Biochemical Characterization of Aspergillus Awamori Exoinulinase: Substrate Binding Characteristics and Regioselectivity of Hydrolysis

Biochimica et Biophysica Acta 1650 (2003) 22–29 www.bba-direct.com

Biochemical characterization of Aspergillus awamori exoinulinase: substrate binding characteristics and regioselectivity of hydrolysis

Anna A. Kulminskayaa, Michael Arandb, Elena V. Eneyskayaa, Dina R. Ivanena, Konstantin A. Shabalina, Sergei M. Shishlyannikova, Andrew N. Savelievc, Olga S. Korneevad, Kirill N. Neustroeva,*

a Division of Molecular and Radiation Biophysics, Petersburg Nuclear Physics Institute, Russian Academy of Sciences, Gatchina, Orlova Rosha, St. Petersburg 188350, Russia b Institute of Pharmacology and Toxicology, University of Wuerzburg, Germany c Biophysics Department, St. Petersburg Technical University, Russia d Voronesh State Technological Academy, Russia Received 4 December 2002; received in revised form 15 April 2003; accepted 13 May 2003

Abstract

1H-NMR analysis was applied to investigate the hydrolytic activity of Aspergillus awamori inulinase. The obtained NMR signals and deduced metabolite pattern revealed that the enzyme cleaves off only fructose from inulin and does not possess transglycosylating activity. Kinetics for the enzyme hydrolysis of inulooligosaccharides with different degree of polymerization (d.p.) were recorded. The enzyme hydrolyzed both h2,1- as well as h2,6-fructosyl linkages in fructooligosaccharides. From the kcat/Km ratios obtained with inulooligosaccharides with d.p. from 2 to 7, we deduce that the catalytic site of the inulinase contains at least five fructosyl-binding sites and can be classified as exo-acting enzyme. Product analysis of inulopentaose and inulohexaose hydrolysis by the Aspergillus inulinase provided no evidence for a possible multiple-attack mode of action, suggesting that the enzyme acts exclusively as an exoinulinase. D 2003 Elsevier B.V. All rights reserved.

Keywords: Exoinulinase; Fructooligosaccharide; Fructosyl-binding site; Multiple attack

1. Introduction sources, represent an effective tool in fructan digestion and nowadays find new industrial applications [5,6]. Mi- Inulin, i.e. 2 ! 1-linked h-D-fructofuranosyl residues crobial inulinases can be divided into two classes according with a terminal glucose unit, and levans consisting predom- to their action on inulin. Exoinulinases (2,1-h-D-fructan inantly of h2 ! 6-fructosyl-fructose linkages (Fig. 1), be- fructohydrolase, E.C. 3.2.1.80) sequentially cleave off ter- long to a group of naturally occurring fructans comprising minal fructose unit from the inulin molecule. Endoinulinase nondigestible fructooligosaccharides, which are commonly (2,1-h-D-fructan-fructanohydrolase, E.C.3.2.1.7) degrades referred to as FOS in the nutrition industry. It was estimated the inulin-type fructans with an endocleavage, producing a that about one-third of the total vegetation on earth consists series of fructooligosaccharides [7,8]. of plants that contain these carbohydrates. Pure D-fructose In our previous paper [9], we reported that a strain of or fructose syrup, which can be easily produced from inulin, Aspergillus awamori var. 2250 produces an extracellular take an important place in human diets today [1,2]. Levan- enzyme during solid-state fermentation converting inulin type FOS have a potential as emulsifying, thickening or and levan to obtain mainly fructose. Based on its DNA stabilizing agents in various food products [3,4]. As such, sequence (EMBL accession number AJ315793), the enzyme fructan-hydrolyzing enzymes, especially from microbial was classified to belong to 32 glycoside hydrolase family Nucleotide Sequence Database (for the classification of glycoside hydrolase families see http://afmb.cnrs-mrs.fr/ * Corresponding author. Tel.: +7-81271-32014; fax: +7-81271-32303. CAZY/db.html) [9]. Despite the numerous reports on isola- E-mail address: [email protected] (K.N. Neustroev). tion and cloning of exoinulinases and exolevanases from

Fig. 1. Structures of h2,6-linked fructan (levan) (A) and h2,1-linked fructan (inulin) (B). fungal and bacterial sources, mainly their enzymatic prop- mass 3850 F 80 Da. 1-Kestose (h-D-Fruf-(2 ! 1)-h-D-Fruf- erties in the hydrolysis of polymeric substrates, inulin and (2 X 1)-a-D-Glcp) was obtained with the aid of the fructo- levan, were described so far [10,11]. Meanwhile, analysis of syltransferase from Aspergillus foetidus according to the the mode of exoinulinase action towards fructooligosacchar- procedure described in Ref. [14]. Inulooligosaccharides with ides with different degree of polymerization (d.p.) is an d.p. from 2 to 4 were produced by digestion of inulin with effective tool for subsite mapping of the active site of this Aspergillus ficuum endoinulinase, which was kindly pro- enzyme. Kinetic studies on the enzymatic hydrolysis of vided by Prof. Chae K.S., Chonbuk National University, oligomeric substrates with different lengths allowed deter- Korea. To produce inulooligosaccharides, 400 mg of inulin mination of affinities and number of subsites in the active was dissolved in 3.5 ml of 20 mM sodium acetate buffer, center of h-glucanases and glucoamylases from various pH 5.0, and incubated with approximately 0.05 U of the sources [12,13]. In the present study, we describe for the endoinulinase per 1 mg of inulin at 37 jC for 60 min. The first time details on the mode of action of the exoinulinase hydrolysate was applied onto a Biogel P2 Super Fine from A. awamori on fructooligosaccharides using 1H-NMR column from Bio-Rad, USA (10 Â 1200 mm, flow rate 9 analysis, thereby estimating the number of fructosyl-binding ml/h) to separate high molecular weight fractions. Specific sites and judging on a possible multiple attack mode of fractions with d.p. from 2 to 4 were rechromatographed on action. a Dextro-Pakk cartridge column (8 Â 10 mm) WATO85650 from Millipore-Waters, USA, using isocratic elution in water (flow rate 1 ml/min) with refractometric detection. To 2. Materials and methods produce inulooligosaccharides with d.p. 5–7, inulin (200 mg) was dissolved with 5% (v/v) of TFA and incubated at 2.1. Substrate production 40 jC for 30 min. After incubation, the mixture was neutralized by mixed Dowex MR-3 (Sigma Chemical Co., Inulin was purchased from Merck (cat. number 1.04733. St. Louis, MO) and concentrated by evaporation. Individual 0010; Germany). Levan from Microbacterium levanifor- fractions of inulooligosaccharides with d.p. 5–7 were iso- mans was a gift from Dr. A. Miasnikov, Danisco Global lated chromatographically as described above. h-2,6-fruc- Innovation, Finland. Average inulin d.p. was calculated by tooligosaccarides were produced by digestion of levan by 1H-NMR from the ratio (n) of the signals H3 (d = 4.3–4.17 levanase [15], which was kindly donated by Dr. Miasnikov, ppm) in fructose and H1 (d = 5.43 ppm) in glucose residues Danisco Global Innovation, Finland. The reaction mixture following the relation d.p. = n + 1. Inulin used in these consisted of levan (100 mg/ml) in 20 mM sodium acetate experiments had d.p. = 23.8 F 0.5 and average molecular buffer, pH 5.0, and was incubated with approximately 24 A.A. Kulminskaya et al. / Biochimica et Biophysica Acta 1650 (2003) 22–29

0.1 unit of levanase per 1 mg of fructan at 37 jC for 60 2.3. Calculation of subsite affinities min. After incubation, specific fractions of levanooligosaccharides with d.p. 2–7 were isolated as described for Kinetic parameters, Michaelis constants (Km) and cata- inulooligosaccharides. Completion of the hydrolysis and lytic rate constants (kcat) of the exoinulinase during the purity of fructooligosaccharides obtained were tested by hydrolysis of h2,1-fructooligosaccharides of d.p. 2–7 were TLC on Kieselgel 60 plates (Merck) with a mobile phase of used to determine subsite affinities of the enzyme. Km and butan-1-ol/acetic acid/water (3:1:1, by vol), HPLC on a kcat were determined by fitting the initial rate of a substrate Dextro-Pac column as described above and by 1H- and 13C- hydrolysis to the Lineweaver–Burk equation. Rates were NMR spectroscopy techniques using data on fructooligo- determined under the assay conditions described above with saccharides given in Refs. [16,17]. The purity of the 10–15 different fructo-oligosaccharide concentrations rang- compounds obtained was greater than 95% as judged by ing from 0.1 Â Km to 4 Â Km. The Michaelis constants were HPLC and NMR analyses. determined by nonlinear regression analysis [21].The equation system developed by Hiromi [22] for exo-acting 2.2. Enzyme and enzyme assays enzymes was used as a base for subsite analysis of the exoinulinase assuming Davies’ nomenclature [23] from À 1 Exoinulinase was isolated from the solid-state culture of to + 1, + 2, etc.: A. awamori var. 2250 according to the procedures described previously [9]. The concentration of pure exoinulinase was lnðkcat=KmÞnþ1 À lnðkcat=KmÞn ¼ÀðAnþ1=RTÞð1Þ estimated by UV absorption spectrophotometry at 280 nm À 1 À 1 using a specific absorption coefficient of 2.2 l g cm . where Km and kcat are the kinetic parameters of the exoinu- Purified exoinulinase contained less than 0.01% of contam- linase-catalyzed hydrolysis of h-2,1-fructooligosacharides inating exo- and endoglycosidase activities. Exoinulinase of d.p. 2–7, and A is the respective subsite affinity. activity was determined by Somogyi–Nelson measurement Assuming that there is only one nonproductive complex, of reducing sugars released from inulin (2.5 mg/ml) in 20 AÀ 1 can be obtained from the Hiromi dependence of 1/kcat mM sodium acetate buffer, pH 4.5, 37 jC, as described in on exp(À An +1/(RT )) [22]. The value for A1 can be esti- Ref. [9]. One unit of enzymatic activity was defined as the mated from Eq. (2) based on preliminary calculated affin- amount of inulinase required to produce 1 Amol of fructose ities A2 –A5: per minute at pH 4.5, 37 jC, using inulin as the substrate at a concentration of 2.5 mg/ml. Alternatively, exoinulinase !55 ! ðKmÞ ¼ : ð2Þ action on fructooligosaccharides was evaluated from analy- n Xn Xnþ1 Ai Ai sis of hydrolytic products by HPLC on a Dextra-Pakk exp À þ exp À þ ... RT RT column using refractometric detection. Products were quan- i¼1 i¼2 tified by integration of the peaks that corresponded to inulooligosaccharides with a d.p. of 2–7 or fructose, using the respective standards. Peak areas were proportional to the 3. 1H-NMR investigation of the exoinulinase hydrolysis amounts of oligosaccharide loaded for standards up to a d.p. of inulin of 7. Activity of the enzyme in the hydrolysis of sucrose was measured following the release of glucose by the glucose 1H-NMR spectra were recorded with an AMX-500 oxidase method [18]. Bruker spectrometer. Inulin, buffer materials and the en- Accompanying exoglycosidase activities in purified zyme were freeze-dried from D2Otwicepriortouse. exoinulinase were measured using appropriate PNP a/h- Reaction was carried out with inulin (5.4–20 mM) in 50 glycosides as substrates according to Ref. [19]. Contami- mM sodium phosphate buffer containing D2O (pH 6.5). nating endo-glycosidase activities were measured by Somo- After the accumulation of the initial spectrum, approx. 10 gyi–Nelson procedure [20] using cellulose, xylan, glucan units of the inulinase was added to complete hydrolysis of from barley, laminarin and galactomannan as substrates. the substrate in 2–20 min. The spectra of the inulin Possible transglycosylating activity was studied in 50 hydrolysis were recorded at 1-min time interval during the mM sodium acetate buffer solutions, pH 4.5 and 5.0, 37 jC, reaction. Concentrations of substrate and products of the and 10 mM sodium phosphate buffer solutions, pH 6.0 and reaction were determined integrating the peaks characteristic 7.0, at 37 jC, using in each case 0.01 U of the enzyme and for each compound: the signal at 5.43 ppm corresponds to up to 100 mM of the respective substrate (sucrose, inulo- H1 proton in glucose residue of inulin and its glucose- triose, inulotetraose and levanotriose). Aliquots of the containing fragments, including sucrose; the signal at 5.23 reaction mixtures were taken after different time intervals ppm corresponds to H1 proton of a free a-glucose; the and the reaction was terminated by freeze-drying. Then, signal at 4.23 ppm corresponds to the H3 proton of the samples were analyzed by TLC in a mobile system of terminal fructofuranose of inulin; the signal at 4.32 ppm acetone/water (87:13, by vol) or using a Dextro-Pakk corresponds to the H3 proton of the fructose residue cartridge column. following the glucose; the signal at 4.29 ppm corresponds A.A. Kulminskaya et al. / Biochimica et Biophysica Acta 1650 (2003) 22–29 25 to H3 protons of inside residues; finally, the signal at 4.28 near completion (20 min, Fig. 2C), are given. The weak ppm corresponds to H3 proton of the terminal fructose. signal with the chemical shift around d = 5.23 ppm corresponds to the anomeric proton of a-glucose. The figure insert shows the time course of the reaction. It demonstrates 4. Results and discussion that the decrease in the concentration of inulin (squares) is closely followed by the increase in the concentration of 4.1. Mode of action of the exoinulinase fructose divided by the d.p. À 1 (circles), indicating that fructose is likely the only major product of the hydrolysis. To investigate the mode of exoinulinase action on inulin Moreover, the absence of signals in the area of 4.3–4.1 ppm as the substrate, 1H-NMR analysis as an effective method that are typical for fructooligosaccharides [17] also confirms was used. Results of the NMR study of hydrolysis of the this suggestion. inulin are shown in Fig. 2 where 1H-NMR spectra of the The possible transglycosylating activity of the exoinuli- reaction mixture at the initial and intermediate stages (0 and nase can be evaluated by the analysis of the product pattern 5 min, Fig. 2A,B, respectively), and after the hydrolysis is of inulin hydrolysis. Transglycosylation is most likely to

Fig. 2. 1H-HMR spectra of the inulin hydrolysis at the initial (A) and intermediate (B) stages of the reaction process and after the hydrolysis is near completion (C). Insert: The time course of inulin hydrolysis catalyzed by exoinulinase: n—inulin degradation; .—formation of fructose during the reaction. The concentrations of inulin and free glucose in the course of the reaction were calculated by integrating the NMR peaks at 4.23 and 5.23 ppm, respectively. The fructose concentration was calculated by summing the concentrations of the pyranose and the furanose forms obtained from integrating the peaks at 3.60 and 3.57 ppm, respectively. 26 A.A. Kulminskaya et al. / Biochimica et Biophysica Acta 1650 (2003) 22–29 occur at high substrate concentrations towards the end of the mediate and final products. The sensitivity of NMR analysis reaction, because under these conditions the concentration gives an indisputable advantage over TLC, HPTLC and of potential acceptor molecules is particularly high. The HPLC methods [7,8] and, moreover, allows to exclude the indicative products of such reaction would be short fructo- presence of more than 1% of contaminating endoinulinase oligosaccharides, the presence of which could be detected activity in the hydrolysis of inulin. by the recording of differential NMR spectra, which are Additionally, transglycosylating activity of the exoinuli- differences between spectrum of the reaction mixture at time nase was investigated in the hydrolysis of sucrose as a t and those at time t0, taking into consideration the degree of substrate in a concentration range from 40 up to 200 mM substrate conversion. At an inulin concentration of 20 mM and at several pH values from 3.5 to 7.0. Despite of 91% and a substrate conversion of 80%, no signals at 4.3–4.1 similarity with A. foetidus fructosyltransferase according to ppm that would be characteristic for fructobioside and their amino acid sequences [9], the A. awamori exoinulinase fructo-oligosaccharides with higher d.p. [17] were observed. was found not to produce 1-kestose, the product of fructo- Instead, the differential 1H-NMR spectra of the reaction syltransferase reaction expected to be most prevalent in the mixture at the intermediate stage of the reaction (1 min, considered substrate concentration range [14], as evidenced Fig. 3A), and after the hydrolysis is near completion (20 by HPLC analysis on a Dextra-Pac column and TLC min, Fig. 3B) closely resemble the spectrum of fructose methods. Moreover, inulotriose, inulotetraose and levano- (Fig. 3C). Therefore, only fructose and glucose are produced triose investigated as substrates in the transglycosylation as the final results of this reaction. Noteworthy, the use of reaction at the concentrations up to 50 mM and pH values 1H-NMR analysis of the reaction mixture allows detection from 3.5 to 7.0 did not yield the predicted corresponding of down to 40–60 Amol of an oligosaccharide formed products, namely inulotetraose, inulopentaose and levanote- during the reaction, which corresponds to less than 0.3% traose. Difructosides were also not formed in the exoinuli- of initial inulin concentration. nase reactions when the enzyme was incubated with The approach introduced here allowed us to employ 1H- fructose in the concentration range of 50–200 mM. In all NMR analysis to study the kinetics of inulin and inulooli- these analyses, the level of substrate hydrolysis was at least gosaccharide hydrolysis carried out by the exoinulinase. The 20–25%. Therefore, based on these results, we have no method can be applied as a continuous assay without the evidence that the exoinulinase from A. awamori 2250 need for termination of the reaction and isolation of inter- possesses transglycosylating activity.

Fig. 3. The differential 1H-NMR spectra of the reaction mixture of the inulin hydrolysis at the intermediate stage (A) and after the hydrolysis is near completion (B). The spectrum (C) is the 1H-NMR spectrum of fructose. A.A. Kulminskaya et al. / Biochimica et Biophysica Acta 1650 (2003) 22–29 27

Fig. 4. (A) Time-dependent HPLC analysis of the exoinulinase-catalyzed hydrolysis of inulopentaose (o); formation of inulotetraose (5); formation of fructose (E). (B) Time-dependent HPLC analysis of the exoinulinase-catalyzed hydrolysis of inulohexaose (n); formation of inulopentaose (o); and formation of fructose (E).

4.2. Multiplicity of the exoinulinase hydrolytic action occurs, k3 becomes greater than zero. Both pathways can exist in parallel and be of similar relevance if constants k2 As it was reported earlier for starch-degrading enzymes, and k3 are in the same range. The extent of multiplicity of endo-acting a-amylases [24,25], exo-acting h-amylases attack by the exoinulinase can be calculated using Eq. (3) and glucoamylase [26–28], hydrolysis of polymeric sub- developed earlier [29]: strate can occur with multiplicity of attack. Here, the primary product of the enzymic action serves as the aN H ¼ ð3Þ secondary substrate for the next act of the hydrolysis. ½SN Àln t After the initial catalytic action, the product can recover ½SN0 itself on the enzyme surface giving a new productive enzyme–substrate complex. This can be repeated several where H is the extent of multiplicity of attack, a is the extent times before the final dissociation of the enzyme–sub- of hydrolysis of the substrate consisting of n fructose units strate complex. Besides polymeric substrates, starch and at the time t, [SN0] and [SNt] are concentrations of the glycogen [24,25], shorter ones, maltooligosaccharides substrate at the initial point t0 and time t. Hydrolysis of [27,28], were used to investigate the phenomenon of inulooligosaccharides with d.p. 4–7 and levanopentaose multiple-attack mode of action. The concept of multiple- was analyzed by HPLC of the reaction products after attack mode of action was studied with the exoinulinase different time intervals (data are given in Fig. 4). The value from A. awamori using fructooligosaccharides as sub- for the multiplicity extent H, which was calculated accord- strates. ing to Eq. (3), was about 1 for all tested compounds (Table Taking the possible multiplicity of attack into account, 1). Therefore, multiple-attack mechanism of the exoinuli- the reaction of fructooligosaccharide hydrolysis effected by nase towards inulooligosaccharides with d.p. 4–6 was not the exoinulinase may occur in two ways: observed.

Table 1 Multiplicity extent of the hydrolysis of fructooligosaccharides by the exoinulinase where (Fru)n is a fructooligosaccharide with d.p. = n, E is the Substrate Multiplicity extent, H enzyme, k+1and kÀ 1 are dissociation constants, (Fru)n À 1 is Inulotetraose 1.05 F 0.04 a fructooligosaccharide with d.p. = n À 1, and Fru is fructose. Inulopentaose 1.03 F 0.04 Such minimal kinetic scheme can be discussed in case of Inulohexaose 1.08 F 0.02 the absence of transglycosylation as was proven above for Inuloheptaose 1.07 F 0.02 F the exoinulinase described here. If multiplicity of attack Levanopentaose 1.15 0.12 28 A.A. Kulminskaya et al. / Biochimica et Biophysica Acta 1650 (2003) 22–29

4.3. Hydrolytic activity and subsite structure of the active Table 3 center Kinetic parameters of the hydrolysis of inulooligosaccharides by the exoinulinase À 1 Previously, we determined kinetic parameters of the Substrate Km [mM] kcat [s ] kcat/Km [s/mM] Aspergillus exoinulinase in the hydrolysis of inulin, levan, Inulobiose 3.72 F 0.19 5060 F 250 1360 F 250 Inulotriose 1.63 F 0.07 6210 F 279 3810 F 279 stachyose and raffinose [9]. The exoinulinase hydrolyzed F F F F À 1 Inulotetraose 0.94 0.04 6240 275 6640 275 sucrose with Km = 40 mM and kcat = 1150 10 s . The Inulopentaose 1.08 F 0.02 7670 F 153 7100 F 153 specificity of the inulinase action was also investigated in Inulohexaose 1.02 F 0.01 6900 F 69 6760 F 69 the hydrolysis of h2,6-fructooligosaccharides (levanooligo- Inuloheptaose 1.10 F 0.01 6670 F 33 6060 F 33 saccharides) with d.p. 2–6. Kinetics of levanooligosaccharide hydrolysis of the exoinulinase followed Michaelis– As seen from Tables 2 and 3, the exoinulinase hydrolyzes Menten kinetics. The values for k and K obtained with cat m levanooligosaccharides and inulooligosaccharides with com- these substrates are given in the Table 2. Rates were parable rate constants suggesting that the enzyme is built to determined under the assay conditions described above with act similarly well on both h-2,6- as well as h-2,1-fructosyl 12–16 different levanooligosaccharide concentrations rang- bonds. However, the rate of the hydrolysis for levan is much ing from 0.1 Â K to 4 Â K . The initial rates of hydrolysis m m lower than that for inulin [9]. This may be explained by of h2,1-fructooligosaccharides (inulooligosaccharides) with differences of space structures of levan and inulin [17]. d.p. 2–7 were measured as a function of substrate concen- Another possible explanation might be interactions between tration. Plots of the hydrolytic rates versus oligosaccharide the enzyme surface and polysaccharides (e.g. inulin and concentrations were linear at the initial stages of the reaction levan) as it is described for the hydrolysis of laminarin by and, therefore, obeyed Michaelis–Menten kinetics ranging Candida albicans h-1,3-exoglucanase [33].Itshouldbe from at least 0.1 Â K to 4 Â K (Table 3). Kinetic analysis m m noted that the K for the hydrolysis of inulin by the of the inulinase indicates that hydrolytic reactions of inu- m exoinulinase has the value of 0.05 mM [9], which is lower looligosaccharides with lower d.p. occur with a lower rate. than the relative constants for the hydrolytic reactions of The catalytic efficiency factor, k /K , also decreases cat m inulooligosaccharides with d.p. 5–7 (Table 2). On the other depending on the d.p. of the substrates in the same way. hand, the application of Hiromi approaches showed the The preference of the exoinulinase for h2,1-fructooligosac- presence of at least five binding sites. Similar dependence charides of increasing chain length suggested that it has an was reported earlier for glucoamylase from fungal sources extended substrate-binding region. This was confirmed by [12,22] in the hydrolysis of maltooligosaccharides and calculation of binding energies based on kinetic results in accordance with Hiromi et al. [12,22] and as presented in Fig. 5. The results indicate that the exoinulinase has an extended active center consisting of at least five fructosyl- binding sites. The highest values for binding energies were at subsites À 1 and + 1. Similar high affinities for subsites À 1 and + 1 were reported for retaining barley h-D-glucosidase/(1,4)-h-D-glucan exohydrolase, which belongs to glycoside hydrolase family 3 [30] and inverting exo-h- 1,3-glucanase from Trichoderma viride [13]; retaining a- glucosidase of family 13 and h-glucosidase of family 1 [31]. Unfortunately, these data do not allow us to speculate about the relationships between retention/inversion mechanism of the A. awamori inulinase action and its binding energies in the particular sites of the active center. However, kinetic and subsite structure data confirm our earlier suggestion that the enzyme can be classified as exo-acting inulinase and not h- fructofuranosidase [32].

Table 2 Kinetic parameters of the hydrolysis of levanooligosaccharides by the exoinulinase À 1 Substrate Km [mM] kcat [s ] F F Levanotriose 0.21 0.01 1840 92 Fig. 5. (A) The subsite structure of the exoinulinase. The subsite affinities F F Levanotetraose 2.74 0.14 1990 89 A are represented by the histogram. (B) Schematic model of the F F i Levanopentaose 10.2 0.31 575 18 exoinulinase-binding site: ( ) nonreducing terminus glucosyl residue; F F . Levanohexaose 1.01 0.02 570 6 (o) glucosyl residues; vertical arrow, catalytic amino acids. A.A. Kulminskaya et al. / Biochimica et Biophysica Acta 1650 (2003) 22–29 29 starch. For both enzymes, the presence of additional non- [13] A.A. Kulminskaya, K.K. Thomsen, K.A. Shabalin, I.A. Sidorenko, catalytic binding sites for polymeric substrates, glycogen and E.V. Eneyskaya, A.N. Savel’ev, K.N. Neustroev, Isolation, enzymatic properties, and mode of action of an exo-1,3-beta-glucanase from starch was well documented [34]. Possibly, the exoinulinase Trichoderma viride, Eur. J. Biochem. 268 (2001) 6123–6131. considered here also has additional binding site for inulin [14] J. Rehm, L. Willmitzer, A.G. Heyer, Production of 1-kestose in trans- located near the active center. Structure investigations of the genic yeast expressing a fructosyltransferase from Aspergillus foeti- enzyme, which are in progress now, will allow to answer this dus, J. Bacteriol. 180 (1998) 1305–1310. question. [15] A.N. Miasnikov, Characterization of a novel endo-levanase and its gene from Bacillus sp. L7, FEMS Microbiol. Lett. 154 (1997) 23–28. [16] H. Hammer, S. Morgenlie, Classification of grass fructans by 13C NMR spectroscopy, Acta Chem. Scand. 44 (1990) 158–160. Acknowledgements [17] S. Baumgartner, T.G. Dax, W. Praznik, H. Falk, Characterization of the high-molecular weight fructan isolated from garlic (Allium sati- The present work was supported in part by grants no. 03- vum L), Carbohydr. Res. 328 (2000) 177–183. 04-48756 and 03-04-49741 from the Russian Foundation [18] M.L. Bruss, A.L. Black, Enzymatic microdetermination of glycogen, Anal. Biochem. 84 (1978) 309–312. for Basic Research, by the grant of Program for Basic [19] T. Akino, N. Nakamara, K. Horikoshi, Characterization of h-manno- Research in Physico-Chemical Biology from the Presidium sidase of an alkalophilic Bacillus sp, Agric. Biol. Chem. 52 (1988) of Russian Academy of Sciences, and a grant from the 6th 1459–1464. Competition–Examination for Young Scientists of RAS on [20] M. Somogyi, Notes on sugar determination, J. Biol. Chem. 195 Fundamental and Applied Researches. (1952) 19–23. [21] R.T. Leatherbarrow, Enzifitter. A Non-Linear Regression Data Anal- ysis Program for IBM-PC, Elsevier Biosoft, Cambridge, 1987. [22] K. Hiromi, Interpretation of dependency of rate parameters on the References degree of polymerization of substrate in enzyme-catalyzed reactions. Evaluation of subsite affinities of exo-enzyme, Biochem. Biophys. [1] E. Menne, N. Guggenbuhl, M. Roberfroid, Fn-type chicory inulin Res. Commun. 40 (1970) 1–6. hydrolysate has a prebiotic effect in humans, J. Nutr. 130 (2000) [23] G.J. Davies, K.S. Wilson, B. Henrissat, Nomenclature for sugar- 1197–1199. binding subsites in glycoside hydrolases, Biochem. J. 321 (1997) [2] I.G. Carabin, W.G. Flamm, Evaluation of safety of inulin and oli- 557–559. gofructose as dietary fiber, Regul. Toxicol. Pharmacol. 30 (1999) [24] J.F. Robyt, D. French, Multiple attack and polarity of action of por- 268–282. cine pancreatic alpha-amylase, Arch. Biochem. Biophys. 138 (1970) [3] H. Hidaka, T. Eida, Y. Saitoh, Industrial production of fructo-oligo- 662–670. saccharides and its application for human and animals, Nippon No- [25] H. Kondo, H. Nakatani, K. Hiromi, R. Matsuno, Multiple attack in geikagaku Kaishi (in Japanese) 61 (1987) 915–923. porcine pancreatic a-amylase-catalyzed hydrolysis of amylose studied [4] I. Vijn, S. Smeekens, Fructans: more than a reserve carbohydrate? with a fluorescence probe, J. Biochem. 84 (1978) 403–417. Plant Physiol. 120 (1999) 351–359. [26] J.M. Bailey, W.J. Whelan, The mechanism of carbohydrase action: 3. [5] N. Kaur, M. Kaur, A.K. Gupta, R. Singh, Properties of h-fructosidase The action pattern of h-amylase, Biochem. J. 67 (1957) 540–547. (invertase and inulinase) of Fusarium oxysporum growth on an [27] T. Suganuma, M. Ohnishi, K. Hiromi, Y. Morita, Evaluation of sub- aqueous extract of Cichorium intybus roots, J. Chem. Technol. Bio- site affinities of soybean h-amylase by product analysis, Agric. Biol. technol. 53 (1992) 279–284. Chem. 44 (1980) 1111 – 1117. [6] A. Pandey, C.R. Soccol, P. Selvakumar, V.T. Soccol, N. Krieger, J.D. [28] H. Kondo, H. Nakatani, R. Matsuno, K. Hiromi, Product distribution Fontana, Recent developments in microbial inulinases. Its production, in amylase-catalyzed hydrolysis of amylose. Comparison of experi- properties, and industrial applications, Appl. Biochem. Biotechnol. 81 mental results with theoretical predictions, J. Biochem. 87 (1980) (1999) 35–52. 1053–1070. [7] S. Baumgartner, W. Praznik, Purification of exo- and endoinulinase [29] J.M. Bailey, D. French, The significance of multiple reactions in from crude inulinase extract for the analysis of fructans, Int. J. Biol. enzyme–polymer systems, J. Biol. Chem. 226 (1957) 1–14. Macromol. 17 (1995) 247–250. [30] M. Hrmova, E.A. MacGregor, P. Biely, R.J. Stewart, G.B. Fincher, [8] R.A.B. Pessoni, R.C.L. Figueredo-Ribeiro, M.R. Braga, Extracellular Substrate binding and catalytic mechanism of a barley h-D-glu- inulinase from Penicillium janczemskii, fungus isolated from the rhizo- cosidase/(1,4)-h-D-glucan exohydrolase, J. Biol. Chem. 273 (1998) sphere of Vernonia herbacea, J. Appl. Microbiol. 87 (1999) 141–147. 11134–11143. [9]M.Arand,A.M.Golubev,J.R.BrandaoNeto,I.Polikarpov,R. [31] T. Yazaki, M. Ohnishi, S. Rokushika, G. Okada, Subsite structure of Wattiez, O.S. Korneeva, E.V. Eneyskaya, A.A. Kulminskaya, K.A. the h-glucosidase from Aspergillus niger, evaluated by steady-state Shabalin, S.M. Shishliannikov, O.V. Chepurnaya, K.N. Neustroev, kinetics with cello-oligosaccharides as substrates, Carbohydr. Res. Purification, characterisation, gene cloning, and preliminary X-ray 298 (1997) 51–57. data of the exo-inulinase from Aspergillus awamori, Biochem. J. [32] M. Warchol, S. Perrin, J.P. Grill, F. Schneider, Characterization of a 362 (2002) 131–135. purified beta-fructofuranosidase from Bifidobacterium infantis ATCC [10] K. Yanai, A. Nakane, A. Kawate, M. Hirayama, Molecular cloning 15697, Lett. Appl. Microbiol. 35 (2002) 462–467. and characterization of the fructooligosaccharide-producing beta-fruc- [33] H.J. Subbs, D.J. Brasch, G.W. Emerson, P.A. Sullivan, Hydrolase and tofuranosidase gene from Aspergillus niger ATCC 20611, Biosci. transferase activities of the h-1,3-exoglucanase of Candida albicans, Biotechnol. Biochem. 65 (2001) 766–773. Eur. J. Biochem. 263 (1999) 889–895. [11] E. Jung Kang, S.O. Lee, J.D. Lee, T.H. Lee, Purification and charac- [34] B.W. Siguskjold, T. Christensen, N. Payre, S. Cottaz, H. Driguez, terization of a levanbiose-producing levanase from Pseudomonas sp. B. Svensson, Thermodynamics of binding of heterobidentate ligans No. 43, Biotechnol. Appl. Biochem. 29 (1999) 263–268. consisting of spacer-connected acarbose and h-cyclodextrin to the [12] K. Hiromi, Y. Nitta, C. Numata, S. Ono, Subsite affinities of glucoa- catalytic and starch-binding domains of glucoamylase from Asper- mylase: examination of the validity of the subsite theory, Biochim. gillus niger shows that the catalytic and starch-binding sites are in Biophys. Acta 302 (1973) 362–375. close proximity in space, Biochemistry 37 (1998) 10446–10452.