Puri®cation and properties of an alkaline proteinase of Fusarium culmorum Anja I. Pekkarinen 1,2 , Berne L. Jones 3 and Marja-Leena Niku-Paavola 2 1 Department of Agronomy, University of Wisconsin-Madison, USA; 2 VTT Biotechnology, Finland; 3 USDA-ARS, Cereal Crops Research Unit, Madison, WI, USA The disease Fusarium head blight (scab) causes severe problems for farmers and for the industries that use cereals. It is likely that the fu ngi that cause scab (Fusarium spp.) use various enzymes when they invade grains. We are studying enzymes that the fungi may use to hydrolyze grain proteins. To do this, Fusarium culmorum was grown in a gluten-con- taining medium from which an alkaline serine proteinase with a molecular mass of 28.7 kDa was puri®ed by size- exclusion and cation exchange chromatographies. The enzyme was maximally active at pH 8.3±9.6 and 50 °C, but was unstable under these conditions. It hydrolyzed the syn- thetic substrates N-succinyl-Ala-Ala-Pro-Phe p-nitroanilide and, to a lesser extent, N-succinyl-Ala-Ala-Pro-Leu p-nitroanilide. It was inhibited by phenylmethanesulfonyl ¯uoride and chymostatin, but not by soybean trypsin or Bowman±Birk inhibitors. Parts of the amino-acid sequence were up to 82% homologous with those of several fungal subtilisins. One of the active site amino acids was d etected and it occupied the same relative position as in the other subtilisins. Therefore, on the basis of these characteristics, the proteinase is subtilisin-like. Puri®cation of the enzyme was c omplicated by the f act t hat, when puri®ed, it apparently underwent autolysis. The presence of extraneous protein stabilized the activity. Keywords: cereal; fungus; chymotrypsin; subtilisin; proteinase. Fusarium head blight (FHB, scab) has for many years been a serious problem for cereal producers and for the various cereal industries. The majority of past FHB epidemics have been caused by the f ungus Fusarium gram inearum (Gibberella zeae), but some infestations have also been due to F. culmorum and/or F. avenaceum (G. avenacea), espe- cially in Europe [1]. FHB causes severe yield losses in w heat and barley and reduces the crop quality by destroying some of the necessary grain components and by producing mycotoxin s. Fusarium contamination of malts is also associated with ÔgushingÕ problems t hat have sometimes plagued the brewing i ndustry [2±4]. The role that fungal proteinases play in the FHB pathogenesis is not known, but there are indications that these enzymes may contribute to some of the problems that are associated with diseased wheat. Electron microscope examinations have indicated that the wheat endosperm protein matrix disappeared when F. graminearum invaded the aleurone layer [5] or the s tarchy endosperm tissue [5,6]. F. graminearum infections also caused a d ecrease in the relative proportions of extractable wheat albumins a nd glutenins [7]. When either F. graminearum or F. culmorum was grown on m edia t hat contained c ereal proteins, it produced proteinases that had predominantly a lkaline pH optima [8]. An a lkaline proteinase activity that was associ- ated with the breakdown of storage proteins has also been detected in FHB-diseased wheat kernels [6]. Cereal grains contain multiple proteins that inhibit the activities of microbial proteinases [9,10] and it seems likely that they make some of these inhibitors to slow or prevent the disruption o f the grain proteins during fungal attacks. We are purifying, identifying and characterizing the pro- teinases that are synthesized by Fusarium fungi when they are grown on grain protein-containing media. These will then be used to probe for barley inhibitors that inactivate Fusarium proteinases, to de®ne the interactions between these enzymes and inhibitors and to ascertain whether or not they occur w ithin infested grains. In this paper we report the puri®cation and characterization of one of the protein- ases that is produced by F. culmorum. MATERIALS AND METHODS Fusarium culture F. culmorum (strain VTT-D-80148) was grown in Arm- strong medium that was modi®ed by replacing its inorganic nitrogen salt with gluten, so that it induced the fungus to Correspondence to A. I. Pekkarinen, USDA-ARS, CCRU, 501 N. Walnut St ., Madison, WI 53705±2334, USA. Fax: + 1 608 264 5528, Tel.: + 1 608 262 4478, E-mail: apekkarinen@facsta.wisc.edu Abbreviations: CMC, carboxymethyl cellulose; E-64, trans-Epoxy- succinyl- L -leucylamido-(4-guanidino)butane; PMSF, phenyl- methanesulfonyl ¯uoride; pAPMSF, p-amidino phenylmethanesulfonyl ¯uoride; CST, chymostatin; STI, soybean trypsin inhibitor; BBI, Bowman±Birk inhibitor; TLCK, Na-p-tosyl- L -lysine chloromethyl ketone; TPCK, N-tosyl- L -phenylalanine chloromethyl ketone; SAAPFpNA, N-succinyl-Ala-Ala-Pro-Phe p-nitroanilide; SAAP LpNA, N-succinyl-Ala-Ala-Pro-Leu p-nitro- anilide; BVGRpNA, N-benzoyl-Val-Gly-Arg p-nitroanilide; GPpNA, N-glutaryl- L -Phe p-nitroanilide; BApNA, Na-benzoyl- L -Arg p-nitro- anilide. Enzymes: chymotrypsin (EC 3.4.21.1); trypsin (EC 3.4.21.4); subtilisin (EC 3.4.21.62); oryzin (EC 3.4.21.63). (Received 16 August 2001, revised 20 November 2001, accepted 23 November 2001) Eur. J. Biochem. 269, 798±807 (2002) Ó FEBS 2002 produce a lkaline proteinases [8]. Two litres of inoculum was grown in Armstrong mineral medium [11] and it was used to start a 30-L culture that was identical, but contained 1 gáL )1 (NH 4 ) 2 SO 4 . This, in turn, was used to inoculate a ®nal 270-L growth medium preparation. The 300-L growth medium was as described previously [8], except that it was prepared with tap water. It contained 8 gáL )1 of an impure gluten preparation (80% protein, 7% fat, Sigma #G5004, St Louis, MO, USA) that had been dry-heat sterilized (10 h at 160 °C). Fermentation was in a N ew Brunswick Scienti®c IF400 fermenter. Throughout the fermentation, the p H of t he culture medium was maintained at 4.5±5.0 by adding either NaOH or H 3 PO 4 as needed, and the temperature was maintained at 18±21 °C. The fungal growth was monitored by measuring the glucose concentration and chymotrypsin/subtilisin pro- teinase activities of samples that were removed from the growth medium 0, 19, 24, 27, 31 and 42 h after the culture was started. The purity of the culture was con®rmed by agar plating and by microscopic examination. When  50% of the glucose had been used up (43 h), the m ycelia were separated by centrifugation with an Alfa Laval Separator Type BPTX 205SGD-30CDP (Sweden). This step yielded 270 L of supernatant that was concentrated to 16.6 L with four PCI modules (PCI, Whitchurch, UK) using ES625 membranes (nominal molecular weight cut-off 10 000 Da, 2.6 m 2 each). The concentrate was divided into appropriate aliquots, frozen, and stored at )20 °C. Analytical methods Nonspeci®c proteinase assay. An azogelatin assay [12] was used for analyzing the total nonspeci®c proteinase activities. Each reaction was started by adding 0.5 mL of enzyme preparation that was diluted (with 30 m M sodium citrate, pH 6.3) to contain 1±2 lgámL )1 of protein, to 2 mL of 12.5 mgámL )1 azogelatin in 100 m M buffer. Unless indicat- ed otherwise, the reactions were carried out at 40 °Cin 80 m M sodium citrate, pH 6.0, buffer. Samples (0.5 mL each) were removed from each reaction at appropriate times (normally 0, 10, 30 and 60 min), mixed with 0.75 mL of 25% trichlo roacetic acid, held in an ice-water b ath for 20 min, and centrifuged at 10 800 g for 10 min, at room temperature. The absorbance values of the supernatants were measured at 440 nm. The enzymatic activity, in arbitrary units (U), was the change in absorbance units per minute multiplied by 100. Each assay was performed in duplicate. Speci®c chymotrypsin/subtilisin-like activity assay.The speci®c chymotrypsin/subtilisin-like activities were mea- sured with the substrate N-succinyl-Ala-Ala-Pro-Phe p-nitroanilide (SAAPFpNA, Sigma, 5 m M ) dissolved in 175 m M Tris/HCl, pH 9 .0. The substrate solution (90 lL) was heated to 28 °C, 10 lL of appropriately diluted enzyme was added and the change in absorbance at 405 nm was monitored for 3 min. The activities were calculated as described earlier and are expressed as nkatámL )1 of sample [8]. Protein assay. For creating the puri®cation table, the protein concentrations of solutions were measured with the Bio-Rad Protein Assay kit (Bio-Rad, Hercules, CA, USA) [13]. Bovine serum albumin (BSA, Pierce, Rockford, IL, USA) was used to prepare a standard curve. In all other cases the protein contents of the enzyme solutions were calculated by measuring their absorbance at 280 nm and assuming that 1 mgámL )1 of protein had an absorbance of 1.0. Puri®cation of the enzyme All puri®cation steps were carried out at room temperature and the chromatography fractions were always collected in glass tubes that had been silanized with trimethylchloro- silane (Sigma #T-4252). For silanization, the tubes were placed in a dessicator with 5 mL of the reagent and the dessicator was evacuated and held at room temperature for 16 h, after which the derivatized tubes were rinsed thor- oughly with MilliQ-water. The concentrated culture medium was centrifuged at 1700 g for 5 min a nd 26.5 mL o f t he supernatant was appliedtoa2.5´ 70 cm Bio-Gel P30 (Bio-Rad) size exclusion column that was equilibrated with 20 m M NH 4 acetate, pH 5.0. The column was elut ed with the same buffer and 200-drop fractions were collected. The absor- bance values of t hese and of the chromatography fractions in subsequent steps were measured at 280 nm and their nonspeci®c proteinase and c hymotrypsin/subtilisin-like activities were both a nalyzed at pH 9.0. The fractions that voided the P-30 column were com- bined and subjected to carboxymethyl cellulose (CM52, Whatman) cation exchange chromatography at pH 5.0 on a 1 ´ 6 cm column. Elution was with a 20-m M to 300 m M NH 4 acetate, pH 5.0, linear gradient (45 mL of each concentration) and 2.8-mL fractions were collected. The fractions having the highest proteolytic activities (19±23) were pooled and the pool was divided into 2-mL aliquots that were stored at )20 °C. The enzyme was stored in this partially puri®ed state because it was not stable when completely puri®ed. Immediately prior to using the enzyme for studies, it was subjected to a ®nal HPLC-cation exchange puri®cation step with a Shodex IEC CM-825, 8 ´ 75 mm column (Phenomenex, Torrance, CA, USA). Each thawed aliquot was ®ltered through an AcrodiscÒ 4CR PTFE 0.45 lm ®lter (Gelman Sciences, Ann Arbor, MI, USA), diluted ®vefold with MilliQ-water, and ap plied to the column, which had been equilibrated with 50 m M NH 4 HCO 3 , pH 8, buffer. The loaded column was washed at 1 mLámin )1 with 6 mL of t he pH 8 buffer a nd the proteins were separated w ith a 12.5-mL linear gradient that ran f rom 50 m M to 175 m M NH 4 HCO 3 .The 280 nm-absorbing fractions were collected and immediately adjusted to pH 4±5 with approximately 40 lL of 2 0% ac etic acid. The activities of the fractions were measured at pH 9.0 and the material that showed chymotrypsin/subtilisin spe- ci®c activity was called the Ôpu ri®edÕ enzyme or ÔCM-HPLCÕ fraction. To ascertain its purity, an aliquot of the CM-HPLC preparation was boiled for 1.5 min with SDS sample buffer and separated on a 12% SDS/PAGE gel [14]. The gel was incubated for 45 min, with shaking, in 50% methanol/12% acetic acid. It was stained in 0.1% Coomassie Brilliant Blue R-250 dissolved in 40% methanol/1% acetic acid and destained with the same solvent. A Precision Protein Standard (Bio-Rad) sample was used to calibrate the gel. Ó FEBS 2002 An alkaline Fusarium proteinase (Eur. J. Biochem. 269) 799 Characterization of the proteinase Eects of pH and temperature on activity. The activities of the partially puri®ed enzyme, obtained by CMC separation at pH 5.0, were measured by using the azogelatin assay at 40 °C and pH 3.6, 4.1, 4.5, 5.0 and 5.5 (Na acetate), pH6.0,6.5and6.9(Nacitrate),pH6.9,7.5,7.9,8.3,8.7 and 9.1 (Tris/HCl) and pH 9.2, 9.6, 10.0 and 10.4 (Caps). All buffers were 80 m M . The activitie s of the CM-HPLC puri®ed proteinase were analyzed at pH 4.6, 6.0, 8.7 and 9.4 inthesamebuffers. The effect of temperature on the proteinase activity was studied at 45, 50 and 56 °CatpH6.0(80m M Na citrate) and at 40, 45 and 50 °CatpH8.7(80m M Tris/HCl). Eects of pH and temperature on the enzyme stability. For measuring the stability of the proteinase at different pH values, the puri®ed enzyme was incubated for 90 min at 40 °Cin30m M buffers: Na acetate, pH 4.1 and 4.9; Na citrate, pH 5.9 and 6.4; or Tris/HCl, pH 7.7 and 8.5. The activity of each sample was m easured after 0 and 90 min of incubation. For measuring its thermal stability, the puri®ed proteinase was incubated in 30 m M Na citrate, pH 6.3, for 50 min, at 24, 40, 50 or 60 °C. Its activity was measured after 0 and 50 min of incubation. The activity retention at each pH or temperature was expressed a s the proportion of the initial activity that remained after the incubation. Eects of class speci®c inhibitors on the enzyme. The mechanistic class of the proteinase was determined by measuring its activity at pH 6.0 in the presence of nine class- speci®c p rotease i nhibitors. Samples of the semipuri®ed (subjected to CMC open-column chromatography) prepa- ration were incubated on ice for 30 min with 50 l M trans- epoxysuccinyl- L -leucylamido-(4-guanidino)butane (E-64), 25 m M EDTA, 5 m M 1,10-phenanthroline in 20% dimethyl- sulfoxide, 50 l M pepstatin A in 20% m eth anol, 0 .5 or 5.0 m M phenylmethanesulfonyl ¯uoride (PMSF) in 20% isopropanol, 8.2 or 82 l M chymostatin (CST) in 20% dimethylsulfoxide, or 5.0 l M soybean trypsin inhibitor (STI, Type II-S, Sigma #T-9128). S amples of the puri®ed enzyme (about 2 lgámL )1 ) were incubated as above in the presence of 165 l M CST, 5.0 m M PMSF, or 5.0 m M 4-amid- inophenylmethanesulfonyl ¯uoride (pAPMSF), each of which was dissolve d in 20% dimethylsulfoxide, o r in a 12.5-l M solution of soybean Bowman±Birk inh ibitor (BBI), or 5.0 l M STI. Control reactions were carried out with enzyme that was preincubated in water or in 20% dimethylsulfoxide, methanol or isopropanol, as appropri- ate. In the ®nal reaction mixtures these enzyme-inhibitor mixtures were all diluted ®vefold with substrate solution. The effects of selected serine class proteinase inhibitors on the hydrolysis of azogelatin at pH 6.0 by several commer- cially available serine proteinases were examined. One set of assays was carried out with 48 lgámL )1 (2.0 l M ) bovine a-chymotrypsin (TLCK treated, Type VII, #C-3142, EC 3.4.21.1) or 1.0 lgámL )1 (0.04 l M ) bovine trypsin (TPCK treated, Type XIII, #T-8642, EC 3.4.21.4) in the presence of 1.0 m M PMSF, 1.0 m M pAPMSF, 33 l M CST, 1.0 l M STI or 2.5 l M BBI. In the other set, t he effects of 1.7 l M CST or 2.5 l M BBI on 0.8 lgámL )1 subtilisin Carlsberg (Type VIII bacterial, Bacillus licheniformis, #P-5380, EC 3.4.21.62) or 1.4 lgámL )1 oryzin (Aspergillus oryzae protease, Type XXIII, #P-4032, EC 3.4.21.63) were studied. The PMSF, pAPMSF and CST were dissolved in 4% dimethylsulfoxide. Appropriate controls were conducted. All of the inhibitors and commercial enzymes were pu rchased from Sigma. To ascertain the effects of CST and STI on the SAAPFpNA hydrolysis activity, samples of the puri®ed Fusarium proteinase were incubated in polypropylene tubes with 16.5 l M CST in 20% dimethylsulfoxide or 10 l M STI, on ice, for 50 min and their activities were measured as described above under Ôspeci®c chymotrypsin/subtilisin-like activity assayÕ except that the substrate buffer contained 4 % dimethylsulfoxide. For control reactions, the enzyme was preincubated with water or 20% dimethylsulfoxide. The ®nal concentrations of STI or CST in the reaction mixtures were 1.0 or 1.7 l M , respectively. Eects of calcium on the activity and stability of the proteinase. The effect of Ca acetate on the proteolytic activity was analyzed in both 80 m M Na citrate, pH 6.0 (0, 5 or 20 m M Ca 2+ )and80m M NH 4 acetate, pH 5.4 (0,1,5or20m M Ca 2+ ) buffers. To determine the effect of calcium on the enzyme stability, the puri®ed enzyme was incubated for 95 min at 40 °Cin30m M Na acetate, pH 4.9, that contained 0 or 100 m M Ca acetate. The activities of the samples were measured after 0 and 95 min of incubation. To ensure that the differing calcium levels did not affect the activity measurements, the Ca acetate concentrations of all of the reactio n mixtures were adjusted to 20 m M . Eect of added protein on the proteinase stability. A solution o f puri®ed enzyme was i ncubated at 40 °C for 90 min with 0, 1.0, 2.5, 5.0 or 10 lgámL )1 of BSA in 30 m M Na citrate, pH 6.3. The activity of the sample without BSA was measured as a control and the activities of all of the samples were analyzed after 90 min of incubation. The concentration of BSA in each of the activity analysis reaction mixtures was a djusted t o 4 lgámL )1 to ensure that the varying BSA levels did not affect the results. Substrate speci®city and kinetic constants for SAAPFpNA. To de®ne the substrate speci®city of the enzyme, its activity was measured in duplicate with the substrates SAAPFpNA, N-succinyl-Ala-Ala-Pro-Leu pNA (SAAPLpNA), N-gluta- ryl- L -Phe pNA (GPpNA), N-benzoyl-Val-Gly-Arg pNA (BVGRpNA), or Na-benzoyl- L -Arg pNA (BApNA). The concentrations of all substrates were 5 m M . The method was as described above under Ôspeci®c chymotrypsin/subtilisin- like activity assayÕ, except that all of the reactions contained 4% dimethylsulfoxide, in which the substrates were dis- solved. The reactions contained 0.06 (SAAPFpNA), 0.22 (SAAPLpNA and BVGRpNA) or 1.1 (BApNA and GPpNA) lgofprotein. The K m value of the proteinase for SAAPFpNA was determined by measuring the activity of the enzyme (0.03 lg protein per reaction) with 0.13±8.0 m M concen- trations of the substrate in pH 9.0, 175 m M Tris/HCl solutions in the presence or absence of 4% dimethylsulf- oxide and at pH 6.0 in 4% dimethylsulfoxide, 175 m M Na citrate. The analyses were carried out in duplicate and the kinetic constants were calculated from Lineweaver±Burk plots. 800 A. I. Pekkarinen et al. (Eur. J. Biochem. 269) Ó FEBS 2002 Eect of azogelatin concentration on the proteolytic activity. The effect of azogelatin on the proteolytic activities was measured with substrate concentrations between 0.5 and 10 mgámL )1 in solutions containing either 80 m M Na citrate, pH 6.0, or 80 m M Tris/HCl, pH 8.7. The enzyme concentrations of the pH 6.0 and 8.7 reaction mixtures were 0.44 and 0.22 lgámL )1 , respectively. Molecular mass analysis. The molecular mass of the proteinase was determined by MALDI-TOF analysis, using a Bruker Bi¯ex III mass spectrometer, at the University of Wisconsin Biotechnology Center, WI, USA. Determination of portions of the amino-acid sequence. CM-HPLC puri®ed enzyme was freeze-dried and submitted to the Protein Chemistry Laboratory of the University of Texas Medical Branch Cancer Center, Galveston, USA, for amino-acid sequence analysis. The enzyme was digested with trypsin, the resulting peptides were separated by reverse phase-HPLC and selected peptides were subjected to amino- acid sequence analysis using the Edman degradation method. RESULTS AND DISCUSSION Puri®cation Depending on which analysis method was used for measuring the activities, the ®nal yield of the proteinase was 5.5 or 11% (Table 1). The speci®c activities increased about fourfold (SAAPFpNA) or eightfold (azogelatin) as the puri®cation process progressed from culture medium concentrate to CM-HPLC preparation. The open column CMC separation gave the largest single puri®cation, but after this step two distinct proteinases were still present. To separate these two enzymes, it was necessary to carry out a ®nal CM-HPLC separation at pH 8 (Fig. 1). No separation was obtained when the HPLC separation was carried out at pH values lower than 8. A typical separation is shown in Fig. 1, where the enzyme of interest is indicated with an arrow. Preliminary studies established that the other major eluting peak contained a trypsin-like proteinase. This trypsin-like enzyme is being studied and will be reported elsewhere. SDS/PAGE analysis of the puri®ed enzyme showed that the predominant p rotein had a molecular mass of  26.8 kDa a nd that a small amount of slightly faster migrating proteins were present (Fig. 2). Mass spectrometry indicated that fresh enzyme preparations contained a protein of mass 28 663  50 Da. Some puri®ed samples also included small amounts of proteins of  4600, 11 180, 17 070 and 17 860 Da. Apparently, contaminants were sometimes present and/or the enzyme underwent partial autolysis. When the puri®ed enzyme was subjected to a Table 1. Puri®cation of the F. culmorum proteinase. Puri®cation step Protein a (mg) Activity Speci®c activity (U b )(lkat c )(Uámg )1 )(lkatámg )1 ) Yield (%) Puri®cation (fold) Crude 4.7 1400 4.7 300 1.0 100 b 100 c 1.0 b 1.0 c P30 pool 1.9 1300 1.8 680 0.95 93 b 38 c 2.3 b 1.0 c CMC pool 0.33 660 1.1 2000 3.3 47 b 23 c 6.7 b 3.3 c CM-HPLC 0.06 150 0.26 2500 4.3 11 b 5.5 c 8.3 b 4.3 c a Based on 26.5 mL of growth medium concentrate. b Activities were measured with the azogelatin assay: U  DA 440 ´ 100 min )1 . c Activities were measured with N-succinyl-Ala-Ala-Pro-Phe p-nitroanilide. Fig. 1. A typical cation exchange HPLC chromatogram of a F. cul- morum proteinase: absorbance at 280 nm (Ð). Thegradient( )was run from 50 m M to 175 m M NH 4 HCO 3 , pH 8. The proteinase that is described in this report is indicated with an arrow. Fig. 2. SDS/PAGE pattern of the puri®ed proteinase. Lanes 1 and 2: 0.5 and 1.5 lg of CM-HPLC puri®ed enzyme; lane 3: molecular mass standards. Ó FEBS 2002 An alkaline Fusarium proteinase (Eur. J. Biochem. 269) 801 second pH 8 HPLC separation, almost all of the proteinase was lost and what remained eluted as a broad peak. This repeated HPLC separation did remove some 17.9 k Da protein whose N-terminal amino-acid sequence was highly homologous with that of a portion of a trypsin-like proteinase from Fusarium ox ysporum (results not shown). Properties of the proteinase; pH optimum and effect of temperature on activity The proteinase hydrolyzed azogelatin between pH 4.0 and 10.5 and was optimally active at pH 8.3±9.6 (Fig. 3). This result was based on data obtained by analyzing the activities of a mixture of the two Fusarium proteinases. However, when the activities of the puri®ed enzyme were measured at pH 4.6, 6.0, 8.7 and 9.4, essentially identical results were obtained (Fig. 3). The enzyme thus functions at pH 6, the physiological pH of grain, but not at its maximal rate. The broad pH optimum of this proteinase resembles those of the alkaline proteinases from some Aspergillus species [15±18], but those of Fusarium sp. ÔS-19-5Õ [19] and F. graminearum [20] have somewhat sharper pH optima, at approximately p H 10. A t rypsin-like proteinase from F. oxysporum had a pH optimum of 8±11 [21]. At pH 8.7 (Fig. 4) the i nitial activity of t he enzyme increasedwithtemperature,upto50°C. However, the enzyme was unstable above 40 °C, so the reaction rate at 50 °C decreased with time. Thus, for the purposes of this assay method, 40 °C was the most appropriate temperature for carrying out the analyses. The proteinase was more temperature stable when assayed at pH 6.0, but even at this pH the activity was slightly un stable at 45 °C and it dropped off quickly at 56 °C. Several alkaline proteinases from Aspergillus and Fusarium species have temperature optima of  40 °C [15,16,18,20,22]. The alkaline proteinase of Fusarium sp. ÔS-19-5Õ was maximally active at 50 °Cand pH 10.5 when its activities were determined with a 20-min assay [19]. Calcium had a negligible effect on the activity of the F. culmorum proteinase in the presence of either Na citrate or NH 4 acetate buffer. The NH 4 acetate analyses w ere carried out to ensure that the calcium concentration of the reaction was not affected by the presence of citrate ion, which is a chelating agent. The enzyme therefore behaved like the alkaline proteinase of Fusarium sp. ÔS-19-5Õ,which also was not affec ted by calcium [19]. Factors affecting the stability of the proteinase The proteinase was heat labile and subject to inactivation at alkaline pH. When the puri®ed proteinase was incu- bated at various temperatures at pH 6.0, the remaining activities were 88, 55, 29 and 0% after 50 min at 24, 40, 50 and 60 °C, respectively. About a third of the enzyme remained active for 90 min at various pH values from 4.1 to 7.7 at 40 °C, but all of the activity was lost at pH 8.5. Calcium did not stabilize the enzyme at pH 4.9. After 95 min of incubation in the presence and absence of Ca 2+ , the remaining activities were 37 and 32%, respec- tively. Tomoda et al. [19] showed previously that calcium stabilized the alkaline proteinase of Fusarium sp. ÔS-19-5Õ at pH 8±9, but not at pH 5 a nd 40 °C. However, the ÔS-19-5Õ enzyme was not as sensitive to inactivation at pH 9 or at elevated temperatures as this F. culmorum proteinase. Bovine chymotrypsin and trypsin are stabilized by calcium [23,24]. The addition o f 2.5 lgámL )1 of BSA per  1 lgámL )1 of proteinase was suf®cient to completely maintain the proteinase activity for at least 90 min at pH 6.0 and 40 °C (Table 2), conditions under which the unp rotected enzyme was almost completely inactivated. BSA, and presumably other proteins, apparently protects the proteinase from autolysis, inhibits conformational changes or prevents it from binding to its containers. Hence, some of the s tability features of the enzyme may be affected by small amounts of contaminating proteins. Fig. 3. The eect of pH on the proteinase activity, measured with the azogelatin method. The enzyme preparations analyzed were: open symbols, a CMC pool or; close d symbols, e nzyme puri®ed by CM-HPLC. All buers were 80 m M and contained: Na acetate (s,d), Na citrate (h,j), Tris/H Cl (n,m)orCaps(e,r). Fig. 4. The eects of temperature on the proteinase activities. The analysis temperat ure s were: 40 °C(s), 45 °C(h,j), 50 °C(n,m), or 56 °C(r). The a ssays were run at pH 8.7 (open symbols) or pH 6.0 (closed symbols). 802 A. I. Pekkarinen et al. (Eur. J. Biochem. 269) Ó FEBS 2002 The enzyme is a serine class proteinase Class speci®c proteinase inhibitors other than those that inactivate serine proteinases had no signi®cant effect on the activities of the CMC preparation ( Table 3). The small effect of soybean trypsin inhibitor (STI) was probably due to some contaminating proteinase that was in the CMC preparation, because the puri®ed proteinase was not inhibited either by it or by a Bowman±Birk type trypsin- chymotrypsin inhibitor (BBI). The very strong inhibitions caused by PMSF and chymostatin (CST) indicated that the enzyme was a serine proteinase and belonged to eithe r the chymotrypsin or subtilisin family. Under the analysis conditions that were used in this experiment, a TLCK- treated bovine a-chymotrypsin was inhibited 64% by BBI and 98% by CST. Neither oryzin nor subtilisin were affected by BBI, but they were almost totally inhibited by CST. In this aspect, the F. culmorum proteinase resembles subtilisins. Howe ver, the a-chymotrypsin was unexpectedly inhibited by 40% in the presence of STI, showing that classi®cation by Ôspeci®cÕ inhibitors is not straightforward and may depend on the analysis method. The activity of the puri®ed F. culmorum proteinase was  25% higher in the presence of STI or BBI than in the control (Table 3). This is probably not due to an activation of the proteinase but rather to the general stabilization of the enzyme by proteins, which was mentioned earlier. Both STI and BBI are proteins. It is also possible that the small amount of contaminating trypsin-like p roteinase caused an inactivation of the ÔsubtilisinÕ proteinase and, when that contaminant was inhibited by STI or BBI, the proteinase remained active. STI and CST both caused similar effects when the enzymatic activities were measured with the synthetic peptide substrate (Table 3). Amino-acid sequence studies Several attempts were made to sequence the N-terminal amino acids of the enzyme, but no data were obtained, indicating that the N-terminus was probably blocked. The enzyme was therefore digested with trypsin, and the resulting peptides were separated by HPLC and some were sequenced. Four of the peptides obtained had sequences of: (1) GSTSYIYDTSAG SGTYAYIVDTGIITSHN; (2) GFNWAANDIISK; (3) SYSNYGTVL and; (4) DIFAPG TSVLSS. These peptides were homologous with sections of other f ungal proteinases ( Table 4). Peptides 3 a nd 4 occupied adjacent areas of th e sequences of several of these proteinases. However, the peptide bond that was cleaved to separate these two peptides connected the amino acids leucine and aspartic acid. Such bonds are not normally cleaved by trypsin, but are by subtilisin. This indicates that these peptides were probably separated by an autolytic e vent rather than by trypsin hydrolysis. In the subsequent discussion the peptides 3 and 4 are considered as a single peptide. Table 4 lists the corresponding amino-acid sequences of several homologous fungal proteinases. The protein that showed the highest homology with all three of the F. c ul- morum peptides was the subtilisin-like proteinase from Cephalosporium acremonium, whose corresponding sequen- ces were 82% identical to those of F. culmorum.The proteinases from several Aspergillus species, from Tricho- derma harzianum, Metarhizium anisopliae, Magnaporthe poae, Tritirachium album, Yarrowia lipolytica and the Fusarium sp. ÔS-19-5Õ and oxysporum contained sequences that were 44±76% identical. Subtilisin-like proteinases from M. poae [25] and F. oxysporum [26] have been detected in infected host plants, although, their roles have not been established. The proteinases that have been cloned f rom Fusarium sp. ÔS-19-5Õ and F. oxysporum showed less homology to the peptides from the F. culmorum proteinase tha n those of several other fungal species (Table 4). This was somewhat surprising, considering that most of the peptide sequences from the Aspergillus species were highly conserved. Also, the proteinases K and R from T. album were 85% identical to each other, showing that similar e nzymes from a single species are often highly homologous. However, such sequences may vary remarkably, as shown by the peptides from A. niger and M. anisopliae. The amino acids that comprise the catalytic triad of the serine proteinases (His, Asp and Ser) occur in d ifferent Table 2. Stabilization of the Fusarium proteinase by BSA. The samples were incubated at 40 °Cfor90minatpH6.0. BSA (lgámL )1 ) Remaining activity (%) 08 1.0 63 2.5 106 5.0 109 10.0 100 Table 3. Inhibition of the Fusarium proteinase activity by various class speci®c inhibitors. Inhibitor Concentration (l M ) Inhibition (%) a Inhibition (%) b E-64 10 2 ± EDTA 5000 8 ± 1,10-Phenathroline 1000 8 ± Pepstatin A 10 16 ± pAPMSF 1000 ± 85 PMSF 100 71 ± 1000 90 100 Chymostatin 1.7 74 100 c 17 74 ± 33 ± 100 Soybean trypsin inhibitor 1.0 35 )22 1.0 ± )38 c Bowman±Birk inhibitor 2.5 ± )25 a Measured with a mixture of two proteinases. b Measured with the puri®ed proteinase. c Measured with the substrate SAAPFpNA at pH 9.0. All of the other measurements were made with azogelatin at pH 6.0. Ó FEBS 2002 An alkaline Fusarium proteinase (Eur. J. Biochem. 269) 803 Table 4. Partial amino-acid sequences of fungal alkaline proteinases that show homology with the F. culmorum enzy me. Identical residues are indic ated with dashes and missing peptides with dots. The bold aspartic acid (D) indicates the position of the active site Asp r esidue that is conserved i n all of these f ungal subtilisins. The percentage of identical amino acids in all three peptides is listed, followed (in parentheses) by the percentages i n the individu al peptid es. Fungus Gene Sequences % Identity Database entry Reference F. culmorum ???GSTSYIYDTSA.GSGTYAYIVDTGIITSHN??? ???GFNWAANDIISK??? ???SYSNYGTVLDIFAPGTSVLSS??? Cephalosporium acremonium ALP 140 D V LE 168 231 V NR242 312-F S I A332 82 (86, 75, 81) SWISS-PROT #P29118 [29] Aspergillus niger PEPD 142S D D E V LAT 170 233 V 244 314-F S-V EQ A334 76 (72, 92, 71) TrEMBL #Q00206 [30] A. niger PEPC 161NFNK-L-ASEG E-VD TI NVD-V189 254-VEY-VQAH-K-265 344YF ECT LNI T364 44 (38, 33, 57) SWISS-PROT #P33295 [31] A. nidulans PRTA 141A T-V E V NAD-E169 232 V 243 313-F S-V QDI A333 73 (69, 92, 67) TrEMBL #Q00208 [32] A. fumigatus ALP 142A D A V S NVN-V170 233 V V 244 314-F S-V QDI A334 71 (69, 83, 67) SWISS-PROT # P28296 [33] A. ¯avus ALP 142P D NG E V I NVD-E170 233 V 244 314YF S-V QNI A334 68 (62, 92, 62) SWISS-PROT # P35211 [34] A. ¯avus None 142Q D E V S-VNVD-E170 233 V 244 314-F F-K-V-V QDI A334 68 (66, 92, 57) TrEMBL #Q9UVU3 [35] A. oryzae ALP 142Q D E V S-VNVD-E170 233 V 244 314-F F-K-V-V QDI A334 68 (66, 92, 57) SWISS-PROT # P12547 [36] Trichoderma harzianum PRB1 141 S A F V S N Q169 232-Y V V 243 318-FT AGV-V VNI 338 69 (76, 75, 57) SWISS-PROT #Q03420 [37] Metarhizium anisopliae PR1B 124 D E CS I DAT-P152 210SNVI M-FVA-221 296 P-V V-I T316 65 (69, 33, 76) TrEMBL #O14410 [38] M. anisopliae PR1D 136NA-G A Q FG-VM RAT-R164 227 D V V 238 314QG P DGIE-A334 61 (55, 75, 62) TrEMBL #Q9P806 [39] Magnaporthe poae mp1 133N D-V PAGL-ADH I LD-E-V162 230 S VK TA-241 311WF P-V-V VA-E-A331 56 (50, 58, 57) TrEMBL #Q9Y778 [25] Tritirachium album PROR 127-TST-R D Q CV-VI VEA P155 218-MDFV YRNR229 299-F S DI T319 56 (52, 33, 76) SWISS-PROT # P23653 [40] T. album PROK 124-TST-Y E Q-SCV-VI EA P152 215-MDFV-S-KNNR226 296-F S G I T316 55 (52, 25, 76) SWISS-PROT # P06873 [41] Fusarium sp. ÔS-19-5Õ ALP 120 A-A T Q-AC VI VEDT-P148 211-MDFV-S-YR-R222 292-F RA DIT-T312 55 (55, 33, 67) TrEMBL #Q02291 [42] F. oxysporum None 131 T N E C VI QLQDF158 215SAVI-GM-FV-G226 302E F-S-V-VL DI T322 52 (59, 25, 57) TrEMBL #O74236 [43] Yarrowia lipolytica XPR2 180-NYA-VRE-VG.KHP-VS-V S R-T-S208 271 T L Y 282 362QG CV-V SDII-S382 50 (38, 75, 52) SWISS-PROT # P09230 [44] 804 A. I. Pekkarinen et al. (Eur. J. Biochem. 269) Ó FEBS 2002 orders in the primary structures of the subtilisin (Asp-His- Ser) and chymotrypsin (His-Asp-Ser) families [27]. Peptide 1 from the F. culmorum proteinase contained the catalytic Asp residue in a position that corresponded to that of the other subtilisin-like enzymes (Table 4). This is another strong indication that the enzyme under study is a subtilisin- like, not a chymotrypsin-like, proteinase. These results therefore su pport t he observations made with the class speci®c inhibitors. However, the classi®cation of this enzyme still needs to be con®rmed by cloning its gene and determining its entire amino-acid sequence. The hydrolytic speci®city of the enzyme When the hydrolytic activities of the proteinase were measured at pH 9.0 with various synthetic substrates, the results listed in Table 5 were obtained. The enzyme hydrolyzed SAAPFpNA faster than the o ther two chy- motrypsin substrates, SAAPLpNA and GPpNA, indicating that it has a preference for phenylalanine over leucine and hydrolyzes small substrates poorly. The K m values for SAAPFpNA were 1.1±3.1 m M , depending on the compo- sition of the substrate buffer (Table 6). The K m values were scarcely affected by pH or the presence of dimethyl- sulfoxide, but the maximal velocity (V max )wastwiceas great at pH 9 as at 6. The alkaline proteinase from Aspergillus fumigatus also showed this preference for the phenylalanine substrate over the leucine one, did not hydrolyze short substrates (succinyl- L -Phe pNA or acetyl- DL -Phe pNA) and, in addition, had a similar K m value (0.62 m M ) for SAAPFpNA [18]. The hydrolysis of N-benzoyl-Val-Gly-Arg pNA (BVGRpNA), a putative trypsin substrate, may have been caused by a 17.9-kDa trypsin-like proteinase contaminant. When the activity of a subtilisin preparation was measured using this substrate in the presence of either STI or CST, only the STI inhibited. Effect of the azogelatin concentration on the enzyme activity As azogelatin is not a homogenous preparation containing only one protein form, but rather a mixture of proteins of varying sizes, a true K m value cannot be calculated. However, a Ôpseudo K m Õ, was computed for this substrate at pH 6.0 by nonlinear regression analysis using the Michaelis±Menten equation (Fig. 5). The ÔK m Õ was 1.6  0.3 mgámL )1 and the maximal activity (V max )was 0.93  0.05 U (DA 440 ´ 100 min )1 )per0.22lgproteinat pH 6.0. The observed maximal activity was  20% lower than the calculated V max . The activities measured at pH 8.7 could not be analyzed using Michaelis±Menten kinetics, because substrate inhibition occurred at concentrations >3 mgámL )1 (Fig. 5). General remarks The stability of the proteinase depended on several condi- tions. Either it adhered to the surfaces of containers or was inactivated by structural changes [28], because its activity was recovered much better from silanized glass tubes than from nonsilanized glass or plastic ones. The puri®ed proteinase could no longer be detected after it had been frozen at )20 °C or freeze-dried. In contrast, up to 80% of the SAAPFpNA hydrolyzing activity was recovered when the puri®ed enzyme preparation was stored on ice f or 2 weeks. However, dilute (less than 10 lgámL )1 )enzyme preparations were u nstable even when stored at 0 °C. CONCLUSIONS A proteinase, whose production by F. culmorum was induced/enhanced in the presence of grain protein, has been puri®ed from growth medium and characterized. The properties and the amino-acid sequence of the e nzyme indicated that it was related to several fungal subtilisins. The role this enzyme plays in F HB pathogenesis remains to be determined. Fig. 5. The eects of a zogelatin concentration on the proteinase activi- ties at pH 6.0 (s) and 8.7 (d). Thedashedlineshowsthenonlinear regression analysis curve for the pH 6.0 data. Table 6. K m values and maximal velocities (V max )oftheFusarium proteinase for N-succinyl-Ala-Ala-Pro-Phe pNA at pH 6 .0 and 9.0, with or without dimethylsulfoxide. pH; dimethylsulfoxide (%) K m (m M ) V max (nkatámg protein )1 ) 6.0; 4 3.1 1130 9.0; 0 1.1 1970 9.0; 4 2.3 2270 Table 5. Hydrolytic activities of the Fusarium proteinase measured at pH9.0with5m M synthetic substrates. Values are shown as mean  SD. Substrate Activity (nkatámg protein )1 ) N-Succinyl-Ala-Ala-Pro-Phe pNA 1360  40 N-Succinyl-Ala-Ala-Pro-Leu pNA 345  32 N-Glutaryl- L -Phe pNA 0.2  0.0 N-Benzoyl-Val-Gly-Arg pNA 145  5 Na-Benzoyl- L -Arg pNA 2.1  1.2 Ó FEBS 2002 An alkaline Fusarium proteinase (Eur. J. Biochem. 269) 805 ACKNOWLEDGEMENTS We thank the Kemira Foundation for funding the proteinase production and Michael Bailey at VTT Biotechnology for culturing the F. culmorum. The ®nancial support of the Tor-Magnus Enari F und, the Raisio Group Research Foundation, the American Malting Barley Association and the Finnish Concordia Association are also greatly appreciated. REFERENCES 1. Parry, D.W., Jenkinson, P. & McLeod, L. (1995) Fusarium ear blight (scab) in small grain cereals ± a review. Plant Path. 44, 207±238. 2. Sloey, W. & Prentice, N. (1962) Eects of Fusarium isolates applied during malting on properties of malt. ASBC Proc. 24±29. 3. Gjertsen, P., Trolle, B. & Andersen, K. (1965) Studies on gushing II. Gushing caused by microorganisms, specially Fusarium specie s. EBC. Proc. 428±438. 4. Schwarz, P.B., Beattie, S. & Casper, H.H. (1996) Relationship between Fusarium infestation o f barley and the gushing potential of malt. J. Inst Brew. 102, 93±96. 5. Bechtel, D.B., Kaleikau, L.A., Gaines, R.L. & Seitz, L.M. (1985) The eects o f Fusarium gramin ear um infection on wheat kernels. Cereal Chem. 62, 191±197. 6. Nightingale, M.J., Marchylo, B.A., Clear, R.M., Dexter, J.E. & Preston, K.R. (1999) Fusarium head blight: eect of fun gal pro- teases on wheat storage proteins. Cereal Chem. 76, 150±158. 7. Boyacioglu, D. & Hettiarachchy, N.S. (1995) Changes in some biochemical compone nts of wheat grain that was infected with Fusarium graminearum. J. Cer. Sci. 21, 57±62. 8. Pekkarinen, A., Mannonen, L., Jones, B.L. & Niku-Paavola, M L. (2000) Production of proteases by Fusarium species grown on barley grains and in media containing cereal proteins. J. Cer. Sci. 31, 253±261. 9. Boisen, S. (1983) Protease inhibitors in cereals. Acta Agric. Scan. 33, 369±381. 10. Ryan, C.A. (1990) Protease inhibitors in plants: genes for improving defenses against insects and pathogens. Ann. Rev. Phytopath. 28, 425±449. 11. Booth, C. (1977) Fusarium La boratory G uide to the I denti®cation of the Major Species, Commonwealth Agricultural Bureaux, Kew, Surrey, UK. 12. Jones, B.L., Fontanini, D., Jarvinen, M. & Pekkarinen, A. (1998) Simpli®ed endoproteinase assays using gelatin or azogelatin. Anal. Biochem. 263, 214±220. 13. Bradford, M. (1976) A rapid and sensitive method for the quanti- tation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248±254. 14. Laemmli, U.K. (1970) Cleavage of structural proteins d uring the assembly of the head of bacteriop hage T4. Nature 227, 680±685. 15. Ohara, T. & Nasuno, S. (1972) Enzymatic properties of alkaline proteinase from Aspergillus candidus. Agr. Biol. Chem. 36, 1797± 1802. 16. Kundu, A.K. & Manna, S. (1975) Puri®cation and characteriza- tion of extracellular proteinases of Aspergillus oryzae. Appl. Microbiol. 30, 507±513. 17. Impoolsup, A., Bhumiratana, A. & Flegel, T.W. (1981) Isolation of alkaline and neutral proteases from Aspergillus ¯avus var. columnaris, a soy sauce koji mold. Appl. Env. Microbiol. 42, 619±628. 18. Larcher, G ., Bouchara, J P., An naix, V., Symoens, F., Chabasse, D. & Tronchin, G. (1992) Puri®cation and characterization of a ®brinolytic serine proteinase from Aspergillus fumigatus culture ®ltrate. FEBS 308, 65±69. 19. Tomoda, K., Miyata, K., Maejima, K., Nakamura, M., Kuno, M . & Isono, M. (1979) Production, puri®cation and g eneral proper- ties of Fusarium alkaline protease. J. Takeda Re s. Labor atory 38, 33±43. 20. McKay, A.M. (1992) Production of an alkaline protease by Fusarium graminearum grownonwhey.Milchwissenschaft 47, 147±148. 21. Rypniewski, W.R., Hastrup, S., Betzel, Ch., Dauter, M., Dauter, Z., P apendorf, G., Br anner, S. & Wilson, K .S. (1993) The sequence and X-ray structure of t he trypsin from Fusarium oxysporum. Prot. Engin. 6, 341±348. 22. Ogundero, V.W. & Osunlaja, S.O. (1986) The puri®cation and activities of an alkaline protease of Aspergillus clavatus from Nigerian poultry feeds. J. Basic Microbiol. 26, 241±248. 23. Wilcox, P.E. (1970) Chymotrypsinogens-chymotrypsins. Methods Enzymol. 19, 64±112. 24. Walsh, K.A. (1970) Trypsinogens and trypsins of various species. Methods Enzymol. 19, 41±63. 25. Sreedhar,L.,Kobayashi,D.Y.,Bunting,T.E.,Hillman,B.I.& Belanger, F.C. (1999) Fungal proteinase expression in the inter- action of th e p lant path ogen Magnaporthe poae with its host. Gene 235, 121±129. 26. Di Pietro, A., Huertas-Gonza  les, M.D., Gutierrez-Corona, J.F., Martõ  nez-Cadena, G ., Me  glecz, E. & Roncero, M.I.G. (2001) Mol. Plant-Microbe Interact. 14, 653±662. 27. Rawlings, N.D. & Barrett, A.J. (1994) Families of serine peptid- ases. Methods Enzymol. 244, 19±61. 28. Oshima, G. (1989) Solid surface-catalyzed inactivation of bovine a-chymotrypsin in dilute solution. Int. J. Biol. Macromol. 11, 43±48. 29. Isogai, T., Fukagawa, M., Kojo, H., Kohsaka, M., Aoki, H. & Imanaka, H. (1991) Cloning and nucleotide sequences of the complementary a nd genomic DNAs for the alkalin e protease from Acremonium chrysogenum. Agric. Biol. Chem. 55, 471±477. 30. Jarai, G. & Buxton, F.P. (1994) Cloning and characterization of thepepDgeneofAspergillus niger which codes for a subtilisin-like protease. Gene 139, 51±57. 31. Frederick, G.D., Rombouts, P. & Buxt on, F.P. (1993) Cloning and characterisation of pepC, a gene encoding a serine protease from Aspergillus niger. Gene 125, 57±64. 32. Katz,M.E.,Rice,R.N.&Cheetham,B.F.(1994)Isolationand characterization of an Aspergillus nidulans gene encoding an alkaline protease. Gene 150, 287±292. 33. Jaton-Ogay, K., Suter, M., Crameri, R., Falchetto, R., Fatih, A. & Monod, M. (1992) Nucleotide seq uence o f a genomic and a cDNA clone encoding an extracellular alkaline protease of Aspergillus fumigatus. FEMS Microbiol. Lett. 71, 163±168. 34. Ramesh, M.V., Sirakova, T. & Kolattukudy, P.E. (1994) Isola- tion, characterization, and cloning of cDNA and the gene for an elastinolytic serine p roteinase f rom Aspe rgillus ¯avus. Infect. Immun. 62, 79±85. 35. Yu,C J.,Chiou,S H.,Lai,W Y.,Chiang,B L.&Chow,L P. (1999) Characterization of a novel allergen, a major IgE-binding protein from Aspergillus ¯avus, as an alkaline serine protease. Biochem. Biophys. Res. Commun. 261, 669±675. 36. Tatsumi, H., Ogawa, Y., Murakami, S., Ishida, Y., Murakami, K., Masaki, A., Kawabe, H., Arimura, H., Nakano, E. & Motai, H. (1989) A full le ngth cDNA clo ne for t he alka line prote ase from Aspergillus oryzae: structural analysis and expression in Sac- charomyces cerevisiae. Mol. Gen. Genet. 219, 33±38. 37. Geremia, R.A., Goldman, G.H., Jacobs, D., Ardiles, W., Vila, S.B., van Montagu, M. & Herrera-Estrella, A. (1993) Molecular characterization of the proteinase-encoding gene, prb1, related to mycoparasitism by Trichoderma harzianum. Mol. Microbiol. 8, 603±613. 38. Joshi, L., St Leger, R.J. & Roberts, D.W. (1997) Isolation of a cDNA enco ding a novel subtilisin-like p rotease (Pr1B) from the 806 A. I. Pekkarinen et al. (Eur. J. Biochem. 269) Ó FEBS 2002 entomopathogenic fungus, Metarhizium anisopliae using dieren- tial display-RT-PCR. Gene 197 , 1±8. 39. Bagga, S., Screen, S.E. & St Leger , R.J. (2000) Isolation and characterization of serine proteases from Metarhizium anisopliae. Submitted to the EMBL/GenB ank/DDBJ databases. 40. Samal, B.B., Karan, B., Boone, T.C., Osslund, T.D., Chen, K.K. & Stabinsky, Y. (1990) Isolation and characterization of the gene encoding a novel, thermostable serine proteinase from the mould Tritirachium album Limb er. Mol. Microbiol. 4, 1789±1792. 41. Gunkel,F.A.&Gassen,H.G.(1989)ProteinaseKfromTriti- rachium album L imber. Characterization of the chromosomal gene and expression of the cDNA in Escherichia coli. Eur. J. Biochem. 179, 185±194. 42. Morita, S., Kuriyama, M., Maejima, K. & Kitano, K. (1994) Cloning and nucleotide sequence of the alkaline protease gene from Fusarium sp. S -19-5 and expression in Saccharomyces cere- visiae. Biosci. Biotechnol. Biochem. 58, 621±626. 43. Huertas-Gonzales, M., Ruiz-Roldan, M., Di Pietro, A. & Ron- cero, M. (1998) A gene encoding an extracellular serine protease o f the vascular wilt pathogen Fusarium oxysporum f. sp. lycopersici is expressed during infection in tomato plants. Submitted to the EMBL/GenBank/DDBJ databases. 44. Davidow, L.S., O'Donnell, M.M., Kaczmarek, F.S., Pereira, D.A., Dezeeuw, J.R. & Franke, A.E. (1987) Cloning and sequencing of the alkaline extracellular protease gene o f Yarrowia lipolytica. J. Bacteriol. 169, 4621±4629. Ó FEBS 2002 An alkaline Fusarium proteinase (Eur. J. Biochem. 269) 807 . Puri®cation and properties of an alkaline proteinase of Fusarium culmorum Anja I. Pekkarinen 1,2 , Berne L. Jones 3 and Marja-Leena Niku-Paavola 2 1 Department of Agronomy, University of Wisconsin-Madison,. espe- cially in Europe [1]. FHB causes severe yield losses in w heat and barley and reduces the crop quality by destroying some of the necessary grain components and by producing mycotoxin s. Fusarium. with an arrow. Fig. 2. SDS/PAGE pattern of the puri®ed proteinase. Lanes 1 and 2: 0.5 and 1.5 lg of CM-HPLC puri®ed enzyme; lane 3: molecular mass standards. Ó FEBS 2002 An alkaline Fusarium proteinase