APPLIED AND ENVIRONMENTAL MICROBIOLOGY, July 1994, p. 2609-2615 Vol. 60, No. 70099-2240/94/$04.00 +0Copyright ©) 1994, American Society for Microbiology

Purification, Characterization, and Substrate Specificities ofMultiple Xylanases from Streptomyces sp. Strain B-12-2

GRAZIANO ELEGIR,1 GEORGE SZAKACS,2 AND THOMAS W. JEFFRIES3*Institute for Microbial and Biochemical Technology, Forest Products Laboratory, Forest Service, U.S. Department of

Agriculture, Madison, Wisconsin 537053; University of Technical Sciences, Budapest Institute ofAgriculturalChemical Technology, Budapest, Hungary2; and Stazione Sperimentale per La Cellulosa,

Carta e Fibre Tessili Vegetali ed Artificiali, Milan 20133, Italy'Received 22 February 1994/Accepted 9 May 1994

The endoxylanase complex from Streptomyces sp. strain B-12-2 was purified and characterized. The organismforms five distinct xylanases in the absence of significant cellulase activity when grown on oat spelt xylan. Thisis the largest number of endoxylanases yet reported for a streptomycete. On the basis of their physiochemicalcharacteristics, they can be divided into two groups: the first group (xyl la and xyl lb) consists oflow-molecular-mass (26.4 and 23.8 kDa, respectively) neutral- to high-pI (6.5 and 8.3, respectively) endoxyla-nases. Group 1 endoxylanases are unable to hydrolyze aryl-o-D-cellobioside, have low levels of activity againstxylotetraose (X4) and limited activity against xylopentaose, produce little or no xylose, and form productshaving a higher degree of polymerization with complex substrates. These enzymes apparently carry outtransglycosylation. The second group (xyl 2, xyl 3, and xyl 4) consists of high-molecular-mass (36.2, 36.2, and40.5 kDa, respectively), low-pI (5.4, 5.0, and 4.8, respectively) xylanases. Group 2 endoxylanases are able tohydrolyze aryl-p-D-cellobioside, show higher levels of activity against X4, and hydrolyze xylopentaosecompletely with the formation of xylobiose and xylotriose plus limited amounts of X4 and xylose. The enzymesdisplay intergroup synergism when acting on kraft pulp. Despite intragroup similarities, each enzyme exhibiteda unique action pattern and physiochemical characteristic. xyl 2 was highly glycosylated, and xyl lb (but noother enzyme) was completely inhibited by p-hydroxymercuribenzoate.

Xylan is the major component of the hemicellulose presentin angiosperm cell walls (32). It is probably the second mostabundant carbohydrate polymer of plants. Xylans are hetero-geneous polysaccharides consisting of a backbone of 13-1,4-linked D-xylopyranosyl residues that often have O-acetyl, ar-abinosyl, and methylglucuronosyl substituents (35). Theypresent a relatively complex substrate that varies greatly fromplant to plant and from tissue to tissue. ao-Arabinofuranosi-dases (EC 3.2.1.55), acetylesterases (8) (EC 3.1.1.6), cx-meth-ylglucuronosidases (20, 26), and feruloyl esterases (15) act inconcert to remove the side chains before xylanases reduce thebackbone to xylooligosaccharides (17).Contemporary interest in xylan-degrading enzymes is due to

new applications in the prebleaching of kraft pulps (34) and totheir possible use in recovering fermentable sugars fromhemicellulose (38). This attention has led to characterizationof many new enzymes, some of which exhibit unusually highlevels of thermal stability (16, 18, 28) or alkaline activity (2, 23).Most of those characterized are endoxylanases (EXs).EXs (1,4-13-D-xylan xylanohydrolase; EC 3.2.1.8) attack the

internal backbone of the polymer, producing xylooligosaccha-rides with different degrees of polymerization. 1-Xylosidases(EC 3.2.1.37) remove successive D-xylose residues from thenonreducing termini (1, 25). Xylanases sometimes exhibitcooperativity in the hydrolysis of xylooligosaccharides andarabinosyl substituents (6, 37).

Notwithstanding much study in recent years, it is not yetknown why microorganisms exhibiting significant hemicellu-lase activity produce several EXs with different specificities (3,

* Corresponding author. Mailing address: Forest Products Labora-tory, One Gifford Pinchot Drive, Madison, WI 53705. Phone: (608)231-9453. Fax: (608) 231-9262. Electronic mail address: TWJeffri@Facstaff.Wisc.edu.

5, 37). The variety of these EXs could reflect the diversestructural features of hemicellulose, it could relate to thekinetics of hydrolysis, or it could reflect the various biophysicalenvironments within which the organisms grow.

Earlier studies have shown that some EXs produce mainlyxylose and xylobiose (X2), while others produce mainly oligo-saccharides with higher degrees of polymerization (DPs) (12,24). These groupings may or may not coincide with otherclassifications based on nucleotide sequences (11) or molecularweights and pl values (37). Because of the difficulties inpurifying multiple xylanases and determining their actionpatterns, there have been relatively few characterizations of allof the EXs produced by any one organism (5, 12, 30, 33).Recent advances in protein purification and product charac-terization techniques have facilitated this task.

In this paper we report the purification and characterizationof five EXs from a highly xylanolytic organism, Streptomyces sp.strain B-12-2, and their action patterns. The enzymes can bedivided into two broad groups on the basis of their molecularweights, isoelectric points, and substrate specificities; however,their action patterns-considered in total-appear to be dis-tinct from one another.

MATERIALS AND METHODS

Organism and culture conditions. Streptomyces sp. strainB-12-2 was isolated from soil (Vesima, Italy) by incubation at45°C. It was cultivated in the medium of Morosoli et al. (22) at45°C in Erlenmeyer flasks (50 or 250 ml) with 1% oat spelt orbirch xylan as the sole carbon source. For enzyme production,a spore suspension prepared from a 1-week-old plate culturewas primed by growth in Trypticase soy broth (Difco) at 45°Cwith shaking at 250 rpm for 24 h. A portion (5% [vol/vol]) ofthis culture was used to inoculate 500 ml of xylanase produc-

2609

2610 ELEGIR ET AL.

tion medium in a 2-liter Erlenmeyer flask. Xylanase produc-tion medium employs defined mineral salts with oat spelt xylanas the sole carbon source. Medium components were auto-claved separately from the xylan (13, 31). Maximum xylanaseactivity was generally detected after 48 h.Medium optimization. Several carbon sources were tested at

1% (wt/vol) each by using Morosoli's basal medium (22). Themedium was inoculated with a spore suspension in sterile water(2 x 107 spores per ml), and xylanase activity was determinedafter 2 and 3 days. Shake flask fermentations were carried outat 45°C and 300 rpm. Four replicate flasks were used for eachmedium. Five inexpensive organic nitrogen sources (defattedsoybean meal, rapeseed meal, sunflower meal, yellow peameal, and corn steep liquor with 50% dry matter) were alsotested at 0.3% (wt/vol) to replace yeast extract and BactoProteose Peptone in xylanase production medium. Experi-ments were performed as described for carbon sources.

Substrates. Acetylglucuronoxylan was prepared from birch-wood holocellulose by dimethyl sulfoxide extraction (14).Arabinoxylan from oat spelt, birch xylan, carboxymethyl cellu-lose, Remazol brilliant blue-xylan, p-nitrophenol (PNP)-,B-D-xyloside, and PNP-,B-D-cellobioside were obtained from Sigma.Xylooligosaccharides (X2 to xylopentaose [X5]) were fromMegazyme (North Rocks, Australia). Unbleached kraft pulpfrom southern red oak was kindly supplied by ConsolidatedPaper Inc. (Wisconsin Rapids, Wis.).Enzymatic assays. Xylanase activity was routinely deter-

mined by measuring the release of reducing sugars from either1% (wt/vol) alkali-soluble or water-soluble oat spelt xylan bythe Somogyi modification of the Nelson method (29) aspreviously described (13). Crude or purified enzyme prepara-tions were diluted appropriately to obtain maximal activityconsistent with a linear response. For comparative purposes,xylanase activities against 1% (wt/vol) water-soluble and water-insoluble xylan from either oat spelts or birch were measured.The dinitrosalicylic acid method of Miller (21) was used toassay xylanase activity during initial screening. Carboxymethylcellulase activity was assayed by replacing 1% xylan with 1%low-viscosity carboxymethyl cellulose. ,-Xylosidase was as-sayed as described by Bachmann and McCarthy (1) by using 5mM p-nitrophenyl-3-D-xylopyranoside in 50 mM potassiumphosphate buffer at pH 7.0. The values of the Michaelisconstant (Kin) and the maximum velocity (Vmax) were deter-mined from a Lineweaver-Burk plot of assayed activities overa range of substrate concentrations. Suitably diluted xylanaseswere incubated with alkali-soluble birchwood or oat spelt xylanat concentrations ranging from 0.5 to 10 mg/ml under the assayconditions described above.

Xylanase purification. The purification was performed ac-cording to a strategy similar to that of Grabski et al. (12).Xylanases from Streptomyces sp. strain B-12-2 were purifiedfrom crude extracellular broth by concentration, clarification,anion-exchange chromatography, and hydrophobic interactionchromatography. Chromatography was performed with a Phar-macia fast protein liquid chromatography system. Cells wereharvested by centrifugation (6,000 x g, 30 min), and thesupernatant solution was concentrated 10-fold by using a10,000-molecular-weight cutoff membrane (Minitan; Milli-pore). The retentate ('100 ml) was treated with 2 to 3%(vol/vol) bioprocessing acid-1000 (Rohm and Haas Co.) toprecipitate pigments and other contaminants (12). BPA-1000is a cross-linked polymer that has strong basic quaternaryammonium functional groups on its surface. The clear super-natant solution was diafiltered with 10 mM bis-Tris buffer (pH6.5) in a stirred ultrafiltration cell (Amicon Division, Grace &Co., Danvers, Mass.) equipped with a YM-3 Amicon disc

membrane (3,000-molecular-weight cutoff). The retentate wascentrifuged (10,000 x g, 10 min). The supernatant (50 to 60 mgof protein, -1,200 IU of xylanase) was applied to a Mono QHR 10/10 (Pharmacia) column. Proteins were separated byusing a discontinuous gradient of buffer A (10 mM bis-Tris, pH6.5) and buffer A plus 1.0 M NaCl (buffer B). The flow rate was4 ml/min, and 4-ml fractions were collected. Elution wasmonitored at 280 nm by using a UV detector. Active fractionswere pooled, concentrated, and diafiltered into 50 mM potas-sium phosphate (pH 7.0). Ammonium sulfate was added to afinal concentration of 1.25 M. Samples were microcentrifugedat 10,000 x g for 5 min. The supernatant solution (20 mg ofprotein per load) was applied (0.5 ml/min) to a PhenylSuperose (Pharmacia) column and eluted with a continuousgradient of buffer C (50 mM potassium phosphate [pH 7.0],1.25 M ammonium sulfate) and buffer D (50 mM potassiumphosphate, pH 7.0). Elution was monitored at 280 nm, and1-ml fractions were collected. Fractions were pooled, concen-trated, and diafiltered into 50 mM potassium phosphate (pH7.0) by using a Centricon-3 microcentrifuge diafiltration unit(Amicon). Purity of the samples was determined by sodiumdodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and isoelectric focusing.

Electrophoretic analysis. SDS-PAGE was performed on aPhast system (Pharmacia, Piscataway, N.J.) by using 10 to 15%gradient polyacrylamide gels. A low-molecular-weight stan-dard mixture (Bio-Rad) containing six proteins in the 14,400-to 97,400-molecular-weight range was used in order to deter-mine the apparent molecular weights (Mrs) of the samples.Protein bands were stained with Coomassie brilliant blueG-250. Isoelectric focusing was performed on a Bromma 2117Multiphor horizontal slab-gel system (Pharmacia LKB, Piscat-away, N.J.) by using Servalyt Precotes (pH 3 to 10; Serva,Heidelberg, Germany), as recommended by the supplier.Protein bands were revealed by staining with Serva Blue W.Zymogram analysis of EX activity in isoelectric focusing gelswas done according to a modification of the method of Biely etal. (7) by using Remazol brilliant blue-xylan (Sigma). Phos-phate buffer (0.1 M, pH 7.0) was used instead of acetate, andthe incubation period for overlaid gels was 15 min at 60°C.

Effect of pH and temperature on xylanase activity. Xylanaseactivity was measured at pHs from 4 to 10 under standard assayconditions with oat spelt xylan as the substrate. Buffers usedwere 100 mM sodium acetate (pH 4 to 5.5), 100 mM potassiumphosphate (pH 6 to 8), and 100 mM glycine-NaOH (pH 9 to10). Enzyme activities were also assayed at temperatures from30 to 70°C at pH 7.0.

Effect of the temperature on xylanase stabilities. Eachpurified enzyme (0.5 IU) was incubated at 60°C in 100 mMpotassium phosphate buffer at pH 7.0 in the absence ofsubstrate. Aliquots were removed at different times between 0and 3 h and immediately cooled on ice. Residual activity wasassayed under standard conditions.

Hydrolysis studies. The extent of hydrolysis of differentxylans was evaluated by measuring the release of reducingsugars from 0.25% substrate solutions in 50 mM phosphatebuffer (pH 7.0) at 40°C in the presence of 0.1 IU of purifiedenzyme in a final volume of 1.0 ml. Time course experimentswere terminated when an increase in reducing sugars was nolonger detected. Hydrolyses of aryl-o-D-xyloside (PNP-r3-D-xyloside) and aryl-p-D-cellobioside (PNP-1-D-cellobioside)were performed at 50°C by using 0.1 IU of purified enzyme and5 mM substrate in a final volume of 1.0 ml of the above-described buffer. The release of PNP groups was determinedspectrophotometrically at 400 nm. Synergism among purifiedxylanases was determined by using alkali-soluble oat spelt

APPL. ENVIRON. MICROBIOL.

XYLANASES FROM STREPTOMYCES SP. STRAIN B-12-2 2611

1 2 3 4 5 6 7 8

97.466.245.0

31.021.514.4

FIG. 1. SDS-PAGE analysis of purified EXs. Lanes 4 and 8,low-molecular-mass standards (Bio-Rad; in kDa); lane 1, crude super-natant; lane 2, xyl la; lane 3, xyl lb; lane 5, xyl 2; lane 6, xyl 3; lane 7,xyl 4.

xylan (1% [wt/vol]) or washed unbleached kraft pulp fromsouthern red oak (Quercus falcata) (10% [wt/vol]). In the caseof oat spelt xylan, hydrolyses were carried out for 24 h at 40°C;in the case of pulp, hydrolyses were carried out for 3 h at 60°C.Turbidity-clearing assays were performed at 50°C by using 0.5IU of each enzyme preparation (as described by Somogyi [29])plus 2 mg of water-insoluble xylan in a final volume of 1.0 mlof 50 mM phosphate buffer, pH 7.0. Turbidities (620 nm) wereread directly for up to 4 h.

Action pattern studies. The hydrolysis of xylans was per-formed as previously reported (24) by using soluble andinsoluble xylan from oat spelt and birchwood. Hydrolysisproducts of xylooligosaccharides were determined by incubat-ing 0.05 IU of purified EXs with 1 mg of X2, xylotriose (X3),xylotetraose (X4), or X5 in 1.0 ml of 50 mM phosphate buffer(pH 7.0) overnight at 50°C. Enzymes were inactivated byboiling for 10 min. Hydrolysates were analyzed by high-performance liquid chromatography with a CARBOPAC PAlcolumn equipped with a pulsed amperometric detector (Di-onex). The standard mixture contained 20 ,uM (each) xylooli-gosaccharide (X2 to X5), xylose, and L-arabinose.Other analyses. Protein concentration was measured by the

method of Lowry et al. (19) with bovine serum albumin as astandard. Proteins were precipitated with trichloroacetic acidto eliminate interfering substances (4). Glycoproteins weredetected on nitrocellulose blots with a glycan detection kitwhich is based on an enzyme immunoassay (Boehringer Mann-heim Corporation, Indianapolis, Ind.). In this method thehydroxyl groups of the sugar moieties are oxidized to aldehydeand linked to digoxigenin. Subsequently the digoxigenin-la-beled glycoproteins are detected by using a digoxigenin-spe-cific antibody conjugated to alkaline phosphatase. Total car-bohydrates were determined by the phenol-H2SO4 methoddescribed by Dubois et al. (10).

RESULTS

Induction and optimization of xylanase production. Variouscarbon sources induced Streptomyces strain B-12-2 to produceextracellular EX. The average titer, 74 IU/ml (as assayed bythe dinitrosalicylic acid method) or 20 IU/ml (as assayed by theSomogyi modification of the Nelson method), was obtainedafter 48 h of cultivation on 1% oat spelt xylan. Since oat speltxylan is expensive, various crude xylan-rich substrates were

TABLE 1. Physiochemical properties of EXs from Streptomyces sp.strain B-12-2

plH Residual Temperature Residual

Enzyme Mra optimum activity at optimum activity atotmm pH 9 (%) OC) 70-C(%

Group 1xyl 1a 26.4 7.5 6.0 8.0 55 16.0xyl1 b 23.8 8.3 6.0 10.0 60 7.0

Group 2xyl 2 36.2 5.4 7.0 32.0 60 46.0xyl 3 36.2 5.0 7.0 43.0 60 48.0xyl 4 40.5 4.8 6.0 1.0 60 34.0a Determined by SDS-PAGE.

evaluated for enzyme production. Several lignocellulosic ma-terials induced xylanase to various extents. Among these,ground, sieved cornstalks and wheat straw (<200 ,um) were thebest substrates. Xylanase activity could not be induced bycellulosic substrates (Avicel PH 102 and Solka Floc), glucose,or xylose. However, a small amount of xylanase activity wasdetected when Solka Floc SW 40 was used. This activity wasprobably due to contamination of the cellulose with xylan (27).Defatted rapeseed meal was a good replacement for both yeastextract and Bacto Peptone. When 2% ground and sievedcornstalks was used in combination with defatted rapeseedmeal, xylanase productivity was comparable to that obtainedwith pure xylan. The level of cellulase activity in the crudepreparation was very low (0.08 IU/ml) even when raw ligno-cellulosic materials were used as a carbon source.

Purification. Xylanases were purified from supernatant so-lutions of cultures grown in oat spelt xylan. Activity stainingafter isoelectric focusing was performed by using Remazolbrilliant blue-xylan. This analysis showed the presence of bothacidic and basic xylanases in the crude broths. No significantdifferences in enzyme activity profiles were found when cul-tures grown with cornstalks as the substrate were used (datanot shown). The supernatant solution of oat spelt xylancultures exhibited a dark coloration which could be eliminatedwith BPA-1000. Following anion-exchange chromatography(Mono Q), xylanase activity was detected primarily in theunbound fraction and in the fractions eluting between 120 and200 mM (i.e., xylanase 2 [xyl 2], xyl 3, and xyl 4). Fractionsshowing EX activity were further purified by hydrophobicinteraction chromatography. The unbound fraction was re-solved into two distinct xylanase peaks (xyl la and xyl lb).Following purification by hydrophobic interaction chromatog-raphy, the proteins were found to be more than 95% homo-geneous by SDS-PAGE (Fig. 1). Total recovery levels wereremarkably good. Approximately 55% of the original activitycould be accounted for by summing the activities in fractionsfollowing Mono Q separation, and 45% was present following

1 2 3 4 5 6 7

FIG. 2. Glycosylation analysis by enzyme immunoassay (glycandetection kit; Boehringer) of purified EXs. Lane 1, xyl la; lane 2, xyllb; lane 3, xyl 2; lane 4, xyl 3; lane 5, xyl 4; lane 6, transferrin (positivecontrol); lane 7, creatinase (negative control).

VOL. 60, 1994

I

2612 ELEGIR ET AL.

TABLE 2. Kinetic properties of EXs from Streptomyces sp. strain B-12-2

Oat spelt xylan Birchwood xylanEnzyme Sp act (nmol Km Vmax Kcata Sp act (nmol K Vmax Kcata

min-' mg-1) (mg/ml) (U mg-1) (s-') min-' mg-') (U mg-,) (U mg-) (s')

xyl la 97.4 5.8 162.5 71.5 105.2 3.4 141.8 62.4xyl lb 216.0 3.4 470.7 162.9 285.2 2.7 410.8 186.7xyl 2 138.3 3.0 210.0 126.7 117.5 1.1 169.0 101.9xyl 3 283.4 1.2 338.0 203.9 268.3 0.8 368.2 222.0xyl 4 224.8 0.8 243.0 164.0 242.7 0.5 279.1 184.9

a Molar turnover number.

phenyl superose separation. xyl 3 accounted for approximatelyhalf of the total recovered activity.

Characterization of the multiple endoxylanases. Table 1summarizes the physiochemical characteristics of five purifiedisoenzymes. Xylanases la and lb (group 1) were low-molecu-lar-weight enzymes with neutral and basic pIs, respectively.Xylanases 2, 3, and 4 (group 2) had higher molecular weightsand acidic pls. These latter enzymes, although they exhibitedsimilar temperature and pH optima, retained greater activity athigher temperatures and more activity at alkaline pH than didxyl la and xyl lb (Table 1). xyl 2, xyl 3, and xyl 4 retained about40% of their original activity after 3 h at 60°C (pH 7.0), whilexyl la and xyl lb were almost completely inactivated after 1 hof incubation. The group 2 enzymes were glycosylated tovarying degrees. xyl 2 is apparently strongly glycosylated, whilexyl 3 and xyl 4 are only weakly glycosylated (Fig. 2). Group 1enzymes are not glycosylated.

Substrate affinities. The kinetic parameters were deter-mined by using both arabinoxylan (oat spelt) and a lesser-substituted glucuronoxylan (birchwood xylan). All isoenzymeswere more active on birch xylan (Table 2). We observedsubstrate inhibition at higher xylan concentrations (>2.5 mg/ml)-especially with the acidic EXs-when we plotted the databy the Lineweaver-Burk method. We obtained the kineticparameters by extrapolating from the linear region of the dataset, and we did not include the rate data obtained at highersubstrate concentrations.

Effectors. We investigated the effects of metal ions (1 mM)and other agents on the activities of purified EXs (Table 3).The isoenzymes were not significantly influenced by Ca2" orMg2+ but were totally inactivated by Hg2' and N-bromosuc-cinamide. Other compounds such as EDTA, SDS, and com-pounds of Cu2+ and Fe2+ showed only partial inhibitiondepending on the enzymes. p-Hydroxymercuribenzoate af-fected only xyl lb, which was completely inactivated. The

TABLE 3. Effects of various reagents on the activityof purified EXs

Relative activity (%) of enzyme:Reagent

xyl la xyl lb xyl 2 xyl 3 xyl 4

CaC12 108 100 100 100 104MgCl2 100 100 100 100 100CuC12 90 85 59 72 74FeSO4 100 100 100 71 89HgCl2 0 0 0 0 0EDTA 80 82 100 100 100SDS 89 100 100 89 89p-Hydroxymercuribenzoate 100 0 90 96 95Phenylmethylsulfonyl fluoride 95 87 100 100 100N-Bromosuccinamide 0 0 0 0 0

different levels of inhibition by HgCl and p-hydroxymercu-ribenzoate suggest the presence of a relatively hydrophobicpocket at the catalytic site of xyl lb. Phenylmethylsulfonylfluoride, a classical inhibitor of serine proteases, did notinfluence the activity of any enzyme, indicating that serineresidues are not involved in the active site. Total inactivationdue to Hg2" and N-bromosuccinamide has already beenreported for xylanases of different origins (3, 9, 23) and, assuggested by Deshpande et al. (9), could be due to thepresence of a tryptophan residue which may be conservedamong all xylanases.

Substrate specificity. None of the purified EXs could releasereducing sugars from carboxymethyl cellulose, filter paper,avicel, or galactomannan (locust bean gum). Basic (group 1)EXs did not hydrolyze PNP-xyloside or PNP-cellobioside. Incontrast, all the acidic (group 2) EXs were able to catalyze thehydrolysis of PNP-cellobioside as was already reported forxylanase A from Streptomyces lividans (5), but none of thexylanases from Streptomyces sp. strain B-12-2 could hydrolyzePNP-xyloside.

Extent of hydrolysis. Acidic EXs (xyl 2, xyl 3, and xyl 4)achieved a higher degree of hydrolysis than basic EXs whenacting on acetylglucuronoxylan, oat spelt arabinoxylan, or birchglucuronoxylan (Table 4). Among the acidic EXs no significantdifferences were found in extents of hydrolysis. In contrast,basic EXs showed different capacities to hydrolyze acetylxylan.In fact, xyl la released almost as much sugar from thissubstrate as the acidic EXs did.

Synergism. The cooperative activity among the differentEXs was investigated by using both oat spelt xylan and a morecomplex substrate (unbleached red oak pulp). When substrate-limiting conditions were used (36), no synergism was detectedamong EXs in the degradation of the oat spelt arabinoxylan.Sugar release was additive and did not exceed the amountobserved with group 2 (xyl 2, xyl 3, and xyl 4) enzymes. Incontrast, experiments carried out with unbleached red oakpulp showed that the isoenzymes have different capacities torelease sugars from this substrate (Table 5). The level of totalsugars released from pulp by a combination of xyl la or xyl lband xyl 3 increased 15 to 25% (as measured by the phenol-H2SO4 method) over that obtained with either enzyme indi-vidually, demonstrating that synergism between these two

TABLE 4. Extent of hydrolysis of various xylans

% of substrate hydrolyzed by enzyme:Substrate

xyl la xyl lb xyl 2 xyl 3 xyl 4

Acetylglucuronoxylan 5.4 3.8 6.8 7.0 6.7Oat spelt 12.0 13.7 19.8 19.2 20.7Birchwood 16.0 16.9 24.2 24.3 26.5

APPL. ENVIRON. MICROBIOL.

XYLANASES FROM STREPTOMYCES SP. STRAIN B-12-2 2613

TABLE 5. Sugar release from hardwood pulp by purified EXsfrom Streptomyces sp. strain B-12-2

Amt of reducing Amt of total

Enzyme added sugars released sugars released Ratio(mg/g of oven- (mg/g of oven-dried pulp) dried pulp)

xyl laa 3.6 17.9 4.9xyl lb' 3.4 14.3 4.2xyl 3' 6.0 18.9 3.1xyl la + xyl lbb 4.2 19.6 4.6xyl la + xyl 3b 6.8 22.9 3.4xyl lb + xyl 3b 6.8 21.9 3.2

a A 5-IU portion of purified enzyme was added per g of oven-dried pulp.b A 2.5-IU portion of enzyme was added per g of oven-dried pulp.

groups of enzymes exists. The ratio of total sugars to reducingsugars revealed that group 1 (xyl la and xyl lb) tends toproduce larger xylooligosaccharides. When a mixture of xyl laand xyl 3 or xyl lb and xyl 3 was used, the ratio of total sugarsto reducing sugars decreased to the value obtained with onlyxyl 3, even though the level of total sugars released was greater.This fact suggests that xylooligosaccharides produced by xyl laand xyl lb can be further degraded by xyl 3 (Table 5).

Solubilization. Group 1 enzymes cleared the turbidity of asuspension of insoluble oat spelt xylan approximately twice asrapidly as group 2 enzymes did (data not shown).

Action patterns. We characterized the action patterns ofeach of the purified enzymes by using water-soluble andwater-insoluble oat spelt arabinoxylan and birch xylan (Fig. 3and 4). The experimental conditions used were selected tominimize the extent of hydrolysis and thus to enable theidentification of any intermediates. Although none of theenzymes were able to release arabinose, remarkable differ-ences were noted when different substrates were used. Thehydrolysis of both soluble and insoluble oat spelt arabinoxylanresulted in larger amounts of xylooligosaccharides with DPs of>5, and X3 was the main product regardless of the enzymeused. X2 was the major product following hydrolysis of solublebirchwood xylan for all enzymes except xyl la.

xyl la produced a greater amount of xylooligosaccharideswith DPs of >5 than the other xylanases on soluble oat

0

20

U)U

a.

C

a-

'a..2

o xi

a X4B 5

El DP >5

Substrate/EnzymeFIG. 4. Hydrolysis of insoluble birch (IB) xylan and insoluble oat

(10) arabinoxylan by the five purified EXs (xyl la, xyl lb, xyl 2, xyl 3,and xyl 4). The relative quantities of oligosaccharide products areshown as the moles percent of the total soluble sugars recovered fromeach analysis.

arabinoxylan or insoluble birch xylan. In contrast, xyl lb and xyl4 tend to accumulate small amounts of these products whenthey act on soluble or insoluble birch xylan. In the case ofinsoluble birchwood xylan, most enzymes formed X3 as themajor hydrolysis product along with X2, X4, and a minoramount of X.. However, xyl lb produced primarily X2 and noX5 from the soluble and insoluble birch substrates. Xyloserepresented only a small percentage of the total but waspresent as a product of all enzymes when they were acting onpolymeric substrates.The action patterns of the EXs were further investigated by

using purified xylooligosaccharides as substrates. None of theEXs showed activity on X2 or X3. They were able to cleave X4and X5 but to different extents. Figure 5 shows the molarpercentages of the products obtained from the hydrolyses ofthese latter substrates. Acidic (group 2) EXs completely hy-drolyzed X5s producing mostly X2 and X3 but also X4 and small

11

O x1

a x

GI DP:,5

cnCOc o c co co co

CO)CO Co CJU)V

Substrate/EnzymeFIG. 3. Hydrolysis of soluble birch (SB) xylan and soluble oat (SO)

arabinoxylan by the five purified EXs (xyl la, xyl lb, xyl 2, xyl 3, and xyl4). The relative quantities of oligosaccharide products are shown as themoles percent of the total soluble sugars recovered from each analysis.

C0U

0 ElxG

0 X4

Eax

>x> S >btaeEzmSubstrate/Enzyme

FIG. 5. Hydrolysis of xylooligosaccharides (xylotetraose, X4; xylo-pentaose, X5) by five purified EXs (xyl la, xyl lb, xyl 2, xyl 3, and xyl4). The concentrations of the individual products (Xl, X29 X3, X4, andX5) are given as the moles percent of the total soluble sugars recoveredfrom each analysis.

VOL. 60, 1994

2614 ELEGIR ET AL.

amounts of xylose. The hydrolysis of X4 and X5 by basic (group1) EXs was much slower, and insignificant amounts of xylosewere formed under the conditions used. The production of X3without the formation of xylose from X4 suggested that thebasic (group 1) enzymes, xyl la and xyl lb, carry out transgly-cosylation.

DISCUSSION

Analysis of the purified proteins by SDS-PAGE and isoelec-tric focusing indicated that the EXs produced by Streptomycesstrain B-12-2 may be classified as low-Mr, basic (group 1) andhigh-Mr, acidic (group 2) xylanases (Table 1) as suggested byWong et al. (37). None of the purified isoenzymes was able torelease arabinose from oat spelt xylan. This is consistent with ahigher affinity (lower Kin) for less-branched substrate (birch-wood xylan) than for the highly substituted oat spelt arabinoxy-lan (Table 2). Nevertheless, in the case of xyl 2, specific activityand V ..ax were higher when oat spelt xylan was used as thesubstrate. This suggests that a substituted substrate may berequired for maximal activity of xyl 2. It is noteworthy thatarabinoxylans (oat spelt and cornstalk) were the best inducersfor overall xylanase activity. In general, group 2 enzymesshowed higher affinities for their substrates than did group 1enzymes.The xylanases isolated from Streptomyces strain B-12-2 are

all endo-acting enzymes as demonstrated by their hydrolysisproducts (Fig. 3 through 5). However, the type of substrateused clearly influences their action patterns and the ratios ofthe main products. The greater amount of higher-DP xylooli-gosaccharides obtained from oat spelt xylan hydrolysis com-pared with that obtained from birchwood hydrolysis is proba-bly due to the higher degree of branching of the formersubstrate.The results obtained from the hydrolysis of the X4 and X5

substrates allowed a clear classification of the isoenzymesbased on their action patterns. We discerned differences in theratios among the products with polymeric substrates, butquantitative analysis was possible only up to a DP of 5. Thegroup 1 EXs (xyl la and xyl lb) appear to have a "more endo"action pattern in that they did not degrade X4 to a significantextent and did not fully hydrolyze X5. The group 2 EXs (xyl 2,xyl 3, and xyl 4) degraded X4 to various extents and completelyhydrolyzed X5, indicating that they are able to act on lower-DPsubstrates. Moreover, the mode of action of the group 2 EXsseemed to be different in that they formed small amounts ofxylose. This is probably due to a lower specificity in thebond-cleavage frequency of the acidic proteins of group 2 (33).The acidic group 2 EXs in the present study produced higherdegrees of hydrolysis with all tested xylans (Table 4).The group 1 enzymes could be clearly distinguished from

those of group 2 by their different action patterns on X4 andX5. In general, group 2 enzymes-having low pIs, high Mrs,and lower Kms4formed lower-DP products from X4 and X5(Fig. 5). In addition, basic (group 1) EXs clarified a suspensionof insoluble xylan much more rapidly and to a greater extentthan did acidic (group 2) EXs (data not shown). When group2 enzymes hydrolyzed birchwood glucuronoxylan, their actionpatterns were not clearly distinguishable from those of group 1enzymes.

xyl lb showed the highest specific activity (Vmax) againstboth birch and oat spelt substrates, but xyl 3 showed thehighest turnover number. When activities were normalized onthe basis of production of reducing groups as determined bythe Somogyi modification of the Nelson method, xyl lb wasfound to produce the most X2 from birch xylan; xyl la generally

exhibited the most extreme endo action pattern by producingoligosaccharides with higher DPs-especially on insolublesubstrates. It had the lowest Vmax and the highest Km.

Within group 1, xyl la and xyl lb were clearly different. xyllb was inactivated byp-hydromercuribenzoate, and its catalyticturnover rate was approximately two to three times higher thanthat of xyl la. Moreover, xyl lb degraded X4 to a greater extentand produced equal amounts of X2 and X3. In contrast, xyl laproduced mostly X3. No xylose was detected among theproducts. This suggests to us that a transglycosylation reactionoccurred. In fact, the production of X3 from X4 is not possiblewithout the formation of xylose. However, when we tested xylla and xyl lb on X2 and X3, transglycosylation was notdetected. Further research using end-labeled substrates wouldbe necessary to resolve the question, but transglycosylation hasbeen reported previously for EXs (33).The acidic xylanases (xyl 2, xyl 3, and xyl 4) were remarkably

similar in most physiochemical and kinetic characteristics. xyl 2could be distinguished from xyl 3 and xyl 4 by its high degreeof glycosylation (Fig. 2), its lower Vmax value on oat spelt orbirch xylan (Table 2), and its greater ability to degrade X4 (Fig.5). xyl 4 was much less active at pH 9 and less thermostable.These three enzymes showed similar product profiles whenused with polymeric xylan substrates, but xyl 4 exhibited amuch lower capability to degrade X4.

In previous work we classified the xylanase isoenzymes fromStreptomyces roseiscleroticus as endo-1 and endo-2 dependingon their hydrolysis products (24). Endo-1 xylanases tended tohydrolyze oat spelt xylan more completely, while endo-2xylanases produced greater amounts of larger oligosacchar-ides. Our current results seem to partially confirm the previousdata. However, with the enzyme complex from Streptomycesstrain B-12-2, using linear birch xylan in place of branched oatspelt arabinoxylan leads to a very different pattern with respectto unhydrolyzed substrate. The endo-I action pattern could beattributed to a greater ability to bind shorter oligosaccharidesand an ability to hydrolyze them (5). The different affinities inbinding could be especially important in the degradation ofcomplex substrates. Indeed, with increased substrate complex-ity, the accessibility of some moieties could be limited toxylanases having a smaller catalytic active site. In this respectwe found a synergistic action between acidic EXs and basicEXs only when a complex substrate (unbleached red oak pulp)was used.There are very few reports dealing with the complete

characterization of xylanase isoenzymes produced by strepto-mycetes. S. lividans is probably the best-characterized strepto-mycete in this respect (5). The results reported herein confirmthe greater catalytic versatility of the high-molecular-mass EXscompared with the low-molecular-mass EXs when they reactwith the substrates tested here. The action patterns and thecatalytic properties are similar among group 2 enzymes. Thissuggests that their multiplicity may be due to other properties.Only one acidic (group 2) enzyme (xyl A) is produced by S.lividans, and it appears to have properties somewhat differentfrom those of the acidic xylanases reported here. In contrast tothat of S. lividans, none of the Streptomyces strain B-12-2 acidicEXs is able to cleave X3 or PNP-xyloside. The shortestxylooligosaccharides hydrolyzed by both low-molecular-massbasic EXs and high-molecular-mass acidic EXs from B-12-2are X4 and X5, respectively. The acidic EXs from Streptomycesstrain B-12-2 are glycosylated. xyl 2 was the most glycosylatedof the Streptomyces strain B-12-2 xylanases; xyl 3 was the mostthermostable and alkali stable (Table 1), and it was not heavilyglycosylated. The production of isoenzymes having different

APPL. ENVIRON. MICROBIOL.

XYLANASES FROM STREPTOMYCES SP. STRAIN B-12-2 2615

physiochemical properties could enable the complex to act indifferent environmental conditions.

ACKNOWLEDGMENTS

We acknowledge Mark Davis for his assistance in analyzing actionpatterns by high-performance liquid chromatography.

This research was supported by USDA competitive grant no.92-37103-7984 from the Program on Improved Utilization of Woodand Wood Fiber and by the Italian National Council of Research.

REFERENCES1. Bachmann, S. L., and A. J. McCarthy. 1989. Purification and

characterization of a thermostable 3-xylosidase from Thermonos-pora fusca. J. Gen. Microbiol. 135:293-299.

2. Balakrishnan, H., M. D. Choudhury, M. C. Srinivasan, and M. V.Rele. 1992. Cellulase-free xylanase production from an alkalo-philic Bacillus species. World J. Microbiol. Biotechnol. 8:627-631.

3. Bastawde, K. B. 1992. Xylan structure, microbial xylanases, andtheir mode of action. World J. Microbiol. Biotechnol. 9:221-224.

4. Bensadoun, A., and D. Weinstein. 1976. Assay of protein in thepresence of interfering materials. Anal. Biochem. 70:241-250.

5. Biely, P., D. Kluepfel, R. Morosoli, and F. Shareck. 1993. Mode ofaction of three endo4- 1,4-xylanases of Streptomyces lividans.Biochim. Biophys. Acta 1162:246-254.

6. Biely, P., C. R. MacKenzie, J. Puls, and H. Schneider. 1986.Co-operativity of esterases and xylanases in the enzymatic degra-dation of acetyl xylan. Biofrechnology 4:731-733.

7. Biely, P., 0. Markovic, and C. Mislovicova. 1985. Sensitive detec-tion of endo-1,4-4-glucanases and endo-1,4-3-xylanases in gels.Anal. Biochem. 144:147-151.

8. Biely, P., J. Puls, and H. Schneider. 1985. Acetyl xylan esterases infungal cellulolytic systems. FEBS Lett. 186:80-84.

9. Deshpande, V., J. Hinge, and M. Rao. 1990. Chemical modificationof xylanases: evidence for essential tryptophan and cysteine resi-dues at the active site. Biochim. Biophys. Acta 1041:172-177.

10. Dubois, M., K. A. Gilles, J. K. Hamilton, P. A. Rebers, and F.Smith. 1956. Colorimetric method for determination of sugars andrelated substances. Anal. Biochem. 28:350-356.

11. Gilkes, N. R., B. Henrissat, D. G. Kilburn, R. C. Miller, Jr., andR. A. J. Warren. 1991. Domains in microbial 3-1,4-glycanases:sequence conservation, function, and enzyme families. Microbiol.Rev. 55:303-315.

12. Grabski, A. C., I. T. Forrester, R. Patel, and T. W. Jeffries. 1993.Characterization and N-terminal amino acid sequence of 1-(1-4)endoxylanases from Streptomnyces roseiscleroticus: purification in-corporating a bioprocess agent. Protein Expression Purif. 4:120-129.

13. Grabski, A. C., and T. W. Jeifries. 1991. Production, purification,and characterization of 3-(1,4)-endoxylanase of Streptomyces ro-seiscleroticus. Appl. Environ. Microbiol. 57:987-992.

14. Hagglund, E., B. Lindberg, and J. McPherson. 1956. Dimethylsul-phoxide, a solvent for hemicellulose. Acta Chem. Scand. 10:1160-1164.

15. Hatfield, R. D., R. F. Helm, and J. Ralph. 1991. Synthesis of methyl5-O-trans-feruloyl-c-L-arabinofuranoside and its use as a substrateto assess feruloyl esterase activity. Anal. Biochem. 194:25-33.

16. Holtz, C., H. Kaspari, and J. H. Klemme. 1991. Production andproperties of xylanases from thermophilic actinomycetes. AntonieLeeuwenhoek 59:1-7.

17. Jeffries, T. W. 1990. Biodegradation of lignin-carbohydrate com-plexes. Biodegradation 1:163-176.

18. Khasin, A., I. Alchanati, and Y. Shoham. 1993. Purification andcharacterization of a thermostable xylanase from Bacillus stearo-thermophilus T-6. Appl. Environ. Microbiol. 59:1725-1730.

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

20. MacKenzie, C. R., D. Bilous, H. Schneider, and K. G. Johnson.1987. Induction of cellulolytic and xylanolytic enzyme systems inStreptomyces spp. AppI. Environ. Microbiol. 53:2835-2839.

21. Miller, G. L. 1959. Use of dinitrosalicylic acid reagent for deter-mination of reducing sugars. Anal. Chem. 31:426-429.

22. Morosoli, R., J. L. Bertrand, F. Mondou, F. Shareck, and D.Kluepfel. 1986. Purification and properties of a xylanase fromStreptomyces lividans. Biochem. J. 239:587-592.

23. Nakamura, S., K. Wakabayashi, R. Nakai, R. Aono, and K.Horikoshi. 1993. Purification and some properties of an alkalinexylanase from alkaliphilic Bacillus sp. strain 41M-1. Appl. Environ.Microbiol. 59:2311-2316.

24. Patel, R. N., A. C. Grabski, and T. W. Jeffries. 1993. Chromophorerelease from kraft pulp by purified Streptomyces roseiscleroticusxylanases. Appl. Microbiol. Biotechnol. 39:405-412.

25. Poutanen, K., and J. Puls. 1988. Characteristics of Trichodemlareesei ,B-xylosidase and its use in the hydrolysis of solubilizedxylans. Appl. Microbiol. Biotechnol. 28:425-432.

26. Puls, J., 0. Schmidt, and C. Granzow. 1987. os-Glucuronidase intwo microbial xylanolytic systems. Enzyme Microb. Technol. 9:83-88.

27. Royer, J. C., J. S. Novak, and J. P. Nakas. 1992. Apparent celluloseactivity of purified xylanase is due to contamination of assaysubstrate with xylan. J. Ind. Microbiol. 11:59-61.

28. Simpson, H. D., U. R. Haufler, and R. M. Daniel. 1991. Anextremely thermostable xylanase from the thermophilic eubacte-rium Thzermotoga. Biochem. J. 277:413-417.

29. Somogyi, M. 1952. Notes on sugar determinations. J. Biol. Chem.195:19-23.

30. Tenkanen, M., J. Puls, and K. Poutanen. 1992. Two majorxylanases of Trichoderma reesei. Enzyme Microb. Technol. 14:566-574.

31. Tien, M., and T. K. Kirk. 1988. Lignin peroxidase of Phanerochaetechrysosporium. Methods Enzymol. 161:238-249.

32. Timell, T. E. 1967. Recent progress in the chemistry of woodhemicellulose. Wood Sci. Technol. 1:45-70.

33. Tuohy, M. G., J. Puls, M. Claeyssens, M. Vrsanska, and M. P.Coughlan. 1993. The xylan-degrading enzyme system of Talaromny-ces emersonii: novel enzymes with activity against aryl j-xylosidesand unsubstituted xylans. Biochem. J. 290:515-523.

34. Viikari, L., M. Pauna, A. Kantelinen, J. Sundquist, and M. Linko.1986. Bleaching with enzymes, p. 67-69. In Proceedings of theThird International Conference on Biotechnology in the Pulp andPaper Industry, Stockholm.

35. Whistler, R. L., and E. L. Richards. 1970. Hemicelluloses, p.447-469. In W. Pigman and D. Horton (ed.). The carbohydrates-chemistry and biochemistry, 2nd ed., vol. 2A. Academic Press,Inc., New York.

36. Wong, K. K. Y., L. U. L. Tan, and J. N. Saddler. 1986. Functionalinteraction among three xylanases from Trichoderma harzanium.Enzyme Microb. Technol. 8:617-622.

37. Wong, K. K. Y., L. U. L. Tan, and J. N. Saddler. 1988. Multiplicityof P3-1,4-xylanases in microorganisms: function and application.Microbiol. Rev. 52:305-317.

38. Woodward, J. 1984. Xylanases: functions, properties and applica-tions. Top. Enzyme Ferment. Biotechnol. 8:9-30.

VOL. 60, 1994