Eur. J. Biochem. 265, 145±151 (1999) q FEBS 1999

Purification and properties of a basic endo-b-1,6-glucanase (BGN16.1)from the antagonistic fungus Trichoderma harzianum

Jesu s de la Cruz and Antonio Llobell

Instituto de BioquõÂmica Vegetal y FotosõÂntesis, Universidad de Sevilla-CSIC, Sevilla, Spain

The antagonistic fungus Trichoderma harzianum CECT 2413 produces at least two extracellular b-1,6-glucanases,

among other hydrolases acting on polysaccharides from fungal cell walls, when grown in chitin as the sole carbon

source. We have previously reported on the purification and biochemical characterization of the major activity,

which corresponds to an acidic enzyme named BGN16.2 [de la Cruz, J., Pintor-Toro, J.A., BenõÂtez, T. & Llobell,

A. (1995) J. Bacteriol. 177, 1864±1871]. In this paper, we report on the purification to electrophoretical

homogeneity of BGN16.1, the second b-1,6-glucanase enzyme. BGN16.1 was purified by ammonium sulfate

precipitation followed by adsorption and digestion of pustulan (a b-1,6-glucan), chromatofocusing and gel-

filtration chromatography. BGN16.1 is a non-glycosylated protein with an apparent molecular mass of 51 kDa

and a basic isoelectric point (pI 7.4±7.7). The enzyme was active toward substrates containing b-1,6-glycosidic

linkages, including yeast cell walls. The Km was 0.8 mg´mL21 with pustulan as the substrate. Reaction product

analysis by HPLC clearly indicated that BGN16.1 has an endo-hydrolytic mode of action. The probable role of

this enzyme in the antagonistic action of T. harzianum is also discussed.

Keywords: antagonism; fungal cell wall; Trichoderma; b-glucanases.

Fungal cell walls contain different kinds of b-glucans asstructural components. These b-glucans are generally com-posed of a major component of predominantly b-1,3-linkedglucans with branches of b-1,6-glycosidic linkages [1,2].Although enzymes hydrolyzing b-1,3-glucans (1,3-b-d-glucanglucanohydrolases; EC 3.2.1.39 or EC 3.2.1.58) have beenextensively studied [3±11], biochemical and molecular proper-ties of b-1,6-glucanases (1,6-b-d-glucan glucanohydrolases;EC 3.2.1.75) have been scarcely described, and unfortunatelythere is almost no information about purification andcharacterization of b-1,6-glucanases [11±17].

Soil-borne fungi of the genus Trichoderma have beenspecially studied for both their cellulolytic activity [18] andtheir antagonistic properties against fungal plant pathogens[19,20]. The anti-fungal mechanism of Trichoderma involvesfungal cell wall-degrading enzymes [19±21], among them,chitinases [22,23], b-1,3-glucanases, b-1,6-glucanases[9,10,12,24] and proteases [25]. As chitin and b-1,3-glucanare the main structural components of fungal cell walls (exceptthose from Oomycetes) [26], chitinases and b-1,3-glucanaseshave been proposed as the key enzymes in the lysis ofphytopathogenic fungal cell walls during the antagonisticaction of Trichoderma [19,21,27]. However, other cellwall-degrading enzymes, including those hydrolyzing minor

polymers (proteins, b-1,6-glucans, a-1,3-glucans, etc.), may beinvolved in the effective and complete degradation of mycelial orconidial walls of phytopathogenic fungi by Trichoderma.

We have previously reported on an extracellular b-1,6-glucanase activity from an antagonistic strain of Trichodermaharzianum growing on chitin as the sole carbon source [27].This activity is due to the presence of at least two b-1,6-glucanase proteins (BGN16.1 and BGN16.2) [12]. We haveanalyzed both BGN16.2 protein and bgn16.2 gene [12,28]. Inthis paper, we report on the purification and biochemicalcharacterization of BGN16.1.

E X P E R I M E N T A L P R O C E D U R E S

Micro-organisms and culture conditions

T. harzianum CECT 2413 was obtained from the Spanish TypeCulture Collection (Burjasot, Valencia, Spain) and maintainedon potato/dextrose/agar medium (Difco, Detroit, MI, USA). Forthe production and purification of b-1,6-glucanases, T. harzia-num was grown as previously described [27]; briefly, 106

conidia were inoculated in 1-L Erlenmeyer flasks containing400 mL Czapek medium supplemented with 10% (w/v) glucoseand incubated for 4 days at 28 8C and 200 r.p.m. (repressedconditions); the mycelia were then washed under sterileconditions, transferred to 1-L Erlenmeyer flasks containing200 mL Czapek medium supplemented with 1.5% (w/v) chitinand 70 mm phosphate/KOH buffer, pH 6.0, and incubated asdescribed above (induction conditions).

Chemicals and reagents

Pustulan (from Umbilicaria papullosa) and pachyman (fromPoria cocos) were from Calbiochem (La Jolla, CA, USA).Aniline blue, CM-cellulose, cello-oligosaccharides, chitin(from crab shells), dextran (from Leuconostoc mesenteroides),

Correspondence to A. Llobell, IBVF, Centro de Investigaciones

CientõÂficas Isla de la Cartuja, Avda. AmeÂrico Vespucio s/n, Isla de la

Cartuja E-41092, Sevilla, Spain. Fax: + 34 95 4460065,

Tel.: + 34 95 4489521, E-mail: Llobell@cica.es

Abbreviations: endo H, endo-b-N-acetylglucosaminidase H; Glcn,

b-1,6-glucan oligomer.

Enzymes: b-1,3-glucanase (1,3-b-d-glucan glucanohydrolase; EC 3.2.1.39);

exo-b-1,3-glucanase (1,3-b-d-glucan glucanohydrolase; EC 3.2.1.58);

b-1,6-glucanase (1,6-b-d-glucan glucanohydrolase; EC 3.2.1.75);

b-glucosidase (b-d-glucoside glucohydrolase; EC 3.2.1.21).

(Received 12 February 1999, accepted 6 July 1999)

146 J. de la Cruz and A. Llobell (Eur. J. Biochem. 265) q FEBS 1999

gentibiose, glucose, glycol-chitosan, horseradish peroxidase-conjugated anti-(mouse IgG), laminarin (from Laminariadigitata), nigeran (from Aspergillus nidulans), ovalbumin,phenylmethanesulfonyl fluoride and soluble starch werepurchased from Sigma Chemical Co. (St Louis, MO, USA).Endo-b-N-acetylglucosaminidase H (Endo-H) and a-mannosi-dase were from Boehringer (Mannhein, Germany). Chemicalsfor electrophoresis and protein assay dye reagent were fromBio-Rad (Richmond, CA, USA). Chemicals for columnchromatography, ampholytes, and pI standard proteins weresupplied by Amersham-Pharmacia (Uppsala, Sweden). Fungalcell walls were prepared as described by Fleet and Phaff [29].Yeast glucan was prepared from baker's yeast as described byRombouts and Phaff [15]. All other chemicals were ofanalytical grade.

b-1,6-Glucanase assay and protein determination

b-1,6-Glucanase activity was routinely assayed by incubating0.8 mL of 5 mg´mL21 pustulan (b-1,6-glucan) in 50 mmpotassium acetate buffer, pH 5.5, with 0.2 mL of enzymesolution appropriately diluted in the same buffer. Reactionmixtures were incubated at 37 8C for 30 min and reactionswere stopped by boiling for 5 min. Then, the reducing sugarcontent was determined by the procedure of Somogyi [30] andNelson [31], using glucose as standard. Enzyme and substrateblanks were included. One unit of b-1,6-glucanase activity wasdefined as the amount of enzyme that releases 1 mmol ofreducing sugar equivalents, expressed as glucose, per min underthe standard assay conditions.

Protein concentration was measured by the method ofBradford [32], with ovalbumin as standard.

Enzyme purification

Step 1. Unless otherwise indicated, all steps were carried out at0±4 8C. T. harzianum cultures grown for 48 h in Czapekmedium with 1.5% chitin were filtered through Whatman no. 1paper and centrifuged (8000 g, 10 min). Then, the supernatant(about 800 mL) was precipitated with solid ammonium sulfate(80% saturation) and the precipitate was recovered bycentrifugation (12 000 g, 20 min), resuspended in the minimalvolume of distilled water and dialyzed extensively against50 mm potassium acetate buffer, pH 5.5. The crude enzymepreparation (about 20 mL) was recovered by centrifugation(12 000 g, 20 min) of the dialyzed fraction.

Step 2. Aliquots of crude enzyme were then adsorbed topustulan as described [12]. Adsorbed fractions were washedthree times with 50 mm potassium acetate buffer, pH 5.5,containing 1 m NaCl, resuspended in 50 mm potassium acetatebuffer, pH 5.5, with 1 mm phenylmethanesulfonyl fluoride and1 mm sodium azide, and incubated overnight at 37 8C forpustulan digestion. The clarified solutions were extensivelydialyzed against 25 mm imidazole/HCl buffer, pH 7.4.

Step 3. The dialyzed solution (named pustulan digest) wasdirectly applied to a Polybuffer Exchanger PBE94 column(1 � 20 cm) equilibrated in 25 mm imidazole/HCl buffer,pH 7.4. Protein elution and pH-gradient conditions(pH 7.4±4) were as previously described [12]. Most activefractions containing basic b-1,6-glucanase activity (voidvolume) were pooled and concentrated to < 0.5 mL onCentricon 10 concentrators (Amicon, Beverley, MA, USA).

Step 4. The concentrated pool was loaded on a SephacrylS-200 H column (1.6 � 40 cm) equilibrated with 100 mmpotassium acetate buffer, pH 5.5, containing 1 m NaCl andeluted with the same buffer at a flow rate of 7 mL´h21.Fractions of 0.65 mL were collected and monitored for protein(A280) and for b-1,6-glucanase activity. Most active fractionswere pooled, washed and concentrated in 50 mm potassiumacetate buffer, pH 5.5, on Centricon 10 concentrators. Thepurified protein was stored at 220 8C, under which conditionsb-1,6-glucanase activity remains unchanged for at least amonth.

Electrophoretical procedures

SDS/PAGE was performed by the method of Laemmli [33] with4% acrylamide in the stacking gel and 12% acrylamide in theseparating gel. Samples were diluted in Laemmli buffer andboiled for 3 min before loading. After electrophoresis, gelswere stained with Coomassie R-250 brilliant blue. Low-molecular-mass standard proteins (Bio-Rad) were used formolecular-mass determination.

Detection of b-1,6-glucanase activity was achieved in agarreplicas of the SDS/polyacrylamide gels exactly as described bySoler et al. [34].

Glycoprotein stain assays were carried out in theSDS/polyacrylamide gels by either the periodic acid/Schiffreagent procedure (Sigma) or the silver nitrate procedure ofDubray and Bezard [35]. Extracellular yeast invertase (Sigma)was used as a glycoprotein-positive control. To remove thecarbohydrate linked to the protein, deglycosylation reactionswere performed before SDS/PAGE with Endo-H or a-manno-sidase as described by VaÂzquez de Aldana et al. [36] andWilliamson et al. [37], respectively. Both b-1,6-glucanase anddeglycosydase enzyme were used as blanks and extracellularyeast invertase as positive control.

Western blots were performed as described by Kombrinket al. [38]. Polyclonal mouse anti-(T. harzianum) BGN16.2antibodies [12], rabbit anti-(T. harzianum) BGN13.1 (A.RincoÂn and T. BenõÂtez, unpublished work), rabbit anti-(tobaccob-1,3-glucanase-PR-2) [4] and rabbit anti-(tobacco basic-b-1,3-glucanase) antibodies [4] were used as primary antibodies(dilution 1 : 1000); peroxidase-conjugated anti-(mouse IgG)and peroxidase-conjugated anti-(rabbit IgG) sera (Sigma) wereused as secondary antibodies (dilution 1 : 1000).

Isoelectrofocusing was performed as described by Robertsonet al. [39]. Proteins were visualized by Coomassie staining.The apparent pI value for BGN16.1 was calculated byreference to standard marker proteins with pI values withinthe range 3.5±9.3 (Amersham-Pharmacia). b-1,6-Glucanaseactivity was detected in isoelectrofocusing gels exactly asdescribed by Soler et al. [34].

Basic chromatofocusing

Analytical basic chromatofocusing was performed on aPolybuffer Exchanger PBE94 column (1 � 20 cm) equilibratedin 25 mm ethanolamine/NaOH, pH 9.4. The column was elutedat a flow rate of 9 mL´h21 with Polybuffer 96, which had beendiluted eightfold with water and adjusted to pH 9.4. Fractionsof volume 1.5 mL were collected. The chromatofocusing wasmonitored by measuring the A280 and the pH of the elutedfractions.

q FEBS 1999 A basic b-1,6-glucanase from Trichoderma harzianum (Eur. J. Biochem. 265) 147

Kinetic parameters, temperature optimum and thermalinactivation

Michaelis±Menten constants were determined from Lineweaver±Burk plots using data obtained by measuring the initial rate ofpustulan hydrolysis and using a range of pustulan concen-trations from 20 to 0.5 mg´mL21. Initial rates of pustulanhydrolysis were calculated by measuring activity under theassay conditions described above at many different time pointsfrom 0 to 30 min.

The temperature optimum was determined by performing thestandard assay within the temperature range 20±80 8C. Thethermal stability was also determined by incubating the purifiedprotein for 30 min at temperatures from 20 to 80 8C in 50 mmpotassium acetate buffer, pH 5.5, and then measuring theremaining activity at 37 8C by adding pustulan (5 mg´mL21) asthe assay substrate. The inactivation temperature was defined asthe temperature at which the specific activity was reduced by50%, under the conditions described above.

Substrate specificity and analysis of b-1,6-glucanasereaction products

In addition to pustulan, the activity of the purified BGN16.1protein was tested on other polymers with a-glycosidic orb-glycosidic bonds at a concentration of 5 mg´mL21. Whenglucans were used, the reaction products were determinedunder the standard assay conditions described above. Chitinaseand chitosanase activities were determined as described by dela Cruz et al. [23].

Substrate hydrolysis by purified BGN16.1 was performed byincubating 4 mg of pustulan or 2 mg of gentibiose (a b-1,6-disaccharide) with 2 mg of purified protein in 1 mL ofdistilled water for various times at 37 8C. Substrate blankswere included in parallel. After hydrolysis, the reactions werestopped by 5 min of boiling and the pustulan oligomers wereanalyzed by HPLC using an HPX 42-A column (Bio-Rad)maintained at 60 8C. Water was used as eluent at a flow rate of0.6 mL´min21. Products were detected on the basis of their A195

and identified by comparison with glucose, gentibiose andcellulose oligosaccharide (rate of oligomerization from 2 to 5)standards.

Hydrolysis of fungal cell walls

To monitor cell wall hydrolytic activity, the purified b-1,6-glucanase (0.5 U) was incubated with 4 mg of lyophilizedcell walls from different fungi in a 1-mL assay in 50 mmpotassium acetate buffer, pH 5.5. The fungi were Botrytiscinerea CECT 2100, Giberella fujikuroi IMI 58289 (ImperialMycological Institute, Kew, UK), Phytophora syringaeCECT 2351 and Saccharomyces cerevisiae (baker's yeast; LaCinta Roja, Spain). Mixtures were incubated at 37 8C for 16 hwith occasional shaking. The reactions were stopped bycentrifugation (10 000 g, 10 min) and the amount of reducingsugars released in the supernatants was determined by themethod of Somogyi [30] and Nelson [31]. Enzyme andsubstrate blanks were included.

Lytic activity on fungal cell walls was also monitored byobserving clearing halli in agar plates containing these fungalcell walls (final concentration 1 mg´mL21), using the cup-plateassay of Tanaka and Phaff [40]. After 48 h of incubation at37 8C, the plates were stained with 0.005% (w/v) aniline blueas described by Grenier et al. [41] and washed extensively withwater. The hydrolytic halli were observed under long-wavelength UV light, and the hallus radii were measured.

R E S U L T S

Enzyme production and purification

We have previously reported on the production of b-1,6-glucanase by T. harzianum CECT 2413 grown in thepresence of chitin as carbon source [27]. The enzymeproduction increased continuously with time to near-maximumlevels after 48 h, which was the usual incubation period [27].We have also reported that the b-1,6-glucanase activitymeasured is the result of at least two b-1,6-glucanase proteinswith different pIs, which were named BGN16.1 (pI . 7) andBGN16.2 (pI 5.8) [12]. In this work, we purified and studied thebiochemical properties of the basic BGN16.1 enzyme.

To purify BGN16.1, filtrates of chitin-supplemented cultures(200 mL) were concentrated by ammonium sulfate precipita-tion. The concentrated crude enzyme was then subjected toaffinity adsorption to pustulan and further digestion of thisb-1,6-glucan. Acidic chromatofocusing separated the pustulandigest into two fractions, a non-adsorbed fraction, whichaccounted for < 15% of the total loaded activity, and a secondpeak at pH 5.8, which contained the remaining activity [12].The latter peak contained the BGN16.2 enzyme [12]. TheBGN16.1 protein was only found in the non-retained peak,which also contained the b-1,3-glucanase activity of BGN13.1[10]. For further purification of BGN16.1, most active fractionsof the non-adsorbed peak were pooled, concentrated and finallypurified by gel filtration on a Sephacryl S-200 H column. Theenzyme purification is summarized in Table 1. The BGN16.1enzyme was purified < 154-fold with a minimal estimatedrecovery of 5%. When analyzed by SDS/PAGE, the finalpurified preparation migrated as a single sharp band with notrace of contaminants (Fig. 1, lane 1). This band showed b-1,6-glucanase activity as demonstrated by pustulan agar replicas ofrenatured SDS/polyacrylamide gels (Fig. 1, lane 2). Weconcluded that the above procedure achieves purification ofBGN16.1 to electrophoretical homogeneity.

Physicochemical parameters

The apparent molecular mass of the purified BGN16.1enzyme was estimated to be < 51 kDa by SDS/PAGE(Fig. 1). When the molecular mass was calculated by gelfiltration on Sephacryl S-200 HR (column calibrationstandards: BSA, 67 kDa; ovalbumin, 43 kDa; carbonicanhydrase, 31 kDa; a-chymotrypsinogen, 25 kDa; lysozyme,14 kDa), the value obtained was 20±25 kDa. The pI of the pure

Fig. 1. SDS/PAGE of the purified BGN16.1 enzyme. std., molecular-

mass standards; lane 1, purified BGN16.1 (5 mg protein); lane 2, b-1,6-

glucanase activity detected on an agarose replica of an SDS/polyacrylamide

gel containing 5 mg of purified BGN16.1. The numbers on the left refer to

the molecular masses of the standards.

148 J. de la Cruz and A. Llobell (Eur. J. Biochem. 265) q FEBS 1999

enzyme, determined by isoelectrofocusing and basic chromato-focusing, was estimated to be pH 7.4 and pH 7.7, respectively.Data were confirmed by enzymatic determination in isoelectro-focusing gels, as the pI of the purified enzyme corresponded tothat of the activity detected [34].

No evidence was found of the presence of carbohydrates inthe purified protein. Staining with periodic acid/Schiff'sreagent or silver nitrate reagent after SDS/PAGE was negative(data not shown). In addition, Endo-H and a-mannosidasetreatments did not alter the apparent size of BGN16.1 inSDS/polyacrylamide gels (data not shown).

Finally, the purified BGN16.1 protein was not detected byimmunoblotting with antiserum against T. harzianum BGN13.1(A. RincoÂn and T. BenõÂtez, unpublished work), T. harzianumBGN16.2 [12], tobacco b-1,3-glucanase PR2 [4] or tobaccobasic b-1,3-glucanase [4] (data not shown). Thus, thepurified enzyme is not immunologically related to any ofthese b-glucanases.

Kinetic parameters

Michaelis constants, determined under standard assay conditionswith pustulan as the substrate, were as follows: Vmax = 312 mmolproduct´min21´(mg protein)21 and Km = 0.8 mg mL21.

The optimal temperature for enzyme activity was calculatedto be 50 8C at pH 5.5. After a 30-min preincubation time atdifferent temperatures, the inactivation temperature was

calculated also as 50 8C. Therefore, pustulan seems to stabilizethe enzyme.

Substrate specificity and action pattern

The substrate specificity of the purified BGN16.1 was assayedwith a variety of glucan substrates (Table 2). The maximalactivity was detected against pustulan and yeast glucan, whichare a linear b-1,6-glucan and a b-1,3:b-1,6 glucan (4 : 1),respectively [2]. Approximately one-fifth of the maximalactivity was observed against laminarin, which is mainlycomposed of b-1,3-glucan with b-1,6-glycosidic linkages asbranches at the ratio 7 : 1 of linkage type [2]. No activity wasdetected with pachyman as substrate which is a linear polymercomposed exclusively of b-1±3-linked b-d-glucose [2], oragainst other a-glucans or b-glucans, chitin or chitosan(Table 2). We conclude that the BGN16.1 enzyme is a specificb-1,6-glucanase.

To determine whether BGN16.1 acts as an endoglucanase oran exoglucanase, the products released from the enzymatichydrolysis of pustulan were analyzed by HPLC (Fig. 2). Thisanalysis revealed an initial release of a range of largeoligosaccharides (Fig. 2, 1 h; data not shown for shorterincubation time points), which were later cleaved to smallsaccharides ranging from glucose to gentitetraose (Glc4),

Table 1. Purification of BGN16.1 from T. harzianum.

Step

Volume

(mL)

Total protein

(g)

Total activity

(U)

Specific activity

(U´mg21)

Yield

(%)

Purification

(fold)

Crude enzyme 16�.0 64 62�.2 0�.97 100 1

Pustulan digestion 9�.0 5�.1 18�.3 3�.58 30 4

Chromatofocusing eluate 0�.4 0�.08 6 75�.0 10 77

Sephacryl

S-200 HR eluate

0�.4 0�.02 3 150�.0 5 154

Table 2. Substrate specificty of the purified BGN16.1 from T. harzia-

num. All substrates were used at a final concentration of 4 mg´mL21. 100%

of activity corresponds to 170 U´(mg protein)21.

Substrate

Main linkage

type

(monomer)

b-1,6-Glucanase

activity

(%)

Pustulan b-1,6 (Glc) 100

Glucan (S. cerevisiae) b-1,3: b-1,6 (Glc) 73

Laminarin b-1,3: b-1,6 (Glc) 22

Pachyman b-1,3 (Glc) 0

CM-cellulose b-1,4 (Glc) 0

Colloidal chitin b-1,4 (GlcNAc) 0

Glycol chitosan b-1,4 (GlcN) 0

Nigeran a-1,3: a-1,4 (Glc) 0

Soluble starch a-1,4: a-1,6 (Glc) 0

Dextran a-1,6 (Glc) 0

B. cinerea cell walls b-Glucan : chitin 0

G. fujikuroi cell walls b-Glucan : chitin 0

P. syringae cell walls b-Glucan : cellulose 0

S. cerevisiae cell walls b-Glucan 20

Fig. 2. HPLC analysis of the action of BGN16.1 on pustulan. Pustulan

(5 mg´mL21) was incubated, as described in Experimental procedures, with

2 mg of the purified enzyme for the indicated times. Gn refer to glucose

oligomers (n = degree of polymerization).

q FEBS 1999 A basic b-1,6-glucanase from Trichoderma harzianum (Eur. J. Biochem. 265) 149

gentibiose (Glc2) being the major final hydrolysis product(Fig. 2, 16 h). When gentibiose was used as the substrate, thedisaccharide was not split by the enzyme (data not shown).Thus, these results indicate that the purified enzyme is anendo-b-1,6-glucanase.

Action on fungal cell walls

To establish the possible role of the purified enzyme in theantagonism of T. harzianum, BGN16.1 was also assayed forfungal cell wall-degrading activity. The enzyme alone was quiteactive against yeast cell walls but not active at all against theother cell walls assayed (Table 2). In addition, no clearingactivity (hydrolytic halli, see Experimental procedures) onyeast or filamentous fungi cell walls was detected (data notshown). However, both clearing and anti-fungal activitiesagainst phytopathogenic fungi were detected when BGN16.1was combined with BGN13.1 and BGN16.2 [12].

D I S C U S S I O N

The study of the expression patterns of Trichoderma fungal cellwall-degrading enzymes in different nutrient sources, and thebiochemical characterization of these enzymes are prerequisitesto understanding the molecular basis of the antagonistic actionof Trichoderma against phytopathogenic fungi. With this aim,we and others have analyzed the production of extracellularchitinases, b-glucanases and proteases by the antagonisticfungus T. harzianum when grown in conditions that wouldresemble antagonism, such as the presence of abundantpolysaccharides of fungal cell walls (i.e. chitin), purified fungalcell walls or autoclaved mycelia as nutrient sources[21,22,24,27]. Some of these hydrolases have been purifiedand implicated as key enzymes during the first steps ofantagonism [9,10,23,25].

Because of the complexity of fungal cell walls, we were alsointerested in hydrolytic activities with the ability to degradeother non-abundant cell wall polysaccharides. It appears fromrecent studies with S. cerevisiae that b-1,6-glucan is the centralmolecule that links the major components of fungal cell walls,including b-1,3-glucans, chitin and mannoproteins [42].Therefore, enzymes hydrolyzing fungal b-1,6-glucans fromT. harzianum grown under the above conditions shouldcontribute, together with chitinases and b-1,3-glucanases, to aneffective breakdown of phytopathogenic cell walls duringantagonism. We have shown evidence of the presence of twoT. harzianum b-1,6-glucanase enzymes and reported on thepurification and characterization of an acidic one, namedBGN16.2 [12,27]. Here, we have purified and biochemicallycharacterized the second b-1,6-glucanase, which was calledBGN16.1. Only very few fungal b-1,6-glucanases have beenextensively described [12,14,17] and to our knowledge this andour previous work [12] are the only reports on purificationand characterization of b-1,6-glucanases from the genusTrichoderma.

The molecular mass of BGN16.1 was 51 kDa, as calculatedusing SDS/PAGE. This value is slightly higher than that ofBGN16.2 (43 kDa) and most other fungal b-1,6-glucanasesanalyzed, which are usually in the range 30±40 kDa. Gelfiltration on Sephacryl S-200 H gave a value of 20±25 kDa.Protein degradation did not account for such a low estimationbecause, once eluted from the gel-filtration column, the proteinrecovered its molecular mass of 51 kDa on SDS/PAGEanalysis. This result rather indicated an affinity for theSephacryl matrix, as previously reported for other fungal cell

wall hydrolases, among them the chitinase CHIT42 [23],BGN16.2 [12] and BGN13.1 [10] from T. harzianum and ab-1,3-glucanase from Trichoderma longibrachiatum [8]. Theaffinity for the Sephacryl matrix did not change when gelfiltration was performed at lower NaCl concentrations (0.5 m or0.1 m; data not shown). Despite this result, BGN16.1 isprobably monomeric because it maintained its activity onSDS/polyacrylamide gels. However, our results do not formallyexclude the possibility that BGN16.1 could be a homomericprotein.

Secreted hydrolases of fungi are normally glycosylated [43].In yeast, heterogeneous glycosylation of the same proteinproduct accounts for two different b-glucanases with differentelectrophoretical mobilities and physicochemical and kineticproperties [44]. Our results, using specific methods forcarbohydrate staining on SDS/polyacrylamide gels and treat-ments with deglycosidases, indicated that either the purifiedbasic BGN16.1 is not glycosylated or the level of glycosylationis so low as to be not detected under the conditions used. Theseresults, however, are similar to those found for otherextracellular fungal cell wall hydrolases from Trichoderma[10,12,23,45]. Immunoblotting experiments showed thatBGN16.1 did not have any common antigenic determinantwith BGN16.2, BGN13.1 and plant b-1,3-glucanases. Theseresults also indicate that BGN16.1 and BGN16.2 are differentand probably encoded by different genes. In full agreementwith this, the N-terminal sequence of BGN16.1 and sequencesof some tryptic peptides are not present in the deduced aminoacid sequence of BGN16.2 ([28]; J. de la Cruz and A. Llobell,unpublished work).

The optimal and the inactivation temperatures (both around50 8C) were similar to those found for other b-1,3-and b-1,6-glucanases from fungi. The Km value for pustulan(0.8 mg´mL21) was lower than that calculated for BGN16.2(2.4 mg´mL21) [12], but in the range calculated for other fungalb-1,6-glucanases also using linear b-1,6-glucans as substrate(1±3 mg´mL21).

The purified BGN16.1 enzyme was found to be specific forb-1,6-linkages in polysaccharides, hydrolyzing pustulan andyeast glucan very efficiently. It exhibited an endo type ofaction, as it gave a series of large b-1,6-oligosaccharides frompustulan as early hydrolytic products and gentitriose, gentibioseand glucose as major final hydrolytic products. The enzymealso displayed significant activity with laminarin. Thismay be due to the ability of the enzyme to hydrolyzeb-1,3-linkages involving 6-substituted glucose residues.This type of action pattern has been suggested for otherextracellular b-1,6-glucanases from certain filamentous fungi[14,16,17,46]. In agreement with this, BGN16.1 was not activewith pachyman, a linear b-1,3-glucan. The enzyme was alsounable to cleave gentibiose, a characteristic that has beenascribed to other non-specific b-1,3/b-1,6-glucanases (1,3-b-d-glucan glucanohydrolase; EC 3.2.1.58) with an exo type ofaction [29,44]. BGN16.1 and BGN16.2 from T. harzianumexhibit some similarities but also differences. BGN16.2 was notactive with laminarin, and it did not release glucose as an endproduct of hydrolysis. Like BGN16.1, it lacks b-glucosidaseactivity [12]. Therefore, a different mode of action with theirbiological b-1,6-glucan substrates would be expected for thetwo enzymes.

We presume that T. harzianum secretes an enzyme systemwith different activities and modes of action, which make itpossible to improve the rate of degradation of b-1,6-glucansubstrates in the phytopathogenic fungal cell walls. Thissystem must include at least BGN16.1 and BGN16.2,

150 J. de la Cruz and A. Llobell (Eur. J. Biochem. 265) q FEBS 1999

together with a b-glucosidase activity (b-d-glucosideglucohydrolase; EC 3.2.1.21), which may release glucosecontinuously from the non-reducing end of the b-glucanoligomers produced by the latter b-endoglucanases (J. de laCruz and A. Llobell, unpublished work). In this sense,BGN16.1 alone is able to generate reducing sugars fromyeast cell walls but unable to degrade filamentous fungal cellwalls. However, when it was combined with other fungal cellwall hydrolases from T. harzianum, such as BGN13.1 andBGN16.2 or chitinases, not only did BGN16.1 show hydrolyticactivity on fungal cell walls [12,27] but it also inhibited growthof phytopathogenic fungi [12]. Similar results have beendescribed for other combinations of Trichoderma fungal cellwall hydrolases [47±49]. Moreover, Trichoderma anti-fungalactivity is synergistically enhanced when hydrolases are used incombination with cell membrane-affecting antibiotics [50,51].Therefore, a similar synergism between fungal cell wall-degrading enzymes and antibiotics is expected to occur in vivoduring the antagonistic action of Trichoderma against phyto-pathogenic fungi [50]. Work is in progress to define further therole of b-1,6-glucanases in this process.

A C K N O W L E D G E M E N T S

We gratefully acknowledge Dr F. DomõÂnguez for his assistance during all

experimental work. We are very grateful to Dr Fritig for generously giving

the two tobacco antisera, A. RincoÂn for generously giving the BGN13.1

antiserum and Dr E. MartõÂnez-Force for help with HPLC experiments. We

thank Dr M. Rey and A. Soler for critical reading of the manuscript. J.d.l.C.

was a recipient of a fellowship from the Ministerio de EducacioÂn y Ciencia

(Spain). This work was supported by grants BIO91-1078 from CICYT

(Spain) and TS3-CT92-0140 from the European Community.

R E F E R E N C E S

1. Peberdy, J.F. (1990) Fungal cell walls: a review. In Biochemistry of Cell

Walls and Membranes in Fungi (Kuhn, P.J., Trinci, A.P.J., Jung, M.J.,

Goosey, M.W. & Copping, L.G., eds), pp. 5±24. Springer-Verlag,

Heidelberg.

2. Hrmova, M. & Fincher, G.B. (1993) Purification and properties of three

(1,3)-b-d-glucanase isoenzymes from young leaves of barley

(Hordeum vulgare). Biochem. J. 289, 453±461.

3. Reese, E.T. & Mandels, M. (1959) b-d-1,3-Glucanases in fungi. Can. J.

Microbiol. 5, 173±185.

4. Kauffmann, S., Legrand, M., Geoffrey, P. & Fritig, B. (1987) Biological

function of `pathogenesis-related' proteins: four PR proteins of

tobacco have b-1,3-glucanase activity. EMBO J. 11, 3209±3212.

5. Mrsa, V., Klebl, F. & Tanner, W. (1993) Purification and characteriza-

tion of the Saccharomyces cerevisiae BGL2 gene product, a cell wall

endo-b-1,3-glucanase. J. Bacteriol. 175, 2102±2106.

6. Watanabe, T., Oyanagi, W., Suzuki, K., Ohnishi, K. & Tanaka, H.

(1992) Three N-terminal domains of b-1,3-glucanase A1 are involved

in binding to insoluble b-1,3-glucan. J. Bacteriol. 174, 186±190.

7. Stahmann, K.P., Schimz, K.L. & Sahm, H. (1993) Purification and

characterization of four extracellular 1,3-b-glucanases of Botrytis

cinerea. J. Gen. Microbiol. 139, 2833±2840.

8. Tangarone, B., Royer, J.C. & Nakas, J.P. (1989) Purification and

characterization of an endo-1,3-b-d-glucanase from Trichoderma

longibrachiatum. Appl. Environ. Microbiol. 55, 177±184.

9. Noronha, E.F. & Ulhoa, C.J. (1996) Purification and characterization of

an endo-b-1,3-glucanase from Trichoderma harzianum. Can. J.

Microbiol. 42, 1039±1044.

10. de la Cruz, J., Pintor-Toro, J.A., BenõÂtez, T., Llobell, A. & Romero,

L.C. (1995) A novel endo-b-1,3-glucanase, BGN13.1, involved in the

mycoparasitism of Trichoderma harzianum. J. Bacteriol. 177, 6937±

6945.

11. BenõÂtez, T., LimoÂn, C., Delgado-Jarana, J. & Rey, M. (1998)

Glucanolytic and other enzymes and their genes. In Trichoderma

and Gliocladium: Enzymes, Biological Control and Commercial

Applications (Harman, G.E. & Kubicek, C., eds), pp. 101±127.

Taylor and Francis Ltd, London, UK.

12. de la Cruz, J., Pintor-Toro, J.A., BenõÂtez, T. & Llobell, A. (1995)

Purification and characterization of an endo-b-1,6-glucanase from

Trichoderma harzianum related to its mycoparasitism. J. Bacteriol.

177, 1864±1987.

13. Reese, E.T., Parrish, F.W. & Mandels, M. (1962) b-D-1,6-Glucanases

in fungi. Can. J. Microbiol. 8, 327±334.

14. Shibata, Y. & Fujimbara, T. (1973) A b-1,6-glucan 6-glucanohydrolase

from Gibberella fijikuroi. J. Ferment. Technol. 51, 216±226.

15. Rombouts, F.M. & Phaff, H.J. (1976) Lysis of yeast cell walls.

Lytic b-1,6-glucanase from Bacillus circulans WL-12. Eur. J.

Biochem. 63, 109±120.

16. Hiura, N., Nakajima, T. & Matsuda, K. (1987) Purification and some

properties of an endo-b-1,6-glucanase from Neurospora crassa.

Agric. Biol. Chem. 51, 3315±3321.

17. Pitson, S.M., Seviour, R.J., McDougall, B.M., Stone, B.A. & Sadek, M.

(1996) Purification and characterization of an extracellular (1!6)-b-

glucanase from the filamentous fungus Acremonium persicinum.

Biochem. J. 316, 841±846.

18. Knowles, J., Lehtovaara, P. & Teeri, T. (1987) Cellulases families and

their genes. Trends Biotechnol. 5, 255±261.

19. Papavizas, G.C. (1985) Trichoderma and Gliocladium: biology,

ecology and potential for biocontrol. Annu. Rev. Phytopathol. 23,

23±54.

20. Chet, I., Benhamou, N. & Haran, S. (1998) Mycoparasitism and lytic

enzymes. In Trichoderma and Gliocladium: Enzymes, Biological

Control and Commercial Applications (Harman, G.E. & Kubicek, C.,

eds), pp. 153±172. Taylor and Francis Ltd, London, UK.

21. Elad, Y., Chet, I. & Henis, Y. (1982) Degradation of plant pathogenic

fungi by Trichoderma harzianum. Can. J. Microbiol. 28, 719±725.

22. Ulhoa, C.J. & Peberdy, J.F. (1991) Regulation of chitinase synthesis in

Trichoderma harzianum. J. Gen. Microbiol. 137, 2163±2169.

23. de la Cruz, J., Hidalgo-Gallego, A., Lora, J.M., BenõÂtez, T., Pintor-

Toro, J.A. & Llobell, A. (1992) Isolation and characterization of

three chitinases from Trichoderma harzianum. Eur. J. Biochem.

206, 856±867.

24. Ridout, C.J., Coley-Smith, J.R. & Lynch, J.M. (1988) Fractionation of

extracellular enzymes from a mycoparasitic strain of Trichoderma

harzianum. Enzyme Microb. Technol. 10, 180±187.

25. GeremõÂa, R., Goldman, G.H., Jacobs, D., Ardiles, W., Vila, S.B., van

Montagu, M. & Herrera-Estrella, A. (1993) Molecular characteriza-

tion of the proteinase-encoding gene, prb1, related to mycoparasitism

by Trichoderma harzianum. Mol. Microbiol. 8, 603±613.

26. Bartnicki-GarcõÂa, S. (1968) Cell wall chemistry, morphogenesis and

taxonomy. Annu. Rev. Microbiol. 22, 87±109.

27. de la Cruz, J., Rey, M., Lora, J.M., Hidalgo-Gallego, A., DomõÂnguez,

F., Pintor-Toro, J.A., Llobell, A. & BenõÂtez, T. (1993) Carbon source

control on b-glucanase, chitobiase and chitinase from Trichoderma

harzianum. Arch. Microbiol. 159, 316±322.

28. Lora, J.M., de la Cruz, J., BenõÂtez, T., Llobell, A. & Pintor-Toro, J.A.

(1995) Molecular characterization and heterologous expression of an

endo-b-1,6-glucanase gene from the mycoparasitic fungus Tricho-

derma harzianum. Mol. Gen. Genet. 247, 639±645.

29. Fleet, G.H. & Phaff, H.J. (1974) Glucanases in Schizosaccharomyces.

Isolation and properties of the cell wall-associated b-1,3-glucanases.

J. Biol. Chem. 249, 1717±1728.

30. Somogyi, M. (1952) Notes in sugar determination. J. Biol. Chem. 195,

19±23.

31. Nelson, N.J. (1955) Colorimetric analysis of sugars. Methods Enzymol.

3, 85±86.

32. Bradford, M.M. (1976) A rapid and sensitive method for the

quantification of microgram quantities of protein utilizing the

principle of protein-dye binding. Anal. Biochem. 72, 248±254.

33. Laemmli, U.K. (1970) Cleavage of structural proteins during the

assembly of the head of the bacteriophage T4. Nature (London) 227,

680±685.

q FEBS 1999 A basic b-1,6-glucanase from Trichoderma harzianum (Eur. J. Biochem. 265) 151

34. Soler, A., de la Cruz, J. & Llobell, A. (1999) Detection of b-1,6-

glucanase isozymes from Trichoderma strains in sodium dodecyl

sulphate-polyacrylamide gel electrophoresis and isoelectrofocusing

gels. J. Microbiol. Methods 35, 245±251.

35. Dubray, G. & Bezard, G. (1982) A highly sensitive periodic acid-silver

stain for 1,2-diol groups of glicoproteins and polysaccharides in

polyacrylamide gels. Anal. Biochem. 119, 325±329.

36. VaÂzquez de Aldana, C.R., Correa, J., San Segundo, P., Bueno, A.,

Nebreda, A.R., Mendez, E. & del Rey, F. (1991) Nucleotide sequence

of the exo-b-1,3-glucanase-encoding gene, EXG1, of the yeast

Saccharomyces cerevisiae. Gene 97, 173±182.

37. Williamson, G., Belshaw, N.J. & Williamson, M.P. (1992) O-glyco-

silation in Aspergillus glucoamilase. Conformation and role in

binding. Biochem. J. 282, 423±428.

38. Kombrink, E., SchroÈder, M. & Hahlbrock, K. (1988) Several

`pathogenesis-related' proteins in potato are 1,3-b-glucanases and

chitinases. Proc. Natl Acad. Sci. USA 85, 782±786.

39. Roberstson, E.F., Dannelly, H.K., Malloy, P.J. & Reeve, H.C. (1987)

Rapid isoelectrofocusing in a vertical polyacrylamide minigel

system. Anal. Biochem. 167, 290±294.

40. Tanaka, H. & Phaff, H.J. (1965) Enzymatic hydrolysis of yeast cell

walls. I. Isolation of wall-decomposing organisms and separation and

purification of lytic enzymes. J. Bacteriol. 89, 1570±1580.

41. Grenier, J., Potvin, C. & Asselin, A. (1993) Barley pathogenesis-related

proteins with fungal cell wall lytic activity inhibit the growth of yeast.

Plant Physiol. 103, 1277±1283.

42. KollaÂr, R., Reinhold, B.B., PetraÂkovaÂ, E., Yeh, H.J.C., Ashwell, G.,

DrgonovaÂ, J., Kapteyn, J.C., Klis, F.M. & Cabib, E. (1997)

Architecture of the yeast cell wall. b-1,6-glucan interconnets

mannoproteins, b-1,6-glucan and chitin. J. Biol. Chem. 272,

17762±17775.

43. Yoshizumi, H. & Ashikari, T. (1987) Expression, glycosylation and

secretion of fungal hydrolases in yeast. Trends Biotechnol. 5,

277±281.

44. Nebreda, A.R., VaÂzquez de Aldana, C.R., Villa, T.G., Villanueva, J.R.

& del Rey, F. (1987) Heterogeneous glycosylation of the EXG1 gene

product accounts for two extracellular exo-b-glucanases of Sac-

charomyces cerevisiae. FEBS Lett. 220, 27±30.

45. Torronen, A., Mach, R., Messner, M., Gonzalez, R., Kalkkinen, N.,

Harkki, A. & Kubicek, C. (1992) The two major xylanases from

Trichoderma reesei: characterization of both enzymes and genes. Bio/

Technol. 10, 1461±1465.

46. Yamamoto, S., Kobayashi, R. & Nagasaki, S. (1974) Purification and

properties of an endo-b-1,6-glucanase from Rhizopus chinensis R-69.

Agric. Biol. Chem. 38, 1493±1500.

47. Lorito, M., Hayes, C.K., di Pietro, A., Woo, S.L. & Harman, G.E.

(1994) Purification, characterization, and synergistic activity of a

glucan 1,3-b-glucosidase and an N-acetyl-b-glucosaminidase from

Trichoderma harzianum. Phytopathology 84, 398±405.

48. di Pietro, A., Lorito, M., Hayes, C.K., Broadway, R.M. & Harman, G.E.

(1993) Endochitinase from Gliocladium virens: isolation, character-

ization and synergistic antifungal activity in combination with

Gliotoxin. Phytopathology 83, 308±313.

49. Sivan, A. & Chet, I. (1989) Degradation of fungal cell walls by lytic

enzymes from Trichoderma harzianum. J. Gen. Microbiol. 135,

675±682.

50. SchirmboÈck, M., Lorito, M., Wang, Y.-L., Arisan-Atac, I., Scala, F. &

Kubicek, C.P. (1994) Parallel formation and synergism of hydrolytic

enzymes and peptaibol antibiotics, molecular mechanisms involved

in the antagonism action of Trichoderma harzianum against

phytopathogenic fungi. Appl. Environ. Microbiol. 60, 4364±4370.

51. Lorito, M., Peterbauer, M., Hayes, C.K., Woo, S.L. & Harman, G.E.

(1994) Synergistic combinations of cell wall degrading enzymes and

different antifungal compounds enhances inhibition of spore

germination. Microbiology 140, 623±629.