Vol. 47, No. 4APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Apr. 1984, p. 647-6520099-2240/84/040647-06$02.00/0Copyright C) 1984, American Society for Microbiology

Colonization of Congenitally Athymic, Gnotobiotic Mice by Candidaalbicans

EDWARD BALISH,* MARION J. BALISH, CINDY A. SALKOWSKI, KENNETH W. LEE,t AND KENNETH F.BARTIZAL

Departments of Surgery and Medical Microbiology, University of Wisconsin Medical School, Madison, Wisconsin 53706

Received 19 September 1983/Accepted 19 January 1984

Colony counts, scanning electron microscopy, and light microscopy were used to assess the capacity ofCandida albicans to colonize (naturally) and infect the alimentary tract of adult and neonatal (athymic[nu/nu] or heterozygous [+/nu] littermates) germfree BALB/c mice. When exposed to yeast-phase C.albicans, the alimentary tract of adult germfree mice (nu/nu or +/nu) is quickly (within 24 to 48 h) colonizedwith yeast cells. Neither morbidity nor mortality was evident in any mice that were colonized with a pure

culture of C. albicans for 6 months. Yeast cells of C. albicans predominated on mucosal surfaces in the oralcavities and vaginas of adult athymic and heterozygous mice. In both genotypes, C. albicans hyphae were

observed in keratinized tissue on the dorsal posterior tongue surface and in the cardial-atrium section of thestomach. Conversely, neonatal athymic or heterozygous mice, born to germfree or C. albicans-colonizedmothers, do not become heavily colonized or infected with C. albicans until 11 to 15 days after birth.Although yeast cells adhered to some mucosal surfaces in vivo, neither widespread mucocutaneouscandidiasis, i.e., invasion of mucosal surfaces with C. albicans hyphae, nor overwhelming systemiccandidiasis was evident in neonatal (nu/nu or +/nu) mice. Thus, even in the absence of functional T-cellsand a viable bacterial flora, athymic and heterozygous littermate mice (adult or neonatal BALB/c) that are

colonized with a pure culture of C. albicans manifest resistance to extensive mucocutaneous and systemiccandidiasis.

Candida albicans is a common component of the mucosalmicroflora of many neonatal and adult warm-blooded mam-mals (25). Under natural conditions, neonates acquire C.albicans from their mothers early in life, and the yeastusually persists as a minor and innocuous component of themucosal flora thereafter. Apparently, the normal bacterialflora inhibits the growth of C. albicans in the alimentarytract since oral therapy with antibacterial antibiotics oftenresults in dramatic increases in the number of viable C.albicans on mucosal surfaces (1, 7, 10, 32).The notoriety of C. albicans stems from the fact that it can

cause mucosal, cutaneous, or even life-threatening systemicdiseases in many patients who are immunologically compro-mised through disease or as a consequence of immunosup-pressive and cytotoxic therapy (8, 15, 22, 30). Mucocutane-ous disease is the most prevalent form of candidiasis inhumans (8, 15, 25). Systemic candidiasis is also thought tooriginate from endogenous populations of C. albicans in thealimentary tract (16, 22). Unfortunately, very little is knownabout host Candida interactions at mucosal surfaces, immu-nity to mucosal candidiasis, or the factors that trigger C.albicans to switch from a commensal to an invasive patho-gen.One of the reasons for the paucity of information on

immunity to mucosal candidiasis is the lack of a good animalmodel of the disease. Most conventional (commercial) miceused in research are the progeny of rodents that wereoriginally obtained by cesarean section, raised in barrierfacilities, and colonized with a poorly defined number oflargely unspeciated bacteria. Since they are purchased either

* Corresponding author.t Present address: Department of Botany and Microbiology,

Arizona State University Tempe, AZ 85281.

specific pathogen free or with a defined flora of bacteria,they usually do not harbor C. albicans on their mucosalsurfaces. It is possible to establish C. albicans in thealimentary tract of conventional neonatal mice by gavagingthem with large doses of viable yeast cells (107 to 108/0.5 ml)after a 6- to 8-h starvation period (9, 10, 13, 27, 28). About 50to 60% of the gavaged pups succumb to the latter traumaticprocedure, but the alimentary tracts of survivors are colo-nized with a low number (103 to 104/g) of C. albicans for atleast 3 to 10 weeks. Recently, Hector and Domer (11)reported that they were able to establish C. albicans in thealimentary tract of conventional neonatal mice by inoculat-ing the mammary glands of the lactating mothers with viableblastospores and allowing the infants to suckle. It is worthnoting that all conventional animal models of mucosal candi-diasis will very likely, as in patients with mucosal candidiasis(3, 23, 24), be contaminated with bacteria.

Germfree (GF) animals are uniquely suited for studies ofmucosal candidiasis. Several investigators have demonstrat-ed that the alimentary tract of adult GF animals is quicklyand permanently colonized after they are inoculated (orally)with a pure culture of C. albicans 1, 2, 4, 12, 14, 17-20, 24,26, 33). Although low numbers (<10/g of tissue) of C.albicans can be isolated from the internal organs of these"C. albicans-monoassociated" rodents (4, 12, 14, 17-20, 24,26, 33), neither morbidity nor extensive mucocutaneous orsystemic disease has been observed in adult nu/nu, +/nu, oreuthymic (+/+) rodents (4, 12, 14, 17-20, 24, 26, 33).No one has assessed the capacity of C. albicans to adhere,

colonize, or infect the skin and mucosal surfaces of GFneonatal (nu/nu, +/nu, or +/+) mice. Children with Di-George syndrome or other defects in thymus-dependent T-cell-mediated immunity are particularly predisposed to de-veloping mucocutaneous candidiasis (15). Conversely,congenitally athymic mice are more resistant to systemic

647

on February 28, 2020 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

candidiasis than their heterozygous littermates (5, 29). Datafrom children with mucocutaneous candidiasis have primari-ly been responsible for the hypothesis that thymus-depen-dent, cell-mediated immunity is essential for human resist-ance to mucosal candidiasis (15, 25).The purpose of this study was to assess the capacity of C.

albicans to colonize mucosal surfaces and skin of adult or

neonatal (nu/nu or +/nu) mice that were born either GF or toC. albicans-colonized (pure culture) mothers and to assess

the capacity of C. albicans to cause mucosal, cutaneous, or

systemic disease under these pure culture conditions in vivo.

MATERIALS AND METHODSMicroorganism. C. albicans B311 (type A) was maintained

in our laboratory by monthly transfers on Sabouraud'sglucose agar slants. Yeast cells were cultured for 24 h at 37°Con Sabouraud's glucose agar and then stored at 4°C. Forexperiments, a single stock was prepared with organismsthat were grown on Sabouraud's glucose agar for 24 h at370C, washed from the slants, and stored at -70°C in 1 to 2ml of Sabouraud's glucose broth at a concentration of 108viable units per ml. The intravenously administered 50%lethal dose of this C. albicans strain is 2 x 104 for flora-defined nude mice and 1.6 x 104 for their heterozygouslittermates.

Colonization of mice with C. albicans. Adult GF mice were

orally inoculated by mixing C. albicans in their drinkingwater at a concentration of approximately 105 viable C.albicans per ml. The mice were allowed access to the C.albicans-contaminated water for 2 to 4 h. Fresh sterile waterbottles were used thereafter. Newborn mice were colonizedby natural contact with' their C. albicans-monoassociatedmothers. Pups born to GF mice were transferred (along withthe GF mother) as soon as they were born or at various timeintervals after birth into flexible film isolators that containedC. albicans-monoassociated mice. Under these condtions,the adult GF mice are quickly (24 h) colonized with C.albicans.

Mice. Congenitally athymic (nude) mice and their thymus-bearing littermates were produced by mating homozygous(nu/nu) males with heterozygous (+/nu) females. All micewere of BALB/c background.GF mice were originally derived from NIH BALB/c nude

mice by cesarean derivation and have since been bred andhoused in flexible film isolators at the University of Wiscon-sin Gnotobiotic Research. Laboratory. The GF or monoasso-ciated status of each experiment was assessed by methodspreviously described (17).Enumeration of viable C. albicans. Mice were killed by

cervical dislocation and dissected immediately. The appro-

priate organs were removed and placed in tissue homogeniz-ers containing 5 ml of sterile Sabouraud's glucose broth.Each organ was homogenized separately, and dilutions were

made of each homogenate. The number of viable units in thehomogenates was determined by plating on Sabouraud'sglucose agar. Colonies were counted after 48 h of incubationat 350C. The viable units are expressed as the number ofCFU per gram (dry weight) of organ.

Because of their small size, it was sometimes necessary toculture the tissues and contents of the whole intestinal tractof neonatal mice. The tissues and areas of the alimentarytract cultured in neonates are detailed in the appropriatetables of data.

Histology. Tissues from C. albicans-monoassociated micewere excised and immediately placed into Hollande-Bouin's

fixative. The tissue was embedded in paraffin, sectioned,and stained with periodic acid-Schiff, azure A-eosin B, orhematoxylin and eosin stain.SEM. Procedures for the preparation and fixing of muco-

sal tissues for scanning electron microscopy (SEM) havebeen described previously (12, 19). Briefly, segments ofmucosal surfaces are quickly removed from the killed animaland placed in vials that contain cold (4°C) 0.1 M cacodylate-buffered glutaraldehyde (2.5%). The glutaraldehyde-fixedsamples are then rinsed in 0.1 M cacodylate buffer, postfixedin osmium tetroxide, again rinsed in cacodylate buffer, anddehydrated in water-ethanol or ethanol-amyl acetate. Tis-sues are then critical point dried, fixed to aluminum stubs,coated with gold-palladium, and examined under a JEM U3or a JEOL 50A scanning electron microscope at 15 to 20 kV.

RESULTSAdult GF athymic and GF thymus-bearing mice quickly

(within 24 to 48 h) become colonized with large numbers (107to 108/g of intestinal contents) of C. albicans. The alimentarytracts of these adult heterozygous (+/nu) or athymic (nu/nu)mice contain consistently large numbers of viable C. albi-cans (-3 x 107/g of cecal contents) for at least 6 months aftercolonization. No viable C. albicans was recovered from thesystemic organs of adult, C. albicans-monoassociated, nude,or heterozygous mice during the first 3 days after oralchallenge. Thereafter, the number of viable C. albicansrecovered from systemic organs was always at the lowerthreshold of detection by our assay (<10 organisms per g).Thus, C. albicans can readily colonize the alimentary tractof adult athymic or heterozygous (+/nu) GF mice. Althoughsmall numbers of viable C. albicans can be isolated fromtheir internal organs, the C. albicans-monoassociated micedid not manifest any obvious morbidity or mortality duringthis 6-month study. Although viable C. albicans was presentin large numbers in the alimentary tract of adult nu/nu or+/nu mice, only minimal mucosal infection (i.e., hyphalinvasion of mucosal surfaces) was found on the keratinizedportion of the cardial-atrium section of the stomach (Fig. 1)and the posterior portion of the dorsal surface of thetongues.SEM of various mucosal surfaces of C. albicans-monoas-

sociated mice revealed yeast cells adherent to the followingepithelial surfaces of adult mice that were colonized for atleast 14 days: ventral surface of the tongue (Fig. 2), cheek(Fig. 3), and the vagina (Fig. 4). Interestingly, the esophagus(a common site of chronic infection in humans), smallintestine (duodenum, jejunum, ileum), cecum, and colon hadvery few yeast cells associated with their mucosal surfaces.Thus, in congenitally athymic or heterozygous littermateBALB/c mice, C. albicans adhered primarily to epithelialsurfaces in the oral cavity, stomach, and vagina. C. albicanscould not be detected in the gingival crevices, on toothsurfaces, or on mucosal surfaces in close proximity to theteeth. In the small and large intestines, C. albicans wasobserved primarily in the intestinal contents. Only occasion-ally were yeast cells seen entrapped in the mucous of theintestinal tract.

Histology of skin from monoassociated athymic or euthy-mic mice indicated that C. albicans did not cause a chronicinfection (yeast or hyphae) in the stratum corneum. Al-though C. albicans could be cultured from swabs of skinsurfaces (abdomen, back, etc.), this probably representedsurface contamination and not a true colonization or infec-tion of murine skin.

648 BALISH ET AL.

on February 28, 2020 by guest

http://aem.asm

.org/D

ownloaded from

MUCOSAL AND SYSTEMIC CANDIDIASIS 649

FIG. 1. Hyphae in cardial-atrium section of stomach from anadult C. albicans-monoassociated mouse. Bar. 10 p.m.

Natural colonization of neonates born to C. albicans-monoassociated heterozygous mothers. C. albicans did notcolonize the alimentary tract of neonatal athymic or hetero-zygous littermate mice that were born to C. albicans-monoassociated heterozygous (+/nu) mothers as quickly asit colonized the alimentary tract of adults. Data in Table 1

FIG. 3. Yeast cells adhering to the cheek mucosal surface of a C.albicans-monoassociated mouse. Bar. 100 ,um.

demonstrate that the stomachs of neonatal (nu/nu or +/nu)mice had relatively low numbers of viable C. albicans for thefirst 10 days after birth. After day 10, the numbers of viableC. albicans in the stomach and the rest of the intestinal tractincreased and were comparable to the viable counts ob-served within 24 to 48 h of exposing adult (nu/nu or +/nu)

FIG. 4. Yeast cells adhering to the vaginal surface of a C.albicans-monoassociated mouse. Bar, 10 ,um.

FIG. 2. Yeast cells adhering to the ventral tongue surface of a C.albicans-monoassociated mouse. Bar, 10 ,um.

VOL. 47, 1984

on February 28, 2020 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

TABLE 1. Natural colonization of neonatal mice with C. albicansfrom their monoassociated mothers

Mean log1oNo.of BALB/c ~~~~(range) of no.

Age (days) Nmice genotype of C. albicansmice genotype ~~~~perg of

stomacha

0-5 9 Nude (nu/nu) 2.4 (0-4.2)13 Heta (+/nu) 1.6 (0-4.0)

6-10 12 Nude (nu/nu) 2.4 (0-4.7)S Het (+/nu) 3.0 (2.0-4.3)

11-15 10 Nude (nu/nu) 5.1 (0-6.0)7 Het (+/nu) 4.5 (3.2-5.8)

16-21 9 Nude (nu/nu) 6.1 (0-7.3)13 Het (+/nu) 7.0 (5.8-8.0)

>30 6 Nude (nu/nu) 7.5 (7.0-8.0)19 Het (+Inu) 6.7 (5.2-8.0)

a Het, Heterozygous.

mice to other C. albicans-monoassociated mice. The livers,kidneys, and spleens of these athymic and heterozygousmice contained very few viable C. albicans (<10/g) fromdays 0 to 30 after birth (Table 2). Thus, for the first 10 daysafter birth, the alimentary tract of neonatal (nu/nu or +/nu)mice does not become heavily colonized (or infected) withC. albicans. SEM of the stomach surface (the cardial-atriumsection) revealed very few yeast cells and no hyphal invasionof tissues by C. albicans for the first 10 days after birth.However, after 11 days of age, SEM revealed the presenceof yeast cells and some invasive hyphae in the cardial-atriumsection of the stomachs of nu/nu and +/nu mice.

Colonization of neonatal GF mice with C. albicans. Toassess whether prior colonization of heterozygous motherswith C. albicans was responsible for inhibiting the capacityof C. albicans to colonize the alimentary tract of their pups,litters of GF (nu/nu or +/nu) mice and their heterozygous GFmother were introduced into an isolator that contained C.albicans-monoassociated mice. The GF mother was thenexposed to C. albicans by putting a C. albicans-monoasso-ciated mouse into the cages of the newly introduced GFmice. The adult mouse was colonized with C. albicanswithin 24 to 48 h. The GF litters introduced into the C.albicans-monoassociated isolators were of different ages (0to 5, 6 to 10, 11 to 15, or 16 to 21 days of age). Three athymicand three heterozygous pups were removed from the isolatorat 3, 10, and 21 days after they were introduced into theisolator and naturally exposed to C. albicans. The mice werekilled immediately after removal from the isolator, and thenumber of viable C. albicans in their alimentary tract wasassessed. Salient features of this experiment (Table 3) wereas follows: (i) older GF pups (i.e., 11 to 21 days of age whenexposed to viable C. albicans) were quickly (within 3 days)colonized with large numbers of viable C. albicans; (ii)younger GF pups (i.e., 0 to 10 days of age when exposed toviable C. albicans) had fewer viable C. albicans in theiralimentary tract (especially in the stomach) than older pups

(11 to 21 days); (iii) although low numbers of C. albicanswere occasionally recovered from the internal organs of thepups, no morbidity or mortality was evident in the athymicor heterozygous C. albicans-monoassociated mice; and (iv)the only invasive hyphae observed by light microscopy and

SEM were in the stomachs of pups that were more than 10days of age.

Induced mucosal candidiasis. When adult C. albicans-monoassociated mice (nu/nu or +/nu) were injected withcyclophosphamide (150 mg/kg intraperitoneally), extensiveinfection (hyphal penetration) of their mucosal surfaces wasevident within 5 to 7 days. Thus the C. albicans strain usedin these studies is capable of causing mucosal candidiasis.Extensive hyphal penetration was observed on the tongue,cheeks, and stomach of the cyclophosphamide-treated mice.

Systemic immunity. Euthymic (+/+) or heterozygous(+/nu) conventional mice are more susceptible to systemic(renal) candidiasis than conventional athymic (nu/nu) mice(5, 29). It has also been demonstrated that GF nude orheterozygous BALB/c mice both manifest resistance to asystemic challenge with C. albicans (17). In this study, wesubjected mice (nu/nu or +/nu) that were monoassociatedwith C. albicans for 4 to 6 weeks to an intravenous challengewith 104 viable C. albicans to assess their resistance tosystemic (renal) candidiasis. The C. albicans-monoassociat-ed mice (nu/nu or +/nu) were capable of clearing a systemicchallenge (Table 4). The heterozygous (+/nu) mice were ableto clear the infectious challenge by day 17 after intravenouschallenge. The athymic mice had low populations of C.albicans in the kidneys on day 33 but still showed asubstantial capacity to decrease the number of viable C.albicans in their kidneys. The kidneys of GF mice (nu/nu or+/nu) challenged intravenously with a similar number ofviable C. albicans remain chronically infected (102 to 103)viable C. albicans per g of tissue) for at least 45 days afterinoculation. Thus, prior colonization with C. albicans appar-ently enhanced the resistance of +/nu, but not nu/nu, mice toa systemic challenge with C. albicans.

DISCUSSIONThese studies demonstrate several important facets of

murine resistance to mucosal and systemic candidiasis.Functional T-cells are thought to play an important role inacquired resistance to human candidiasis. It is alreadyknown that congenitally athymic mice are more resistant torenal candidiasis than their heterozygous littermates (5, 29).In this study, athymic and littermate (+/nu) mice, eventhough heavily colonized with C. albicans on many mucosal

TABLE 2. Systemic candidiasis in C. albicans-monoassociatedneonatal mice

Mean log1o of no. of C. albicansAge (days) BALB/c No. of per g of:genotype mice

Kidney Spleen Liver

0-5 Nude 11 0 0 0.3Heta

6-10 Nude 20 0.4 0.6 0.5Het

11-15 Nude 13 0 0 0Het

16-21 Nude 13 0.3 0.3 0.1Het 11 0.1 0.4 0

>30 Nude 6 0.6 0 0.9Het 6 0.4 0 0.4

a Het, Heterozygous.

650 BALISH ET AL.

on February 28, 2020 by guest

http://aem.asm

.org/D

ownloaded from

MUCOSAL AND SYSTEMIC CANDIDIASIS 651

surfaces (oral cavities, alimentary tracts, and vaginas), donot manifest any extensive mucosal, cutaneous, or systemicinfections with C. albicans. Some hyphal penetration was

observed in the cardial-atrium section of the stomach and inkeratinized dorsal tongue surfaces of athymic mice andthymus-bearing littermates, but it was minimal. Thus, thy-mus-matured T-cells do not appear to be necessary formurine resistance to either systemic, cutaneous, or mucosalcandidiasis. Low numbers of C. albicans could be isolatedfrom the skin and from the internal organs of monoassociat-ed athymic and heterozygous littermate mice; however, no

obvious cutaneous or systemic disease was observed in our

studies, which were carried out over a period of 6 months.Obviously, factors other than thymus-matured T-cells are

able to maintain the resistance of these congenitally athymicmice to mucocutaneous and systemic candidiasis.

C. albicans-monoassociated mice (nu/nu or +/nu) are ableto decrease the number of viable C. albicans that invadetheir kidneys after an intravenous challenge. The C. albi-cans-monoassociated +/nu mice were able to clear an intra-venous challenge with C. albicans faster than the C. albi-cans-monoassociated athymic mice, indicating that the C.albicans monoassociation of thymus-bearing mice enhancedtheir resistance to systemic C. albicans infections. Con-versely, prior colonization with C. albicans did not appear toenhance the resistance of athymic mice to renal candidiasis.

It was also obvious from these studies that neonatal mice,whether born GF or to C. albicans-monoassociated mothers,were not colonized (alimentary tract) as quickly with C.albicans as were adult mice. Since C. albicans is easilyisolated from the skin of adult C. albicans-monoassociatedmice, it is very likely that the neonatal mice were continu-ously exposed to viable C. albicans. Heavy alimentary tractcolonization occurred within 24 to 72 h in mice that were 10days of age and older and appeared to coincide with the timeperiod when the pups were more actively consuming labora-tory chow. Since athymic pups are born to heterozygous(i.e., with a thymus) mothers, we thought that the transmis-sion of cellular or humoral immune factors in the colostrumand milk of the +/nu mothers might account for the apparent

TABLE 3. Susceptibility of young GF mice (athymic or

heterozygous) to systemic candidiasis after monoassociation withC. albicans

Mean log10 of no. of C. albicans per g of tissue'Age at Days Small

exposure after Kidney Liver Spleen Stomach intestine(days) exposure

e N H N H N H N H N H

0-5 3 0 0 0 0 NDb 0 0 3.5 2.8 4.210 0 0.3 0.5 1.3 0 0 6.2 4.7 6.7 6.5

6-10 3 0 0 0 0 0 0 0 0 3.2 3.710 0 0.3 1.5 0.3 0 0.3 4.0 5.8 6.9 6.2

11-15 3 0 0 0 0.2 0.5 1.2 2.5 5.5 5.0 6.810 0.3 0 0 0 0 0.2 5.5 7.3 6.4 7.521 OC 0' 0' 0' 0' 0' 5.0 4.5 6.1 5.4

16-21 3 0 0 0 0 0 0 5.2 5.5 7.3 7.010 0 0 0.3 0 0 0 6.9 7.9 7.3 7.421 0.3 0.1 0.1 0 0.3 0.4 6.1 6.1 7.3 7.3

a Three to five mice per group. N, Athymic; H, heterozygous.b ND, Not done.

' Values represent pooled data from nude and heterozygous mice.

TABLE 4. Resistance of C. albicans-monoassociated mice to asystemic challenge with C. albicans

Time Log 10 of no. of C. albicans per g (dryafter wt) of kidney homogenateaintra- Heterozy-venous Athymic gous

challenge (nu/nu) (+/nu)

4 h 3.6 0.4 3.1 ± 0.13 days 2.7 1.4 2.3 ± 1.55 days 2.5 ± 0.5 2.7 ± 1.510 days 6.2 0.4 1.1 ± 1.117 days 3.1 + 2.1 NDb33 days 1.3 ± 0.7 NDb

a Each data point represents the mean ± standard error of themean of three mice ranging in age from 7 to 10 weeks.

b No viable C. albicans detected at 1/10 dilution of kidneyhomogenate.

resistance of their pups to C. albicans. We tried to have C.albicans-monoassociated athymic females deliver and raisepups; however, congenitally athymic mothers are poor lacta-tors and are unable to take care of their pups for more than acouple of days. We were unable to detect (by culture andSEM) C. albicans in the alimentary tract of athymic pups (24to 48 h old) that were born to C. albicans-monoassociatedathymic females. We are still not certain as to why C.albicans does not quickly colonize the alimentary tract ofgnotobiotic pups in large numbers since other investigatorsare apparently able to colonize pups (5 days of age) born toconventional female mice (10, 11, 27, 28).

This delayed colonization in gnotobiotic neonates is notunusual since others have demonstrated that pups born tognotobiotic mothers that are monoassociated with Clostridi-um botulinum (2), Listeria monocytogenes (6), or Roseburiacecicola (31) also appear to be slowly colonized with bacte-ria.

It is commonly believed that the microbial flora in conven-tional animals plays an important role in protecting the hostfrom mucosal candidiasis (8, 12, 25, 30). When conventionalmammals (with complex flora) are treated with oral, broad-spectrum antibiotics, C. albicans increases in numbers andcan invade the mucosal tissues in the alimentary tract and inthe vagina and at times can even cause systemic disease (15).In GF mice (nu/nu or +/nu) which have had no priorexperience with viable microbes (bacteria, fungi, etc.), C.albicans is unable to cause any extensive invasion of theskin, mucosal tissues, or systemic organs even though largepopulations of yeast cells readily adhere to many of theirmucosal surfaces. The C. albicans strain we used in thisstudy is capable of invading mucosal tissue, since treatmentwith cyclophosphamide (150 mg/kg of body weight, intra-peritoneally) caused a massive increase in mucosal candidia-sis (hyphal penetration of mucosal surfaces).

This gnotobiotic model has many advantages for studieson immunity to mucosal candidiasis over conventional ani-mal models. There are no viable bacteria present to interferewith the growth of C. albicans on mucosal surfaces or tocontaminate experimentally induced mucosal or systemiccandidiasis. The adult GF alimentary tracts and vaginas ofrodents are easily and quickly colonized with C. albicans(yeast phase) without any associated lethality, and neonatalmice undergo a slow alimentary tract colonization that doesnot result in any overt systemic or mucosal disease. Thus, C.albicans-colonized mice (nu/nu or +/nu) represent a unique

VOL. 47, 1984

on February 28, 2020 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

animal model that can be manipulated to facilitate pureculture studies on a number of factors (e.g., nutrition,endocrine function, immunosuppression, etc.) that play arole in the resistance or susceptibility to either mucosal orsystemic candidiasis.

ACKNOWLEDGMENTSThis work was supported by Public Health Service grant AI-14537

from the National Institutes of Health.We acknowledge the secretarial expertise of Donna Brackett.

LITERATURE CITED1. Balish, E., and A. W. Phillips. 1966. Growth, morphogenesis,

and virulence of Candida albicans after oral inoculation in thegermfree and conventional chick. J. Bacteriol. 91:1736-1743.

2. Balish, E., and A. W. Phillips. 1966. Growth and virulence ofCandida albicans after oral inoculation in the chick with amonoflora of either Escherichia coli or Streptococcus faecalis.J. Bacteriol. 91:1744-1749.

3. Chappler, R. R., H. I. Maibach, and R. Aly. 1978. Mucocutane-ous candidiasis or mucocutaneous microbiosis. J. Am. Med.Assoc. 239:428-429.

4. Clark, J. D. 1976. Influence of antibiotics or certain intestinalbacteria on orally administered Candida albicans in germ-freeand conventional mice. Infect. Immun. 4:731-736.

5. Cutler, J. G. 1976. Acute systemic candidiasis in normal andcongenitally thymus-deficient (nude) mice. J. Reticuloendothel.Soc. 19:121-124.

6. Czuprynski, C. J., and E. Balish. 1981. Pathogenesis of Listeriamonocytogenes for gnotobiotic rats. Infect. Immun. 32:323-331.

7. DeMaria, A., H. Buckley, and F. Von Lichtenberg. 1976. Gastro-intestinal candidiasis in rats treated with antibiotics, cortisone,and azathioprine. Infect. Immun. 13:1761-1770.

8. Drake, T. E., and H. I. Maibach. 1973. Candida and candidiasis.II. Clinical manifestations and therapy of candidal disease.Postgrad. Med. 53:121-127.

9. Field, L. H., L. M. Pope, G. T. Cole, and M. N. Guentzel. 1981.Persistence and spread of Candidta ailbiccans after intragastricinoculation of infant mice. Infect. Immun. 31:783-791.

10. Guentzel, M. N., and C. Herrera. 1982. Effects of compromisingagents on candidosis in mice with persistent infections initiatedin infancy. Infect. Immun. 35:222-228.

11. Hector, R. F., and J. E. Domer. 1982. Mammary gland contami-nation as a means of establishing long-term gastrointestinalcolonization of infant mice with Candida albicans. Infect.Immun. 38:788-790.

12. Helstrom, P. B., and E. Balish. 1979. Effect of oral tetracycline,the microbial flora, and the athymic state on gastrointestinalcolonization and infection of BALB/c mice with Candida albi-cans. Infect. Immun. 23:764-774.

13. Herrera, C., and N. Guentzel. 1982. Mice with persistentgastrointestinal Candida albicans as a model for antifungaltherapy. Antimicrob. Agents Chemother. 21:51-53.

14. Hummel, R. P., B. A. Bestreither, M. P. Maley, and B. G.MacMillan. 1973. Inhibition of Candida albicans by Escherichiacoli in 4itro and in the germfree mouse. J. Surg. Res. 15:53-59.

15. Kirkpatrick, C. H., R. R. Rick, and J. E. Bennett. 1971. Chronicmucocutaneous candidiasis. Model-building in cellular immuni-

ty. Ann. Intern. Med. 74:955-978.16. Krause, W., M. Matheis, and K. Wolf. 1981. Fungemia and

funguria after oral administration of Candida albicans. Lanceti:598-599.

17. Lee, K. W., and E. Balish. 1981. Resistance of germfree mice tosystemic candidosis. J. Reticuloendothel. Soc. 29(3):241-248.

18. Lee, K. W., and E. Balish. 1982. Effect of T-cells and intestinalbacteria on resistance of mice to candidosis. J. Reticuloen-dothel. Soc. 31:233-240.

19. Lee, K. W., N. J. Balish, and E. Balish. 1981. Resistance ofgermfree nude and thymus-bearing mice to systemic and gastro-intestinal candidosis, p. 424-432. In S. Sasaki (ed.), Recentadvances in germfree research. Tokai University Press, Tokyo,Japan.

20. Liljemark, W. F., and R. J. Gibbons. 1973. Supression ofCandida albicans by human oral streptococci in gnotobioticmice. Infect. Immun. 8:846-849.

21. Moberg, L. J., and H. Sugiyama. 1979. Microbial ecologicalbasis of infant botulism as studied with germfree mice. Infect.Immun. 25:653-657.

22. Munier-Carpentier, F., F. E. Kiehn, and D. Armstrong. 1981.Fungemia in the immunocompromised host. Am. J. Med.71:363-370.

23. Nilsson, E., and C. Henning. 1977. The bacteriological flora incandidosis of the skin. Curr. Therapeutic Res. 22:27-32.

24. Nishikawa, T., H. Hotane, N. Ohnishi, S. Sasaki, and T. Nomu-ra. 1969. Establishment of Candida albiccans in the alimentarytract of the germfree mice and antagonism with Escherichia coliafter oral inoculation. Jpn. J. Microbiol. 13:263-276.

25. Odds, F. C. 1979. Ecology and epidemiology of Candida, p. 50-74. In Candida and candidosis. University Park Press, Balti-more, Md.

26. Phillips, A. W., and E. Balish. 1966. Growth and invasiveness ofCandida albicans in the germfree and conventional mouse afteroral challenge. Appl. Microbiol. 14:737-741.

27. Pope, L. M., and G. T. Cole. 1981. SEM studies of adherence ofCandida albicans to the gastrointestinal tract of infant mice.Scanning Electron Microsc. 3:73-80.

28. Pope, L. M., G. T. Cole, M. N. Guentzel, and L. J. Berry. 1979.Systemic and gastrointestinal candidiasis of infant mice afterintragastric challenge. Infect. Immun. 25:702-707.

29. Rogers, T. J., E. Balish, and D. D. Manning. 1976. The role ofthymus-dependent cell-mediated immunity in resistance to ex-perimental disseminated candidiasis. J. Reticuloendothel. Soc.20:291-295.

30. Singer, C. M., M. K. Kaplan, and D. Arnmstrong. 1977. Bacte-remia and fungemia complicating neoplastic disease. Am. J.Med. 62:731-741.

31. Stanton, T. B., and D. Savage. 1983. Colonization of gnotobioticmice by Roseburia cecicola, a motile, obligality anaerobicbacterium from murine ceca. Appl. Environ. Microbiol.45:1677-1684.

32. Turner, J. R., T. F. Butler, M. E. Johnson, and R. S. Gordee.1976. Colonization of the intestinal tract of conventional micewith Candida albicans and treatment with antifungal agents.Antimicrob. Agents Chemother. 9:787-792.

33. Wagner, M., and K. K. Srivastava. 1975. Decontamination ofgnotobiotic mice experimentally monoassociated with Candidaalbicans. Infect. Immun. 12:1401-1404.

34. Wilborn, W. H., and L. F. Montes. 1980. Scanning electronmicroscopy of oral lesions in chronic mucocutaneous candidia-sis. J. Am. Med. Assoc. 244:2294-2297.

652 BALISH ET AL.

on February 28, 2020 by guest

http://aem.asm

.org/D

ownloaded from