Cellulase-poor xylanases produced by Trichoderma reesei RUT C-30 on hemicellulose substrates

Appl Microbiol Biotechnol(1992) 38:315-322

Appfied Microbiology Biotecknology © Springer-Verlag 1992

Cellulase-poor xylanases produced by Trichoderma reesei RUT C-30 on hemicellulose substrates Gernot Gamerith 1, Rene Groicher 1, Susanne Zeilinger 2, Petra Herzog 2, Christian P. Kubicek 2 1 LenzingAG, Department of Research and Development, A-4860 Lenzing,Austria 2 Abteilung ffir Mikrobielle Biochemie, Institut ft~r BiochemischeTechnologicund Mikrobiologie, TU Wien, Getreidemarkt9, A-1060 Wien, Austria Received: 11 May 1992/Accepted: 27 July 1992

Abstract. Hemicellulose components from industrial viscose fibre production are characterized by a lower cellulose content than commercial xylan and the presence of a carboxylic acid fraction originating from the alkaline degradation of carbohydrates during the process. This substrate, after neutralization, can be used by Trichoderma reesei RUT C-30 for the production of cellulase-poor xylanases, useful for the pulp and paper industry. The yields of xylanase ranged up to almost 400 units/ml, with a ratio of carboxymethylcellulase/xylanase of less than 0.015. This crude xylanase enzyme mixture was shown to be superior to that obtained on beech-wood xylan when used for bleaching and, particularly, upgrading of hard-wood chemical pulp by selective removal of the xylan components. Biochemical studies indicate that the low cellulase production by T. reesei grown on these waste hemicelluloses is the result of a combination of at least three factors: (a) the comparatively low content of cellulose in these hemicellulosic wastes, (b) the inhibitory action of the carboxylic acid fraction present in the hemicellulosic wastes on growth and sporulation of T. reesei, and (c) the use of a mycelial inoculum that is unable to initiate the attack on the cellulose components within the carbon source.

Introduction Xylanases in the pulp industry Recently, the production of xylanolytic enzymes has received considerable attention because of their potential application in the pulp and paper industry (Paice and Jurasek 1984, Viikari et al. 1991). One of the approaches for reducing environmental pollution, particularly the release of chlorine-containing compounds during bleaching, is to use a xylanase pretreatment o f the pulp. This approach allows a significant reduction in

Correspondence to: G. Gamerith

bleaching chemicals, particularly chlorine compounds, in the following bleaching steps and thus lowers the pollution with absorbable organic halogenated (AOX) compounds (Koponen 1991) or even avoids any further chlorine need in bleaching (Eriksson 1992). The main problems with enzyme bleaching that the pulp industry is facing at present are the availability and cost of the enzymes, as well as the quality standards and residual cellulase contamination, which may pose severe problems in bleaching processes on technical scale. This is particularly important in the production of chemical pulp, where the selective removal of hemicellulose is an additional target of pulp bleaching. Oxygen or ozone, which are presently considered and investigated to reduce or eliminate chlorine compounds in chemical pulp bleaching, result in some degradation of cellulose by oxidative attack. With hot or cold caustic extraction, the degradation reactions of carbohydrates are even more pronounced. Therefore, treatment with a specific enzyme for xylan removal is advantageous to avoid losses in yield and quality.

Production o f xylanases The ability of various microorganisms to secrete enzymes capable of degrading xylans from plant material is well documented (Wong et al. 1988). Xylanases for pulp bleaching may be produced in different ways, most of them being proprietary and thus not published in detail. To keep production costs low, agricultural wastes are frequently suggested or used (Wizani et al. 1990). These substrates contain significant amounts of cellulase inducers and therefore may not be used for culturing a cellulase-excreting microorganism. Unfortunately, organisms with high xylanase productivity, particularly Trichoderrna sp., and Aspergillus sp., are also very good cellulase producers (Poutanen et al. 1987, Bailey and Poutanen 1989). Keeping this in mind, any method of enhancing the xylanase/cellulase ratio is an important goal of technical development. One approach is the construction of recombinant strains unable

316

to produce cellulases (Paice et al. 1988). In view of the multiplicity of the cellulase genes in most organisms (Beguin 1990), this is, however, a time-consuming task. On the other hand, it may theoretically be possible to induce xylanases selectively or to repress the cellulase by appropriate growth conditions. Although the regulation of xylanase formation in these organisms is still quite poorly understood, this can probably be achieved only by the use of pure xylan substrates, which have not been available in sufficient amounts for large-scale fermentation so far. In this paper we describe the successful use of waste hemicelluloses from viscose fibre production in the formation of xylanases by T. reesei RUT C-30 (Gamerith et al. 1990). Results are also be presented to explain the reason for the superiority of this substrate, which will aid in the development of further strategies in the production of cellulase-free xylanases.

107 conidia/ml. Cultivation was carried out at 280 C on an orbital shaker, operating at 120 rpm, for up to 7 days. For cultivation in fermentors, 101 of the same medium was inoculated with 1 1 of a 7-day-old preculture, grown on beechwood xylan (10 g/l) as a carbon source. The pH was maintained at 5.0-5.5 with 1 M NaOH. The air supply was controlled at 50% PO2, impeller speed held at 400 rpm, and temperature maintained at 30°C. For cultivation of resting mycelia in replacement media, the fungus was harvested after growth on glycerol (10 g/l) as a carbon source for 24 h, and then aseptically transferred to 100-ml erlenmeyer flasks containing 10 ml replacement medium as described by Sternberg and Mandels (1979). Cultivation was continued at 28°C for up to 24 h. For cultivation on solid media, the medium was solidified with 2.5% (w/v) agar, and 10-ml aliquots poured into petri dishes (11 cm diameter). After solidification, a hole of 3 mm diameter was drilled into the centre of the plate, 20 ~tl of a conidial suspension (108 conidia/ml) was applied into the hole, and the plates then incubated at 28°C for up to 7 days.

Enzyme assays. Xylanase and carboxymethylcellulase (CMC-ase)

Materials and methods Substrates. Xylan and xylan-containing solutions were derived from the dialysis plant of Lenzing. The first step in viscose fibre production is the formation of the sodium salt of cellulose, which results in (a) dissolving the hemicellulose components and (b) partiai degradation of these solubilized components to various hydroxy carboxylic acids (Fengel and Wegener 1984). The lye is pressed off and recycled in the process, resulting in an accumulation of the soluble organic compounds. To keep the concentration of these compounds within reasonable limits, part of the lye has to be dialysed against water to recover some sodium hydroxide and to remove the organic matter, which is emitted as a waste stream. In terms of fibre production, the alkali-soluble organics (hemicellulose) can be differentiated to acid-precipitable B-cellulose consisting of carbohydrate oligo- and polymers and y-cellulose, which is the acid-soluble fraction, chemically composed mainly of organic acids; In hardwood pulp the hemicellulose is constituted predominantly of xylan polymers (Gamerith and Strutzenberger 1992). Thus the alkaline lye contains primarily solubilized xylan and its degradation products. This solution or the xylan prepared from it was used in this work for preparation of a substrate for xylanase production. To obtain a pure xylan fraction, the lye was acidified with sulphuric acid to a pH of 2.5, centrifuged and washed with distilled water to remove the salts. The material consists of some 80% xylan and some 5% glucan, together with other sugars and inorganic compounds. This material is indicated as beech-wood xylan (Lenzing) or simply "xylan" in the following text. Alternatively, the lye is treated by a proprietary process to remove part of the inorganics, leaving a mixture of soluble carboxylic acids and xylan~ which can be used directly for preparing the growth medium. This carbon source is called substrate 2 in the text.

Organism and conditions f o r cultivation. The organism used in this study was T. reesei RUT C-30 (ATCC 56765). Stock cultures were maintained on potato dextrose agar slants at 28 ° C. For growth of the organism, the following medium was used (g/l)." ammonium sulphate, 1.5; urea, 0.3; M g s o n . 7 H 2 0 , 0.5; KH2PO4, 10; peptone, 6; carbon source as indicated. Tap water was used to prepare the medium; only when substrate 2 was used as a carbon source was tap water omitted and substrate 2 used to dissolve the medium. After adjusting to pH 5.0, the medium was distributed in 300-ml aliquots into 1-1 wide-mouthed erlenmeyer flasks, sterilized (121 ° C, 15 min), and then inocula~ted with a conidial suspension of T. reesei to give an initial inoc~lum density of

activities were assayed by essentially the same method, using beech-wood xylan (Lenzing AG HC 6/88, Austria) and carboxymethylcellulose (CMC; Serva no. 16110, Heidelberg, FRG) as substrates, respectively. For this purpose, 1% (w/v) suspensions of the respective substrate were prepared in 50 mM sodium citrate buffer, pH 4.8 and 1-ml aliquots preincubated in 22-ml test tubes for 2 min at 50 ° C. Then 0.5 ml enzyme solution was added, and incubation continued for a further 15 (xylanase) or 30 (CMC-ase) min. The incubation was terminated by the addition of 1 ml of 2.5 M N a O H and thorough vortexing. The amount of reducing sugars released was determined by the 3,5-dinitrosalicylic acid (DNS) method (Miller 1959), using glucose and xylose as calibration standards, respectively, The turbidity of the suspension was also considered in the calculations. Alternatively, the enzyme assays were performed using a microplate reader system. To obtain reproducible results, 100 Ixl of 1% xylan suspension or 133 ~tl of a 0.75% CMC solution were preincubated in a microplate as above followed by addition of 50 ~1 enzyme solution. The mixture was mixed for 5 min on a plate shaker and then incubated at 50 ° C for 30 min. Addition of 150 Ixl DNS reagent stopped the reaction. For xylanase assay it is necessary to add 50 ~tl of 2.5 M NaOH to solubilize the remaining xylan completely. After mixing for 5 min the microplates were incubated at 100 ° C for 20 min (xylanase) or 30 min (CMC-ase) and cooled. Absorbance was measured at 570 nm. The enzyme solution was diluted to obtain an optical density of 0.2 absorbance units.

Analytical Methods. Reducing sugars in the growth medium were quantified according to Miller (1959) or by H P L C after complete hydrolysis of oligomers. Hydrolysis was performed by adding 1.5 mt sulphuric acid (73.3%) to 25 ml of sample and boiling for 3 h. After cooling to room temperature the solution was diluted to a final volume of 50 ml and the sugars quantified by HPLC. For quantitative determination of extracellular protein, the bicichoninic acid-copper sulphate method (Smith 1985) was used. The organic compounds in substrate 2 were determined as total hemicellulose and differentiated as 13-cellulose (acid-insoluble) and y-cellulose (acid-soluble) by oxidation with K2Cr207 according to CCA 10/41 (CCA 1941).

Preparation o f carboxylic acid fraction. To obtain a pure carboxylic acid fraction the acids in the y-cellulose fraction were separated from residual carbohydrates by ion exchange using Lewatit MP 500. A sufficient volume of substrate 2 was treated with the exchange resin at pH 12.0, the carbohydrates removed with NaOH (pH 10.0), and the absorbed acids eluted with sulphuric acid. The yield, determined as y-cellulose content, was 40%. This procedure resulted in a sugar-free solution of carboxylic acids suited for incubation experiments.

317 B

A 400-

@

0

4

-'=-300. E []

oDD

[]

0

O

0

[]

•

[] 0

200

0

b3

0 O0

•

[]

x 100

i

C

6

E 4 0

o~

-~ "~

00 I

~2 • 00 0000

•

0 @0

0

O

0

•

•

• ,i

0

5'0 Fermentation time (h)

1 O0

20 18 16 14 12 10 8, 6 4 2 0

D

@ []

000 0000

[] [] []

5'0

0

Fermentation time (hi

Enzymatic treatment of pulp. To evaluate the effect of xylanases on the composition of pulp and to identify quality criteria for various enzymes, samples of Mg-bisulphite pulp from beech-wood (Lenzing) were treated with different quantities of enzymes, given as international units (IU)/g of pulp. Filtrates of various growth experiments were used for this purpose and added to raw pulp or by oxygen-peroxide extraction (EOP)-prebleached pulp samples to a 5°70content in erlenmeyer flasks. The pH was set to 5.0 by addition of sulphuric acid to the pulp suspension without any further buffer system. The pulp was shaken on a rotary shaker (120 rpm) at 50° C for 48 h. After filtration the carbohydrates in the filtrate were hydrolysed and the released sugars quantified by HPLC. The removal of carbohydrates after enzymatic treatment is given in grams of xylan or giucan released per kilogram pulp.

Electrophoretic methods. Isoelectric focusing (IEF) of the proteins from the culture filtrates was carried out using a Phast-System equipment (Pharmacia, Uppsala, Sweden). The gels were subjected to silver staining (Heukeshoven and Dernick 1985). Sodium dodecylsulphate-polyacrylamide gel electrophoresis (SDS-PAGE) under denaturing conditions was carried out according to Laemmli (1970). Separated proteins were stained by the Coomassie Blue procedure. Alternatively, the gels were blotted on nitrocellulose (Burnette 1981), and then subjected to immunostaining (Kubicek et al. 1987), using previously decribed monoclonal antibodies to detect cellobiohydrolase I (CBH I), cellobiohydrolase II (CHB II) and endoglucanase I (EG I) (Mischak et al. 1989; Luderer et al. 1991). The origin and properties of the monoclonal antibody KA 33.1 for the detection of endoxylanase I (IP 9.0, 22 kDa) will be described elsewhere.

Immunological quantification of xylanases and cellulases. For quantification of individual cellulases, an enzyme-linked immunoabsorbent assay (ELISA) method was used, essentially as described by Kolbe and Kubicek (1990). The modification of this method and the respective antibodies used to quantify T. reesei xylanases will be described elsewhere.

[] i

IO 0

Fig. 1A-D. Formation of biomass A, xylanase B, carboxymethylcellulase CMCase C, and uptake of substrate D during growth of Trichoderma reesei on beechwood xylan (123)and substrate 2 (0). Values given are means of two independently performed growth experiments in 10-1 fermentors: U, units

Results Growth and production o f xylanases and cellulases by T. reesei on beech-wood xylan and hemicellulosic wastes

In order to compare the suitability of the hemicellulosic wastes (substrate 2) with that of beech-wood xylan with respect to the production of xylanases and cellulases, T. reesei was pregrown on xylan as carbon source and then transferred to 10-1 fermentors containing one of both substrates, respectively, as carbon source (Fig. 1). Biomass was formed at a slower rate on substrate 2, which was reflected in a correspondingly slower substrate consumption rate. The formation of xylanases was also somewhat slower on substrate 2, but the final titres [up to 350-400 units (U)/ml after 100 h of cultivation] were roughly comparable. The formation of ceUulase displayed a more complex picture: on substrate 2, the activity of cellulase introduced with the inoculum decreased slowly throughout cultivation, probably due to enzyme inactivation. In contrast, during cultivation on xylan a sharp increase in cellulase activity was observed between 30 and 50 h of cultivation. No further cellulase formation occurred thereafter. This correlated with the time when most of the xylan had already been used up and the organisms had entered the stationary phase. Since the formation of cellulases by T. reesei requires the presence of an inducer and is not simply promoted by carbon depletion (Messner and Kubicek 1991), it was assumed that this temporary increase in cellulase formation on xylan was due to the consumption of cellulose after xylan, and that the presence of xylan apparently prevented the induction of cellulases in T. reesei. In order to support this assumption, an experiment was designed in which the carbon source used, either xylan or

318 Table 1. Effect of fed-batch experiments with xylan and substrate 2 (see Materials and methods) as carbon sources on the xylanase/ cellulase ratio

Table 2. Effect of substrate used for enzyme production on results

of beech sulphite pulp treated with xylanase Parameters

% CMC-ase

Xylan

Substrate 2

Batch Fed-batch

Batch Fed-batch

1.19

0.65

0.39

Xylanase amount in pulp treatment (U/g pulp) 21 100 30 100

0.32

Values are means of at least two independent growth experiments in fermentors and are percentage carboxymethylcellulase (% CMC-ase) calculated in relation to xylanase

Xylanase obtained by growing fungus on Xylan Substrate 2

Xylan removed (g/kg pulp) Glucan removed (g/kg pulp)

1.1 6.0 2.4 50.4

1.6 3.2 0.7 1.5

1.5

Enzymes were produced by batch-growth experiments in fermentots: U, units

~ 0.5

quisite for an economic process and cellulose solubilization has to be kept well below xylan removal. A similar effect of substrate influence on the properties of xylanase was observed when using. T. reesei QM 9414 instead of T. reesei Rut C-30 (data not shown).

>

0

Electrophoretic characterization o f the xylanases and cellulases formed on xylan and substrate 2 0

50 Incubation Time (hi

100

Fig. 2. Effect of xylan feeding (arrow) on cellulase formation (A) in a shake-flask culture; O, control without feeding

substrate 2 was fed to T. reesei cultures after one or two days of growth. The results, given in Table 1, are completely consistent with the above assumption, even in the case of substrate 2, where the fed-batch carbon source reduced the percentage of residual cellulase to even lower levels than without feeding. The effect of xylan feeding on the repression of cellulase f o r m a t i o n was also demonstrated in a shake-flask culture when the fungus was grown on xylan (Fig. 2).

Effect o f enzyme treatment o f sulphite pulp with xylanases produced on xylan and substrate 2 When comparing the efficiency of xylanase and the effect of residual cellulase content on the upgrading of beech sulphite pulp, the enzyme samples obtained by growth on xylan released some 20 times more glucose oligomers than those samples derived f r o m growth on substrate 2. Thus, even relatively small differences in residual cellulase content influence the properties and composition of enzyme-treated pulp in a very significant manner. Table 2 compares the effect of the enzymatic treatment on xylan and glucan (cellulose) release with different enzyme loads, using enzymes produced with conventional batch fermentation. Obviously, the xylan-removing effect is enhanced by partial cellulose degradation, but for enzymatic upgrad~ ing of pulps the selectivity for xylan removal is a prere-

IEF and S D S - P A G E were used to characterize the cellulases and xylanases formed on the two different xylancontaining carbon sources (Fig. 3). Evidence was obtained that growth on sustrate 2 preferentially reduced the secretion of proteins with an isoelectric point of 4.05.5 (Fig. 3a), and with molecular masses f r o m 40 to 65 k D a (Fig. 3b). Western blotting and immunostaining of the proteins f r o m S D S - P A G E confirmed that these protein bands were due to the presence of C B H I, C B H II and E G I (Fig. 3c). All three cellulases appeared in concomitantly reduced amounts on substrate 2. In contrast, the m a j o r xylanase, with an isolelectric point of 9.0 (Fig. 3a) and a molecular mass of 22 kDa, was formed in similar amounts on xylan as on substrate 2. This was supported by ELISA, yielding a mean value for endoxylanase I o f 125-145 ~tg/ml under both conditions.

Biochemical mechanisms involved in cellulase supression on substrate 2 Formation o f cellulases by T. reesei on media containing mixtures o f xylan and substrate 2. Xylan and substrate 2 have different contents of polysaccharides, and it is especially noted that substrate 2 contains a somewhat lower portion of cellulose when calculated in relation to total organic material (Table 3). In order to find out whether these different cellulose concentrations are responsible for the formation of different amount of cellulases, various total concentrations and ratios of xylan and substrate 2 were used as carbon sources for T. reesei (Table 4). The data obtained were basically consistent with this assumption, as they demonstrate an increase in cellulase activity correlating with an increase in the total cellulose

319

Fig. 3a-c. Electrophorefic characterization of the extracellular proteins secreted by T. reesei during growth on beech-wood xylan (lanes X) and substrate 2 (lanes S). The position of proteins of known isoelectric point (pI) or size (marker proteins, lanes C) is indicated by arrows, a Isoelectric focusing, b Sodium dodecyl sul-

phate-polyacrylamide gel electrophoresis, c Western blotting and immunostaining of proteins separated as in b for EX I, CBH II, CBH I and EG I, using monoclonal antibodies CH-6, CE-16 and EG1, respectively (Mischak et al. 1989; Luderer et al. 1991)

Table 3. Carbohydrate composition of carbon sources (xylan and substrate 2)

Table 4. Effect of different concentrations and ratios of xylan and substrate 2 on xylanase and cellulase formation by Trichoderma reesei RUT C-30

Component

Carbon source Substrate

Hemicellulose I~-Cellulose y-Cellulose Xylan Glucan

Xylan (°70 of OTS)

Substrate 2 (g/l) (°70of OTS)

(= 100) n.a. n.a. 83.0 6.9

27.0 11.0. 16.0 13.4 1.5

(= 100) 41.1 58.9 49.5 5.9

n.a., not applicable; OTS, organic dry matter

content o f the medium. However, the relatively small differences in cellulose content o f the two substrates alone is p r o b a b l y n o t sufficient to explain the significant variation in cellulase activities. Thus the presence o f other c o m p o n e n t s in substrate 2, which suppress the form a t i o n o f cellulases, is also apparent f r o m these data.

Induction o f cellulase formation by cellobiose and sophorose in the presence o f substrate 2. In order to find out whether c o m p o n e n t s f r o m substrate 2 repress cellulase f o r m a t i o n or interfere with their induction, two inducers o f T. reesei cellulase f o r m a t i o n - - s o p h o r o s e (Sternberg and Mandels 1979) and cellobiose (Fritscher et al. 1 9 9 0 ) - - w e r e added to cultures growing on substrate 2 (Table 5) or to resting mycelia replaced on substrate 2 (Fig. 4). In the latter case, also a carboxylic acid

Conc (o70 w/v)

Enzyme activities a (U/mg) Xylanase

Cellulase

0.5 1.0 2.0

23.4 19.8 27.9

1.2 1.7 2.3

0.5/0.5 0.5/1.0 1.0/0.5 1.0/1.0

25.4 30.2 21.3 24.5

1.4 1.1 1.6 2.0

0.5 1.0

23.8 24.7

0.3 0.6

Xylan

Xylan + Substrate 2

Substrate 2

a Enzyme activities were measured after 92 h of growth in shake flasks

fraction isolated f r o m substrate 2 (see below) was used. The results clearly show that substrate 2 had no effect on the ability o f T. reesei to f o r m cellulases. The induction o f cellulase by cellobiose was also not effected by substrate 2. This could be d e m o n s t r a t e d by growing the fungus on lactose or xylan with or without

320 Table 5. Effect of cellobiose on cellulase and xylanase formation by T. reesei RUT C-30 during growth on xylan or lactose, with and without substrate 2

25

Carbon source

20

Addition

Enzyme activity (U/mg) ~ Xylanase Cellulase

15 >

Xylan (2%, w/v) Xylan (2%, w/v) Xylan - substrate 2 (1/1%, w/v) Xylan - substrate 2 (1/1%, w/v) Lactose (2%, w/v) Lactose (2%, w/v) Lactose - substrate 2 (I/1%, w/v) Lactose - substrate 2 (1/1%, w/v)

23.4 1 mM cellobiose 29.1

0.7 1.6

-

28.7

0.2

1 mM cellobiose 29.2 8.7 1 mM cellobiose 10.5

2.5 1.7 2.5

5

-

34.6

1.1

0

1 mM cellobiose 36.1

1.9

£9

10

0

~ Values were measured a{ter 120 h of growth in shake flasks

50 Incubation Time, (h)

i00

50 i00 Incubation Time (h)

150

3O s 25

0.8

R >,20 a_l .¢.-i

~ 0.6

>

-~ 15 .1_1 ~.] -~

>

.,--~ .m]

0

0.4

~ 10

~5

.~ 02

0 0

0

5 i0 15 20 Incubation Time (h)

25

Fig. 4. Induction of cellulase formation by sophorose ( 4 mM) in T. reeseiafter replacement in a resting cell medium in the presence of substrate 2 (A), or the carboxylic acid fraction isolated from substrate 2 (©): D, control without any addition of inducer

Fig. 5. Induction of xylanase and cellulase using either a spore inoculum (a) or a mycelial inoculum ~). Induction was performed either with xylan (©), substrate 2 (A) or chemical pulp (I~). Enzyme activities are plotted as filled symbols (xylanase) or empty symbols (cellulase) #

addition of substrate 2, where the cellulase formation was induced with cellobiose (Table 5). The results confirm the inability of substrate 2 to repress cellulase induction.

Effect of a carboxylic acid fraction from substrate 2 on growth of T. reesei. Due to its origin from alkaline solubilization, substrate 2 contained a considerable amount of organic acids, created during alkalization by wellknown peeling reactions (LOwendahl et al. 1976). Since these acids represented the major difference between xylan and substrate 2, they were isolatedas a subfraction and their effect studied separately. When used either in submerged liquid culture or o n agar plates at a final concentration comparable to that in substrate 2, the carboxylic acid fraction inhibited growth by approximately

40% (data not showr0. In addition, on solid medium it almost completely repressed sporulation and thus stimulated colony growth only.

Growth and enzyme formation of conidial and mycelial inocula on xylan, substrate 2 and cellulose. In order to find out whether the growth- and sporulations-inhibiting properties of substrate 2 contributed to any extent to the suppression of cellulase formation, we investigated whether there is any difference in using either a mycelial or a conidial inoculum for the induction of xylanases and cellulases on xylan, substrate 2 and pulp. As expected, T. reesei formed xylanases at comparable rates on any of these carbon sources, and irrespective of the type of inoculum used (Fig. 5). Cellulase activity, however, was only formed in notable amounts on pulp, and

321 this was considerably delayed when a mycelial inoculum was used. Lowest activities were found with mycelial inocula on substrate 2. These findings, therefore, are consistent with the assumption that the use of a mycelial inoculum contributes to the low cellulase activity obtained during cultivation on substrate 2 as a carbon source.

Discussion When using xylanases for upgrading of chemical pulps, higher quantities of enzymes are necessary than in prebleaching of kraft pulps. Therefore the production costs as well as the residual a m o u n t of cellulase are even more important as quality criteria. The large effect of even small amounts of residual cellulase in our application clearly demonstrated the importance of a low cellulase level in the xylanase. There was, unexpectedly, even a marked difference between xylanases grown on isolated xylan versus xylan not isolated (substrate 2), when applied on chemical pulps. Thus enzyme samples containing some 1.2%0 residual CMC-ase (calculated to xylanase) are not well suited to this application, Whereas those samples grown on substrate 2, containing approx. 0.7% residual CMC-ase, showed only a little effect on the cellulose and are well suited to the selective removal of xylan. For pretreatment of kraft pulp, the difference in the effect of residual cellulase will not be as marked, due to a lower enzyme load. The substrates investigated in this work, particularly substrate 2, are well suited to the production of inexpensive, high-quality xylanases for pulp bleaching and upgrading processes by means of a well-known group of organisms. Furthermore, they are, in principle, available in amounts high enough to allow large-scale production of xylanases. Concerning the biochemical reasons for the improved suitability of the waste hemicelluloses, even compared to isolated xylan, evidence was obtained f r o m the present study that this is due to a decreased formation o f cellulases, rather than enhanced production of xylanases. The observed restriction of cellulase formation concerned all three m a j o r cellulases, C B H I, C B H II, and E G I, and is apparently not due to repression of synthesis or interference with induction, since addition of inducers (sophorose, cellobiose) rapidly switched on cellulase biosynthesis also in cultures growing on the waste hemicelluloses. F r o m the results obtained, the most plausible interpretation is that it is the use of the mycelial inoculum as well as the low content of cellulose present in substrate 2 that synergistically restricts cellulase induction. The role of cellulose is well supported by data f r o m the experiments, in which cellulase/xylanase formation was followed on different concentrations of xylan and substrate 2. The effect of different amounts of cellulose within substrate 2 and xylan is enhanced by the apparent inability o f mycelia to attack cellulose (Kubicek 1987). The hiphasic consumption of xylan and cellulose by a mycelial

inoculum of T. reesei has been well documented in this study, and it is especially stressed that sporulation had to occur before the fungus was able to attack cellulose after consumption of xylan. Finally, the carboxylic acids present in substrate 2 were also demonstrated to retard fungal growth and sporulation. A growth-inhibitory effect of other weak organic acids has also been observed with other microorganisms (Cherrington et al. 1991). During growth on substrate 2, the carboxylic acid fraction is apparently responsible for the decreased rate of growth on xylan, and the lack of initiation of growth on cellulose. Acknowledgements. This work was supported by gram no. 3/8018

from Forschungsf6rderungsfond for die gewerbliche Wirtschaft to Lenzing AG. We also acknowledge the skilful assistance of Sabine Stix and Agnes Metsz6ssy in enzyme assays and electrophoretic separations, and Hans Strutzenberger in preparing the various xylan-containing substrates.

References Bailey MJ, Poutanen K (1989) Production of xylanolytic enzymes by strains of Aspergillus. Appl Microbiol Biotechnol 30:5-10 Beguin P (1990) Molecular biology of cellulose degradation. Annu Rev Microbiol 44:219-248 . Burnette WN (1981) "Western blotting,: electrophoretic transfer of proteins from sodium dodecyl sulfate gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A. Anal Biochem 112:195-203 CCA (1941) Determination of beta and gamma cellulose CCA 10, accepted 1941. The Swedish Association of Pulp and Paper Engineers, Technical Communications, series CCA, Stockholm Cherrington CA, Hinton M, Mead GC, Chopra I (1991) Organic acids: chemistry, antibacterial activity and practicai applications. Adv Microb Physiol 32:87-105 Eriksson KEL (1992) Biotechnology will play key role in future non-chlorine bleaching. Pulp Paper 66, 2 1992:149-150 Fengel D, Wegener G (eds) (1984) Wood: chemistry, ultrastructure, reactions. Walter de Gruyter, Berlin, pp 296-303 Fritscher C, Messner R, Kubicek CP (1990) Cellobiose metabolism and cellobiohydrolase I biosynthesis by Trichoderma reesei. Exp Mycol 14: 405-415 Gamerith G, Strutzenberger H (1992) Fate of xylan during pulp and viscose production and it's use for xylanase manufacture. In: Visser J (ed) Xylan and xylanases. Elsevier, Amsterdam, in press Gamerith G, Groicher R, Prucha E (1990) Verfahren zur Herstellung einer ceIlulasearmen XylanaselOsung. AT patent 1036/90 Heukeshoven J, Dernick R (1985) Simplified method for silver staining of proteins in polyacrylamide gels and the mechanism of silver staining. Elektrophoresis 6:103-112 Kolbe J, Kubicek CP (1990) Quantification and identification of the main components of the Trichoderma cellulase complex with monoclonai antibodies using an enzyme-linked immunosorbent assay. Appl Microbiol Biotechnol 34:26-30 Koponen R (1991) Enzyme systems prove their potentiai. Pulp a. Paper International 53, 11:20-25 Kubicek CP (1987) Involvement of a conidial endoglucanase and a plasma membrane bound ~-glucosidase in the induction of endoglucanase synthesis by cellulose in Trichoderma reesei. J Gen Microbiol 133 : 1481-1487 Kubicek CP, Panda T, Schreferl-Kufiar G, Gruber F, Messner R (1987) O-Linked but not N-linked glycosylation is necessary for the secretion of endoglucanases I and II by Trichoderma reesei. Can J Microbflo[ 33 : 698-703

322 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227 : 680-685 LOwendahl L, Petersson G, Samuelson O (1976) Formation of carboxylic acids by degradation of carbohydrates during kraft cooking of pine TAPPI 59:118-121 Luderer MEH, Hofer F, Hagspiel K, Allmaier G, Blaas D, Kubicek CP (1991) A re-appraisal of multiplicity of endoglucanase I from Trichoderma reesei using monoclonal antibodies and plasma desorption mass spectrometry. Biochim Biophys Acta 1076: 427-434 Messner R, Kubicek CP (1991) Carbon source control of cellobiohydrolase I and II formation by Trichoderma reesei. Appl Environ Microbiol 57 : 630-635 Miller GL (1959) Use of dinitrosalicylic acid reagent for determination of reducing sugar. Anal Chem 31:426-428 Mischak H, Hofer F, Messner R, Weissinger E, Hayn M, Tomme P, Esterbauer H, Kachler E, Claeyssens M, Kubicek CP (1989) Monoclonal antibodies against different domains of cellobiohydrolase [ and II from Trichoderma reesei. Biochim Biophys Acta 990:1-7 Paice MG, Jurasek L (1984) Removing hemicellulose from pulps

by specific enzymic hydrolysis. J Wood Chem Technol 4: 187198 Paice MG, Bernier R, Jurasek L (1988) Viscosity-enhancing bleaching of hardwood kraft pulp with xylanase from a cloned gene. Biotechnol Bioeng 32: 235-239 Poutanen K, R/~tt6 M, Puls J, Viikari L (1987) Evaluation of different microbial xylanolytic systems, J. Biotechnol 6:49-60 Smith PK (1985) Measurement of protein using bichinoninic acid. Anal Biochem 150: 76-78 Sternberg D, Mandels G (!979) Induction of cellulolytic enzymes in Trichoderma reesei by sophorose. J. Bacteriol !39: 761769 Viikari L, Sundquist J, Kettunen J (1991) Xylanase enzymes promote pulp bleaching. Paperi ja Pu - Paper and Timber 73 : 385-389 Wizani W~ Esterbauer H, Steiner W, Gomes J (1990) Verfahren zur Herstellung exo- und endocellulasefreier Xylanase. AT patent 1030/90 Wong KKY, Tan LUL, Saddler JN (1988) Multiplicity of [~-l,4xylanase in microorganisms: function and application. Microbiol Rev 52: 305-317