WO2003002728A2 - Prevention of retrogradation of starch - Google Patents

Abstract

The invention provides an isolated or recombinant nucleic acid derived from a nucleic acid encoding a polypeptide essentially having alpha-glucanotransferase activity but having essentially no hydrolysing activity, said isolated or recombinant nucleic acid encoding a polypeptide with hydrolytic activity.

Description

Title: Prevention of retrogradation of starch.

Starch containing crops form an important constituent of the human diet and a large proportion of the food consumed by the world's population originates from them. Besides the use of the starch-containing plant parts directly as a food source, starch is harvested and used as such or chemically or enzymatically processed into a variety of different products such as starch hydrolysates, glucose syrups, fructose, starch or maltodextrin derivatives, or cyclodextrins. In spite of the large number of plants able to produce starch, only a few plants are important for industrial starch processing. The major industrial sources are maize, tapioca, potato, and wheat. In the European Union, 3.6 million tons of maize starch, 2 million tons of wheat starch, and 1.8 million tons of potato starch were produced in 1998 (DeBaere, 1999). Starch is found as granules containing polymers of glucose linked to one another through the Cl oxygen, known as the glycosidic bond. This glycosidic bond is stable at high pH but hydrolyses at low pH. At the end of the polymeric chain, a latent aldehyde group is present. This group is known as the reducing end. Two types of glucose polymers are present in starch granules: (i) amylose and (ii) amylopectin. Amylose is a hnear polymer consisting of up to 6,000 glucose units with alpha, 1-4 glycosidic bonds. The number of glucose residues, also indicated with the term DP (degree of polymerization), varies with the origin. Amylose from e.g. potato or tapioca starch has a DP of 1,000 - 6,000 while amylose from maize or wheat amylose has a DP varying between 200 and 1,200. The avarage amylose content in starches can vary between almost 0 to 75%, but a typical value is 20-25%. Amylopectin consists of short alpha, 1-4 linked hnear chains of 10-60 glucose units and alpha, 1-6 linked side chains with 15-45 glucose units. The average number of branching points in amylopectin is 5%, but varies with the botanical origin. The complete amylopectin molecule contains on avarge about 2,000,000 glucose units, thereby being one of the largest molecules in nature. The most commonly accepted model of the structure of amylopectin is the cluster model, in which the side chains are ordered in clusters on the longer backbone chains (see Buleon et al., 1998; Myers et al., 2000).

While amylopectin is soluble in water, amylose and the starch granule itself are insoluble in cold water. This makes it relatively easy to extract starch granules from their plant source. When a water-starch slurry is heated, the granules first swell untill a point is reached at which the swelling is irreversible. This swelling processes is termed gelatinization. During this process, amylose leaches out of the granule and causes an increase in the viscocity of the slurry. Further increase in temperature then leads to maximum swelling of the granules and increased viscosity. Finally, the granules break apart resulting in a complete viscous colloidal dispersion. Subsequent cooling results in association of the amylose chains, resulting in the formation of insoluble aggregates. In diluted starch suspensions these aggregates precipitate, cooling of a concentrated colloidal starch dispersion results in the formation of an elastic gel. This retrogradation is primarily caused by the amylose; amylopectin, due to its highly branched organization, is less prone to retrogradation.

A large variety of bacteria employ extracellular or intracellular enzymes able to convert starch or glycogen, that thus can serve as energy and carbon sources (Fig. 2). There are basically four groups of starch-converting enzymes: (i) endoamylases; (ii) exoamylases; (iii) debranching enzymes; and (iv) transferases.

Endoamylases are able to cleave alpha, 1-4 glycosidic bonds present in the inner part (endo-) of the amylose or amylopectin chain, alpha- Amylase (EC 3.2.1.1) is a well known endoamylase. It is found in a wide variety of microorganisms, belonging to the Archaea as well as the Bacteria (Pandey et al., 2000). The end products of alpha- amylase action are oligosaccharides with varying length with an alpha-configuration and alpha-limit dextrins, which constitute branched oligosaccharides.

Enzymes belonging to the second group, the exoamylases, either exclusively cleave alpha, 1-4 glycosidic bonds such as b-amylase (EC 3.2.1.2) or cleave both alpha, 1-4 and alpha, 1-6 glycosidic bonds like amyloglucosidase or glucoamylase (EC 3.2.1.3) and alpha-glucosidase (EC 3.2.1.20). Exoamylases act on the external glucose residues of amylose or amylopectin and thus produce only glucose (glucoamylase and alpha-glucosidase), or maltose and b-limit dextrin (b-amylase).

The third group of starch-converting enzymes are the debranching enzymes that exclusively hydrolyse alpha, 1-6 glycosidic bonds: isoamylase (EC 3.2.1.68) and pullanase type I (EC 3.2.1.41). Pullulanases hydrolyse the alpha, 1-6 glycosidic bond in pullulan and amylopectin, while isoamylase can only hydrolyse the alpha, 1-6 bond in amylopectin. These enzymes exclusively degrade amylopectin, thus leaving long hnear polysaccharides. There are also a number of pullulanase type enzymes that hydrolyse both alpha, 1- 4 and alpha, 1-6 glycosidic bonds. These belong to the group II pullulanase and are also referred to as alpha-amylase-pullulanase or amylopullulanase. The main degradation products are maltose and maltotriose. The fourth group of starch-converting enzymes are transferases that cleave an alpha, 1-4 glycosidic bond of the donor molecule and transfer part of the donor to a glycosidic acceptor with the formation of a new glycosidic bond. Enzymes such as amylomaltase (EC 2.4.1.25) and cyclodextrin glycosyltransferase (EC 2.4.1.19) form a new alpha, 1-4 glycosidic bond while branching enzyme (EC 2.4.1.18) forms a new alpha, 1,6 glycosidic bond.

Cyclodextrin glycosyltransferases have a very low intrinsic hydrolytic activity and make cyclic oligosaccharides with 6,7,or 8 glucose residues and highly branched high molecular weight dextrins, the cyclodextrin glycosyl-transferase limit dextrins. Cyclodextrins are produced via an intramolecular transglycosylation reaction in which the enzyme cleaves an alpha, 1-4 glycosidic bond and concomitantly hnks the reducing to the non-reducing end (Takaha and Smith, 1999; van der Veen et al., 2000a).

Amylomaltases are very similar to cyclodextrin glycosyltransferases with respect to the type of enzymatic reaction. The major difference is that amylomaltase performs a transglycosylation reaction resulting in a hnear product while cyclodextrin glycosyltransferase gives a cyclic product. Another difference is that they essentially do not hydrolyse starch. Amylomaltases have been found in different microorganisms in which it is involved in the utilization of maltose or the degradation of glycogen (Takaha and Smith, 1999). Glucan branching enzymes are invloved in the synthesis of glycogen in many microorganisms. They are responsible for the formation of alpha, 1-6 glycosidic bonds in the side chains of glycogen and in general do not hydrolyse either. Although glycogen has been found in a large number of microorganisms (Preiss, 1984), only a limited number of microbial glucan branching enzymes have been characterized (Kiel et al., 1991; Kiel et al., 1992; Takata et al., 1994; Binderup and Preiss, 1998).

Most of the enzymes mentioned above belong to one family based on amino acid sequence homology: the alpha-amylase family or family 13 glycosyl hydrolases according to the classification of Henrissat (1991). This group comprises those enzymes that have the following features: (i) they possess a (bete/alpha)₈ or TIM barrel (Fig. 3) structure containing the catalytic residues; (ii) they have four highly conserved regions in their primary sequence (Table 1) which contain the amino acids that form the catalytic site, as well as some amino acids that are essential for the stability of the conserved TIM barrel topology (Kuriki and Imanaka, 1999); (iii) they act on alpha-glycosidic bonds and hydrolyse or trans gly cosylate this bond with retention of the alpha-anomeric configuration through a double displacement mechanism.

The enzymes that match the above mentioned criteria and belong to the alpha- amylase family are listed in Table 2. During the last three decades, alpha-amylases have been exploited by the starch- processing industry as a replacement of acid hydrolysis in the production of starch hydrolysates. This enzyme is also used for removal of starch in beer, fruit juices, or from clothes and porcelain. Another starch-hydrolysing enzyme that is used on a large scale is thermostable pullulanase for the debranching of amylopectin. A recent apphcation is directed at the use of maltogenic amylases as an anti-stahng agents to prevent the retrogradation of starch in bakery products.

The baking industry is a large consumer of starch and starch modifying enzymes. Bread baking starts with dough preparation by mixing flour, water, yeast and salt and possibly additives. Flour consists mainly of gluten, starch, non-starch polysaccharides and lipids. Immediately after dough preparation, the yeast starts to ferment the available sugars into alcohols and carbon dioxide, which causes rising of the dough. Amylases can be added to the dough to degrade the damaged starch in the flour into smaller dextrins, which are subsequently fermented by the yeast. The addition of malt or fungal α-amylase to the dough results in increased loaf volume and improved texture of the baked product

After rising, the dough is baked. When the bread is removed from the oven, a series of changes start which eventually leads to deterioration of quality. These changes include increase of crumb firmness, loss of crispness of the crust, decrease in moisture content of the crumb and loss of bread flavor. All undesirable changes that do occur upon storage together are called staling. Retrogradation of the starch fraction in bread is considered to be very important in stahng (Kulp and Ponte, 1981). Especially the extent of amylopectin retrogradation strongly correlates with the firming rate of bread (Champenois et al., 1999). Stahng is of considerable economic importance for the baking industry since it hmits the shelf life of baked products. In the USA, for instance, bread worth more than 1 billion US$ is discarded annually.

To delay stahng, to improve texture, volume and flavor of bakery products, several additives may be used in bread baking. These include chemicals, small sugars, enzymes or combinations of these. Well known additives are: milk powder, gluten, emulsifiers (mono- or diglycerides, sugar esters, lecithin etc.), granulated fat, oxidant (ascorbic acid or potassium bromate), cysteine, sugars or salts (Spendler and Jørgensen, 1997). Rapid advances in biotechnology have made "new" enzymes available for the industry. Since enzymes are produced from natural ingredients, they will find greater acceptance by the consumers as they demand for products without chemicals. Several enzymes have been suggested to act as dough and/or bread improvers, by modifying one of the major dough components. Examples are glucose oxidase, hemicellulase, lipase, protease and xylanase. These enzymes, however, do not act on the starch fraction itself. Enzymes active on starch have been suggested to act as anti-stahng agents. Examples are: α-amylases (De Stefanis and Turner, 1981; Cole, 1982), branching (Okada et al., 1984) and debranching (Carroll et al., 1987) enzymes, maltogenic amylases (Olesen, 1991), β-amylases (Wύrsch and Gumy, 1994), and amyloglucosidases (Vidal and Gerrity, 1979). Present nti-staling agents, however, often act to fast. Originally, α-amylases were added during dough preparation to generate fermentable compounds. Besides generating fermentable compounds, α-amylases also have an anti-stahng effect in bread baking, and they improve the softness retention of baked goods (De Stefanis and Turner, 1981; Cole, 1982; Sahlstrόm and Brathen, 1997). Despite a possible anti-stahng effect, the use of α-amylases as anti-stahng agent is not widespread because even a shght overdose of α-amylase results in a sticky bread. Positive effects of delayed staling, on the contrary, are measured only after 3 to 4 days (Olesen, 1991). The increased gummyness of α-amylase treated bread is associated with the production of branched maltodextrins of DP20-100 (De Stefanis and Turner, 1981). Debranching enzymes are claimed to strongly decrease the problems associated with the use of α-amylases as anti-stahng agents in baking. In this method a thermostable pullulanase, and an α-amylase are used together. The pullulanase rapidly hydrolyzes the branched maltodextrins of DP20-100 produced by the α-amylase, while they have httle effect upon the amylopectin itself (Carroll et al., 1987). Pullulanase thus specifically removes the compound responsible for the gummyness associated with α-amylase treated bakery products.

Branching enzyme is claimed to increase shelf life and loaf volume of baked goods (Okada et al., 1984; Spendler and Jorgensen, 1997). These effects are achieved by modifying the starch material in the dough during baking. Improved quahty of baked products is also obtained when the branching enzyme is used in combination with other enzymes, such as α-amylase, maltogenic amylase, cyclodextrin glycosyltransferase, β-amylase, cellulase, oxidase and/or hpase (Spendler and Jorgensen, 1997). The use of cyclodextrin glycosyltransferase as dough additive is claimed to increase loaf volume of the backed product (Van Eijk and Mutsaers, 1995). The effect is suggested to result from the gradual formation of cyclodextrins in the dough after mixing.

Exo-amylases, such as β-amylase and amyloglucosidase, shorten the external side chains of amylopectin by cleaving of maltose or glucose molecules, respectively. Both enzymes are suggested to delay bread stahng by reducing the tendency of the amylopectin compound in bakery products to retrograde (Wϋrsch and Gumy, 1994). Anti-stahng effects of amylo-glucosidase in baking are claimed in a few patents (Van Eijk, 1991; Vidal and Gerrity, 1979). The synergetic use of an α- and a β-amylase is also claimed to increase the shelf hfe of baked goods (Van Eijk, 1991).

Since α-amylases cause stickiness of backed goods, especially when overdosed, it was suggested that these problems could be solved using an exo-amylase, since they do not produce the branched maltooligosaccharides of DP20-100. Such enzymes, called maltogenic amylases, produce hnear oligosaccharides of 2 to 6 glucose residues. Maltogenic amylases producing maltose (Olesen, 1991), maltotriose (Tanaka et al., 1997) and maltotetraose (Shigeji et al., 1999a; Shigeji et al., 1999b) are claimed to increase the shelf hfe of bakery products by delaying retrogradation of the starch compound. Currently, a thermostable maltogenic amylase of Bacillus stearothermophilus (Diderichsen and Christansen, 1988) is used commercially in the bakery industry. Although this enzyme has some endo-activity (Christophersen et al., 1998), it does act as an exo-acting enzyme during baking, modifying starch at a temperature when most of the starch starts to gelatinize (Olesen, 1991). Cherry et al. (1999) described in detail the 3D-structure of the maltogenic alpha- amylase and used this to suggest specific amino acid modifications to obtain variants of the enzyme with improved product specificity, altered pH optimum, improved thermostabihty, increased specific activity, altered cleavage pattern and thus have an increased ability to reduce retrogradation of starch or stahng of bread.

Cyclodextrins are cyclic alpha, 1-4 hnked oligosaccharides mainly consisting of 6, 7,or 8 glucose residues. The glucose residues in the rings are arranged in such a manner that the inside is hydrophobic thus resulting in an apolar cavity while the outside is hydrophihc. This enables cyclodextrins to form inclusion complexes with a variety of hydrophobic guest molecules. Specific cyclodextrins are required for complexation of guest molecules of specific sizes. The formation of inclusion complexes leads to changes in the chemical and physical properties of the guest molecules, such as stabilization of light- or oxygen sensitive compounds, stabilization of volatile compounds, improvement of solubility, improvement of smell or tast, or modification of hquid compounds to powders. These altered characteristics of the encapsulated compounds have led to various applications of cyclodextrins in analytical chemistry (Armstrong, 1988; Loung et al., 1995), agriculture (Saenger, 1980; Oakes et al., 1991), biotechnology (Allegre and Deratani, 1994; Szejth, 1994), pharmacy (Albers and Muller, 1995; Thompson, 1997), food (Allegre and Detrani, 1994; Bicchi et al., 1999) and cosmetics (Allegre and Detrani, 1994).

A major drawback for the apphcation of cyclodextrins on a large scale is that all enzymes used today produce a mixture of cyclodextrins. Two different industrial approaches are used to purify the cyclodextrin mixtures: selective crystallization of beta-cyclodextrin, which is relatively poorly water-soluble, and selective complexation with organic solvents. Major disadvantages of the latter method is the toxitiy, flammabihty, and need for solvent recovery (Pedersen et al., 1995). This makes the production of cyclodextrins too costly for many applications. Additionally, the use of organic solvents limits applications involving human consumption.

For the industrial production of cyclodextrins, starch is first liquefied by a heat- stable alpha-amylase and then the cyclization occurs with a cyclodextrin glycosyltransferase from Bacillus macerans (Riisgaard, 1990) sp. A major drawback of this process is that the cychzation reaction has to be performed at lower temperatures than the initial hquefaction because of the low thermostabihty of the bacillus cyclodextrin glycosyltransferase. The use of cyclodextrin glycosyltransferase from thermophilic microoganisms can solve this problem. Thermostable cyclodextrin glycosyltransferases have been found in a Thermoanaerobacter species (Starnes, 1990; Norman and Jørgensen, 1992; Pedersen et al., 1995), Thermoanaero-bacterium thermosulfurogenes (Wind et al., 1995), and Anaerobranca bogoriae (Prowe et al., 1996).

Cyclodextrin glycosyltransferases can also be used for the production of novel glycosylated compounds, making use of the transglycosylation activity. A commercial application is the glycosylation of the intense sweetener stevioside, isolated from the leaves of the plant Stevia rebaudania, thereby increasing solubility and decreasing the bitterness (Pedersen et al. 1995).

Other cyclic products that can be generated from starch are cycloamyloses. These large cyclic glucans (DP >20) contain antiparalel helices, providing long cavities with a diameter similar to that of alpha-cyclodextrin. Unlike cyclodextrins, cycloamylose is formed by aU transglycosylating enzymes of the alpha-amylase family (Takata et al., 1996; Terada et al., 1997; Terada et al., 1999). Formation of cyclodextrins occurs by an intramolecular transglycosylation reaction whereas the formation of large cycloamylose molecules is the result of an intramolecular transglycosylation. To form cycloamylose, low concentrations of high molecular weight amylose are incubated with a relatively high amount of enzyme. This reaction is therefore not based on a novel catalytic mechanism but is a direct effect of the hmited availability of acceptor molecules. Production of cycloamylose is currently not done on an industrial scale. alpha-Amylase, pullulanase, cyclodextrin glycosyltransferase, and maltogenic amylase are nowadays widely used by industry in various applications (Table 3). alpha-Amylase probably has the most wide-spread use. Besides their use in hydrolysis leading to the saccharification or hquefaction of starch, these enzymes are also used for the preparation of viscous, stable starch solutions used for the warp sizing of textile fibers, the clarification of haze formed in beer or fruit juices, or for the pretreatment of animal feed to improve the digestibility. A growing new area of apphcation of alpha- amylases is in the fields of laundry and dish-washing detergents. A modern trend among consumers is to use colder temperatures for doing the laundry or dish-washing. At these lower temperatures the removal of starch from cloth and porcelain becomes more problematic. Detergents with alpha-amylases optimally working at moderate temperatures and alkaline pH can help to solve this problem. Two starch-modifying enzymes of the alpha-amylase family that do not find large scale apphcation yet are amylomaltase and branching enzyme. Apphcation of branching enzymes is hmited by the lack of commercially available enzymes that are sufficiently thermostable. A potentially interesting industrial apphcation of amylomaltase is the production of thermoreversible starch gels. As already indicated above, a normal untreated starch gel cannot be dissolved in water after it has retrograded. However, starch that has been treated with amylomaltase has obtained thermoreversible gelling characteristics: it can be dissolved numerous times upon heating. This behaviour is very similar to gelatine. Van der Maarel et al. (2000) described this process using the amylomaltase from the hyperthermophihc bacterium Thermus thermophilus. Currently, no amylomaltases are commerciaUy available and the thermoreversible starch gel is not produced on an industrial scale.

Table 1. The four conserved regions and the corresponding b-sheets found in the amino acid sequence of amylomaltase and alpha-amylase family enyzmes. Highhghted are the conserved active site amino acid residues. The following enzymes were used for the alignment: amylomaltase of Thermus aquaticus (Terada et al. 1999); amylosucrase of Neisseria poly saccharea (Buttcher et al. 1997); CGTase: cyclodextrin glucosyltranferase oi Bacillus circulans 251 (Lawson et al. 1994); CMDase: cyclomaltodextrinase of Clostridium thermohydrosulfuricim 39E (Podkovyrov & Zeikus 1992); BE: branching enzyme oi Bacillus stearothermophilus (Kiel et al. 1991); isoamylase of Pseudomonas amyloderamosa (Amemura et al. 1988); M. amylase: maltogenic alpha-amylase oi Bacillus stearothermophilus (Cha et al. 1998); pullulanase oi Bacillus flaυocaldarius KP 1228 (Kashiwabara et al. 1999); sucrose Pase: sucrose phosphorylase of Escherichia coli K12 (Aiba et al. 1996); Taka- amylase: alpha-amylase of Aspergillus oryzae (Matsuura et al. 1980). b2, b4, b5, and b7 indicate the beta-sheet in which this region is present.

Region Ib2 Region IIb4 Region IIIb5 Region IVb7

Amylomaltase EALGIRIIGDMPIFVAED LFHLVRIDHFRG VPVLAEDLGVI

WYTGTHDNDT Amylosucrase HEAGISAWDFIFNHTSN GVDILRMDAVAF VFFKSEAIVHP

VNYVRSHDDIG

CGTase HAKNIKVIIDFAPNHTSP GIDGIRMDAVKH VFTFGEWFLGV

VTFIDNHDMER

CMDase HDNGIKNIFDAVFNHCGY DIDGWRLDVANE AIIVGEVWHDA FNLIGSHDTER

BE HQAGIGVILDWVPGHFCK HVDGFRVDAVAN ILMIAEDSTDW

FILPFSHDEW

Isoamylase HNAGIKVYMDWYNHTAE GVDGFRFDLASV LDLFAEPWAIG

INFIDVHDGMT M.amylase HQKAIRVMLDAVFNHSGY DIDGWRLDVANE AYILGEIWHDA

FNLLGSHDTPR

Pullulanase HAHGVRVILDGVFNHTGR GVDGWRLDVPNE AYIVGEIWEEA

MNLLTSHDTPR Sucrose Pase LGECSHLMFDFVCNHMSA GAEYVRLDAVGF TVIITETNVPH

FNFLASHDGIG

Taka-amylase HERGMYLMVDWANHMGY SIDGLRIDTVKH VYCIGEVLDGD

GTFVENHDNPR

Table 2. Enzymes of the alpha-amylase family that act on glucose-containing substrates, their corresponding E.C. number, the domain organization as far as it has been described, and main substrates.

Enzyme E.C. number Domains Main substrate amylosucrase 2.4.1.4 sucrose sucrose phosphorylase 2.4.1.7 sucrose glucan branching enzyme 2.4.1.18 A, B, F starch, glycogen cyclodextrin glycosyltransferase 2.4.1.19 A, B, C, D, E starch amylomaltase 2.4.1.25 A, Bl, B2 starch, glycogen maltopentaose-forming amylase 3.2.1.- A, B, I starch alpha-amylase 3.2.1.1 A, B, C starch ohgo- 1,6-glucosidase 3.2.1.10 A, B amylopectin alpha-glucosidase 3.2.1.20 starch amylop ullulanase 3.2.1.41 or A, B, H, G, 1 pullulan cyclomaltodextrinase 3.2.1.54 A, B cyclodextrins isopullulanase 3.2.1.57 pullulan isoamylase 3.2.1.68 A, B, F, 7 amylopectin maltotetraose-forming amylase 3.2.1.60 A, B, C, E starch glucodextranase 3.2.1.70 starch trehalose-6-phosphate 3.2.1.93 trehalose maltohexaose -forming amylase 3.2.1.98 starch maltogenic amylase 3.2.1.133 A, B, C, D, E starch neopullulanase 3.2.1.135 A, B, G pullulan malto-oligosyl trehalose 3.2.1.141 trehalose malto-ohgosyl threhalose 5.4.99.15 maltose

Table 3. Different fields of apphcation of enzymes belonging to the alphalpha-amylase family

Apphcation Enzyme

Starch hquefaction alpha-amylase Starch saccharification amyloglucosidase, pullulanase, maltogenic alpha-amylase, alpha-amylase, isoamylase

Laundry detergent and cleaners; alpha-amylase reduction of haze formation in juices, baking, brewing, digestibility of animal feed, fiber and cotton desizing, sanitary waste treatment Cyclodextrin production cyclodextrin glycosyltransferase Thermoreversible starch gels amylomaltase Cycloamylose amylomaltase, branching enzyme, cyclodextrin glycosyltransferase

The invention provides an isolated or recombinant nucleic acid encoding a 4- alpha- or 6-alpha-glucanotransferase, which, in a preferred embodiment, is provided with hydrolytic activity, a or functional fragment thereof. In one embodiment, the invention provides such a nucleic acid encoding an amylomaltase, the wild types of which are generaly not known for any hydrolysing activity. 4-α-Glucanotransferase (e.g. EC 2.4.1.25, amylomaltase (AMase) or D-enzyme) forms a separate family (77) of glycosyl hydrolases. However, it is closely related to the alpha-amylase family or family 13 of glycosyl hydrolases. Unlike most members of this family of enzymes 4-α- glucanotransferase is not directly involved in starch degradation, but promotes metabohsm of starch degradation products inside the cell (AMase), or is involved in starch biosynthesis (D-enzyme). Recently, however, the action of amylomaltase from Thermus thermophilus on starch has been described, resulting in the production of a thermoreversible gel. To investigate the enzymatic properties responsible for this action the T. thermophilus mαZQ gene has been cloned and expressed in E. coli, and its sequence as here been provided, allowing purification of large amounts of enzyme, and manipulation of the gene.

In order to determine the AMase reaction specificity its action on maltoohgosaccharides and soluble starch was analyzed. Although the enzyme is closely related to the α-amylase family, of the wild type enzyme no hydrolyzing activity could be detected. In the disproportionation reaction the enzyme was found to prefer longer ohgosaccharides as donor substrates, while shorter oligosaccharides are efficiently used as acceptors. As observed for other amylomaltases, maltose is not cleaved off and hardly used as acceptor by the enzyme. The complete lack of hydrolyzing activity of wild type AMase and its specificity for donor and acceptor substrates makes it a very interesting enzyme to be studied regarding reaction and product specificity.

In another embodiment, the invention provides a nucleic acid encoding a enzyme or polypeptide derived from said non-hydrolysing enzyme, now provided with hydrolysing acitivity. For example, interaction with hydrophobic amino acids, such as F366, which is highly conserved in amylomaltases, is involved in the reaction specificity of the enzyme. Hydrolyzing activity can be introduced by mutating this residue, or other hydrophobic residues such as F251 or Y54. This hydrolyzing activity has significant effects on product profiles of the enzyme, indicating the necessity of complete absence of hydrolysis for the function of the wild type enzyme (the production of longer oligosaccharides from short substrates). Now that the enzyme has been provided with hydrolysing activity, it can be used in preventing retrogradation of starch. Especially useful in such prevention is the use of a newly hydrohsing enzyme as provided herein that is derived from thermostable transferase, which can be found in a thermophilic micro-organism Particulary provided is such an enzyme wherein said micro-organism comprises Thermus thermophilus, Thermus aquaticus or Aquifex aeolicus.

Also, the branching enzyme (BE) gene from Aquifex aeolicus (BE Aae) was cloned, sequenced ( for he amino acid sequence see fig 4) and overexpressed in E. coli. The thermostable branching enzyme was purified to homogeneity, and biochemically characterized. The temperature optimum for activity was 80 °C, which is the highest optimum known for branching enzymes as compared with the other known thermostable branching enzyme from Bacillus stearothermophilus (BE Bst) which has a temperature optimum of 50 °C. This higher temperature optimum is very useful in hydrolysing starch. Furthermore, BE Aae was determined to be thermostable up to 90 °C compared with approximately 60 °C for BE Bst. Branching enzymes (BE) catalyze the formation of alpha- 1,6-glucosidic linkages in two steps (pres. via cov. interm.). The first step is the cleavage of an alpha- 1,4-glucosidic hnkage followed by a transfer of the oligosaccharide to the 6-position of another glucose present within an alpha- 1,4 glucosidic chain. This results in the branching points present in starch and glycogen. It has been shown that a lot of organisms are capable of producing starch or glycogen and express BE in order to do so. From various sources the BE has been cloned and characterized. It has been shown that BE's belong to the alpha-amylase family and that they posses the four conserved regions present within the family. A 3D model of the BE from Aquifex aeolicus has been designed. The crystal structure of isoamylase from Pseudomonas was used for modelling using the program Swiss-Pdb viewer. All amino acids that are conserved in the catalytic center within the alpha- amylase family were present in the active site of the 3D-model of BE from Aquifex aeolicus. The most striking feature was the present of hydrophobic residues (see fig 5) at the putative acceptor site. Alignment of branching enzymes showed that these residues are highly conserved (see fig 4) . These residues are mutated to more hydrophilic residues, for example according to the table below

Table 4 Active site residue mutagenesis of BE. function mutant acceptor site W276Q W367Q

W385Q M387S F458S Y460S donor site Y512S catalytic site D311N E362Q D430N

Now that a branching enzyme has been provided with hydrolysing activity, it can be used in preventing retrogradation of starch. Especially useful in such prevention is the use of a newly hydrohsing enzyme as provided herein that is derived from thermostable transferase, which can be found in a thermophilic micro-organism Particulary provided is such an enzyme wherein said micro-organism comprises Thermus thermophilus, Thermus aquaticus or Aquifex aeolicus.

In overview, the invention provides modified a transferase that is derived from or has an activity of an enzyme known under EC number 2.4.1.25 or 2.4.1.18, with added hydrolysing activity. These are derivable from a nucleic acid according to the invention provided with a mutation leading to an alteration or loss of a codon originally encoding a hydrophobic amino acid located in or around a acceptor, a donor or a catalytic site extending from a TIM barrel structure 'of said transferase. Such mutation is preferably provided by site-directed mutagenesis, wherein said codon originally encoding a hydrophobic amino acid is altered into a codon encoding an amino acid which is substantially less hydrophobic. Preferably, the hydrophobic amino acid to be changed comprises phenylalanine, tryptophan or tyrosine, and is located at aor around the positions as indicated herein in the (beta/alpha)β or TIM barel structure of the enzyme. For example, a nucleic acid is provided wherein said change in hydrophobic amino acid is located at or around an amino acid position essentiaUy corresponding to amino acid position 54, 251, 258 or 366 of amylomaltase of Thermus thermophilus HB8. Furthermore, the invention provides a vector comprising a nucleic acid according to the invention and a host ceh comprising a vector or a nucleic acid according to the invention.

As said, and further explained in the detaUed descriptuion herein, the invention provides a method for providing a polypeptide or fragment thereof essentiaUy having alpha glucanotransferase acitivity but having essentiaUy no hydrolysing activity with specific hydrolysing activity said method comprising providing a nucleic acid encoding such a transferase with a mutation leading to an alteration or loss of a codon originaUy encoding a hydrophobic amino acid located in or around a acceptor, a donor or a catalytic site extending from a TIM barrel structure of said transferase, and provides a polypeptide obtainable therewith.

The invention also provided use a polypeptide or fragment according to the invention in reducing retrogradation of starch, such as in reducing retrogradation of amylopectine, particularly in reducing long-term retrogradation of amylpectine. The invention provides specific enzymes provided with one of more specific amino acid modifications to obtain variants of the enzyme with hydrlolysing activity, and thus with improved product specificity, altered pH optimum, improved thermostabihty when strating with a thermostable enzyme as provided herein, increased specific activity, altered cleavage pattern. An enzyme as provided herein has increased ability to reduce retrogradation of starch or stahng of bread.

Also, the invention provides use such a polypeptide or fragment in hydrolysing starch, said uses for example apphed in the prevention or at least temporarily avoiding of stahng of bakery products such as bread, or as a replacement of acid hydrolysis in the production of starch hydrolysates. Such prevention of staling comprises use of a method for reducing retrogradation of starch comprising treating said starch with a polypeptide or fragment, such as a amylomaltase or branching enzyme provide with hydrolysing activity according to the invention. Improved quahty of baked products is further obtained when the alpha-glucanotransferase (e.g. amylomaltase or branching enzyme) provided with hydrolysing activity according to the invention is used in combination with other enzymes, such as α-amylase, maltogenic amylase, cyclodextrin glycosyltransferase, β-amylase, ceUulase, oxidase and/or hpase Furthermore, the invention provides a bakery ingredient comprising a polypeptide according to the invention and a bakery product such as bread comprising a polypeptide according to the invention. The invention is further explained in the detailed description provided herewith.

Figure legends

Figure 1: OveraU secondary structure of the amylomaltase from Thermus thermophilus. The central (b/a)s barrel is shown; this barrel consists of 8 β-sheets, depicted as arrows, surrounded by 8 α-helices, depicted as spirals. The amino acid residues constituting the catalytic site extend from this barrel into the active site surrounded by subdomains Bl, B2 and B3 respectively. Amino acid residues involved in binding of the donor and acceptor substrates are located in and extending from subdomain Bl and loops protruding from the (b/a)s barrel.

Figure 2: A model showing the binding of a maltoheptaose substrate in the active site of the T. thermophilus amylomaltase. The sugar residues are numbered according to the general subsite labehng scheme proposed for aU glycosyl hydrolases by Davies et al. (Biochem. J. 1997, 321: 557-559), in which the glycosidic bond between -1 and +1 is the bond which is cleaved, and the substrate reducing end is at position +3. The positively numbered subsites constitute the acceptor binding site. The following amino acid residues are shown: (i) The catalytic residues Asp293 and Glu340; (ii) those involved in interactions with the substrate by hydrogen bonds, which are indicated by dotted hnes; (hi) the aromatic amino acids involved in hydrophobic stacking interactions, being Tyr54, Trp258, Phe251, and Phe366. The model was constructed manuaUy with the program O (Jones et al. 1991 Acta Crystallogr. D55, 849-861) on basis of the 3D structures of a porcine pancreatic a- amylase-hexasaccharide complex (Machius et al. 1996, J. Mol. Biol. 260, 409-421) and a cyclodextrin glycosyltransferase-maltononaose complex (Uitdehaag et al. 1999, Nature Struct. Biol. 6, 432-436). For clarity the model does not show the conserved catalytic site residues Tyr 59, Arg 291, His 294 and Asp 395.

Figure 3: The amino acid sequence ahgnment of Aquifex aeolicus branching enzyme (glgB Aqu) with Pseudomonas amyloderamosa isoamylase (isoamyla) used for constructing the 3-D model of the Aquifex aeolicus branching enzyme. Symbols represent the foUowing: dots, functionaUy simhar amino acids; *, identical amino acids; s, amino acids present in a β-sheet; amino acids present in an -helix. bl-b8 and al-a8 represent the alternating β-sheets and α-hehces, respectively, comprising the (β/α)₈ barrel.

Figure 4: Detailed overview of the active site of Aquifex aeolicus branching enzyme, showing the catahtyc amino acid residues Asp311 (D311 cat.res.), Glu362 (E362 cat.res.), and Asp430 (D430 cat.res.) and the hydrophobic amino acid residues surrounding the catalytic site Trp276 (W276), Trp367 (W367), Trp385 (W385), Met387 (M387), Phe458 (F458), Tyr460 (Y460), and Tyr512 (Y512).

DetaUed description

Kinetic analysis of amylomaltase from Thermus thermophilus HB8: donor and acceptor specificities

FamUy 77 of glycosyl hydrolases consists of a single group of enzymes; 4-α- glucanotransferases (EC 2.4.1.25, amylomaltase (AMase) or D-enzyme). AMase is found in prokaryots and promotes metabohsm of starch degradation products inside the ceU as shown for Escherichia coli. In other organisms, lacking other enzymes required for growth on ohgosaccharides (p.e. maltodextrin phosphorylase), it may be involved in glycogen metabohsm as suggested for Aquifex aeolicus. D-enzyme is found in plants and is reported to be involved in in starch metabohsm. Recent studies on Chlamydomonas rheinhardtii show that D-enzyme is essential for biosynthesis of starch. Sequence comparisons and 3-D structure sir larities show that AMase is closely related to the a-amylase famUy or famUy 13 of glycosyl hydrolases. The a- amylase family is a very diverse group of enzymes that have the ability to modify and degrade starch. In the past, many 3D structures of enzymes from the a-amylase family have been elucidated, showing that aU members share an (alpha/beta)₈-barrel architecture of the catalytic domain, containing a conserved active site that comprises seven amino acid residues. For this reason, it is thought that aU members of the a- amylase famUy catalyze the same reaction cycle. This is suggested to proceed according to a two-step a-retaining mechanism. In the first step an a-glycosidic bond is cleaved in the substrate and a covalently bound enzyme -glycosyl intermediate is formed. In the second step, the leaving group is exchanged for an acceptor molecule, which is then linked via a new a-glycosidic bond to the intermediate.

Recently, amylomaltases from thermophile organisms like Thermus aquaticus and thermophilus HB8 have been isolated. These enzymes have a high thermostabihty, which makes them suitable for industrial apphcations, such as the production of large cyclic glucans and the production of thermoreversible gels from starch. A 2.0 A 3D structure of the amylomaltase from Thermus aquaticus shows that the enzyme consists of a compact (alpha/beta)s-barrel catalytic domain with three loop excursions that are probably responsible for part of the enzyme's specificity. In the catalytic site, 6 out of the 7 conserved residues of the a-amylase famUy are present, showing the close relatedness between amylomaltase and the alpha-amylase famUy.

Here we describe the cloning and characterization of the T. thermophilus AMase. Further glycosyl hydrolase families 13 and 77 are compared regarding reaction (mechanism and) specificity.

EXPERIMENTAL PROCEDURES

Escherichia coli TOP10 was used for recombinant DNA manipulations. AMase (mutant) proteins were produced with E. coli BL21 (DE3).

DNA manipulations - Restriction endonucleases and DNA polymerase were purchased from Pharmacia LKB Biotechnology, Sweden, and used according to the manufacturer's instructions. DNA manipulations and calcium chloride transformation of E. coli strains were as described {350}. Cloning and expression of the T thermophilus MalQ gene - A T. thermophilus gene library was constructed by inserting the 4-8 kb fragments of a partial Sau3A digest of genomic DNA in the BamHI site of pZerO. This construct was transformed to E. coli TOP10 ceUs and plated on LB agar plates. After replicaplating the transformants were screened for amylomaltase activity by overlaying the motherplate with 5 ml of a 0.5 % soluble starch solution, incubating for 24 h at 70°C, and staining with 4 ml Lugol solution. Positive colonies showed a shift from blue to red staining due to the disproportionation of the starch chains by amylomaltase. The DNA sequence of one of these clones was determined using the dideoxynuleotide chain termination method on a cycle sequencer (Pharmacia) The malQ gene was amphfied with PCR using the foUowing primers:

Forward: GGCAGCCATATGGAGCTTCCCCGCGCTTTCGG Reverse: GCAGCCAGATCTAGAGCCGTTCCGTGGCCTCGGC

The PCR product was digested with Ndel (CATATG) and Bglll (AGATCT, overhang compatible with BamHI) and ligated with either plasmid pET9c or plasmid pETlδb digested with Ndel and BamHI, resulting in pGJ6002 or pCCBmalQ, respectively. Transformation of these plasmids resulted in expression of the native enzyme (pGJ6002) or of the amylomaltase with an Ν-terminal Hisβ-tag (pCCBmalQ).

Production and purification of AMase - For the production of AMase protein E. coli BL21(DE3), containing the pCCBmalQ vector, was grown overnight in a 1 1 flask with 250 ml LB medium containing ampicilin.

Protein determination - Protein concentrations were determined with the Bradford method {63} using the Bio-Rad reagent and bovine serum albumin as a standard (Bio-Rad Laboratories, Richmond, CA, USA). Enzyme assays - AU assays were performed in a 25 mM sodium maleate buffer

(pH 6.5) at 70 °C.

Disproportionation reaction - Disproportionation activities were determineded using the ability of AMase to release glucose from ohgosaccharides. Various concentrations (upto 50 M) of (mixtures of) ohgosaccharides (G2-G7) were incubated with appropriatly dUuted enzyme. For the determination of donor specificity different concentrations of maltooligosaccharides as donor and methyl-α- D-glucose as acceptor.At regular time intervals 50 U samples were taken and added to 200 il GOD-PAP reagent (Roche) to measure the amount of glucose released.

Hydrolyzing activities were measured as described earher using 1% soluble starch (Lamers & Pleuger, Belgium) as substrate and dinitrosalicyhc acid to determine the number of reducing ends {680}.

In above assays 1 U of activity is defined as the amount of enzyme required for the processing of 1 :mole of donor substrate per minute. Kinetic parameters were fitted using the computer program Sigma Plot (Jandel Scientific). Product formation from oligosaccharides was analyzed by HPLC. For this purpose 1 ml of a 25 mM G3, G5, or G7 solution was incubated with 0.1 U AMase at 70 °C for 8 h. Samples were taken at regular time intervals and the products formed were applied to a 25 cm Econosphere-NH₂ 5 micron column (AUtech Associates Inc. USA) eluted with acetonitrUe/water (60/40, v/v) at a flow rate of 1 ml per min.

In the assay for the disproportionation reaction various oligosaccharides (G2- G7) were used as single (donor and acceptor) substrate. The KM and Vmax values for the formation of glucose varied with the different ohgosaccharides. The highest Vmax is observed for G4, which also shows the highest affinity. No activity on G2 was observed. Adding G3 to the G5 reaction mixture resulted in a further increase in activity (Fig. 2), whereas the addition of G2 had no effect (not shown). At high G3 concentrations a decrease in activity is observed, indicating competition between G3 and G5.

The donor specificity of AMase was further investigated using the various ohgosaccharides as donor and M-α-DG as acceptor substrates. Fig 3 shows that the addition of this monosaccharide clearly affects disproportionation activities, especiaUy with the lower concentrations of ohgosaccharide (donor) substrates. At higher M-α-DG concentrations the monosaccharide has an inhibitory effect. At lower concentrations, however, it can efficiently be used as acceptor, aUowing a clearer determination of affinities of the different ohgosaccharides for the donor binding site. G5, showing the lowest KM value, clearly is the best donor substrate. Combined with the above observation of the stimulating effect of G3 on disproportionation of G5, this suggests that G3 is a better acceptor substrate. Hydrolyzing activity on soluble starch was investigated, but even overnight incubation did not result in an increase of reducing ends, thus no hydrolyzing reaction is performed by the enzyme. Furthermore the enzyme was incubated with 4- nitrophenyl-α-D-maltoheptaoside-4-6-O-ethylidene (EPS; Boehringer Mannheim) (a maltoheptasaccharide which is blocked at the non-reducing end and with a para- nitrophenyl group at its reducing end). This compound is generaUy used for the detection of α-amylase activity. However also with this substrate no hydrolyzing activity was observed. Furthermore it reacted very weakly when accepting ohgosaccharides were added, suggesting that amylomaltase is an exo-acting enzyme, requiring the presence of a non-reducing end glucose. The oligosaccharide formation of AMase was analyzed with HPLC (Fig. 4).

With G3 as substrate the initial products were Gl and G5 (see Fig 4. a). After an initial lagfase, the production of Gl increased, whUe G3 decreased, however with a less significant increase of G5. Various larger oligosaccharides are produced, indicating that the initial product (G5) is used as donor and G3 is mainly used as acceptor. With G5 as substrate the predominant initial products were G3 and G7, although also considerable amounts of Gl, G4, G6, and G9 were formed. In both cases httle maltose is produced initiaUy, as observed previously (T. aquaticus, potato). The final production of maltose is probably caused by the tranfer of glucose from the donor to a glucose acceptor. This supported by the early formation of G4 from G5, which indicates transfer of a glucose moyety, and by the above results with MαDG, which indicate that glucose can indeed be used as acceptor.

AMase is the ultimate disproportionating enzyme, producing a variety of (long) ohgosaccharides from short substrates. In the disproportionation reaction the enzyme has a preference for longer oligosaccharides to be used as donor whUe shorter ohgosaccharides except maltose are efficiently used as acceptor. One of the requirements of doing this efficiently is a low hydrolyzing activity, which is extremely weU met in AMase. The complete lack of hydrolyzing activity of this enzyme makes it a very interesting enzyme to be studied regarding reaction specificity in the α- amylase famUy. Reaction kinetics of the disportionation of ohgosaccharides

DP Km Vmax

2 nd nd

3 10 90

4 3.4 281

5 4.6 235

6 3.5 143

7 4.5 108

nd = not detectable

Affinity constants for ohgosaccharides using MαDG as acceptor

DP O mM 4 mM 10 mM

3 8.0 4.7 12.2

4 3.4 3.0 2.9

5 4.6 1.5 2.7

6 3.5 2.5 2.8

Kinetic analysis of amylomaltase from Thermus thermophilus HB8: donor and acceptor specificities

Family 77 of glycosyl hydrolases consists of a single group of enzymes; 4-α- glucanotransferases (EC 2.4.1.25, amylomaltase (AMase) or D-enzyme). AMase is found in prokaryots and promotes metabohsm of starch degradation products inside the ceU as shown for Escherichia coli. In other organisms, lacking other enzymes required for growth on ohgosaccharides (p.e. maltodextrin phosphorylase), it may be involved in glycogen metabohsm as suggested for Aquifex aeolicus. D-enzyme is found in plants and is reported to be involved in in starch metabohsm. Recent studies on Chlamydomonas rheinhardtii show that D-enzyme is essential for biosynthesis of starch.

Sequence comparisons and 3-D structure simUarities show that AMase is closely related to the a-amylase famUy or famUy 13 of glycosyl hydrolases. The a- amylase family is a very diverse group of enzymes that have the ability to modify and degrade starch. In the past, many 3D structures of enzymes from the a-amylase famUy have been elucidated, showing that aU members share an (alpha beta)s-barrel architecture of the catalytic domain, containing a conserved active site that comprises seven amino acid residues. For this reason, it is thought that aU members of the a- amylase family catalyze the same reaction cycle. This is suggested to proceed according to a two-step a-retaining mechanism. In the first step an a-glycosidic bond is cleaved in the substrate and a covalently bound enzyme-glycosyl intermediate is formed. In the second step, the leaving group is exchanged for an acceptor molecule, which is then hnked via a new a-glycosidic bond to the intermediate.

Recently, amylomaltases from thermophUe organisms like Thermus aquaticus and Thermus thermophilus HB8 have been isolated. These enzymes have a high thermostabihty, which makes them suitable for industrial applications, such as the production of large cyclic glucans and the production of thermoreversible gels from starch. A 2.0 A 3D structure of the amylomaltase from Thermus aquaticus shows that the enzyme consists of a compact (alpha/beta)₈-barrel catalytic domain with three loop excursions that are probably responsible for part of the enzyme's specificity. In the catalytic site, 6 out of the 7 conserved residues of the a-amylase famUy are present, showing the close relatedness between amylomaltase and the alpha-amylase family.

Here we describe the cloning and characterization of the T. thermophilus AMase. Further glycosyl hydrolase families 13 and 77 are compared regarding reaction (mechanism and) specificity. EXPERIMENTAL PROCEDURES

Escherichia coli TOP10 was used for recombinant DNA manipulations. AMase (mutant) proteins were produced with E. coli BL21 (DE3). NA manipulations - Restriction endonucleases and DΝA. polymerase were purchased from Pharmacia LKB Biotechnology, Sweden, and used according to the manufacturer's instructions. DΝA manipulations and calcium chloride transformation of E. coli strains were as described {350}.

Cloning and expression of the T thermophilus MalQ gene - A T. thermophilus gene library was constructed by inserting the 4-8 kb fragments of a partial Sαu3A digest of genomic DΝA in the BamHI site of pZerO. This construct was transformed to E. coli TOP10 ceUs and plated on LB agar plates. After rephcaplating the transformants were screened for amylomaltase activity by overlaying the motherplate with 5 ml of a 0.5 % soluble starch solution, incubating for 24 h at 70°C, and staining with 4 ml Lugol solution. Positive colonies showed a shift from blue to red staining due to the disproportionation of the starch chains by amylomaltase. The DΝA sequence of one of these clones was determined using the dideoxynuleotide chain termination method on a cycle sequencer (Pharmacia)

The malQ gene was amplified with PCR using the foUowing primers: Forward: GGCAGCCATATGGAGCTTCCCCGCGCTTTCGG Reverse: GCAGCCAGATCTAGAGCCGTTCCGTGGCCTCGGC

The PCR product was digested with Ndel (CATATG) and Bglll (AGATCT. overhang compatible with BamHI) and hgated with either plasmid pET9c or plasmid pETlδb digested with Ndel and BamHI, resulting in pGJ6002 or pCCBmalQ, respectively. Transformation of these plasmids resulted in expression of the native enzyme (pGJ6002) or of the amylomaltase with an Ν-terminal Hisβ-tag (pCCBmalQ).

Production and purification of AMase - For the production of AMase protein E. coli BL21(DE3), containing the pCCBmalQ vector, was grown overnight in a 1 1 flask with 250 ml LB medium containing ampicilin.

Protein determination - Protein concentrations were determined with the Bradford method {63} using the Bio-Rad reagent and bovine serum albumin as a standard (Bio-Rad Laboratories, Richmond, CA, USA). Enzyme assays - AU assays were performed in a 25 M sodium maleate buffer (pH 6.5) at 70 °C.

Disproportionation inaction - Disproportionation activities were determineded using the ability of AMase to release glucose from ohgosaccharides. Various concentrations (upto 50 mM) of (mixtures of) ohgosaccharides (G2-G7) were incubated with appropriatly diluted enzyme. For the determination of donor specificity different concentrations of maltooligosaccharid.es as donor and methyl-α- D-glucose as acceptor.At regular time intervals 50 U samples were taken and added to 200 U GOD-PAP reagent (Roche) to measure the amount of glucose released. Hydrolyzing activities were measured as described earher using 1% soluble starch (Lamers & Pleuger, Belgium) as substrate and dinitrosahcyhc acid to determine the number of reducing ends {680}.

In above assays 1 U of activity is defined as the amount of enzyme required for the processing of 1 :mole of donor substrate per minute. Kinetic parameters were fitted using the computer program Sigma Plot (Jandel Scientific).

Product formation from oligosaccharides was analyzed by HPLC. For this purpose 1 ml of a 25 mM G3, G5, or G7 solution was incubated with 0.1 U AMase at 70 °C for 8 h. Samples were taken at regular time intervals and the products formed were apphed to a 25 cm Econosphere-NH₂ 5 micron column (AUtech Associates Inc. USA) eluted with acetonitrUe/water (60/40, v/v) at a flow rate of 1 ml per min.

In the assay for the disproportionation reaction various ohgosaccharides (G2- G7) were used as single (donor and acceptor) substrate. The KM and Vmax values for the formation of glucose varied with the different ohgosaccharides. The highest Vmax is observed for G4, which also shows the highest affinity. No activity on G2 was observed. Adding G3 to the G5 reaction mixture resulted in a further increase in activity (Fig. 2), whereas the addition of G2 had no effect (not shown). At high G3 concentrations a decrease in activity is observed, indicating competition between G3 and G5. The donor specificity of AMase was further investigated using the various ohgosaccharides as donor and M-α-DG as acceptor substrates. Fig 3 shows that the addition of this monosaccharide clearly affects disproportionation activities, especiaUy with the lower concentrations of oligosaccharide (donor) substrates. At higher M-α-DG concentrations the monosaccharide has an inhibitory effect. At lower concentrations, however, it can efficiently be used as acceptor, aUowing a clearer determination of affinities of the different ohgosaccharides for the donor binding site. G5, showing the lowest KM value, clearly is the best donor substrate. Combined with the above observation of the stimulating effect of G3 on disproportionation of G5, this suggests that G3 is a better acceptor substrate.

Hydrolyzing activity on soluble starch was investigated, but even overnight incubation did not result in an increase of reducing ends, thus no hydrolyzing reaction is performed by the enzyme. Furthermore the enzyme was incubated with 4- nitrophenyl-α-D-maltoheptaoside-4-6-O-ethylidene (EPS; Boehringer Mannheim) (a maltoheptasaccharide which is blocked at the non-reducing end and with a para- nitrophenyl group at its reducing end). This compound is generally used for the detection of α-amylase activity. However also with this substrate no hydrolyzing activity was observed. Furthermore it reacted very weakly when accepting oligosaccharides were added, suggesting that amylomaltase is an exo-acting enzyme, requiring the presence of a non-reducing end glucose.

The oligosaccharide formation of AMase was analyzed with HPLC. With G3 as substrate the initial products were Gl and G5 After an initial lagfase, the production of Gl increased, whUe G3 decreased, however with a less significant increase of G5. Various larger ohgosaccharides are produced, indicating that the initial product (G5) is used as donor and G3 is mainly used as acceptor. With G5 as substrate the predominant initial products were G3 and G7, although also considerable amounts of Gl, G4, G6, and G9 were formed. In both cases httle maltose is produced initiaUy, as observed previously (T. aquaticus, potato). The final production of maltose is probably caused by the tranfer of glucose from the donor to a glucose acceptor. This supported by the early formation of G4 from G5, which indicates transfer of a glucose moyety, and by the above results with MαDG, which indicate that glucose can indeed be used as acceptor.

AMase is the ultimate disproportionating enzyme, producing a variety of (long) oligosaccharides from short substrates. In the disproportionation reaction the enzyme has a preference for longer oligosaccharides to be used as donor whUe shorter ohgosaccharides except maltose are efficiently used as acceptor. One of the requirements of doing this efficiently is a low hydrolyzing activity, which is extremely weU met in AMase. The complete or near complete lack of hydrolyzing activity of this enzyme makes it a very interesting enzyme to be studied regarding reaction specificity in the α-amylase famUy.

Reaction kinetics of the disportionation of ohgosaccharides

DP Km Vmax

2 nd nd

3 10 90

4 3.4 281

5 4.6 235

6 3.5 143

7 4.5 108

nd = not detectable

Affinity constants for ohgosaccharides using MαDG as acceptor

DP O mM 4 mM 10 mM

3 8.0 4.7 12.2

4 3.4 3.0 2.9

5 4.6 1.5 2.7

6 3.5 2.5 2.8

STRUCTURES OF THE THERMOSTABLE AMYLOMALTASE FROM THERMUS THERMOPHILUS HB8 IN TWO DIFFERENT SPACE GROUPS

Enzymes from the alpha-amylase family, or glycosyl hydrolase famUy 13, are a very diverse group of starch-converting enzymes, which have a common architecture of their catalytic site. Many enzymes from the alpha-amylase famUy are used in industrial starch processing, and many have been structuraUy characterized with the aim of improving them for specific apphcations. Because of a lack of sufficient homology to allow Molecular Replacement, the phase problem for most of these structures has been solved by using multiple isomorphous replacement (MIR) or multi wavelength anomalous dispersion (MAD) approaches.

Here we show that the phase problem in the alpha-amylase famUy can be solved by using six 'high potential heavy atom compounds that bind to conserved elements in the family. The effectiveness of this strategy was demonstrated by the elucidaton of the structure of the amylomaltase from Thermus thermophilus HB8, which is the most divergent member of the alpha-amylase family.

The structure of the amylomaltase from Thermus thermophilus HB8 was solved in space groups P2ι2ι2 and C2, whereas the highly (99.8%) identical amylomaltase from Thermus aquaticus was solved earher in space group P6 . A comparison of these three structures shows that the core of the enzyme is highly rigid, whereas some loops around the catalytic site can vary in conformation.

1. Introduction

The alpha-amylase family is a very diverse group of enzymes that have the abUity to modify and degrade starch. Some well-known members of this famUy, such as bacterial alpha-amylases, cyclodextrin glycosyltransferase, and iso-amylase are used in industrial starch processing. Other enzymes, such as human sahvary and pancreatic alpha-amylases are therapeutic targets in the treatment of diabetes, whereas insect alpha-amylases are useful targets in the development of crop protectants. In the past, many 3D structures of enzymes from the alpha-amylase famUy have been elucidated (Table 1), showing that all members share an (alpha/beta)s-barrel architecture of the catalytic domain, in which a conserved active site is that comprises seven amino acid residues. For this reason, it is thought that aU members of the alpha-amylase famUy catalyze the same reaction cycle. This is suggested to proceed according to a two-step a-retaining mechanism. In the first step an a-glycosidic bond is cleaved in the substrate and a covalently bound enzyme- glycosyl intermediate is formed. In the second step, the leaving group is exchanged for an acceptor molecule, which is then hnked via a new a-glycosidic bond to the intermediate.

The most divergent member of the alpha-amylase famUy is, on basis of sequence comparisons, the enzyme amylomaltase. Amylomaltase is a 57 kDa intraceUular enzyme that is also known as 4a-glucanotransferase in bacteria and D-enzyme in plants. Investigations with Escherichia coli have estabhshed that amylomaltase is the product of the MalQ gene and is essential for the growth on maltose. Presumably, the function of the enzyme is to synthesize long amylose-like ohgosaccharides from shorter ohgosaccharides, which can then be further catabohzed. This synthesizing capacity of amylomaltase is probably related to the enzyme's high transglycosylation activity. This forms an interesting contrast with the activity of 'classical' alpha- amylases that degrade starch and mainly perform hydrolysis.

Recently, amylomaltases from thermophUe organisms like Thermus aquaticus and Thermus thermophilus HB8 have been isolated. These enzymes have a high thermostabihty, which makes them suitable for industrial apphcations, such as the production of large cychc glucans and the production of thermoreversible gels from starch. A 2.0 A 3D structure of the amylomaltase from Thermus aquaticus shows that the enzyme consists of a compact (alphalpha/betaeta)₈-barrel catalytic domain with three loop excursions that are responsible for part of the enzyme's specificity. In the catalytic site, 6 out of the 7 conserved residues of the alpha-amylase famUy are present, establishing amylomaltase as a member of the alpha-amylase famUy. We have investigated two 3D structures of the amylomaltase from Thermus thermophilus HB8 (TTHB8), determined from data in space groups P2ι2ι2 and C2 to 2.3 A and 3.1 A, respectively. The TTHB8 enzyme has a 99.8% sequence identity to the amylomaltase from Thermus aquaticus, which crystallizes in space group P6₄. However, the structures of the TTHB8 enzyme were solved independently using a MIRAS strategy with general applicability for alpha-amylase enzymes. 2. Materials and methods

2.1 Crystallization and data coUection

The amylomaltase from Thermus thermophilus HB8 was cloned and expressed in E. coli, and purified by a series of standard chromatographic steps untU aU heterogeneities had dissappeared as judged from sUver-stained SDS page. The isolation and characterization of the enzyme wUl be described in detaU in another publication. For crystallization an enzyme preparation in 25 mM Tris-HCl, pH 7.5 was used, which was concentrated to 2.5 mg/ml using a FUtron 30K system. The TTHB8 amylomaltase was crystallized at 293 K with the hanging drop vapor diffusion method, using a reservoir solution of 12% (w/v) PEG 20000 and 100 mM MES (2- [N- morpholino] ethanesulfonic acid) buffer at pH 6.8. Crystals appeared after five days, in the form of very thin plates with dimensions 0.15 x 0.15 x 0.04 mm³. Prior to data coUection, they were frozen in a cryo-mother hquor consisting of 25% (v/v) glycerol, 10% (w/v) PEG 20000 and 100 mM MES buffer at pH 6.8.

Due to the small dimensions of these crystals, the diffraction of the amylomaltase crystals on a rotating anode source was hmited to 8 A resolution. However, by using synchrotron radiation, a complete dataset could be obtained to 2.3 A resolution. The intensity distribution of the data was very anisotropic, most likely because of the non- uniform dimensions of the crystals. The space group of the crystals was P2ι2ι2, with cell dimensions a=115.2 A, b=93.7 A, c=53.5 A.

In addition to this crystal form, smaU microneedles (0.04 x 0.04 x 0.20 mm³) were found growing in hanging drops at 12% (w/v) PEG 20000 and 100 mM maleate at pH 6.8 and 0.1% (w/v) maltotriose. These crystals were frozen by transferring them to an identical solution to which 20% (v/v) glycerol was added, and subsequently dipping them in liquid nitrogen. The frozen crystals were exposed to synchrotron radiation and belong to space group C2, with ceU dimensions a=104.9 A, b=52.5 A, c=104.9 A, and 3=90°^= 96.4°, and g=90°. Unfortunately, their diffraction was hmited to 3.1 A, therefore we performed further soaking experiments with the better- diffracting P2ι2ι2 crystal form.

2.2 Phasing

Because sequence comparisons suggest that amylomaltase is a member of the alpha-amylase famUy, we initially attempted to solve the phase problem for amylomaltase by Molecular Replacement using a poly-alanine TIM barrel domain as search model. Several models were tried, originating from cyclodextrin glycosyltransferase (CGTase) and Aspergillus oryzae (Taka) alpha-amylase, but aU attempts faUed. This is not surprising since simUar Molecular Replacement attempts were also problematic in cases in the alpha-amylase famUy where model and target had much more structural homology. Therefore, as a next strategy, we decided to use ab initio phasing with multiple isomorphous replacement combined with anomalous scattering (MIRAS).

To determine a suitable MIRAS strategy, an overview was made of the compounds that were used in the past to solve structures of enzymes from the alpha-amylase famUy. It appears that many structures have been solved using the same heavy atom compounds. Out of 14 cases, a HgCl₂ derivative was useful 8 times, a K2PtCl derivative 7 times, a UO2²" derivative 6 times and a Sm³⁺ derivative 3 times. This suggests that these compounds bind to conserved features in alpha-amylase famUy enzymes and thus would have general applicabUity within the family.

To check this hypothesis and to solve the structure of the TTHB8 amylomaltase, we used these compounds to soak crystals and collected data at the EMBL beamline BW7B at DESY, Hamburg and the EMBL beamhne ID 14-3 of the ESRF, Grenoble (Table 3). Despite non-isomorphism in the length of the longest ceU axis, aU the four above-mentioned compounds turned out to be useful derivatives. In addition we found an ethylmercury phosphate derivative.

From these data, heavy atom sites were located using the program Solve and subsequently refined with the program Sharp. Solvent flattening resulted in an experimental electron density map in which secondary structure elements were weU discernible. Model buUding was was performed with the program O. To facilitate model building, we used the sequence and structure of amylomaltase from Thermus aquaticus as a template.

2.3 Refinement of the P2ι2ι2 crystal form Our initial model was refined against our best data, those from a HgCl₂ soak which diffracted to 2.3 A. Refinement was performed using the program CNS version 1.0. After initial rigid body refinement, fuU coordinate refinement, grouped B-factor refinement and individual atomic B-factor refinement against the CNS maximum likelihood target were applied. Solvent molecules were placed at peaks of at least 3.0 s in Fo-Fc difference electron density maps, at positions where they could form at least one hydrogen bond. This was done using in combination with refinement using the iterative procedure implemented in CNS. Manual rebuUding was done in SA-weighted Fo-Fc, 2Fo-F_c and OMIT F₀-F_c,and 2F₀-F_C maps, calculated with CNS. During rebuUding, a very strong peak close to a smaU peak in an F₀-F_c difference electron density was observed in the active site. A peak in a simUar position was observed in an anomalous difference map (F₀ ⁺-Fo", where + ^"and - reflections are Bijvoet mates) from the HgCk data. Moreover, the program Sharp had interpreted the position of this peak as a heavy atom binding site. From this we concluded that a HgC ion bound in the active site should be included in our model.

The stereochemistry of the final model was checked with the programs Procheck and Whatcheck. The final model contains no residues in disaUowed regions of the Ramachandran plot, in contrast to the structure of the Thermus aquaticus enzyme. The atomic coordinates and structure factors have been deposited at the Protein Data Bank (www.rcsb.org, code 1FP8).

2.4 Refinement of the C2 crystal form

In order to study the influence of crystal contacts on the conformation of the enzyme, we also determined the 3.1 A structure of TTHB8 amylomaltase in the maltotriose-dependent C2 crystal form. An initial model was obtained from the structure in P2ι2ι2, by Molecular Replacement with the program AMoRe. This model was refined using CNS as outlined above. The final refinement step consisted of a few rounds of individual B-factor refinement, which was stopped after the free R- factor started to increase. No solvent molecules were incorporated. Although the crystals were grown in the presence of sugars, we found no evidence for the presence of maltotriose or any other oligosaccharide in the electron density maps. Final model statistics, coordinates and structure factors have been deposited at the Protein Data Bank (www.rcsb.org, code 1FP9).

2.5 Binding locations of heavy atom hgands

To solve the phase problem for amylomaltase we used heavy atom compound with a high success rate in the alpha-amylase famUy, under the assumption that they bind to conserved features within the family. To check whether this is true, we investigated their location using anomalous difference electron density maps computed with phases from refined models. As indicated above, the HgC soak resulted in a HgCh ion bound in the conserved catalytic site of the alpha-amylase famUy. At that position, the C atom binds to Tyr 59 with a typical hahde-aryl interaction, whereas the Hg²⁺ atom is bound by the conserved acidic residues Asp 395, Glu 340 and Asp 293.

In addition to HgCk, the ethylmercury phosphate soak also resulted in an active site complex in which an Hg²⁺ moiety is bound by acidic residues. However, the other soaks (UO2AC2, K2PtCl4, SmCta) resulted in heavy atoms bound in non-conserved regions (Table 3). This contrasts with other reports. UO2AC2 was observed to bind in the active site of the CGTase from Bacillus circulans strain 251, and UO₂(NO3)2 was observed in the active site of Taka alpha-amylase. K^PtCU was observed to bind close to the catalytic site in the CGTase from Bacillus circulans strain 8, near residue His 233 in the sugar binding subsite +1. In Taka alpha-amylase, K₂PtCl4 was observed to bind close to the catalytic site. To explain this discrepancy, we suggest that the binding of UO2AC2 andK2PtCl₄ in the active site of amylomaltase is hindered by the presence of a low concentration of HgCl₂ that was apphed to stabihze the crystals. The HgCb might compete with the other compounds for binding.

Thus, we show that there exist 'high potential' compounds, which are much more succesful than average in forming heavy atom derivatives of a crystallized alpha- amylase famUy enzymes. Most of these compounds were reported to bind in the conserved catalytic site, though this could not always be reproduced for TTHB8 amylomaltase. Nevertheless, it was shown that with these compounds, the phase problem for alpha-amylase-famUy enzymes can be quickly and efficiently solved.

3. Results

3.1 Secondary structure

The three-dimensional structure of the amylomaltase from Thermus thermophilus HB8 in its P2ι2ι2 crystal form is depicted in Figure 1. It is simUar to the Thermus aquaticus amylomaltase and consists of a central (alpha/beta)₈ or TIM-barrel domain from which three other smaU domains protrude. Although the (alpha/beta)₈-barrel domain is a feature that is shared by aU enzymes from the alpha-amylase, a superposition of the (alpha/beta)β-barrel domain in amylomaltase with those from cyclodextrin glycosyltransferase (CGTase) and Taka alpha-amylase shows large differences in the position, length and orientation of the a-helices that surround the central b-barrel. These differences explain the difficulty of solving the phase problem by using TIM-barrels from alpha-amylase famUy enzymes as templates for a Molecular Replacement search. Moreover, they shows that the folding pattern of the (alpha/beta)₈-barrel is more conserved than the precise three-dimensional orientation of its constituent secondary structure elements.

From the central (alpha/beta)s-barrel domain in amylomaltase three subdomains protrude that are labeUed Bl, B2 and B3. Subdomain B2 comprises residues 68 to 179 and protrudes at the third beta-strand of the TIM barrel, which makes this subdomain the structural homolog of domain B in CGTases and alpha-amylases. Subdomain Bl comprises residues 222 to 272 and 294 to 320, and subdomain B3 comprises residues 398 to 427. Both these latter domains are unique to amylomaltase.

3.2 The active site of amylomaltase

Another determinant of alpha-amylase family membership is the presence of seven conserved residues in the catalytic site in a characteristic orientation. The catalytic site of the TTHB8 amylomaltase is compared with the catalytic site of CGTase, a representative member of the alpha-amylase famUy. It appears that the nucleophihc catalytic residue Asp 229 in CGTase, and the acid/base catalyst Glu 257 have amylomaltase equivalents in Asp 293 and Glu 340. Residues Arg 227, His 327 and Tyr 100, which are important in strabilization of the transition state and the covalent intermediate have equivalents in amylomaltase in Arg 291, His 394 and Tyr 59, respectively. Interestingly however, of two residues in CGTase that are important for distortion of a bound substrate, Asp 328 and His 140, only Asp 328 has an equivalent in amylomaltase in Asp 293, whereas the position of His 140 is taken by Asn 260. In this respect, amylomaltase is different from aU other members of the alpha-amylase family. Interestingly, when His 140 is replaced in CGTase or alpha-amylase, the activity decreases 50-100x times. However, amylomaltase has an optimal enzymatic rate that is comparable to that of other alpha-amylases. This might indicate that amylomaltase has found a way of compensating for the absence of a His 140 equivalent by an unknown mechanism. 3.3 Putative sugar binding sites

In addition to the catalytic site, amylomaltase possesses at least seven sugar binding subsites that assist in substrate processing. We attempted to identify these sugar binding subsites by a crystal-soaking procedure, in which P2ι2ι2 crystals of TTHB8 amylomaltase were subjected to a stabilizing solution containing the oligosaccharide inhibitor acarbose. This inhibitor is known to bind strongly in the catalytic site of alpha-amylase-fam y enzymes and in adjacent sugar binding subsites. Unfortunately, after subsequent data coUection on these crystals, this inhibitor could not be observed in the electron density, and therefore had not bound inside the crystals. Probably, the active site of amylomaltase in its P2ι2ι2-crystaUine form is not accessible to ohgosaccharide binding.

To nevertheless estimate the location of extra sugar binding subsites, we constructed a model of sugar binding. We superimposed the 3D structure of a maltohexaose inhibitor in complex with Porcine pancreatic alpha-amylase on amylomaltase on basis of the conserved active site in both enzymes. The torsion angles of the glycosidic bonds in the maltohexaose inhibitor were subsequently adjusted to improve its fit in the active site of amylomaltase. This remodelling was aided by comparisons with the conformations of other oligo-saceharides in comple with alpha-amylase famUy enzymes, such as maltononaose bound to CGTase. The final model is schematicaUy drawn in Figure 3, and is the first detaUed model of how amylomaltase might bind an ohgosaccharide, and is provides the guidance needed for site-directed mutagenesis experiments that alter the properties of amylomaltase in a desired fashion.

3.4 Crystal contacts in the P2ι2ι2 and C2 crystal forms

The structure of TTHB8 amylomaltase was determined to high resolution in a P2ι2ι2 crystal form, and to lower resolution in a C2 crystal form. This aUows us to establish whether the conformation of amylomaltase is influenced by the crystaUine packing of the molecules. In the P2ι2ι2 form, crystal contacts are formed in three regions. In the first a loop of residues Gly 149-Gly 153 is grabbed by residues Gly 422-Arg 426 and the C- terminus Ala 492-Leu 500. A second, weaker contact is formed between residues Gin 27-Glu 38 and the two stretches Glu 313-:Lys 318, Gly 343-Val 349. Interestingly, a third contact is formed by the only cysteine residue in amylomaltase. Cys 308 is at a distance of 5.2 A of a Cys 308 of another amylomaltase molecule inside the crystal. This suggests that one crystal contact is formed by an intramolecular disulphide bond. The possibility of a disulphide-hnked crystal contact is corroborated by the electron density at this location, which suggest the

(partial) presence of a disulphide bond. This suggests that the crystal lattice consists of a mixture of disulphide-bonded dimers and monomeric units. Dynamic hght scattering experiments with our sample support the presence of a smaU amount of dimers (results not shown) mixed with monomers. It is not unlikely that the dimeric impurities enforce the presence of a disulphide-bonded crystal contact, inhibiting the formation of other (stronger) contacts, such as present for example in the P6₄ crystal form of the Thermus aquaticus amylomaltase. The crystals in space group C2 show a similar disulphide-linked crystal contact. However, at the other crystal contacts, there are significant differences. The stretch of residues Lys 148-Glu 173 binds to Pro 378-Gly 385, and the residues

Gly 26-Asp 31 and Leu 74- Gly 89 bind to Gly 26-Asp 31 and Leu 74- Gly 89 in another molecule. Due to these differences, the C2 crystal form can be regarded as independent from the P2ι2ι2 crystal form.

3.5 Comparison of amylomaltases from T aquaticus and T. thermophilus HB8

We have determined two structures of the amylomaltase from Thermus thermophilus HB8. Earher, the 3D structure of the amylomaltase from Thermus aquaticus was determined, which has a sequence identity to the TTHB8 amylomaltase of 99.8%. Only Gin 27 and Leu 154 in the TTHB8 enzyme have been substituted by Arg 27 and Pro 154 in the Thermus aquaticus enzyme. Strangely, the pubhshed amino acid sequence of the Thermus aquaticus enzyme, which gives Pro 154, does not correspond to the sequence derived from the 3D structure of the Thermus aquaticus amylomaltase, which gives Leu 154. If this is interpreted as a correction on a sequencing error, both amylomaltases only differ in amino acid sequence at position 27. Therefore, for aU practical purposes these structures can be regarded as independently solved structures of the same enzyme in different space groups. A comparison could reveal interesting areas of flexibility. 3.6 Conformational differences between the three structures of amylomaltase Since the structures of T. thermophilus HB8 amylomaltase in space groups C2 and

P2ι2ι2 and the structure of T aquaticus amylomaltase in space group P6₄ can be regarded as three structures of the same enzyme in different crystal packing environments, differences between these structures can show how crystal contacts influence the conformation of the enzyme, and in which areas it is very flexible or very rigid.

If we take the 'P2ι2ι2-structure' as basis, and superimpose the 'C2-structure', we observe that the position of most amino-acids is identical (r.m.s.d. 0.5 A). However, two loops in the active site cleft have a significantly different conformation. First, the loop that comprises residues Tyr 141 to Ala 170 has shifted in the C2-form towards the active site (maximally 1.5 A). Secondly, the loop of residues Val 242-Leu 262 (and its adjacent loop Tyr 301-Val 317), which cover the active site cleft, have shifted ~0.5 A outwards in the C2 form, thereby opening the cleft a httle. When the structure of T. aquaticus amylomaltase is superimposed on the 'P2ι2ι2- structure', this shows that they have an almost identical conformation (r.m.s.d. 0.4 A). Interestingly, also in the T. aquaticus enzyme the loop of residues 141-170 has a position that is oriented more toward the active site (maximum difference

1.3 A). This position resembles the conformation of this loop in the C2-crystal form.

The flexibility of amylomaltase was further studied through the atomic temperature factors. In general, aU three structures show a simUar temperature factor distribution, indicating only a marginal influence of crystal packing contacts. In aU cases amylomaltase appears to be rigid, with specific areas having higher temperature factors, and thus higher flexibility. These include four loop stretches near the catalytic site comprising residues 80-93, 114-125, 342-348 and most strongly 249-253. Thus, in general amylomaltase appears to have a rigid, weU-determined conformation, which might in part explain the enzyme's thermostabihty. However, when information on conformational variability and temperature factor distributions is combined, it appears that there are two interesting regions in the enzyme. The first is the loop 242-262 (comprising 249-253) that can have different conformations and is also very flexible (high B-factors). This loop incorporates residues Tyr 250 and Phe 251, which might be involved in substrate binding (Figure 3). The second is the loop 141-170, which is conformationaUy variable but has a very low temperature factor. Therefore, this loop is not flexible, but can 'switch' between two rigid conformations. As was observed for other alpha-amylase famUy enzymes, such conformational variations could play an important role in promoting catalysis.

Overview of heavy atom compounds used to solve 3D structures of alpha-amylase- famUy proteins

Enzyme method⁰ heavy atom compounds used

Animal alpha-amylases porcine pancreas MIE (2x)^a OCMP^b/K^PtCl4/K₂Hgl4/PbNθ3/HgAc2/U₂θ7 human salivary MR human pancreas MR yellow meal worm MR

Fungal and plant alpha-amylases

Taka (Aspergillus oryzae) MIR HgCl2/Uθ2(Nθ3)2/AgNθ3 K2PdCl4/K2PtCl4/K2Pt(CN)4/KAu(C3SI)2

Aspergillus niger MIR/MR HgCl2/SmAc3 K2PtCl₆/PbAc2

Barley MIR HgCl₂/Eu(N0₃)3/K2PtCl4 bacterial alpha-amylases

B. licheniformis MIR (2x) Uθ2Ac2/Pb(CH₃)3Ac/HgCl2/IC2PtCl4/K2PtCl6^a

B. subtilis MIR K2PtCl /HgCl2

P. stutzeri MIR KaUOzFβ/SmCh

Alteromonas haloplanctis MR cyclodextrin glycosyltransferases !

B. circulans strain 8 MIR K2PtCl4/cis-(NH₃)2PtCl2/U0₂C2θ4

B. stearothermophilus

B. sp. 1011 MR

B. circulans strain 251 SIRAS UO2AC2 other enzymes

B. cereus oligo-l,6-glucosidase HgCl2/Uθ2(N0₃)2/Sm(Nθ3)3

P. amyloderamosa iso-amylase MIR NaAuCU/HgCk

B. stearothermophilus MR maltogenic alpha-amylase

Thermoactinomyces υulgaris MIRAS PbAc2/C₂H₅Hg-

A47 alpha-amylase II

Thermus strain maltogenic MIR/MR Se-Met/PtC12(NH₃)2/HoCl2 alpha-amylase

Thermus aquaticus MIR PCMBS/HgCl2/K2PtCl4/EAu(CN)2/K2Pt(SCN)₆/Pb(CH3)3Ac amylomaltase ^aused in the most recent report. ^bOCMP means ortho-chloromercuriphenol. PCMBS means para-chloromercuriphenylsulfonic acid. ^CMIR(AS) means Multiple isomorphous replacement (with anomalous scattering), MR means Molecular Replacement, SIR(AS) means single isomorphous replacement (with anomalous scattering).

Data coUection and refinement statistics

^bBecause HgC appeared to stabUize the crystals, soakings with this compound was preceded by a soaking in HgC . The site labeUed 'b' is therefore probably a Hg²⁺ ion. ^CPP: Phasing Power. ^dBefore solvent flattening. ^eRegarding Bijvoet mates as separate reflections. Generation of mutants

FamUy 77 of glycosyl hydrolases consists of a single group of enzymes; 4- - glucanotransferases (EC 2.4.1.25, amylomaltase (AMase) or D-enzyme). AMase is found in prokaryots and promotes metabohsm of starch degradation products inside the ceU as shown for Escherichia coli. In other organisms, lacking other enzymes required for growth on oligosaccharides (p.e. maltodextrin phosphorylase), it may be involved in glycogen metabohsm as suggested for Aquifex aeolicus. D-enzyme is found in plants and is reported to be involved in in starch metabohsm. Recent studies on Chlamydomonas rheinhardtii show that D-enzyme is essential for biosynthesis of starch. In each case the role of AMase is based on its transglycosylating activity, which enables the enzyme to produce long ohgosaccharides from short chained substrates or transfer ohgosaccharides to branched polymers (glycogen, amylopectin). The synthesizing capacity of wid type amylomaltase is probably related to the enzyme's high transglycosylation activity and lack of hydrolyzing activity. This forms an interesting contrast with the activity of 'classical' a-amylases that degrade starch and mainly perform hydrolysis.

Sequence comparisons and 3-D structure simUarities show that AMase is closely related to the alpha-amylase famUy or famUy 13 of glycosyl hydrolases. The a- amylase famUy is a very diverse group of enzymes that have the ability to modify and degrade starch. In the past, many 3D structures of enzymes from the a-amylase family have been elucidated, showing that aU members share an (alpha/beta)s-barrel architecture of the catalytic domain, containing a conserved active site that comprises seven amino acid residues. For this reason, it is thought that aU members of the a- amylase famUy catalyze the same reaction cycle. This is suggested to proceed according to a two-step a-retaining mechanism. In the first step an a-glycosidic bond is cleaved in the substrate and a covalently bound enzyme-glycosyl intermediate is formed. In the second step, the leaving group is exchanged for an acceptor molecule, which is then linked via a new a-glycosidic bond to the intermediate. Recently, amylomaltases from thermophUe organisms like Thermus aquaticus and Thermus thermophilus HB8 have been isolated. These enzymes have a high thermostabihty, which makes them suitable for industrial apphcations, such as the production of large cychc glucans and the production of thermoreversible gels from starch. A 2.0 A 3D structure of the amylomaltase from Thermus aquaticus shows that the enzyme consists of a compact (alpha/beta)s-barrel catalytic domain with three loop excursions that are probably responsible for part of the enzyme's specificity. In the catalytic site, 6 out of the 7 conserved residues of the a-amylase famUy are present, showing the close relatedness between amylomaltase and the a-amylase famUy.

EXPERIMENTAL PROCEDURES

Bacterial strains and plasmids - Escherichia coli TOP10 (Invitrogen) was used for recombinant DNA manipulations. AMase (mutant) proteins were produced with E. coli BL21(DE3) (Stratagene). The malQ gene was amplified with PCR using the foUowing primers:

Thermus thermophilus:

Forward: GGCAGCCATATGGAGCTTCCCCGCGCTTTCGG Reverse: GCAGCCAGATCTAGAGCCGTTCCGTGGCCTCGGC

Aquifex aeolicus: Forward: GGCAGCCATATGAGATTGGCAGGTATTTTAC Reverse: GCAGCCGGATCCTTAAACTTCTCTTCCG

The PCR product was digested with Ndel (CATATG), and Bgllϊ (AGATCT. overhang compatible with BamHI, T. thermophilus) or BamHI (GGATCC. A. aeolicus) and ligated with plasmid pETlδb (Novagen), digested with Ndel and BamHI. The resulting construct (pCCBmalQ) encodes the amylomaltase with an N-terminal His6- tag.

Site-directed mutagenesis - For site-directed mutagenesis a method based upon PCR reactions using PWO-DNA polymerase was used. In a first PCR reaction a mutagenesis primer together with the reverse primer was used. The product of this reaction was used as a primer in a second PCR reaction together with the forward primer. This PCR product was cloned in pETlδb using the same strategy as for the wUd type. The foUowing mutagenesis primers were used to produce the mutations:

Tliermus thermophilus: F251L/S: 5'-CCC'CCC'GAC'TAC'TYG'AGC'GAG'ACC'GGT'CAG'CGC'TGG' GGC-3', F366L/S: 5'-AAG'GTC'CTG'CA^rYG'GCC'TTT'GAC'GAC-3'

Aquifex aeolicus: F244L/S: 5'-CCT'CCT'GAT'TTC'TYG'AGTAAA'ACG'GG-3' F359L/S:5'-GTT'ATT'GAG'TYG'GCC'TTC'TAC'G-3' In these primers Y= T (F-L) of C (F-S). Successful mutagenesis resulted in appearance of the underlined restriction sites, aUowing rapid screening of potential mutants. For F251L/S this restriction site was Agel (ACCGGT); for F251S an additional Xhol site (CTCGAG) was introduced. Mutation F366L/S caused deletion of a Pstl site; for F366L an additional Muni site (CAATTG) was introduced. For F244S an Xhol site (CTCGAG) was introduced. AU mutations were confirmed by restriction analysis and DNA sequencing.

DNA manipulations - Restriction endonucleases were purchased from Pharmacia LKB Biotechnology, Sweden; NEB; or Boehringer, and used according to the manufacturer's instructions. DNA manipulations and calcium chloride transformation of E. coli strains were as described.

Growth conditions - Plasmid carrying bacterial strains were grown on LB medium containing 50 ig/ml ampicUin ( E. coli TOP10) or 50 ig/ml ampicilin and 50 ig/ml chloramphenicol (E. coli DE3(RP)). For the production of (mutant) AMase proteins E. coli DE3(RP), containing the pCCBmalQ vector, was grown in a 1 1 flask with 250 ml LB medium containing 50 ig/ml ampicihn.

Protein determination - Protein concentrations were determined with the Bradford method {63} using the Bio-Rad reagent and bovine serum albumin as a standard (Bio-Rad Laboratories, Richmond, CA, USA).

Enzyme assays - AU assays were performed in a 25 mM sodium maleate buffer (pH 6.5) at 70 °C.

Disproportionation reaction - Disproportionation activities were determineded using the ability of AMase to release glucose from ohgosaccharides. Various concentrations (upto 50 mM) of (mixtures of) ohgosaccharides (G2-G7) were incubated with appropriatly chluted enzyme. For the determination of donor specificity different concentrations of maltoohgosaccharides as donor and methyl-α- D-glucose as acceptor .At regular time intervals 50 U samples were taken and added to 200 microliter GOD-PAP reagent (Roche) to measure the amount of glucose released. Hydrolyzing activities were measured as described earlier using 1% soluble starch (Lamers & Pleuger, Belgium) as substrate and dinitrosaUcyhc acid to determine the number of reducing ends .

In above assays 1 U of activity is defined as the amount of enzyme required for the processing of 1 mmole of donor substrate per minute. Kinetic parameters were fitted using the computer program Sigma Plot (Jandel Scientific).

Product formation from oligosaccharides was analyzed by HPLC. For this purpose 1 ml of a 25 mM G3, G5, or G7 solution was incubated with 0.1 U AMase at 70 °C for 8 h. Samples were taken at regular time intervals and the products formed were apphed to a 25 cm Econosphere-NH₂ 5 micron column (AUtech Associates Inc. USA) eluted with acetonitrUe/water (60/40, v/v) at a flow rate of 1 ml per min.

Sequence alignments - Sequence ahgnments of various amylomaltases indicate that the two phenylalanines selected based on the structure of amylomaltase and the model of the maltoheptaose bound in the active site are (functionaUy) conserved in aU amylomaltases.

T.thermophUus GVPPDYFSETGQRWGNP

T.aquaticus {4} GVPPDYFSETGQRWGNP

Synechocystis {101} GVPPDYFSATGQLWGNP

Aaeolicus {97} GVPPDFFSKTGQLWGNP S.tuberosum {25} GVPPDAFSETGQLWGSP

C.butyricum {11} GCPPDAFSETGQLWGNP

S.pneumoniae {55} GCPPDEFSVTGQLWGNP

M.tuberculosis {96} GAPPDEFNQLGQDWSQP

H.influenzae {103} GAPPDPLGPVGQNWNLP E.coh {40} GAPPDILGPLGQNWGLP

C.pneumoniae {94} GAPPDLYNSEGQNWHLP

C.psittaci {99} GAPPDIYNTEGQNWHLP

C.trachomatis {95} GAPPDLYNAEGQNWHLP

T.thermophUus LAEDLGVITPEVEALRDRFGLPGMKVLQFAF T. aquaticus LAEDLGVITPE VEALRDRFGLPGMKVLQFAF Synechocystis VAEDLGVITPEVEALRDEFNFPGMKVLHFAF Aaeolicus IAEDLGFITDEVRYLRETFKIPGSRVIEFAF

S.tuberosum IAEDLGVITEDWQLRKSIEAPGMAVLQFAF C.butyricum IAEDLGYLTEETLEFKKRTGFPGMKIIQFAF S.pneumoniae IAEDLGFMTDEVIELRERTGFPGMKILQFAF M.tuberculosis VGEDLGTVEPWVRDYLLLRGLLGTSILWFEQ H.influenzae IGEDLGTVPDEVRWKLNEFQIFSYFVLYFAQ E.coli IGEDLGTVPVEIVGKLRSSGVYSYKVLYFEN C.pneumoniae IGEDLGIIPQDVKTTLTHLGICGTRIPRWER C.psittaci IGEDLGSVPTDVKETLVKLGICGTRIPRWER

C.trachomatis IGEDLGTIPSDVKRMLESFAVCGTRIPRWER Construction of mutant enzymes - One mutant (F366L, Thermus) has been constructed and confirmed by sequence analysis. Other mutants have been constructed (for example F366S (Thermus), F359L/S (Aquifex))

Disproportionation activity - Mutant F366L has been analyzed concerning the disproportionation of maltotriose. The activity (25 U/mg) was four times lower than that of the wUd type, whereas the affinity (Km = 3.5) was threefold higher than the wUd type.

Hydrolyzing activity - As for the wUd type, no hydrolyzing activity could be determined during incubation of soluble starch, even with large amounts of enzyme. However, contrary to the wild type enzyme, an increase in reducing power of the reaction mixture after overnight incubation was detected, indicating that hydrolysis had taken place.

Product formation from maltotriose - HPLC analysis of the products formed during incubation of the enzyme with maltotriose clearly shows that hydrolysis takes place. Whereas the wild type produces essentiaUy no maltose, which cannot be cleaved of by the enzyme, the mutant produces maltose as one of the main compounds.

Interaction with hydrophobic amino acids, such as F366, which is highly conserved in amylomaltases, is involved in the reaction specificity of the enzyme. Hydrolyzing activity can be introduced by mutating this residue or other hydrophobic residues. This hydrolyzing activity has significant effects on product profiles of the enzyme, indicating the necessity of essentiaUy complete or practicaUy complete absence of hydrolysis for the function of the wUd type enzyme (the production of longer ohgosaccharides from short substrates).

Alignement of Branching Enzymes

BstearothermophUus

Bcaldolyticus BsubtUis mycobacterium MSRSEKLTGEH-LAPEPA-—

EMARLVAGT

Streptomyces

MSAARQPSPTVRDKAAPEPAAPAAPKGARAPP1ARRAAPPHGVRPAPALAAEERAR LLEGR

E. MSDRIDRDVTNALIAGH

H.influenzae MTTAv QAIΪDGFFDAS

Agro.tume MKKPLNSAEEKKTGDITKAEIEAIKSGL

Aquifex_a. Synechococcus TGTTPLPSSSLSVEQVNRIASNQ

Synechocystis MTYTINADQVHQIVHNL

Butyrivibrio

CHLAMYDIA MDPFFLNTQHVELLVSGK

BstearothermophUus

Bcaldolyticus

BsubtUis mycobacterium HHNPHGILGAHEYDDHTVIR AFRPHAVEWALVGK— DRFSLQHLD-SGLFAVA

Streptomyces HHDPHAVLGARTQRGGVAFR VLRPYAKAVTWAKG—

LRTELVDEG-DGLFSGL

E. FADPFSVLGMHKTTAGLEVR

ALLPDATDVWNIEPKTGRKLAKLECLDSRGFFSGV H.influenzae ΝGDPFATLGMHETEQGIEIR

TLLPDAΝRMWIERESGKEITELDCVDERGFFVGV

Agro.tume HSΝPFQIIPLHETPEGFSAR CFIPGAEEVSVLTLD-

GΝFVGELKQIDPDGFFEGR

Aquifex_a. Synechococcus EQNPFDILGPHPYEHEGQAG-WVTRAYLPEAQEAAVICPAL-

RREFAMHPVHHPHFFETW

Synechocystis

HHDPFEVLGCHPLGDHGKVNQWVIRAYLPTAEAVTVLLPTD- RREVΓMTTVHHPNFFECV

Butyrivibrio

CHLAMYDIA QSSPQDLLGIVS-ESLNQDR— IVLFRPGAETVFVELRG— -

KIQQAESHHSGIFSLP

BstearothermophUus

MIAANPTDLEVYLFHEGSLYKSYELFGAHV- Bcaldolyticus

MIAANPTDLEVYLFHEGRLYQSYELFGAHV- Bsubtihs

MAAASPTAHDVYLFHEGSLFKSYQLFGSHY- mycobacterium LPFVD-

LIDYRLQVTYEGCEPHTVADAYRFLPTLGEVDLHLFAEGRHERLWEVLGAHPRS

Streptomyces LPLTG-VPDYRLLVTYDSDE- IEVHDPYRFLPALGELDLHLIGEGRHEELWTALGSQP-

E. IPRRKNFFRYQLAWWHGQQ-

NLIDDPYRFGPLIQEMDAWLLSEGTHLRPYETLGAHA-

H.influenzae IPNCRQFFAYQLQVFWGNEA-

QIIEDPYRFHPMIDDLEQWLLSEGSMLRPYEVLGAHF- Agro.tume IDLSK-RQPVRYRACRDDAE-

WAVTDPYSFGPVLGPMDDYFVREGSICGYSTGWARIP-

Aquifex_a.

MKKFSLISDYDVYLFKEGTHTRLYDKLGSHV-

Synechococcus VPEET- LEIYQLRITEGERERIIYDPYAFRSPLLTDYDIHLFAEGNHHRIYEKLGAHP--

Synechocystis LELEE-

PE-NYQLRITENGHERVΓYDPYGFKTPKLTDFDLHVFGEGNHHRΓYEKLGAHL--

Butyrivibrio

MSQKVFISEDDEYLFGQGTHYDΓYDKLGAHP- CHLAMYDIA VMKGISPQDYRVYHQN-G—

LLAHDPYAFPLLWGEIDSFLFHEGTHQRIYERMGAIP-

BstearothermophUus -INEGG-

KVGTRFCVWAPHAREVRLVGSFNDWDGTDFRLEKVND-EGVWTIWPENLEGH

Bcaldolyticus -IRGGG-

AVGTRFCVWAPHARENRLVGSFNDWNGTNSPLTKVND-EGVWTIWPENLEGH

BsubtUis -RELNG- KSGYEFCNWAPHASEVRVAGDFNSWSGEEHVMHRVND-NGIWTLFIPGIGEKE mycobacterium

FTTADGWSGVSFAVWAPNAKGVSLIGEFNGWNGHEAPMRVLGP-

SGVWELFWPDFPCDG

Streptomyces -MEHQG- VAGTRFTVWAPNALGVRVTGDFSYWDAVAYPMRSLGA-SGVWELFLPGVAEGA

E. -DTMDG-

VTGTRFSNWAPNARRVSWGQFNYWDGRRHPMRLRKE-SGIWELFIPGAHNGQ

H.influenzae -MECDG-

VSGVNFRLWAPNARRVSIVGDFNYWDGRRHPMRFHSK-SGVWELFLPKASLGQ Agro.tume -LKLEG-

VEGFHFAVWAPNGRRVSWGDFNNWDGRRHVMRFRKD-TGIWEIFAPDVYA-C

Aquifex_a. -IELNG-

KRYTFFAVWAPHADYVSLIGDFNEWDKGSTPMVKREDGSGIWEVLLEGDLTGS

Synechococcus -CELEN- VAGVNFAVWAPSARNVSILGDFNSWDGRKHQMAR-RS-NGIWELFIPELTVGA

Synechocystis -MTVDG-

VKGVYFAVWAPNARNVSILGDFNNWDGRLHQMRK-RN-NMVWELFIPELGVGT

Butyrivibrio -SEEKG-

KKGFFFAVWAPNAADVHVVGDFNGWDENAHQMKRSKT-GNIWTLFIPGVAIGA CHLAMYDIA -CEIDG-

VPGVRFIVWAPHAQRVSVIGDFNGWHGLVNPLHKVSD-QGVWELFVPGLTAGA

* - "k BstearothermophUus LYKYErvTPDGQVL-

FKADPYAFYSELRPHTASIAYDLKGYQWNDQSWKRKKRRKRIYDQ

Bcaldolyticus LYKYEIITPDGRVL-

LKADPYAFYSELRPHTASIVYDLKGYEWNDSPWQRKKRRKRIYDQ BsubtUis RYKYEIVTNNGEIR-

LKADPYAIYSEVRPNTASLTYDLEGYSWQDQKWQKKQKAKTLYEK mycobacterium LYKFRVHGADGWT-DRADPFAFGTEVPPQTASRVT-

SSDYTWGDDDWMAGRALRNPVNE

Streptomyces LYKYEITRPDGGRT-LRADPMARYAEVPPANASIVT- ASRYEWQDAEWMARRGALAPHQA

E. LYKYEMIDANGNLR-LKSDPYAFEAQMRPETASLIC-

GLPEKWQTEERKKANQFDA-

H.influenzae LYKFELIDCHGNLR-LKADPFAFSSQLRPDTASQVS-

ALPNWEMTEARKKANQGNQ- Agro.tume

AYKFEILGANGELLPLKADPYARRGELRPKNASVTAPELTQKWEDQAHREHWAQ

VDQRRQ

Aquifex_a. KYKYFIKNGNYEVD-KSDPFAFFCEQPPGNASWW-

KLNYRWNDSEYMKKRKRVNSHDS Synechococcus AYKYEIKNYDGHIYE-

KSDPYGFQQEVRPKTASIVADLDRYTWGDADWLERRRHQEPLRQ

Synechocystis SYKYEIKNWEGHIYE-

KTDPYGFYQEVRPKTASIVADLDGYQWHDEDWLEARRTSDPLSK

Butyrivibrio LYKFLITAQDGRKLY- KADPYANYAELRPGNASRTTDLSGFKWSDSKWYESLKGKDMNRQ

CHLAMYDIA CYKWEMVTESGQVL-IKSDPYGKFFGPPPWSVSWI-

DDSYEWTDSEWLEERIKKTEG-

BstearothermophUus PMVIYELHFGSWKKK— - DGRFYTYREMADELISYVLDH

Bcaldolyticus PMVIYELHFGSWKKKP

DGRFYTYREMADELIPYVLER

BsubtUis PVFIYELHLGSWKKHS

DGRHYSYKELSQTLIPYIKKH mycobacterium AMSTYEVHLGSWRP

GLSYRQLARELTDYIVDQ

Streptomyces PMSVYELHLASWRP

GLSYRQLAEQLPAYVKEL E. PISIYEVHLGSWRRH TDNN

FWLSYRELADQLVPYAKWM

H.influenzae PISIYEVHLGSWRRN LENN

FWLDYDQIADELIPYVKEM

Agro.tume PISIYEVHAGSWQR SEDG TFLSWDELEAQLIPYCTDM

Aquifex_a. PISIYEVHVGSWRRVP EEGN

RFLSYRELAEYLPYYVKEM

Synechococcus

PISVYEVHLGSWMHASSDAIATDAQGKPLPPVPVADLKPGARFLTYRELADRLIPY VLDL

Synechocystis PVSVYELHLGSWLHTAYDEPVKTLHGEGVP-

VEVSEWNTGARFLTYYELVDKLIPYVKEL

Butyrivibrio PIAIYECHIGSWMKHP DGTEDG

FYTYRQFADRIVEYLKEM CHLAMYDIA PMNIYEVHVGSWRWQE

GQPLNYKELADQLALYCKQM

BstearothermophUus GFTHIELLPLVEHPLDRSWGYQGTGYYAVTSRYGTPHDFMYFVDRCHQAGIGVIM

DWVPG

Bcaldolyticus

GFTHIELLPLVEHPLDRSWGYQGTGYYSVTSRYGTPHDFMYFVDRCHQAGLGVII

DWVPG Bsubtihs

GFTHIELLPVYEHPYDRSWGYQGTGYYSPTSRFGPPHDLMKFVDECHQQNIGVIL

DWVPG mycobacterium

GFTHVELLPVAEHPFAGSWGYQVTSYYAPTSRFGTPDDFRALVDALHQAGIGVIVD

WVPA

Streptomyces GFTHVELMPVAEHPFGGSWGYQVTGFYAPTSRMGTPDDFRFLVDALHRAGIGVIV

DWVPA

E.

GFTHLELLPINEHPFDGSWGYQPTGLYAPTRRFGTRDDFRYFIDAAHAAGLNVILD

WVPG H.influenzae

GFTHIEFLPLSEFPFDGSWGYQPLGLYSPTSRFGSPEAFRRLVKRAHEAGINVILD

WVPG

Agro.tume

GFTHIEFLPITEHPYDPSWGYQTTGLYAPTARFGDPEGFARFVNGAHKVGIGVLLD WVPA

Aquifex_a.

GFTHVEFLPVMEHPFYGSWGYQITGYFAPTSRYGTPQDFMYLIDKLHQEGIGVILD

WVPS

Synechococcus GYSHIELLPIAEHPFDGSWGYQVTGYYAATSRYGSPEDFMYFVDRCHQNGIGVILD

WVPG

Synechocystis

GYTHIELLPIAEHPFDGSWGYQVTGYYAPTSRFGSPEDFMYFVDQCHLNGIGVIID

WVPG Butyrivibrio

KYTHIELIGIAEHPFDGSWGYQVTGYYAPTARYGEPTDFMYLINQLHKHGIGVILD

WVPA

CHLAMYDIA

HYTHVELLPVTEHPLNESWGYQTTGYYAPTSRYGSFEDLQYFIDTMHQHGIGVIL DWVPG BstearothermophUus

HFCKDAHGLYMFDGAPTYEYANEKDRENYVWGTANFDLGKPEVRSFLISNALFW

LEYYHI

Bcaldolyticus HFCKDAHGLYMFDGAPTYEYANEKDRENYVWGTANFDLGKPEVRSFLISNALFW

LEYYHV

BsubtUis

HFCKDAHGLYMFDGEPLYEYKEERDRENWLWGTANFDLGKPEVHSFLISNALY

WAEFYHI mycobacterium

HFPKDAWALGRFDGTPLYEHSDPKRGEQLDWGTYVFDFGRPEVRNFLVANALY

WLQEFHI

Streptomyces

HFPRDDWALAEFDGRPLYEHQDPRRAAHPDWGTLEFDYGRKEVRNFLVANAVY WCQEFHV

E.

HFPTDDFALAEFDGTNLYEHSDPREGYHQDWNTLIYNYGRREVSNFLVGNALYW

IERFGI

H.influenzae HFPSDTHGLVAFDGTALYEHEDPREGYHQDWNTLIYNYGRNEVKNFLSSNALYW

LERFGV

Agro.tume

HFPTDEHGLRWFDGTALYEHADPRQGFHPDWNTAIYNFGRIEVMSYLINNALYW

AEKFHL Aquifex_a.

HFPTDAHGLAYFDGTHLYEYEDWRKRWHPDWNSFVFDYGKPEVRSFLLSSAHF

WLDKYHA

Synechococcus

HFPKDGHGLAFFDGTHLYEHADSRQGEHREWGTLVFNYGRHEVRNFLAANALF WFDKYHI

Synechocystis

HFPKDGHGLAFFDGTHLYEHGDPRKGEHKEWGTLIFNYGRNEVRNFLVANALF

WFDKYHI Butyrivibrio

HFCPDEFGLACFDGTCΓYEDPDPRKGEHPDWGTKIFNLAKPEVKNFLIANALYWI

RKFHI

CHLAMYDIA HFPIDSFAMSGFDGTPLYEYTRNPSPLHPHWHTYTFDYAKPEVCNFLLGSVLFWI

DKMHV

** * . *** ** . * . .. . ** .* .*

BstearothermophUus DGFRVDAVANMLYWPNNDRL YE— NPYAVEFLRKLNEAVFAYDPNALMIAED

Bcaldolyticus DGFRVDAVANMLYWPNNDRL YE—

NPYAVEFLRQLNEAVFAYDPNVWMIAED

BsubtUis DGFRVDAVANILYWPNQDER HT— -

NPYAVDFLKKLNQTMREAYPHVMMIAED mycobacterium

DGLRVDAVASMLYLDYSRPEGGWTPNVHGGRENLEAVQFLQEMNATAHKVAPGΓ

VTIAEE

Streptomyces

DGLRADAVASMLYLDYSRDEGDWSPNAHGGREDLDAVALLQEMNATVYRRFPGV VTIAEE

E.

DALRVDAVASMIYRDYSRKEGEWIPNEFGGRENLEAIEFLRNTNRILGEQVSGAVT

MAEE

H.influenzae DGIRVDAVASMIYRDYSRAEGEWIPNQYGGRENLEAIEFLKHTNWKIHSEMAGAI

SIAEE

Agro.tume

DGLRVDAVASMLYLDYSRKEGEWIPNEYGGRENLESVRFLQKMNSLVYGTHPGV

MTIAEE Aquifex_a. DGLRVDAVASMLYLDYSRKE-

WVPNIYGGKENLEAIEFLRKFNESVYRNFPDVQTIAEE

Synechococcus

DGIRVDAVASMLYLDYNRKEGEWIPNEYGGRENIEAADFLRQVNHLIFSYFPGALS

IAEE Synechocystis DGMRVDAVASMLYLDYCREEGEWVANEYGGRENLEAADFLRQVNSWYSYFPGI

LSIAEE

Butyrivibrio

DGLRVDAVASMLYLDYGKI03GQWVPNKYGDNKNLDAIEFFKHFNSNVRGTYPNI

LTIAEE

CHLAMYDIA

DGIRVDAVSSMLYLDYGRYAGEWVPNRYGGRENLDAIRFLQQFNTVIHEKYPGNL

TFAEE

BstearothermophUus STDWPKVTAPTYEGGLGFΝYKΛVΝMGWMΝDMLKYMETPPYERRHVHΝQVTFSL

LYAYSEΝF Bcaldolyticus

STDWPRVTAPTYDGGLGFΝYKWΝMGWMΝDMLKYMETPPHERKYAHΝQVSFSL

LYAYSEΝF

BsubtUis

STEWPQVTGAVEEGGLGFHYKWΝMGWMΝDVLKYMETPPEERRHCHQLISFSLL YAFSEHF mycobacterium

STPWSGVTRPTΝIGGLGFSMKWΝMGWMHDTLDYVSRDPVYRSYHHHEMTFSML

YAFSEΝY

Streptomyces STAWDGVTRPTDSGGLGFGLKWΝMGWMHDTLRYVSKEPVHRKYHHHDMTFGM

VYAFSEΝF

E.

STDFPGVSRPQDMGGLGFWYKWΝLGWMHDTLDYMKLDPVYRQYHHDKLTFGI

LYΝYTEΝF H.influenzae

STSFAGVTHPSEΝGGLGFΝFKWΝMGWMΝDTLAYMKLDPIYRQYHHΝKMTFGM

VYQYSEΝF Agro.tume

STSWPKVSQPVHEGGLGFGFKWNMGFMHDTLSYFSREPVHRKFHHQELTFGLL

YAFTENF

Aquifex_a. STAWPMNSRPTYVGGLGFGMKWNMGWMNDTLFYFSKDPIYRKYHHEVLTFSIW

YAFSENF

Synechococcus

STSWPMVSWPTYNGGLGFNLKWNMGWMHDMLDYFSMDPWFRQFHQNNVTFSI

WYAFSENF Synechocystis

STSWPMVSWPTYVGGLGFNLKWNMGWMHDMLDYFSMDPWFRQFHQNSITFSM

WYNHSENY

Butyrivibrio

STAvVPKVTAPPEEDGLGFAFKWNMGWMHDFCEYMKLDPYFRQGAHYMMTFAM SYNDSENY

CHLAMYDIA

STTFPKITVSVEEGGLGFDYKWNMGWMHDTLHYFEKDFPYRPYHQSDLTFPQW

YAFSERF

BstearothermophUus

ILPFSHDEVVHGKKSLLNKMPGSYEEKFAQLRLLYGYMMAHPGKKLLFMGNEFA

QFDEWK

Bcaldolyticus

ILPFSHDEVVHGKKSLLNKMPGSYEEKFAQLRLLYGYMMAHPGKKLLFMGSEFA QFDEWK

BsubtUis

VLPFSHDEVVYGKKSLLNKMPGDYWQKFAQYRLLLGYMTVHPGKKLIFMGSEFA

QFDEWK mycobacterium VLPLSHDEVVHGKGTLWGRMPGNNHVKAAGLRSLLAYQWAHPGKQLLFMGQEF

GQRAEWS

Streptomyces

VLPISHDEWHGKRSLVSKMPGDWWQQRATHRAYLGFMWAHPGKQLLFMGQEF

AQGSEWS E.

VLPLSHDEWHGKKSILDRMPGDAWQKFANLRAYYGWMWAFPGKKLLFMGNEF

AQGREWN

H.influenzae VLPLSHDEVVHGKYSLLGKMPGDTWQKFANLRAYYGYMWGYPGKKLLFMGNEF

AQGREWN

Agro.tume

VLPLSHDEVVHGKGSLIAKMSGDDWQKFANLRSYYGFMWGYPGKKLLFMGQEF

AQWSEWS Aquifex_a.

VLPLSHDENVHGKGSLIGKMPGDYWQKFANLRALFGYMWAHPGKKLLFMGGEF

GQFKEWD

Synechococcus

MLALSHDEWHGKSNLIGKMPGDEWQKFANLRCLLGYMFTHPGKKTLFMGMEF GQWAEWN

Synechocystis

Ml^LSHDEVVHGKSNMLGKMPGDEWQKYANVRALFTYMFTHPGKKTMFMSME

FGQWSEWN

Butyrivibrio ILPLSHDEVVHLKCSMVEKMPGYKVDKYANLRVGYTYMFGHSGKKLLFMGQDF

GQEREWS

CHLAMYDIA

LLPFSHDEWHGKRSLIGKMPGDAWRQFAQLRLLLGYQICQPGKKLLFMGGEFG

QGREWS

.^•&^■ .-k-k-k-k- -k . * . .* -k . - it . it-k . . ick ,"k it ^{■ •}

BstearothermophUus^" FEDELDWVLFDF

ELHRKMNDYMKELIACYKRYKPFYELDHDPQGFEWIDVHNAEQ

Bcaldolyticus FAEELDWVLFDF ELHRKMDEYVKQLIACYKRYKPFYELDHDPRGFEWIDVHNAEQ

Bsubtihs DTEQLDWFLDSF

PMHQKASVFTQDLLRFYQKSKILYEHDHRAQSFEWIDVHNDEQ mycobacterium EQRGLDWFQLDE— -

NGFSNGIQRLVRDINDIYRCHPALWSLDTTPEGYSWIDANDSAN Streptomyces

ETYGPDWWVLDSSYPAAGDHLGVRSLVRDLNRTYTASPALWERDSVPEGFAWVE

ADAADD

E. HDASLDWHLLEG- GDNWHHGVQRLVRDLNLTYRHHKAMHELDFDPYGFEWLWDDKER

H.influenzae YEESLDWFLLDENI-

GGGWHKGVLKLVKDLNQIYQKNRPLFELDNSPEGFDWLWDDAAN

Agro.tume EKGSLDWNLRQY

PMHEGMRRLVRDLNLTYRSKAALHARDCEPDGFRWLWDDHEN Aquifex_a. HETSLDWHLLEY

PSHRGIQRLVKDLNEVYRREKALHETDFSPEGFEWVDFHDWEK

Synechococcus VWGDLEWHLLQY

EPHQGLKQFVKDLNHLYRNAPALYSEDCNQAGFEWIDCSDNRH

Synechocystis VWGDLEWHLLNF PPHQQLKQFFTELNHLYKNEPALYSNDFDESGFQWIDCSDNRH

Butyrivibrio EKRELDWFLLEN

DLNRGMIiDYVGKLLEIYRKYPALYEVDNDWGGFEWINADDKER

CHLAMYDIA PGRELDWELLDI

SYHQGVHLCSQELNALYVQSPQLWQADHLPSSFRWVDFSDVRN

.^•k . * . * . *.

BstearothermophUus SIFSFIRRGKKED-DVLVTVCNFTNQAYDDYKVGVP-

LLVPYREVLNSDAVTFGGSGHVN

Bcaldolyticus SIFSFIRRGKKEG-DVLVIVCNFTNQAYDDYKVSVP- LLAPYREVLNSDAAEFGGSGHVN

BsubtUis SIFSFIRYGQKHG-EALVIICNFTPWYHQYDNGVP-

FFTQYIEVLNSDSETYGGSGQIN mycobacterium NVLSFMRYGSDG-SVLACVFNFAGAEHRDYRLGLP-

RAGRWREVLNTDATIYHGSGIGN Streptomyces NVFAFLRFARDG-

SPLLCVSNFSPWRHGYRIGVPQEVGQWREVLNTDLEPYGGSGVHH

E. SVLIFVRRDKEG-NEIIVASNFTPVPRHDYRFGIN-

QPGKWREILNTDSMHYHGSNAGN H.influenzae SVLAFERRSSNG-ERIIWSNFTPVPRHNYRIGVN-

VAGKYEEILNTDSMYYEGSNVGN

Agro.tume SVFAWLRTAPGE-KPVAVICNLTPVYRENYYVPLG-

VAGRWREILNTDAEIYGGSGKGN Aquifex_a. SVISFLRKDKSGK-EIILWCNFTPVPRYDYRVGVP-

KGGYWREIMNTDAKEYWGSGMGN

Synechococcus SIVSFIRRAHESD-RFLVWCNFTPQPHAHYRIGVP-

VAGFYREIFNSDARSYGGSNMGN

Synechocystis SWSFIRRAKNSA-EFWTICNFTPQPHSHYRVGVP- VPGFYTELFNSDARQYGGSNMGN

Butyrivibrio STYSFYRRASNGK-DNILFVLNMTPMERKGFKVGVP-

FDGTYTKILDSAKECYGGSGSSV

CHLAMYDIA GWAYLRFADADAKKALLCVHHFGVGYFPHYLLPIL-

PLESCDLLMNTDDTRFGGSGKGF

BstearothermophUus GKR-LSAFNEPFHGK P-

YHVRMTIPPFGISILRPVQKRGERKRNEK

Bcaldolyticus GKR-LPAFSEPFHGK P- YHVRMTIPPFGISILRPVQKRGERKQNEE

BsubtUis KKP-LSAKKGALHHK P-

CYITMTIPPYGISILRAVKKRGEIKR— mycobacterium LGG-VDATDDPWHGR P-

ASAVLVLPPTSALWLTPA Streptomyces ARA-LRPEPVPAQGR A-VSLRMTLPPMATVWLRP-

E. GGT-VHSDEIASHGR Q-HSLSLTLPPLATIWLVREAE

H.influenzae FGC-VASEQIESHGR E-NSISVSIPPLATVYLRLKTK-

Agro.tume GG— -RVQAVDAGG E-IGAMLVLPPLATIMLEPEN— -

Aquifex_a. LGG-KEADKIPWHGR K-FSLSLTLPPLSVIYLKHEG- Synechococcus LGG-KWTDEWSCHNR P-

YSLDLCLPPLTTLVLELASGPES—-LS

Synechocystis LGG-KWTEEWSFHEQ P-

YSLDLCLPPLSVLVLKLSQNAEENTVPAE Butyrivibrio PDK-IKAVKGLCDYK D-

YSIEFDLPPYGAEVFVFQTKKTKN

CHLAMYDIA REPEILTPEIARQEREAAGLIEADDESGPDCWGLDIELPPSATLIFSVTLQ-

BstearothermophUus EMHRHVIGRRARKSASLADDKHR

Bcaldolyticus EVHRHVIGRRARKPASLADEKHRETSRAVWGEVPDH

Bsubtihs mycobacterium

Streptomyces

E.

H.influenzae

Agro.tume

Aquifex_a.

Synechococcus EAANSPL

Synechocystis EASNIA

Butyrivibrio

CHLAMYDIA

Ahgnement of BE and isoamylases.

BstearothermophUus

Bcaldolyticus

BsubtUis mycobacterium -MSRSEKLTGEH-LAPEPA--

EMARLVAGT

Streptomyces

MSAARQPSPTVRDIKAAPEPAAPAAPKGARAPRARRAAPPHGWPAPALAAEERAR

LLEGR

E. MSDRIDRDVTNALIAGH

H.influenzae MTTAVTQAIIDGFFDAS

Agro.tume -MKKPLNSAEEKKTGDITKAEIEAIKSGL

Aquifex_a.

Synechococcus -TGTTPLPSSSLSVEQVNRIASNQ

Synechocystis MTYTINADQVHQIVHNL

Butyrivibrio

CHLAMYDIA -MDPFFLNTQHVELLVSGK

BstearothermophUus

Bcaldolyticus

BsubtUis mycobacterium HHNPHGILGAHEYDDHTVIR AFRPHAVEWALVGK-

DRFSLQHLD-SGLFAVA

Streptomyces HHDPHAVLGARTQRGGVAFR VLRPYAKAVTWAKG-

LRTELVDEG-DGLFSGL

E. FADPFSVLGMHKTTAGLEVR ALLPDATDVWNIEPKTGRKLAKLECLDSRGFFSGV

H.influenzae ΝGDPFATLGMHETEQGIEIR

TLLPDAΝRMWIERESGKEITELDCVDERGFFVGV

Agro.tume HSΝPFQIIPLHETPEGFSAR CFIPGAEEVSVLTLD-

GΝFVGELKQIDPDGFFEGR Aquifex_a.

Synechococcus EQNPFDILGPHPYEHEGQAG-WVIRAYLPEAQEAAVICPAL-

RREFAMHPVHHPHFFETW

Synechocystis

HHDPFEVLGCHPLGDHGKVNQWVIRAYLPTAEAVTVLLPTD-

RREVIMTTVHHPNFFECV

Butyrivibrio

CHLAMYDIA QSSPQDLLGIVS-ESLNQDR— IVLFRPGAETVFVELRG— -

KIQQAESHHSGIFSLP

BstearothermophUus

MIAANPTDLEVYLFHEGSLYKSYELFGAHV- Bcaldolyticus MIAANPTDLEVYLFHEGRLYQSYELFGAHV- Bsubtilis

MAAASPTAHDVYLFHEGSLFKSYQLFGSHY" mycobacterium LPFVD-

LIDYRLQVTYEGCEPHTVADAYRFLPTLGEVDLHLFAEGRHERLWEVLGAHPRS Streptomyces LPLTG-VPDYRLLVTYDSDE-

IEVHDPYRFLPALGELDLHLIGEGRHEELWTALGSQP-

E. IPRRKNFFRYQLAWWHGQQ-

NLIDDPYRFGPLIQEMDAWLLSEGTHLRPYETLGAHA-

H.influenzae IPNCRQFFAYQLQVFWGNEA- QIIEDPYRFHPMIDDLEQWLLSEGSMLRPYEVLGAHF-

Agro.tume IDLSK-RQPVRYRACRDDAE-

WAVTDPYSFGPVLGPMDDYFVREGSICGYSTGWARIP-

Aquifex_a.

MKKFSLISDYDVYLFKEGTHTRLYDKLGSHV- Synechococcus VPEET-

LEIYQLRITEGERERIIYDPYAFRSPLLTDYDIHLFAEGNHHRIYEKLGAHP-

Synechocystis LELEE-

PK ^QLRITENGHERVIYDPYGFKTPKLTDFDLHVFGEGNHHRIYEKLGAHL-- Butyrivibrio

MSQKVFISEDDEYLFGQGTHYDIYDKLGAHP- CHLAMYDIA VMKGISPQDYRVYHQN-G—

LLAHDPYAFPLLWGEIDSFLFHEGTHQRIYERMGAIP-

BstearothermophUus -INEGG-

KVGTRFCNWAPHAREVRLVGSFNDWDGTDFRLEKVND-EGVWTIVVPENLEGH

Bcaldolyticus -IRGGG- AVGTRFCVWAPHAREVRLVGSFNDWNGTNSPLTKVND-EGVWTIWPENLEGH

BsubtUis -RELNG-

KSGYEFCNWAPHASEVRVAGDFNSWSGEEHVMHRVND-NGIWTLFIPGIGEKE mycobacterium

FTTADGWSGVSFAVWAPNAKGVSLIGEFNGWNGHEAPMRVLGP- SGVWELFWPDFPCDG

Streptomyces -MEHQG-

VAGTRFTVWAPNALGVRVTGDFSYWDAVAYPMRSLGA-SGVWELFLPGVAEGA

E. -DTMDG-

VTGTRFSVWAPNARRVSWGQFNYWDGRRHPMRLRKE-SGIWELFIPGAHNGQ H.influenzae -MECDG-

VSGVNFRLWAPNARRVSIVGDFNYWDGRRHPMRFHSK-SGVWELFLPKASLGQ

Agro.tume -LKLEG-

VEGFHFAVWAPNGRRVSWGDFNNWDGRRHVMRFRKD-TGIWEIFAPDVYA-C

Aquifex_a. -IELNG- KRYTFFAVWAPHADYVSLIGDFNEWDKGSTPMVKREDGSGIWEVLLEGDLTGS

Synechococcus -CELEN-

VAGVNFAVWAPSARNVSILGDFNSWDGRKHQMAR-RS-NGIWELFIPELTVGA

Synechocystis -MTVDG-

VKGVYFANWAPNARNNSILGDFNNWDGRLHQMRK-RN-NMVWELFIPELGVGT Butyrivibrio -SEEKG-

KKGFFFAVWAPNAADVHWGDFNGWDENAHQMKRSKT-GNIWTLFIPGVAIGA

CHLAMYDIA -CEIDG-

VPGVRFIVWAPHAQRVSVIGDFNGWHGLVNPLHKVSD-QGNWELFVPGLTAGA BstearothermophUus LYKYEIVTPDGQVL-

FKADPYAFYSELRPHTASIAYDLKGYQWNDQSWKRKKRRKRIYDQ

Bcaldolyticus LYKYEIITPDGRVL- LKADPYAFYSELRPHTASIVYDLKGYEWNDSPWQRKKRRKRIYDQ

BsubtUis RYKYEIVTNNGEIR-

LKADPYAIYSEVRPNTASLTYDLEGYSWQDQKWQKKQK iTLYEK mycobacterium LYKFRVHGADGWT-DRADPFAFGTEVPPQTASRVT-

SSDYTWGDDDWMAGRALRNPVNE Streptomyces LYKYEITRPDGGRT-LRADPMARYAEVPPANASIVT-

ASRYEWQDAEWMARRGALAPHQA

E. LYKYEMIDANGNLR-LKSDPYAFEAQMRPETASLIC-

GLPEKWQTEERKKANQFDA—

H.influenzae LYKFELIDCHGNLR-LKADPFAFSSQLRPDTASQVS- ALPNWEMTEARKKANQGNQ-

Agro.tume

AYKFEILGANGELLPLKADPYARRGELRPKNASVTAPELTQKWEDQAHREHWAQ

VDQRRQ

Aquifex_a. KYKYFIKNGNYEVD--KSDPFAFFCEQPPGNASWW- KLNYRWNDSEYMKKRKRVNSHDS

Synechococcus AYKYEIKNYDGHIYE-

KSDPYGFQQEVRPKTASIVADLDRYTWGDADWLERRRHQEPLRQ

Synechocystis SYKYEIKNWEGHIYE-

KTDPYGFYQEVRPKTASIVADLDGYQWHDEDWLEARRTSDPLSK Butyrivibrio LYKFLITAQDGRKLY-

KADPYANYAELRPGNASRTTDLSGFKWSDSKWYESLKGKDMNRQ

CHLAMYDIA CYKWEMVTESGQVL-IKSDPYGKFFGPPPWSVSWI-

DDSYEWTDSEWLEERIKKTEG-

^■kit. . ..** * A BstearothermophUus PMVTYELHFGSWKKK- DGRFYTYREMADELISYVLDH

Bcaldolyticus PMVTYELHFGSWKKKP

DGRFYTYREMADELIPYVLER BsubtUis PVFIYELHLGSWKKHS

DGRHYSYKELSQTLIPYIKKH mycobacterium AMSTYEVHLGSWRP

GLSYRQLARELTDYΓVDQ Streptomyces PMSVYELHLASWRP

GLSYRQLAEQLPAYVKEL

E. PISIYEVHLGSWRRH TDNN

FWLSYRELADQLVPYAKWM

H.influenzae PISIYEVHLGSWRRN LENN FWLDYDQIADELIPYVKEM

Agro.tume PISIYEVHAGSWQR SEDG

TFLSWDELEAQLIPYCTDM

Aquifex_a. PISIYEVHVGSWRRVP EEGN

RFLSYRELAEYLPYYVKEM Synechococcus

PISVYEVHLGSWMHASSDAIATDAQGKPLPPVPVADLKPGARFLTYRELADRLIPY

VLDL

Synechocystis PVSVYELHLGSWLHTAYDEPVKTLHGEGVP-

VEVSEWNTGARFLTYYELVDKLIPYVKEL Butyrivibrio PIAIYECHIGSWMKHP DGTEDG

FYTYRQFADRIVEYLKEM

CHLAMYDIA PMNIYEVHVGSWRWQE

GQPLNYKELADQLALYCKQM

^•kit * -kit

BstearothermophUus

GFTHIELLPLVEHPLDRSWGYQGTGYYAVTSRYGTPHDFMYFVDRCHQAGIGVIM

DWVPG

Bcaldolyticus GFTHIELLPLVEHPLDRSWGYQGTGYYSVTSRYGTPHDFMYFVDRCHQAGLGVπ

DWVPG

BsubtUis

GFTHIELLPVYEHPYDRSWGYQGTGYYSPTSRFGPPHDLMKFVDECHQQNIGVIL

DWVPG mycobacterium

GFTHVELLPVAEHPFAGSWGYQVTSYYAPTSRFGTPDDFRALVDALHQAGIGVrVD

WVPA

Streptomyces GFTHVELMPVAEHPFGGSWGYQVTGFYAPTSRMGTPDDFRFLVDALHRAGIGVIV

DWVPA

E.

GFTHLELLPINEHPFDGSWGYQPTGLYAPTRRFGTRDDFRYFIDAAHAAGLNVILD

WVPG H.influenzae

GFTHIEFLPLSEFPFDGSWGYQPLGLYSPTSRFGSPEAFRRLVKRAHEAGINVILD

WVPG

Agro.tume

GFTHIEFLPITEHPYDPSWGYQTTGLYAPTARFGDPEGFARFVNGAHKVGIGVLLD WVPA

Aquifex_a.

GFTHVEFLPVMEHPFYGSWGYQITGYFAPTSRYGTPQDFMYLIDKLHQEGIGVILD

WVPS

Synechococcus GYSHIELLPIAEHPFDGSWGYQVTGYYAATSRYGSPEDFMYFVDRCHQNGIGVILD

WVPG

Synechocystis

GYTHIELLPIAEHPFDGSWGYQVTGYYAPTSRFGSPEDFMYFVDQCHLNGIGVIID

WVPG Butyrivibrio

KYTHIELIGIAEHPFDGSWGYQVTGYYAPTARYGEPTDFMYLINQLHKHGIGVILD

WVPA

CHLAMYDIA

HYTHVELLPVTEHPLNESWGYQTTGYYAPTSRYGSFEDLQYFIDTMHQHGIGVTL DWVPG

..-k.it.. . k -k it-k-k-kk .. k A -k . .. A . it ..iekkit BstearothermophUus

HFCKDAHGLYMFDGAPTYEYANEKDRENYVWGTANFDLGKPEVRSFLISNALFW

LEYYHI

Bcaldolyticus HFCKDAHGLYMFDGAPTYEYANEKDRENYVWGTANFDLGKPEVRSFLISNALFW

LEYYHV

BsubtUis

HFCKDAHGLYMFDGEPLYEYKEERDRENWLWGTANFDLGKPEVHSFLISNALY

WAEFYHI mycobacterium

HFPKDAWALGRFDGTPLYEHSDPKRGEQLDWGTYVFDFGRPEVRNFLVANALY

WLQEFHI

Streptomyces

HFPRDDWALAEFDGRPLYEHQDPRRAAHPDWGTLEFDYGRKEVRNFLVANAVY WCQEFHV

E.

HFPTDDFALAEFDGTNLYEHSDPREGYHQDWNTLIYNYGRREVSNFLVGNALYW

IERFGI

H.influenzae HFPSDTHGLVAFDGTALYEHEDPREGYHQDWNTLIYNYGRNEVKNFLSSNALYW

LERFGV

Agro.tume

HFPTDEHGLRWFDGTALYEHADPRQGFHPDWNTAIYNFGRIEVMSYLINNALYW

AEKFHL Aquifex_a.

HFPTDAHGLAYFDGTHLYEYEDWRKRWHPDWNSFVFDYGKPEVRSFLLSSAHF

WLDKYHA

Synechococcus

HFPKDGHGLAFFDGTHLYEHADSRQGEHREWGTLVFNYGRHEVRNFLAANALF WFDKYHI

Synechocystis

HFPKDGHGLAFFDGTHLYEHGDPRKGEHKEWGTLIFNYGRNEVRNFLVANALF

WFDKYHI Butyrivibrio

HFCPDEFGI CFDGTCIYEDPDPRKGEHPDWGT FNIiAKPEVKNFLIANALYWI RKFHI CHLAMYDIA HFPIDSFAMSGFDGTPLYEYTRNPSPLHPHWHTYTFDYAKPEVCNFLLGSVLFWI DKMHV

A* k . *^•*^■■*^■ -kk . -k . .. . -kit .* .#

BstearothermophUus DGFRVDAVANMLYWPNNDRL YE— - NPYAVEFLRKLNEAVFAYDPNALMIAED

Bcaldolyticus DGFRVDAVANMLYWPNNDRL YE—

NPYAVEFLRQLNEAVFAYDPNVWMIAED

BsubtUis DGFRVDAVANILYWPNQDER HT— -

NPYAVDFLKKLNQTMREAYPHVMMIAED mycobacterium

DGLRVDAVASMLYLDYSRPEGGWTPNVHGGRENLEAVQFLQEMNATAHKVAPGI

VTIAEE

Streptomyces

DGLRADAVASMLYLDYSRDEGDWSPNAHGGREDLDAVALLQEMNATVYRRFPGV VTIAEE

E.

DALRVDAVASMIYRDYSRKEGEWIPNEFGGRENLEAIEFLRNTNRILGEQVSGAVT

MAEE

H.influenzae DGIRVDAVASMΓYRDYSRAEGEWIPNQYGGRENLEAIEFLKHTNWKIHSEMAGAI

SIAEE Agro.tume

DGLRVDAVASMLYLDYSRKEGEWIPNEYGGRENLESVRFLQKMNSLVYGTHPGV

MTIAEE Aquifex_a. DGLRVDAVASMLYLDYSRKE-

WVPNIYGGKENLEAIEFLRKFNESVYRNFPDVQTIAEE

Synechococcus

DGIRVDAVASMLYLDYNRKEGEWIPNEYGGRENIEAADFLRQVNHLIFSYFPGALS

IAEE Synechocystis

DGMRVDAVASMLYLDYCREEGEWVANEYGGRENLEAADFLRQVNS YSYFPGI

LSIAEE

Butyrivibrio

DGLRVDAVASMLYLDYGKKDGQWVPNKYGDNKNLDAIEFFKHFNSVVRGTYPNI

LTIAEE

CHLAMYDIA

DGIRVDAVSSMLYLDYGRYAGEWVPNRYGGRENLDAIRFLQQFNTVIHEKYPGVL

TFAEE

BstearothermophUus

STDWPKVTAPTYEGGLGFNYKWNMGWMNDMLKYMETPPYERRHVHNQVTFSL

LYAYSENF Bcaldolyticus

STDWPRVTAPTYDGGLGFNYKWNMGWMNDMLKYMETPPHERKYAHNQVSFSL

LYAYSENF

BsubtUis

STEvvTQVTGAVEEGGLGFHYKWNMGWMNDVLKYMETPPEERRHCHQLISFSLL YAFSEHF mycobacterium

STPWSGVTRPTNIGGLGFSMKWNMGWMHDTLDYVSRDPVYRSYHHHEMTFSML

YAFSENY

Streptomyces STAWDGVTRPTDSGGLGFGLKWNMGWMHDTLRYVSKEPVHRKYHHHDMTFGM

VYAFSENF

E.

STDFPGVSRPQDMGGLGFWYKWNLGWMHDTLDYMKLDPVYRQYHHDKLTFGI

LYNYTENF H.influenzae

STSFAGVTHPSENGGLGFNFKWNMGWMNDTl^YMKLDPIYRQYHHNKMTFGM

VYQYSENF Agro.tume

STSWPKVSQPVHEGGLGFGFKWNMGFMHDTLSYFSREPVHRKFHHQELTFGLL

YAFTENF

Aquifex_a. STAWPMVSRPTYVGGLGFGMKWNMGWMNDTLFYFSKDPIYRKYHHEVLTFSIW

YAFSENF

Synechococcus

STSWPMVSWPTYVGGLGFNLKWNMGWMHDMLDYFSMDPWFRQFHQNNVTFSI

WYAFSENF Synechocystis

STSWPMNSWPTYVGGLGFNLKWNMGWMHDMLDYFSMDPWFRQFHQNSITFSM

WYNHSENY

Butyrivibrio

STAWPKVTAPPEEDGLGFAFKWNMGWMHDFCEYMKLDPYFRQGAHYMMTFAM SYNDSENY

CHLAMYDIA

STTFPKITVSVEEGGLGFDYKWNMGWMHDTLHYFEKDFPYRPYHQSDLTFPQW

YAFSERF

AA . .. AAAA AAA . A . * .* A A . . .A A .A . BstearothermophUus

ILPFSHDEWHGKKSLLNKMPGSYEEKFAQLRLLYGYMMAHPGKKLLFMGNEFA

QFDEWK

Bcaldolyticus

ILPFSHDEWHGKKSLLNKMPGSYEEKFAQLRLLYGYMMAHPGKKLLFMGSEFA QFDEWK

BsubtUis

VLPFSHDEWYGKKSLLNKMPGDYWQKFAQYRLLLGYMTVHPGKI^LIFMGSEFA

QFDEWK mycobacterium VLPLSHDEVVHGKGTLWGRMPGNNHVKAAGLRSLLAYQWAHPGKQLLFMGQEF

GQRAEWS

Streptomyces

VLPISHDEWHGKRSLVSKMPGDWWQQRATHRAYLGFMWAHPGKQLLFMGQEF

AQGSEWS E.

VLPLSHDEWHGKKSILDRMPGDAWQKFANLRAYYGWMWAFPGKKLLFMGNEF

AQGREWN

H.influenzae VLPLSHDEVVΉGKYSLLGKMPGDTWQKFANLRAYYGYMWGYPGKKLLFMGNEF

AQGREWN Agro.tume

VLPLSHDEVVHGKGSLLAKMSGDDWQKFANLRSYYGFMWGYPGKKLLFMGQEF

AQWSEWS Aquifex_a.

VLPLSHDEVVHGKGSLIGKMPGDY VQKFANLRALFGYMWAHPGKKLLFMGGEF

GQFKEWD

Synechococcus

MLALSHDEWHGKSNLIGKMPGDEWQKFANLRCLLGYMFTHPGKKTLFMGMEF GQWAEWN

Synechocystis

MLALSHDEVVHGKSNMLGKMPGDEWQKYANVRALFTYMFTHPGKKTMFMSME

FGQWSEWN

Butyrivibrio ILPLSHDEVVHLKCSMVEKMPGYKVDKYANLRVGYTYMFGHSGKKLLFMGQDF

GQEREWS

CHLAMYDIA

LLPFSHDEVVHGKRSLIGKMPGDAWRQFAQLRLLLGYQICQPGKKLLFMGGEFG

QGREWS

,k .-kkititic-k. * . .* * . k i: . kk. .itit .* k kk

BstearothermophUus FEDELDWVLFDF

ELHRICMNDYMKELIACYKRYKPFYELDHDPQGFEWIDVHNAEQ

Bcaldolyticus FAEELDWVLFDF ELHRKMDEYVKQLIACYKRYKPFYELDHDPRGFEWIDVHNAEQ

BsubtUis DTEQLDWFLDSF

PMHQKASVFTQDLLRFYQKSKILYEHDHRAQSFEWIDVHNDEQ mycobacterium EQRGLDWFQLDE— -

NGFSNGIQRLVRDINDIYRCHPALWSLDTTPEGYSWIDANDSAN Streptomyces

ETYGPDWWVLDSSYPAAGDHLGVRSLVRDLNRTYTASPALWERDSVPEGFAWVE

ADAADD

E. HDASLDWHLLEG— GDNWHHGVQRLVRDLNLTYRHHKAMHELDFDPYGFEWLWDDKER

H.influenzae YEESLDWFLLDENI-

GGGWHKGVLKLVKDLNQrYQKNRPLFELDNSPEGFDWLWDDAAN

Agro.tume EKGSLDWNLRQY

PMHEGMRRLVRDLNLTYRSKAALHARDCEPDGFRWLWDDHEN Aquifex_a. HETSLDWHLLEY

PSHRGIQRLVKDLNEVYRREKALHETDFSPEGFEWVDFHDWEK

Synechococcus VWGDLEWHLLQY

EPHQGLKQFVKDLNHLYRNAPALYSEDCNQAGFEWIDCSDNRH

Synechocystis VWGDLEWHLLNF PPHQQLKQFFTELNHLYKNEPALYSNDFDESGFQWIDCSDNRH

Butyrivibrio EKRELDWFLLEN

DLNRGMKDYVGKLLEIYRKYPALYEVDNDWGGFEWINADDKER

CHLAMYDIA PGRELDWELLDI

SYHQGVHLCSQELNALYVQSPQLWQADHLPSSFRWVDFSDNRN

.-k k . -k . -k.

BstearothermophUus SIFSFIRRGKKED-DVLVIVCNFTNQAYDDYKVGVP-

LLVPYREVLNSDAVTFGGSGHVN

Bcaldolyticus SIFSFIRRGKKEG-DVLVIVCNFTNQAYDDYKVSVP- LLAPYREVLNSDAAEFGGSGHVN

BsubtUis SIFSFIRYGQKHG-EALVπCNFTPWYHQYDVGVP-

FFTQYIEVLNSDSETYGGSGQIN mycobacterium NVLSFMRYGSDG-SVLACVFNFAGAEHRDYRLGLP-

RAGRWREVLNTDATIYHGSGIGN Streptomyces NVFAFLRFARDG-

SPLLCVSNFSPWRHGYRIGVPQEVGQWREVLNTDLEPYGGSGVHH

E. SVLIFVRRDKEG-NEIIVASNFTPVPRHDYRFGIN-

QPGKWREILNTDSMHYHGSNAGN H.influenzae SVLAFERRSSNG-ERIIWSNFTPVPRHNYRIGVN-

VAGKYEEILNTDSMYYEGSNVGN

Agro.tume SVFAWLRTAPGE-KPVAVICNLTPVYRENYYVPLG-

VAGRWREILNTDAEΓYGGSGKGN

5 Aquifex_a. SVISFLRKDKSGK-EIILWCNFTPVPRYDYRVGVP-

KGGYWREIMNTDAKEYWGSGMGN

Synechococcus SIVSFIRRAHESD-RFLWVCNFTPQPHAHYRIGVP-

VAGFYREIFNSDARSYGGSNMGN

Synechocystis SWSFIRRAKNSA-EFWTICNFTPQPHSHYRVGVP-

10 VPGFYTELFNSDARQYGGSNMGN

Butyrivibrio STYSFYRRASNGK-DNILFVLNMTPMERKGFKVGVP-

FDGTYTKILDSAKECYGGSGSSV

CHLAMYDIA GWAYLRFADADAKKALLCVHHFGVGYFPHYLLPIL-

PLESCDLLMNTDDTRFGGSGKGF ι c . -k . .. . . .... . " it

BstearothermophUus GKR-LSAFNEPFHGK P-

YHVRMTIPPFGISILRPVQKRGERKRNEK

Bcaldolyticus GKR-LPAFSEPFHGK P- 0 YHVRMTIPPFGISILRPVQKRGERKQNEE

BsubtUis KKP-LSAKKGALHHK P-

CYITMTIPPYGISILRAVKKRGEIKR— mycobacterium LGG-VDATDDPWHGR P-

ASAVLVLPPTSALWLTPA 5 Streptomyces ARA-LRPEPVPAQGR A-VSLRMTLPPMATVWLRP-

E. GGT-VHSDEIASHGR Q-HSLSLTLPPLATIWLVREAE

H.influenzae FGC-VASEQIESHGR E-NSISVSIPPLATVYLRLKTK- 0

Agro.tume GG— -RVQAVDAGG E-IGAMLVLPPLATIMLEPEN— -

Aquifex_a. LGG-KEADKIPWHGR K-FSLSLTLPPLSVIYLKHEG- Synechococcus LGG-KWTDEWSCHNR P-

YSLDLCLPPLTTLVLELASGPES--LS

Synechocystis LGG-KWTEEWSFHEQ P-

YSLDLCLPPLSVLVLKLSQNAEENTVPAE Butyrivibrio PDK-IKAVKGLCDYK D-

YSIEFDLPPYGAEVFVFQTKKTKN

CHLAMYDIA REPEILTPEIARQEREAAGLIEADDESGPDCWGLDIELPPSATLIFSVTLQ-

BstearothermophUus EMHRHVIGRRARKSASLADDKHR

Bcaldolyticus EVHRHVIGRRARKPASLADEKHRETSRAVWGEVPDH

BsubtUis mycobacterium

Streptomyces

E.

H.influenzae

Agro.tume

Aquifex_a.

Synechococcus EAANSPL

Synechocystis EASNIA

Butyrivibrio

CHLAMYDIA

Nucleotide sequence of T. thermophilus AMase

1 ATGGAGCTTC CCCGCGCTTT CGGTCTGCTT CTCCACCCCA CGAGCCTCCC CGGCCCCTAC 61 GGCGTCGGCG TCCTGGGCCA GGAGGCCCGG GACTTCCTCC GCTTCCTCAA GGAGGCAGGG

121 GGGCGGTACT GGCAGGTCCT CCCCTTGGGC CCCACGGGCT ATGGCGACTC CCCCTACCAG

181 TCCTTCAGCG CCTTCGCCGG AAACCCCTAC CTCATAGACC TGAGGCCCCT CGCGGAAAGG

241 GGCTACGTGC GCCTGGAGGA CCCCGGCTTC CCCCAAGGCC GGGTGGACTA CGGCCTCCTC

301 TACGCCTGGA AGTGGCCCGC CCTGAAGGAG GCCTTCCGGG GCTTCAAGGA AAAGGCCTCC 361 CCGGAGGAGC GGGAGGCCTT CGCCGCCTTC CGGGAGAGGG AGGCCTGGTG GCTCGAGGAC

421 TACGCCCTCT TCATGGCCCT GAAGGGGGCG CACGGGGGGC TTCCCTGGAA CCGGTGGCCC

481 CTTCCCCTGC GGAAGCGGGA AGAGAAGGCC CTTAGGGAGG CGAAAAGCGC CTTGGCCGAG

541 GAGGTGGCCT TCCACGCCTT CACCCAGTGG CTCTTCTTCC GCCAGTGGGG GGCCTTGAAG

601 GCGGAGGCCG AGGCGTTGGG CATCCGGATC ATCGGGGACA TGCCCATCTT CGTGGCCGAG 661 GACTCCGCCG AGGTCTGGGC CCACCCCGAG TGGTTTCACC TGGACGAGGA GGGCCGCCCC

721 ACGGTGGTGG CGGGGGTGCC CCCCGACTAC TTCTCGGAGA CGGGCCAGCG CTGGGGCAAC

781 CCCCTTTACC GCTGGGACGT TTTGGAGCGG GAGGGGTTCT CCTTCTGGAT CCGCCGTCTG

841 GAGAAGGCCC TGGAGCTCTT CCACCTGGTG CGCATAGACC ACTTCCGCGG CTTTGAGGCC

901 TACTGGGAGA TCCCCGCAAG CTGCCCCACG GCGGTGGAGG GGCGCTGGGT CAAGGCCCCG 961 GGGGAGAAGC TCTTCCAGAA GATCCAGGAG GTCTTCGGCG AGGTCCCCGT CCTCGCCGAG

1021 GACCTGGGGG TCATCACCCC CGAGGTGGAG GCCCTGCGCG ACCGCTTCGG CCTTCCCGGG

1081 ATGAAGGTCC TGCAGTTCGC CTTTGACGAC GGGATGGAAA ACCCCTTCCT CCCCCACAAC

1141 TACCCTGCCC ACGGCCGGGT GGTGGTCTAC ACCGGCACCC ACGACAACGA CACCACCCTG

1201 GGCTGGTACC GCACGGCCAC CCCCCACGAG AAGGCCTTCA TGGCGCGGTA CCTGGCGGAC 1261 TGGGGGATCA CCTTCCGGGA AGAGGAGGAG GTGCCCTGGG CCCTGATGCA CCTGGGGATG

1321 AAGTCCGTGG CCCGGCTCGC CGTCTACCCG GTGCAGGACG TCCTGGCCCT GGGCAGCGAG

1381 GCCCGGATGA ACTACCCGGG AAGGCCCTCG GGGAACTGGG CCTGGCGGCT CCTCCCGGGG 1441 GAGCTTTCCC CGGAGCACGG GGCGAGGCTT AGGGCCATGG CCGAGGCCAC GGAACGGCTC 1501 TAG

Amino acid sequence of T. thermophilus AMase

1 MELPRAFGLL LHPTSLPGPY GVGVLGQEAR DFLRFLKEAG GRYWQVLPLG

PTGYGDSPYQ 61 SFSAFAGNPY LIDLRPLAER GYVRLEDPGF PQGRVDYGLL YAWKWPALKE

AFRGFKEKAS

121 PEEREAFAAF REREAWWLED YALFMALKGA HGGLPWNRWP

LPLRKREEKA LREAKSALAE

181 EVAFHAFTQW LFFRQWGALK AEAEALGIRI IGDMPIFVAE DSAEVWAHPE WFHLDEEGRP

241 TWAGVPPDY FSETGQRWGN PLYRWDVLER EGFSFWIRRL EKALELFHLV

RIDHFRGFEA

301 YWEIPASCPT AVEGRWVKAP GEKLFQKIQE VFGEVPVLAE DLGVITPEVE

ALRDRFGLPG 361 MKVLQFAFDD GMENPFLPHN YPAHGRVWY TGTHDNDTTL

GWYRTATPHE KAFMARYLAD

421 WGITFREEEE VPWALMHLGM KSVARLAVYP VQDVLALGSE

ARMNYPGRPS GNWAWRLLPG

481 ELSPEHGARL RAMAEATERL

Nucleotide sequence of A. aeolicus MTase

1 ATGAGATTGG CAGGTATTTT ACTTCACGTA ACTTCACTTC CCTCTCCTTA

CGGGATAGGG 61 GATCTCGGAA AAGAAGCCTA CAGGTTTCTG GACTTCTTAA AGGAGTGCGG

TTTTAGCCTT 121 TGGCAGGTTC TACCTCTGAA CCCCACTTCA CTTGAGGCGG GAAACTCACC

CTACAGTTCA

181 AACTCCCTCT TCGCGGGCAA TTACGTACTA ATAGACCCTG AAGAATTATT GGAGGAGGAC

241 TTAATAAAAG AAAGGGACTT AAAAAGATTT CCCTTGGGTG AAGCCCTTTA

CGAAGTCGTG

301 TACGAGTATA AAAAAGAGTT GCTCGAAAAA GCCTTTAAAA ATTTCAGGAG

ATTTGAACTG 361 CTTGAAGATT TTCTGAAGGA ACACTCTTAC TGGCTCAGAG ATTACGCACT

TTACATGGCT

421 ATAAAAGAAG AAGAGGGAAA GGAGTGGTAT GAATGGGATG

AAGAATTGAA GAGGAGAGAA

481 AAAGAGGCTT TAAAAAGGGT GTTAAATAAG TTAAAGGGGA GGTTTTACTT CCACGTATTC

541 GTCCAGTTTG TTTTCTTCAA GCAGTGGGAA AAACTGAGAA GATACGCAAG

GGAAAGGGGG

601 ATAAGCATAG TTGGAGATCT TCCAATGTAC CCCTCGTACT CAAGTGCGGA

CGTGTGGACA 661 AATCCTGAAC TTTTTAAACT GGACGGAGAT TTAAAACCCC TTTTTGTAGC

GGGTGTTCCT

721 CCTGATTTTT TCAGTAAAAC GGGACAGCTG TGGGGAAATC CCGTTTACAA

CTGGGAAGAA

781 CACGAAAAGG AAGGCTTCAG ATGGTGGATA AGGAGAGTTC ATCACAACTT AAAACTCTTT

841 GACTTTTTAA GACTTGACCA CTTCAGGGGA TTTGAGGCGT ACTGGGAGGT

TCCTTACGGT

901 GAAGAAACTG CGGTAAACGG AAGGTGGGTA AAGGCTCCCG

GAAAGACACT ATTTAAAAAA 961 CTCTTATCAT ACTTCCCGAA GAACCCATTC ATAGCGGAGG ACTTAGGTTT

TATAACGGAC

1021 GAAGTGAGGT ACTTGAGGGA AACTTTTAAA ATCCCGGGAA

GCAGAGTTAT TGAGTTTGCC

1081 TTCTACGATA AGGAAAGTGA GCACCTTCCC CACAACGTTG AAGAGAACAA CGTTTACTAC

1141 ACTTCAACTC ATGACCTTCC TCCGATAAGA GGATGGTTTG AGAATTTAGG

AGAAGAATCA

1201 AGAAAACGAT TATTTGAATA CTTGGGAAGG GAGATTAAAG

AGGAAAAAGT TAACGAGGAG 1261 CTTATAAGAC TCGTTTTAAT CTCAAGGGCG AAGTTCGCAA TAATCCAGAT

GCAGGACTTA

1321 CTCAATCTCG GCAATGAAGC GAGGATGAAT TACCCCGGAA

GACCTTTCGG AAATTGGAGG

1381 TGGAGAATAA AGGAAGATTA CACACAAAAG AAGGAATTTA TTAAAAAACT CCTCGGAATT

1441 TACGGAAGAG AAGTTTAA Amino acid sequence of A aeolicus MTase

1 MRLAGILLHV TSLPSPYGIG DLGKEAYRFL DFLKECGFSL WQVLPLNPTS LEAGNSPYSS

61 NSLFAGNYVL IDPEELLEED LIKERDLKRF PLGEALYEW YEYKKELLEK AFKNFRRFEL

121 LEDFLKEHSY WLRDYALYMA IKEEEGKEWY EWDEELKRRE KEALKRVLNK LKGRFYFHVF 181 VQFVFFKQWE KLRRYARERG ISIVGDLPMY PSYSSADVWT NPELFKLDGD LKPLFVAGVP

241 PDFFSKTGQL WGNPVYNWEE HEKEGFRWWI RRVHHNLKLF DFLRLDHFRG FEAYWEVPYG

301 EETAVNGRWV KAPGKTLFKK LLSYFPKNPF IAEDLGFITD EVRYLRETFK IPGSRVIEFA

361 FYDKESEHLP HNVEENNVYY TSTHDLPPIR GWFENLGEES RKRLFEYLGR EIKEEKVNEE

421 LIRLVLISRA KFAIIQMQDL LNLGNEARMN YPGRPFGNWR WRIKEDYTQK KEFIKKLLGI 481 YGREV

Nucleotide sequence of A. aeolicus BE

1 ATGAAGAAGT TCAGTCTCAT CAGTGATTAC GACGTTTACC TCTTTAAGGA GGGAACGCAC 61 ACGAGACTTT ACGATAAACT TGGCTCCCAC GTTATAGAAC TAAACGGGAA AAGGTATACC

121 TTCTTTGCGG TTTGGGCACC CCACGCGGAT TACGTATCAC TTATAGGCGA TTTTAACGAA

181 TGGGATAAAG GTTCTACTCC CATGGTAAAG AGGGAGGACG GCTCCGGAAT ATGGGAGGTT

241 TTACTTGAAG GAGACCTGAC TGGTTCAAAG TACAAGTACT TTATAAAGAA CGGGAATTAC

301 GAAGTTGATA AGTCCGATCC CTTCGCATTT TTCTGTGAGC AACCCCCCGG AAACGCTTCC 361 GTAGTGTGGA AGCTCAATTA CAGGTGGAAC GACTCCGAAT ACATGAAAAA GAGGAAAAGA

421 GTAAACTCAC ACGACTCGCC TATATCCATA TACGAAGTTC ACGTGGGTTC TTGGAGGAGA

481 GTTCCAGAAG AGGGAAACAG ATTTTTGAGC TATAGGGAAC TTGCCGAATA CCTCCCATAC

541 TACGTAAAAG AGATGGGATT TACTCACGTT GAGTTCTTAC CCGTTATGGA ACATCCCTTT

601 TACGGCTCTT GGGGCTACCA GATAACGGGC TACTTCGCTC CGACTTCCAG ATACGGAACT 661 CCTCAGGACT TTATGTACTT AATAGACAAA CTTCATCAAG AAGGGATAGG TGTGATACTA

721 GACTGGGTTC CCTCTCACTT TCCCACCGAT GCCCACGGGC TCGCATACTT TGACGGGACT

781 CACCTTTACG AGTACGAGGA CTGGAGAAAG AGGTGGCATC CCGACTGGAA CAGCTTTGTT

841 TTTGATTACG GAAAACCGGA AGTTCGCTCC TTTCTCCTGA GTTCTGCCCA CTTCTGGCTC

901 GACAAGTACC ACGCAGACGG TCTCAGAGTG GATGCAGTTG CTTCAATGCT TTACCTAGAT 961 TACTCTAGGA AAGAATGGGT TCCAAACATA TACGGAGGGA AAGAAAACCT CGAGGCTATA

1021 GAATTCCTCA GGAAGTTTAA CGAAAGCGTT TACAGAAATT TTCCAGACGT CCAGACAATA

1081 GCGGAGGAAT CAACAGCCTG GCCTATGGTG TCCAGACCTA CATACGTGGG GGGACTGGGA

1141 TTTGGAATGA AGTGGAATAT GGGTTGGATG AACGACACAC TCTTTTACTT TTCAAAGGAT

1201 CCCATCTACA GGAAGTACCA CCATGAAGTC CTCACTTTCA GTATATGGTA CGCTTTTTCC 1261 GAGAACTTCG TCCTTCCACT ATCCCACGAT GAAGTTGTTC ACGGAAAGGG TTCTCTGATA

1321 GGGAAGATGC CAGGAGATTA CTGGCAGAAG TTTGCAAACC TTAGAGCCCT TTTCGGATAC

1381 ATGTGGGCAC ACCCAGGGAA AAAACTCCTC TTTATGGGGG GAGAGTTCGG ACAGTTTAAG 1441 GAATGGGATC ACGAAACGAG TCTCGACTGG CACCTCTTGG

AATACCCTTC TCACAGAGGT

1501 ATTCAGAGAT TAGTTAAGGA CTTAAACGAA GTTTACAGGA

GGGAAAAGGC TTTGCACGAA 1561 ACGGATTTTT CACCTGAGGG CTTTGAGTGG GTAGACTTCC

ACGACTGGGA AAAGAGCGTT

1621 ATATCCTTCT TGAGAAAGGA CAAAAGCGGT AAGGAAATTA TACTCGTAGT

TTGCAACTTC

1681 ACACCCGTTC CGAGATACGA TTACAGGGTA GGTGTACCGA AAGGCGGATA CTGGAGGGAG

1741 ATAATGAATA CCGATGCAAA GGAGTACTGG GGCTCCGGAA

TGGGAAATCT GGGTGGAAAA

1801 GAGGCTGATA AAATCCCGTG GCACGGAAGA AAATTCTCAC TTTCACTTAC

CCTGCCTCCC 1861 CTTTCCGTGA TCTATTTAAA GCACGAAGGA TGA

Amino acid sequence of A. aeolicus BE

1 MKKFSLISDY DVYLFKEGTH TRLYDKLGSH VIELNGKRYT FFAVWAPHAD YVSLIGDFNE

61 WDKGSTPMVK REDGSGIWEV LLEGDLTGSK YKYFIKNGNY EVDKSDPFAF

FCEQPPGNAS

121 WWKLNYRWN DSEYMKKRKR VNSHDSPISI YEVHVGSWRR

VPEEGNRFLS YRELAEYLPY 181 YVKEMGFTHV EFLPVMEHPF YGSWGYQITG YFAPTSRYGT PQDFMYLIDK

LHQEGIGVIL

241 DWVPSHFPTD AHGLAYFDGT HLYEYEDWRK RWHPDWNSFV

FDYGKPEVRS FLLSSAHFWL

301 DKYHADGLRV DAVASMLYLD YSRKEWVPNI YGGKENLEAI EFLRKFNESV YRNFPDVQTI

361 AEESTAWPMV SRPTYVGGLG FGMKWNMGWM NDTLFYFSKD

PΓYRKYHHEV LTFSIWYAFS

421 ENFVLPLSHD EWHGKGSLI GKMPGDYWQK FANLRALFGY

MWAHPGKKLL FMGGEFGQFK 481 EWDHETSLDW HLLEYPSHRG IQRLVKDLNE VYRREKALHE

TDFSPEGFEW VDFHDWEKSV

541 ISFLRKDKSG KEIILWCNF TPVPRYDYRV GVPKGGYWRE IMNTDAKEYW

GSGMGNLGGK

601 EADKIPWHGR KFSLSLTLPP LSVIYLKHEG

Claims

Claims

1. An isolated or recombinant nucleic acid derived from a nucleic acid encoding a polypeptide essentiaUy having alpha-glucanotransferase activity but having essentiaUy no hydrolysing activity, said isolated or recombinant nucleic acid encoding a polypeptide with hydrolytic activity.

2. A nucleic acid according to claim 1 wherein said transferase comprises amylomaltase or branching enzyme.

3. A nucleic acid according to claim 1 or 2 wherein said transferase comprises a thermostable transferase.

4. A nucleic acid according to anyone of claims 1 to 3 wherein said transferase is derived from a thermophilic micro-organism.

5. A nucleic acid according to claim 4 wherein said micro-organism comprises Thermus thermophilus, Thermus aquaticus or Aquifex aeolicus.

6. A nucleic acid according to anyone of claims 1 to 5 wherein said transferase is known under EC number 2.4.1.25 or 2.4.1.18

7. A nucleic acid according to anyone of claims 1 to 6 provided with a mutation leading to an alteration or loss of a codon originaUy encoding a hydrophobic amino acid located in or around a acceptor, a donor or a catalytic site extending from a TIM barrel structure of said transferase.

8. A nucleic acid according to claim 7 wherein said codon originaUy encoding a hydrophobic amino acid is altered into a codon encoding an amino acid which is substantiaUy less hydrophobic.

9. A nucleic acid according to claim 7 or 8 wherein said hydrophobic amino acid comprises phenylalanine, tryptophan or tyrosine.

10. A nucleic acid according to anyone of claims 7 to 9 wherein said hydrophobic amino acid is located at or around an amino acid position essentiaUy corresponding to amino acid position 54, 251, 258 or 366 of amylomaltase of Thermus thermophilus HB8.

11. A vector comprising a nucleic acid according to anyone of claims 1 to 10.

12. A host ceU comprising a vector according to claim 11 or a nucleic acid according to anyone of claims 1 to 10.

13. A method for providing a polypeptide or fragment thereof essentiaUy having alpha glucanotransferase acitivity but having essentiaUy no hydrolysing activity with hydrolysing activity said method comprising providing a nucleic acid encoding such a transferase with a mutation leading to an alteration or loss of a codon originaUy encoding a hydrophobic amino acid located in or around a acceptor, a donor or a catalytic site extending from a (alphalpha/betaeta)₈ barrel structure of said transferase.

14. A polypeptide, or an enzymaticaUy functional fragment thereof encoded by a nucleic acid according to anyone of claims 1 to 10 or obtainable by a method according to claim 13.

15. Use a polypeptide or fragment according to claim 14 in reducing retrogradation of starch.

16. Use according to claim 15 in reducing retrogradation of amylopectine.

17. Use according to claim 16 in reducing long-term retrogradation of amylpectine.

18. Use a polypeptide or fragment according to claim 14 in hydrolysing starch.

19. A method for reducing retrogradation of starch comprising treating said starch with a polypeptide or fragment according to claim 14.

20. A method for hydrolysing starch comprising treating said starch with a polypeptide or fragment according to claim 14.

21. A bakery ingredient comprising a polypeptide or fragment according to claim 14.

22. A bakery product such as bread comprising a polypeptide or fragment according to claim 14.