocompetent flies

demonstrates that the Toll-mediated immune response is not ef-fectually fungicidal and might have a fungistatic effect, although,as discussed earlier, a dynamic equilibrium cannot be excluded.Tens of genes are activated on stimulation of the Toll pathway,with the antimicrobial peptides being strongly upregulated. Yet,no efficient effectors against yeasts have been described. Droso-mycin and Metchnikowin are active only on filamentous fungi (48,49). Moreover, in vitro fungicidal assays failed to reveal any ac-tivity of Drosomycin on C. albicans and C. glabrata yeasts (50,51). In vivo, the ubiquitous overexpression of Drosomycin orMetchnikovin in spatzle Toll mutant flies did not rescue theirsusceptibility phenotype to C. albicans infection (data not shown).The effect of CecropinA has not been tested in vivo, andCecropinA activity in vitro remains controversial. It has beenpreviously reported that the yeast S. cerevisiae and the filamentousfungi Metarhizium anisopliae were susceptible to CecropinA (52).However, other in vitro analyses have established that DrosophilaCecropinA is inactive against S. cerevisiae, C. glabrata, and C.albicans, but effective against some filamentous fungi (Neuros-pora crassa, Fusarium oxysporum, Fusarium culmorum) (50, 51).In this study, we have shown that the IMD pathway key mutantswere resistant to C. glabrata infection and that Dif mutant flieswere hardly affected by the infection, yet our preliminary dataindicate that the induction of CecropinA is blocked in thesemutants, in keeping with similar results obtained with other mi-crobial challenges (31, 40). It is therefore unlikely that this anti-microbial peptide plays a major role in the host defense against C.glabrata. Overall, our results clearly point out the importance ofan intact and functional Toll pathway to counteract C. glabratainfection, but its effectors remain to be discovered.Altogether, our data indicate that C. glabrata is able to persist in

wild-type flies and that this phenomenon depends on yapsins (Fig.4C). The observation that phagocytosis is not required for C.glabrata persistence opens the possibility that the ability of thisopportunistic yeast to persist in its host over long periods is notdirectly linked to its capacity to proliferate in vertebrate macro-phages ex vivo.

Attributes of the Drosophila host response to C. glabrata

C. glabrata is detected both through the GNBP3 and PSH armsof the fungal detection system. Although b-(1,3)-glucans are theligands of GNBP3 (24), the C. glabrata protease that activates Tollthrough PSH remains undefined as we have shown in this articlethat yapsins are not involved in this process. This function mightbe achieved through a subtilisin-like protease identified in thegenome (T. Gabaldon, personal communication).An unexpected finding was the almost normal resistance of psh

mutants to a C. glabrata challenge as compared with GNBP3hades

mutants, even though both affect Toll signaling to the same extent.This is reminiscent of the C. albicans case as exposed in the in-troduction to this article (30). However, agglutination and melani-zation cannot account for this disparity in the survival phenotypesas regards C. glabrata. One interesting possibility is that GNBP3might function as an opsonin because, in contrast with the situationfor C. albicans (30), phagocytosis is a relevant host defense inToll pathway mutants. Another hypothesis would be that GNBP3musters an effective attack complex on C. glabrata (30).C. glabrata is thought to evade detection by the immune system

in mammals (10), which may account for its persistence. In flies,

C. glabrata is efficiently detected by the systemic surveillanceapparatus. However, it is not agglutinated, as is the case for C.albicans, and it does not fully trigger the PO cascade, a phenom-enon that might be linked to the limited role of DIF in triggeringthis cascade, as discussed later. Overall, C. glabrata hardly eludesimmune detection in Drosophila.A peculiarity of the C. glabrata infection model regards the role

of the DIF transcription factor. It is not required for PO activa-tion, unlike the C. albicans case (30), and the susceptibility of thismutant to a C. glabrata challenge was variable. This is likelylinked to a redundancy with the Dorsal transcription factor. Weenvision two possibilities. In the first model, the Toll pathwaywould be activated differently after a C. glabrata challenge ascompared with a C. albicans challenge, for instance, by modifyingthe mode of activation of the Toll receptor by amplitude or fre-quency modulation (53). A second model is that the set of ef-fectual effector genes may vary between C. glabrata and C.albicans, and that their promoters have different requirements interms of the actual NF-kB dimers that promote their transcription.For instance, in the case of the PO cascade, it is hypothesized thatthe Toll pathway regulates the expression of an activator of the POcascade. This activator might be distinct in the case of C. glabrata(Fig. 5). A genome-wide transcriptomic analysis will be requiredto resolve this issue. Thus, the C. glabrata model raises an in-teresting problem with respect to Toll signaling.

Cellular defense against C. glabrata

As for mammals, we show in this article that C. glabrata cells arephagocytosed by hemocytes in the body cavity of the adult fly.This process is essential to fight infection when the flies are de-prived of the systemic immune response mediated by the Tollpathway. However, phagocytosis of C. glabrata does not representthe limiting step for the clearance of C. glabrata as the yeastspersist in wild-type flies deprived of a functional phagocytic ap-paratus. It has been reported in larvae that Toll-dependent anti-microbial production by the fat body requires the Spatzle ligandsecreted by hemocytes (54). However, a role of phagocytes in theinduction of the systemic humoral Toll response in adult flies hasnever been proved. Our observations that wild-type flies presentno susceptibility phenotype to C. glabrata infection upon phago-cytosis blockade allow us to exclude a role of phagocytosis in Tollpathway activation.We have observed in this study that C. glabrata remains present

in the hemocytes of adult Drosophila up to 7 d postinfection andthat hemocytes form an essential defense in Toll pathway-deficientflies. Yapsin proteases are required for the yeast proliferation inmammalian phagocytes (8) but do not counteract the cellular re-sponse in Drosophila. Had this been the case, then yapsin mutantswould have displayed a full virulence phenotype in phagocytosis-impaired flies (55). This rescue of the virulence phenotype hasbeen observed for another C. glabrata mutant we have identified(A. Taravaud, J. Bourdeaux, and D. Ferrandon, unpublished ob-servations). It will be interesting to determine whether the C.glabrata catalase is required for the pathogen to withstand thecellular immune response, because it is currently unknownwhether Drosophila hemocytes produce an oxidative shock to killingested microbes (56). The existence of an oxidative burst canalso be tested on the host side by reducing genetically the ex-pression of the Dual oxidase gene or that of the NADPH oxidasegene in hemocytes.In conclusion, we have developed a model in which C. glabrata

kills its host likely as a result of its unchecked proliferation in thebody cavity. Interestingly, a strain presenting an attenuated viru-lence displays only a limited growth in the hemolymph. This

8 C. GLABRATA PERSISTENCE IN DROSOPHILA

by guest on January 28, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

model is qualitatively distinct from current mice models in whichfungal load is monitored in several organs. The identification ofseveral attenuated strains validates this model for further large-scale screens that can be achieved only in Drosophila. It will alsobe interesting to test mutants affecting the C. glabrata–specificgenes that are absent in other Candida species, as well as potentialvirulence strategies unique to this pathogen. This may include C.glabrata tolerance to high copper concentrations (57), this ionbeing present in the hemolymph because it is required for POactivity. The challenge will then be to determine whether thevirulence factors identified by these approaches will be relevant toan understanding of C. glabrata pathogenesis in patients. Indeed,the Drosophila model is adequate to identify “public” virulencefactors, that is, general virulence factors that are required forpathogenicity independently of the host in opposition to “private”virulence factors, such as thermotolerance for mammalian hosts,which involves the calcineurin pathway in C. glabrata (14). Thus,it remains to be determined whether the HOG pathway belongs tothe “public” or to the Drosophila “private” category of virulencefactors. In support of the former possibility, we note that HOGpathway mutants in Cryptococcus neoformans displayed an at-tenuated virulence in mammals (58).

AcknowledgmentsWe thank Drs. Rupinder Kaur, Brendan P. Cormack, Karl Kuchler, Bernard

Hube, and the late Barbara Winsor for the kind gift of C. glabrata strains.

We are also thankful to Dr. Karl Kuchler for providing primary Abs raised

against whole Candida.

DisclosuresThe authors have no financial conflicts of interest.

References1. Marriott, D. J., E. G. Playford, S. Chen, M. Slavin, Q. Nguyen, D. Ellis, and

T. C. Sorrell; Australian Candidaemia Study. 2009. Determinants of mortality innon-neutropenic ICU patients with candidaemia. Crit. Care 13: R115.

2. Sipsas, N. V., R. E. Lewis, J. Tarrand, R. Hachem, K. V. Rolston, I. I. Raad, andD. P. Kontoyiannis. 2009. Candidemia in patients with hematologic malignanciesin the era of new antifungal agents (2001-2007): stable incidence but changingepidemiology of a still frequently lethal infection. Cancer 115: 4745–4752.

3. Pfaller, M. A., G. J. Moet, S. A. Messer, R. N. Jones, and M. Castanheira. 2011.Geographic variations in species distribution and echinocandin and azole anti-fungal resistance rates among Candida bloodstream infection isolates: reportfrom the SENTRYAntimicrobial Surveillance Program (2008 to 2009). J. Clin.Microbiol. 49: 396–399.

4. Pfaller, M. A., S. A. Messer, G. J. Moet, R. N. Jones, and M. Castanheira. 2011.Candida bloodstream infections: comparison of species distribution and resis-tance to echinocandin and azole antifungal agents in Intensive Care Unit (ICU)and non-ICU settings in the SENTRY Antimicrobial Surveillance Program(2008-2009). Int. J. Antimicrob. Agents 38: 65–69.

5. Pfaller, M. A., S. A. Messer, L. Boyken, C. Rice, S. Tendolkar, R. J. Hollis, andD. J. Diekema. 2003. Caspofungin activity against clinical isolates offluconazole-resistant Candida. J. Clin. Microbiol. 41: 5729–5731.

6. Barns, S. M., D. J. Lane, M. L. Sogin, C. Bibeau, and W. G. Weisburg. 1991.Evolutionary relationships among pathogenic Candida species and relatives. J.Bacteriol. 173: 2250–2255.

7. Gow, N. A., F. L. van de Veerdonk, A. J. Brown, and M. G. Netea. 2012. Candidaalbicans morphogenesis and host defence: discriminating invasion from colo-nization. Nat. Rev. Microbiol. 10: 112–122.

8. Kaur, R., B. Ma, and B. P. Cormack. 2007. A family of glycosylphosphatidylinositol-linked aspartyl proteases is required for virulence of Candida glabrata. Proc. Natl.Acad. Sci. USA 104: 7628–7633.

9. Cormack, B. P., N. Ghori, and S. Falkow. 1999. An adhesin of the yeast pathogenCandida glabratamediating adherence to human epithelial cells. Science 285: 578–582.

10. Seider, K., S. Brunke, L. Schild, N. Jablonowski, D. Wilson, O. Majer, D. Barz,A. Haas, K. Kuchler, M. Schaller, and B. Hube. 2011. The facultative intracel-lular pathogen Candida glabrata subverts macrophage cytokine production andphagolysosome maturation. J. Immunol. 187: 3072–3086.

11. Tsoni, S. V., A. M. Kerrigan, M. J. Marakalala, N. Srinivasan, M. Duffield,P. R. Taylor, M. Botto, C. Steele, and G. D. Brown. 2009. Complement C3 playsan essential role in the control of opportunistic fungal infections. Infect. Immun.77: 3679–3685.

12. Bourgeois, C., O. Majer, I. E. Frohner, I. Lesiak-Markowicz, K. S. Hildering,W. Glaser, S. Stockinger, T. Decker, S. Akira, M. Muller, and K. Kuchler. 2011.

Conventional dendritic cells mount a type I IFN response against Candida spp.requiring novel phagosomal TLR7-mediated IFN-b signaling. J. Immunol. 186:3104–3112.

13. Jawhara, S., E. Mogensen, F. Maggiotto, C. Fradin, A. Sarazin, L. Dubuquoy,E. Maes, Y. Guerardel, G. Janbon, and D. Poulain. 2012. Murine model ofdextran sulfate sodium-induced colitis reveals Candida glabrata virulence andcontribution of b-mannosyltransferases. J. Biol. Chem. 287: 11313–11324.

14. Chen, Y. L., J. H. Konieczka, D. J. Springer, S. E. Bowen, J. Zhang, F. G. Silao,A. A. Bungay, U. G. Bigol, M. G. Nicolas, S. N. Abraham, et al. 2012. Con-vergent evolution of calcineurin pathway roles in thermotolerance and virulencein Candida glabrata. G3 (Bethesda) 2: 675–691.

15. Mylonakis, E., F. M. Ausubel, R. J. Tang, and S. B. Calderwood. 2003. The art ofserendipity: killing of Caenorhabditis elegans by human pathogens as a model ofbacterial and fungal pathogenesis. Expert Rev. Anti Infect. Ther. 1: 167–173.

16. Pukkila-Worley, R., E. Holson, F. Wagner, and E. Mylonakis. 2009. Antifungaldrug discovery through the study of invertebrate model hosts. Curr. Med. Chem.16: 1588–1595.

17. Limmer, S., J. Quintin, C. Hetru, and D. Ferrandon. 2011. Virulence on the fly:Drosophila melanogaster as a model genetic organism to decipher host-pathogeninteractions. Curr. Drug Targets 12: 978–999.

18. Alarco, A. M., A. Marcil, J. Chen, B. Suter, D. Thomas, and M. Whiteway. 2004.Immune-deficient Drosophila melanogaster: a model for the innate immuneresponse to human fungal pathogens. J. Immunol. 172: 5622–5628.

19. Chamilos, G., M. S. Lionakis, R. E. Lewis, J. L. Lopez-Ribot, S. P. Saville,N. D. Albert, G. Halder, and D. P. Kontoyiannis. 2006. Drosophila melanogasteras a facile model for large-scale studies of virulence mechanisms and antifungaldrug efficacy in Candida species. J. Infect. Dis. 193: 1014–1022.

20. Glittenberg, M. T., S. Silas, D. M. MacCallum, N. A. Gow, and P. Ligoxygakis.2011. Wild-type Drosophila melanogaster as an alternative model system forinvestigating the pathogenicity of Candida albicans. Dis. Model. Mech. 4: 504–514.

21. Lemaitre, B., E. Nicolas, L. Michaut, J. M. Reichhart, and J. A. Hoffmann. 1996.The dorsoventral regulatory gene cassette spatzle/Toll/cactus controls the potentantifungal response in Drosophila adults. Cell 86: 973–983.

22. Gottar, M., V. Gobert, A. A. Matskevich, J. M. Reichhart, C. Wang, T. M. Butt,M. Belvin, J. A. Hoffmann, and D. Ferrandon. 2006. Dual detection of fungalinfections in Drosophila via recognition of glucans and sensing of virulencefactors. Cell 127: 1425–1437.

23. Jang, I. H., N. Chosa, S. H. Kim, H. J. Nam, B. Lemaitre, M. Ochiai, Z. Kambris,S. Brun, C. Hashimoto, M. Ashida, et al. 2006. A Spatzle-processing enzymerequired for toll signaling activation in Drosophila innate immunity. Dev. Cell10: 45–55.

24. Mishima, Y., J. Quintin, V. Aimanianda, C. Kellenberger, F. Coste, C. Clavaud,C. Hetru, J. A. Hoffmann, J. P. Latge, D. Ferrandon, and A. Roussel. 2009. TheN-terminal domain of Drosophila Gram-negative binding protein 3 (GNBP3)defines a novel family of fungal pattern recognition receptors. J. Biol. Chem.284: 28687–28697.

25. Ligoxygakis, P., N. Pelte, J. A. Hoffmann, and J. M. Reichhart. 2002. Activationof Drosophila Toll during fungal infection by a blood serine protease. Science297: 114–116.

26. Chosa, N., T. Fukumitsu, K. Fujimoto, and E. Ohnishi. 1997. Activation ofprophenoloxidase A1 by an activating enzyme in Drosophila melanogaster. In-sect Biochem. Mol. Biol. 27: 61–68.

27. De Gregorio, E., S. J. Han, W. J. Lee, M. J. Baek, T. Osaki, S. Kawabata,B. L. Lee, S. Iwanaga, B. Lemaitre, and P. T. Brey. 2002. An immune-responsiveSerpin regulates the melanization cascade in Drosophila. Dev. Cell 3: 581–592.

28. Ligoxygakis, P., N. Pelte, C. Ji, V. Leclerc, B. Duvic, M. Belvin, H. Jiang,J. A. Hoffmann, and J. M. Reichhart. 2002. A serpin mutant links Toll activationto melanization in the host defence of Drosophila. EMBO J. 21: 6330–6337.

29. Tang, H., Z. Kambris, B. Lemaitre, and C. Hashimoto. 2006. Two proteasesdefining a melanization cascade in the immune system of Drosophila. J. Biol.Chem. 281: 28097–28104.

30. Matskevich, A. A., J. Quintin, and D. Ferrandon. 2010. The Drosophila PRRGNBP3 assembles effector complexes involved in antifungal defenses indepen-dently of its Toll-pathway activation function. Eur. J. Immunol. 40: 1244–1254.

31. Rutschmann, S., A. C. Jung, C. Hetru, J.-M. Reichhart, J. A. Hoffmann, andD. Ferrandon. 2000. The Rel protein DIF mediates the antifungal but not theantibacterial host defense in Drosophila. Immunity 12: 569–580.

32. Rutschmann, S., A. Kilinc, and D. Ferrandon. 2002. Cutting edge: the Tollpathway is required for resistance to gram-positive bacterial infections in Dro-sophila. J. Immunol. 168: 1542–1546.

33. Gobert, V., M. Gottar, A. A. Matskevich, S. Rutschmann, J. Royet, M. Belvin,J. A. Hoffmann, and D. Ferrandon. 2003. Dual activation of the Drosophila Tollpathway by two pattern recognition receptors. Science 302: 2126–2130.

34. Gottar, M., V. Gobert, T. Michel, M. Belvin, G. Duyk, J. A. Hoffmann,D. Ferrandon, and J. Royet. 2002. The Drosophila immune response againstGram-negative bacteria is mediated by a peptidoglycan recognition protein.Nature 416: 640–644.

35. Leclerc, V., N. Pelte, L. El Chamy, C. Martinelli, P. Ligoxygakis, J. A. Hoffmann,and J. M. Reichhart. 2006. Prophenoloxidase activation is not required for sur-vival to microbial infections in Drosophila. EMBO Rep. 7: 231–235.

36. Tauszig-Delamasure, S., H. Bilak, M. Capovilla, J. A. Hoffmann, and J. L. Imler.2002. Drosophila MyD88 is required for the response to fungal and Gram-positive bacterial infections. Nat. Immunol. 3: 91–97.

37. Elrod-Erickson, M., S. Mishra, and D. Schneider. 2000. Interactions between thecellular and humoral immune responses in Drosophila. Curr. Biol. 10: 781–784.

The Journal of Immunology 9

by guest on January 28, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

38. Meng, X., B. S. Khanuja, and Y. T. Ip. 1999. Toll receptor-mediated Drosophilaimmune response requires Dif, an NF-kappaB factor. Genes Dev. 13: 792–797.

39. Manfruelli, P., J. M. Reichhart, R. Steward, J. A. Hoffmann, and B. Lemaitre.1999. A mosaic analysis in Drosophila fat body cells of the control of antimi-crobial peptide genes by the Rel proteins Dorsal and DIF. EMBO J. 18: 3380–3391.

40. Rutschmann, S., A. C. Jung, R. Zhou, N. Silverman, J. A. Hoffmann, andD. Ferrandon. 2000. Role of Drosophila IKK g in a toll-independent antibac-terial immune response. Nat. Immunol. 1: 342–347.

41. Yanez, A. M., C. Murciano, S. Llopis, T. Fernandez-Espinar, M. L. Gil, and D.Gozalbo. 2009. In vivo and in vitro studies on virulence and host responses toSaccharomyces cerevisiae clinical and non-clinical isolates. Open Mycol. J. 3:37–47.

42. Lemaitre, B., J. M. Reichhart, and J. A. Hoffmann. 1997. Drosophila host de-fense: differential induction of antimicrobial peptide genes after infection byvarious classes of microorganisms. Proc. Natl. Acad. Sci. USA 94: 14614–14619.

43. Domergue, R., I. Castano, A. De Las Penas, M. Zupancic, V. Lockatell,J. R. Hebel, D. Johnson, and B. P. Cormack. 2005. Nicotinic acid limitationregulates silencing of Candida adhesins during UTI. Science 308: 866–870.

44. Ma, B., S. J. Pan, M. L. Zupancic, and B. P. Cormack. 2007. Assimilation ofNAD(+) precursors in Candida glabrata. Mol. Microbiol. 66: 14–25.

45. Gregori, C., C. Schuller, A. Roetzer, T. Schwarzmuller, G. Ammerer, andK. Kuchler. 2007. The high-osmolarity glycerol response pathway in the humanfungal pathogen Candida glabrata strain ATCC 2001 lacks a signaling branchthat operates in baker’s yeast. Eukaryot. Cell 6: 1635–1645.

46. Apidianakis, Y., L. G. Rahme, J. Heitman, F. M. Ausubel, S. B. Calderwood, andE. Mylonakis. 2004. Challenge of Drosophila melanogaster with Cryptococcusneoformans and role of the innate immune response. Eukaryot. Cell 3: 413–419.

47. Jacobsen, I. D., S. Brunke, K. Seider, T. Schwarzmuller, A. Firon, C. d’Enfert,K. Kuchler, and B. Hube. 2010. Candida glabrata persistence in mice does notdepend on host immunosuppression and is unaffected by fungal amino acidauxotrophy. Infect. Immun. 78: 1066–1077.

48. Fehlbaum, P., P. Bulet, L. Michaut, M. Lagueux, W. F. Broekaert, C. Hetru, andJ. A. Hoffmann. 1994. Insect immunity. Septic injury of Drosophila induces thesynthesis of a potent antifungal peptide with sequence homology to plant anti-fungal peptides. J. Biol. Chem. 269: 33159–33163.

49. Levashina, E. A., S. Ohresser, P. Bulet, J.-M. Reichhart, C. Hetru, and J. A. Hoffmann.1995. Metchnikowin, a novel immune-inducible proline-rich peptide fromDrosophila with antibacterial and antifungal properties. Eur. J. Biochem.233: 694–700.

50. Lamberty, M., S. Ades, S. Uttenweiler-Joseph, G. Brookhart, D. Bushey,J. A. Hoffmann, and P. Bulet. 1999. Insect immunity. Isolation from the lepi-dopteran Heliothis virescens of a novel insect defensin with potent antifungalactivity. J. Biol. Chem. 274: 9320–9326.

51. Lamberty, M., D. Zachary, R. Lanot, C. Bordereau, A. Robert, J. A. Hoffmann,and P. Bulet. 2001. Insect immunity. Constitutive expression of a cysteine-richantifungal and a linear antibacterial peptide in a termite insect. J. Biol. Chem.276: 4085–4092.

52. Ekengren, S., and D. Hultmark. 1999. Drosophila cecropin as an antifungalagent. Insect Biochem. Mol. Biol. 29: 965–972.

53. Dolmetsch, R. E., R. S. Lewis, C. C. Goodnow, and J. I. Healy. 1997. Differentialactivation of transcription factors induced by Ca2+ response amplitude andduration. Nature 386: 855–858.

54. Shia, A. K., M. Glittenberg, G. Thompson, A. N. Weber, J. M. Reichhart, andP. Ligoxygakis. 2009. Toll-dependent antimicrobial responses in Drosophilalarval fat body require Spatzle secreted by haemocytes. J. Cell Sci. 122: 4505–4515.

55. Limmer, S., S. Haller, E. Drenkard, J. Lee, S. Yu, C. Kocks, F. M. Ausubel, andD. Ferrandon. 2011. Pseudomonas aeruginosa RhlR is required to neutralize thecellular immune response in a Drosophila melanogaster oral infection model.Proc. Natl. Acad. Sci. USA 108: 17378–17383.

56. Cuellar-Cruz, M., M. Briones-Martin-del-Campo, I. Canas-Villamar, J. Montalvo-Arredondo, L. Riego-Ruiz, I. Castano, and A. De Las Penas. 2008. High resistanceto oxidative stress in the fungal pathogen Candida glabrata is mediated by a singlecatalase, Cta1p, and is controlled by the transcription factors Yap1p, Skn7p,Msn2p, and Msn4p. Eukaryot. Cell 7: 814–825.

57. Zhou, P., and D. J. Thiele. 1993. Rapid transcriptional autoregulation of a yeastmetalloregulatory transcription factor is essential for high-level copper detoxi-fication. Genes Dev. 7: 1824–1835.

58. Bahn, Y. S., K. Kojima, G. M. Cox, and J. Heitman. 2005. Specialization of theHOG pathway and its impact on differentiation and virulence of Cryptococcusneoformans. Mol. Biol. Cell 16: 2285–2300.

10 C. GLABRATA PERSISTENCE IN DROSOPHILA

by guest on January 28, 2018http://w

ww

.jimm

unol.org/D

ownloaded from