Deprecated: Creation of dynamic property cls_session::$session_data_table is deprecated in /www/sites/www.188bio.com/index/systems/cls_session.php on line 49
(PDF) Sabate188bio精品生物—专注于实验室精品爆款的电商平台 - 蚂蚁淘旗下精选188款生物医学科研用品
您好,欢迎您进入188进口试剂采购网网站! 服务热线:4000-520-616
蚂蚁淘商城 | 现货促销 | 科研狗 | 生物在线

(PDF) Sabate

RESEARCH PAPERCharacterization of hyperbranched glycopolymers producedin vitro using enzymesAgnès Rolland-Sabaté Sophie Guilois Florent Grimaud Christine Lancelon-Pin Xavier Roussel Sandrine Laguerre Anders Viksø-Nielsen Jean-Luc Putaux Christophe D’Hulst Gabrielle Potocki-Véronèse Alain BuléonReceived: 19 July 2013 /Revised: 19 September 2013 /Accepted: 27 September 2013#Springer-Verlag Berlin Heidelberg 2013Abstract Asymmetrical flow field flow fractionation (AF4)has proven to be a very powerful and quantitative method forthe determination of the macromolecular structure of highmolar mass branched biopolymers, when coupled withmulti-angle laser light scattering (MALLS). This work de-scribes a detailed investigation of the macromolecular struc-ture of native glycogens and hyperbranched α-glucans(HBPs), with average molar mass ranging from 2× 106to4.3× 107gmol−1, which are not well fractionated by meansof classical size-exclusion chromatography. HBPs were enzy-matically produced from sucrose by the tandem action of anamylosucrase and a branching enzyme mimicking in vitro theelongation and branching steps involved in glycogen biosyn-thesis. Size and molar mass distributions were studied byAF4, coupled with online quasi-elastic light scattering(QELS)andtransmissionelectronmicroscopy.AF4-MALLS-QELS has shown a remarkable agreement betweenhydrodynamic radii obtained by online QELS and by AF4theory in normal mode with constant cross flow. Molar mass,size, and dispersity were shown to significantly increase withinitial sucrose concentration, and to decrease when thebranching enzyme activity increases. Several populations withdifferent size range were observed: the amount of small sizemolecules decreasing with increasing sucrose concentration.The spherical and dense global conformation thus highlightedwas partly similar to native glycogens. A more detailed studyof HBPs synthesized from low and high initial sucrose con-centrations was performed using complementary enzymatichydrolysis of external chains and chromatography. It empha-sized a more homogeneous branching pattern than nativeglycogens with a denser core and shorter external chains.Keywords Hyperbranched α-glucan .Glycogen .Amylosucrase .Branching enzyme .Asymmetrical flow fieldflow fractionation .Hydrodynamic radius .Macromolecularsize distributionIntroductionGlycogen is, with starch, the major energy reserve in mostliving organisms [1,2]. Glycogen is made up of α-D-glucosylunits joined by α(1,4) glycosidic linkages, with 7–10 % α(1,6)branch points evenly distributed within the glycogen particles,Published in the topical collection Field-Flow Fractionation with guesteditors S. Kim R. Williams and Karin D. Caldwell.Electronic supplementary material The online version of this article(doi:10.1007/s00216-013-7403-2) contains supplementary material,which is available to authorized users.A. Rolland-Sabaté (*):S. Guilois :A. BuléonUR1268 Biopolymères Interactions Assemblages, INRA,F-44300 Nantes, Francee-mail: Agnes.Sabate@nantes.inra.frF. Grimaud :S. Laguerre :G. Potocki-VéronèseUniversité de Toulouse; INSA, UPS, INP; LISBP, 135 Avenue deRangueil, 31077 Toulouse, FranceF. Grimaud :S. Laguerre :G. Potocki-VéronèseINRA, UMR792 Ingénierie des Systèmes Biologiques et desProcédés, 31400 Toulouse, FranceF. Grimaud :S. Laguerre :G. Potocki-VéronèseCNRS, UMR5504, 31400 Toulouse, FranceC. Lancelon-Pin :J. L. PutauxCERMAV-CNRS (affiliated with Université Joseph Fourier, memberof Institut de Chimie Moléculaire de Grenoble and Institut CarnotPolyNat), BP 53, 38041 Grenoble cedex 9, FranceX. Roussel:C. D’HulstUGSF, UMR 8576 CNRS—Université Lille 1, sciences ettechnologies, Bât. C9, 59655 Villeneuve d’Ascq, FranceA. Viksø-NielsenNovozymes A/S, Krogshoejvej, 36, 2880 Bagsvaerd, DenmarkAnal Bioanal ChemDOI 10.1007/s00216-013-7403-2 leading to average branch chain length of 10–14 glucose resi-dues [3,4]. A glycogen-type polymer, referred to asphytoglycogen, has also been isolated from higher plants lack-ing an isoamylase-type starch debranching enzyme [5]. It has aregularly branched structure with α(1,6) linkages every 8–12glucose residues. Recently, a biomimetic system reproducingin vitro the activities involved in the formation of α(1,4) andα(1,6) glycosidic linkages during glycogen biosynthesis [6,7]was used to produce hyperbranched α-glucans from sucrose asunique substrate, in one step [8]. This method was based on theuse of two bacterial transglucosidases: the amylosucrase fromNeisseria polysaccharea (NpAS) and the branching enzyme(BE) from Rhodothermus obamensis (RoBE) [9]. The influ-ence of RoBE/NpAS activity ratio (aRoBE/aNpAS) and initialsucrose concentration on the structure of the resulting glucanwas studied in detail. The morphology of these hyperbranchedα-glucans seemed to be close to that of natural glycogens with10–13 % α(1,6) branch points and a branch chain length around10 [8], but the mechanisms of their in vitro synthesis appearedto be different from those involved in the biosynthesis of nativeones. Ciric and Loos also obtained branched α-glucans withtunable degree of branching, using α(1,4) glucan phosphory-lase for elongation and a BE from Deinococcus geothermalisfor branching [10].All these macromolecules have a high average molar massand a broad distribution of molar mass, which makes theirdetailed characterization difficult. The distribution broadnessof the molar mass can be approached by the ratio of thenumber-average molar mass (Mn) and the weight-averagemolar mass (Mw), i.e., the dispersity (Mw=Mn); howevercomplete distributions are more informative to depict the sam-ple content. Macromolecule size can be expressed by its radiuswhich can be defined in different forms [11,12]. Among them,radius of gyration (RG, (1)) which is geometrically defined andhydrodynamic radius (RH, (2)) which results frommacromolecule/solvent interactions are widely used [12,13].RG≡1=nþ1ðÞðÞXiri−RcmðÞ2hi1=2ð1Þwhere nis the number of bonds in the polymer backbone, rithe location of an individual group of atoms, and Rcmis thelocation of the center of mass of the particle.RH≡kBT=6πηDTð2Þwhere kBis Boltzman’sconstant,Tthe temperature, ηtheviscosity of the solvent, and DTthe translational diffusioncoefficient. RG/RHinforms on the conformation of macromol-ecules in solution [14,15].The polymer chain distribution is generally characterizedby high-performance size-exclusion chromatography(HPSEC) coupled with multi-angle laser light scattering(MALLS). The limiting factors for the fractionation of largeand branched polymers in HPSEC columns are (1) their lowexclusion limit regarding large polymer size, (2) the columnretention due to several interactions linked to the ramifica-tions, (3) the shear scission [16], and (4) the band broadeningwhich could cause limited resolution [17]. Asymmetrical flowfield flow fractionation (AF4), which is a liquid separationtechnique without stationary phase, limits the interactionswith the sample [18] and has already been used to investigatethe fractionation of branched polysaccharides of average mo-larmassupto108gmol−1, including starches and glycogens[12,19,20]. The separation takes place in a laminar flow andis caused by a flow perpendicular to the carrier flow called thecross flow (Fc). With AF4, there can be also some limitingfactors to the separation of such large branched macromole-cules, such as (1) retention on the membrane, (2) loss of someof the lower molar mass fractions through the membrane, and(3) possible hybrid (normal vs. steric elution) elution effectsthat could occur due to particularly high molecular sizes.Nevertheless, AF4 which separates without any shear scissionis the most powerful technique for separating branched mac-romolecules with molar mass of 107–108gmol−1and the onlyavailable for particles. AF4 selectivity is based on the diffu-sion coefficient. The hydrodynamic radius (RH)isthenthehard-sphere size related to this diffusion coefficient by theStokes–Einstein equation [18]. For branched polymers, unlikelinear chains, it is impossible to obtain a true molar massdistribution by size separation. The determination of sizedistributions is then preferred to the molar mass distributions[21,22]. According to the AF4 theory in the normal modeoperation, under which small particles are eluted first, theretention time (elution time) of the slice i(tri) is related tothe translational diffusion coefficient Diof the slice i[20,23]:t0tri≈6DiV0Fcw2ð3Þwhere t0is the void time(i.e., the time for the carrier solventtopass through the channel from inlet to outlet), V0the voidvolume (i.e., the geometric volume of the channel), and wthechannel thickness. trican also be expressed in the form of aretention level (RL=tri/t0) which characterizes the quality ofthe fractionation as it is directly linked to resolution, bandbroadening, and plate height. For an optimal fractionation, RLhas to be 5 to avoid low resolution and 40 to avoid exces-sive broadening and peak tailing [24]. From Eq. 3,itisthenpossible to have the relation linking DTand trand consecu-tively, the size (using Eq. 2)andtr. With AF4 working innormal mode, size distributions could then be obtained bydetermining the relationship between DTand treither by usingthe equation established for constant Fcmethods (Eq. 3)[19,23] and its derivatives for decaying Fc[25,26], or by meansof calibration curves [12]. Alternatively, RHcan beA. Rolland-Sabaté et al. determined by means of online quasi-elastic light scattering(QELS).The objective of the present study was to investigate deeplythe structural characteristics of several hyperbranched poly-mers (HBPs) enzymatically synthetized as described byGrimaud et al., using different sucrose concentrations andaRoBE/aNpAS. This knowledge is necessary for a better under-standing of the mechanisms involved in the one-pot synthesisof HBPs [8]. Our approach was based on the determination oftheir structural heterogeneity notably by determining theircomplete size distributions. For this aim, we used AF4-MALLS-QELS, and complementary enzymatic, chromato-graphic (high-performance anion-exchange chromatographycoupled with pulsed amperometric detection, HPAEC-PAD),and imaging techniques.Materials and methodsMaterialsN. polysaccharea amylosucrase (NpAS, Protein IDCAA09772.1) was produced as a fusion protein glutathione-S-transferase/amylosucrase, by recombinant Escherichia colistrain BL21 carrying the pGST-NpAS plasmid and purified aspreviously described [9]. R. obamensis branching enzyme(RoBE, Protein ID BAB69858.1) was kindly provided byNovozymes (Bagsvaerd, Denmark). The enzymes activitieswere determined as previously described [8]. Isoamylase fromPseudomonas sp. (210 U mg−1)andβ-amylase (2,484 U mg−1)fromBacillus cereus were purchased fromMegazyme International (Ireland). The β-amylase wasdesalted against 40 mM phosphate buffer, pH 7.2, by usingdialysis tubing cellulose membranes (MW 10233) purchasedfrom Sigma (Saint Quentin Fallavier, France).Oyster glycogen (OGLY) was purchased by Sigma (SaintQuentin Fallavier, France). Phytoglycogen (PHY) wasextracted from maize sugary-1 provided by INRA (PlantBreeding Department, Clermont-Ferrand, France) as previ-ously described [27]. Monodisperse latex spheres with diam-eters of 50, 100, and 200 nm (referred to L50, L100, andL200, respectively) were purchased from Duke Scientific(Palo Alto, California, USA).The water used for samples preparation was producedusing a RiOsTMand Synergy purification system (Millipore,Bedford, MA, USA).MethodsEnzymatic synthesisHBPs syntheses were carried out at 30 °C, in 50 mM Tris–HClbuffer, pH 7.0, containing NpAS, RoBE, and sucrose aspreviously described [8]. Two RoBE/NpAS activity ratios wereused (referred to as aRoBE/aNpASin the following). aRoBE/aNpAS=4 corresponded to a NpAS activity of 1 U mL−1anda RoBE activity of 4 U mL−1and aRoBE/aNpAS=19 to 1 and19 U mL−1for NpAS and RoBE, respectively. The impact ofsucrose concentration on size and structure of the synthesizedparticles was studied by using different initial sucrose concen-trations (Sucr): from 17.1 to 205.2 g L−1. However, in order tocheck the impact of higher aRoBE/aNpASon the glucan struc-ture, an experiment with aRoBE/aNpAS=19and51.3gL−1su-crose was also performed. The synthesized HBPs were sepa-rated from other reaction products (glucose, fructose, sucroseisomers, and maltooligosaccharides) by ethanol 70 % (v/v)precipitation and freeze-dried as previously described [8].HBPs global structure analysisMacromolecular characteristics of HBPs and glycogens weredetermined by AF4-MALLS-QELS. All α-glucans weredissolved at a concentration of 0.5 g L−1in Millipore waterand kept 16 h at room temperature. They were then filteredthrough 0.45-μm Durapore™membranes (Waters, Bedford,MA, USA) before injection into the AF4-MALLS-QELS de-vice. Sample recovery yields were calculated from the ratio ofthe mass before and after filtration, whereas elution recoveriescorresponded to the ratio of the mass eluted from the AF4channel (integration of the DRI signal) and the injected mass,determined using the sulfuric acid–orcinol colorimetric meth-od [12,19]. The AF4 equipment (Consenxus, Ober-Hilbersheim, Germany) and configuration was exactly thesame as previously described [12,19]. A Dawn®Heleos™MALLS system fitted with a K5 flow cell and a GaAs laser(λ=658 nm) from Wyatt Technology Corporation (SantaBarbara, CA, USA) and an RID-10A refractometer fromShimadzu (Kyoto, Japan) were used as detectors. OnlineQELS measurements were performed at 142.5° for a timeinterval of 7 s using a WyattQELS®system (Wyatt TechnologyCorporation) [15]. Before use, the mobile phase (Milliporewater containing 0.2 g L−1sodium azide) was carefullydegassed and filtered through Durapore GV (0.1 μm) mem-branes from Millipore and eluted at a flow rate of 1 mL min−1.The Fcwasinitiallysetat1mLmin−1and the channel flowrate (Fout) at 0.2 mL min−1for the sample introduction and therelaxation/focusing period. One hundred microliters of eachsample at approximately 0.2 mg mL−1was injected at0.1 mL min−1for 600 s (injected mass of 20 μg). After theinjection pump was stopped, the samples were allowed to relaxand focus for 60 s. For elution, Foutwas set at 1 mL min−1andFcwas maintained at 0.8 mL min−1for 2274 s. AF4 elutionmethod was optimized in order to determine the optimumseparation conditions for the synthetized HBPs (see ElectronicSupplementary Material, Section 1, including Fig. S1).Latex standards were diluted in the AF4 eluent.Characterization of hyperbranched glycopolymers Mi, the molar mass of the slice i, was calculated using theAstra®software (Wyatt Technology Corporation, version5.3.4.20 for Windows) as previously described [19]. A valueof 0.145 mL g−1was used as the refractive index increment(dn/dc) for glucans. The normalization of photodiodes wasachieved using a low molar mass pullulan standard (P20). TheRHof the slice i (RHi) was calculated using the well-knownStokes–Einstein relation, DTobtained from online QELSmeasurement, as previously described [20]. Mn, Mw,thedispersity (Mw=Mn), RGand RH(in nanometer) wereestablished by adjusting the integration edges in order to retainthe most reliable data.Fine structure analysis of HBPIn order to determine the structure of the more internal layers inHBPs, their external chains were hydrolyzed of by β-amylase.β-Amylase is an exoenzyme that releases maltose from thenon-reducing end of α-glucans by cleavage of α(1,4) linkages.This enzyme is unable to hydrolyze α(1,6) branch points andstops two to three residues before the branch points. Therefore,the resulting glycogens or HBPs β-limit dextrins do not containany outer chains. The purified α-glucans (500 mg) weredissolved in 90 % Me2SO (50 mg mL−1)for3daysatroomtemperature. The solution was diluted five times in phosphatebuffer at a final concentration of 40 mM, pH 7.2. The prepara-tion was incubated with B. cereus β-amylase (15 U mg−1)at40 °C for 24 h. At the end of the hydrolysis, the reactionmediumwasboiledinawaterbathfor10minat95°Ctoinactivate β-amylase and centrifuged for 10 min at 4 °C. Thesupernatant obtained after the first β-amylolysis was dialyzedagainst water for 24 h and then against 40 mM phosphate bufferat pH 7.2. For complete enzymatic hydrolysis, the β-limitdextrins were incubated with B. cereus β-amylase(15 U mg−1) at 40 °C for a further 24 h. At the end of thereaction, the reaction media wereboiledinawaterbathfor10 min at 95 °C. The final β-limit dextrins were dialyzedagainst water for 24 h and precipitated with 10 volumes ofethanol. The precipitate was isolated by centrifugation (10 min,10,000×g). The β-limit dextrins were washed three times with1 volume of absolute ethanol at 4 °C. The resulting precipitatewas resuspended in milliQ water and freeze-dried.Native and β-amylolyzed HBP17, HBP205, OGLY, andPHY were then submitted to isoamylase debranching to de-termine their chain length (CL) distribution by HPAEC-PADon a 4×250 mm Dionex Carbopac PA100 column, using thesame debranching method and HPAEC-PAD gradient asGrimaud et al. [8]. Debranched samples were then dilutedtwo times in 1 M NaOH (final concentration) and directlyinjected into the HPAEC-PAD system. The concentration ofeach chain was estimated by using the linear relationshipbetween the detector response per mole of α(1,4) chains andCL as previously described [28]. The linear curve coefficientswere determined from maltooligosaccharide standards of CLbetween 2 and 7 and were used for longer compound quanti-fication (see Electronic Supplementary Material Fig. S2).Number and weight-average branched chain length (BCLnand BCLw, respectively) and dispersity (d ¼BCLw=BCLn)were calculated as previously described [29].β-Amylolyzed HBP17, HBP205, OGLY, and PHY werealso injected in the A4F-MALLS-QELS setup.The amount of α(1–6) linkages in the native and β-amylolyzed glucans were also determined by1H NMR aspreviously described [8].Results and discussionDetermination of size distributionIn order to establish the size distributions according to onesingle parameter, the hydrodynamic radius, RH, obtainedcombining the Eqs. 2and 3(valid for separations operatingwith constant Fcand RL≥5) [20,23] and measured by onlineQELS was plotted versus tr. An effective channel thick-ness (w)of256μm was used for the calculations andwas determined as shown in the Electronic SupplementaryMaterial, Section 3.Figure 1, which shows evolution of RHas a function of tr,reported the following: (1) for latex spheres L50 and L100,experimental trcorresponding to the apex of differential re-fractometric index (DRI) and light scattering (LS) peaks andRHcertified by the supplier (which fitted very well the onlineQELS measurements, see Electronic Supplementary Material,Section 3); (2) for glycogens and HBPs, RHdetermined byQELS in stop flow mode at trcorresponding to the apex of theLS peak; (3) for the glycogens and HBPs, RHimeasured byonline QELS. All the reported values were an average of threemeasurements. RHof HBPs and glycogens was determinedusing online QELS measurements and QELS in stop flowmode (at trcorresponding to the apex of the LS peak): thetwo modes yielded identical values of DTand RH(Fig. 1)showing that the online determination of these parameters didnot introduce any deviation for these samples. The standarddeviation for RHvalues measured by online QELS was 2.8 %for RHbetween 5 and 40 nm and 3.4 % forRHbetween 40 and55 nm.RHcalculated from trusing Eqs. 2and 3with effectivechannel thickness of 256 μm (determined from latexes data)fitted very well the experimental RHdetermined by onlineQELS at small values of retention (Fig. 1). However, attr 18.5 min (RL 28), the measured RHbecame lower thanthe calculated RHat the same tr. The observed deviation be-tween measured and calculated RHfor tr 18.5 min was higherthan 5 % and increased up to 10 % for the larger species. It wasthen statistically significant. This deviation may be due to aA. Rolland-Sabaté et al. slight retention (as the elution recoveries were higher than 95 %)of the largest particles on the membrane, causing slight delay intheir elution.Consequently, making a calibration curve with RHmea-sured by online QELS was better to ensure a more accuratedetermination of the size distribution for samples exhibitingRL 28. Consequently, the trand RHvalues of the slices imeasured by online QELS corresponding to glycogens andHBPs (for RL≥5) were used to establish the following equa-tion (R2=0.9931):RHi ¼−0:0195 tri2þ2:6651 tri −0:5901 ð4ÞThe Eq. 4was subsequently used to determine the sizedistributions according to RH.Finally, use the online QELS allowed to confirm the effi-ciency of the AF4 retention Eqs. 2and 3for the determinationof the size distributions but also to point out some of theirlimits for hyperbranched α-glucans.Structural characterization of glycogens and HBPsEffect of the sucrose concentration and the ratio aRoBE/aNpASThe dissolution recoveries in water are presented in Table 1.These recoveries varied from 71 to 97 % and depended onSucr. They decreased when Sucr increased from 17.1 to51.3 g L−1and then increased when Sucr increased from51.3 to 205.2 g L−1. The residual insoluble fraction couldcontain linear glucose chains, synthetized by NpAS [8]andknown to be less soluble. At high Sucr (from 34.2 to102.6 g L−1), the concentrated linear chains produced byNpAS could precipitate before being used as substrate byRoBE, aRoBE/aNpASbeing too low to produce only solubleparticles. This hypothesis was supported by the fact that thesample prepared with higher aRoBE/aNpAS(HBP51BE19) hada dissolution recovery in water of 94 % (Table 1). It could befurther confirmed by redissolving the residual insoluble frac-tion in better solvants (e.g., KOH 0.1 M or DMSO with salts)andcheckingchain-lengthdistributionbySEC.However,thisstudy was focused on the analysis of the hyperbranchedTabl e 1 Syntheses conditions, dissolution recoveries, weight-average molar mass (Mw), z-average radius of gyration (RG), z-average hydrodynamicradius (RH), and dispersity index (Mw=Mn) determined by AF4-MALLS-QELS for HBPsReference Sucrose(g L−1)RoBE/NpASactivity ratioDissolutionrecovery (%) Mw10−6(g mol−1)RG(nm) RH(nm) Mw=MnHBP17 17.1 4 95.4 2.36 12.6 16.3 2.22HBP34 34.2 4 78.2 5.44 15.4 19.9 2.00HBP51 51.3 4 71.4 7.10 18.3 23.7 2.30HBP103 102.6 4 79.3 18.65 24.7 32.4 2.39HBP154 153.9 4 94.7 30.62 28.9 36.8 2.39HBP205 205.2 4 96.8 42.37 31.9 37.7 2.34HBP51BE19 51.3 19 93.9 3.42 ND 17.8 2.19Mw,RG,RH,andMw=Mnvalues were taken over the whole peak. The experimental uncertainties were 5 %ND not determined01020304050600 5 10 15 20 25 30RH(nm)tr(min)Fig. 1 Hydrodynamic radius (RH) versus retention time (tr). Experimen-tal tr(taken at the apex of DRI and LS peaks) and RHcertified by thesupplier for latex spheres L50 and L100 (black squares), glycogens, andHBPs RHvalues of the slices idetermined by online QELS (blue dash),RHof glycogens and HBPs determined by QELS in stop flow mode at trcorresponding to the apex of the LS peak (red circles); theoretical curveestablished from trusing Eqs. 2and 3with nominal channel thickness(brown thick line) and effective channel thickness of 256 μm(black thinline), and calibration curve (RHi=−0.0195 tri²+2.6651tri−0.5901,R2=0.9931) obtained by fitting experimental RHvalues of the slices idetermined by online QELS for glycogens and HBPs (thin gray line).Reported experimental values were an average of three measurements.Errors bars (in black) are showed for the trof latex spheres, and for theRHof glycogens and HBPs determined by QELS in stop flow modeCharacterization of hyperbranched glycopolymers fraction and not on the minor linear fraction. Consequently,the linear fraction was not analyzed.AF4 analyses of HBPs exhibited elution recoveries higherthan 95 %. The high sample recovery values indicated that thefractionation response was quantitative for all the samples.These dissolution and analysis mode were thus considered asto enable the representative structural characterization ofHBPs. The plots of LS signal versus trfor HBPs (Fig. 2a)showed a shift of the peak to higher trwith increasing Sucr,meaning that the higher the Sucr, the higher is the molecularsize. Size distributions established for each sample, by usingthe RHcalibration curves previously determined, are reportedin Fig. 2b.DRI=f(tr) elugrams (results not shown) and sizedistributions showed that (1) the LS maxima did not elute atthe same tras the modal fraction of the sample; (2) the peakbroadness increased with Sucr. When Sucr increased, the RHmodal values were shifted to higher values, as already ob-served for LS and DRI signals. The distributions were bimod-al for Sucr 34gL−1,withRHmodal values of 7–10 and 20–25 nm, and trimodal for Sucr ≥34 g L−1with RHmodal valuesof 5, 12, and 25–30 nm for HBP51; 5, 22, and 35 nm forHBP103, and 5, 22, and 45 nm for HBP154 and HBP205(Fig. 2b). These results agreed with the different size distribu-tionsobservedontransmissionelectronmicroscopy(TEM)micrographs. In fact, the average values of RHobtained byonline QELS were very close tothe average radii measured byparticle counting from TEM micrographs (see Electronic Sup-plementary Material Table S3). Moreover, the size distribu-tions obtained by TEM (see Electronic Supplementary Mate-rial Fig. S3) coincided, for the overall size range, with thosedetermined by AF4.Figure 2b clearly showed that the amount of RH≤13 nmfractions (including the 5 nm population) decreased withincreasing Sucr, while the amount of RH 13 nm increased.Fractions with RH 40 nm were only obtained for Sucr ≥102gL−1. These observations were in line with the work ofGrimaud et al. which showed that HBP17 contained a higheramount of low molar mass polymer and a lower amount ofhigh molar mass polymer than HBP205 [8]. These resultsshowed that a critical Sucr value had to be used to get an optimalchain length for BE activity and thus a quasi-complete transferof linear chains towards the branched particles with RH 30 nm,the BE affinity for branched substrate being higher than forlinear ones [9]. Fraction with RH≤13 nm did not correspondto linear chains. The molar mass of fractions with RH=5 and13 nm was 1.37× 105gmol−1(DP 846) and 1.48× 106gmol−1(DP 9115), respectively, whereas for linear chains with similarRH, molar masses of 4.00 ×104gmol−1(DP 247) and 2.50×105gmol−1(DP 1543) would be expected, respectively(according to the relation: RHw ¼2:60 10−2Mw0:50 deter-mined by Rolland-Sabate et al. [15] for linear amyloses inwater). These results were different from the observations madeby Ciric et al. [30] in the case of in vitro synthesis of HBPs usingglucose-1-phosphate as a substrate, phosphorylase, and D.geothermalis BE. They reported an increase of RH 15 nmchains (considered as quasi-linear ones) and a decrease ofRH 15 nm chains (considered as branched ones) when theglucose-1-phosphate concentration increased. This differencewas probably due to a different mechanism of polymerizationcatalyzed by the enzymes used.Mw, RGand RHincreased with Sucr from 2.36×106to4.24× 107gmol−1, 12.6 to 31.9 nm, and 16.3 to 37.7 nm,respectively (Table 1). When plotted as a function of the realsucrose amount available for the synthesis (Fig. 3), Mwincreased linearly with corrected Sucr and RGand RHlogarithmically. Corrected values of Sucr were obtained bysubtraction of the amount of oligosaccharides having a DP≤7(determined by Grimaud et al. [8] to be about 70.4 % of thetotal sugars at the end of the synthesis) and the amount ofinsoluble particles (i.e., dissolution recoveries in water). Asimilar behavior was observed for Mw, RG,andRHcorre-sponding to the population of HBP with the largest size (taken0 10203040500.00.10.20.30.40.50.60.70.80.91.01.10 5 10 15 20 25 30 35RLNormalized LS90 signalt0ab00 1020304050607010 20 30 40 50RL0.00.10.20.30.40.50.60.70.80.91 1.E+091.E+081.E+071.E+061.E+051.E+04Normalized chain concentrationMolar mass (g mol-1)RH (nm)tr (min)t0Fig. 2 Elugrams, size, and molar mass distributions for HBPssynthetized with various initial sucrose concentrations. HPB17 (red ),HBP34 (orange), HBP51 (yellow), HBP103 (green), HBP154 (brown ),and HBP205 (black) corresponding to initial sucrose concentrations of17.1, 34.2, 51.3, 102.6, 153.9, and 205.2 g L−1, respectively. Lightscattering signal at 90° (LS90) versus retention time (tri)(a), and DRIsignal and molar masses versus RH(b)A. Rolland-Sabaté et al. at the apex of the LS peak). This confirmed the linear increaseof Mwwith Sucr that had already beensuggested byGrimaudet al. [8]. Thus, the HBP molecular size and mass can be easilycontrolled by varying the Sucr.For each fraction, structural information can be determinedfrom the exponent νusing the power-law equations: RGi=KGMiνGand RHi=KHMiνH.νvalues depend on polymer shape,temperature, and polymer–solvent interactions: νG=νH=0.33for a compact sphere, 0.50–0.60 for a linear random coil, and1.00 for a rod. A very good fit of experimental data wasachieved by using these power law equations. HBPs νGandνHvalues ranged from 0.36 to 0.41 and 0.34 to 0.38, respec-tively (Table 2). All νvalues calculated were very similar andclose to the theoretical value for a sphere.The size distribution of HBP51BE19, synthesized with ahigher aRoBE/aNpAS, was very different from that of HBP51(Fig. 4). HBP51BE19 had smaller molar mass and size (Table 1)and a narrower distribution than HBP51. Its size distribution alsoshowed less molecules with RH≤13 nm and more moleculeswith RH 13 nm, but less molecules of RH 23 nm than HBP51.In this case, all linear chains were probably used by RoBE andthe RHpeaks at 10 and 18 nm also corresponded to branchedchains. This higher aRoBE/aNpASvalue led to reduced Mw, size,and dispersity. νHvalue (Table 2) was the same as HBP51 (0.37)which meant that even if increasing aRoBE/aNpASchanged thesize of the molecule it did not change its global conformation.The structure factor ρ¼RG=RHdid not vary much forHBPs (0.68–0.73, Table 2) and was close to the theoreticalvalues of a very dense spherical structure (0.778) or microgel(0.3–0.6) [31]. The branching degrees (BD) and BCLnwerecalculated by using the model previously established byGrimaud et al. [8] for the different HBPs (see Electronic Sup-plementary Material, Section 5). BCLnincreased and BDdecreasedslowlywhenincreasingSucr up to 102.6 g L−1andthen BCLndecreased and BD increased slowly up to205.2 g L−1(Table 2). The measured BD values for HBP17to HBP205 were very close to those calculated using theGrimaud et al. model [8], thus confirming its validity. Themeasured BD value for HBP51BE19 was higher than for theTabl e 2 Structural characteristics of HBPs: νG(slope of the loglog plotof radius of gyration versus molar mass), νH(slope of the loglog plot ofhydrodynamic radius versus molar mass), structure factor ρLS(=RG/RH),and apparent molecular densities (dGappLSand dHappLS) determined byAF4-MALLS-QELS; average chain length (BCLn) and branchingdegree (BD, percent of α(1,6)-linkages) determined experimentally byHPAEC-PAD and1H NMR, respectively; average chain length (BCLn)and branching degree (BD, percent of α(1,6)-linkages) calculated byusing the model previously established [8]Measured values Calculated valuesReference νGνHρLS=RG/RHdGappLS(g mol−1nm−3)dHappLS(g mol−1nm−3)BCLnaBD (%) BCLnBD (%)HBP17 0.36 0.34 0.68 543.2 171.5 6.9 12.6a6.0a11.8aHBP34 0.39 0.36 0.70 512.0 177.6 ND 11.0 6.6 10.7HBP51 0.41 0.37 0.69 445.9 144.1 ND 9.1 7.1 9.8HBP103 ND 0.37 0.70 432.6 147.5 ND 7.9 8.1 8.4HBP154 0.41 0.38 0.72 443.6 162.3 ND 8.4 8.3 8.6HBP205 0.38 0.38 0.73 529.1 203.9 8.3 10.6a7.8a10.7aHBP51BE19 ND 0.37 0.73 592.8 ND 7.2 13.2a47.1a24.5aνvalues were taken over the whole peak, the dGappLSand dHappLSand the structure factor ρLS(=RG/RH) were taken at the apex of the light scatteringpeak (i.e., the fraction corresponding to the largest sizes). dGappLSand dHappLSwere calculated using the following relations: dGapp=Mw/(4π/3) RG3and dHapp=Mw/(4π/3) RH3. The experimental uncertainties were 5 %ND not determinedaValues reported from [8]1015202530354045501.E+041.E+072.E+073.E+074.E+075.E+070 10203040506070Radii (nm)Molar mass (g mol-1)Corrected sucrose concentration used for HBP synthesis (g L-1)Fig. 3 Relationship between macromolecular values and corrected sucroseconcentration (corr Sucr) used for HBP synthesis. Mw(blue triangles), RG(red dots), and RH(purple dots). The equations corresponding to the fulllines were obtained by fitting the data: Mw¼7:04 105corr Sucr–5:63105R2¼0:9922;RG¼7:4103 ln corrSucrðÞþ1:3781 R2¼0:9874Þ;RH¼8:9531 ln corrSucrðÞþ2:3612 R2¼0:9855Characterization of hyperbranched glycopolymers other HBPs but the BD value calculated for HPB51BE19 waserroneous as it was out of the model range. The apparentmolecular density (Table 2) taken at the apex of the LS peak(dGappLSand dHappLS), i.e., corresponding to the largest sizefraction, was very high and similar for all HBPs, thusconfirming their similar and very densely branched structure.Nevertheless, dGappLSand dHappLSvaried as BD with Sucr,whereas νGvaries inversely, suggesting that HBP51, HBP103,and HBP154 had a slightly less densely branched structure. Asexpected, dGappLSmeasured for HBP51BE19 was higher thanfor the other HBPs, confirming its higher branching density, inline with its BD. As expected, νGvalue very slightly decreasedwhen BD value increased (Table 2). Nevertheless, the confor-mation plot showed a very similar behavior for all the HBPs(see Electronic Supplementary Material Fig. S4) and did notallow discriminating the HBPs according to their branchingdensity. The HPBs had very similar global conformations andthe slight differences in νGand density could be explained bydifferences in their internal structure.Indeed, the BD values decreased initially for Sucr up to103gL−1and increased for higher Sucr up to 205 g L−1.These variations were related to the enzymatic mechanism ofsynthesis. At low Sucr, the chains synthesized by NpAS werelonger and less concentrated than those synthesized at higherSucr [32]. This difference resulted in distinct aggregationkinetics and, consequently, products with different morphol-ogies. At medium Sucr (103 g L−1), the synthesis catalyzed byNpAS conducted to the production of high concentration oflong chains. In that case, the fraction of linear chains couldprecipitate before being used as substrate by RoBE. At low(17 g L−1) and high Sucr (205 g L−1), the conditions ofsynthesis were unfavorable to chain precipitation dueto the low concentration and the low length of chainssynthetized by NpAS, respectively. Increasing theamount of soluble chains allowed increasing the numberof chains that are used as substrate by RoBE, andconsequently, their BD value.In-depth analysis of HBPs structureIn order to determine in more details the internal structure ofHBPs, HBP17 and HBP205 were submitted to β-amylasehydrolysis which removed their external chains and comparedto glycogens.Firstly, the molar mass and size distributions of nativeHBP17, HBP205, OGLY, and PHY and their β-limit dextrins(referred to as HBP17-B, HBP205-B, OGLY-B, and PHY-B,respectively) were analyzed by AF4-MALLS-QELS. The dis-solution yields in water and elution recoveries were between 95and 100 %, and 94 and 100 %, respectively, giving a represen-tative response. Size distribution established for each sample byusing the RHcalibration curves previously determined are re-ported in Fig. 5. OGLY, OGLY-B, PHY, and PHY-B size distri-butions were monomodal (Fig. 5a) whereas those of HBP17,HBP17-B, HBP205, and HBP205-B were multimodal (Fig. 5b).As expected, the RHmodal values were shifted to lower valuesafter β-amylolysis. In particular, there were no more moleculeswith RH 20nminHBP17-BandwithRH 45nminHBP205-B because of their size reduction by β-amylase (Fig. 5b).Mw, RG,andRHdecreased from 1.49×107and 6.34×106gmol−1, from 22.0 to 18.9 nm and from 28.0 to 24.0 nm,for PHY and Mwand RHdecreased from 4.97× 106and 1.52×106gmol−1, and from 20.0 to 14.8 nm for OGLY (Table 3).All the νvalues calculated for HBPs, glycogens, and theirβ-limit dextrins were very close together and close to thetheoretical value for a compact sphere (0.33), except forOGLY and PHY-B. νvalues slightly decreased or remainedstable after β-amylolysis, excepted for PHY-B which exhibit-ed higher νvalues than PHY. ρvalues remained stablethroughout the distributions as well, around 0.788 for thetwo glycogens and HBPs (Fig. 5c), in line with theoreticalvalues for a compact spherical structure [31]. This meant thatthe structure of these polymers was dense, spherical, andremains the same whatever the size of the particle. PHYseemed to be less homogeneous as the higher νvalues ofPHY-B would account for a less dense internal structure thanHBPs. These results were in good agreement with dataobtained for glycogens and HBPs in the literature [8,12,15]. Except for the νHvalues of HBPs and OGLYobtained by HPSEC-MALLS: using AF4-MALLS theywere higher (0.34–0.39) than the values obtained usingHPSEC-MALLS (0.29–0.34). This could be due to anincomplete fractionation obtained by HPSEC: slicedispersity would induce lower νvalues, as was alreadyobserved for amylopectins [12].The percentage of β-amylolysis was evaluated by using thefollowing relation:Fig. 4 Elugrams, size, and molar mass distributions for HPBssynthetized with two aRoBE/aNpASratios. HBP51 (yellow)andHBP51BE19 (blue), corresponding to aRoBE/aNpASratios of 4 and 19,respectivelyA. Rolland-Sabaté et al. %β−AmylolysisMALLS ¼MnHBP−MnHBPBMnHBP100 ð5Þwhere MnHBP and MnHBPB are the Mnof HBP and itscorresponding β-limit dextrin, respectively, determined byAF4-MALLS. It was 60.1, 69.5, 44.4, and 54.6 % for PHY,OGLY, HBP17, and HBP205, respectively (Table 3). OGLY wasthe most susceptible to β-amylolysis which would account forlonger external linear chains. The β-amylolysis rate was higherfor HBP205 than for HBP17 which could be due either to morenumerous external linear chains or to longer external linearchains. In fact, the BD increased after β-amylolysisasexpected(Table 3), in line with the loss of external linear chains. The BDof HBP-Bs (23–24 %) were higher than those of OGLY-B andPHY-B (16–18 %), accounting for more densely branched inter-nal structures for HBPs compared to glycogens. The CL distri-butions obtained by HPAEC-PAD after debranching of the na-tive and β-amylolyzed glycopolymers are shown in Fig. 6.OGLY-B and PHY-B showed two maxima at DP 2 and DP 6andchainsuptoDP34–36, whereas HBP17-B and HBP205-Bshowed two maxima at DP 3 and DP 5 and no chains of DP 22.The CL distribution differences (Fig. 6)showedthatβ-limitdextrins had more very short chains (essentially DP 2-3 forOGLY-B and PHY-B, and DP 3-4 for HBP17-B and HBP205-B), in line with the hydrolysis specificity of the enzyme whichreduces the external chains to DP 2 and 3, and less long chainswith a DP 5 compared to the non β-amylolyzed samples. Theimportance of the negative peaks in the difference HBP205-HBP205-B compared to HBP17 showed that a higher amountand longer chains have been removed by the β-amylase inHBP205. This was in line with its higher susceptibility to β-amylolysis. For glycogens, the height of the negative peaks in thedifference native-β-amylolyzed glucopolymers was low, butlonger chains were concerned, meaning that glycogens lose lessbut longer chains than HBPs.The percentage of β-amylolysis was also evaluated byusing the following relation [33,34]:%β−AmylolysisNMR ¼ECL−2CL*100 ð6Þwhere CL and ECL are the average chain length and theaverage external chain length of HBP. CL was calculatedusing the equation: CL ¼100=BD (BD being determined byNMR) and ECL ¼CL−CLHBPB þ2, where CLHBPB is theaverage chain length of β-amylolyzed HBP. Finally, the inter-nal chain length of the polymers ICL was determined asfollows: ICL ¼CLECL−1.The%β-amylolysisNMRwas 56.7, 58.5, 47.9, and 54.1 % for PHY, OGLY, HBP17,and HBP205, respectively (Table 3). The % β-amylolysisNMRand % β-amylolysisMALLSvalues varied in the same way andwere in the same order of magnitude excepted for OGLY(Table 3). That might be due to the high dispersity of OGLYand the discrepancies between the averages obtained by A4F-MALLS, on one hand, and NMR, on the other hand.By examining the chain lengths reported in Table 3,onecould see that PHY and GLY had higher linear chain length(CL , ECL , and ICL ) than HBPs, meaning that they were lessabcFig. 5 AF4 elugrams (DRI responses) and molar masses versus calcu-lated RHfor native and β-amylolyzed natural glycogens (a)andHBPs(b), and structure parameter ρ(=RG/RH) versus molar mass (c). a,bThethin lines represent DRI responses and the thick lines the molar massesfor PHY (yellow), PHY-B (orange), OGLY (green ), OGLY-B (blue),HPB17 (red ), HBP17-B (brown), HBP205 (black), and HBP205-B(gray). cDotted and full lines correspond to ρof 0.9 (soft sphere)and0.788 (compact sphere), PHY (yellow ), PHY-B (orange), OGLY(green), OGLY-B (blue), HPB17 (red ), HBP17-B (brown), HBP205(black), and HBP205-B (gray)Characterization of hyperbranched glycopolymers densely branched overall the particle. Among glycogens,PHY exhibited the lowest BD (determined by NMR) and thehighest CL , ECL , and ICL , in line with BCLnof PHY andPHY-B determined by HPAEC-PAD but unexpectedly withregard to the higher % β-amylolysis observed for OGLY.Moreover, % β-amylolysis, CL , ECL , and ICL values werein the same range but not completely in line with literaturedata reporting about glycogens [33,34], especially concerningthe % β-amylolysis, which had been reported to be muchsmaller for oyster glycogen and phytoglycogen. This couldbe due to different extraction conditions of glycogens. AmongHBPs, HBP205 exhibited the lowest BD and the highest CL ,ECL , and ICL . HBP17 had then the more densely branchedstructure.These glycopolymers are constituted by two different kindof chains: A chains do not bear any other chain whereas Bchains bear at least one other chain [33,34]. The values of A/Bchain ratio could give insight of the structure of the HBPs andwere determined from NMR and HPAEC-PAD data as previ-ously described [34] by using the following relations:a¼%DP2þDP3ðÞ1002:5CL 100−%βamylolysisðÞ100 ð7ÞA:B chain ratio ¼a1−að8ÞThe A/B chain ratio was 0.80, 0.87, 1.00, and 0.83 forHBP17, BHP205, PHY, and OGLY, respectively. These lasttwo values were in agreement with literature data reported forglycogens [34]. PHY had thus the highest amount of A chainswhereas HBP205 had more A chains than HBP17.The higher νvalues, the higher % β-amylolysis, the higheramount, and longer chains removed by the β-amylase shownby HPAEC-PAD analyses of HBP205 compared to HBP17were in agreement with a higher amount and longer externalchains in HBP205 compared to HBP17. Moreover, the BDand ICL of β-amylolyzed HBP17 and HBP205 were veryclose, meaning that they have a similar internal density, andthe external layer appeared to be less dense for HBP205 thanfor HBP17 confirming the hypothesis of longer chains in theexternal layer of HBP205.The in-depth analysis of structural characteristics ofthese newly synthetized polymers allowed to show theirspherical and dense global conformation by analogy toglycogens and also to point out some important differ-ences among glycogens and HBP fine structure. First,HBPs exhibited not only a more dense branched struc-ture than OGLY and PHY but also a more homogeneousbranched structure. Second, they had shorter externalchains. In particular, a large difference between theexternal (9.8–10.3) and the internal (2.5–3.4) chainlength was observed for OGLY and PHY, whereas HBPsTabl e 3 Structural characteristics of native and β-amylolyzed HBPs andglycogens: weight-average molar mass (Mw), z-average radius of gyration(RG), z-average hydrodynamic radius (RH), dispersity index (Mw=Mn), nG(slope of the loglog plot of radius of gyration versus molar mass) and nH(slope of the loglog plot of hydrodynamic radius versus molar mass)determined by AF4-MALLS-QELS, amounts of β-amylolysis, averagechain length (BCLn) and branching degree (BD, percent of α(1,6)-linkages) determined experimentally by HPAEC-PAD and1H NMRrespectively, and calculated chain lengths (CL,ECL,ICL)Reference Mw10−6(g mol-1)RG(nm) RH(nm) Mw=MnνGνH%β-amylolysisMALLS BCLnBD (%)NMR%β-amylolysisNMR CL ECL ICLHBP17 2.36 12.6 16.3 2.22 0.36 0.34 44.4 6.9a12.6a47.9 7.9 5.8 1.1HBP17-B 1.38 ND 13.8 1.97 0.36 0.33 –4.2 24.2 –4.1 2.0 1.1HBP205 42.37 31.9 37.7 2.34 0.38 0.38 54.6 8.3a10.6a54.1 9.4 7.1 1.3HBP205-B 11.96 24.3 28.6 1.80 0.37 0.35 –4.2 23.1 –4.3 2.0 1.3PHY 14.90 22.0 28.0 1.34 0.38 0.34 60.1 10.6a6.8a56.7 14.7 10.3 3.4PHY-B 6.34 18.9 24.0 1.39 0.42 0.36 –5.8 15.7 –6.4 2.0 3.4OGLY 4.97 15.1 20.0 1.27 0.51 0.39 69.5 11.0a7.5a58.5 13.3 9.8 2.5OGLY-B 1.52 ND 14.8 1.28 ND 0.36 –5.2 18.1 –5.5 2.0 2.5The Mw,Mw=Mn,RG,RH,andνvalues were taken over the whole peak. Their experimental uncertainties were 5 %.cThe β-amylolysis percentagewas evaluated by using the following relation: %β−AmylolysisMALLS ¼MnHBP−MnHBPBMnHBP 100 .WhereMnH BP and MnHBPB are the Mnof HBP and itscorresponding β-limit dextrin, respectively, determined by A4F-MALLS. The average polymer chain length (CL ) was calculated using the equation:CL ¼100=BD . The average external chain length of the polymers ECL:ECL ¼CLHBP−CLHBPB þ2. Finally, the internal chain length of thepolymers ICL was determined as follows: ICL ¼CL−ECL−1[33,34]. The amount of β-amylolysis was also evaluated by using the followingrelation: %β−AmylolysisNMR ¼ECL−2CL *100ND not determinedaValues reported from [8]A. Rolland-Sabaté et al. .CL concentration CL concentrationCL concentration00,050,10,15CL concentration00,050,10,15CL concentration00,050,10,1500,050,10,1500,050,10,150,000,050,100,1500,050,10,15-0,06-0,010,040,090,1400,050,10,15-0.06-0.010.040.090.14-0,06-0,010,040,090,14-0,06-0,010,040,090,14CL concentrationCL concentrationCL concentrationChain length Chain length Chain length Chain length Chain length Chain length Chain length Chain length Chain length Chain length Chain length Chain length CL concentrationCL concentrationCL concentrationCL concentrationCL7CL22CL7CL22CL7CL22CL7CL22CL7CL22CL7CL22CL7CL22CL7CL22CL7 CL22CL7 CL22CL7 CL22CL7 CL22........................................abcdefghijklFig. 6 Chain length (CL) distribution of native glycogens andhyperbranched products before and after β-amylolysis. Chain length(CL) distribution was determined after deramification by isoamylaseand HPAEC-PAD analysis. aCL of native HBP205. bCL of HBP205-B. cDifference plots between diagrams shown in band a(profile b–profile a). dCL of native HBP17. eCL of HBP17-B. fDifference plotsbetween diagrams shown in eand d(profile e–profile d). gCL of nativeOGLY. hCL of OGLY-B. iDifference plots between diagrams shown inhand g(profile h–profile g). jCL of native PHY. kCL of PHY-B. lDifference plots between diagrams shown in kand j(profile k–profile j).The concentration (in milligram per milliliter) of each chain is expressedwith respect to 1 mg of total productCharacterization of hyperbranched glycopolymers exhibited a more homogeneous branched structure withadensercore(ICL , 1.1–1.3; ECL , 5.8–7.1; Table 3).General conclusionsAF4-MALLS has proven to be very powerful for the deter-mination of the macromolecular structure of high molar massbranched biopolymers [12,19,20]. In this work, it has beenapplied very successfully to the determination of the molarmass and size distributions of a series of glycogens and HBPsproduced in vitro by enzymatic synthesis, and presentingvarious size and branching degree. Agreement between ex-perimental data, obtained by online QELS, and AF4 theory innormal mode with constant cross flow, was noteworthy. AF4-MALLS-QELS coupling allowed to confirm the efficiency ofthe AF4 retention equations for the calculation of the size ofglycopolymers and to highlight some of their limits.For HBPs, the solubility, branching degree, molar mass,and size distributions were shown to greatly depend on theSucr and the aRoBE/aNpASratio. Mwincreases linearly withSucr with a joint increase in molecular size, whereas Mw,size, and dispersity decrease with increasing aRoBE/aNpAS.Amore detailed structural characterization of native glycogensand two HBPs synthesized using low and high Sucr was doneby combining AF4 with enzymatic and other chromatographictechniques. Although glycogen-like particles were observedin all cases, we outlined differences in terms of branchingpattern in relation with Sucr and glycogen source.This work confirms the very high efficiency of AF4 for thefractionation of high molar mass and highly branched α-glucans with different size and branching scheme, and thedetermination of their size distribution and structure in solu-tion. AF4-MALLS-QELS coupling is therefore a very prom-ising tool for the thorough determination of the macromolec-ular characteristics, which can be applied to many other highmolar mass native or enzymatically synthesized biopolymersin the growing field of green chemistry.Acknowledgments The authors thank the Agence Nationale de laRecherche for financial support (grant number ANR-09-CP2D-07-01),as well as Marion De Carvahlo, Bruno Pontoire and Nelly Monties fortechnical assistance.References1. Buleon A, Colonna P, Planchot V, Ball S (1998) Int J Biol Macromol23:85–1122. Ball SG, Morell MK (2003) Annu Rev Plant Biol 54:207–2333. Shearer J, Graham TE (2002) Can J Appl Physiol 27:179–2034. Manners DJ (1991) Carbohydr Polym 16:37–825. Lin Q, Huang B, Zhang M, Zhang X, Rivenbark J, Lappe RL, JamesMG, Myers AM, Hennen-Bierwagen TA (2012) Plant Physiol 158:679–6926. James MG, Denyer K, Myers AM (2003) Curr Opin Plant Biol 6:215–2227. Kotting O, Kossmann J, Zeeman SC, Lloyd JR (2010) Curr OpinPlant Biol 13:321–3298. Grimaud F, Lancelon-Pin C, Rolland-Sabate A, Roussel X,Laguerre S, Vikso-Nielsen A, Putaux J-L, Guilois S, BuleonA, D’Hulst C, Potocki-Veronese G (2013) Biomacromolecules14:438–4479. Roussel X, Lancelon-Pin C, Vikso-Nielsen A, Rolland-Sabate A,Grimaud F, Potocki-Veronese G, Buleon A, Putaux JL, D’Hulst C(2013) BBA General Subjects 1830:2167–217710. Ciric J, Loos K (2013) Carbohydr Polym 93(1):31–3711. Burchard W (1999) Adv Polymer Sci 143:113–19412. Rolland-Sabate A, Guilois S, Jaillais B, Colonna P (2011) AnalBioanal Chem 399:1493–150513. Striegel AM (2005) Anal Chem 77:104A–113A14. Morris GA, Ang S, Hill SE, Lewis S, Schafer B, Nobbmann U,Harding SE (2008) Carbohydr Polym 71:101–10815. Rolland-Sabate A, Mendez-Montealvo MG, Colonna P, Planchot V(2008) Biomacromolecules 9:1719–173016. Cave RA, Seabrook SA, Gidley MJ, Gilbert RG (2009)Biomacromolecules 10:2245–225317. Vilaplana F, Gilbert RG (2010) J Sep Sci 33:3537–355418. Martin M (1998) Adv Chromatogr 39:1–3819. Rolland-Sabate A, Colonna P, Mendez-Montealvo MG, Planchot V(2007) Biomacromolecules 8:2520–253220. Wittgren B, Wahlund KG (1997) J Chromatogr A 791:135–14921. Gaborieau M, Gilbert RG, Gray-Weale A, Hernandez JM,Castignolles P (2007) Macromol Theory Simul 16:13–2822. Gidley MJ, Hanashiro I, Hani NM, Hill SE, Huber A, Jane J-L, LiuQ, Morris GA, Rolland-Sabate A, Striegel AM, Gilbert RG (2010)Carbohydr Polym 79:255–26123. Wittgren B, Wahlund KG, Derand H, Wesslen B (1996) Macromol-ecules 29:268–27624. Wahlund KG (2013) J Chromatogr A 1287:97–11225. Hakansson A, Magnusson E, Bergenstahl B, Nilsson L (2012) JChromatogr A 1253:120–12626. Magnusson E, Hakansson A, Janiak J, Bergenstahl B, Nilsson L(2012) J Chromatogr A 1253:127–13327. Rolland-Sabate A, Colonna P, Potocki-Veronese G, Monsan P,Planchot V (2004) J Cereal Sci 40:17–3028. Koch K, Andersson R, Aman P (1998) J Chromatogr A 800:199–20629. Montesanti N, Veronese G, Buleon A, Escalier P-C, Kitamura S,Putaux J-L (2010) Biomacromolecules 11:3049–305830. Ciric J, Oostland J, de Vries JW, Woortman AJJ, Loos K (2012) AnalChem 84:10463–1047031. Burchard W (1983) Adv Polym Sci 48:1–12432. Potocki-Veronese G, Putaux J-L, Dupeyre D, Albenne C, Remaud-Simeon M, Monsan P, Buleon (2005) Biomacromolecules 6:1000–101133. Calder PC (1991) Int J Biochem 23:1335–135234. Yun SH, Matheson NK (1993) Carbohydr Res 243:307–321A. Rolland-Sabaté et al.Citations (0)References (39)ResearchGate has not been able to resolve any citations for this publication.Starch granules: structure and biosynthesisArticleFull-text availableSep 1998INT J BIOL MACROMOL Alain Buléon Paul ColonaVéronique Planchot Steven G BallThe emphasis of this review is on starch structure and its biosynthesis. Improvements in understanding have been brought about during the last decade through the development of new physicochemical and biological techniques, leading to real scientific progress. All this literature needs to be kept inside the general literature about biopolymers, despite some confusing results or discrepancies arising from the biological variability of starch. However, a coherent picture of starch over all the different structural levels can be presented, in order to obtain some generalizations about its structure. In this review we will focus first on our present understanding of the structures of amylose and amylopectin and their organization within the granule, and we will then give insights on the biosynthetic mechanisms explaining the biogenesis of starch in plants.ViewShow abstractNew Perspectives on the Storage and Organization of Muscle GlycogenArticleFull-text availableMay 2002Can J Appl Physiol Jane Shearer Terry GrahamDue to its large mass, skeletal muscle contains the largest depot of stored carbohydrate in the body in the form of muscle glycogen. Readily visualized by the electron microscope, glycogen granules appear as bead-like structures localized to specific subcellular locales. Each glycogen granule is a functional unit, not only containing carbohydrate, but also enzymes and other proteins needed for its metabolism. These proteins are not static, but rather associate and dissociate depending on the carbohydrate balance in the muscle. This review examines glycogen-associated proteins, their interactions, and roles in regulating glycogen metabolism. While certain enzymes such as glycogen synthase and glycogen phosphorylase have been extensively studied, other proteins such as the glycogen initiating and targeting proteins are just beginning to be understood. Two metabolically distinct forms of glycogen, pro- and marcoglycogen have been identified that vary in their carbohydrate complement per molecule and have different sensitivities to glycogen synthesis and degradation. Glycogen regulation takes place not only by allosteric regulation of enzymes, but also due to other factors such as subcellular location, granule size, and association with various glycogen-related proteins.ViewShow abstractTheory of field flow fractionationArticleJan 1998ADV CHROMATOGR Michel MartinViewElongation and insolubilisation of α-glucans by the action of Neisseria polysaccharea amylosucraseArticleJul 2004J CEREAL SCI Agnès Rolland-Sabaté Paul Colona Gabrielle VeroneseVéronique PlanchotAmylosucrase (E.C. 2.4.1.4) from Neisseria polysaccharea catalyses a transglycosylation reaction using sucrose as a glucose donor to synthesise an insoluble (1→4)-α-glucan, releasing fructose in the solution. If glycogen is used as a glucosyl unit acceptor, amylosucrase elongates polymer branches from their non-reducing ends. Testing of a large series of acceptors with various molar masses, glycosidic linkages, branching degrees and solubilities showed that chain elongation occurred only on polymers with (1→4)-α- or (1→4)-α- and (1→6)-α- linkages. Synthesis yields were better for completely soluble polymers (1.45 glucosyl units grafted per glucosyl unit for polymers with (1→4)-α- and (1→6)-α- linkages), than for partially soluble polymers (0.27–0.97 glucosyl units grafted per glucosyl unit for polymers with either (1→4)-α- linkages or (1→4)-α- and (1→6)-α- linkages). Elongation occurred randomly at the non-reducing ends of some external chains. Elongation of soluble branched molecules led to rapid gel formation favourable to gelation control inside suspensions and solutions. These polymers showed resistant starch contents (22% (w/w) to 57%) after elongation with amylosucrase.ViewShow abstractGlycogen structure and biogenesisArticleFeb 1991Int J Biochem Philip C CalderViewStarch synthesis in the cereal endospermArticleJul 2003CURR OPIN PLANT BIOL Martha G JamesKay Denyer Alan M MyersThe pathway of starch synthesis in the cereal endosperm is unique, and requires enzyme isoforms that are not present in other cereal tissues or non-cereal plants. Recent information on the functions of individual enzyme isoforms has provided insight into how the linear chains and branch linkages in cereal starch are synthesized and distributed. Genetic analyses have led to the formulation of models for the roles of de-branching enzymes in cereal starch production, and reveal pleiotropic effects that suggest that certain enzymes may be physically associated. For the first time, tools for global analyses of starch biosynthesis are available for cereal crops, and are heralded by the draft sequence of the rice genome.ViewShow abstractFrom Bacterial Glycogen to Starch: Understanding the Biogenesis of the Plant Starch GranuleArticleFeb 2003ANNU REV PLANT BIOL Steven G Ball Matthew MorellPlants, green algae, and cyanobacteria synthesize storage polysaccharides by a similar ADPglucose-based pathway. Plant starch metabolism can be distinguished from that of bacterial glycogen by the presence of multiple forms of enzyme activities for each step of the pathway. This multiplicity does not coincide with any functional redundancy, as each form has seemingly acquired a distinctive and conserved role in starch metabolism. Comparisons of phenotypes generated by debranching enzyme-defective mutants in Escherichia coli and plants suggest that enzymes previously thought to be involved in polysaccharide degradation have been recruited during evolution to serve a particular purpose in starch biosynthesis. Speculations have been made that link this recruitment to the appearance of semicrystalline starch in photosynthetic eukaryotes. Besides the common core pathway, other enzymes of malto-oligosaccharide metabolism are required for normal starch metabolism. However, according to the genetic and physiological system under study, these enzymes may have acquired distinctive roles.ViewShow abstractJan 2009BIOMACROMOLECULES2245-2253R A CaveS A SeabrookM J GidleyR G GilbertCave RA, Seabrook SA, Gidley MJ, Gilbert RG (2009) Biomacromolecules 10:2245-2253Jan 2007BIOMACROMOLECULES2520-2532A Rolland-SabateP ColonnaM G Mendez-MontealvoV PlanchotRolland-Sabate A, Colonna P, Mendez-Montealvo MG, Planchot V (2007) Biomacromolecules 8:2520-2532Jan 1993307-321S H YunN K MathesonYun SH, Matheson NK (1993) Carbohydr Res 243:307-321Show moreLinked ResearchCharacterization of hyperbranched glycopolymers produced in vitro using enzymes Field-Flow FractionationNovember 2013Agnès Rolland-Sabaté · Sophie Guilois · Florent Grimaud · Christine Lancelon-Pin · Alain BuléonAdvertisementRecommended publicationsDiscover moreSponsored contentINRAE is hiring 10 research scientists - Call for research projects (CRCN)October 2020INRAE is hiring researchers who have already shown their ability to produce research of excellence under supervision, attested by high-level publications. Candidates must be prepared to work independently and propose an ambitious research project in INRAE’s main areas of research: Agriculture,...View postSponsored contentINRAE is hiring 45 Scientists through open competitions and offering permanent positions.January 2020On January 1, 2020, INRA (French National Institute for Agronomical Research) and IRSTEA (National Research Institute of Science and Technology for Environment and Agriculture) merged to become the National Research Institute for Agriculture, Food and the Environment – INRAE.In 2020, INRAE...View postArticleCharacterization of hyperbranched glycopolymers produced in vitro using enzymes Field-Flow Fractiona...November 2013 · Analytical and Bioanalytical Chemistry Agnès Rolland-SabatéSophie Guilois Florent Grimaud[...] Alain BuléonAsymmetrical flow field flow fractionation (AF4) has proven to be a very powerful and quantitative method for the determination of the macromolecular structure of high molar mass branched biopolymers, when coupled with multi-angle laser light scattering (MALLS). This work describes a detailed investigation of the macromolecular structure of native glycogens and hyperbranched α-glucans (HBPs), ... [Show full abstract] with average molar mass ranging from 2 × 10(6) to 4.3 × 10(7) g mol(-1), which are not well fractionated by means of classical size-exclusion chromatography. HBPs were enzymatically produced from sucrose by the tandem action of an amylosucrase and a branching enzyme mimicking in vitro the elongation and branching steps involved in glycogen biosynthesis. Size and molar mass distributions were studied by AF4, coupled with online quasi-elastic light scattering (QELS) and transmission electron microscopy. AF4-MALLS-QELS has shown a remarkable agreement between hydrodynamic radii obtained by online QELS and by AF4 theory in normal mode with constant cross flow. Molar mass, size, and dispersity were shown to significantly increase with initial sucrose concentration, and to decrease when the branching enzyme activity increases. Several populations with different size range were observed: the amount of small size molecules decreasing with increasing sucrose concentration. The spherical and dense global conformation thus highlighted was partly similar to native glycogens. A more detailed study of HBPs synthesized from low and high initial sucrose concentrations was performed using complementary enzymatic hydrolysis of external chains and chromatography. It emphasized a more homogeneous branching pattern than native glycogens with a denser core and shorter external chains.Read moreArticleIn Vitro Synthesis of Hyperbranched α-Glucans Using a Biomimetic Enzymatic ToolboxJanuary 2013 · Biomacromolecules Florent Grimaud Christine Lancelon-Pin Agnès Rolland-Sabaté[...] Gabrielle VeroneseGlycogen biosynthesis requires the coordinated action of elongating and branching enzymes, of which the synergetic action is still not clearly understood. We have designed an experimental plan to develop and fully exploit a biomimetic system reproducing in vitro the activities involved in the formation of α(1,4) and α(1,6) glycosidic linkages during glycogen biosynthesis. This method is based on ... [Show full abstract] the use of two bacterial transglucosidases, the amylosucrase from Neisseria polysaccharea and the branching enzyme from Rhodothermus obamensis. The α-glucans synthesized from sucrose, a low cost agroresource, by the tandem action of the two enzymes, have been characterized by using complementary enzymatic, chromatographic and imaging techniques. In a single step, linear and branched α-glucans were obtained, whose proportions, morphology, molar mass and branching degree depended on both the initial sucrose concentration and the ratio between elongating and branching enzymes. In particular, spherical hyperbranched α-glucans with a controlled mean diameter (ranging from 10 to 150 nm), branching degree (from 10 to 13%), and weight-average molar mass (3.7X10^6 to 4.4X10^7 g.mol-1) were synthesized. Despite their structure that is similar to that of natural glycogens, the mechanisms involved in their in vitro synthesis appeared to be different from those involved in the biosynthesis of native hyperbranched α-glucans.Read moreArticleCharacterization of substrate and product specificity of the purified recombinant glycogen branching...October 2012 · Biochimica et Biophysica Acta (BBA) - General Subjects Christine Lancelon-Pin Xavier RousselAnders Viksø-Nielsen[...] Christophe D HulstBackground: Glycogen and starch branching enzymes catalyze the formation of α(1→6) linkages in storage polysaccharides by rearrangement of preexisting α-glucans. This reaction occurs through the cleavage of α(1→4) linkage and transfer in α(1→6) of the fragment in non-reducing position. These enzymes define major elements that control the structure of both glycogen and starch.Methods: The ... [Show full abstract] kinetic parameters of the branching enzyme of Rhodothermus obamensis (RoBE) were established after in vitro incubation with different branched or unbranched α-glucans of controlled structure.Results: A minimal chain length of ten glucosyl units was required for the donor substrate to be recognized by RoBE that essentially produces branches of DP 3-8. We show that RoBE preferentially creates new branches by intermolecular mechanism. Branched glucans define better substrates for the enzyme leading to the formation of hyper-branched particles of 30-70nm in diameter (dextrins). Interestingly, RoBE catalyzes an additional α-4-glucanotransferase activity not described so far for a member of the GH13 family.Conclusions: RoBE is able to transfer α(1→4)-linked-glucan in C4 position (instead of C6 position for the branching activity) of a glucan to create new α(1→4) linkages yielding to the elongation of linear chains subsequently used for further branching. This result is a novel case for the thin border that exists between enzymes of the GH13 family.General significance: This work reveals the original catalytic properties of the thermostable branching enzyme of R. obamensis. It defines new approach to produce highly branched α-glucan particles in vitro.Read moreDataSupplementary MaterialFebruary 2014 Agnès Rolland-SabatéSophie Guilois Florent Grimaud[...]Alain BuléonRead moreDiscover the world s researchJoin ResearchGate to find the people and research you need to help your work.Join for free ResearchGate iOS AppGet it from the App Store now.InstallKeep up with your stats and moreAccess scientific knowledge from anywhere orDiscover by subject areaRecruit researchersJoin for freeLoginEmail Tip: Most researchers use their institutional email address as their ResearchGate loginPasswordForgot password? Keep me logged inLog inorContinue with GoogleWelcome back! Please log in.Email · HintTip: Most researchers use their institutional email address as their ResearchGate loginPasswordForgot password? Keep me logged inLog inorContinue with GoogleNo account? Sign upCompanyAbout usNewsCareersSupportHelp CenterBusiness solutionsAdvertisingRecruiting© 2008-2021 ResearchGate GmbH. All rights reserved.TermsPrivacyCopyrightImprint

新闻动态
行业前沿
技术文章
最新产品

188进口试剂采购网 www.188bio.cn -中国试剂网,试剂网,化学试剂网,国药试剂,抗体公司,试剂公司,试剂盒公司,苏州试剂公司,北京化学试剂公司,天津化学试剂,试剂商城,试剂代理,流式抗体 细胞库查询 sitemap