INFECTION AND IMMUNITY, Mar. 1977, p. 978-987Copyright X) 1977 American Society for Microbiology

Vol. 15, No. 3Printed in U.S.A.

Cell Wall Studies ofHistoplasma capsulatum andBlastomyces dermatitidis Using Autologous

and Heterologous EnzymesTHOMAS E. DAVIS, JR., JUDITH E. DOMER,* AND YU-TEH LI

Department of Microbiology and Immunology* and the Department ofBiochemistry, Tulane UniversitySchool of Medicine, New Orleans, Louisiana 70112

Received for publication 27 August 1976

Enzymes capable of hydrolyzing cell walls of Blastomyces dermatitidis andchemotypes I and II of Histoplasma capsulatum were prepared in the laboratoryor obtained from commercial sources. They included chitinases, /8-1,3-glucan-ases, 83-1,6-glucanase, and Pronase. Monosaccharides and disaccharides of glu-cose released from the cell walls by the enzymes were determined qualitativelyby paper and gas-liquid chromatography, and monosaccharides were quanti-tated by the latter technique as well. An enzyme system isolated from Strepto-myces sp. containing both chitinase and glucanase released maximum amountsof glucose and N-acetylglucosamine from the cell walls of H. capsulatumchemotype I. A chitinase preparation, free of glucanase, from Serratia marces-

cens released only chitobiose and N-acetylglucosamine from chemotype I cellwalls, but the total quantity of N-acetylglucosamine released was about 60%less than that released by the Streptomyces system. A /8-1,3-glucanase fromBacillus circulars hydrolyzed the cell walls of H. capsulatum chemotype I, buta 8-1,6-glucanase failed to release glucose from the same walls. Autolyticenzymes, viz., /8-1,3-glucanases and several glycosidases, were detected as con-

stitutive enzymes in both yeast and mycelial phases of B. dermatitidis and H.capsulatum chemotypes I and II. No difference in the amount of activity wasfound between cell sap and culture filtrate preparations. The /8-glucanasesprepared from the Histoplasma and Blastomyces strains were active on the cellwalls of the yeast phases of H. capsulatum chemotypes I and II, releasinglaminaribiose and glucose, but were essentially inactive on the cell walls of B.dermatitidis. Chitinase, 3-1,6-glucanase, a-glucanase, and a-glucosidase activi-ties were absent from these fungal enzyme preparations.

The cell wall of Histoplasma capsulatum, adimorphic fungus that infects both man andother animals, consists predominantly of chitinand glucan (12-14, 23, 24). Based on the ratio ofchitin to glucan in the walls of the yeast phase,those strains of H. capsulatum studied to datecan be divided into two chemotypes, I and II(14). Chemotype I cell walls contain more chitinand less glucan than chemotype II. Moreover,the evidence suggests that the glucan in chemo-type II cell walls is predominantly linked in thea configuration with few /8-glucans present,whereas the glucan of chemotype I is entirely,3-linked.

Lytic enzymes have been used in the past toinvestigate the cell wall of H. capsulatum, al-beit to a limited extent. For example, Domer etal. (12, 14) used a crude chitinase to attack thecell wall of H. capsulatum, but since the prep-aration contained not only chitinase but alsoglucanase activity, no conclusions concerning

the role of chitin or glucan in wall structurecould be drawn. Additionally, Kanetsuna et al.(23) employed a chitinase as well as a prepara-tion containing several glucanases to lyse cellwalls of H. capsulatum. They suggested thatthe wall of H. capsulatum consisted of an in-ner layer of chitin and an outer layer of glucan.

In the present investigation we decided toexpand the studies of H. capsulatum byDomer (12, 14) in two directions: (i) using par-tially purified enzymes produced by other mi-croorganisms in an attempt to further definethe cell wall and (ii) searching for glucan hy-drolases produced by the fungus itself. For com-parative purposes, cell walls and cultures ofBlastomyces dermatitidis were included in thestudies as well.

MATERIALS AND METHODSMicroorganisms and cultivation. H. capsulatum

SwA, H. capsulatum G184A, and B. dermatitidis978

on Novem

ber 22, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

CELL WALLS OF HISTOPLASMA AND BLASTOMYCES 979

3489 were clinical isolates obtained from H. C.Sweany (Missouri State Sanatorium), M. D. Ber-liner (Harvard School of Public Health), and J. D.Schneidau, Jr. (Tulane University), respectively,and were maintained in the yeast phase by weeklytransfer on brain heart infusion agar slants (Difco)at 370C. H. capsulatum SwA was representative ofchemotype I, whereas H. capsulatum G184A wasrepresentative of chemotype II. Cell walls as well asenzymes were derived from yeast- or mycelial-phasecultures incubated in modified Sabouraud liquidbroth (Baltimore Biological Co.) that had been dis-pensed in 100-ml amounts in 500-ml Erlenmyerflasks. Additionally, enzymes were prepared fromthe yeast phase of H. capsulatum SwA grown in amodified synthetic medium (33) consisting of: 0.7 gof K2HPO4, 0.3 g of KH2PO4, 0.58 g of MgSO4 7H20,0.018 g of ZnSO4 7H2O, 0.01 g of FeSO4 7H2O, and1.0 g of (NH4)2SO4 dissolved in 1 liter of distilledwater. The pH of the latter medium was adjusted to7.0, and chitin or laminarin at a final concentrationof 1% was added as the sole carbon source.

Inocula for the mass production of yeasts or my-celia were prepared in the following manner. Twoslants of brain heart infusion agar were seeded fromthe stock culture and incubated for 5 days at 370C.To prepare yeasts, the growth from both slants waswashed from the surface and used to inoculate asingle flask of modified Sabouraud liquid broth,which was incubated at 370C for 3 days on a gyratoryshaker at 160 rpm. A 5-ml portion ofthis culture wasthen inoculated into each of four flasks of modifiedSabouraud liquid broth and incubated as above.Cultures were harvested by centrifugation at1,000 x g. To prepare mycelia, the brain heart infu-sion agar slants were inoculated with yeast and usedto seed the modified Sabouraud liquid broth, but theculture was subsequently incubated at 25°C for 14days on a gyratory shaker. Complete conversion tothe mycelial phase had occurred by the end of 14days, and then four additional flasks of mediumwere inoculated from the 14-day-old culture. Thefresh cultures were incubated for only 7 days andharvested by centrifugation.

Protoplasts of the H. capsulatum chemotype I iso-late were produced and harvested by the method ofCarbonell et al. (8), and the degree of protoplastformation was determined by phase-contrast mi-croscopy and by the use of Calcofluor White M-2Rfluorescent brightener (19).

Source of substrates and biochemicals. Cell wallswere obtained from yeast-phase cells of all threestrains, which had been disrupted mechanically in amodel MSK Braun cell homogenizer (Bronwill Sci-entific Co.) operating at 4,000 cycles/min, the innerchamber of which was cooled by intermittent blastsof carbon dioxide. Glasperlen beads (0.25 to 0.30 mmin diameter, Bronwill Scientific Co.) were added tocells in glass bottles in a ratio of approximately 3:1(wt/vol), i.e., 30 g of beads per 10 g (wet weight) offungus suspended in approximately 10 ml of 0.1 Nsodium phosphate buffer at pH 7.0. Virtually 100%breakage of the yeast cells occurred after 4 min ofagitation. The degree of breakage was determinedby Gram staining, with unbroken cells being grampositive and broken cells being gram negative, and

by culture of the debris on blood agar plates. Cellwalls were separated from the mixture by centrifu-gation at 1,000 x g and were washed at least 10times with 0.2 M NaCl, once with 1 M NaCl, and 20times with distilled water before lyophilization andstorage in a vacuum desiccator at 25°C. A glycopro-tein fraction, soluble in water, was extracted withethylenediamine from these cell walls after lipidextraction (14).

Laminaribiose was prepared according to Wol-from and Franks (40), but the remainder of thesubstrates were commercial preparations or gifts.Chitin, laminarin, p-nitrophenyl phosphate, 2-deox-yglucose, and p-nitrophenyl glycosides were pur-chased from Sigma Chemical Co., and pustulan,Azocoll, and Lipostrate-CB were purchased fromCalbiochem. Laminarin is a 18-1,3-glucan, and pus-tulan is a 13-1,6-glucan. Chitobiose was preparedaccording to the method of Bailey and Pridham (3).Calcofluor White M-2R was a gift of the AmericanCyanimid Co.

Preparation and assay of heterologous enzymes.A chitinase was prepared from the culture filtrate ofStreptomyces sp. ATCC 11238 (obtained from theAmerican Type Culture Collection [ATCC]) by amethod similar to that of Skujins et al. (36, 37) inwhich the crude enzyme preparations were chromat-ographed on a Sephadex G-100 column. A secondchitinase was isolated from the culture medium ofSerratia marcescens by precipitation with solid am-monium sulfate (32), followed by column chromatog-raphy on Sephadex G-100. Both chitinases were pre-pared from cultures incubated with chitin as the solecarbon source. To obtain /8-1,3- and 3-1,6-glucan-ases, Bacillus circulans was incubated in a mediumcontaining laminarin as the sole carbon source andthe glucanases were purified on a Sephadex G-100column, using the method of Fleet and Phaff (17).Finally, a crude pullulanase containing a-1,4 anda-1,6 activities was prepared from Aerobacter aero-genes as previously described (5).

Several enzymes were purchased from commer-cial sources: chitinase (Calbiochem 22071, lot400891); laminarinase (Calbiochem 427999, lot400266); protease type VI from Streptomyces griseus(Sigma P-5130, lot 114C-0168); and grade B Pronase(Calbiochem 53702, lot 502133).Enzymatic hydrolyses of cell walls, chitin, lami-

narin, chitobiose, or pustulan were performed in aphosphate-acetate buffer, pH 5.0 to 5.5, at 37°C. A0.5- to 2.0-mg portion of substrate was added to 2 mlof buffer in screw-capped tubes fitted with Teflon-lined lids. A given quantity of enzyme, determinedby preliminary assays to be the amount required formaximum hydrolysis, was added to the substrateand incubated for 24 to 48 h. The products of thehydrolyses were determined by paper and gas-liquidchromatography (see below). Since some of the sub-strates were insoluble, the tubes were incubated ona reciprocating shaker operating at approximately100 excursions/min to allow maximum contact ofenzyme and substrate. The total distance traveledper excursion was approximately 1.5 in (ca. 3.81cm). Phosphatase, lipase, and protease activitieswere determined using p-nitrophenyl phosphate,Lipostrate-CB, and Azocoll, respectively, as sub-

VOL. 15, 1977

on Novem

ber 22, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

980 DAVIS, DOMER, AND LI

strates according to the instructions of the manufac-turers for the conditions of hydrolysis.

Preparation and assay of autologous enzymes.

Glucanases were extracted from the culture filtrate(extracellular), the cell wall (cell wall-associated),and the cell sap (intracellular) fractions of all threestrains of fungi in the following manner. Cell sap

(i.e., all cellular constituents other than cell wallsremaining in the supernatant after low-speed cen-

trifugation after the homogenization of yeast cells)and culture filtrate were fractionated at 40C by theaddition of solid ammonium sulfate. The precipitateat 40% saturation was tested for glucanase activity,found to be negative, and discarded, and the concen-

tration of ammonium sulfate was then raised to 75%saturation. The resultant precipitate was resus-

pended in 5 to 10 ml of 0.1 N sodium phosphatebuffer, pH 7.0, dialyzed for 4 h against distilledwater, lyophilized, and stored at -200C. Cell wall-associated glucanase was removed from suspensionsof cell walls by incubation under sterile conditionsin 0.1 N sodium phosphate buffer, pH 7.0, for 7 daysat 230C. After centrifugation to remove the cellwalls, the supernatant was fractionated and storedas above. Where appropriate, protein estimateswere made by the method of Lowry et al. (29).The enzymatic hydrolyses were performed as pre-

viously described (12), and the reaction was stoppedby heating in a boiling-water bath. The residue wasremoved by centrifugation, and 50 pl ofthe superna-tant was used for paper chromatography. A 1-mlportion of the remaining hydrolysate was dividedequally between two vacuoles (Wheaton Glass Co.)and lyophilized. The lyophilized material in eachvacuole was dissolved in 1 ml ofwarm 1.0 N metha-nolic HCl and incubated at 80'C for 3 h (10). Themethanol was subsequently removed under nitro-gen, and O-trimethylsilyl ether derivatives were

prepared for gas-liquid chromatography (26, 38), us-

ing mannitol as an internal standard for quantita-tion. Using this two-step procedure, i.e., enzymehydrolysis followed by acid hydrolysis, all of theglucose in the substrate could be quantitated, even

that which was released by the enzyme in the formof oligomers.

Substrate specificity was determined by measur-

ing the activity of enzyme preparations against a

battery of p-nitrophenyl glycosides (28) as well as

against laminarin, chitin, pustulan, and cell wallpreparations. The effect ofpH on glucanase activitywas observed between pH 3.5 and 7.5, using 0.1 Nsodium phosphate and 0.05 M phosphate citrate andphosphate acetate buffers. The effect oftemperatureon glucanase activity was observed between 23 and80'C, and the effect of metal ions on enzyme activitywas observed using 5 x 10-3 M Ca2+, 5 x 10-3 MCu2+, 5 x 10-3 M Hg2+, 5 x 10-3 M Mg2+, or 5 x 10-3M Mn2+. The effect of 5 x 10-6 M ethylenediamine-tetraacetic acid was also determined.

Chromatography. Carbohydrates released by en-

zymatic attack were determined qualitatively by de-scending paper chromatography. A 50-ml portion ofthe hydrolysate was spotted on Whatman no. 1 chro-matography paper and developed for 12 to 16 h in a

system consisting of 1-butanol-pyridine-0.1 N HCl

INFECT. IMMUN.

(5:3:2, vol/vol). The chromatograms were then airdried, dipped in benzidine reagent, and heated at1000C for 1 min to visualize the carbohydrate prod-ucts of the hydrolysis (6).The monosaccharides and disaccharides released

were also detected as O-trimethylsilyl ether deriva-tives in a Barber-Coleman gas-liquid chromato-graph equipped with a flame ionization detector anda glass tubular column (6 ft [ca. 182.88 cm]) contain-ing 3% OV-1 on 80/100 Chromosorb W-HP (Supelco,Inc.). Nitrogen was the carrier gas. N-acetylglucos-amine and glucose were the only carbohydrates thathad been quantitated. They were quantitated bythis method, using mannitol as the internal stan-dard. The detector response to glucose was equiva-lent to that of mannitol on a weight-to-weight basisso that calculations could be done directly withoutthe use of a factor. The detector response to an equalamount of N-acetylglucosamine, however, was 2.1times less than the response to mannitol. That fac-tor, therefore, was used in all calculations involvingN-acetylglucosamine.When the hydrolytic mixture contained both

mono- and disaccharides, they were separated in asingle operation with the following program. A con-stant temperature of 170'C was used to elute themonosaccharides from the column during the first 15min of operation. At the end of that time, the tem-perature was increased to 220'C at a rate of 10'C/min and held there for 40 min, during which timethe disaccharides were eluted. Table 1 contains theretention indices, based on a-glucose, which was ar-bitrarily assigned the value of 1.0, of 10 mono- anddisaccharides that were separated by this procedure.One or two major peaks were observed with eachcompound, as indicated in the table. Several minorpeaks occurred when mannose was tested, but theywere proportionately insignificant compared to themajor peak. When cell walls were hydrolyzed withmethanolic HCl, however, only one anomeric formof mannose was observed. No minor peaks wereobserved with glucose, N-acetylglucosamine, or thedisaccharides.

RESULTSEnzymatic hydrolyses with heterologous en-

zymes. The products of enzymatic hydrolyses ofthe yeast phase cell walls of H. capsulatum andof selected additional substrates of known com-position, which were determined using the het-erologous enzymes prepared as described aboveor purchased from commercial sources, are pre-sented in Table 2. The products were identifiedby paper and gas chromatography after 4 h ofincubation. If the mixtures were incubated for48 h, only monosaccharides could be detected.The chitinase produced by Streptomyces sp.ATCC 11238 was contaminated with glucanaseactivity, and attempts to purify it by columnchromatography were unsuccessful. Therefore,N-acetylglucosamine, chitobiose, and glucosewere released from the fungal cell walls. The

on Novem

ber 22, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

CELL WALLS OF HISTOPLASMA AND BLASTOMYCES 981

same enzyme system released glucose and lam-inaribiose from laminarin and N-acetylglucosa-mine and chitobiose from chitin. Larger oligo-mers of both glucose and N-acetylglucosaminewere released from laminarin and chitin, re-

spectively, during the early hours of incubationbut had been hydrolyzed to monomers or di-mers by 24 h. A commercial chitinase also con-

tained glucanase activity and produced a simi-lar hydrolysis pattern from the cell walls.Three of the enzyme preparations isolated in

this study, a chitinase-glucanase mixture fromStreptomyces sp. ATCC 11238, a chitinase from

TABLE 1. Retention indices of0-trimethylsilyl etherderivatives of mono- and disaccharides separated in

a single run by gas-liquid chromatography

Sugar Retention index

Mannose ...... ........ 0.70a-Glucose ...... ........ 1.00Mannitol .............. 1.25,8-Glucose ...... ........ 1.50N-acetylglucosamine .............. 2.32Sucrose ..... ......... 3.40Cellobiose ...... ........ 4.00, 6.00Laminaribiose ........... ... 7.00, 7.40Gentibiose ...... ........ 8.80, 9.20Chitobiose .............. 11.10

a The following chromatographic conditions andequipment were used: flame ionization detector; 3%OV-1 column on 80/100 Chromosorb W-HP; isother-mal operation, 15 min at 1700C followed by pro-grammed increase, 10'C/min, to 220'C; injector tem-perature, 2650C; detector temperature, 2350C.

S. marcescens, and a f-1,3-glucanase from B.circulans, were used to quantitate by gas-liquidchromatography the release of glucose and N-acetylglucosamine from the cell walls of theyeast phase of H. capsulatum chemotype I (Ta-ble 3). The Streptomyces mixture released 9 to12% of the cell wall as glucose and 25 to 38% as

N-acetylglucosamine. Chitobiose could be de-tected in hydrolysates, as verified by co-gas-liquid chromatography, when the incubationperiod was less than 24 h, but it was not quanti-tated. Chitobiase activity was demonstrated inthe enzyme mixture, since N-acetylglucosa-mine was released from chitobiose. Moreover,the enzyme mixture from Streptomyces wasfree of lipase, phosphatase, and protease activ-ity.The chitinase system prepared from S. mar-

cescens was free of glucanase activity, and onlyN-acetylglucosamine was released from the cellwalls ofH. capsulatum after 24 h (Table 3). Aswith the Streptomyces mixture, the chitinasefrom S. marcescens contained chitobiase andwas free of lipase, phosphatase, and proteaseactivity. Using the S. marcescens system, how-ever, only 12 to 15% of the cell wall was hydro-lyzed to N-acetylglucosamine. Pretreatment ofcell walls with 83-1,3-glucanase from B. circu-lans prior to hydrolysis with this chitinase re-sulted in only a moderate increase in theamount ofN-acetylglucosamine released.The glucanase prepared from B. circulans

contained no chitinase but did contain both ,B-1,3- and 8-1,6-glucanase activities. The two ac-

TABLE 2. Products resulting after 4 h of incubation of selected enzyme preparations with variouspolysaccharide substratesa

Substrates"Source of enzymes

Cell walls Chitin Laninarin Pustulan

ChitinasesStreptomyces sp. Glc(tr), GlcNAc GlcNAc Glc(tr) None

(GlcNAc), (GlcNAc)0S. marcescens, GlcNAc GlcNAc None NoneCommercial (Calbiochem) Glc(tr), GlcNAc GlcNAc Glc(tr) None

(GlcNAc)2 (GlcNAc)2GlucanasesB. circulans (crude) Glc, GlcNAc(tr) None Glc, (Glc)0 GlcB. circulans (,8-1,3-glucan- Glc, GlcNAc(tr) None Glc, (Glc)0 None

ase)B. circulans (,8-1,6-glucan- None None None Glc

ase)Commercial (Calbiochem) Glc, GlcNAc(tr) None Glc, (Glc)0 Glc(tr)

PronasesCommercial (Sigma) Glc, (Glc)0 None Glc, (Glc),, NoneCommercial (Calbiochem) None None None Nonea Products were identified by paper chromatography. Monosaccharides and disaccharides were confirmed

by gas chromatography.b Abbreviations: Glc, glucose; (Glc)0, laminaribiose and higher oligomers of glucose; GlcNAc, N-acetyl-

glucosamine; (GlcNAc)2, chitobiose; (GlcNAc)0, higher oligomers ofN-acetylglucosamine; and tr, trace.

VOL. 15, 1977

on Novem

ber 22, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

982 DAVIS, DOMER, AND LI

TABLE 3. Quantities ofglucose and N-acetylglucosamine released from the yeast-phase cell

walls ofH. capsulatum by selected enzymes

Quantities released (mg/2 mgEnzymes of cell walls)

N-acetyl-D-glu-Glucose cosamine

Chitinase-glucanase 0.25a 0.75(Streptomyces sp.)

Chitinase (S. marcescens) 0.00 0.25,8-1,3-Glucanase (B. circu- 0.50 0.13

lans)a Sugars were detected and quantitated by gas

chromatography. See legend to Table 1 for condi-tions.

tivities were subsequently separated by columnchromatography on Sephadex G-100. Both par-tially purified glucanases were free of protease,lipase, and phosphatase activities. When cellwalls were hydrolyzed with partially purified,f-1,3-glucanase, 22 to 23% of the wall was re-leased as glucose and approximately 2% wasreleased asN-acetylglucosamine (Table 3). Pre-treatment of cell walls with the Serratia chiti-nase prior to attack by the 18-1,3-glucanase didnot result in increased amounts of glucosebeing released. The 13-1,6-glucanase did notrelease glucose from cell walls, and neither itnor the crude B. circulans enzyme from whichit was derived released gentiobiose. Both prepa-rations, however, released glucose and gentio-biose from pustulan after short periods of incu-bation. The 8-1,3-glucanase released glucoseand p8-1,3-oligomers of glucose from laminarinbut released only glucose from cell walls.A commercial laminarinase, which should

theoretically be active only on laminarin, wasactive both on laminarin and pustulan. It re-leased glucose and a trace of N-acetylglucosa-mine from the cell walls of H. capsulatum. Acrude pullulanase prepared from A. aerogeneswas ineffective against the cell walls.A commercial Pronase (Sigma P-5130) was

also found to contain 3-1,3-glucanase activity.It released laminaribiose from both laminarinand cell walls, but only trace amounts of glu-cose appeared in the hydrolysates, even after 48h of incubation. A second Pronase (Calbiochem53702), free of glucanase, released neither glu-cose nor N-acetylglucosamine from cell walls.Pretreatment of cell walls with the latter Pro-nase did not significantly enhance cell wall hy-drolysis by either glucanase or chitinase.Enzymatic hydrolyses with autologous en-

zymes. Initially, glucanase activity was dem-onstrated in the culture filtrate of the chemo-type I isolate ofH. capsulatum when it was in a

synthetic medium with laminarin as the solecarbon source. Subsequently, however, glucan-ase was found to be a constitutive enzyme, notrequiring laminarin for expression, and glucan-ase activity was demonstrated in the culturefiltrates of both the yeast and mycelial phasesof the same isolate as well as in the other twostrains examined, viz., an isolate of H. capsu-

latum chemotype II and one of B. dermatitidis.The culture filtrate of the mycelial phase of thechemotype I isolate of H. capsulatum containedapproximately 20% more activity than did thecorresponding yeast-phase culture filtrate.Moreover, protoplasts of the yeast phase of thatsame isolate liberated glucanase into the cul-ture medium as well. None of the fungi was

ever successfully cultivated using chitin as thesole carbon source, and chitinase activity was

never detected in cultures of fungi grown on

laminarin or glucose.Glucanase was also detected in the cell sap

derived from both yeast- and mycelial-phasecells of all three strains. The cells had beenthoroughly washed to remove any extracellularenzyme prior to breakage. The recovery of 8-1,3-glucanase activity, as measured by the hy-drolysis of laminarin, from the yeast-phasecells of the chemotype I strain of H. capsula-tum is presented in Table 4. It should be notedthat approximately 60% of the activity was notliberated into the culture medium. Glucanaseclosely associated with the cell wall was de-tected in suspensions of washed walls that hadbeen incubated for 7 days, so that at least someof the glucanase remained firmly attached tothe wall after homogenization of the wholecells. Glucose was present in the suspensions atthe end of the incubation period as well, per-haps the result of the autolytic property of theglucanase. Such activity, however, could onlybe detected with cell walls that had not beenlyophilized but had been incubated directly fora week after their preparation. The incubationof lyophilized cell wall preparations with en-zymes not capable of attacking the walls; i.e.,Pronase from Calbiochem andB-1,6-glucanasefrom B. circulans did not result in the release

TABLE 4. (3-1,3-Glucanase activity from yeast-phasecultures ofH. capsulatum chemotype I

Units (,mol ofSource of activity glucose from Protein Sp actlaminarin/30 (mg)a

min)

Culture filtrate 586.0 12.5 46.8Cell sap fraction 729.0 14.9 48.7Cell wall-associ- 61.0 4.3 14.2

ateda Lowry method (reference 29).

INFECT. IMMUN.

on Novem

ber 22, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

CELL WALLS OF HISTOPLASMA AND BLASTOMYCES 983

of glucose from the walls after 48 h of incuba-tion.

All ofthe fungal glucanases hydrolyzed lami-narin, a 83-1,3-glucan, but not pustulan, a 38-1,6-glucan, when tested over a wide range oftemperature and pH values. A comparison ofthe activity of one of the fungal glucanases,that from the yeast phase of the chemotype Iisolate of H. capsulatum with a 3-1,3-glucan-ase prepared from B. circulans, is presented inTable 5. It can be seen that both enzymes hy-drolyzed laminarin and a glycoprotein ex-tracted from the cell wall of H. capsulatumchemotype I and that neither hydrolyzed pustu-lan. Moreover, the enzyme isolated from B.circulans was moderately active on the cellwalls of the yeast phase of H. capsulatumchemotype I, whereas that isolated from thefungus released only trace amounts of glucosefrom its own cell wall. To determine the glycosi-dase activities of the crude glucanase prepara-tions, the enzyme preparations from culturefiltrates of all three strains were tested againsta battery of p-nitrophenyl derivatives (Table6). a-Glucosidase activity was not exhibited byany ofthe fungi. Both isolates ofH. capsulatummaximally hydrolyzed p-nitrophenyl-f3-D-glu-coside, and H. capsulatum chemotype I and B.dermatitidis 3489 significantly hydrolyzed p-ni-trophenyl-a-D-mannoside as well. The wide dis-parity of activities was suggestive of more thanone glycosidase being produced by each fungus.Protease activity was present in all prepara-tions, particularly in the cell sap fractions.The glucanase preparations from the culture

filtrates of all three fungi liberated glucose andlaminaribiose from the yeast phase cell walls ofboth isolates of H. capsulatum, but, at most,only a trace ofglucose could be removed fromB.dermatitidis cell walls by any of the enzymes.The glucanase prepared from H. capsulatumchemotype I hydrolyzed the glycoprotein frac-tion from its own cell walls and that from thecell walls of H. capsulatum chemotype II butfailed to hydrolyze the glycoprotein preparedfrom the walls of B. dermatitidis (Fig. 1). Theglucanase from H. capsulatum chemotype II(not pictured) exhibited the same pattern of

hydrolysis toward the glycoprotein fractions as

did the other two strains. The enzyme preparedfrom B. dermatitidis also hydrolyzed the glyco-protein fractions from both H. capsulatumstrains but not that prepared from its own

walls. All three enzyme preparations hydro-lyzed laminarin under the same conditions.N-acetylglucosamine and chitobiose were not

liberated from any cell walls incubated withany of the fungal glucanase preparations. Thehydrolysis of the glycoprotein fraction from thecell walls ofH. capsulatum chemotype I by itsown glucanase preparations was monitored atselected intervals during incubation, and it wasfound that laminaribiose and higher oligomersof glucose appeared after 1 or 2 h of incubationbut that laminaribiose and glucose were thepredominant sugars thereafter (Fig. 2). Thesame enzyme hydrolyzed laminarin even morevigorously when examined at the same timeintervals. Therefore, a,8-1,3-glucanase and a,8-glucosidase were present, but there may alsohave been an exo-f3-1,3-glucanase in the prepa-

ration. Glucose and laminaribiose were demon-strated by paper and gas-liquid chromatogra-phy in the hydrolysates ofthe fungal walls, butmannose, galactose, gentiobiose, and othermono- or disaccharides were not present.The glucanase obtained from the culture fil-

trate of the yeast-phase culture of H. capsula-tum chemotype I was active over a broad pHrange, but optimum activity occurred at pH 5.0(Fig. 3A). Enzymes stored for 1 h at selectedtemperatures before assay showed rapid inacti-vation above 60°C (Fig. 3B). Optimum hydroly-sis of laminarin occurred at 55°C (Fig. 3C). Theglucanase was inhibited by Hg2+ but not byethylenediaminetetraacetic acid. Two attemptsto purify the crude glucanase by chromatogra-phy on a Sephadex G-100 column were unsuc-

cessful, since /8-D-glucosidase activity was re-

covered and no (8-1,3-glucanase was found inthe eluent.

DISCUSSIONThe carbohydrate components ofthe cell wall

of the yeast phase of H. capsulatum chemotypeI were confirmed as (8-1,3-linked glucan and (8-

TABLE 5. Hydrolyses ofvarious substrates by ,f31 ,3-glucanases isolated from B. circulars or the yeast phase ofH. capsulatum chemotype I

SubstrateEnzyme source

Unextracted wallb Wall glycoprotein Laminarin Pustulan

H. capsulatum 0.05 0.75 1.05 0.00B. circulars 0.42 0.78 1.48 0.00

a Milligrams of glucose released per 2 mg of substrate after 24 h of incubation at 370C and pH 5.5 using 100,ug of protein per assay.

" Both substrates were isolated from H. capsulatum chemotype I.

VOL. 15, 1977

on Novem

ber 22, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

984 DAVIS, DOMER, AND LI

1,4-linked chitin, as suggested by Domer (12)and Kanetsuna et al. (23), inasmuch as dimersofboth ,8-1,3-linked glucose and ,3-1,4-linkedN-acetylglucosamine were isolated from cell wall

TABLE 6. Activity of enzymes isolated from filtratesofyeast-phase cultures on p-nitrophenyl substrates

Source of enzymes (/.mol of nitro-phenol/30 min)

p-Nitrophenyl H. capsulatumsubstrates B. dermati-

Chemo- Chemo- tidistype I type II

Phosphate 0.04 0.00 0.00N-Acetyl-,3-D-glu- 0.40 0.40 0.40

cosaminidea-D-Mannoside 1.60 0.00 1.60a-D-Glucoside 0.00 2.00 0.00/B-D-Glucoside 2.80 2.00 0.04a-D-Galactoside 0.80 0.00 0.01,8-D-Galactoside 0.00 0.00 0.00

a One unit was equivalent to 0.1 ,umol of nitrophe-nol/30 min at 370C at pH 5.5 with 25 ,ug ofprotein perassay.

H ISTO PLAS MA

GkcNAc E NZYME

hydrolysates and identified by paper and gas-liquid chromatography. No a-1,4- or a-1,6-linked glucose was detected, and a-glucan hasnot been found by others (12, 23). H. capsula-tum chemotype II walls, on the other hand,appear to contain a mixture of a- and 8-glucans(23), and although there was no enzymatic datapresented here to confirm those conclusions,the acid solubility characteristics of the cellwalls would support that contention (12).The architectural arrangement of the chitin

and glucan within the cell walls remains un-solved, although Kanetsuna et al. (23) sug-gested that there was an inner layer of chitin inthe wall of the yeast phase of H. capsulatum.Our own studies indicated that at least part ofthe N-acetylglucosamine in the yeast-phasewall was linked to glucan in some fashion,since a small amount of N-acetylglucosaminewas released from the cell walls when theywere hydrolyzed by a glucanase free of chiti-nase activity and since maximal amounts ofN-acetylglucosamine were released only in thepresence of both glucanase and chitinase. On

BLA STOMYCES

E N Z Y ME

GI c

Lamb

G e n b

Std B.dexrm H.capsII

H.caps La m B.derm H.caps H.capsI

L amII I

FIG. 1. Hydrolysis of wall glycoprotein from the yeast phases ofH. capsulatum chemotype I, H. capsulatumchemotype II, and B. dermatitidis 3489 by the extracellular glucanases prepared from H. capsulatum chemo-type I and B. dermatitidis 3489. Substrates were incubated with the enzyme at 370C, pH 5.5, for 24 h. Abbre-viations: Glc, glucose; GlcNAc, N-acetylglucosamine; Lamb, laminaribiose; Genb, gentiobiose; I, II, B. derm,glycoproteins extracted from H. capsulatum chemotypes I and II and B. dermatitidis, respectively; Lam,laminarin; Std, standard.

INFECT. IMMUN.

on Novem

ber 22, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

CELL WALLS OF HISTOPLASMA AND BLASTOMYCES 985

LA MI NARI N G LYC O PR OT E IN

A.06L.A~i.-il

Std 0 1 2 4 0 2 4H O U R S

FIG. 2. Hydrolysis of laminarin and wall glycoprotein from the yeast phase ofH. capsulatum chemotype Iby a glucanase prepared from the culture filtrate of the yeast phase ofH. capsulatum chemotype I. Substrateswere incubated with the enzyme for 0, 1, 2, and 4 h at 379C, pH 5.5. Abbreviations: Glc, glucose; GlcNAc, N-acetylglucosamine; Lamb, laminaribiose; Genb, gentiobiose; Std, standard.

60 ~

.0~~~~~

5 6 7 25 so0 75 25 so is

pH IE"PfR^TU~~f (C PER.("TUal to

FIG. 3. Effect of various pH values (A), varioustemperatures of storage (B), and various tempera-tures ofincubation (C) on the extracellularglucanaseprepared from H. capsulatum chemotype I.

the contrary, maximal amounts of glucose werereleased in the presence or absence of chitinase.Protein, if present, did not interfere with therelease of maximal quantities of glucose and N-acetylglucosamine, since prior treatment of thecells with protease did not alter the amounts ofeither sugar released. An outer layer of glucanhas been reported for two other fungi, Neuro-spora (21) and Aspergillus (36), but many ofthestudies designed to determine the cell wallstructure ofH. capsulatum with the use of lyticenzymes were complicated by the use ofenzyme

preparations containing extraneous activities(12, 14, 23). In fact, most of the commercialenzymes tested in this study were contami-nated by one or more additional activities.B. dermatitidis and representatives of chem-

otypes I and II ofH. capsulatum produced 13-1,

3-glucanase and several glycosidases constitu-tively when growing in either the yeast or my-celial phase. j8-1,6-Glucanase, chitinase, 13-glu-canase, and a-glucosidase could not be de-tected. One might expect to find f8-1,3-glucan-ase in both phases of the two chemotypes of H.capsulatum, since each has 13-glucan in itsyeast-phase cell (14) and some hydrolysis ofglucan would be expected to occur during bud-ding or hyphal extension, but it was somewhatsurprising to find high levels of /3-1,3-glucanaseactivity in the yeast phase of B. dermatitidis,since its cell walls contain little /8-glucan. Infact, Kanetsuna et al. (22) found less than 5% ofthe glucan in the cell wall of the yeast phase ofB. dermatitidis 3489 to be 13-linked. Perhapsthe ability to produce such high levels of 13-

glucanase activity in the virtual absence of 8-glucan in the cell wall represents an evolution-ary link between Histoplasma and Blasto-

G I cNAc

G I c

L a mb

G e n b

VOL. 15, 1977

I

on Novem

ber 22, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

986 DAVIS, DOMER, AND LI

myces. We anticipated finding a-glucanase aswell as /8-glucanase activity in the yeast phaseof H. capsulatum chemotype II, since the cellwalls of that strain contain both a- and 38-glucans. Further, glucan in the cell walls oftheyeast phase of B. dermatitidis is 95% a-linked(22), so one would also expect to find a-glucan-ase activity there.

Autolytic glucanases are produced by a widevariety of fungi (15, 16, 18, 20, 32; A. E. Stoneand B. A. Stone, Biochem. J. 96:793, 1965), andthe glucanases detected in H. capsulatum andB. dermatitidis are similar to those demon-strated in other fungi in that they are producedconstitutively and are most active at moder-ately high temperatures. Since the 13-glucan-ases are more active at temperatures compati-ble with yeast-phase growth, their activity mayresult in diminished amounts of 8-glucan in theyeast wall when compared to that of the myce-lial-phase cell wall. Using electron microscopy,Carbonell (9) observed the migration of vesiclesinto the area of regenerating cell walls of H.capsulatum and showed by chemical analysisthat the newly formed cell wall contained chitinbundles and random strands of glucan. Kregerand Kopecka (25), studying regeneration of cellwalls ofSaccharomyces cerevisiae, carried suchstudies one step further and found that thenewly formed chitin and glucan were resistantto attack by autolytic enzymes. Such enzymeshydrolyzed the intact wall, however, and itwould appear that part ofthe cell wall glucan ofthe yeast might always be resistant to autolyticglucanases. It was demonstrated here, for ex-ample, that intact cell walls of H. capsulatumwere relatively resistant to autologous 18-glu-canases but that extraction of the walls withethylenediamine yielded a fraction that washighly susceptible to attack.Morphogenesis of the fungal cell wall is

thought to require the cooperation of both lyticand synthetic enzymes (7, 27), and glucanasesbut not chitinases have been found to be local-ized on or near the fungal cell wall (16, 18).Moreover, Sentandreu et al. (35) isolated a glu-can synthetase from cultures of S. cerevisiae.Bacon (2) in a recent review suggested that thepresence of enzymes involved in glucan hydrol-ysis may indicate that the state of the glucanand not chitin determines the mechanical prop-erties and integrity of the wall.There exists some disagreement in the litera-

ture on the percentage of hexose in the yeast-phase cell wall of H. capsulatum (1, 12, 23, 24),but there is general agreement that the yeast-phase cell walls of H. capsulatum chemotype Iand B. dermatitidis contain less 8-glucan thanthe mycelial-phase cell walls. Since H. capsula-

tum chemotype I strains do not possess a-glu-cans in their walls, one might assume thatdecreasing amounts of a-glucan in the yeast-phase wall are replaced by chitin. In chemotypeII cell walls, a-glucan is replaced by f-glucanand/or chitin when growing in the yeast phase.A similar replacement might also be true forother dimorphic fungi. DeVries and Wessels(11) have demonstrated that the synthesis of f3-glucan is required for the initiation of hyphalmorphogenesis in Schizophyllum commune,but whether this is due to decreased synthesis,increased hydrolysis, or a combination of bothhas yet to be determined. One might speculatethat extracellular or cell wall-associated glu-canases result in lesser amounts of 13-glucan inthe yeast phases of Histoplasma and Blasto-myces isolates and could play a role in wallmorphogenesis. Alteration in glucan composi-tion could affect cell wall thickness or stackingof wall components, or possibly even be a factorin determining whether the cell grows as ayeast or mycelium. However, since the diver-sity of fungal cell wall composition is wellknown (4, 30, 39), it is probable that the archi-tectural arrangement of components and theenzymes involved in morphogenesis are equallydiverse.Further examination of the extracellular glu-

canases of these pathogenic fungi is warrantedbecause: (i) characterization of individual gly-cohydrolases could offer insight into cell wallstructure and (ii) purified glucanases would beuseful in the preparation of cell wall glucans forlinkage analysis. Moreover, the manipulationof enzyme activity in growing cells could pro-vide information on the mechanisms involvedin dimorphism. Recently, a purified enzyme ofSchistosoma mansoni was used as a skin testreagent for the diagnosis of schistosomiasis(34), and since extracellular glucanases such asthose produced by H. capsulatum and B. der-matitidis might stimulate the defense mecha-nism of the infected host, the purified enzymescould potentially be excellent antigens for de-tection of infection.

ACKNOWLEDGMENTSThis investigation was supported by Public Health Serv-

ice research grant AI-08588 from the National Institute ofAllergy and Infectious Diseases. T. D. was a mycologytrainee supported by Public Health Service training grant5-TO1-AI00003 from the National Institute of Allergy andInfectious Diseases.

LITERATURE CITED

1. Anderson, K. L., R. W. Wheat, N. R. Conant, and W.Clingenpeel. 1974. Composition of the cell wall andother fractions of the autolyzed yeast form of Histo-plasma capsulatum. Mycopathol. Mycol. Appl.54:439-451.

INFECT. IMMUN.

on Novem

ber 22, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

CELL WALLS OF HISTOPLASMA AND BLASTOMYCES 987

2. Bacon, J. S. D. 1973. The contribution of 83-glucanasesto the lysis of fungal cell walls, p. 61-74. In J. R.Villanueva, I. Garcia-Acha, S. Gascon, and F. Uru-buru (ed.), Yeast, mould, and plant protoplasts. Aca-demic Press Inc., New York.

3. Bailey, R. W., and J. B. Pridham. 1962. The separationand identification of oligosaccharides. Chromatogr.Rev. 4:114-136.

4. Bartnicki-Garcia, S. 1968. Cell wall chemistry, mor-phogenesis, and taxonomy of fungi. Annu. Rev.Microbiol. 22:87-108.

5. Bender, H., and K. Wallenfels. 1966. Pullulanase (anamylopectin and glycogen debranching enzyme) fromAerobacter aerogenes, p. 555-559. In E. F. Neufeldand V. Ginsburg (ed.), Methods in enzymology, vol.8. Academic Press Inc., New York.

6. Block, R. J., R. LeStrangel and G. Zweig. 1952. Paperchromatography, a laboratory manual. AcademicPress Inc., New York.

7. Cabib, E. 1975. Molecular aspects of yeast morphogene-sis. Annu. Rev. Microbiol. 29:191-214.

8. Carbonell, L. M., M. D. Berliner, and F. Gil. 1972.Kinetic and morphological observations on the yeastphase of Histoplasma capsulatum during protoplastformation. J. Bacteriol. 110:731-738.

9. Carbonell, L. M., F. Gil, and F. Yegres. 1973. Cell wallregeneration of protoplasts of Histoplasma capsula-tum, p. 93-103. In R. J. Villanueva, I. Garcia-Acha,S. Gascon, and F. Uruburu (ed.), Yeast, mould, andplant protoplasts. Academic Press Inc., New York.

10. Chambers, R. E., and J. R. Clamp. 1971. An assess-ment of methanolysis and other factors used in theanalysis of carbohydrate-containing materials. Bio-chem. J. 125:1009-1018.

11. DeVries, 0. M. H., and J. G. H. Wessels. 1975. Chemi-cal analysis of cell wall regeneration and reversion ofprotoplasts from Schizophyllum commune. Arch. Mi-crobiol. 102:209-218.

12. Domer, J. E. 1971. Monosaccharide and chitin contentof cell walls of Histoplasma capsulatum and Blasto-myces dermatitidis. J. Bacteriol 107:870-877.

13. Domer, J. E., and J. G. Hamilton. 1971. The readilyextracted lipids ofHistoplasma capsulatum and Blas-tomyces dermatitidis. Biochim. Biophys. Acta231:465-478.

14. Domer, J. E., J. G. Hamilton, and J. C. Harkin. 1967.Comparative study of the cell walls of the yeastlikeand mycelial phase of Histoplasma capsulatum. J.Bacteriol. 94:466-474.

15. Farkas, V., P. Biely, and S. Bauer. 1973. Extracellular,8-glucanases of the yeast Saccharomyces cerevisiae.Biochim. Biophys. Acta 321:246-255.

16. Fleet, G. H., and H. J. Phaff. 1974. Glucanases inSchizosaccharomyces. J. Biol. Chem. 249:1717-1728.

17. Fleet, G. H., and H. J..Phaff. 1974. Lysis of yeast cellwalls: glucanases from Bacillus circulans WL-12. J.Bacteriol. 119:207-219.

18. Fleet, G. H., and H. J. Phaff. 1975. Glucanases inSchizosaccharomyces. Isolation and properties of anexo-,8-glucanase from the cell extracts and culturefluid of Schizosaccharomyces japonicus var. versatil-lis. Biochim. Biophys. Acta 401:318-332.

19. Gull, K., P. M. Moore, and A. P. J. Trinci. 1972. Prepa-ration of protoplasts from Goetrichum lactis and useof fluorescence to detect cell walls. Trans. Br. Mycol.Soc. 59:79-85.

20. Holten, V. Z., and S. Bartnicki-Garcia. 1972. Intracel-

lular f3-glucanase activity of Phytophthora palmi-vora. Biochim. Biophys. Acta 276:221-227.

21. Hunsley, D., and J. H. Burnett. 1970. The ultrastruc-tural architecture of the walls of some hyphal fungi.J. Gen. Microbiol. 62:203-218.

22. Kanetsuna, F., and L. M. Carbonell. 1971. Cell wallcomposition of yeastlike and mycelial forms ofBlasto-myces dermatitidis. J. Bacteriol. 106:946-948.

23. Kanetsuna, F., L. M. Carbonell, F. Gil, and I. Azuma.1974. Chemical and ultrastructural studies on the cellwalls of Histoplasma capsulatum. Mycopathologia54:1-13.

24. Kobayashi, G. S., and P. L. Guiliacci. 1967. Cell wallstudies of Histoplasma capsulatum. Sabouraudia5:180-188.

25. Kreger, D. R., and M. Kopecka. 1975. On the natureand formation of the fibrillar nets produced by proto-plasts ofSaccharomyces cerevisiae in liquid media: anelectron-microscopic, X-ray diffraction and chemicalstudy. J. Gen. Microbiol. 92:207-220.

26. Laine, R. A., W. J. Esselman, and C. C. Sweeley. 1972.Gas-liquid chromatography of carbohydrates, p. 159-167. In D. B. McCormick and L. D. Wright (ed.),Methods in enzymology, vol. 18. Academic Press Inc.,New York.

27. Lampen, J. 0. 1968. Yeast cell wall. Antonie van Leev-enhoek J. Microbiol. Serol. 34:1-18.

28. Li, Y.-T., and M. R. Shetlar. 1964. Occurrence of a-galactosidase from Calvatia cyathiformis. Arch. Bio-chem. Biophys. 108:523-530.

29. Lowry, 0. H., N. Rosebrough, A. Farr, and J. Randall.1951. Protein measurement with the Folin phenolreagent. J. Biol. Chem. 193.265-275.

30. Matile, P., H. Moor, and C. F. Robinow. 1969. Yeastcytology, p. 220-302. In A. H. Rose and J. S. Harrison(ed.), The yeasts, vol. 1. Academic Press Inc., NewYork.

31. Monreal, J., and E. T. Reese. 1969. The chitinase ofSerratia marcescens. Can. J. Microbiol. 15:689-696.

32. Page, W. J., and J. J. Stock. 1974. Sequential action ofcell wall hydrolases in the germination and out-growth of Microsporum gypseum macroconidia. Can.J. Microbiol. 20:483-489.

33. Reynolds, D. M. 1954. Exocellular chitinase from aStreptomyces sp. J. Gen. Microbiol. 11:150-159.

34. Senft, A. W., and S. E. Maddison. 1975. Hypersensitiv-ity to parasite proteolytic enzyme in schistosomiasis.Am. J. Trop. Med. Hyg. 24:83-89.

35. Sentandreu, R., M. V. Elorza, and J. R. Villaneuva.1975. Synthesis of yeast wall glucan. J. Gen. Micro-biol. 90:13-20.

36. Skujins, J. J., H. J. Potgieter, and M. Alexander. 1965.Dissolution of fungal cell walls by a Streptomycetechitinase and P-1,3-glucanase. Arch. Biochem. Bio-phys. 111:358-364.

37. Skujins, J., A. Pukite, and A. D. McLaren. 1970. Chiti-nase of Streptomyces sp: purification and properties.Enzymologia 39:353-370.

38. Stimson, W. H. 1971. Gas-liquid chromatography ofhexosamines. FEBS Lett. 13:17-20.

39. Taylor, I. E. P., and D. S. Cameron. 1973. Preparationand quantitative analysis of fungal cell walls: strat-egy and tactics. Annu. Rev. Microbiol. 27:243-259.

40. Wolfrom, J. L., and N. E. Franks. 1965. Partial acidhydrolysis. Preparation of polymer-homologous oligo-saccharides; xyloglycoses from xylan. Methods Car-bohydr. Chem. V:276-280.

VOL. 15, 1977

on Novem

ber 22, 2020 by guesthttp://iai.asm

.org/D

ownloaded from