Modeling of α(1–4) chain arrangements in α(1–4)(1–6) glucans: The action and outcome of β-amylase and Pseudomonas stutzeri amylase on an α(1–4)(1–6) glucan model

Simulated enzymic debranching of a β-limit dextrin model, prepared from a computed construct made by random extension and branching, and given the CCL value of w-maize amylopectin (and equal amounts of external chains with ECL values of 2 and 3) has been related to experimental chromatograms of the debranched β-limit dextrin of the amylopectin. The profile was similar to those from gel chromatograms and IEC-PAD chromatography.The equivalent lengths in glucosyl units of grid-links (g-links) of internal and external chains in constructs were calculated from the ICL and ECL values of amylopectin and models produced from the constructs with the appropriate lengths for internal and external chains. These derived models were subjected to simulated hydrolysis by Pseudomonas stutzeri amylase and the products compared with those of the experimental distribution from w-maize amylopectin. With the model the amounts of maltotetraose and maltodextrins released were similar to the experimental values but the distribution of branched maltodextrins was quite different. Unlike w-maize amylopectin – a polymer with the cluster structure – which has given a profile of molecular sizes of maltodextrins with low amounts of single and small numbers of internal chains and with a peak at a M_W of about 14,000 (13 chains), in the model the proportion of maltodextrin with one internal chain was high and as d.p. increased the amounts decreased exponentially. This would be expected if the distribution of internal chains in the core was random. It is suggested that in the core of a model prepared from a construct made with alternating probabilities of extension – one in which this probability is high relative to branching, and a second in which it is low – may give clusters of branched maltodextrins with short internal chains which are joined by longer chains; more closely approximating the distribution of internal chains of different lengths in amylopectin.An arrangement for amylopectin molecules in the starch granule has been proposed. In this, they have a wafer-like, discoidal shape, composed of the amorphous zone overlain with the double helical, crystalline region. The flat macromolecules are concentrically layered with the former on the inside and the latter oriented to the outside of the granule.     

Multiple branching in glycogen and amylopectin

The β-amylase limit dextrins of glycogen and amylopectin are completely debranched by joint action of isoamylase and pullulanase. Action of isoamylase alone results in incomplete debranching as a consequence of the inability of this enzyme to hydrolyze those A-chains that are two glucose units in length (half the total number of A-chains). From the reducing powers released by isoamylase acting (a) alone and (b) in conjunction with pullulanase, the relative numbers of A- (unsubstituted) and B- (substituted) chains in the β-dextrins, and therefore in the native polysaccharides themselves, can be calculated. Examination of a series of glycogens and amylopectins in this way showed that the ratio of A-chains: B-chains is markedly higher in amylopectins (1.5–2.6:1) than in glycogens (0.6–1.2:1). Glycogen typically contains A-chains and B-chains in approximately equal numbers; amylopectin typically contains approximately twice as many A-chains as B-chains. These polysaccharides therefore differ in degree of multiple branching as well as in average chain length. A consequence of these findings is that amylopectin cannot be formed in vivo by debranching of a glycogen precursor, as proposed by Erlander, since it is impossible to increase the A:B chain ratio by action of a debranching enzyme.     

Hydrazinolysis and nitrous acid deamination of the carbohydrate moiety of alpha1-acid glycoprotein.

A limit dextrinase, free from contaminating carbohydrases, has been purified from malted sorghum flour. The enzyme readily hydrolysed α-limit dextrins having maltosyl or maltotriosyl side-chains, pullulan, and amylopectin β-limit dextrin. Glycogen β-limit dextrin and amylopectin were more slowly hydrolysed, the detection of the hydrolysis of amylopectin being dependent on enzyme concentration. No significant debranching of glycogen could be detected.     

A comparison of the size and shape of β-limit dextrin and amylopectin using pulsed field-gradient nuclear magnetic resonance and analytical ultracentrifugation

The size and shape of β-limit dextrin have been investigated by using pulsed, field-gradient nuclear magnetic resonance and analytical ultracentrifugation. In addition, the β-limit dextrin has been compared with the amylopectin from which it was derived by enzymic hydrolysis. When measuring size and shape, dimethyl sulfoxide was used as the solvent, in order to avoid problems of polymer agggregation. The results suggest that β-limit dextrin is an oblate ellipsoid with an axial ratio of ∼5:1, and the corresponding amylopectin molecule is even flatter. This indicates that the linear segments beyond the final branch-points of amylopectin lie in the plane of its branched core. The study also demonstrated that the density of packing of polymer chains in this branched core is much greater than at the periphery of amylopectin, and that the latter region is the location of the great majority of the nonreducing chains cleaved by beta amylase. Furthermore, the different sized molecules in amylopectin samples appear to undergo the same degree of degradation by this enzyme.     

The cluster structure of barley amylopectins of different genetic backgrounds

The unit chains of amylopectin are organized into clusters. In this study, the cluster structure was analysed in detail in four different genotypes of barley, of which two possessed the amo1 genetic background. Amylose content of the barley starches differed from 0 to 32.6%. Isolated amylopectin was hydrolysed with α-amylase from Bacillus subtilis into domains, defined as groups of clusters, which were size-fractionated by methanol. The domain fractions were further treated with α-amylase to release single clusters. Amylopectin, domains and clusters were subsequently treated with phosphorylase and β-amylase to produce φ,β-limit dextrins and the detailed internal structures of these different structure levels were investigated. Analysis was performed with gel-permeation and anion-exchange chromatography. Equal amount of A-chains were detected in all barleys, but the distribution of B-chains differed. At least two types of domain structures were identified in all four barley varieties. Large domains were built up by large clusters and small domains by small clusters. In all four barley samples the number of long chains was small suggesting that shorter chains with a degree of polymerization of 25-35 also are involved in the interconnection of clusters. The cluster structure of the amylopectin correlated with the genetic background. The two barley samples with amo1 genetic background possessed a more dense structure. Internal chain lengths in these two barleys were shorter resulting in larger domains built up by larger clusters.     

Roles of the N-terminal domain and remote substrate binding subsites in activity of the debranching barley limit dextrinase

Barley limit dextrinase (HvLD) of glycoside hydrolase family 13 is the sole enzyme hydrolysing α-1,6-glucosidic linkages from starch in the germinating seed. Surprisingly, HvLD shows 150- and 7-fold higher activity towards pullulan and β-limit dextrin, respectively, than amylopectin. This is investigated by mutational analysis of residues in the N-terminal CBM-21-like domain (Ser14Arg, His108Arg, Ser14Arg/His108Arg) and at the outer subsites +2 (Phe553Gly) and +3 (Phe620Ala, Asp621Ala, Phe620Ala/Asp621Ala) of the active site. The Ser14 and His108 mutants mimic natural LD variants from sorghum and rice with elevated enzymatic activity. Although situated about 40 Å from the active site, the single mutants had 15–40% catalytic efficiency compared to wild type for the three polysaccharides and the double mutant retained 27% activity for β-limit dextrin and 64% for pullulan and amylopectin. These three mutants hydrolysed 4,6-O-benzylidene-4-nitrophenyl-6³-α-d-maltotriosyl-maltotriose (BPNPG3G3) with 51–109% of wild-type activity. The results highlight that the N-terminal CBM21-like domain plays a role in activity. Phe553 and the highly conserved Trp512 sandwich a substrate main chain glucosyl residue at subsite +2 of the active site, while substrate contacts of Phe620 and Asp621 at subsite +3 are less prominent. Phe553Gly showed 47% and 25% activity on pullulan and BPNPG3G3, respectively having a main role at subsite +2. By contrast at subsite +3, Asp621Ala increased activity on pullulan by 2.4-fold, while Phe620Ala/Asp621Ala retained only 7% activity on pullulan albeit showed 25% activity towards BPNPG3G3. This outcome supports that the outer substrate binding area harbours preference determinants for the branched substrates amylopectin and β-limit dextrin.     

Growth Inhibition of Lactobacillus casei Phage J1 by Dehydroacetate

The fine structures of amylopectin and intermediate material characteristic of amylomaize starch were investigated by chemical and enzymatic means. In comparison with waxy-maize amylopectin, that of amylomaize starch was found to possess a img/ approximately 10 glucose units longer. Unit-chain profiles of waxy and amylomaize amylopectins revealed that the clear difference lay simply in the relative amounts of two unit-chain fractions. By fractionations of debranched β-limit dextrins, it was demonstrated that the img/ of the internal chains in amylomaize amylopectin was 9 glucose units longer than that in waxy-maize amylopectin. In addition, the proportion of maltose and maltotriose fractions in the debranched dextrin for amylomaize amylopectin was noticeably smaller than found for waxy-maize amylopectin. These data suggest a lesser branching frequency of outer branches in amylomaize amylopectin, confirming the previous proposal that this amylopectin has longer inner and outer branches than those of normal amylopectin.

As for amylomaize intermediate material, the average degree of polymerization was estimated to be 250 to 300 glucose units per molecule. It was also indicated that there were 5 or 6 glucose residues corresponding to the non-reducing end in the molecule. The unit-chain profile of the intermediate material implied that this molecule was mainly composed of branches with img/ around 50. Moreover, the presence of only small amounts of maltose and maltotriose fractions was demonstrated by the unit-chain distribution of this β-limit dextrin. These findings indicate that amylomaize intermediate material is totally consistent with a branched glucan having a low molecular weight, proposing that this anomalous glucan has such a fine structure that four or five branches with img/ around 50 are linked to a main linear chain of 100 to 150 glucose units.     

Synthesis of phenyl 2-acetamido-2-deoxy-4-O-α-l-fucopyranosyl-β-d-glucopyranoside and p-nitrophenyl 2-acetamido-2-deoxy-6-O-α-l-fucopyranosyl-β-d-glucopyranoside

Maize and potato amylopectin (57 and 64%, respectively) were recovered as non-cyclic products from 4-h digests of the starches with cyclodextrin glycosyltransferase {(1→4)-α-d-glucan:[(1→4)-α-d-glucopyranosyl]transferase (cyclising), EC 2.4.1.19} from Klebsiella pneumoniae M 5 al. Besides smaller saccharides, highly branched fragments of different sizes (average d.p. 40–140) were obtained by fractionation. The extents of beta-amylolysis varied between 24 and 37%, indicating that the clusters were not equally susceptible to attack by cyclodextrin glycosyltransferase. The fragments of potato amylopectin still contained larger amounts of material of high molecular weight. Accordingly, part of the longer B-chains of the basic structure were protected from the enzymic attack, presumably because of interchain branches. By debranching with pullulanase, it was evident that the beta-limit dextrins of the fragments of potato amylopectin were composed of longer B-chains (average chainlength 17.8) than those of maize amylopectin (average chain-length 14.1). The A/B-chain ratios, which were calculated from h.p.l.c. data for the debranched beta-limit dextrins, were 1.22 (maize) and 1.06 (potato). Some structural differences between potato and maize amylopectin are discussed.     

Structures of clusters in sweetpotato amylopectin

Sweetpotato amylopectin was subjected to partial hydrolysis by α-amylase from Bacillus amyloliquefaciens to release the clusters. Clusters were then fractionated and precipitated by methanol and structurally characterized by gel-permeation chromatography and high-performance anion-exchange chromatography. An initial stage of α-amylolysis on the amylopectin isolated mostly domains but also clusters. A second stage of α-amylolysis on the domains and clusters further isolated their respective clusters and sub-clusters. All the domains, clusters and sub-clusters were sequentially subjected to phosphorolysis and β-amylolysis to obtain their internal part. The degree of polymerization of the clusters in the form of φ,β-limit dextrins were from 58 to 86. Each domain contained 2–8 clusters. Two types of clusters were structurally identified. Type A clusters were larger and contained about 12 chains per cluster with higher degree of branching (DB), whereas those of type B were smaller and contained about eight chains per cluster with lower DB.     

Monte Carlo simulation of the α-amylolysis of amylopectin potato starch

The branched structure of potato amylopectin (degree of polymerization ~200,000) was modeled in a computer matrix. The chain-length distribution and the length and width of a cluster of the amylopectin molecule were used as input variables in the model. Independent literature values related to the structure of amylopectin (percentage #-hydrolysis and ratio of A- to B-chains) were used for evaluation of the branching characteristics (length of branch area and chance of branching) of the modeled amylopectin. The structural parameters predicted by the model agreed very well with data from the literature. The chain-length distribution and values for the percentage of #-hydrolysis were the two most important parameters required to model the structure of amylopectin. This computer-generated model of potato amylopectin in solution can be used to simulate various enzymatic (i.e., !-amylase, #-amylase, glucoamylase, pullunanase) or chemical reactions (i.e., acid hydrolysis, hypochlorite oxidation). The modeling approach described in this paper is also suitable for starches from other botanical sources (i.e., corn, wheat, tapioca).     

A comparative study of edible canna (Canna edulis) starch from different cultivars. Part II. Molecular structure of amylose and amylopectin

Molecular structures of starches isolated from Japanese-green, Thai-green and Thai-purple cultivars of edible canna (Canna edulis Ker) were investigated. The absolute amylose content ranged from 19 to 25%. Degrees of polymerization (DPn) values of amylose determined by fluorescence-labeling method were 1590 for Thai-purple, 1620 for Japanese-green and 1650 for Thai-green cultivars. Mole% of branched fraction of amyloses from edible canna starches examined by a HPLC system after β-amylolysis of labeled amyloses was 13–16%. Branch chain-length distributions of amylopectin analyzed by HPSEC after debranching with isoamylase, followed by fluorescence-labeling of unit chain, showed bimodal distribution with the DPn range of 25–28. The amylopectin of edible canna starches contained high amounts of organic phosphorus (391–420 ppm). The distribution profile of phosphorylated chains, separated from non-phosphorylated chains by DEAE-Sephadex A-50 chromatography, indicated that the phosphate groups were located mostly in long B-chains of amylopectin molecules.     

Hydrolysis of amylopectin by amylolytic enzymes: structural analysis of the residual amylopectin population

Amylopectin fine structures were studied following limited hydrolysis of gelatinised waxy maize starch by amylases with a different level of inner chain attack (LICA). This was done by size exclusion chromatography as well as by debranching the (partially hydrolysed) amylopectin samples and studying the size distributions of the released chains. α-Amylases from Bacillus amyloliquefaciens and Aspergillus oryzae, with a relatively high LICA, drastically altered amylopectin chain length distribution and reduced the amylopectin molecular size (MS) significantly even at a low to moderate degree of hydrolysis (DH). Porcine pancreatic α-amylase (PPA), with a rather low LICA but a high multiple attack action on amylose, reduced the amylopectin MS much slower. Following hydrolysis by PPA to a DH of 10% and enzymic debranching of the amylopectin residue, several subpopulations of chains consisting of 2-12 glucose units were detected, indicating a multiple attack action on the amylopectin side chains. During the early stages of hydrolysis, the maltogenic Bacillus stearothermophilus α-amylase (BStA) preferentially hydrolysed the exterior chains of amylopectin. However, during the later phases, BStA also hydrolysed inner chains, presumably with a high multiple attack action. The present results clearly show that different enzymes can be used for (limited) conversion of amylopectin into structures differing in molecular weight and chain length distributions.     

Enzyme-Synthesized Highly Branched Maltodextrins Have Slow Glucose Generation at the Mucosal α-Glucosidase Level and Are Slowly Digestible In Vivo

For digestion of starch in humans, α-amylase first hydrolyzes starch molecules to produce α-limit dextrins, followed by complete hydrolysis to glucose by the mucosal α-glucosidases in the small intestine. It is known that α-1,6 linkages in starch are hydrolyzed at a lower rate than are α-1,4 linkages. Here, to create designed slowly digestible carbohydrates, the structure of waxy corn starch (WCS) was modified using a known branching enzyme alone (BE) and an in combination with β-amylase (BA) to increase further the α-1,6 branching ratio. The digestibility of the enzymatically synthesized products was investigated using α-amylase and four recombinant mammalian mucosal α-glucosidases. Enzyme-modified products (BE-WCS and BEBA-WCS) had increased percentage of α-1,6 linkages (WCS: 5.3%, BE-WCS: 7.1%, and BEBA-WCS: 12.9%), decreased weight-average molecular weight (WCS: 1.73×10⁸ Da, BE-WCS: 2.76×10⁵ Da, and BEBA-WCS 1.62×10⁵ Da), and changes in linear chain distributions (WCS: 21.6, BE-WCS: 16.9, BEBA-WCS: 12.2 DP_w). Hydrolysis by human pancreatic α-amylase resulted in an increase in the amount of branched α-limit dextrin from 26.8% (WCS) to 56.8% (BEBA-WCS). The α-amylolyzed samples were hydrolyzed by the individual α-glucosidases (100 U) and glucogenesis decreased with all as the branching ratio increased. This is the first report showing that hydrolysis rate of the mammalian mucosal α-glucosidases is limited by the amount of branched α-limit dextrin. When enzyme-treated materials were gavaged to rats, the level of postprandial blood glucose at 60 min from BEBA-WCS was significantly higher than for WCS or BE-WCS. Thus, highly branched glucan structures modified by BE and BA had a comparably slow digesting property both in vitro and in vivo. Such highly branched α-glucans show promise as a food ingredient to control postprandial glucose levels and to attain extended glucose release.     

Purification and properties of starch hydrolyzing enzymes in mature roots of sugar beets

Mature roots of sugar beets, which accumulate large amounts of sucrose but not starch, nevertheless contained acid and neutral amylases, judging from their pH optima, as well as pullulanase. Acid and neutral amylases were partially purified by procedures including fractionation with ammonium sulfate, ion exchange column chromatography, and gel filtration. Acid amylase was classified as an exoamylase, since it produced only glucose from soluble starch, amylopectin. β-limit dextrin, and rabbit liver glycogen. Neutral amylase was classified as an endoamylase, since it liberated maltose as the main product plus a small amount of glucose and oligosaccharides, and was capable of hydrolyzing β-limit dextrin. Pullulanase was purified to apparent homogeneity by procedures including fractionation with ammonium sulfate, Diethylaminoethyl-cellulose column chromatography and affinity chromatography. Pullulanase was capable of hydrolyzing soluble starch, amylopectin, β-limit-dextrin, and pullulan. Debranching of amylopectin was further evident by an increase in extinction coefficient, and by a shift of λ_max from 530 to 560 nm when the debranched amylopectin formed a complex with I₂-KI.     

Comparison of the Unit-chain Distributions of Glycogens from Different Biological Sources,Revealed by Anion Exchange Chromatography

The precise distributions of α-1,4-unit-chains of several kinds of glycogens, average chain length 7–13, from different sources, e.g., mammals, shellfish, mushrooms, and microorganisms, were compared using high performance anion exchange chromatography (HPAEC). These glycogens were completely debranched, and a series of maltosaccharides derived from individual unit-chains were analyzed. Most glycogens had similar chromatographic profiles, with degrees of polymerization (DPs), from 4–5 to 35; oyster glycogen contained shorter chains of DP 2–3. On the other hand, molar-based distribution patterns of these glycogens appeared to be specific to their origins. Ratios of their A- to B-chains varied from 0.6: 1 to 1.2: 1, as estimated by quantitative HPAEC of maltosaccharides by debranching of the β-limit dextrins. Possible chain lengths of the exterior A-chains were also discussed.     

Fine structure characterization of amylopectins from grain amaranth starch

The aim of this study was to determine the fine structure of amylopectin from grain amaranth. Amaranthus amylopectin was hydrolyzed with α-amylase, and single clusters and a group of clusters (domain) were isolated by methanol precipitation. The domain and the clusters were treated with phosphorylase a and then β-amylase to remove all external chains, whereby the internal structure was obtained. The ,β-limit dextrins were analyzed on Sepharose CL 6B. The average DP (degree of polymerization) and peak-DP values of fractions of clusters were 57 and 82, respectively; the values of the domain were 137 and 309, respectively. The unit chain length profiles were analyzed by high-performance anion-exchange chromatography with pulsed amperometric detector (HPAEC–PAD). The results showed that the domain fraction contained 2.2 clusters, and single clusters were composed of 13 chains. The ,β-limit dextrins of the clusters were further hydrolyzed with α-amylase to characterize their building block composition. The average DP of the branched blocks was 11 and they contained on average 2.5 chains. Their average chain length, internal chain length, and degree of branching were approximately 4.3, 2.8, and 14, respectively. A cluster consisted of 6 branched blocks, and the internal chain length between the blocks was 6.8.     

A debranching enzyme deficiency in endosperms of the sugary-1 mutants of maize

Many of the sugary-1 mutants of maize (Zea mays L.) have the highly branched water-soluble polysaccharide, phytoglycogen, in quantities equal to or greater than starch as an endosperm storage product in mature seeds. We find that all sugary mutants investigated are deficient in debranching enzyme [α-(1, 6)-glucosidase] activity in endosperm tissue 23 days postpollination and suggest that this deficiency is the primary biochemical lesion leading to phytoglycogen accumulation in sugary endosperms. This would indicate that the amylopectin component of starch depends on an equilibrium between the activities of branching enzymes introducing α-1,6 branch points into the linear α-1,4 glucans and debranching enzymes. The debranching enzyme activities from nonsugary endosperms can be separated into three peaks on a hydroxyapatite column. The sugary endosperm extracts lack one of these peaks of activity while the other two fractions have much reduced activity. The embryos of developing seeds (23 days after pollination) from both sugary and nonsugary genotypes have equivalent debranching activity. The debranching enzyme activity of developing endosperms is proportional to the number of copies (0 to 3) of the nonmutant (Su) allele present suggesting that the Su allele may be the structural gene for this debranching enzyme, although this is not definitive. This identification of debranching enzyme activity as being the biochemical lesion in sugary endosperms is consistent with several previous observations on the mutant.     

Chemical and physical properties of ginkgo (Ginkgo biloba) starch

Starch isolated from mature Ginkgo biloba seeds and commercial normal maize starches were subjected to α-amylolysis and acid hydrolysis. Ginkgo starch was more resistant to pancreatic α-amylase hydrolysis than the normal maize starch. The chain length distribution of debranched amylopectin of the starches was analyzed by using high performance anion-exchange chromatography equipped with an amyloglucosidase reactor and a pulsed amperometric detector. The chain length distribution of ginkgo amylopectin showed higher amounts of both short and long chains compared to maize starch. Naegeli dextrins of the starches prepared by extensive acid hydrolysis over 12 days demonstrated that ginkgo starch was more susceptible than normal maize to acid hydrolysis. Ginkgo dextrins also demonstrate a lower concentration of singly branched chains than maize dextrins, and unlike maize dextrin, debranched ginkgo shows no multiple branched chains. The ginkgo starch displayed a C-type X-ray diffraction pattern, compared to an A-type pattern for maize. Ginkgo starch and maize starch contained 24.0 and 17.6% absolute amylose contents, respectively.     

Mode of action of glycogen branching enzyme from Neurospora crassa

Neurospora crassa branching enzyme [EC 2.4.1.18] acted on potato amylopectin or amylose to convert them to highly branched glycogen-type molecules which consisted of unit chains of six glucose units. The enzyme also acted on the amylopectin beta-limit dextrin, indicating that the enzyme acted on internal glucose chains as well as outer chains. By the combined action of N. crassa glycogen synthase [EC 2.4.1.11] and the branching enzyme, a glycogen-type molecule was formed from UDP-glucose. In the presence of primer glycogen, the glucose transfer reaction was accelerated by the addition of branching enzyme. On the other hand, the glucose transfer reaction by glycogen synthase did not occur without primers. When the branching enzyme was added, the glucose transfer occurred after a short time lag. This recovery of the glucose transfer reaction did not occur upon addition of the inactivated branching enzyme. The structure of the product formed by the combined action of the two enzymes was different from that of the intact N. crassa glycogen with respect to the distribution patterns of the unit chains.     

The building block structure of barley amylopectin

Amylopectin branchpoints are present in amorphous lamellae of starch granules and organised into densely branched areas, referred to as building blocks. One single amylopectin cluster contains several building blocks. This study investigated the building block structure of domains (groups of clusters) and clusters in four different barley genotypes. Two of the barleys possessed the amo1 mutation, Glacier Ac38 and the double recessive SW 49427 with both wax and amo1 mutations, and were compared with the two waxy type barleys Cinnamon and Cindy. A previous detailed study on these four barley genotypes showed that the amo1 mutation affected the internal structure of amylopectin as manifested in the composition of clusters. In this work the building blocks were isolated from domains and clusters by extensive treatment with liquefying α-amylase of Bacillus subtilis and structurally characterised with enzymatic and chromatographic techniques. The proportion of large building blocks with a high number of chains was increased in the amo1 barleys, and the chain length between the blocks was short, which explained the previous findings of large clusters with more dense structure in the amo1 amylopectins.