INFECTION AND IMMUNITY, Dec 1982, p. 1020-10280019-9567/82/121020-09$02.00/0Copyright © 1982, American Society for Microbiology

Vol. 38, No. 3

Immune Responses to Candida albicans in GeneticallyDistinct Mice

RICHARD F. HECTOR,* JUDITH E. DOMER, AND EMILY W. CARROWDepartment of Microbiology and Immunology, Tulane University School of Medicine, New Orleans,

Louisiana 70112

Received 5 August 1982/Accepted 3 September 1982

Mice from six genetically distinct strains were examined for their immuneresponses to Candida albicans in in vitro and in vivo assays, and naive mice andmice immunized with the fungus were challenged intravenously with threedifferent doses of C. albicans to determine differences in susceptibility. Naivemice from the six groups showed substantial differences in resistance to challengebased on mortalities and quantitative cultures of kidneys, with mice from strainsC57BL/6J and BALB/cByJ showing the most resistance; mice from strains A/J,C3H/HeJ, and CBA/J showing moderate susceptibility; and mice from strainDBA/2J showing the highest degree of susceptibility to challenge. Unimmunizedmice from strains C57BL/6J and BALB/cByJ did not produce detectable levels ofCandida-specific antibody by the end of the 28-day observation period whenchallenged intravenously, but the other strains did. Immunized mice showed adegree of protection to challenge, with all groups except mice from strainBALB/cByJ showing a reduction of two to three log units in the level ofcolonization in their kidneys and all strains producing significant levels ofantibody. Additionally, the immunized mice of all strains developed substantiallevels of delayed-type hypersensitivity and demonstrated nearly identical lympho-cyte proliferative responses to Candida antigens. The results indicate thatresistance to systemic candidiasis is dependent upon a combination of innatefactors, predominately an intact complement system, and the acquisition of animmune response, most likely of a cell-mediated type. Additionally, the findingssuggest that genetic control of acquired resistance to C. albicans may not beassociated with the H-2 complex.

Demonstration in the laboratory of the precisemechanisms of resistance, either innate or ac-quired, to disease caused by Candida albicanshas met with only limited success. Innately,polymorphonuclear leukocytes (PMN) are clear-ly one of the first lines of defense (18, 19, 41),and complement has been demonstrated to playa major role by acting as an opsonin for phago-cytosis by this cell type (16, 24, 25). The role ofmacrophages is much less clear, as studies withagents known to stimulate macrophages haveprovided conflicting results (14, 22, 37, 46).Studies in our laboratory in the past have beenoriented predominately toward the mecha-nism(s) of acquired immunity rather than theinnate response. Using T-cell-depleted mice (11)and mice treated with cyclophosphamide (28),we obtained data which implicated the T-lym-phocyte in immunity, but its precise involve-ment, e.g., at the level of cell-mediated immuni-ty or in a helper function in antibody formation,was not determined. Furthermore, data pub-lished by others attempting to transfer protec-

tion, either with cell suspensions (13, 23, 32, 40)or serum (17, 29, 32), to demonstrate the natureof acquired immunity are conflicting and lessthan convincing in many cases. Our own unpub-lished transfer studies with cells from the spleen,lymph node, or peritoneal cavity, as well ashyperimmune mouse serum, have been unsuc-cessful despite the fact that resistance to anintravenous challenge in immunized donors canbe clearly demonstrated.

Since the use of various syngeneic and con-genic strains of mice to investigate the immuno-logical and genetic factors involved in determin-ing resistance to various infectious agents hasmet with success in recent years, we began asearch for mouse strains susceptible and resist-ant to C. albicans. By using such an approach,we would not have the disadvantages of modelsinvolving immunological manipulations whereinmultiple and somewhat uncontrollable effectscould influence the data derived from suchwork. We report here, therefore, our initialstudies involving in vitro and in vivo responses

1020

on Septem

ber 26, 2018 by guesthttp://iai.asm

.org/D

ownloaded from

IMMUNE RESPONSES TO C. ALBICANS 1021

to immunization via the cutaneous route as wellas responses to intravenous challenge of bothimmune and naive animals in six geneticallydistinct strains of mice.

MATERIALS AND METHODSMice. Six- to eight-week-old male mice of the fol-

lowing strains were obtained from Jackson Labora-tories, Bar Harbor, Maine: BALB/cByJ, C57BL/6J,DBA/2J, A/J, C3H/HeJ, and CBA/J. Outbred CD-1mice were obtained from Charles River Farms, Wil-mington, Mass. All mice were maintained under con-ventional conditions and fed mouse chow and water adlibitum.

Culture methods. C. albicans B311, originally ob-tained from H. Hasenclever, National Institutes ofHealth, was used throughout these studies. It wasmaintained at 4°C by monthly transfer on Sabourauddextrose agar (Scott Laboratories, Inc., Fiskville,R.I.). Viable blastospores for inoculation into micewere grown in tryptic soy dialysate broth (35) for 18 hat 37°C on a gyratory shaker operating at approximate-ly 165 rpm before harvesting by centrifugation. Afterthree washes with sterile nonpyrogenic saline (NPS)(Cutter Laboratories, Inc., Berkeley, Calif.), the blas-tospores were counted in a hemacytometer and sus-pended in NPS at the desired concentration. Eachsuspension was further diluted and spread onto Sa-bouraud agar plates to determine the viable count.Viable blastospores from which antigens were pre-pared were inoculated in soy dialysate broth as above,but after harvest they were washed four times withphosphate-buffered saline (PBS) (0.15 M, pH 7.2)containing 0.001 M phenylmethylsulfonyl fluoride andstored at -20°C in the same buffer until fractionated.

Fractionation procedures and source of test antigens.Subcellular components were prepared from blasto-spores suspended in PBS-phenylmethylsulfonyl fluo-ride by disruption with ballistic action in a Braunhomogenizer followed by differential centrifugation asdescribed previously (7). The two fractions used in thisstudy are designated as B-HEX and SCS. The SCS,soluble cytoplasmic substances, was the dialyzed andlyophilized supernatant remaining after a final centrif-ugation at 144,000 x g for 2 h. B-HEX was preparedfrom a 'membrane-mitochondria fraction which hadbeen extracted with butanol (28) and hot PBS (7)before the precipitation of protein with 100%o saturatedammonium sulfate. The precipitate was redissolved inand dialyzed against NPS and stored at -70°C. Proteincontent was determined by the method of Lowry et al.(21), using bovine serum albumin as a standard.Animal immunization and challenge. Mice were im-

munized as described previously (10) by two intracuta-neous inoculations of 106 live C. albicans B311 blasto-spores suspended in 0.05 ml of NPS. When testing fora protective respon'se, control and immunized animalswere challenged intravenously via the lateral tail veinwith viable blastospores suspended to the appropriateconcent,ration in 0.5 ml of NPS. Animals were moni-tored for mortality for 28 days, bled, and then sacri-ficed for the quantitative culture of brains and kidneys.Footpad testing. A 0.02-ml portion of test prepara-

tion, adjusted to 20 F.g of protein per test dose, wasinjected into each footpad with a micrometer syringe.The footpads had been measured with Schnelltaster

calipers (H. Kroplein GmbH, Schluchtern, West Ger-many) before the injection of antigen. At 15 min and 4,7, 24, and 48 h after injection, footpads were remea-sured, and the mean net increase in thickness wasdetermined for each group. In each experiment theanimals were randomized within each strain of mouse,coded and injected by one individual, and then mea-sured by a second individual to whom the code wasunknown. Uninfected animals were used as controls.Lymphocyte stimulation. Inguinal lymph nodes

draining the site of cutaneous inoculation of threeinfected animals from each strain of mouse and thecorresponding nodes in three uninfected"'mice wereexcised and pooled in their respective groups. Single-cell suspensions were prepared by teasing the nodesapart in cold RPMI 1640 containing 25 mM N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid(HEPES; GIBCO Laboratories, Grand Island, N.Y.)supplemented with 20 mM L-glutamine, 100 U ofpenicillin, and 100 p.g of streptomycin per ml; 5%(vol/vol) heat-inactivated normal pooled horse serum(GIBCO); and 5 x 10-5 M 2-mercaptoethanol (Calbio-chem, La Jolla, Calif.) (4). After filtering throughsterile gauze, the cells were washed in completemedium and suspended to a concentration of 2.5 x 106viable cells per ml, as determined by hemacytometercounts of cells stained with trypari blue (GIBCO).Microcultures were initiated in flat-bottomed micro-titer trays (Microtest II, Falcon Plastics, Oxnard,Calif.) by delivering 0.1 ml of cells (2.5 x 105 per well)and 0.1 ml of complete medium alone or completemedium containing the appropriate concentrations ofantigen or mitogen to each well. The mitogens usedwere purified phytohemagglutinin (PHA) (WellcomeReagents, Ltd., Beckenham, England), 0.5 ,ug perwell, and Salmonella typhosa 0901 lipopolysaccharideB (LPS) (Difco Laboratories, Detroit, Mich.), 12.5 ,ugper well. The antigens used were B-HEX at 10, 20, or40 ,ug of protein per well, and SCS at 50, 100, or 200 ,ugof protein per well. All conditions were prepared intriplicate, and trays were incubated at 37?C in ahumidified atmosphere in 5% CO2 in air. Eighteenhours before harvesting with a MASH II unit (Micro-biological Associates, Bethesda, Md.), each well waspulsed with 0.5 ,uCi of [methyl-3H]thymidine (specificactivity 6.7 Ci/mmol; New England Nuclear Corp.,Boston, Mass.) delivered in 10 IlI of NPS. Mitogenswere harvested on the 4th day of incubation andantigens on the 5th day of incubation. After drying at100°C for 20 min, the glass fiber disks were punchedout into plastic liquid scintillation vials (New EnglandNuclear), and 10 ml of liquid scintillation solutioncontaining 3.66 g of 2,5-diphenyoxazole (New EnglandNuclear) per liter of scintillation-grade toluene (J. T.Baker Chemical Co., Phillipsburg, N.J.) was added.Samples were counted in a Beckman LS250 for aminimum of 10 min or 10,000 counts, and results wereexpressed as the mean of the three samples in countsper minute.

Antibody detection. Anti-Candida antibody in thesera of infected mice was determined by a solid-phase,double antibody enzyme-linked immunosorbent assay.The method used was adapted from several publishedtechniques; therefore, the procedure is described here-in. SCS, the preparation used as the Candida-specificantigen, was dissolved in 0.05 M sodium carbonatebuffer (pH 9.6) to achieve a protein concentration of 50

VOL. 38, 1982

on Septem

ber 26, 2018 by guesthttp://iai.asm

.org/D

ownloaded from

1022 HECTOR, DOMER, AND CARROW

,ug/ml, and 200 p.l per well was added to 96-wellpolyvinyl microtiter plates (Dynatech Laboratories,Inc., Alexandria, Va.). The plates were incubated for 2h at 37°C and then washed four times with 0.15 M PBS(pH 7.2) with 0.05% Tween 20. All wells were thencoated with a solution of 1 mg of bovine serumalbumin per ml (Sigma Chemical Co., St. Louis, Mo.)in carbonate buffer, 200 ,ul per well, and incubated andwashed as before. Sera to be tested were seriallydiluted with PBS-Tween 20 supplemented with 1%fetal calf serum (M.A. Biologicals), starting with a 1:50dilution followed by twofold dilutions to a final dilu-tion of 1:6,400. A positive control, consisting of a poolof sera from immunized mice and a negative controlfrom unimmunized mice, were diluted in like manner.A 150-,l portion of each serum dilution was placed inthe appropriate well, with each plate containing atleast one positive control series, and incubated over-night at 4°C. The plates were then washed as before,and 200 ,ul of a 1:800 dilution (optimal concentrationdetermined by checkerboard titrations) of peroxidase-labeled affinity-purified goat anti-mouse immunoglob-ulin G and immunoglobulin M heavy and light chains(Kirkegaard and Perry Labs, Gaithersburg, Md.) inPBS with 1% fetal calf serum (no Tween 20) was thenadded to each well before incubation for 2 h at 37°C. A200-p.l portion of the substrate for the enzyme system,which consisted of 0.01% hydrogen peroxide (Sigma)and 0.01% 2,2'-azino-di-[3-ethylbenzthiazoline sulfo-nate] (Sigma) in 0.1 M sodium citrate buffer (pH 4.0)prepared immediately before use, was added to eachwell, and the plates were incubated for 30 min at roomtemperature in the dark. The reaction was stopped bythe addition of 50 ,ul of a solution of 10 parts 6 mMNaOH and 0.18% hydrofluoric acid and 1 part 0.4 MNaOH and 78 mM EDTA. The plates were allowed tosit for an additional 10 min before the optical density ofthe samples was measured at 417 nm. Values forrepresentative positive control results were plotted asoptical density versus log2 titer to determine the 50%endpoint value, and the corresponding optical densityvalue was used to determine the titers of the samples.Complement assay. A hemolytic complement assay

was performed according to published methods (15,45). Briefly, ox erythrocytes were sensitized withrabbit anti-ox erythrocyte antibody, kindly suppliedby W. K. Anderson, Tulane University, and suspend-ed to a 2% solution. Sera to be analyzed were dilutedwith barbital buffer in twofold dilutions, starting with a1:4 dilution. A 200-,ul portion of each dilution wasdispensed, in duplicate, into glass tubes, and then 50,ul of the sensitized erythrocytes were added. Thetubes were incubated at 37°C for 30 min, centrifuged at400 x g for 10 min, and the supernatants read at 415nm. The results were plotted, and the CH50 value wascalculated as described (45).

Calculations and statistical analyses. The total colo-ny-forming units per organ were determined by stan-dard dilution calculations and were expressed as loga-rithmic numbers. Within each group, values forkidneys or brains were averaged to obtain a geometricmean. Because deaths before the specific time ofsacrifice would bias such results, a normalizationfactor, representing the maximum values seen for eachorgan, was included in the calculations for each animalthat had died.Where applicable, data were analyzed by one-way

INFECT. IMMUN.

STRAIN

BA Li Byl

C5711/6)

OBA/2 J

A/J

C 3N/NeJ

C A/i

;-~~~~~~-

0. 0.2 03 o 0. 0 0.7 50" 1 15 1.2 13 . 1.5

INCREASE IN FOOTPAD THICKNESS (mm)

FIG. 1. Footpad responses to 20 p.g of protein ofthe antigen B-HEX in immunized (closed bars) andcontrol mice (stippled bars). Values represent theaverage net swelling ± the standard error of the meanfor eight mice at 24 h after injection for each of twoexperiments.

analysis of variance or Tukey's w procedure (42) forthe comparison of means. Significance was deter-mined at the P = 0.05 level.

RESULTSIn vivo cellular immunity. To detect delayed

hypersensitivity, mice from each strain werefootpad tested with B-HEX 7 days after a sec-ond cutaneous inoculation of C. albicans, thetime at which the response is known to bemaximum for CBA/J mice (7). The results fromtwo experiments are summarized in Fig. 1. Al-though the footpads were measured at five dif-ferent intervals, only the 24-h response is pre-sented. There was a large immediate-typeresponse in all immunized animals 15 min afterthe injection of antigen, but this had diminishedby 4 h to levels well below either the immediateresponse or the subsequent delayed-type re-sponse. The largest responses were seen in micefrom strains BALB/cByJ, C57BL/6J, andDBA/2J, and statistical analyses showed no sig-nificant differences among these strains whenresults from both experiments were analyzed byTukey's w procedure. Mice from strainsC3H/HeJ, CBA/J, and A/J showed moderateresponses which were statistically different fromthe highly reactive group. When the experimentwas repeated a third time with a different lot ofantigen, the responses were, with one excep-tion, similar, although all were slightly in-creased. The exception was the immunizedC3H/HeJ mice which were as responsive as the

on Septem

ber 26, 2018 by guesthttp://iai.asm

.org/D

ownloaded from

IMMUNE RESPONSES TO C. ALBICANS 1023

BALI/cIyi C5711/6S DIAl A/i C3N/NeJ CIA/I BALS/cSyi C5781/6J DIA/2i A/i C3H/eJ CIA/I

STRAIN

FIG. 2. Uptake of tritiated thymidine in response to LPS and PHA by lymph node cells from control mice(stippled bars) and from mice immunized with two cutaneous inoculations of C. albicans (closed bars). Valuesrepresent the averages of three experiments ± the standard error of the mean.

three strains which had segregated into the high-ly reactive group in the previous two experi-ments.

In vitro cellular immunity. Lymph node cellsfrom control mice or from mice infected twicecutaneously, the second inoculation having been10 days before sacrifice, were stimulated withthe mitogens LPS and PHA and with two anti-gen preparations derived from C. albicans, B-HEX and SCS. Ten days after the second cuta-neous inoculation was selected because that isknown to be the time of maximum response inCBA/J mice (7). The results of three experi-ments using the mitogens and antigens are sum-marized in Fig. 2 and 3, respectively. Responsesto LPS were minimal in all six strains, perhapsdue to the low percentage of B-cells found inlymph nodes; but on a relative basis, theC57BL/6J cells were consistently more respon-sive than those from the other strains. Cells fromstrains BALB/cByJ, A/J, and C3H/HeJ were theleast responsive to LPS. There were good,though variable, responses by all cells to PHA,with C57BL/6J and DBA/2J cells being the leastresponsive. Only cells from strains C3H/HeJand CBA/J showed significant differences be-tween the responses of control and infectedanimals. The lack of differences between controland infected cell responses in the majority ofstrains suggest there was no suppression in anyof the cell cultures.

In contrast to the results obtained with themitogens, the cell suspensions from immunizedmice responded equally well to the Candidaantigens, regardless of the mouse strain. Al-though the results of the individual experimentsvaried somewhat, when the data were pooled

there were no statistical differences among theresponses from immunized animals to either B-HEX or SCS. With the exception of DBA/2J andA/J cells, most lymphocytes from unimmunizedanimals were unresponsive to the antigens. It isnot known if the response of naive DBA/2J orA/J cells was a mitogenic effect or if these strainsof mice might have had as part of their normalflora an organism containing antigens cross-reactive with those of C. albicans.

Susceptibility of mice to intravenous challenge.Two weeks after receiving a second immunizinginoculation of C. albicans blastospores, the timeat which CBA/J mice are known to be resistantto challenge (10), immunized and control micefrom each of the six strains were challengedintravenously with 4 x 103, 5 x 103, or 7.5 x 103viable C. albicans blastospores in three separateexperiments. The mortality results after 28 daysof observation, together with the cultural dataon kidneys and brains, are presented in Table 1.Naive mice segregated into sensitive, intermedi-ate, and resistant groups, with C57BL/6J andBALB/cByJ being highly resistant. The otherfour strains were more susceptible, with higherlevels of colonization in the kidneys and somedeaths occurring before the time of sacrifice.DBA/2J mice were the most susceptible, withgreater than six log units of yeast in their kid-neys and a high percentage of deaths occurring.In response to immunization, however, five ofthe six strains were able to limit multiplication inkidneys to a level two to three log units lowerthan that observed in unimmunized animals.

In contrast to the significant differences ob-served in colony counts among the kidney cul-tures when naive mice were challenged intrave-

VOL. 38, 1982

T

on Septem

ber 26, 2018 by guesthttp://iai.asm

.org/D

ownloaded from

1024 HECTOR, DOMER, AND CARROW

IAL/ItyJ C571141 I IAJ A/i C3MINJ CAM BALUIAly C571/iJ IAl1 A/J UI/Uei CIA/i

STAlIN

FIG. 3. Uptake of tritiated thymidine in response to 20 ,ug of protein per well of B-HEX (BEX) or 50 ,ug ofprotein per well of SCS, two antigens from C. albicans, by lymph node cells from control mice (stippled bars) andfrom mice immunized with two cutaneous inoculations of C. albicans (closed bars). Values represent theaverages of three experiments ± the standard error of the mean.

nously, there were few differences in thenumbers of Candida recovered from brain. Inmost cases, however, there was a statisticallysignificant difference between the levels of yeastseen in naive versus immunized mouse brains ineach strain in the three experiments (Table 1).Antibody levels. At the time of sacrifice for

cultures of kidneys and brains, blood was col-lected from each mouse for determinations ofanti-Candida antibody by an enzyme-linked im-munosorbent assay. The results of the threeexperiments are presented in Fig. 4.Although virtually all strains of mice which

had been sensitized before challenge had highlevels of circulating antibody to C. albicans, theunimmunized mice varied considerably in theirproduction of antibody in response to intrave-nous infection. Mice from strains C57BL/6J andBALB/cByJ had essentially no detectable anti-body at the end of the 28-day period of infection,whereas the other four strains were able tomount low to moderate responses. A pool ofsera from cutaneously infected but not intrave-nously challenged CBAIJ mice was used as a

positive control for the assay, and typically gavetiters of 1:200. Not surprisingly, the process ofintravenous infection boosted antibody produc-tion in immunized mice, as the mean titers forthe immunized and challenged CBA/J mice werebetween 1:400 and 1:800. Because of the differ-ences in the numbers of samples in the variousgroups, no meaningful statistical analyses were

possible.Complement assay. Simple hemolytic assays

were performed on all strains of mice in thisstudy to verify known deficiencies in the C5

component in strains DBA/2J and A/J (5). Al-though the assay as used was incapable ofresolving concentration differences in each ofthe nine components of the classical pathway, itnonetheless offered indirect verification of thecomplement status of those two strains. Anoutbred CD-1 strain, which was assumed tohave intact complement components, was in-cluded as a positive control, and the results werecomputed as CH50 U/ml (45). The CD-1 micehad the highest levels of hemolytic activity, with32.2 U/ml, and the six strains of inbred mice hadmoderate to negligible levels of hemolytic activi-ty. As expected, the DBA/2J and A/J had essen-tially little or no demonstrable levels, with 0.0and 3.2 U/ml, respectively. C3H/HeJ mice werealso essentially negative, with a value of 0.2U/ml. C57BL/6J and BALB/cByJ mice had mod-erate levels of activity, with 25.3 and 10.3 U/ml,respectively, and CBA/J mice had 6.5 U/ml.

DISCUSSIONNaive mice from the six strains of inbred mice

used in this study showed substantial differencesin susceptibility to C. albicans when challengedintravenously with the fungus. Based on theresults of both the quantitative cultures of kid-neys, the target organ in systemic candidiasis,and the mortality data, strains BALB/cByJ andC57BL/6J had a high degree of native resistanceto challenge; strains A/J, C3H/HeJ, and CBA/Jwere moderately susceptible, and strain DBA/2Jwas highly susceptible. In an effort to determinethe mechanisms which might be responsible forthe observed differences, features of innate and

INFECT. IMMUN.

on Septem

ber 26, 2018 by guesthttp://iai.asm

.org/D

ownloaded from

IMMUNE RESPONSES TO C. ALBICANS 1025

TABLE 1. Mean colony-forming units in kidneys and brains after intravenous challenge of unimmunizedmice or mice immunized by the cutaneous inoculation of viable C. albicans blastospores

Intra- Log1o colony-forming units'veou Mortaliv

Mouse strain challenge Kidney Braindose Immune Control Immune Control Immune Control

C57BL/6J 4 x 103 0.15 ± 0.15 3.54 ± 0.70' 0 1.93 ± 0.18'' 0/8 0/85 x 103 0.24 ± 0.13 3.36 ± 0.48'' 0.11 ± 0.11 2.18 ± 0.04' 0/8 0/87.5 x 103 0 2.51 ± 0.42c 0.20 ± 0.20 0.95 ± 0.25c 0/8 0/8

BALB/cByJ 4 X 103 1.98 ± 0.82 2.00 ± 0.25 0.08 ± 0.08 0.87 ± 0.10'' 0/8 0/75 x 103 0.54 ± 0.37 2.19 ± 0.50c 0 1.36 ± 0.12' 0/8 0/87.5 x 103 2.52 ± 0.32 2.20 ± 0.07 0.32 ± 0.32 1.23 ± 0.23'' 0/8 0/8

DBA/2J 4 x 103 4.27 ± 0.74 6.71 ± 0.39c 0.41 ± 0.30 1.38 ± 0.40 1/8 2/85 X 103 3.84 ± 1.08 6.48 ± 0.41c 0.84 ± 0.37 2.10 ± 0.30c 2/8 7/87.5 x 103 5.49 ± 0.68 6.23 ± 0.44 1.16 ± 0.36 2.06 ± 0.19c 1/8 6/8

A/J 4 x 103 1.42 ± 0.52 4.25 ± 0.61' 0 1.02 ± 0.34c 0/8 1/85 x 103 2.24 ± 0.22 4.92 ± 0.62c 0.48 ± 0.27 1.61 ± 0.20c 0/8 2/87.5 x 103 3.24 ± 0.61 2.48 ± 0.32 1.38 ± 0.39 1.06 ± 0.28 0/8 0/8

C3H/HeJ 4 X 103 1.85 ± 0.81 3.76 ± 0.38 0.43 ± 0.31 1.13 ± 0.25 1/8 0/85 X 103 2.55 ± 0.64 6.40 ± 0.10'' 0.26 ± 0.18 2.16 ± 0.44c 0/8 2/77.5 x 103 3.07 ± 0.81 5.72 ± 0.14c 0 1.95 ± 0.18c 0/8 0/8

CBA/jd 5 X 103 3.30 ± 1.00 6.01 ± 0.39' 0.78 ± 0.33 2.00 ± 0.14'' 1/8 1/87.5 x 103 1.20 ± 0.37 4.90 ± 0.39' 0.20 ± 0.16 1.60 ± 0.16' 0/8 0/8

a Geometric mean ± standard error of eight animals per group with the two exceptions noted in the mortalitycolumn; where animals had died a normalization factor was included for each animal.

b Survival was monitored for 28 days after intravenous challenge, at which time survivors were sacrificed forculture.

c Signifies a statistical difference between immune and control values determined by one-way analysis ofvariance with significance determined at the P = 0.05 level.

d The 4 x 103 challenge was not given to the CBA/J strain.

acquired resistance were considered.The results of this study suggest a major role

for complement in the innate resistance to C.albicans in the mouse model. The DBA/2Jstrain, in particular, and the A/J strain to a lesserextent, were found to be susceptible to systemiccandidiasis, and both have been reported to bedeficient in the C5 component of complement(5). Although we did not determine the levels ofC5 specifically, both strains were deficient inhemolytic complement activity. Thus, our re-sults confirm the in vivo studies of Morelli andRosenberg (24), who used outbred mice, andGelfand et al. (9), who used guinea pigs, inwhich animals lacking an intact alternative path-way of complement did not survive as long asanimals with an intact complement system. Invitro studies, on the other hand, have either notsatisfactorily confirmed the in vivo data or haveresulted in conflicting data. Morelli and Rosen-berg (25), for example, reported in a secondstudy that sera from C5 deficient mice did notsupport phagocytosis of the fungus in an in vitrophagocytic assay as well as sera from mice withC5. This effect, however, was only evident at

certain periods during incubation, and not atothers. In contrast to the latter study, it hasrecently been reported (27) that the sera frommouse strains DBA/2J and D1OD2/oSn, whichare deficient in the C5 component, were able toopsonize and promote phagocytosis of C. albi-cans blastospores by mouse PMN in vitro on alevel equal to that of serum from mice withnormal levels of C5. Furthermore, through de-pletion of factor C3 from normal serum, theywere able to demonstrate a reduction in phago-cytosis of yeast and concluded that C3 was theprimary opsonin of C. albicans for phagocytosisby PMN. Interestingly, it has been reported thatmannan extracted from the cell walls of C.albicans is able to activate the alternative path-way of complement in human serum with aconcomitant release of chemotactic factors (34).Since it has been demonstrated that activated C5but not C3 has chemotactic activity towardsPMN in both in vitro and in vivo systems (8, 39),it is tempting to speculate that the increasedsusceptibility to candidiasis seen in C5 deficientmice in our studies, as well as in those of Morelliand Rosenberg (26), may not have been due to

VOL. 38, 1982

on Septem

ber 26, 2018 by guesthttp://iai.asm

.org/D

ownloaded from

1026 HECTOR, DOMER, AND CARROW

6400 I

32001 A

16001'

I-I-

-me

800So

400 '

DOOA

AO A 00a r--5O033 --- 0&A 0a&

r

000 00 0~~~~~~a&00 000 Do*°° 0°°' 0 °° m° ° AO r-4- 3 - *______ _ _ ~L J _____J

200 F 0h00

1001'

50?o

<so

AAA&0000 0000

0

OAAO 0

0000000An

A A 000 Ama,&On

r------ AA 0000 Go0A00

a WC -~ 000U *AAALAm& *DAAA r--.AAAS..-.A a

a i in---. EaseA 009 00

0ep

BALB/cByJ C57B1/6J DBA/2J A/J C3H/HeJ CBA/JSTRAIN

FIG. 4. Antibody titers as determined by enzyme-linked immunosorbent assay procedure in immunized(open symbols) or control (closed symbols) mice 28 days after intravenous challenge with 4 x 103 (0, 0), 5 x 103(O, *), or 7.5 x 103 (A, A) blastospores of C. albicans. The mean values for immunized (*) or control (**)mice are represented by the dashed lines. In the calculation of means, titers were converted to log values beforeaveraging, with titers of less than 1:50 assigned a value of zero.

the inability to opsonize and phagocytose thefungus, but rather to the inability to mount anadequate inflammatory response.The role of PMN in defense against candidia-

sis is well known, its importance having beendemonstrated in clinical studies in humans (18,19, 41) as well as in studies with experimentalanimals (13, 30, 38). Thus, in the infected ani-mal, the ability to opsonize the organism as wellas promote chemotaxis of PMN into regions offungal invasion may be of equal importance, andthe high levels of susceptibility seen in theDBA/2J mice, which might be deficient in theability to induce chemotaxis, could be attributedto a lack of chemotactic activity. The increasedsusceptibility ofDBA/2J mice does not appear tobe related to the numbers of circulating PMN,since it has been reported in the literature (6)that that strain has relatively high levels ofcirculating PMN, whereas the more resistantstrains have lower counts. Our own limiteddeterminations (data not shown) confirmed theliterature reports, in that the DBA/2J mice, themost susceptible strain, had over twice the num-ber of circulating PMN as mice from strainC57BL/6J, a highly resistant strain.

Elucidation of the factors responsible for ac-

quired immunity to candidiasis in inbred mice

has proven to be somewhat difficult. All strainsof mice used in this study were capable ofdeveloping some measure of resistance to intra-venous challenge with C. albicans if they wereimmunized by the intracutaneous inoculation ofviable cells before challenge. In fact, five of thesix strains showed a reduction of two to threelog units in the levels of colonization in thekidneys in immunized versus naive animals.That mice from strain BALB/cByJ did not showsuch reductions in two of the three experimentsis not surprising in view of the already low levelsof colonization in the naive animals. The resultsof the antibody assays in naive animals wouldseem to indicate that antibody played no role innative resistance and by inference, therefore, inacquired resistance. The two most resistantstrains of mice, C57BL/6J and BALB/cByJ, hadessentially no detectable levels of antibody atthe end of the 28-day infection period whennaive animals were challenged intravenously,whereas the more susceptible strains had de-monstrable titers of antibody. Thus, the differ-ences observed in antibody titers may simply bea case of differences in the antigen load, suchthat susceptible strains with their higher levelsof colonization produced more antibody in re-sponse to the higher antigenic insult. It has been

INFECT. IMMUN.

on Septem

ber 26, 2018 by guesthttp://iai.asm

.org/D

ownloaded from

IMMUNE RESPONSES TO C. ALBICANS 1027

demonstrated previously that in in vitro phago-cytosis assays, specific antibody plays little orno role in the phagocytosis and subsequentkilling of C. albicans (16, 25, 27).

If antibody does not play a role in acquiredimmunity, then perhaps one must postulate aprotective role for cell-mediated immunity, ashas been clearly demonstrated for coccidioido-mycosis (2). The results of the lymphocyte pro-liferation and footpad assays demonstratedclearly that all six strains were capable ofmounting cell-mediated responses to antigensfrom C. albicans. Correlation of these T-cellresponses with the ability to resist infection bycell-mediated immunity is more difficult, howev-er. Although Hurtrel et al. (13) and Moser andDomer (28) reported a lack of correlation be-tween delayed-type hypersensitivity and protec-tion in systemic candidiasis, Kagaya et al. (17)felt the temporal relationship between acquisi-tion of a delayed-type response and resistance toinfection observed in their mouse model wasgood evidence for their correlation. Althoughour results suggest the relative importance of acell-mediated versus a humoral response in ac-quired immunity to candidiasis, actual proof ofthe dominance of either aspect of the immuneresponse would be dependent on cell transferexperiments. Attempts at transferring protectionto naive mice against infection with C. albicanswith either immune serum (17, 29, 32) or cellsfrom immune donors (13, 23, 32, 40) have led toconflicting and less than convincing results.There have been only a few publications in

which the responses of various inbred strains ofmice to fungal agents were reported. In the mostrecent, Bistoni et al. (3) reported strain differ-ences in susceptibility to intravenous challengewith C. albicans slightly different than ours inthat at certain doses, C3H/HeJ mice were asresistant to challenge as BALB/c mice. Theirdata are difficult to compare with ours, however,because the authors used much larger inoculaand reported only the mean survival time as abasis for resistance. Earlier, resistance to Cryp-tococcus neoformans was found to be undersingle-gene control, with the gene product reput-ed to be C5 (36). Finally, Morozumi et al. (26)reported wide variations in susceptibility of nineinbred strains of mice to Blastomyces dermatiti-dis given intranasally, but no obvious reasonsfor such differences were evident.The use of inbred mice to study infectious

agents other than fungi has been exploited withsuccess. Interestingly, studies with parasitic (1,20, 31, 44, 47) and bacterial agents (12, 33, 43)have shown that there are no uniform trends inthe ability of selected strains of mice to resist allmembers of a given class of pathogen. Addition-ally, in most of the studies cited above, the

determinants of resistance were reported tohave no close linkage to the H-2 locus. Thiswould appear to be the case with C. albicans, asBALB/cByJ and DBA/2J mice, representing themost resistant and susceptible mice in our study,respectively, share the same H-2 haplotype.Confirmation of the genetic restrictions in candi-diasis will, however, require further studies.

ACKNOWLEDGMENTS

This investigation was supported by Public Health Servicegrants Al-12806 and AI-07152 from the National Institute ofAllergy and Infectious Diseases.R.F.H. was a trainee in medical mycology.

LITERATURE CITED

1. Albright, J. W., and J. F. Albright. 1981. Differences inresistance to Trypanosoma musculi infection amongstrains of inbred mice. Infect. Immun. 33:364-371.

2. Beaman, L., E. Benjamini, and D. Pappaglanls. 1981. Roleof lymphocytes in macrophage-induced killing of Cocci-dioides immitis in vitro. Infect. Immun. 34:347-353.

3. Bistoni, F., P. Marconi, L. Frati, E. Bma , and E.Garaci. 1982. Increase of mouse resistance to Candidaalbicans infection by Thymosin a,i. Infect. Immun.36:609-614.

4. Brummer, E., T. W. Vris, and H. S. Lawrence. 1977. Amicroculture system for the measurement of antigen-induced murine lymphocyte proliferation: advantages of5% horse serum and 5 x 10-5 M mercaptoethanol. J.Immunol. Methods 17:319-327.

5. Cinander, B., S. Dubiski, and A. C. Wardlaw. 1964.Distribution, inheritance, and properties of an antigen,MUB1, and its relation to hemolytic complement. J. Exp.Med. 120:897-924.

6. Crispens, C. G. (ed.). 1975. Handbook on the laboratorymouse. Charles C. Thomas, Publisher, Springfield, Ill.

7. Domer, J. E., and S. A. Moser. 1978. Experimental mu-rine candidiasis: cell-mediated immunity after cutaneouschallenge. Infect. Immun. 20:88-91.

8. Fernandez, H. N., P. M. Henson, A. Otani, and T. E.Hugli. 1978. Chemotactic response to human C3a and C5aanaphylatoxins. I. Evaluation of C3a and C5a leukotaxisin vitro and under simulated in vivo conditions. J. Immu-nol. 120:109-115.

9. Gelfand, J. A., D. L. Hurley, A. S. Fauci, and M. M.Frank. 1978. Role of complement in host defense againstexperimental disseminated candidiasis. J. Infect. Dis.138:9-16.

10. Giger, D. K., J. E. Domer, and J. T. McQuitty, Jr. 1978.Experimental murine candidiasis: pathological and im-mune responses to cutaneous inoculation with Candidaalbicans. Infect. Immun. 19:499-509.

11. Giger, D. K., J. E. Domer, S. A. Moser, and J. T.McQuitty, Jr. 1978. Experimental murine candidiasis:pathological and immune responses in T-lymphocyte-depleted mice. Infect. Immun. 21:729-737.

12. Gros, P., E. Skamene, and A. Forget. 1981. Geneticcontrol of natural resistance to Mycobacterium bovis(BCG) in mice. J. Immunol. 127:2417-2421.

13. Hurtrel, B., P. H. Lagrange, and J.-C. Michel. 1981.Absence of correlation between delayed-type hypersensi-tivity and protection in experimental systemic candidiasisin immunized mice. Infect. Immun. 31:95-101.

14. Jakoniuk, P., J. Borowski, W. Jablonska-Strynkowska,and W. Badmajew. 1978. Effect of lipids from Listeriamonocytogenes on bacterial and fungal infections in mice.Arch. Immunol. Ther. Exp. 26:637-640.

15. Kabat, E. A., and M. M. Mayer (ed.). 1961. ExperimentalImmunochemistry, p. 133-240. Charles C. Thomas, Pub-lisher, Springfield, Ill.

VOL. 38, 1982

on Septem

ber 26, 2018 by guesthttp://iai.asm

.org/D

ownloaded from

1028 HECTOR, DOMER, AND CARROW

16. Kagaya, K., and Y. Fukazawa. 1981. Murine defensemechanisms against Candida albicans infection. II. Mi-crobiol. Immunol. 25:807-818.

17. Kagaya, K., T. Shinoda, and Y. Fukazawa. 1981. Murinedefense mechanism against Candida albicans infection. I.Microbiol. Immunol. 25:647-654.

18. Laforce, F. N., D. M. Mills, K. Iverson, R. Cousins, andE. D. Everett. 1975. Inhibition of leukocyte candidacidalactivity by serum from patients with disseminated candi-diasis. J. Lab. Clin. Med. 86:657-666.

19. Lehrer, R. I. 1970. Measurement of candidacidal activityof specific leukocyte types in mixed cell populations. I.Normal, myeloperoxidase-deficient, and chronic granulo-matous disease neutrophils. Infect. Immun. 2:42-47.

20. Levine, R. F., and J. M. Mansfield. 1981. Genetics ofresistance to African trypanosomes: role of the H-2 locusin determining resistance to infection with Trypanosomarhodesiense. Infect. Immun. 34:513-518.

21. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J.Randall. 1951. Protein measurement with the Folin phenolreagent. J. Biol. Chem. 193:265-275.

22. Marra, S., and E. Balish. 1974. Immunity to Candidaalbicans induced by Listeria monocytogenes. Infect. Im-mun. 10:72-82.

23. Miyake, T., K. Takeya, K. Nomoto, and J. Muraoka. 1977.Cellular elements in the resistance to Candida infection inmice. I. Contribution of T lymphocytes and phagocytes atvarious stages of infection. Microbiol. Immunol. 21:703-725.

24. Morelli, R., and L. T. Rosenberg. 1971. Role of comple-ment during experimental Candida infection in mice.Infect. Immun. 3:521-523.

25. MoreNi, R., and L. T. Rosenberg. 1971. The role ofcomplement in the phagocytosis of Candida albicans bymouse peripheral blood leukocytes. J. Immunol. 7:476-480.

26. Morozumi, P. A., J. W. Halpern, and D. A. Stevens. 1981.Susceptibility differences of inbred strains of mice toblastomycosis. Infect. Immun. 32:160-168.

27. Morrison, R. P., and J. E. Cutler. In vitro studies of theinteraction of murine phagocytic cells with Candida albi-cans. RES. J. Reticuloendothel. Soc. 29:23-34.

28. Moser, S. A., and J. E. Domer. 1980. Effects of cyclo-phosphamide on murine candidiasis. Infect. Immun.27:376-386.

29. Mourad, S., and L. Friedman. 1968. Passive immunizationof mice against Candida albicans. Sabouraudia 6:103-105.

30. Mukherji, A. K., and K. C. Basu Mallick. 1972. Dissemi-nated candidiasis in cyclophosphamide induced leuco-penic state: an experimental study. Indian J. Med. Res.60:1584-1591.

31. Niederkorn, J. Y., J. A. Shadduck, and E. C. Schmidt.1981. Susceptibility of selected inbred strains of mice toEncephalitozoon cuniculi. J. Infect. Dis. 144:249-253.

32. Pearsall, N., B. Adams, and R. Bunni. 1978. Immunologicresponses to Candida albicans. III. Effects of passive

transfer of lymphoid cells or serum on murine candidiasis.J. Immunol. 120:1176-1180.

33. Plant, J., and A. A. Glynn. 1976. Genetics of resistance toinfection with Salmonella typhimurium in mice. J. Infect.Dis. 133:72-78.

34. Ray, T. L., A. Hanson, L. F. Ray, and K. D. Wuepper.1979. Purification of a mannan from Candida albicanswhich activates serum complement. J. Invest. Dermatol.73:269-274.

35. Restrepo-Moreno, A., and J. D. Schneidau, Jr. 1967. Na-ture of the skin-reactive principle in culture filtratesprepared from Paracoccidioides brasiliensis. J. Bacteriol.93:1741-1748.

36. Rhodes, J. C., L. S. Wicker, and W. J. Urba. 1980. Genet-ic control of susceptibility to Cryptococcus neoformans inmice. Infect. Immun. 29:494-499.

37. Rogers, T., and E. Balish. 1977. The role of activatedmacrophages in resistance to experimental renal candidia-sis. RES. J. Reticuloendothel. Soc. 22:309-318.

38. Ruthe, R. C., B. L. Anderson, B. L. Cunningham, andR. B. Epstein. 1978. Efficacy of granulocyte transfusion incontrol of systemic candidiasis in leukopenic host. Blood52:493-498.

39. Snyderman, R., H. S. Shin, J. K. Phillips, H. Gewurz, andS. E. Mergenhagen. 1969. A neutrophil chemotactic factorderived from C'5 upon interaction of guinea pig serumwith endotoxin. J. Immunol. 103:413-422.

40. Sohnle, R. G., M. M. Frank, and C. H. Kirkpatrick. 1976.Mechanisms involved in elimination of organisms fromexperimental cutaneous Candida albicans infection inguinea pigs. J. Immunol. 117:523-530.

41. Staples, P. J., J. Boujak, R. G. Douglas, Jr., and J. P.Leddy. 1977. Candidiasis in a previously healthy girl-implication of a leukocyte candidacidal defect. Clin. Im-munol. Immunopathol. 7:157-167.

42. Steel, R. G. D., and J. H. Torrie. (ed). 1960. Principlesand procedures of statistics. p. 109-111. McGraw-HillBook Co., New York.

43. Stevenson, M. M., P. A. L. Kongshavn, and E. Skamene.1981. Genetic linkage of resistance to Listeria monocyto-genes with macrophage inflammatory responses. J. Im-munol. 127:402-407.

44. Tanowitz, H. B., N. Minato, R. Lalonde, and M. Wittner.1981. Trypanosoma cruzi: correlation of resistance andsusceptibility in infected inbred mice with the in vivoprimary antibody response to sheep red blood cells. Exp.Parasitol. 52:233-242.

45. Weir, D. M. (ed.). 1978. Handbook of experimental immu-nology, vol. I, p. SA.8-5A.13. Blackwell Scientific Publi-cations.

46. Williams, D., J. Cook, E. Hoffman, and N. DiLuzio. 1978.Protective effect of glucan in experimentally inducedcandidiasis. RES. J. Reticuloendothel. Soc. 23:479-490.

47. Williams, D. M., F. C. Grumet, and J. S. Remington.1978. Genetic control of murine resistance to Toxoplasmagondii. Infect. Immun. 19:416-420.

INFECT. IMMUN.

on Septem

ber 26, 2018 by guesthttp://iai.asm

.org/D

ownloaded from