The Production of Fructooligosaccharides from Inulin or Sucrose Using Inulinase or Fructosyltransferase

Transcription

1 103 [Denpun Kagaku, Vol.36, No.2, p.103 `111 (1989)] The Production of Fructooligosaccharides from Inulin or Sucrose Using Inulinase or Fructosyltransferase from Aspergillus ficuumt Barrie E. NORMAN and Birgitte HOJER-PEDERSEN Novo Industri A/S (Novo Alle, DK-2880 Bagsvaerd, Denmark) A crude "inulinase" system from A. ficuum has been partly characterized and shown to con tain at least 3 major enzyme components: Exo-inulinase [EC ], Endo-inulinase [EC ] and fructosyltransferase [EC ]. The fructosyltransferase has a ph-optimum of about 5.0 and a temperature optimum of 50 Ž. It can be used to produce "Neosugar" (a mixture of 1-kestose, nystose etc.) from sucrose. The endo-inulinase has a ph-optimum of about 5.0 and a temperature optimum of Ž. It can convert inulin to a mixture of fructooligosaccharides, mainly inulotriose and inulotetraose. When used in combination with exo-inulinase, it provides an efficient system for producing fructose from inulin. INTRODUCTION Inulin. Inulin is a linear ƒà-(2 1) linked fructose polymer which is terminated by a single glucose unit (Fig. 1). The degree of polymerization depends upon the source of the inulin, but typical average values of are often quoted. It is found as a food reserve in a number of plants including Jerusalem artichoke(helianthus tuberosus L.) and chicory (Cichorium intybus L.), which are particularly rich sources. The tubers of Jerusalem artichoke are reported to contain up to about 65% inulin, and the carbohydrate production level per hectare is similar to that of sugar beet or potato.l, 2) Several enzymes which are capable of de grading inulin have been described in the liter ature. These may be classified into three main groups2),3): T Exo-inulinases [EC ] (Invertase [EC ] ) U Endo-inulinases [EC ] V Inulin fructotransf erase [EC ] "Proceedings of the Amylase Symposium, 1988" The application of endo-/exoinulinase from A. ficuum (ATCC 16882) for the production of fructose syrup has been described by Zittan.4) More recently, we observed that when this crude inulinase system was incubated with sucrose, there was a build-up of oligosaccharides. This indicated that in addition to the two types of inulinase, there was also fructosyltransf erase activity. CHARACTERIZATION OF A. FIG UUM INULINASE In order to characterize the different enzyme components present, the following purification process was carried out. The crude fermen tation liquor was first centrifuged to remove the mycelia and other debris, and the supernatant concentrated by ultrafiltration. This concentrate was then subjected to a fractionated ammonium sulphate precipitation. The mate rial which precipitated at 40-80% saturation was redissolved in 0.05 M acetate buffer, ph 4.1 and desalted by dialysis against the buffer. The dialyzed fraction was then applied onto a CM-Sepharose CL 6B column which was equilibrated with 0.05 M acetate buffer, ph

2 104 Denpun Kagaku, Vol.36, No.2 (1989) Fig. 2. HPLC analysis of fructooligosaccharides. Fig. 1. Generalized structure of inulin The fractions which were collected from the column were monitored for enzyme activity using sucrose or inulin. One peak, which was not retained on the column, was charac terized as ENDOINULINASE as it could hy drolyze inulin, but had little action on sucrose. Two other peaks exhibiting activity were eluted from the CM-Sepharose CL 6B column using a sodium chloride gradient from M. The peak which was eluted at the lowest ionic strength was characterized as FRUCTOSYL TRANSFERASE, as it had no action on inulin, but formed f ructooligosaccharides when incu bated with sucrose. The highest specific activity in the top fraction was found to be 240 units/ mg protein. The third peak could hydrolyze both sucrose and inulin, and was therefore characterized as EXOINULINASE, with a specific activity of approximately 300INU/mg protein. The endoinulinase activity which was not bound to the CM-Sepharose CL 6B column was purified as follows. The front fractions from the column were pooled, and after the ph and ionic strength had been adjusted, the sample was applied onto a DEAE-Sepharose CL 6B column which was equilibrated with 10 mm phosphate buffer ph 7.0. The endoinu linase, which was retained on the column was eluted using a M sodium chloride gra dient. A single band could be seen when the fractions were examined by SDS-PAGE electro phoresis. The specific activity was found to be 310INU/mg protein. The molecular weights determined by SDS- PAGE electrophoresis were as follows: ENDOINULINASE 67,000 EXOINULINASE 58,000 FRUCTOSYLTRANSFERASE 52,000. FRUCTOOLIGOSACCHARIDE ANALYSIS The fructooligosaccharide samples were an alyzed by liquid chromatography. Glucose, fructose and f ructooligosaccharides up to DP4 can be separated on a cation exchange resin using pure water as the eluent. Prepacked columns containing an 8% crosslinked sulfonic acid type cation exchange resin in the Ca2+ form (e.g., Bio-Rad HPX-87C) have been found suitable (Fig. 2). Although this system is very useful for rou tine analysis it is not possible to fully resolve 1-kestose/6-kestose/inulotetraose, sucrose/inu lotriose, or difructosedianhydride/inulobiose/ glucose. Fig. 3. HPLC analysis of onion extract.

3 The Production of Fructooligosaccharides 105 Fig. 6. The action of fructosyltransferase. Fig. 4. Gel-chromatogram of onion extract. PROPERTIES OF A. FICUUM FRUCTOSYLTRANSFERASE Determination of activity. Fructosyltrans ferase acts on sucrose by cleaving the ƒà-(2 1) linkage, and transferring the f ructosyl moiety to a suitable acceptor molecule (Fig. 6), releas ing glucose.5) The glucose formed can be used as a measure of enzyme activity. One f ructosyltransf erase unit (FTU) is defined as the amount of enzyme which, under standard conditions, generates 1 pmol of glucose Fig. 5. HPLC analysis of DP3 fraction from a: Neosugar, b: onion (for analytical con ditions, see Fig. 3). An amine modified silica column (e.g., Bio- Rad Bio-Sil Amino 5 S) with an acetonitrile/ water eluent gives better resolution, but higher oligosaccharides are eluted only very slowly from the column (Fig. 3). Gel-permeation chromatography can be used to study the higher oligosaccharides. A column, 1.5 ~100 cm packed with polyacrylamide gel (Bio-Gel P 2, -400 mesh) and heated at 65 Ž was employed. The eluent is pure water, and the carbohydrates are monitored with a differ ential refractive index detector (Fig. 4). This column can also be used for preparative work in combination with a fraction collector. Oligosaccharide fractions are collected and con centrated by freeze-drying. Positive identifica tion can then be made by 13C-NMR spectros copy. Using this technique we have shown that the DP3 fraction of Neosugar is 1-kestose, whereas a similar fraction collected from onion contained almost equal amounts of 1-kestose and 6-kestose (Fig. 5). per minute. The activity is determined by incubating a reaction mixture consisting of 1 ml 5% w/v sucrose in 0.1 M phosphate/citrate buffer at ph 6.0 and 1 ml of a suitably diluted enzyme solution for 20 min at 50 Ž. The reac tion is stopped by the addition of base, and the glucose formed determined using Boehr ingers Glucose GOD-PeridR reagent. The effect of temperature and ph on ac tivity. The effect of temperature on fructosyl transf erase activity at ph 5.7 is illustrated in Fig. 7(A). The optimum temperature under these conditions is 50 Ž. Figure 7(B) illustrates the effect of ph on f ructosyltransf erase activity at 50 Ž. Under these conditions the optimum ph is 5.2. The effect of temperature and ph on fructo oligosaccharide formation is illustrated in Fig. 8(A) and Fig. 8(B). An ficuum f ructosyltrans ferase was incubated with 60% sucrose sub strates at different temperatures and ph's. The amount of f ructooligosaccharides formed, determined by HPLC is shown as a function of ph and temperature. From these data it can be seen that the optimum application conditions for the enzyme are 55 Ž, ph 6-7. Fructooligosaccharide production. The

4 106 Denpun Kagaku, Vol.36, No.2 (1989) Fig. 7. The effect of temperature activity. (A) and ph (B) on Am fccuum fructosyltransferase Fig. 8. The effect of temperature (A) and ph (B) on fructooligosaccharide formation. Table 1. The effect of A. ficuum fructosyltransferase dosage on fructooligosaccharide formation. a) Including traces of inulotriose.

5 The Production of Fructooligosaccharides 107 Fig. 9. The effect of temperature (A) and ph (B) on Arn ficuum endo-inulinase activity. Table 2. The effect of ph on carbohydrate composition. Substrate: 20% w/w inulin (Sigma), 60 Ž. Enzyme dosage: 0.5 INU/g DS. Table 3. The effect of temperature on carbohydrate composition. Substrate: 20% w/w inulin (Sigma), ph 4.5. Enzyme dosage: 0.5INU/g DS.

6 108 Denpun Kagaku, Vol.36, No.2 (1989) effect of enzyme dosage on fructooligosaccharide formation is illustrated in Table 1. A 60% w/w sucrose substrate was incubated with different amounts of A. f cuum fructosyltrans ferase at ph 6. 0, 55 Ž and the amount of fructooligosaccharides formed was determined by HPLC. Under these conditions, a max imum of 61-62% was obtained. If the reaction time is prolonged, or the enzyme dosage increased, the f ructooligosaccharide level decreases. Uses of fructooligosaccharides. Fructooligo saccharides have been available in Japan for a number of years. The Meiji Seika company have been offering two forms : Neosugar P and Neosugar G.6) There is also an increasing interest in Europe. Fructooligosaccharides have a pleasant sweet taste and are essentially non-cariogenic. They are not digested by the enzymes of the human gut, but are selectively utilized by Bifidobac terium sp. and other "beneficial" bacteria. Intake of these fructooligosaccharides therefore improves the intestinal flora." If f ructooligosaccharides are incorporated into animal feed, diarrhoea and weight loss can be prevented in young animals after weaning.8) PROPERTIES OF A. FICUUM ENDO-INULINASE Determination of activity. Inulinase activity is determined by incubating a reaction mixture consisting of 1 ml 5% w/v inulin (Sigma) in 0.1 M acetate buffer, ph 4.7 with 1 ml of a suitably diluted enzyme solution for 20 min at 50 Ž. The reaction is stopped by the addition of base, and the reducing sugars formed measured according to the Somogyi-Nelson method, using D-glucose as standard. One inulinase unit (INU) is defined as the amount of enzyme which liberates 1 pmol reducing sugar per minute under the above standard conditions. The effect of temperature and ph on activity. The effect of temperature on endo -inulinase activity at ph 4.7 is illustrated in Fig. 9(A). The optimum temperature under the conditions of the analysis is Ž. Fig. 10. Gel-chromatogram showing the action of Am ficuum endo-inulinase on inulin. The effect of ph on endo-inulinase activity at 50 Ž is illustrated in Fig. 9(B). More than 80% activity is exhibited between ph 4.0 and 6.0. The effect of temperature and ph on hydrolysis. Hydrolysis experiments were carried out under standard conditions at differ ent ph's (Table 2). The rate of hydrolysis at ph 4.5 and 5.0 was similar, but fructose formation is greater at the lower ph, presum ably due to acid catalyzed hydrolysis. The

7 The Production of Fructooligosaccharides 109 Table 4. Tentative identification (HPLC analysis) of fructooligosaccharides isolated from inulin (Dahlia) hy drolyzed with A. frcuum endo-inu linase. Fig. 11. Gel-chromatograms illustrating the composition of a: Neosugar P and b: hydrolyzed inulin. enzyme is sufficiently heat stable to be used at 62.5 Ž (Table 3), but fructose formation is greater than at 55 Ž. In choosing the optimum conditions for an industrial process, it is important to remember that the ƒà-(2 1) f ructoside link is very labile. A ph of >4.5 and a temperature of Ž should be used. Action on inulin. The action pattern of A. ficuum endo-inulinase was studied by incubat ing a 20% w/w inulin (Sigma) substrate at 60 Ž, ph 4.5 with the enzyme, and following the reaction by gel-permeation chromatography. Initially, a whole series of fructooligosaccharides are formed, the major products being DP3 and DP4 (Fig. 10). Reverse phase chromatography analysis indicated that the DP6 and DP7 peaks contained at least two major components. We therefore carried out preparative gel-permeation chro matography as described above and isolated fractions corresponding to DP1-DP7. By ion exchange chromatography the identities of DP1- DP4 have been determined (Table 4). The concentration of 1-kestose and other f ructooligosaccharides with a terminal glucose unit will depend upon the source of the inulin and the average degree of polymerization. The inulin used in these experiments was obtained from Dahlia tubers. If the reaction is prolonged, DP, is hy drolyzed. The rate of hydrolysis of DP6 appears to be very slow. Fructooligosaccharide production. The effect of reaction time on f ructooligosaccharide production is illustrated in Table 5. Using a dosage of 0.5INU/g DS, maximum DP3/DP4 was obtained in hr under the given con ditions. If the enzyme dosage is increased, or the reaction time prolonged, these oligosac charides will gradually be hydrolyzed to fructose. The fructooligosaccharide syrup produced from inulin has an interesting composition because it is low in monosaccharides and Table 5. The effect of reaction time on carbohydrate composition. Substrate: 20% w/w inulin (Sigma), 60 Ž, ph 4.5. Enzyme dosage: 0.5INU/g DS.

8 110 Denpun Kagaku, Vol.36, No.2 (1989) disaccharides. At first sight it is similar to the chromatographically purified Neosugar P (Fig. 11). However, the composition is different, because few of the fructooligosac charides have a terminal glucose group. We have not studied the properties of this product in our laboratories, but we would expect it to be similar to Neosugar.9) CONCLUSION The original aim of our project was to develop an efficient process for the conversion of inulin to fructose using a crude "inulinase" preparation from A. ficuum. However, by identifying and characterizing some of the different enzyme components present, we were also able to study the formation of fructooligo saccharides from inulin and sucrose. Industrial processes for the production of fructooligosaccharides from sucrose are well established. A. ficuum fructosyltransferase has also been found suitable for this application, provided that exo-inulinase and invertase activities are absent. The production of fructooligosaccharides from inulin using purified endoinulinase provides an interesting alternative, as a syrup low in mono and disaccharides can be made without expen sive chromatographic separation. The majority of the oligosaccharides will not be terminated by glucose, but it is expected that inulotriose and inulotetraose etc. will find similar appli cations to the 1-kestose, nystose series. It will not be feasible to produce pure enzyme components on an industrial scale, using the chromatographic techniques described here. In a recent patent, Meiji Seika have described a process for separating endo- and exoinulinase from an A. niger strain using solvent precipitation.10) It should also be possible, using classical mutation techniques to remove the exoinulinase/invertase activities from our strains. Alternatively, the genes coding for one or more of the enzyme components could be transferred separately to a suitable host using recombinant DNA techniques. REFERENCES 1) S. E. FLEMING and I. W. D. GROOT WASSINK: in CRC Critical Reviews in Food Science and Nutrition, T. E. FURIA, end., CRC Press, Florida, p.1-23 (1979). 2) A. FUCHS: Stdrke, 39, (1987). 3) T. UCHIYAMA, K. TANAKA and M. KAWAMURA : Denpun Kagaku, 35, (1988). 4) L. ZITTAN: Starke, 33, (1981). 5) T. ADACHI and H. HIDAKA: US Patent 4, 681, 771 (1987). 6) H. HIDAKA, M. HIRAYAMA and N. SUMI: Agric. Biol. Chem., 52, (1988). 7) H. HIDAKA, T. EIDA, T. TAKIZAWA, T. TOKUNAGA and Y. TASHIRO: Biidobacteria Micro flora, 5, (1986). 8) H. HIDAKA, T. EIDA, K. HASHIMOTO and T. NAKAZAWA: European Patent (1987). 9) T. NAKAMURA and S. NAKATSU: Denpun Kagaku, 35, (1988). 10) T. NAKAMURA, T. TAKIZAWA, Y. KAMO and H. HIDAKA: European Patent Application A2 State J. Robyt You state that inulin would be the preferred substrate over sucrose for the enzymatic production of fructooligosaccharides. Yet, on a commercial scale, sucrose is a much cheaper agricultural commodity than inulin. So, would it not be that sucrose would be the preferred substrate for the commercial production of fructooligosaccharides? Inulin is an attractive raw material for fructooligosaccharide production. If we employ pure endoinulinase for the hydrolysis, then the final product will be low in mono- and disaccharides. If we use sucrose, and fructo syltransferase, then a chromatographic separa tion will be necessary to remove the monoand disaccharides. The carbohydrate yields for Jerusalem arti choke, sugar beet and potato per ha are very similar, but it will be difficult for inulin to compete with cane sugar as a substrate for fructooligosaccharide production.

9 The Production of Fructooligosaccharides 111 The Aspergillus strain we worked with elaborated both exo- and endoinulinase. Exo inulinase will hydrolyze fructooligosaccharides to fructose, so that its presence is detrimental when using either the sucrose/fructosyltrans ferase or the inulin/endoinulinase process. Exoinulinase can be removed by cation exchange chromatography as described above, solvent precipitation (Ref. 10) or possibly classical mutation. Your fructosyltransferase seems to be inter esting as an application enzyme. On this subject, I have two questions to ask you. 1) What sugars are available to accept the fructosyl residue transferred by the enzyme? 2) What action does the enzyme take place in solutions of low concentrations of sucrose? 1) We have not investigated the possibility of using Am ficuum fructosyltransferase to produce heterofructooligosaccharides, using acceptor molecules other than sucrose. 2) Our work has been directed towards the production of Neosugar from sucrose, where we have been using very high substrate con centrations (50-70% w/w). At low solids e.g., 1-2%, transfer to water (i.e., hydrolysis) dominates and Neosugar production is signifi cantly lower. In the US, crystalline fructose is produced from high fructose syrup which has been chromatographically enriched to>90% fructose. It is possible to obtain 90-95% fructose, by hydrolyzing long-chain inulin with the crude A. ficuum inulinase, without chromatographic enrichment. This could be an attractive alter native route to crystalline fructose production, but we have not looked at the economics of the process. (Received January 19, 1989)