MXPA00010936A - Functional sugar polymers from inexpensive sugar sources and apparatus for preparing same - Google Patents

Abstract

A process for preparing functional sugar polymers comprising transferring a monosaccharide or oligosaccharide to an acceptor, removing by-products, separating polymers which have not achieved the desired chain length and recycling these underdeveloped polymers, and an apparatus for producing same.

Description

FUNCTIONAL SUGAR POLYMERS FROM LOW COSTLY SUGAR SOURCES. AND APPARATUS TO PREPARE THEMSELVES FIELD OF THE INVENTION This invention relates to a process for preparing functional sugar polymers and to an apparatus useful for the synthesis of said sugar polymers.

BACKGROUND OF THE INVENTION Sugar polymers produced by microorganisms and plants, mainly straight and branched glucans and fructans, have a long history of applicability of food products such as gums, fillers, bulk sources and other similar use. In addition to its uses in foods, some of these sugars have been reported for a long time that impart health benefits. See, McAuliffe, J., Hindsgaul, O .; "Carbohydrate Drugs - An Ongoing Challenge", Chemistry and Industry, March 3, 1997. More recently, an increasing number of these polymers (whether of plant origin, of microorganisms or of mammals), especially those that have unique branching or that contain other sugars such as fructose, galactose, sugar amines or sialic acid has been cited for imparting a variety of health benefits (see, for example, US Patent 5,514,660). Although a large amount of data is known related to the specific use and functionality of many of the significant natural sugar polymers, the complexity of these compounds and the resulting difficulty and high cost of preparing structures that may exist in nature has limited the knowledge of the general structure / function or relationships with health benefits. The sugar polymers are generally isolated from plants or microorganisms by extraction, and purified by a variety of techniques. Significant work has been carried out to increase the availability and functionality and to decrease the cost of these polymers. This work has focused on the following areas: Better sources of plant sugars have been found through the discovery of new species, selective breeding and genetic manipulation. Improved processes have been developed to extract the sugar polymers. The sugar polymers have been separated into fractions of individual chain length and / or "refined" to a more limited specification. Simple chemical modifications of some plant sugars have been developed.

Work is taking place and accelerating in all these areas. An important area with a significant ongoing effort is the genetic modification of organisms to improve the production of existing sugars or to obtain new sugars. Methods for obtaining sugar polymers without using plants are also known. Both chemical and enzymatic syntheses have been developed (see, for example, Whitesdie, G. M., et al., "Enzyme-Catalyzed Synthesis of Carbohydrates", Tetrahedron, 45 (17), pp. 5365-5422, 1989). Chemical synthesis, although possible, is hindered by three factors. First, each of the monomer units has multiple reaction sites. In normal chemical synthesis, multiple reactive sites always add complexity, but with carbohydrates this complexity is extremely significant. A polymer with only four "monomer" units of sugar can be assembled in -270,000 possible ways. This complexity can be overcome by blocking and selective unblocking of reactive sites, but only at the cost of low performance and high processing costs. Second, the links between sugars are very fragile under typical processing conditions, again limiting synthetic selections and decreasing yields. Finally, many chemical reactions are not stereo-specific, creating a problem of separation and reductions in additional yields. The real results are high cost and a general lack of availability. The chemical synthesis is ^ a ^ aMk kAriMHIIia? lta clearly suitable for the production of sugar polymers only when the product has a very high value. Procedures that use enzymes as selective catalysts offer more promise for the synthetic production of sugar polymers. Recent advances in identification and production of these enzymes, as well as systems in which multiple step coupling reactions are achieved, have reduced the cost of producing sugar polymers by many orders of magnitude. These techniques are applicable to both "natural" polymers and novel polymers based on the coupling of their "natural" carbohydrate units in new forms. Enzymes have been identified that allow the copulation of somewhat expensive starting materials such as sucrose, making this type of process potentially very economical. Enzyme-based production provides both coupling and stereoselectivity, which minimize purification costs (see U.S. Patent No. 5,288,637, Roth, S., "Apparatus for the Synthesis of Saccharide Compositions"; US patent A. # 5,180,674, Roth, S., "Saccharide Compositions, Methods, and Apparatus for Their Synthesis"; PCT / US92 / 10891, Roth, S., "A Method for Obtaining Glycosyltransferases for Biosynthesis of Oligosaccharides and Genes for Encoding Them"; PCT / US94 / 07807, Roth, S., "A Method for Synthesizing Saccharide Compositions"). Although these techniques allow great control in the coupling reactions, the control of the chain length or the distribution of branching in polymers, have not been solved.

An example of a sugar polymer is inulin. This polymer consists of fructose units, linearly linked in a β- [2- »1] form to a terminal -D-glucopyranosyl unit. In many plants, inulin is a main energy storage carbohydrate. It is highly regarded as a naturally low calorie food product, which can be used to provide bulk to high intensity sweeteners, as a source of soluble fiber, as a fat replacement and as a promoter of bifidobacteria in the digestive tract. Inulin is commonly produced by the extraction of plants, notably chicory and aguaturma. The material extracted from these plants contains polymers with from 1 to more than 60 units of fructose bound to a terminal glucose unit. Polymer distributions vary with the plant source, the location of the plant and the time of harvest. Although the raw extract has application, the product with 3 to 10 units of fructose provides the most "sucrose-like" yield. These polymers are ideal bulk agents and sources of fiber since they have low caloric density, are soft and have many of the functional attributes of sucrose. The shorter polymers, although functional, provide a disproportionate increase in calories and sweetness. The longer polymers are less water soluble and have adequate fat replacement properties but decreased sucrose replacement functionality. Various approaches have been used to provide improved functionality and a consistent inulin product over a growing season. The first approach involves using a plant source containing relatively long chain nullin and using selective enzymes to reduce the chain size in a controlled manner. EP 440074 relates to the selective hydrolysis of long chain nulin. This allows some control over temporary variations at the cost of adding a procedural step, and does not help eliminate short-chain and high-calorie products. Another approach is to fractionate raw inulin into cuts with different functionality. Although this solves the problems of functionality and temporal variation, it does so at the expense of creating by-products. This approach also provides for the removal of short chain polymers. Since it only includes fractioning, it does not add "foreign" substances to the product. The two approaches can be combined in several ways to optimize the recovery of the fractional product. In both of these cases, the agricultural source grows specifically for the inulin product with all the risks associated with any agricultural program. These approaches also present the problem of the disposal of a large mass of unused plant material. Finally, except for very high use sugars such as sucrose, it is difficult to generate any economies of scale when a plant is grown for a specific product. Another method for producing fructooligosaccharides of nulin is to accumulate the fructose polymer from a source of a little expensive fructose, such as sucrose, using coupling enzymes (see Jong Won Yun, "Fructooligosaccharides-Occurrence, Preparation, and Application ", Enzyme and Microbial Technology 19: 107-117, 1996, and U.S. Patent No. 4,681,771). This approach suffers from the inhibition of the enzymatic reaction by a by-product of the coupling, glucose, which limits the conversion of the intake of sucrose and the length of the chains produced. These inulin fructooligosaccharide compositions commonly have an average chain length of less than 5 fructose units. The reactions catalyzed by enzymes in biological systems are regulated by the inhibitory effects of the enzyme by means of the accumulation of by-products. This is used for the benefit of the organism to avoid the accumulation of unused reaction products. When the enzyme is used as part of a cascade of reactions, the product of an enzyme is removed by the actions of the next enzyme in the cascade. When enzymes are used in man-made reactors, alternative methods should be employed. One approach used to remove the inhibition is to purify the glucose that inhibits the forward reaction by converting it into another by-product. Although this may improve performance, it does little to improve the chain length selectivity. Other techniques that remove glucose, such as permeation through a selective membrane are also possible, but again only result in improved performance and can not produce the desired polymers. Another refinement, commonly used in multi-step enzyme reactors, is to use a series of reactors with by-product removal between reactors and, if desirable, the addition of feedstocks.

For a significant increase in system complexity, this type of system can result in improved performance. Although it also allows some control over the average polymer size, it offers very little control over the distribution of the chain length. The distribution of the chain length can not be predicted using the Míchaelis-Menten kinetics (F. Ouarne and A. Guibert, "Fructooligosaccharides Enzymic Synthesis from Sucrose." Zuckerindustrie (Berlin) (1995), 120, pp 793-798). The chain lengthening of the fructooligosaccharide chains by the sequential transfer of fructose from sucrose by the fructosyltransferase of A. niger, is reported by M.

Hirayama and H. Hidaka (Chapter 9, "Production and Utilization of Microbial Fructans "in Science and Technology of Fructans, CRC Press, 1993): 2 GF? G + GF2 GF + GF2? G + GF3 GF + GF3? G + GF4, etc. As used above, the term GF refers to sucrose , and the term GFn refers to a linked β-2,1 fructose oligosaccharide, which is linked via its reducing end to an alpha-D-glucopyranosyl moiety.A different model for the processing of fructooligosaccharide chains is proposes in KJ Duan, JS Chen and DC Cheu., "Kinetic Studies and a Mathematical Model for Enzymatic Production of Fructooligosaccharides from Sucrosee; Fructooligosaccharide Sweetener Production Using Aspergillus japonicus beta-D-fructofuranosidase. "Enzyme Microb. Technol., 4, 334-339 (1994).) Two chains of similar size react with the enzyme to give a shorter and longer chain, and a step of hydrolysis that released fructose from the tetrasaccharide GF3: 2 GFn? GFn-i + GFn +? for n = 1 to 3 GF3? F + GF2 F. Ouarne and A. Guibert ("Fructooligosaccharides: Enzymic Synthesis from Sucrose". Zuckerindustrie (Berlin), 120, 793-798 (1995)) describe a mathematical model based on the previous reaction pattern, and add the fructose transfer of the trisaccharide GF2 to glucose to produce two moles of sucrose and the hydrolysis reaction of sucrose up to its monosaccharide components: G + GF2? 2 GF GF? F + G The model described in Zuckerindustrie showed good agreement with experimental values observed in the production of the different components in the reaction mixture, but did not prevent the wing chain by sequential transfer of fructose to sucrose as a mechanism of possible contribution. The model described in Zuckerindustrie is based on well-documented enzyme kinetics, and predicts the reported yield for the production of fructooligosaccharides. A quick analysis of the model will reveal why these chain length distributions are obtained. The speed of formation of GFn +? it depends on the concentration of GFn. In both plug flow and intermittent flow reaction systems, this relationship establishes the type of distributions that will be observed. Taking the reaction further through increased concentrations, catalyst levels or residence time does not affect the relative distribution, only its position. In the case of fructose polymers, this means that increasing the degree of the reaction can shift the product distribution mainly from GF2 with increasingly smaller amounts of GF3I GF4, etc., to almost GF3 progressively with smaller amounts of chains longer pushing the reaction harder. The plugged-type reactor configurations or multiple back-mixing reactors used in series increase the amounts of higher polymers, but again they can not have an impact on the basic ratio. If the polymerization reaction is considered essentially irreversible, the average length is clearly limited by these relationships. When the supply of GF is exhausted, no additional growth will take place. If more sucrose is added, the reaction will progress to GF2. Assuming that the problem of glucose inhibition can be overcome, this can be improved by adding sucrose between the stages of a series of flow reactors by sealing or backmixing. Although this is possible, it would certainly add significant costs and complexity to the system to achieve any significant improvements. When reversible polymerization is possible, it might be possible to take the reaction further, but there would be a new set of limitations. In this case, a level of equilibrium of products must be achieved, assuming that a sufficient type of residence is given. Adding more sucrose would increase the total material produced, but not the product ratios. Also, other reactor configurations would not change the final reaction mixture either.

BRIEF DESCRIPTION OF THE INVENTION The invention relates to a process for preparing a straight or branched chain sugar polymer of a desired chain length, preferably four or more units, which comprises: a) transferring a monosaccharide or oligosaccharide residue from a carbohydrate donor to an acceptor by means of selective catalytic transfer, preferably by reacting with an enzyme; b) optionally, remove byproducts from step a) that could inhibit yield or selectivity; c) separating polymers that have not reached the desired chain length; and d) recirculating the polymers that have not reached the desired chain length. In another aspect, the invention relates to an apparatus for producing straight or branched sugar polymers of a desired chain length, which comprises: a) a reactor in which a monosaccharide or oligosaccharide residue is transferred by selective catalytic transfer of a carbohydrate donor to an acceptor; b) means for removing byproducts of the reaction of step a); c) means for separating polymers that have not reached the desired chain length; and d) means for recirculating the polymer chains that have not reached the desired chain length back to the reactor. It has been found that combining means for separating underdeveloped polymers and by-products, for example, by selective membrane such as nano-filters, with an enzyme-based reactor, not only increases the yield, but also the resulting product is designed to a highly specific scale. The use of membranes in conjunction with the enzyme system provides two key benefits. First, a membrane is used to remove the byproduct, glucose, without converting it into another entity. This allows the reaction to proceed to high yields without adding reaction steps to the process. A second membrane is added to remove chains of the desired length of the system, while shorter (underdeveloped) chains are recirculated back for further growth. This novel combination results in a system that outperforms conventional approaches by a significant margin. Again, returning to the kinetic model, the reasons why this novel combination offers superior performance are clear. When a separation system is added as described herein, the entire dynamic is changed. For example, the GF3's produced are no longer removed from the system, but are the source to produce GF4 which are taken as a product. The concentrations in the reactor necessary to produce longer chains can be present without having an impact on the distribution of the product. The fundamental relationships in the reactor remain, but do not affect the product mix. This is true regardless of whether the reaction is considered reversible or irreversible. Through judicious membrane choice, any average polymer chain length distribution can be achieved, while maintaining virtually 100% performance.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 illustrates the general scheme of the invention. Figure 2 illustrates a configuration of a membrane / enzyme system combination. Other possible configurations using intermittent, semi-intermittent and continuous systems will be apparent to those skilled in the art. Figure 3 is a graph illustrating the average chain length and the remaining sucrose enzyme hours.

Figure 4 is a graph illustrating the average chain length over enzyme hours. Figure 5 is a graph illustrating the average chain length and monosaccharides over time. Figure 6 is a graph illustrating the average DP length for DP3 + chains on the number of recirculations.

DETAILED DESCRIPTION OF THE INVENTION As used herein, the term "acceptor" will mean a carbohydrate that receives a monosaccharide or oligosaccharide from a carbohydrate donor., said acceptor can be a monosaccharide, disaccharide, oligosaccharide or polysaccharide. According to preferred embodiments, the "carbohydrate donor" can be, for example, a sugar nucleotide, a glycosyl phosphate, a disaccharide or an oligosaccharide. Examples of suitable monosaccharides include glucose, galactose, fructose, fucose, sialic acid, N-acetyl glucosamine, N-acetyl galactosamine, glucuronic acid, iduronic acid, xylose and mannose. The invention relates to a process for preparing a straight or branched sugar polymer of a desired chain length, preferably four or more units, which comprises: ^^^ J ^^^^^ tjé ^ '"" • "* - a) transfer by selective catalytic transfer, preferably reacting with an enzyme, a monosaccharide residue or oligosaccharide from a carbohydrate donor to an acceptor; b) remove byproducts that could inhibit the yield or selectivity of the reaction of step a), c) separate polymers that have not reached the desired chain length, and d) recirculate polymers that have not reached the desired chain length. Another aspect, the invention relates to an apparatus for producing straight or branched sugar polymers of a desired chain length, which comprises: a) a reactor in which a monosaccharide or oligosaccharide residue is transferred by selective catalytic transfer of a donor carbohydrates to an acceptor; b) means for removing byproducts of the reaction of step a); c) means for separating polymers that have not reached the desired chain length; and d) means for recirculating the polymer chains that have not reached the desired chain length back to the reactor. As shown in Figure 1, a sugar (carbohydrate) donor (1) is fed to the reactor (2), where the sugar polymer is made. The crude product mixture is transported to a separator (3) in ü iiuui ^ where the raw product is separated into a product of sugar polymer (4), the underdeveloped polymer (5), which is taken back to the reactor (2), and a byproduct (6) of the polymerization. Figure 2 illustrates an example of how the general arrangement 5 described in Figure 1 can be used to produce a fructose polymer from sucrose. A solution of sucrose (7) and enzyme (8) is fed in a controlled manner to the reactor (2), wherein the sucrose is converted into the polymer of fructose and glucose (by-product). A pump (9) doses the crude mixture to the separating membranes (3a, 3b and 3c). The membrane (3a) permeates mainly the glucose byproduct (6). The concentrated material is brought to a membrane (3b), wherein the sucrose and the underdeveloped polymer and the polymer chains, ie, the polymer chains that do not have the desired length (underdeveloped fructose polymers 5a), are returned to the reactor. The concentrate containing product and enzyme is leads directly to a third membrane (3c) separating the product (polyfructose (4)) from the enzyme (10) in the stream of concentrated material. Flow control valves (11 a-d) are placed to make possible the control of the flow velocities of the effluents that come from the separating membranes (3a, 3b, 3c). 20 The following key advantages can be achieved with this invention: 1. The system allows the use of an easily available and non-captive economic raw material. - '- * - - - > • '2.- Fundamentally, a production of 100% of the length distribution of any polymer can be achieved. 3.- Additional reaction phases can be eliminated to avoid inhibition. 4.- The system can use a series of available economic membranes or a continuous chromatography system. 5.- The membranes or chromatographic conditions can be modified quickly and economically or reconfigured to produce any distribution of desired product. 6.- The system can be operated continuously allowing consistent and easy operation. Other benefits will be evident for those familiar with sugar processing. The system is also very flexible as to how an enzyme can be used. In one form, a fresh enzyme is added as an ingredient enzyme and deactivated, periodically removed using a membrane to recover the sugars carried with the purge. In other cases, the organism that produces the enzyme can be grown in the reactor. This is the case in which the organism can be cultured under the reaction conditions and when it is naturally overexpressed and secretes the required enzyme. In the case of inulin, the yeasts that produce the fructose polymerization enzyme, for example; Aspergillus niger, A. japonicus, Pullularia or Aureobasidium pullulans, Saccharomyces cerevisiae, behave like & ^ ^^^^^ just described. In addition, they are grown in the byproduct of the reaction path, glucose. The methods using fructosyl transferases derived from Pullularia or Aureobasidium pullulans, and certain Aspergillus useful in the preparation of fructans, are described in the following patents: US Pat. No. 4,309,505; patent of E.U.A. No. 4,317,880; patent of E.U.A. No. 4,335,207; patent of E.U.A. No. 4,356,262; patent of E.U.A. No. 4,423,150; and patent of E.U.A. No. 4,849,356. It has also been discovered that the chromatography process can be used as an alternative to the membranes for carry out the separation. This would allow a narrower fractionation at a slightly higher cost of capital prices. Other systems can also be used that would offer a classification of polymers by means of chain lengths, such as selective precipitation. Since this new system has been described in terms of nullin, it can easily be seen that it is applicable to the production of other linear and branched sugar polymers. In addition, the system is useful for controlling the degree to which other functional sugars are added to the polymer chain. It is interesting that this global system simulates the operations that normally occur in a plant. For example, in the chicory plant, a se of enzymatic reactions converts C02 and water into simple sugars which in turn, through another se of enzymatic reactions, are converted into sugar polymers. These polymers are stored in the tuber of the chicory plant for use in the next season of ^ j ^^ g ^^ growth as a source of energy to start the next growth cycle. In the present system, other plants, such as sugar cane, provide simple sugar, and enzymes and reaction equipment are provided to make the polymerization. In the case of sugar polymers, this combination brings out the best part of the natural system and improves the "production logistics" of the plant. Consider a group of products based on fructose polymers. It is conceivable to "design" a plant that produces each of these polymer products. Unlike inulin, which the plant produces by its own need, the yields of these novel polymers would be based on how well the biochemical system of the plant is manipulated to produce the desired material. A method to control the chain length would also have to be built into the natural "regulator" system of the plant. Then, the plant could be harvested and the polymer extracted. Each of these products would be subject to all the general costs associated with its production and the risks associated with the development of agricultural products. If the use of any of these products is large enough, these costs and risks can be effectively absorbed, but for low volume products, the cost could imply significant limitations. The advantage of separating the "mass-produced" sucrose from the "custom-made" polymers is clear. The economy of agricultural and scale risk reduction is captured in the common intermediate, sucrose, to all low volume polymers, while custom manufacturing is limited to only one part of the scheme, «MÁÉi limiting general costs, capital, and the elimination of all agricultural risks. The following examples are intended to illustrate the invention, but not to limit it: EXAMPLE 1 Enzymatic synthesis of inulin as an intermittent procedure Fructosyl transferase (1000 Units / I) was added to a 70% by weight sucrose solution. The mixtures were then diluted with the water necessary to maintain 47% by weight of sucrose, and incubated at 50 ° C for 3.33, 6.67 or 18 hours. Boiling for 10 minutes stopped the reactions. The carbohydrate composition was analyzed by means of CLAR. The carbohydrate compositions were determined to include sucrose, glucose, fructose, difructose and unreacted inulin of chain length DP3-DP7 (DP refers to the degree of polymerization). Figure 3 shows the effect of the enzyme hours on the length of chains (*) and the remaining sucrose (M). Enzyme hours are defined as the enzyme concentration (units / kg of sucrose) per residence time (in hours).

EXAMPLE 2 Enzymatic synthesis of inulin as a continuous process An aqueous solution of sucrose (50% by weight) was pumped into a jacketed flask containing fructosyl transferase and sucrose solution. The content of the jacketed flask was recirculated through a 30K MWCO hollow fiber membrane chosen to allow the passage of all carbohydrates and the retention of fructosyl transferase. The flow passing through the hollow fiber membrane is regulated to equalize the solution flow of sugar in the flask with shirt. The permeate flow is directed either to a fraction collector, as in the case of a simple ("CSTR"), or to another jacketed flask containing fructosyl transferase, as in the case of a multiple CSTR. The contents of the jacketed flask were maintained at 50 ± 2 ° C. A series of simple and multiple CSTRs were performed varying the concentration of enzyme and residence time. The carbohydrate composition was analyzed by means of CLAR in each complete residence. It was determined that the carbohydrate compositions included sucrose, glucose, fructose, difructose and unreacted inulin, chain length DP3-DP7. Figure 4 shows the effect of enzyme hours (product of the concentration of enzymes in units / l per reaction time or residence time in hours) over the chain length for simple CSTR (A), double CSTR (•), triple CSTR (X) and intermittent reactions (*). a * a? ^^ kiMÍB EXAMPLE 3 Enzymatic synthesis of inulin during the elimination of the by-product Fructosyl transferase (40 units / 1) was added to a 5-neck flask containing a sucrose solution (50% by weight). The reaction was diluted with IM citrate to a pH of 5.5 and distilled water to obtain a solution of 20% sucrose by weight, 20mM citrate, at a pH of 5.5 with an enzyme concentration of 15 units / kg of sucrose. The final solution was maintained at 50 ° C for 24 hours while stirring constantly. At the end of the After a 24-hour reaction, the content of the jacketed flask was recirculated through a membrane selected to remove glucose selectively from the glucose, fructose, sucrose, difructose and inulin mixture of DP3-8 chain length. The permeate flow (glucose) was removed from the flask and the volume lost by the elimination of glucose was replaced with 20mM citrate. at a pH of 5.5. Under this configuration, the enzymatic reaction was allowed to continue during the glucose elimination. Figure 5 shows an average chain length (M) increasing during glucose removal (*). -and jM ^ .í- - EXAMPLE 4 Sugars recirculated by continuous separation using a membrane separator The crude reaction mixture as described in Examples 1 or 2 was passed through a membrane of 5000 NMCOW (Nominal Molecular Weight Cut) to retain fructosyl transferase and permeate all carbohydrates. The carbohydrate flow was diluted with an equal volume of water and then concentrated in a spiral membrane chosen in this way so that molecules with a DP 3 or greater were mostly retained and with DP2 or less were mostly permeated. In this case OSMONICS / DESAL G10 was used. The degree of separation between the permeate and the concentrate was calculated for each oligomer in the mixture. The calculated and observed values coincided in the system error. Three cycles of dilution and concentration in the retentate were completed. The final concentrate contained about 9-10% of mono- and di-saccharides. The concentrate was diluted and reconcentrated three more times using a slightly narrower membrane (OSMONICS / DESAL G5) until the content of mono- and di-saccharides was reduced to less than 5% of the total sugars in the solution. The combined permeate (approximately 80% of mono- and disaccharides based on the dissolved sugars) was concentrated using a membrane chosen so that only the monosaccharides permeated (OSMONICS / DESAL DL). The retentate contained disaccharides and smaller amounts of larger molecules that had infiltrated through the anterior membrane, and less than 60% of monosaccharides. This was returned directly to the reactor, along with the additional sucrose. The content of monosaccharides can be further reduced by repeating the dilution and reconcentration cycles. Other membranes, such as OSMONICS / DESAL G5 may be used in place of, or sequentially with, the DL to achieve an optimum degree of monosaccharide removal and recovery of desirable sugars in the recirculation flow.

EXAMPLE 5 Sugars recirculated by continuous separation using a chromatographic separator The crude reaction mixture as described in Examples 1 or 2 was passed through a 5000 MNWCO membrane to retain fructosyl transferase and permeate all carbohydrates. The chromatographic separator used was a simulated moving bed separator (SMB) that uses ten columns (2.54 cm X 48.26 cm). The supplies and extractions started at the points indicated below in table 1: TABLE 1 The separator was operated in the traditional way for SMB so that the supply and extraction positions would advance in a column with a fixed frequency (interval time, 560 sec.), While maintaining the same relative positions. A continuous recirculation of approximately 26 ml / min was maintained during the system. As is common for separations with SMB, the relative proportion of flows was adjusted periodically to maintain or change the desired separation. The composition of the outflows is shown below in table 2: TABLE 2 twenty • ÜMMÜÜIIÍi EXAMPLE 6 Enzymatic synthesis of nulin with recirculation of underdeveloped polymers A series of inulin synthesis reactions at 50% by weight and 28 units of fructosyl transferase / g carbohydrate were carried out with an incubation of 18 hours at 50 ° C. The carbohydrate composition of the starting material for the first reaction was 100% sucrose. The reaction was stopped by boiling, and the final carbohydrate composition was analyzed by means of HPLC. The glucose by-product was removed by means of chromatography. The remaining carbohydrates including glucose, fructose, sucrose and unreacted inulin, of residual chain length DP3-7, were concentrated in a rotary vaporizer, and used as a starting material in a subsequent inulin synthesis reaction. The carbohydrate compositions of subsequent reactions were about 20% recirculated carbohydrates and 80% sucrose. In total eleven reactions were performed. Ten reactions contained recirculated carbohydrates. Figure 6 shows the increase in polymer chain length with each reaction containing recirculated carbohydrates.

Claims (16)

NOVELTY OF THE INVENTION CLAIMS

1. A process for producing straight or branched chain carbohydrate polymers, characterized in that it comprises: a) the transfer, by selective catalytic transfer, of a monosaccharide or oligosaccharide residue from a donor carbohydrate to an acceptor; b) the elimination of any by-products of the reaction mixture formed in step a), which may inhibit the yield or selectivity; c) the separation of polymers that have not reached the desired chain length; and d) the recirculation of polymers that have not reached the desired chain length.

2. The process according to claim 1, further characterized in that the selective catalyst is an enzyme.

3. The process according to claim 2, further characterized in that the donor carbohydrate is a sugar nucleotide, a glycosyl phosphate, a disaccharide or an oligosaccharide.

4. The method according to claim 3, further characterized in that the monosaccharide or oligosaccharide is selected from glucose, galactose, fructose, fucose, sialic acid, N-acetylglucosamine, N-acetylgalactosamine, glucuronic acid, iduronic acid, xylose and mannose . »» ** »-» - "« »--- • - '•

5. The method according to claim 1, further characterized in that the enzyme used is produced in situ by means of a suitable organism. The method according to claim 1, further characterized in that the carbohydrate polymer produced is inulin 7. An apparatus for the production of straight or branched chain carbohydrate polymers of a desired chain length, characterized in that it includes: a ) a reactor to which a monosaccharide or oligosaccharide residue is transferred by means of selective catalytic transfer from a donor carbohydrate to an acceptor, b) means for the removal of byproducts of the reaction from step a); ) means for separating polymers that have not reached the desired chain length, and d) means for recirculating the polymer chains that have not reached the desired chain length of vu. 8. The apparatus according to claim 7, further characterized in that the means for separating polymers are at least one membrane. 9. The apparatus according to claim 8, further characterized in that the reactor and the membrane housing are a single unit. «. .. *. **. «.» »......, - .., ... ^^ ". ^ - ^ _ ^ 10. The apparatus according to claim 7, further characterized in that the means for the separation of polymers employ chromatography. 11. The apparatus according to claim 7, further characterized in that the means for the separation of polymers employ chromatography selected from ligand exchange resin, size exclusion and ion exchange. 12. The apparatus according to claim 11, further characterized in that the selective catalyst is an enzyme. 13. The apparatus according to claim 12, further characterized in that the enzyme is suspended in, or adhered to, the * resin 14. The apparatus according to claim 7, further characterized in that the means for the separation of polymers 15 employ selective crystallization. 15. The apparatus according to claim 7, further characterized in that the means for the separation of polymers employ selective precipitation. 1

6. The method according to claim 1, further characterized in that the method is continuous, intermittent or semi-intermittent.