Regulation of serine-type exoproteinases by endogenous inhibitors present in exoantigens of the mycelial form of Paracoccidioides brasiliensis

Regulation of serine-type exoproteinases by endogenousinhibitors present in exoantigens of the mycelial form ofParacoccidioides brasiliensis

E. A. ZAMBRANO, I. RODRIÂGUEZ, M. MENDOZA, C. SANTAELLA, M. LOÂ PEZ, E. DIÂAZ & M. ALBORNOZ

SeccioÂn de MicologõÂa, Instituto de Biomedicina, Facultad de Medicina, Universidad Central de Venezuela, Caracas, Venezuela

We have partially characterized some biochemical properties of exoproteinasessecreted into culture medium by the mycelial form of Paracoccidioides brasiliensis, adimorphic fungus that causes human disease in Latin America. Proteinase activitywas analyzed in solid- and liquid-phase systems using zymography and Azocoll,respectively. Minimal or no gelatinase activity was observed by zymography in thecrude �ltrates among proteins with a relative mobility greater than 200 kDa. Whenthe crude �ltrate was fractionated by isoelectric focusing or ion exchangechromatography, we observed striking activation of gelatinases, both those of highapparent molecular mass and alkaline isoelectric points (pI), as well as those oflower molecular mass and acidic pI. The apparent high molecular mass gelatinases,pI 10, showed optimal activity at pH 7¢0. They were totally inhibited byphenylmethylsulfonyl�uoride and partially inhibited by incubation with previouslyneutralized fractions of pI 5¢4 and 6¢1. The latter inhibition could be reversed byexposure to 10% isopropanol. These results provide evidence of regulatorymechanisms controlling proteinase activity in secreted proteins. The principalmechanism appears to be the formation of reversible complexes with endogenousinhibitors.

Keywords endogenous inhibitors, exoproteinases, Paracoccidioides brasiliensis,regulation

Introduction

Paracoccidioides brasiliensis (Pb) is a dimorphic fungusthat causes paracoccidioidomycosis, an endemic diseasecharacterized by multiple clinical forms affecting humansin Latin America [1–3]. The biochemical characteristicsof proteinases present in crude exoantigens of the fungushave received relatively little attention.

As important biochemical components of all micro-organisms, proteinases represent a diverse group ofenzymes that catalyze the hydrolysis of peptide bonds inproteins. In addition to their role in nutrition andcatabolism of proteins, they may be pathogenicity factors

and may in�uence ecological distribution [4–6]. In recentyears the important role of proteinases as specializedregulatory enzymes controlling physiological processeshas been demonstrated in many organisms. This reg-ulatory activity requires ef�cient control mechanisms,fundamental for maintenance of biological integrity [7].

The proteinases of Pb have not been well de�ned.Recently, a study of intracellular proteinases hasrevealed serine-, cysteine- and metallo-proteinases inboth mycelial and yeast forms of the microorganisms;their activity was greater in the mycelial forms [8].Proteinases secreted into culture media have beendetected in both mycelial and yeast cultures; however,they have been characterized only partially [9,10]. Anexoproteinase of mixed serine-thiol type has beenreported in cultures of the yeast form [11], but theexoproteinases of the mycelial form have not beencharacterized.

ã 2001 ISHAM, Medical Mycology, 39, 359±368

Medical Mycology 2001, 39, 359±368 Accepted 14 July 2000

Correspondence: Dr Edgar Armando Zambrano, Seccion deMicologõ a, Instituto de Biomedicina, Apartado 4043, Caracas1010A, Venezuela. Tel.: ‡58 212 8625326; fax: ‡58 212 8611258;e-mail: [email protected]

ã 2001 ISHAM

at IVIC

on August 22, 2014

http://mm

y.oxfordjournals.org/D

ownloaded from

http://mmy.oxfordjournals.org/

In the present study, the effect of isoelectric focusingon the activation of proteinases present in crudeexoantigens was investigated. Our objective was toanalyze possible regulatory functions as well as bio-chemical properties of the proteinases present in crudeexoantigens of the mycelial form of Pb, includingisoelectric point, susceptibility to classical inhibitors,and pH-dependency of activity.

Materials and methods

Cultivation, harvesting of ®ltrates and preparation ofcrude exoantigens

The study was carried out in culture �ltrates of isolate B-339 of Pb, kindly provided by Dr A. Restrepo, Medellin,Colombia and maintained in our laboratory by passageon Sabouraud agar. Four fragments of about 0¢5 cm3

from a Sabouraud agar culture were transferred to125-ml �asks containing 50 ml of Sabouraud broth (2%glucose, 1% peptone) or Medium 199 (M199; GibcoBRL, Grand Island, NY, USA) and subcultured twicefor 7 days at room temperature with shaking at 100 rpm.On day 7 of the second subculture, 5 ml of the mycelialsuspension was transferred to 500-ml �asks containing125 ml of Sabouraud broth supplemented with 18 mM

asparagine and 430 m M thiamine-HCl (Sabouraud thia-mine asparagine medium [STA]) [2] and M199 supple-mented with 4 mM NaHCO3 [12]. The cultures in STAwere harvested after 21 days, and after 35 days in M199.Filtrates were obtained by successive passage throughWhatman No. 1 �lter paper (Whatman, Spring�eld Mill,UK) and Millipore (Millipore Corporation, Bedford,MA, USA) membranes with 0¢45 m m pores. The sterile�ltrates were dialyzed against 40 volumes of distilledH2O with three changes at 8-h intervals using dialysistubing with a 10 kDa cut-off. The preparations wereconcentrated by pervaporation and the protein concen-tration was determined by the Lowry method [13]. Theresults summarize data from �ve independent exoanti-gen preparations, three from growth STA and two fromM199.

SDS-PAGE and Western blot

The presence of gp43 (the main serum diagnosticexoantigen of paracoccidioidomycosis) was corroboratedby sodium dodecyl sulfate-polyacrylamide gel electro-phoresis (SDS-PAGE) (10%) [14] and Western blot [15].Brie�y, 150 m g of crude exoantigen was solubilized inbuffer containing 62¢5 mM Tris-HCl (Sigma Co, St. Louis,MO, USA), pH 6¢8, 25% v/v glycerol, 2% w/v SDS and0¢01% w/v bromphenol (Sigma), a formula referred to as‘sample buffer’ throughout this paper, to which 5% v/v

2-mercaptoethanol (2-ME) was added in the SDS-PAGEprotocol. The samples were heated for 3 min at 90 oC.High range protein standards (Gibco BRL, GrandIsland, NY, USA) within the range 14¢3–200 kDa wererun simultaneously to estimate molecular weight (MW)of the exoantigens. For the immunoblotting assay,polyclonal rabbit anti-gp43 antibodies diluted 1:1000 orsera from patients with paracoccidioidomycosis diluted1:100 in 6¢7 mM phosphate buffered saline, pH 7¢2(0¢85% w/v NaCl) with 0¢1% v/v Tween 20 (PBST) wereused.

Gelatinase activity

Proteinase activity was investigated by SDS-PAGE usinggelatin as the proteinase substrate (an assay hereaftercalled ‘gel zymography’) [14,16]. Activity was measuredin 150 m g of protein from the crude �ltrates and in 50 m gof the fractions described below. The samples weredissolved for 1 h at room temperature in sample bufferwithout 2-ME. Gelatin (Sigma) was co-polymerized at0¢2% in 8% SDS-PAGE gels. After electrophoreticseparation, the gels were washed for 30 min at 4 oC withTriton-X100 (Sigma) at 2¢5% v/v and three 10-minwashes with cold dH2O. They were then incubated for18 h at 35 oC in incubation buffer, 100 mM N-2-hydro-xyethylpiperazine-N 0-2-ethane sulfonic acid (HEPES),pH 7¢0; 1 mM CaCl2; 0¢025% NaN3. Proteinase activitywas revealed as clear zones after gel staining with 0¢1%Coomassie Brilliant Blue R-250 (Gibco BRL). Highrange standards (Gibco BRL), dissolved in sample buffercontaining 2-ME and heated for 3 min at 90 oC, wereincluded in all zymograms to estimate the apparentrelative mobility (Mr) of the gelatinases under theexperimental conditions. The Mr is the mobility of theproteinase with reference to a marker protein, and hasalso been referred to as apparent molecular mass. Theposition of these protein bands could be identi�ed bypartially decolorizing the gels.

Isoelectric focusing

To determine the isoelectric point (pI) of the proteinasesdetected in the culture �ltrates, isoelectric focusing wascarried out under native conditions in a mini-Rotoforapparatus (Bio-Rad Laboratories, Hercules, CA, USA)according to the manufacturer’s instructions. Brie�y,90–180 mg of protein were dissolved in 2% v/v of wide-range ampholytes, pH 3¢5–10, 15% v/v ultra-pureglycerol and double-distilled H2O, to a �nal volume of20 ml. Focusing was carried out at 4 oC, 12 W constantpower. Variations of current and voltage were registeredeach 30 min. Stabilization of the current and voltage wasused to determine the endpoint of focusing, usually 3¢5 h.

360 Zambrano et al.


at IVIC

on August 22, 2014

http://mm


ownloaded from


The Rotofor fractions were collected and pH wasdetermined. The distinct Rotofor fractions, each contain-ing 50 m g of protein, were analyzed by two-dimension gelzymography after solubilization for 1 h at room tem-perature in sample buffer without 2-ME.

pH dependency of activity

The dependence of proteinase activity on pH wasmeasured for pooled Rotofor fractions with pI 9 and10. The pH pro�les were investigated by zymographyusing a preparative comb. Forty micrograms of totalpooled protein were dissolved in sample buffer without2-ME and separated in an 8% acrylamide gel containing0¢2% gelatin. After separation and washing, the gel wasdivided into six portions 1¢3 cm wide by 5¢5 cm long. Thegel strips were incubated for 18 h at 35 oC in separatebuffer solutions with pH intervals between 4¢0 and 9¢0,using 100 mM buffers which also contained 1 mM CaCl2and 0¢025% NaN3. The buffers were as follows: pH 4¢0and 5¢0, acetate/acetic acid; pH 6¢0, phosphate buffer;pH 7¢0 HEPES; pH 8¢0 and 9¢0, Tris-HCl. Additionally,the hydrolysis of azo-dye-impregnated collagen (Azo-coll; Sigma) was studied as described by Chavira et al.[17]. The pooled protein samples were incubated at37 oC for 18 h with shaking. The absorbance of thesupernates at 520 nm was measured in triplicate inreference to spontaneous hydrolysis of Azocoll in thecorresponding buffers in the absence of proteinases. Aunit of activity (UA) was de�ned as the quantity ofenzyme capable of producing an increase of absorbanceof 0¢01 per hour, and the units of speci�c activity (SA)were de�ned as UA per milligram of protein.

Susceptibility to classical proteinase inhibitors

Crude exoantigens and Rotofor fractions containing 150and 50 m g of protein, respectively, were incorporatedinto a �nal volume of 0¢02 ml of a speci�c incubationbuffer (20 mM HEPES, pH 7¢0; 1 mM CaCl2; 0¢025%NaN3) and were pre-incubated with speci�c proteinaseinhibitors for 1–12 h at 4 oC. After pre-incubation thesamples were solubilized in sample buffer without 2-MEfor 1 h at room temperature and then analyzed byzymography as described above. The following protei-nase inhibitors were studied [18,19]: phenylmethyl-sulphonyl �uoride (PMSF), ranging in concentration0¢1–10 mM in the presence or absence of 10 mM

dithiothreitol (DTT) for serine-type proteinases; trans-epoxysuccinyl-L-leucylamido-[4-guanidino]butane (E64)ranging 0¢01–0¢1 mM as well as iodoacetic acid at5–10 mM for cysteine-type proteinases; Pepstatin A0¢1–5 m M for aspartic-type proteinases; and o-phenan-throline 1–5 mM for metalloproteinases. PMSF, Pepsta-

tin A and o-phenanthroline were dissolved in dimethylsulfoxide (DMSO). The volume of the inhibitors did notexceed 5% of the �nal volume. Controls consisting of 5%DMSO alone were used to rule out action of this solventon the proteinases. The zymograms were analyzed bydensitometry, using both a densitometer model GS-690and Multi-Analyst Ò software (Bio-Rad).

Effect of pre-incubation on proteinase activity of crudepreparations at distinct pH levels

Crude preparations of exoantigens produced in STAwhich had been stored at ¡20 oC for periods rangingfrom 1 week to 1 year were analyzed. A 150 m g sample ofcrude exoantigen was pre-incubated in separate buffersolutions with the pH ranging 4¢0–10¢0. The buffers(10 mM) were prepared as described above; additionally,carbonate/bicarbonate buffer, pH 10, was used. Pre-incubation was carried out in a volume of 0¢02 ml at 4 oCovernight. After pre-incubation, samples were solubi-lized in sample buffer without 2-ME for 1 h at roomtemperature and then analyzed by zymography at pH 7¢0as described.

Effect of Rotofor fractions of acidic pI on gelatinases ofpI 10

Rotofor protein fractions in the pI range 3¢0–6¢5 wereobtained from cultures grown in M199. Equal quantitiesof these fractions were pre-incubated for 3 h at roomtemperature with a fraction of pI 10. The pre-incubationwas carried out in a volume of 0¢02 ml of sample bufferwithout 2-ME. In other studies, 0¢5 ml volumes ofRotofor fractions with pI 5¢4 and 6¢1 were subjected toa process in which the incubation buffer was changed.A 0¢2 ml aliquot of each of the Rotofor fractions of pI5¢4, 6¢1 and 10¢0 were washed four times with a total ofsix volumes of buffer, pH 7¢0, containing 10 mM ultra-pure HEPES (Calbiochem, San Diego, CA, USA), 1 mM

CaCl2 and 0¢025% NaN3, using Ultrafuge centrifuge�lters (OSMONICS; Micron Separations Inc., St Paul,MN, USA) 0¢5 ml capacity with a cut-off of 10 kDa, spunat 2000 g for 45 min. Equivalent quantities of proteinretained by the �lter of the fractions of pI 5¢4 and 6¢1, aswell as the �ltrate, were pre-incubated with fraction pI 10in a volume of 0¢02 ml of 20 mM HEPES buffer,pH 7¢0 for 2–3 h at 37 oC, with gentle agitation. Finally,0¢01 ml of sample buffer without 2-ME was added andthe results were analyzed by zymography. The corre-sponding controls consisted of samples of the pI 10¢0fraction alone, pre-incubated in 20 mM HEPES buffer,pH 7¢0, as well as in phosphate buffer, pH 6¢0, and inacetic acid/acetate buffer, pH 5¢0. The last two bufferssimulated the conditions of the more acidic isoelectricpoints.

Regulation of serine proteinases of P. brasiliensis 361


at IVIC

on August 22, 2014

http://mm


ownloaded from


Ion exchange chromatography

A column 1 cm in diameter was prepared using 1 ml ofdiethylaminoethyl (DEAE)-sepharose (Sigma), pre-equilibrated with 50 mM Tris-HCl, pH 7¢0; 90 mg ofcrude concentrated �ltrate was fractionated. Elution wasmonitored with a UV detector (280 nm) adjusted to asensitivity of 0¢5 units of absorbance (optical density).Crude �ltrate was passed through DEAE and unboundproteins (�ow-through) were collected. The proteinsbound to the exchanger were eluted with a step-wisesaline gradient corresponding to 20, 30, 40, 50, 70, 80,100, 150, 200, 300, 400, 500, 700 and 1000 mM of NaCl in50 mM Tris-HCl buffer, pH 7¢0, at a �ow of 1¢6 ml min¡1.A total of 32 ml of buffer was used in each step.Fractions of 8 ml corresponding to the different zones ofmaximum protein concentration (peaks) were collected,dialyzed against dH2O using membrane tubing with acut-off of 7¢0 kDa (PIERCE Snake Skin, Rockford, IL,USA) and concentrated with polyethylene glycol. Thefractions were �nally supplemented with 10 mM NaCland 1 mM CaCl2 and analyzed by SDS-PAGE andzymography.

Effect of isopropanol

Portions of fractions of pI 10¢0 that were pre-incubatedwith pI 5¢4 or 6¢1 fractions as described above, were post-treated with isopropanol in concentrations from 2¢5 to10% v/v in a volume of 0¢02 ml of 20 mM HEPES buffer,pH 7¢0, for 3 h at 37 oC. Finally, 0¢01 ml of sample bufferwithout 2-ME was added and the products analyzed byzymography. Additionally, 150 m g of the crude exoanti-gen was incubated under the same conditions. Thecorresponding controls consisted of the pI 10¢0 protei-nase fraction pre-incubated with 10% isopropanol as

well as and crude exoantigen pre-incubated in 20 mM

HEPES buffer, pH 7¢0, without isopropanol.

Results

The presence of gp43 as the predominant protein in allexoantigens produced in both media was con�rmed bySDS-PAGE (10%) and immunoblotting (data notshown). The one-dimensional zymogram correspondingto these preparations revealed, in some cases, minimalgelatinase activity limited to the top of the gel; in others,no gelatinase activity was detectable (data not shown).

Figure 1a shows a two-dimensional zymogram corre-sponding to exoantigen produced in STA, Rotofor-separated fractions, pH 3¢1–5¢2. The unfocused sampleappears in lane C, and fractions in adjacent lanes are asindicated in the �gure. Several gelatinases are observedwhich were not detectable in the crude preparation. Themost obvious are indicated and are differentiated byrelative mobility (Mr) and pI. Activity was detected inthe following fractions: Mr near 200 kDa and pI between4¢0 and 5¢2; Mr 97 kDa and pI between 3¢1 and 3¢9; Mr

43 kDa and pI between 3¢1 and 3¢9. The gelatinaseactivity of fractions with a pI between 5¢6 and 10¢2 isshown in Figure 1b. Figure 1c shows the upper portion ofthe two-dimensional zymogram loaded with exoantigenproduced in M199. In this zymogram, as in that derivedfrom STA growth, the proteinases of apparent highmolecular mass (low mobility) are evident beginningwith the fraction of pI 4¢6. Gelatinase activity isintensi�ed in the basic fractions of pI 9¢0–10¢3. Incontrast to the exoantigen produced in STA, that fromM199 showed little gelatinase activity in the mobilityzone between 97¢4 and 43¢0 kDa protein standards (datanot shown).

362 Zambrano et al.


Fig. 1 Two-dimensional zymograms of P. brasiliensis exoantigens isolated from STA and M199. (a and b) STA fractions produced byisoelectric focusing, pH indicated above each lane; C, crude exoantigen. (c) Upper portion of zymogram with M199 exoantigen, pH indicatedabove each lane; C, crude exoantigen. MW, molecular weight of standard proteins.

at IVIC

on August 22, 2014

http://mm


ownloaded from


In the two-dimensional zymograms shown in Figure 1band 1c, gelatinases are seen as three bands of apparentmolecular mass greater than the 200 kDa standard, witha pI between 9 and 10. One had an Mr near that of the200 kDa standard; the other two are referred to asproteinases of apparent molecular mass 240 kDa and 270kDa. The greater activity shown by the proteinases oflow mobility in the zymograms compared to theunfocused sample (lane C) is clear, suggesting theactivation of these enzymes during isoelectric focusing.Proteinases of Mr ¶200 kDa with isoforms of pI between9 and 10 were chosen for further analysis, since thesewere detected in both culture media in a morereproducible manner than the other proteinases whichwere observed; furthermore they showed an adequaterelative gelatinase activity.

In order to classify the proteinases according to theircatalytic mechanism, their sensitivities to classic inhibi-tors were analyzed. PMSF at 10 mM totally inhibitedthe low mobility gelatinases (data not shown). E64,o-phenanthroline, Pepstatin A and iodoacetic acid hadno effect on these gelatinases (data not shown). SincePMSF may inhibit some cysteine-type proteinases as wellas serine proteinases, we studied the inhibition of theproteinases by 10 mM PMSF in the presence of 10 mM

DTT. The same result was obtained. This con�rmed theserine character of these proteinases, because PMSFdoes not inhibit cysteine-type proteinases in the presenceof DTT 18.

In order to investigate the possible role of pH in theactivation process, we investigated the effect of pre-incubation at distinct pH levels on gelatinase activity ofcrude �ltrate. The results were variable depending on the

time that the preparations had been stored at ¡20 oC.One-dimensional zymogram analyses of fresh prepara-tions (1–3 weeks) at pH 7¢0 showed no activation.Zymograms made from material stored for 2 months at¡20 oC showed an increase in gelatinase activity afterpre-incubation at pH 8¢0 (Fig. 2a). In preparations storedfor 1 year at ¡20 oC, the increase was observed after pre-incubation at pH 10¢0 (Fig. 2b). It is important toemphasize that the apparent molecular mass of theproteinases activated by pre-incubation at basic pHsdiffered, with one proteinase about 270 kDa (Fig. 2a)and the other 240 kDa (Fig. 2b). In all cases, detectableactivation by pre-incubation at different pHs wascon�ned to alkaline pHs and to proteinases of apparenthigh molecular mass. Proteinases of apparent lowmolecular mass were not activated by pre-incubationover the pH range studied.

The possible alkaline character of pI 10 proteinaseswas investigated by analysis of pH-dependent activity.Figure 3 shows the results of a typical experiment usingAzocoll. The relative maximum activity of the gelati-nases of pI 10¢0 occurs at pH 7¢0, while at pH 8¢0 and 9¢0,the activity was still signi�cant, corresponding to 88 and81% of the maximum, respectively. Similar pro�les andoptimal activity were obtained with pI 10¢0 gelatinases inzymography (not shown). In all of the analyses, therelative activity at pH 4¢0 was relatively low.

To establish the requirement for alkaline pH condi-tions in the activation observed in the pH pre-incubationand after isoelectric focusing experiments, we studiedwhether activation could take place at neutral pH.We fractionated the crude �ltrate by ion exchangechromatography at pH 7¢0. Figure 4 shows the activity



Fig. 2 Effect of the pre-incubation of two crude STA exoantigens at different pH levels on gelatinase activity. (a) Crude exoantigen storedfor 2 months at ¡20 oC. Activity after pre-incubation at pH 7¢0–10¢0, preparation 1. (b) Crude exoantigenstored at ¡20oC for 1 year. Activityat pH 7¢0 and 10¢0, preparation 2. Zymographs were carried out at pH 7¢0. The corresponding densitometric analyses are shown in graphicform.

at IVIC

on August 22, 2014

http://mm


ownloaded from


of distinct fractions separated by DEAE ion exchangechromatography. The recovery of gelatinase activity ofapparent high molecular mass is observed principally inthe �ow-through fraction. The zymogram shows activa-tion of gelatinases with an Mr of about 43 kDa in theeluates corresponding to 300 and 400 mM NaCl.

To investigate possible endogenous inhibitors of pI 10gelatinases, acidic fractions (fractions with pI 3¢0–6¢5)and a pI 10 fraction with gelatinase activity, that hadbeen obtained by Rotofor separation of M199 exoanti-gen, were neutralized and pre-incubated together. Theresults were analyzed by zymography. Figure 5 showsthat this treatment reduced the pI 10 gelatinase activityto minimum values, similar to those observed in theunfractionated material. This effect was not observedwhen the pI 10 fraction was pre-incubated with pH 5¢0and 6¢0 buffers, so it cannot be attributed to pH alone.Additionally, the inhibitory effect was not observed withother acidic fractions from the Rotofor, nor with the lowmolecular weight �ltrates from the pI 5¢4 and 6¢1

fractions (results not shown). Taking into account thatsome acidic pI fractions with an Mr of around 270 kDaalso show gelatinase activity, and that the incubationwith pI 10 fractions results in notably diminishedgelatinase activity in both fractions, a reciprocal inhibi-tion of these fractions is suggested.

In order to demonstrate the reversible character of thisinhibition, the pI 10 fraction inhibited by acidic pI 6¢1fractions was post-incubated in 10% v/v isopropanol,with the purpose of dissociating endogenous inhibitorsand reactivating the proteinases. Figure 6 shows the

364 Zambrano et al.


Fig. 3 pH dependence of pI 10¢0 exoproteinase activity. Absor-bance was measured at 520 nm after incubation with Azocoll atdifferent pH levels as indicated. Each point represents the averageof three experiments. Bars represent standard deviation.

Fig. 4 Zymogram of P. brasiliensis exoantigen fractionated at pH7¢0 on DEAE-Sepharose. Lane 1, crude exoantigen; lane 2, �owthrough fraction; lanes 3–12, fractions eluted with 20, 70, 80, 100,150, 200, 300, 400, 500 and 700 mM NaCl, respectively; lane 13, �ow-through fraction.

Fig. 5 Effect of pre-incubation at pH 5 and 6 on pI 10 proteinasesand inhibition by incubation with neutralized pI 5¢4 and 6¢1fractions. A, B and C refer to fraction pI 10 pre-incubated at pH 7¢0,5¢0 and 6¢0, respectively; D, fraction pI 10 plus pI 5¢4; E, fraction pI10 plus pI 6¢1. The densitometry analyses are shown in the graph.

Fig. 6 Partial inhibition of gelatinase activity of the pI fraction bypI 6¢1 and reversibility after post-incubation with isopropanol. Aand B, fraction pI 10 incubated at pH 7¢0 in the absence or presenceof isopropanol, respectively; C, activity of pI 10 incubated withpI 6¢1; D, incubation of the mixture C with 10% isopropanol.The densitometry analyses are shown in the graph.

at IVIC

on August 22, 2014

http://mm


ownloaded from


results of a typical experiment. The reversibility ofinhibition is evident.

Taken together, these results strongly suggest thepresence of inhibitory proteins of molecular mass greaterthan 10 kDa in the pI 5¢4 and 6¢1 fractions.

Discussion

The present experimental evidences are consistent withthe existence of endogenous inhibitors of proteinases incrude exoantigens of Pb mycelial forms.

Two-dimensional zymograms made with fresh exoan-tigen preparations and developed at pH 7¢0 showdramatic changes in gelatinase activity. This activitywas detectable in various Rotofor fractions with acidicand basic pI, some of which were not observed in thezymogram of the crude extract. The activation seen wasnot mediated merely by pH, because the optimal pH ofthe pI 10¢0 gelatinases is 7¢0. Enzymatic activationthrough conformational changes or autocatalytic pro-cesses mediated by variations in pH may have beenresponsible. Such effects have been described for someproteinases, particularly those the aspartic type [20–22].When we studied the pre-incubation effect of the crudeextract in pH ranging 4¢0–10¢0, the results were variabledepending upon storage time. In fresh preparations, noeffect was observed; whereas in some preparationsstored at ¡20 oC, an increase in relative activity wasobserved when the crude extracts were pre-incubated atpH 8¢0; in others the increment was produced after pre-incubation at pH 10¢0. In the cases where activationoccurred, the pre-incubated quantities of crude extractwere equivalent, ruling out the possibility that differ-ences in concentration were causal and suggesting thatpH has a modulatory role affecting proteinase activity.In the present study, however, pH related activationseemed to be restricted to proteinases pre-incubated atalkaline pH. It was also only observed in proteinases ofapparent high molecular mass.

When the crude extract of fresh preparations wasfractionated by anion exchange chromatography atpH 7¢0 and the resulting fractions were analyzed byzymography at the same pH, a pattern of activation wasobserved that was similar to that produced by iso-electric focusing. This result may indicate that analkaline pH is not indispensable for proteinase activa-tion. Alternatively, the result may suggest a mechanismof activation that involves the dissociation of aproteinase–inhibitor complex whose subunits are oppo-sitely charged at neutral pH. Competitive displacementof the inhibitor or electrostatic repulsion of theproteinase by the exchanger may occur during chroma-

tography, resulting in the dissociation of the complex.Similar reasoning may explain the activation of theseproteinases during isoelectric focusing, in which the pHgradient and electrical potential differential may dis-sociate the oppositely charged subunits. If this were thecase, and if the dissociation were reversible, then itshould be possible to demonstrate the existence ofinhibitors in some of the Rotofor or DEAE fractions.In fact, the inactivation of gelatinase activity in thepI 10 fraction by incubation with neutralized proteinfractions of pI 5¢4 and 6¢1 and the reversal of thisinactivation by isopropanol appear to con�rm theexistence of dissociable regulatory subunits.

How can the activation by preincubation at pH 8 and10 after storage at ¡20 oC be interpreted? We do nothave de�nitive answers, although a possible auto-catalytic activation of zymogens, mediated by basic pH,can not be excluded [22]. Alternatively, if we postulatethe existence of a proteinase-inhibitor complex whoseassociation is mediated by non-covalent interactions ofthe electrostatic dipole–dipole type, in which titratablecomponents are involved, then storage time could alteror degrade inhibitors and thus alter their interactionswith the proteinase in a way that affected its activitydifferently at various pH levels. An additional con-sideration is necessary: the dissociated inhibitory subunitshould be irreversibly inactivated at alkaline pH. If thatwere not so, it would reassociate during preincubation inthe pH 6¢8 sample buffer and the activity detected in thecrude extract preincubated at pH 8 or 10 would not beobserved. Endogenous protein inhibitors resistant todenaturation at acid pH but susceptible at alkaline pHhave been reported previously [23–25].

The effects of inhibitors provide the most reliableinformation concerning the possible catalytic type ofthese proteinases. The absence of inhibition by E64 andiodoacetic acid excludes the participation of a thiolgroup in the active site. The absence of an effect ofPepstatin A and phenanthroline rules out aspartic ormetalloproteinases. The gelatinase activity of the pI 10fraction was inhibited by PMSF, indicating the participa-tion of the amino acid serine in the active site.

Serine proteinases with alkaline pIs have beencharacterized previously [26–28]. Also, endogenousprotein inhibitors with acidic pIs have been detectedthat act on trypsin and chymotrypsin, two serineproteinases [26,27] and on other serine proteinasesproduced by Schizosaccharomyces pombe [29]. InCoccidioides immitis, a protein inhibitor of a serineprotease has been identi�ed [30]. The anionic characterof some serine proteinase inhibitors is presupposed inthe use of anionic exchangers in their puri�cation [24,31–33].



at IVIC

on August 22, 2014

http://mm


ownloaded from


Although the majority of the serine proteinasesexhibit optimal activity at alkaline pH, some such assubtilisin [34] as well as hepsin serine proteinase [35] andproteinases from Geotrichum candidum [36] have aoptimal activity at neutral pH.

No precise mechanism has been described by whichthe types of protein inhibitors seen here interact withtheir proteinase targets [7,37,38]. Nevertheless, itappears that the stabilization of the proteinase-inhibitorcomplex involves hydrogen bonding, dipole–dipoleinteractions and hydrophobic interactions betweenamino acid residues belonging to domains situated onsterically complementary surfaces of the inhibitor andthe proteinase [39]. These forces produce complexes thatremain stable in SDS [40–42]. This would explain whyminimal gelatinase activity was observed both whenzymographs of the crude extracts were analyzed and alsoin re-association experiments, even though the extractswere solubilized in SDS buffer.

In cases where protein–protein complexes are dis-sociated by organic solvents such as glycerol, DMSO orisopropanol, the participation of water molecules insome domains of the interacting surfaces has beenpostulated [43,44]. When organic solvents occupy thesedomains the water is displaced, altering the network ofhydrogen bonds between the interfaces and promotingthe dissociation of the complex [45]. Such an effect mayhave been operative in our own isopropanol-mediatedenzyme reactivations.

The mechanisms of in vivo activation of these proteinsare purely speculative at present. One might postulate adecrease in the biosynthesis of the inhibitor as a switchactivating the enzymes. Alternatively, as with the effectof DEAE, the enzymes could be activated throughcompetitive bonding of the inhibitory subunit by basicproteins, produced when growth reached the stationaryphase [46]. In some cases, the dissociation could occur asa result of pH changes in the micro-environment.

The results presented here provide the �rst evidencefor the existence of proteinase–inhibitor complexes inPb crude �ltrates, indicating the presence of a post-translational mechanism of proteinase regulation. Theseinhibitors could interfere with analyses of proteinases incrude exoantigens.

The activation of proteinases in de�ned culture mediasuch as M199 requires special commentary. The use ofchemically de�ned media promotes the establishment ofreproducible conditions for protein biosynthesis [47]. Itis generally accepted that exoproteinases participateprimarily in nutrition and not in regulatory processes[4,48]. The exoproteinases are produced under condi-tions of nitrogen or sulfur, limitation, or during thestationary growth after glucose has been exhausted, in

order to assimilate the protein component included inthe medium [46,48]. Based on these scenarios theinducible biosynthesis of exoproteinases functioning inassimilation would not be expected in M199 which doesnot contain proteins. Contrary to what might beexpected, in M199, as in STA, the activation of secretedproteinases was con�rmed. Two possibilities may explainthis observation. First, perhaps exoantigens were har-vested during stationary growth, and the proteinasesseen are intracellular enzymes liberated into the mediumvia cellular autolysis. Alternatively, the proteinasescould be constitutive components of growth relatedphysiology. If so, they should be produced duringexponential growth. Although we did not determinethe growth phase of the cultures from which exoantigenswere harvested, studies by other authors have estab-lished that gp43 is produced and accumulates exclusivelyduring exponential growth [49].

The predominance of gp43 in the crude exoantigensstudied and the minimal gelatinase activity detected incrude �ltrates studied suggests that the �ltrates in bothculture media have been harvested during exponentialphase growth. This conclusion is consistent with theproduction of regulated proteinases during this growthphase. The production of proteinase-inhibitor complexesin this phase suggests that these enzymes couldparticipate in functions other than those involvingnutrition. For example, they may play a role in post-translational modi�cation of proteins involved in cellwall morphogenesis [50].

Acknowledgements

This study was �nanced by the National Council ofScienti�c and Technological Research (CONICIT),Venezuela, Grant No. S1-97002420.

References

1 Hamilton AJ. Serodiagnosis of histoplasmosis, paracoccidioido-mycosis and penicilliosis marneffei; current status and futuretrends. Med Mycol 1998; 36: 351–364.

2 De Camargo ZP, Unterkircher C, Campoy SP, Travassos LR.Production of Paracoccidioides brasiliensis exoantigens forimmunodiffusion tests. J Clin Microbiol 1988; 26: 2147–2151.

3 Travassos LR, Puccia R, Cisalpino P, et al. Biochemistry andmolecular biology of the main diagnostic antigen of Paracocci-dioides brasiliensis. Arch Med Res 1995; 26: 297–304.

4 Rao MB, Tanksale AM, Ghatge MS, Deshpande VV. Molecularand biotechnological aspects of microbial proteases. MicrobiolMol Rev 1998; 62: 597–635.

366 Zambrano et al.


at IVIC

on August 22, 2014

http://mm


ownloaded from


5 Maeda Hirosshi, Yamamoto Tetsuro. Pathogenic mechanisminduced by microbial proteases in microbial infections. BiolChem Hoppe Seyler 1996; 377: 217–226.

6 Hogan LH, Klein BS, Levitz SM. Virulence factors of medicallyimportant fungi Clin Microbiol Rev 1996; 9: 469–488.

7 Bode W, Fernandez C, Nagase H, Maskos K. Endoproteinase–protein inhibitor interactions. APMIS 1999; 107: 3–10.

8 San-Blas G, Sorais F, Nino-Vega G, Mendez C, San-Blas F.Cytosolic neutral proteinases of Paracoccidiodes brasiliensis.Current Microbiol 1998; 37: 141–143.

9 Bedoya-Escobar VI, Naranjo-Mesa MS, Restrepo-Moreno A.Detection of proteolytic enzymes released by the dimorphicfungus Paracoccidioides brasiliensis. J Med Vet Mycol 1993; 31:229–304.

10 Vaz CAC, MacKenzie DWR, Hearn V, et al. Gelatinase activityof exoantigens from virulent and non virulent isolates ofParacoccidioides brasiliensis. J Med Vet Mycol 1994; 32: 65–69.

11 Carmona AK, Puccia R, Oliveira MCF, Rodrigues EG, JulianoL, Travassos LR. Characterization of an exocellular serine-thiolproteinase activity in Paracoccidioides brasiliensis. Biochem J1995; 309: 209–214.

12 Goihman-Yahr M, Pine L, Albornoz MC, et al. Studies onplating ef�ciency and estimation of viability of suspension ofParacoccidioides brasiliensis yeast cells. Mycopathologia 1980;71: 73–83.

13 Lowry OH, Rosenbrough NJ, Lewis Faar A, Randall RJ.Protein measurement with the Folin phenol reagent. J BiolChem 1951; 193: 265–275.

14 Laemmli UK. Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature 1970; 227:680–685.

15 Towbin H, Theophil S, Gordon J. Electrophoretic transfer ofprotein from polyacrylamide gels to nitrocellulose sheets:procedure and some applications. Proc Natl Acad Sci USA1979; 76: 4350–4354.

16 Heussen C, Dowdle EB. Electrophoretic analysis of plasmino-gen activators in polyacrylamide gels containing sodium dodecylsulfate and copolymerized substrates. Anal Biochem 1980; 102:196–202.

17 Chavira R, Burnett TJ, Hageman JH. Assaying proteinases withAzocoll. Anal Chem 1984; 136: 446–450.

18 Dunn B.M. Determination of protease mechanisms. In: BeynonRJ, Bond JS, eds. Proteolytic Enzymes: a Practical Approach.1st edn. Oxford: IRL press at Oxford University Press, 1989: 57–81.

19 Salvensen G, Nagase H. Inhibition of proteolytic enzymes. In:Beynon RJ, Bond JS, eds. Proteolytic Enzymes: a PracticalApproach. 1st edn. Oxford: IRL press at Oxford UniversityPress, 1989: 83–104

20 Lee AY, Gulnik SV, Erickson JW. Conformational switching inan aspartic proteinase. Natl Struct Biol 1998; 10: 866–871.

21 McEwen RK, Young TW. Secretion and pH-dependent self-processing of the pro-form of the Yarrowia lipolytica acid-extracellular protease. Yeast 1998; 14: 1115–1125.

22 Khan AR, James MN. Molecular mechanism for conversion ofzymogens to active proteolytic enzymes. Protein Sci 1998; 7:815–836.

23 Melrose J, Ghosh P, Taylor TK, Andrew JL. The serineproteinase inhibitory proteins of the human inverterbral disc:their isolation, characterization and variation with aging anddegeneration. Matrix 1992; 12: 456–470.

24 Belew M, Porath J, Sundberg L. The trypsin and chymotrypsininhibitor in chick peas (Cicer arietinum L.). Puri�cation andproperties of the inhibitors. Eur J Biochem 1975; 60: 247–258.

25 Veeraragavan K, Singh K, Wachter E, Hochstrasser K.Characterization of trypsin inhibitor from equine urine. Bio-chem Int 1992; 26: 405–413.

26 Sokolova EA, Starkova NN, Vorotyntseva TI, ZamolodchikovaTS. A serine protease from the bovine duodenal mucosachymotrypsin-like duodenase. Eur J Biochem 1998; 255: 501–507.

27 Shirai T, Suzuki A, Yamane T, et al. High-resolution crystalstructure of M-protease: phylogeny aided analysis of the high-alkaline adaptation mechanism. Protein Eng 1997; 10: 627–634.

28 Zamolodchikova TS, Vorotyntseva TI, Antonov VK. Duode-nase, a new serine protease of unusual speci�city from bovineduodenal mucosa. Puri�cation and properties. Eur J Biochem1995; 227: 866–872.

29 Escudero B, Parra F, Suarez-Rendueles P. Puri�cation andcharacterization of endogenous inhibitor for proteinase B fromSchizosaccharomyces pombe. Biochimie 1993; 75: 855–859.

30 Yuan L, Cole GT. Characterization of a proteinase inhibitorisolated from the fungal pathogen Coccidioides immitis.Biochem J 1989; 257: 729–736.

31 Lepley RA, Pampusch M, Dayton WR. Puri�cation of a high-molecular-weight inhibitor of the calcium-activated proteinase.Biochim Biophys Acta 1985; 828: 95–103.

32 Graber M, Condoret JS. Preparative anion-exchange chromato-graphy of soybean trypsin inhibitor: the alternative of column–overload methods. J Chromatogr 1992; 584: 115–120.

33 Watanabe E, Tsoka S, Asenjo JA. Selection of chromatographyprotein puri�cation operations based on physicochemicalproperties. Ann N Y Acad Sci 1994; 721: 348–64.

34 Babe LM, Schmidt B. Puri�cation and biochemical analysis ofWprA, a 52-kDa serine protease secreted by B. subtilis as anactive complex with its 23-kDa propeptide. Biochim BiophysActa 1998; 1386: 211–219.

35 Zhukov A, Hellman U, Ingelman-Sundberg M. Puri�cation andcharacterization of hepsin from rat liver microsomes. BiochimBiophys Acta 1997; 1337: 85–95.

36 Litthauer D, Louw CH, du Toit PJ. Geotrichum candidum P-5produces an intracellular serine protease resembling chymo-trypsin. Int J Biochem Cell Biol 1996; 28: 1123–1130.

37 Laskowski M Jr. Protein inihibitors of serine proteinases –mechanisms and classi�cation. Bioessays 1996; 18: 453–464.

38 Plotnick MI, Mayne L, Schechter NM, Rubin H. Distortion ofthe active site of chymotrypsin complexed with a serpin.Biochem 1996; 35: 7586–7590.

39 Lee AY, Smitka TA, Bonjouklian R, Clardy J. Atomic structureof the trypsin-A90720A complex: a uni�ed approach tostructure and function. Chem Biol 1994; 2: 113–117.

40 Olson ST, Swanson R, Patson PA, Bjork I. Apparent formationof sodium dodecyl sulfate-stable complexes between serpins and3,4-dichloroisocoumarin-inactivated proteinases is due to re-generation of active proteinase from the inactivated enzyme. JBiol Chem 1997; 272: 13338–13342.

41 Wright HT. The structural puzzle of how serpin serineproteinase inhibitors work. Bioessays 1996; 18: 453–464.

42 Gettins P, Patson PA, Schapira M. The role of conformationalchange in serpin structure and function. Bioessays 1993; 15: 461–467.

43 Polticelli F, Ascenzi P, Bolognesi M, Honig B. Structuraldeterminants of trypsin af�nity and speci�city for cationicinhibitors. Protein Sci 1999; 8: 2621–2629.



at IVIC

on August 22, 2014

http://mm


ownloaded from


44 Kuznetsova N, Rau DC, Parsegian VA, Leikin S. Solventhydrogen-bond network in protein self-assembly: solvation ofcollagen triple helices in nonaqueous solvents. Biophys J 1997;72: 353–362.

45 Kornblatt JA, Kornblatt MJ, Hoa GH, Mauk AG. Responses oftwo protein-protein complexes to solvent stress: does water playa role at the interface? Biophys J 1993; 65: 1059–1065.

46 Kim IS, Lee KJ. Physiological roles of leupeptin and extra-cellular proteases of Streptomyces exfoliatus SMF13. Microbiol1995; 141: 1017–1025.

47 Restrepo A, Jimenez B. Growth of Paracoccidioides brasiliensisyeast phase in a chemically de�ned culture medium. J ClinMicrobiol 1980; 12: 279–281.

48 North M. Comparative biochemistry of the proteinases ofeucaryotic microorganisms. Microbiol Rev 1982; 46: 308–340.

49 Stambuk BU, Puccia R, DeAlmeida MLC, Travassos LR,Schenkman SS. Secretion of the 43 kDa glycoprotein antigen byParacoccidioides brasiliensis. J Med Vet Mycol 1988; 26: 367–373.

50 Lipke PN, Ovalle R. Cell wall architecture in yeast: Newstructure and new challenges. J Bacteriol 1998; 180: 3735–3740.

368 Zambrano et al.


at IVIC

on August 22, 2014

http://mm


ownloaded from


Regulation of serine-type exoproteinases by endogenous inhibitors present in exoantigens of the mycelial form of Paracoccidioides brasiliensis

Documents