research article glycosaminoglycan monosaccharide blocks...

Research ArticleGlycosaminoglycan Monosaccharide Blocks Analysis byQuantum Mechanics Molecular Dynamics and NuclearMagnetic Resonance

Sergey A Samsonov1 Stephan Theisgen2 Thomas Riemer2 Daniel Huster2

and M Teresa Pisabarro1

1 Structural Bioinformatics BIOTEC TU Dresden Tatzberg 47-51 01307 Dresden Germany2 Institute of Medical Physics and Biophysics University of Leipzig Hartelstr 16-18 04107 Leipzig Germany

Correspondence should be addressed to Sergey A Samsonov sergeysbiotectu-dresdende

Received 17 December 2013 Accepted 22 January 2014 Published 7 April 2014

Academic Editor Dieter Scharnweber

Copyright copy 2014 Sergey A Samsonov et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

Glycosaminoglycans (GAGs) play an important role in many biological processes in the extracellular matrix In a theoreticalapproach structures of monosaccharide building blocks of natural GAGs and their sulfated derivatives were optimized bya B3LYP6311ppddB3LYP6-31+G(d) method The dependence of the observed conformational properties on the appliedmethodology is described NMR chemical shifts and proton-proton spin-spin coupling constants were calculated using the GIAOapproach and analyzed in terms of the methodrsquos accuracy and sensitivity towards the influence of sulfation O1-methylationconformations of sugar ring and 120596 dihedral angle The net sulfation of the monosaccharides was found to be correlated withthe 1H chemical shifts in the methyl group of the N-acetylated saccharides both theoretically and experimentally The 120596 dihedralangle conformation populations of free monosaccharides and monosaccharide blocks within polymeric GAG molecules werecalculated by a molecular dynamics approach using the GLYCAM06 force field and compared with the available NMR andquantummechanical data Qualitative trends for the impact of sulfation and ring conformation on the chemical shifts and proton-proton spin-spin coupling constants were obtained and discussed in terms of the potential and limitations of the computationalmethodology used to be complementary to NMR experiments and to assist in experimental data assignment

1 Introduction

Glycosaminoglycans (GAGs) represent a class of linearanionic heteropolysaccharides containing repeating disac-charide units made up of a hexose or a hexuronic acidlinked to a hexosamine by a 1-3 or 1-4 glycosidic linkageHydroxyl groups of these saccharides can be sulfated atdifferent positions Being localized in the extracellularmatrixGAGs play a crucial role in cell adhesion and proliferation[1] by involvement in key molecular regulatory mechanisms[2] As for all saccharides GAGs are very flexible and adopt anumber of energetically similar conformational states underphysiological conditions which render structural studies ofGAGs challenging from both the experimental [3 4] and

the computational [5] points of view Solvent is suggested toplay an indispensable role for the structure and dynamicsof saccharides due to the tight coupling of solvent andsolute dynamics their interactions [6ndash10] and the effectsof electrostatic polarization [11] In addition the highlycharged nature ofGAGsmakes their interactionswith solventmolecules by hydrogen bonding even more important forthe exploration of their conformational space [10 12ndash15]Therapid exchange of the intramolecular and solvent-mediatedhydrogen bonds does not allow experimental techniques suchas nuclear magnetic resonance (NMR) to gain a deep viewon the hydrogen bonds formation in GAGs and thereforecomputational methods as molecular dynamics (MD) simu-lations are very useful to analyze GAGs structural properties

Hindawi Publishing CorporationBioMed Research InternationalVolume 2014 Article ID 808071 11 pageshttpdxdoiorg1011552014808071

2 BioMed Research International

inmore detail [16] Regarding the sulfation patterns of GAGscombination ofNMRwithMD [17] and quantummechanical(QM) [18] approaches were successfully applied to revealthe impact of sulfation effects on GAGs structure in termsof dynamic behaviour of glycosidic linkages However notonly glycosidic linkage conformations but also sugar ringpuckering could be decisive for the biologic relevance andthe specificity of GAGprotein interactions [19] In the case ofheparin it is supposed that the conformational flexibility ofthe free heparin molecule is not dramatically affected by theflexibility of the IdoUA(2S) sugar rings [20] Nevertheless itwas reported that in the complex of heparin pentasaccharidewith FGFR one of the IdoUA(2S) adopted the 2S

0ring

conformation whereas the rest of IdoUA(2S) residues were inthe 1C

4ring conformation providing high specificity for the

formation of this GAGprotein complex [21] Therefore it isimportant to understand the basic rules governing the ringconformation preferences for individual monosaccharideblocks of the GAGmoleculesThe ring conformational spacefor several GAG mono- and disaccharides for GlcNAc andits N- 3-O and 6-O sulfated derivatives [22] GlcUA IdoUA[23 24] IdoUA(2S) [23 25 26] and heparin disaccharides[27 28] was extensively analyzed in recent studies by meansof MD QM and NMR approaches demonstrating agree-ment and complementarity of these methodologies Thissuggests a high potential of the use of theoretical approachesfor the assistance in interpretation of NMR experimentaldata Interestingly despite the above-mentioned importantrole of solvent the use of an implicit solvent model (incontrast to the use of explicit solvent molecules) does notimprove agreement between spin-spin coupling parame-ters calculated by QM and measured experimentally byNMR [26]

In addition to natural GAGs artificial GAGs withdistinct sulfation patterns are promising components forfunctional biomaterials targeted for extracellular artificialmatrix engineering since additional sulfate groups couldmodulate specific binding of growth factors and therebyinfluence wound healing [29ndash31] Unfortunately sometimesonly the net sulfation degree of GAGs used in the experi-ments but not the exact sulfation pattern is known whichrenders assignment of NMR spectra for the following eluci-dation of structure-function relationships more challengingTherefore theoretical analysis of the structural propertiesof sulfated GAG monosaccharides and calculation of theirNMR chemical shifts and spin-spin coupling constants couldbe essential for the assistance in NMR experimental datainterpretation Along these lines it was shown that spe-cific sulfation patterns of some GlcNAc derivatives inducechanges in ring puckering preferences [22] Here we sys-tematically study sulfated derivatives of GlcNAc GalNAcIdoUA and GlcUA with varying degrees of net sulfationwhich represent the building blocks for heparin hyaluronicacid chondroitin sulfate and dermatan sulfate In particularwe analyze conformational preferences of the sugar ringsusing a QM approach the impact of sulfation and usedpolymerization models Furthermore we calculate NMRparameters using several computational models which pro-vide GAG monosaccharide conformational QM dictionary

data For GlcNAc GlcNAc(6S) GalNAc GalNAc(4S) andGalNAc(6S) we compare our calculated parameters withexperimental data on 13C and 1H chemical shifts and proton-proton spin-spin coupling constants ( 3JH-H) and we discussthe potential accuracy of this methodology For GlcNAc andGalNAc sulfated derivatives the conformational space ofthe dihedral angle around the C5ndashC6 bond is analyzed andcomparedwithinQMandMDapproachesThe data obtainedin this work help to get a deeper insight in the potential andlimitations of state-of-the-art computational methods usedto complement NMR experiment interpretation for GAGmolecules

2 Materials and Methods

21 Quantum Mechanical Calculation Thefollowingmonosaccharides (Figure 1) and their O1-methylatedvariants (abbreviated with M- in Tables 1 and 2) wereused for QM calculations GlcNAc (120573-D-N-acetylglu-cosamine) GlcNAc(4S) (4-O-sulfo-120573-D-N-acetylglucosa-mine) GlcNAc(6S) (6-O-sulfo-120573-D-N-acetylglucosamine)GlcNAc(46S) (46-O-disulfo-120573-D-N-acetylglucosamine)GalNAc (120573-D-N-acetylgalactosamine) GalNAc(4S) (4-O-sulfo-120573-D-N-acetylgalactosamine) GalNAc(6S) (6-O-sulfo-120573-D-N-acetylgalactosamine) GalNAc(46S) (46-O-disulfo-120573-D-N-acetylgalactosamine) GlcUA (120573-D-glucuronic acid)GlcUA(2S) (2-O-sulfo-120573-D-glucuronic acid) GlcUA(3S)(3-O-sulfo-120573-D-glucuronic acid) GlcUA(23S) (23-O-disulfo-120573-D-glucuronic acid) IdoUA (120572-L-iduronic acid)IdoUA(2S) (2-O-sulfo-120572-L-iduronic acid) and IdoUA(3S)(3-O-sulfo-120572-L-iduronic acid) IdoUA(23S) (23-O-disulfo-120572-L-iduronic acid)

First the molecules were built in MOE [32] in 1C4 4C1

and 2S0ring conformations For GlcGalNAc sulfated deriva-

tives gt tg and gg conformations were built correspond-ing to the values of dihedral angle 120596 = (O6ndashC6ndashC5ndashO5)of sim300∘ sim60∘ and sim180∘ respectively [10] Na+ counterionswere manually added to the systems with a nonzero netcharge and their positions were subsequently optimized byAMBER99 force field in MOE The geometry optimizationof these structures was carried out with GAUSSIAN 09 [33]using B3LYP functional [34] with 6-31+G(d) basis set Singlepoint energies were calculated using the B3LYP6311ppddmethod which was shown to be appropriate for energy cal-culations for carbohydrates [35] For each monosaccharidethe relative energies were calculated using the energy of themost stable conformation as reference GIAO methodologyimplemented within GAUSSIAN [36] was used to calculateNMR parameters B3LYP6311+G(2dp) for chemical shiftsandB3LYPaug-cc-pVDZ for spin-spin coupling constants asthese levels of theory demonstrated highest reliability in thecalibration studies [37 38] TMS (tetramethylsilane)was usedas a reference to calculate 13C- and 1H-chemical shifts For thecalculations carried out in solvent PCM solvent model [39]was used

22 Molecular Dynamics Calculations For MD simulationsthe GLYCAM06 force field [40] implemented in the AMBER

BioMed Research International 3

H7

C6H6

C5

H4C1

H5

H2

C4 C3C2

H1

H3

H8 H11

H10H9

H7

C6H6

C5

H4 C1

C4H5

H2

C3C2 H1

H3

H8

H11

H10

H9

GlcNAc GalNAc

C6

C5H4

C4H5

H2

C3C2

C1

H1

H3

C6

C5

C1

H4

C4

H5

H2

C3

C2 H1

H3

GlcUA IdoUA

Figure 1 Chemical structure of 120573-D-GlcNAc 120573-D-GalNAc 120573-D-GlcUA and 120572-L-IdoUA with numbering used throughout the paper

11 package [41] was used for GAGs For sulfated residuessulfate atomic charges for HA and CS derivatives residuelibraries were obtained from RESP by fitting calculations atthe level of 631(d)G for methylsulfate and introduced into thecorresponding GLYCAM libraries All the monosaccharideswere modeled in the 4C

1ring conformation as it was

suggested by our results (see Section 31) Prior to thesimulation GAG monosaccharides were solvated withinan octahedral TIP3PBOX of 15 A distance to the sidesof the periodic unit and counterions were added whenrequired The system was minimized and equilibrated asdescribed before [42] and simulated for 50 ns in NTPensemble For MD simulations of GAGs hexasaccharidesthe structures available in the PDB for octameric HA (PDBID 2BVK NMR) and hexameric CS4 (PDB ID 1CS4fiber diffraction) were used as templates for modeling of

the following HA and CS derivatives (GlcUA-GlcNAc)3

(GlcUA-GlcNAc(4S))3 (GlcUA-GlcNAc(6S))

3 (GlcUA-

GlcNAc(46S))3 (GlcUA(2S)-GlcNAc(46S))

3 (GlcUA(3S)-


3 (GlcUA-

GalNAc)3 (GlcUA-GalNAc(4S))

3 (GlcUA-GalNAc(6S))

3

(GlcUA-GalNAc(46S))3 (GlcUA(2S)-GalNAc(46S))

3 (Glc-

UA(3S)-GalNAc(46S))3 and (GlcUA(23S)-GalNAc(46S))

3

The MD simulations for these GAGs were carried out for20 ns and the obtained data for three monosaccharide unitswithin each hexasaccharide were averaged to be comparedwith the data on free monosaccharides The trajectoriesanalysis was done using the ptraj module of AMBER 11 Forthe analysis of the dihedral angle 120596 = (O6ndashC6ndashC5ndashO5) forGlcGalNAc sulfated derivatives gg gt and tg conformationswere defined for 120596 in the ranges of [minus120∘ 0∘) [0∘ 120∘) and[minus180∘ minus120∘) cup [120∘ 180∘) respectively


Table 1 B3LYP6311ppddB3LYP6-31+G(d) relative energies for ring and gggttg conformations of GlcNAc and GalNAc derivatives

Moleculeconformation Δ119864(4C1) kcalmol Δ119864 (1C

4) kcalmol Δ119864 (2S

0) kcalmol

Saccharides 119892119892 119892119905 119905119892 119892119892 119892119905 119905119892 119892119892 119892119905 119905119892

GlcNAc 202 177 0 1017 685 1031 281 428 1166M-GlcNAc 340 272 109 567 201 604 167 0 740GlcNAcPCM 039 028 0 783 639 829 404 425 965M-GlcNAcPCM 402 255 0 311 428 697 481 302 822GlcNAc(4S) 777 830 0 1919 1461 1470 700 1505 477M-GlcNAc(4S) 1646 1376 0 1937 1511 1510 171 1594 822GlcNAc(4S)PCM 331 331 0 1004 915 838 398 733 438M-GlcNAc(4S)PCM 526 283 0 861 661 644 459 565 628GlcNAc(6S) 022 0 1014 898 1684 1947 2293 1408 1449M-GlcNAc(6S) 694 345 0 130 1249 1176 1504 1153 1606GlcNAc(6S)PCM 027 0 152 718 528 683 851 462 606M-GlcNAc(6S)PCM 0 035 246 267 266 256 4C

141 579

GlcNAc(46S) 0 613 189 4414 2903 2105 1759 3731 1450M-GlcNAc(46S) 1196 790 0 2316 2490 1229 4C

12870 049

GlcNAc(46S)PCM 161 378 0 1245 927 828 944 1152 1078M-GlcNAc(46S)PCM 371 465 0 886 889 456 733 832 175GalNAc 0 203 367 929 888 1210 078 698 1095M-GalNAc 0 207 281 399 360 595 343 451 409GalNAcPCM 0 104 240 855 840 868 378 664 939M-GalNAcPCM 0 199 370 574 550 611 736 677 698GalNAc(4S) 777 830 0 1919 1461 1470 700 1505 477M-GalNAc(4S) 499 080 0 1043 2657 1524 722 2422 779GalNAc(4S)PCM 0 296 209 957 1248 928 390 1129 500M-GalNAc(4S)PCM 0 031 309 725 1113 831 621 859 445GalNAc(6S) 0 469 1580 1042 2096 2091 2399 2026 2175M-GalNAc(6S) 314 0 586 1179 1101 544 1210 1036 1023GalNAc(6S)PCM 0 081 020 609 513 544 1288 771 742M-GalNAc(6S)PCM 082 313 0 404 439 544 948 542 309GalNAc(46S) 0 615 643 560 2994 2515 348 1974 313M-GalNAc(46S) 314 875 0 511 2236 1303 459 1243 1529GalNAc(46S)PCM 150 209 0 694 919 1643 455 1368 490M-GalNAc(46S)PCM 0 235 274 513 653 1039 1172 632 299Relative energies were calculated using the energy of the most stable conformation for the same molecule as a reference for in vacuo and PCM solvent model(marked with PCM subscript)4C1 The ring conformation changed to 4C1 during geometry optimization

23 NMR Measurements All NMR spectra were measuredon a Bruker Avance III 600MHz spectrometer operatingat 60013MHz 1H resonance frequency equipped with a5mm TBI triple resonance probehead with Z-gradient oron a Bruker Avance I 700MHz spectrometer operating at70018MHz 1H resonance frequency equipped with a tripleresonance cryo-probehead at 37∘C in D

2O with TSP as a

reference (set to 0 ppm for 1H and 13C chemical shifts) Theresonance assignments were based on COSY J-modulatedandHSQC 2D spectra To account for strong coupling effectsthe chemical shifts and 3JH-H were extracted by fitting the

experimental 1D spectra with a self-written Octave script[43]

Statistical analysis of data was carried out with the R-package [44]

3 Results and Discussion

31 Conformational Preferences of the Analyzed Monosac-charides The geometries of GlcNAc GalNAc GlcUA andIdoUA monosaccharides and their sulfated derivatives wereoptimized and their single point energies were calculated for


Table 2 B3LYP6311ppddB3LYP6-31+G(d) relative energies for ring conformations of IdoUA and GlcUA derivatives

Moleculeconformation Δ119864 (4C1) kcalmol Δ119864 (1C


0) kcalmol

GlcUA 088 0 342M-GlcUA 0 070 512GlcUAPCM 0 169 311M-GlcUAPCM 0 623 693GlcUA(2S) 111 400 0M-GlcU(2S) 0 529 684GlcUA(2S)PCM 125 751 0M-GlcU(2S)PCM 0 371 784GlcU(3S) 444 0 193M-GlcU(3S) 0 234 073GlcU(3S)PCM 678 365 0M-GlcU(3S)PCM 0 1109 129GlcUA(23S) 794 1314 0M-GlcUA(23S) 0 1486 583GlcUA(23S)PCM 582 1089 0M-GlcUA(23S)PCM 0 887 162IdoUA 0 200 653M-IdoUA 0 233 664IdoUAPCM 0 154 559M-IdoUAPCM 0 154 401IdoUA(2S) 028 0 578M-IdoUA(2S) 277 0 718IdoUA(2S)PCM 0 471 718M-IdoUA(2S)PCM 150 0 350IdoUA(3S) 0 1964 2268M-IdoUA(3S) 0 2136 2417IdoUA(3S)PCM 0 199 294M-IdoUA(3S)PCM 0 187 294IdoUA(23S) 0 070 046M-IdoUA(23S) 0 1118 644IdoUA(23S)PCM 0 555 757M-IdoUA(23S)PCM 0 782 438Relative energies were calculated using the energy of the most stable conformation for the same molecule as a reference for in vacuo and PCM solvent model(marked with PCM subscript)

three ring conformations (4C1 1C4 2S0) and in addition

for the gggttg conformations for N-acetylated saccharides(Tables 1 and 2) The obtained results represent in vacuoand PCM implicit solvent models for nonmethylated andO1-methylated monosaccharides where the latter is used as thesimplest model for the glycosidic linkage in GAGs Usingthe same level of theory for single point energy calculations(B3LYP6311ppdd) but a different level for geometry optimiza-tion (6-31+G(d) versus B3LYP6311ppdd) we were able tonicely reproduce relative energies forM-IdoUA(2S) ring con-formers obtained in the work of Hricovıni [26] (7690 versus718 2775 versus 277 kcalmol for the differences betweenthe most stable 1C

4and 2S

0 4C1conformations resp) The

positions of counterions were also predicted very similarly tothe positions in the aforementioned study If the counterionswere not used for the calculations though the geometry ofM-IdoUA(2S) was correctly obtained energetic comparison

of the conformations failed For example when not usingcounterions the 4C

1ring conformation was observed to be

the most stable (data not shown) This suggests a strongimpact of the ions on the ring puckering due to the net elec-trostatic effect in a not neutralized system Interestingly finalpoint energies are also affected by the counterion positionsoccupied after the geometry minimization In case of manynegatively charged groups as for example for double sulfatedGlcUA or IdoUA these positions could be not unique Thispoint is important to consider when quantitatively analyzingthe results represented in Tables 1 and 2

For GlcNAc and GalNAc derivatives all the data showthe preference for the 4C

1ring conformation (except for M-

GlcNAc in vacuo where 2S0was found to be the most stable

with a relatively low difference of 109 kcalmol to the 4C1

conformation) (Table 2) This agrees with the previous longMD studies for GlcNAc [22] and the experimental structures


of free chondroitin sulfate 4 (PDB ID 1CS4) and hyaluronicacid (PDB IDs 1HYA 2HYA 3HYA 4HYA 1HUA 2BVK)In general the probability of adopting 2S

0was calculated to

be higher than for 1C4for the optimized structures When

analyzing gggttg conformations of 120596 dihedral angle thereis an essential dependence on the model used for the calcula-tions Nevertheless for both in vacuo and implicit solvent cal-culations GlcNAc derivatives prefer tg conformation whileGalNAc derivatives are more prone to the gg conformationwhich is not in agreement with the expected gauche effect forthese molecules This inconsistency of QMmethods was alsoobserved previously in the work of Kirschner and WoodsThese authors explain the limitation of this methodologyin terms of disregarding explicit solvent [10] Indeed thetg conformation for structure of GlcNAc obtained by QMis favourable due to the formation of a hydrogen bondbetweenO6 andHO4 atoms which in the presence of explicitsolvent could be disrupted due to the interaction with watermolecules In case of GlcNAc(4S) andGlcNAc(6S) the stronginteraction between sulfate groups and hydrogens of thehydroxyl groups defines themost favourable120596 dihedral angleconformation Besides that the positions of counterions areespecially important for GlcNAc(46S) two Na+ ions arecoordinated between sulfate groups in the positions 4 and6 and therefore strongly stabilize the tg conformation ForGalNAc the formation of the hydrogen bond between O6andHO4 atoms is more probable in the gg conformation Forthe sulfated GalNAc derivatives the positions of Na+ ions arecrucial for the selection of themost energetically favourable120596dihedral angle conformationMethylation of theO1 decreasesthe opportunity for intramolecular hydrogen bonding andtherefore also influences the gggttg conformational distri-bution

For GlcUA sulfated derivatives the conformationaldependence on both sulfation pattern and the model usedfor calculations is clearly observed (Table 2) For GlcUAmonosaccharide all methods find the 4C

1ring conformation

highly probable In vacuo calculations propose coexistence ofthis conformation together with the 1C

4conformation with

the prevalence of the latter for the case when O1 positionis not methylated (difference in energy of two conforma-tions of 1 kcalmol corresponds to their probabilities ratioof 85 15) This agrees with the experimental structures offree hyaluronic acid (PDB IDs 1HYA 2HYA 3HYA 4HYA1HUA 2BVK) and data from MD simulations [23] For sul-fated derivatives of GlcUA all O1-methylated monosaccha-rides are found to be the most stable in the 4C

1conformation

independently of the solvent use in the calculations For M-GlcUA(3S) the differences in energies between 2S

0and 4C

1

are quite low suggesting possible coexistence of these confor-mations For all unmethylated sulfated derivatives of GlcUA(except for in vacuo calculation for GlcUA(3S) where the 1C

4

conformation was the most stable) the 2S0conformation is

preferredFor IdoUAderivatives there is amuch higher consistency

within the results obtained by different methods though therelative differences between the conformations stabilities fordifferent methods are still relatively high (eg IdoUA(3S) in

vacuo versus implicit solvent) (Table 2) All the derivativesexcept IdoUA(2S) prefer the 4C

1conformation whereas

IdoUA(2S) prefers the 1C4conformation Interestingly in the

free heparin crystal structure (PDB ID 1HPN) IdoUA(2S)monosaccharide units are observed in 2S

0and 1C

4but not in

the 4C1conformation

All these QM data for the analyzed monosaccharidessuggest that the results obtained for distinct models (withrespect to solvent and O1-methylation) should be consideredwith caution especially when compared to the data onconformational preferences for sugar rings and gggttg forthese monosaccharides within long GAG polymers

32 MD Conformational Analysis of 120596 Dihedral forGlcGalNAcDerivatives As it was pointed out in the previoussection QM approaches experience severe difficulties in thequantitative description of 120596 dihedral angle conformationsIn contrast to QM calculations MD simulations are able notonly to take into account solvent explicitly but also to gaininsights into internalmotions of themolecules and thereforeyield more complete information about the conformationalspace than the data from QM or NMR experiments

According to the experimental data glycopyranosylderivatives tend to adopt gg and gt conformations (knownas gauche effect) with the ratios of gggttg in the percent-age range sim60ndash70 30ndash40 0ndash5 per conformation whereasgalactopyranosyl derivatives adopt less gg and gain in tgconformational content with the corresponding ratios of 10ndash20 45ndash55 30ndash40 [45ndash47] MD simulations for GlcNAc andGalNAc nonmethylated derivatives in general agree with thistrend and are able to reproduce the gauche effect (Table 3) ForGlcNAc derivatives sulfation in the 4th position increasesthe preference to the gt conformation and sulfation inthe 6th position makes gg more favourable For GalNAcderivatives sulfation in the 4th position does not make anysignificant effect while sulfation in position 6 makes thetg conformation significantly more favourable This couldbe explained in terms of the dipole interactions betweenthe sulfate and hydroxyl groups in positions 4 and 6 Forthe data interpretation several aspects should be taken intoaccount (i) the sulfation of the hydroxyl group changes thedirection of the corresponding dipole to the opposite one(ii) the absolute value of the OH group dipole is lower thanthe one of OndashSO

3 (iii) C4ndashO4 is in equatorial configuration

for GlcNAc and in axial configuration for GalNAc (iv) theflexibility of the group in position 6 is higher than theflexibility of the group in position 4 (v) the repulsive strengthof the dipole-dipole interaction is defined by the dipolersquosabsolute value the distance between them and their mutualorientation With these considerations the increase of the ggconformation population for GlcNAc(6S) in comparison toGlcNAc could be explained by more favourable interactionof O4ndashH dipole with O6ndashSO

3dipole in comparison to

weak repulsion between two OndashH dipoles in the nonsulfatedderivative since the angle and distance between these dipolesare lower in gg in comparison to the gt conformation Onthe contrary the decrease of the gg conformation populationof GlcNAc(46S) in comparison to GlcNAc(6S) could be


Table 3 120596 dihedral angle gggttg () conformations distributionfor GlcNAcGalNAc monosaccharide derivatives

Monosaccharide 119892119892 119892119905 119905119892

GlcNAc 50 48 2GlcNAc(4S) 38 59 3GlcNAc(6S) 78 12 11GlcNAc(46S) 59 29 12GalNAc 7 76 16GalNAc(4S) 5 81 14GalNAc(6S) 3 19 78GalNAc(46S) 1 31 68

Table 4 120596 dihedral angle gggttg () conformations distribu-tion for GlcNAcGalNAc monosaccharide units within hexamericGAGs1GAG gg gt tgHA 53 45 2HA4 61 37 2HA6 87 11 2HA46 57 40 4HA4621015840 83 16 1HA4631015840 59 30 11HA462101584031015840 34 54 12CS de 8 80 12CS4 6 86 9CS6 6 56 38CS46 2 77 21CS4621015840 0 81 19CS4631015840 2 53 45CS462101584031015840 2 57 42HA HA4 HA6 HA46 HA4621015840 HA4631015840 HA462101584031015840 CS CS4 CS6CS46 CS4621015840 CS4631015840 CS462101584031015840 stay for (GlcUA-GlcNAc)3 (GlcUA-GlcNAc(4S))3 (GlcUA-GlcNAc(6S))3 (GlcUA-GlcNAc(46S))3 (GlcUA(2S)-GlcNAc(46S))3 (GlcUA(3S)-GlcNAc(46S))3 (GlcUA(23S)-GlcNAc-(46S))3 (GlcUA-GalNAc)3 (GlcUA-GalNAc(4S))3 (GlcUA-GalNAc(6S))3(GlcUA-GalNAc(46S))3 (GlcUA(2S)-GalNAc(46S))3 (GlcUA(3S)-GalNAc-(46S))3 and (GlcUA(23S)-GalNAc(46S))3 respectively

explained by the stronger repulsion between two O4O6ndashSO3dipoles in the case of the GlcNAc(46S) molecule For

GalNAc derivatives the gg conformation is sterically lessaccessible because the C4ndashO4 configuration is different toGlcNAc and the O4ndashH group would overlap with the O6ndashSO3group in this conformation Here the tg conformation

is favourable when both O4 and O6 are sulfated because theangle between these two dipoles is closer to 90∘ than for thegt conformation However this explanation in terms of onlytwo dipoles interaction cannot be used when comparing thedifferences in preferences of GalNAc(4S) and GalNAc(6S)

We also compared these data for monosaccharides MDsimulations with the results obtained for the correspond-ing monosaccharide blocks within the hexameric GAGs(Table 4) When only the sulfation of GAGs changes thepopulations of the 120596 dihedral angle conformation changevery similarly to the ones in monosaccharides for HA

HA4 HA6 and HA46 and for CS de CS4 CS6 and CS46respectively Sulfation of the GlcUA within hexameric HAand CS derivatives affects the conformations of GlcNAcand GalNAc46 for GlcUA(2S) and GlcUA(3S) respectivelyThis suggests a pronounced mutual influence of electrostaticenvironment of themonosaccharide unitswithin the polymeron their120596 dihedral angle conformations but a weak influenceof the polymerization via O1 and O3 respectively

33 Chemical Shifts and 3119869119867-119867 Calculations For 13C and

1H chemical shifts and 3JH-H GIAO calculations we usedGlcNAc GalNAc GlcUA and IdoUA monosaccharidesand their sulfated derivatives both nonmethylated and O1-methylated in the conformations listed in Section 31 Thesedata (see Supplementary Tables 1ndash17 available online athttpdxdoiorg1011552014808071) could be used as QMNMR parameters dictionary for monosaccharides with adifferent sulfation pattern

Our analysis of 13C and 1H chemical shifts yields nosignificant dependence of these calculated NMR parametersneither on the conformations nor on O1-methylation (exceptfor C1) (Supplementary Tables 1ndash12) In contrast thereare changes of chemical shifts occurring upon sulfation Inparticular 13C chemical shifts increase more than 5 ppm forC4 and slightly less than 5 ppm for C6 in case of GlcNAcand GalNAc sulfated monosaccharides The same trend isobserved in the experiments (Table 5) where the chemicalshifts of the sulfated carbons C4 and C6 increased by 6to 7 ppm When C4 is sulfated calculated chemical shiftsof the adjacent C3 and C5 slightly drop which is alsoobserved in the experiments where the similar chemicalshift changes are observed for C5 when C6 is sulfatedIn case of GlcUA and IdoUA sulfated derivatives C2 andC3 chemical shifts similarly increase upon sulfation whilechemical shifts of adjacent carbons also drop Our resultsallow for the conclusion that the calculated changes in thechemical shifts upon sulfation agree well with the experimen-tal data However the variance within the values correspond-ing to different individual conformations is substantial Forexample for GlcUA 1C

4conformation the increase of the

chemical shift for C3 is observed upon the sulfation of C3which contradicts the general observation derived from theaveraging per all conformations This makes the use of thesevalues challenging for the practical purposes of the directNMR spectra assignment Based on the presented data theexpected error range for 13C chemical shifts is up to 5 ppmwhich is similar to the differences found for the sulfatedcarbons For the 1H chemical shifts we also observe thesignificant increase of about 05 ppm for the values of thehydrogens bound to the carbons being sulfated as well asa slight increase of chemical shifts of the hydrogens boundto the carbons adjacent to the sulfated ones Our experi-mental data for the C4-sulfation of GlcNAc and GalNAcqualitatively support the data obtained in our calculations(Table 5) For both 13C and 1H chemical shifts we clearlyobserve experimentally validated qualitative trends whichmight further allow for a quantitative comparison of QMobtained values with experimental data The variance of


Table 5 Experimentally versus computationally obtained chemical shifts (ppm)

Atom GlcNAc GlcNAc(6S) GalNAc GalNAc(4S) GalNAc(6S)Exp GIAO Exp GIAO Exp GIAO Exp GIAO Exp GIAO

C1 9777 10700 9790 10427 9827 10712 9816 10140 9828 10779C2 5960 6430 5958 6523 5662 6265 5698 6046 5645 6280C3 7674 8251 7661 8245 7399 7719 7281 8051 7379 7732C4 7276 8069 7258 7870 7077 7461 7864 7733 7048 7260C5 7879 8415 7669 8711 7803 7854 7722 8188 7563 7938C6 6364 6917 7013 7115 6406 6844 6385 6528 7030 7057C7 17759 18304 mdash 18456 mdash 18275 17753 18210 mdash 18271C8 2503 2396 2493 2401 2497 2408 2508 2496 2508 2407H1 472 465 474 488 465 445 472 494 467 446H2 367 378 370 367 388 394 389 400 389 390H3 354 336 356 341 373 351 388 388 375 342H4 346 317 352 327 394 435 470 521 400 424H5 347 329 368 382 370 391 382 395 394 364H6 375 384 422 425 377 415 mdash 384 420 412H7 391 399 435 469 380 433 mdash 427 423 477H9 10 11 205 205 205 212 206 205 206 211 206 202

Table 6 Experimentally versus computationally obtained 3JH-H (Hz)

Protonpair GlcNAc GlcNAc(6S) GalNAc GalNAc(4S) GalNAc(6S)Exp GIAO Exp GIAO Exp GIAO Exp GIAO Exp GIAO

H1-H2 842 559 832 574 817 553 830 527 809 563H2-H3 1046 768 1039 742 1092 787 mdash 850 1086 769H3-H4 882 589 899 607 336 390 250 371 350 373H4-H5 953 669 1002 741 103 208 mdash 285 085 194H5-H6 240 256 184 054 441 565 mdash 684 477 414H5-H7 567 670 559 460 779 640 mdash 490 755 789

the observed chemical shifts within the groups of differentconformations is about 05 ppm and therefore compara-ble to the experimentally observed differences induced bysulfation

For 3JH-H we obtain slight qualitative differences depend-ing on the sulfation pattern (Supplementary Tables 13ndash17)which is similarly observed by the experiment (Table 6) Thevariance of the 3JH-H for the protons bound to the carbonsC1ndashC5 of the ring is up to 2-3 whereas the variance of 3JH-Hfor other proton pairs (available for GlcNAc and GalNAcderivatives) is higher and reaches the values of 4-5 The vari-ance of the 3JH-H grouped by ring conformations decreasesdown to 1 for the protons bound to the carbons C1ndashC5 whichis similar to the corresponding experimental accuracy In caseof GlcNAc derivatives especially high variance for the 2S

0

ring conformation is observedThis is due to the fact that thegeometry optimization starting from this ring conformationfor M-GlcNAc(6S) and M-GlcNAc(46S) ended up in the4C1conformation whereas for some molecules dramatic

geometrical distortions were found (they correspond to highenergies in Table 2) Except for GlcNAc derivatives in the2S0conformation clear trends for 3JH-H of H1-H2 H2-

H3 H3-H4 and H4-H5 are observed which allows signif-icantly distinguishing different ring conformationswithin theapplied method In particular GlcUA and IdoUA derivativeshave four high 3JH-H (sim5ndash8) that correspond to the 4C

1

conformation four low (sim2ndash4) to the 1C4conformation and

three low and one high to the 2S0conformation respec-

tively For GlcNAc and GalNAc derivatives epimeric C4could be clearly distinguished for the corresponding H3-H4 and H4-H5 3JH-H for the most energetically favourable4C1conformation which is qualitatively in agreement with

the experimental data but quantitatively underestimated(Table 6)The values for gggttg conformations for each ringconformation clearly differ (Supplementary Tables 13ndash16)which corresponds to different geometries and could be usedas a dictionary for these parameters Therefore calculationsof 3JH-H could be used for assistance in NMR assignments incases where ring conformation and sulfation patterns are notwell defined

NMR parameters calculated by GIAO approaches (chem-ical shifts 3JH-H) qualitatively reflect the sulfation ringand 120596 dihedral conformations ( 3JH-H) However the directand quantitative use of the calculated NMR parameters for


Table 7 Comparison of the experimental and theoretical data on NMR parameters

GlcNAc GlcNAc(6S) GalNAc GalNAc(4S) GalNAc(6S) All119877Pearson

13C 0999 (0998) 0994 0995 0993 (0998) 0995 0994 (0996)119877Spearman

13C 1000 1000 1000 0893 (0929) 1000 0959 (0965)119877Pearson

1H 0980 0980 0932 0991 0939 0961119877Spearman

1H 1000 0976 0905 0886 0929 0933ΔΔppm 13C 566 (563) 519 396 313 (439) 379 434 (438)ΔΔppm 1H 006 004 015 017 002 005119877Pearson

3JH-H 0836 0985 0933 mdash 0923 0899119877Spearman

3JH-H 0657 1000 0829 mdash 0829 0840ΔΔppm 3JH-H 17 22 07 09 08 13The analysis for 13C chemical shifts is done without the consideration of C7 chemical shifts The values obtained with the consideration of C7 chemical shiftsfrom GlcNAc and GalNAc(4S) are given in the parenthesis

experimental data assignment could be limited due to theintrinsic error of the method

34 Comparison of the Calculated NMR Parameters withExperimental Data for GlcNAc GlcNAc(6S) GalNAcGalNAc(4S) and GalNAc(6S) In order to estimate thepractical applicability of the used computational methods forNMR parameter calculations we compared the calculatedchemical shifts and 3JH-H with the available experimentaldata obtained by NMR for 120573-GlcNAc 120573-GlcNAc(6S) 120573-GalNAc 120573-GalNAc(4S) and 120573-GalNAc(6S) (Tables 5 and6) The Pearson and Spearman correlations between thecalculated and experimental data and the mean error of thetheoretical methods are 0994 0959 and 434 ppm for 13Cchemical shifts 0961 0933 and 005 ppm for 1H chemicalshifts and 0899 0840 and 13Hz for 3JH-H respectively(Table 7) For 13C chemical shifts the theoretically obtainedabsolute values for all analyzed saccharides are systematicallyoverestimated for C1ndashC7 except for C4 of GalNAc(4S) andunderestimated for about 1 ppm for C8 (Table 5)The Pearsoncorrelations between experimental and theoretical valuesare very high for 13C chemical shifts which neverthelesscould be partially explained in terms of high differencesbetween the values for C8 in comparison to other valuesSpearman correlations were found to be 10 for four outof five monosaccharides which represents a promisingresult for ranking the peaks for 13C spectra At the sametime the mean error could be too high for distinguishingcarbons C1ndash6 in case the most probable conformationof the molecule is a priori unknown For 1H chemicalshifts we obtained systematic underestimation of H3 andoverestimation of H7 chemical shifts by the applied GIAOapproach while for other protons both overestimation andunderestimation of experimental values were observed(Table 5) Both Pearson and Spearman correlations for1H chemical shifts are lower than for 13C chemical shiftsbut the low mean error seems to be more promising forpotential use of these chemical shifts for assisting NMRassignment In addition if more NMR data for the same classof molecules would be available the mean and intercept of

the linear regression between theoretical and experimentalvalues could be used for scaling and interextrapolation ofcomputational data in order to further minimize the meanerror of the predicted values For 3JH-H the correlationsare slightly lower and mean errors are higher The valuescalculated by GIAO 3JH-H values for H1-H2 H2-H3 for allanalyzed molecules and for H4-H5 for GlcNAcGlcNAc(6S)are underestimated while the corresponding 3JH-H valuesfor H4-H5 for GalNAcGalNAc(6S) are overestimated(Table 6) For H3-H4 H5-H6 and H5-H7 both theoreticaloverestimation and underestimation in comparison to theexperiment were obtained Scaling of the calculated 3JH-Hbased on the further obtained experimental data would assistthe creation of a quantitative procedure to be used in NMRassignment for this class of molecules

35 Sulfation Degree and Methyl-Group Chemical Shifts inAcetyl Group of GlcGalNAc Derivatives According to ourcomputational data despite the limitations for the chemicalshifts calculations described above we can clearly see ageneral increase of the H9H10H11 chemical shift valueaveraged for all ggtggt conformations with an increasein the sulfation of the monosaccharides (SupplementaryTable 18) whereas for protons H10 and H11 there isonly one significant increase of the chemical shift whena monosaccharide is sulfated once the increase of thechemical shift value for proton H9 is significant in theorder GlcGalNAc GlcGalNAc(4S) GlcGalNAc(6S) andGlcGalNAc(46S) (Supplementary Table 19) These resultsshow that despite the expected moderate accuracy in theprediction of chemical shifts the trend for such an importantparameter as net sulfation of the monosaccharide beinganalyzed by the calculations of the methyl-group chemicalshifts in the acetyl group of GlcGalNAc derivatives agreeswith the trend observed by NMR experimental data for thepolymeric GAGs with different net sulfation degree

4 Conclusions

In this work we applied QM methodology in order toanalyze the conformational space and NMR parameters


(chemical shifts and 3JH-H) of GAG monosaccharide blocksWe investigated perspectives and limitations of the appli-cability of GIAO methodology for the assistance to NMRanalysis of GAGs We observed that in such conformationalanalysis the choice of the model for QM calculation has asignificant impact on the results Comparison of our QMand MD results for gggttg conformations distribution forGlcNAc and GalNAc stressed the importance of the use ofexplicit solvent for conformational analysis of saccharides bytheoretical approaches We found that calculated chemicalshifts could be used for the analysis of the sulfation positionof GAG monosaccharide blocks as well as the net sulfationwhereas 3JH-H could be useful for both sulfation positionand ring conformation analysis Despite being promising ourresults suggest that more experimental data are needed foroptimization of the theoretically obtained parameters beforebeing used to support NMR assignment

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Authorsrsquo Contribution

Sergey A Samsonov and Stephan Theisgen contributedequally to this paper

Acknowledgments

This work was supported by the German Research Council(SFB-TRR67 A2 A6 A7) The authors would like to thankthe ZIH at TU Dresden for providing high-performancecomputational resources and Ralf Gey for technical assis-tance They acknowledge support by the German ResearchFoundation and the Open Access Publication Fund of the TUDresden

References

[1] R L Jackson S J Busch and A D Cardin ldquoGlycosamino-glycans Molecular properties protein interactions and role inphysiological processesrdquo Physiological Reviews vol 71 no 2 pp481ndash539 1991

[2] U Lindahl and J-P Li ldquoInteractions between heparan sulfateand proteins-design and functional implicationsrdquo InternationalReview of Cell andMolecular Biology vol 276 pp 105ndash159 2009

[3] M L DeMarco and R J Woods ldquoStructural glycobiology agame of snakes and laddersrdquo Glycobiology vol 18 no 6 pp426ndash440 2008

[4] T R Rudd M A Skidmore M Guerrini et al ldquoThe conforma-tion and structure of GAGs recent progress and perspectivesrdquoCurrentOpinion in Structural Biology vol 20 no 5 pp 567ndash5742010

[5] E Fadda and R J Woods ldquoMolecular simulations of car-bohydrates and protein-carbohydrate interactions motivationissues and prospectsrdquo Drug Discovery Today vol 15 no 15-16pp 596ndash609 2010

[6] Q Liu and J W Brady ldquoAnisotropic solvent structuring inaqueous sugar solutionsrdquo Journal of the American ChemicalSociety vol 118 no 49 pp 12276ndash12286 1996

[7] H A Taha P-N Roy and T L Lowary ldquoTheoretical investiga-tions on the conformation of the 120573-D-arabinofuranoside ringrdquoJournal of Chemical Theory and Computation vol 7 no 2 pp420ndash432 2011

[8] S E Pagnotta S E McLain A K Soper F Bruni and M ARicci ldquoWater and trehalose How much do they interact witheach otherrdquo Journal of Physical Chemistry B vol 114 no 14 pp4904ndash4908 2010

[9] A Vila Verde and R K Campen ldquoDisaccharide topologyinduces slowdown in local water dynamicsrdquo Journal of PhysicalChemistry B vol 115 no 21 pp 7069ndash7084 2011

[10] K N Kirschner and R J Woods ldquoSolvent interactions deter-mine carbohydrate conformationrdquo Proceedings of the NationalAcademy of Sciences of the United States of America vol 98 no19 pp 10541ndash10545 2001

[11] Y Zhong B A Bauer and S Patel ldquoSolvation propertiesof N-acetyl-120573-glucosamine molecular dynamics study incor-porating electrostatic polarizationrdquo Journal of ComputationalChemistry vol 32 no 16 pp 3339ndash3353 2011

[12] A Almond J K Sheehan and A Brass ldquoMolecular dynamicssimulations of the two disaccharides of hyaluronan in aqueoussolutionrdquo Glycobiology vol 7 no 5 pp 597ndash604 1997

[13] A Almond A Brass and J K Sheehan ldquoDeducing polymericstructure from aqueous molecular dynamics simulations ofoligosaccharides predictions from simulations of hyaluronantetrasaccharides compared with hydrodynamic and X-ray fibrediffraction datardquo Journal of Molecular Biology vol 284 no 5pp 1425ndash1437 1998

[14] A Almond and J K Sheehan ldquoGlycosaminoglycan conforma-tion do aqueous molecular dynamics simulations agree with x-ray fiber diffractionrdquo Glycobiology vol 10 no 3 pp 329ndash3382000

[15] A Almond and J K Sheenan ldquoPredicting the molecular shapeof polysaccharides from dynamic interactions with waterrdquoGlycobiology vol 13 no 4 pp 255ndash264 2003

[16] A Almond A Brass and J K Sheehan ldquoDynamic exchangebetween stabilized conformations predicted for hyaluronantetrasaccharides comparison of molecular dynamics simula-tions with available NMR datardquo Glycobiology vol 8 no 10 pp973ndash980 1998

[17] B M Sattelle J Shakeri I S Roberts and A Almond ldquoA3D-structural model of unsulfated chondroitin from high-fieldNMR 4-sulfation has little effect on backbone conformationrdquoCarbohydrate Research vol 345 no 2 pp 291ndash302 2010

[18] G Cilpa M T Hyvonen A Koivuniemi and M-L RiekkolaldquoAtomistic insight into chondroitin-6-sulfate glycosaminogly-can chain through quantummechanics calculations andmolec-ular dynamics simulationrdquo Journal of Computational Chemistryvol 31 no 8 pp 1670ndash1680 2010

[19] S A Samsonov and M T Pisabarro ldquoImportance of IdoAand IdoA(2S) ring conformations in computational stud-ies of glycosaminoglycan-protein interactionsrdquo CarbohydrateResearch C vol 381 pp 133ndash137 2013

[20] L Jin M Hricovıni J A Deakin M Lyon and D UhrınldquoResidual dipolar coupling investigation of a heparin tetrasac-charide confirms the limited effect of flexibility of the iduronicacid on the molecular shape of heparinrdquo Glycobiology vol 19no 11 pp 1185ndash1196 2009


[21] J Angulo R Ojeda J-L De Paz et al ldquoThe activation of Fibrob-last Growth Factors (FGFs) by glycosaminoglycans influenceof the sulfation pattern on the biological activity of FGF-1rdquoChemBioChem vol 5 no 1 pp 55ndash61 2004

[22] B M Sattelle and A Almond ldquoIs N-acetyl-d-glucosamine arigid4C1 chairrdquoGlycobiology vol 21 no 12 pp 1651ndash1662 2011

[23] B M Sattelle S U Hansen J Gardiner and A Almond ldquoFreeenergy landscapes of iduronic acid and related monosaccha-ridesrdquo Journal of the American Chemical Society vol 132 no 38pp 13132ndash13134 2010

[24] V Babin andC Sagui ldquoConformational free energies ofmethyl-120572-L-iduronic and methyl-Β-D-glucuronic acids in waterrdquo Jour-nal of Chemical Physics vol 132 no 10 Article ID 104108 2010

[25] N S Gandhi and R L Mancera ldquoCan current force fieldsreproduce ring puckering in 2-O-sulfo-120572-l-iduronic acid Amolecular dynamics simulation studyrdquo Carbohydrate Researchvol 345 no 5 pp 689ndash695 2010

[26] M Hricovıni ldquoB3LYP6-311++Glowastlowast study of structure and spin-spin coupling constant in methyl 2-O-sulfo-120572-l-iduronaterdquoCarbohydrate Research vol 341 no 15 pp 2575ndash2580 2006

[27] M Hricovıni E Scholtzova and F Bızik ldquoB3LYP6-311++Glowastlowaststudy of structure and spin-spin coupling constant in heparindisacchariderdquo Carbohydrate Research vol 342 no 10 pp 1350ndash1356 2007

[28] M Hricovıni ldquoEffect of solvent and counterions upon structureand NMR spin-spin coupling constants in heparin disaccha-riderdquo Journal of Physical Chemistry B vol 115 no 6 pp 1503ndash1511 2011

[29] A van der Smissen V Hintze D Scharnweber et al ldquoGrowthpromoting substrates for human dermal fibroblasts providedby artificial extracellular matrices composed of collagen I andsulfated glycosaminoglycansrdquo Biomaterials vol 32 no 34 pp8938ndash8946 2011

[30] J Salbach T D Rachner M Rauner et al ldquoRegenerativepotential of glycosaminoglycans for skin and bonerdquo Journal ofMolecular Medicine vol 90 no 6 pp 625ndash635 2011

[31] J Salbach S Kliemt M Rauner et al ldquoThe effect of the degreeof sulfation of glycosaminoglycans on osteoclast function andsignaling pathwaysrdquo Biomaterials vol 33 no 33 pp 8418ndash84292012

[32] Chemical Computing Group Inc MOE v200506 2006[33] M J Frisch G W Trucks H B Schlegel et al ldquoGaussian 09

Revision A1rdquo Gaussian Inc Wallingford CT 2009[34] A D Becke ldquoA new mixing of Hartree-Fock and local density-

functional theoriesrdquoThe Journal of Chemical Physics vol 98 no2 pp 1372ndash1377 1993

[35] F A Momany M Appell J L Willett U Schnupf and WB Bosma ldquoDFT study of 120572- and 120573-D-galactopyranose at theB3LYP6-311++Glowastlowast level of theoryrdquo Carbohydrate Research vol341 no 4 pp 525ndash537 2006

[36] J R Cheeseman M J Frisch F J Devlin and P J StephensldquoAb initio calculation of atomic axial tensors and vibrationalrotational strengths using density functional theoryrdquo ChemicalPhysics Letters vol 252 no 3-4 pp 211ndash220 1996

[37] R Jain T Bally and P R Rablen ldquoCalculating accurate protonchemical shifts of organic molecules with density functionalmethods and modest basis setsrdquo Journal of Organic Chemistryvol 74 no 11 pp 4017ndash4023 2009

[38] T Bally and P R Rablen ldquoQuantum-chemical simulation of1H NMR spectra 2 Comparison of DFT-based proceduresfor computing proton-proton coupling constants in organic

moleculesrdquo Journal of Organic Chemistry vol 76 no 12 pp4818ndash4830 2011

[39] M Cossi N Rega G Scalmani and V Barone ldquoEnergiesstructures and electronic properties of molecules in solutionwith the C-PCM solvation modelrdquo Journal of ComputationalChemistry vol 24 no 6 pp 669ndash681 2003

[40] K N Kirschner A B Yongye S M Tschampel et al ldquoGLY-CAM06 a generalizable biomolecular force field carbohy-dratesrdquo Journal of Computational Chemistry vol 29 no 4 pp622ndash655 2008

[41] D Case T Darden T Cheatham et al ldquoAmber 11 UsersrsquoManualrdquo

[42] A Pichert S A Samsonov S Theisgen et al ldquoCharacterizationof the interaction of interleukin-8 with hyaluronan chondroitinsulfate dermatan sulfate and their sulfated derivatives byspectroscopy and molecular modelingrdquo Glycobiology vol 22no 1 pp 134ndash145 2012

[43] Octave Community GNUOctave 2012 httpwwwgnuorgsoftwareoctave

[44] RDevelopment Core Team R A Language and Environment forStatistical Computing R Foundation for Statistical ComputingVienna Austria 2006

[45] R Marchessault and S Perez ldquoConformations of the hydrox-ymethyl group in crystalline aldohexopyranosesrdquo Biopolymersvol 18 no 9 pp 2369ndash2374 1979

[46] Y Nishida H Ohrui andHMeguro ldquo1H-NMR studies of (6r)-and (6s)-deuterated d-hexoses assignment of the preferredrotamers about C5C6 bond of D-glucose and D-galactosederivatives in solutionsrdquo Tetrahedron Letters vol 25 no 15 pp1575ndash1578 1984

[47] H Ohrui Y Nishida M Watanabe H Hori and H Meguroldquo1H-NMR Studies on (6R)- and (6S)-deuterated (1ndash6)-linkeddisaccharides assignment of the preferred rotamers aboutC5-C6 bond of (1ndash6)-disaccharides in solutionrdquo TetrahedronLetters vol 26 no 27 pp 3251ndash3254 1985

Submit your manuscripts athttpwwwhindawicom

ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

CorrosionInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Polymer ScienceInternational Journal of



CeramicsJournal of


CompositesJournal of

NanoparticlesJournal of



International Journal of

Biomaterials


NanoscienceJournal of

TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Journal of

NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

CrystallographyJournal of


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


CoatingsJournal of

Advances in

Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Smart Materials Research



MetallurgyJournal of


BioMed Research International

MaterialsJournal of


Nano

materials


Journal ofNanomaterials


inmore detail [16] Regarding the sulfation patterns of GAGscombination ofNMRwithMD [17] and quantummechanical(QM) [18] approaches were successfully applied to revealthe impact of sulfation effects on GAGs structure in termsof dynamic behaviour of glycosidic linkages However notonly glycosidic linkage conformations but also sugar ringpuckering could be decisive for the biologic relevance andthe specificity of GAGprotein interactions [19] In the case ofheparin it is supposed that the conformational flexibility ofthe free heparin molecule is not dramatically affected by theflexibility of the IdoUA(2S) sugar rings [20] Nevertheless itwas reported that in the complex of heparin pentasaccharidewith FGFR one of the IdoUA(2S) adopted the 2S

0ring

conformation whereas the rest of IdoUA(2S) residues were inthe 1C

4ring conformation providing high specificity for the

formation of this GAGprotein complex [21] Therefore it isimportant to understand the basic rules governing the ringconformation preferences for individual monosaccharideblocks of the GAGmoleculesThe ring conformational spacefor several GAG mono- and disaccharides for GlcNAc andits N- 3-O and 6-O sulfated derivatives [22] GlcUA IdoUA[23 24] IdoUA(2S) [23 25 26] and heparin disaccharides[27 28] was extensively analyzed in recent studies by meansof MD QM and NMR approaches demonstrating agree-ment and complementarity of these methodologies Thissuggests a high potential of the use of theoretical approachesfor the assistance in interpretation of NMR experimentaldata Interestingly despite the above-mentioned importantrole of solvent the use of an implicit solvent model (incontrast to the use of explicit solvent molecules) does notimprove agreement between spin-spin coupling parame-ters calculated by QM and measured experimentally byNMR [26]

In addition to natural GAGs artificial GAGs withdistinct sulfation patterns are promising components forfunctional biomaterials targeted for extracellular artificialmatrix engineering since additional sulfate groups couldmodulate specific binding of growth factors and therebyinfluence wound healing [29ndash31] Unfortunately sometimesonly the net sulfation degree of GAGs used in the experi-ments but not the exact sulfation pattern is known whichrenders assignment of NMR spectra for the following eluci-dation of structure-function relationships more challengingTherefore theoretical analysis of the structural propertiesof sulfated GAG monosaccharides and calculation of theirNMR chemical shifts and spin-spin coupling constants couldbe essential for the assistance in NMR experimental datainterpretation Along these lines it was shown that spe-cific sulfation patterns of some GlcNAc derivatives inducechanges in ring puckering preferences [22] Here we sys-tematically study sulfated derivatives of GlcNAc GalNAcIdoUA and GlcUA with varying degrees of net sulfationwhich represent the building blocks for heparin hyaluronicacid chondroitin sulfate and dermatan sulfate In particularwe analyze conformational preferences of the sugar ringsusing a QM approach the impact of sulfation and usedpolymerization models Furthermore we calculate NMRparameters using several computational models which pro-vide GAG monosaccharide conformational QM dictionary

data For GlcNAc GlcNAc(6S) GalNAc GalNAc(4S) andGalNAc(6S) we compare our calculated parameters withexperimental data on 13C and 1H chemical shifts and proton-proton spin-spin coupling constants ( 3JH-H) and we discussthe potential accuracy of this methodology For GlcNAc andGalNAc sulfated derivatives the conformational space ofthe dihedral angle around the C5ndashC6 bond is analyzed andcomparedwithinQMandMDapproachesThe data obtainedin this work help to get a deeper insight in the potential andlimitations of state-of-the-art computational methods usedto complement NMR experiment interpretation for GAGmolecules

2 Materials and Methods

21 Quantum Mechanical Calculation Thefollowingmonosaccharides (Figure 1) and their O1-methylatedvariants (abbreviated with M- in Tables 1 and 2) wereused for QM calculations GlcNAc (120573-D-N-acetylglu-cosamine) GlcNAc(4S) (4-O-sulfo-120573-D-N-acetylglucosa-mine) GlcNAc(6S) (6-O-sulfo-120573-D-N-acetylglucosamine)GlcNAc(46S) (46-O-disulfo-120573-D-N-acetylglucosamine)GalNAc (120573-D-N-acetylgalactosamine) GalNAc(4S) (4-O-sulfo-120573-D-N-acetylgalactosamine) GalNAc(6S) (6-O-sulfo-120573-D-N-acetylgalactosamine) GalNAc(46S) (46-O-disulfo-120573-D-N-acetylgalactosamine) GlcUA (120573-D-glucuronic acid)GlcUA(2S) (2-O-sulfo-120573-D-glucuronic acid) GlcUA(3S)(3-O-sulfo-120573-D-glucuronic acid) GlcUA(23S) (23-O-disulfo-120573-D-glucuronic acid) IdoUA (120572-L-iduronic acid)IdoUA(2S) (2-O-sulfo-120572-L-iduronic acid) and IdoUA(3S)(3-O-sulfo-120572-L-iduronic acid) IdoUA(23S) (23-O-disulfo-120572-L-iduronic acid)

First the molecules were built in MOE [32] in 1C4 4C1

and 2S0ring conformations For GlcGalNAc sulfated deriva-

tives gt tg and gg conformations were built correspond-ing to the values of dihedral angle 120596 = (O6ndashC6ndashC5ndashO5)of sim300∘ sim60∘ and sim180∘ respectively [10] Na+ counterionswere manually added to the systems with a nonzero netcharge and their positions were subsequently optimized byAMBER99 force field in MOE The geometry optimizationof these structures was carried out with GAUSSIAN 09 [33]using B3LYP functional [34] with 6-31+G(d) basis set Singlepoint energies were calculated using the B3LYP6311ppddmethod which was shown to be appropriate for energy cal-culations for carbohydrates [35] For each monosaccharidethe relative energies were calculated using the energy of themost stable conformation as reference GIAO methodologyimplemented within GAUSSIAN [36] was used to calculateNMR parameters B3LYP6311+G(2dp) for chemical shiftsandB3LYPaug-cc-pVDZ for spin-spin coupling constants asthese levels of theory demonstrated highest reliability in thecalibration studies [37 38] TMS (tetramethylsilane)was usedas a reference to calculate 13C- and 1H-chemical shifts For thecalculations carried out in solvent PCM solvent model [39]was used

22 Molecular Dynamics Calculations For MD simulationsthe GLYCAM06 force field [40] implemented in the AMBER


H7

C6H6

C5

H4C1

H5

H2

C4 C3C2

H1

H3

H8 H11

H10H9

H7

C6H6

C5

H4 C1

C4H5

H2

C3C2 H1

H3

H8

H11

H10

H9

GlcNAc GalNAc

C6

C5H4

C4H5

H2

C3C2

C1

H1

H3

C6

C5

C1

H4

C4

H5

H2

C3

C2 H1

H3

GlcUA IdoUA







3 (GlcUA-


3 (GlcUA(3S)-


3 (GlcUA-



3


3 (Glc-


3






0) kcalmol

Saccharides 119892119892 119892119905 119905119892 119892119892 119892119905 119905119892 119892119892 119892119905 119905119892


141 579


12870 049



2O with TSP as a










0) kcalmol




4and 2S













0was calculated to




1ring conformation


4conformation with


1conformation


0and 4C

1


4








0and 1C

4but not in

the 4C1conformation










Monosaccharide 119892119892 119892119905 119905119892






















0




1









13C 0999 (0998) 0994 0995 0993 (0998) 0995 0994 (0996)119877Spearman

13C 1000 1000 1000 0893 (0929) 1000 0959 (0965)119877Pearson

1H 0980 0980 0932 0991 0939 0961119877Spearman








4 Conclusions








Acknowledgments


References

























































CeramicsJournal of







Biomaterials




Journal of


Journal of





CoatingsJournal of

Advances in








MaterialsJournal of


Nano

materials




H7

C6H6

C5

H4C1

H5

H2

C4 C3C2

H1

H3

H8 H11

H10H9

H7

C6H6

C5

H4 C1

C4H5

H2

C3C2 H1

H3

H8

H11

H10

H9

GlcNAc GalNAc

C6

C5H4

C4H5

H2

C3C2

C1

H1

H3

C6

C5

C1

H4

C4

H5

H2

C3

C2 H1

H3

GlcUA IdoUA







3 (GlcUA-


3 (GlcUA(3S)-


3 (GlcUA-



3


3 (Glc-


3






0) kcalmol

Saccharides 119892119892 119892119905 119905119892 119892119892 119892119905 119905119892 119892119892 119892119905 119905119892


141 579


12870 049



2O with TSP as a










0) kcalmol




4and 2S













0was calculated to




1ring conformation


4conformation with


1conformation


0and 4C

1


4








0and 1C

4but not in

the 4C1conformation










Monosaccharide 119892119892 119892119905 119905119892






















0




1









13C 0999 (0998) 0994 0995 0993 (0998) 0995 0994 (0996)119877Spearman

13C 1000 1000 1000 0893 (0929) 1000 0959 (0965)119877Pearson

1H 0980 0980 0932 0991 0939 0961119877Spearman








4 Conclusions








Acknowledgments


References

























































CeramicsJournal of







Biomaterials




Journal of


Journal of





CoatingsJournal of

Advances in








MaterialsJournal of


Nano

materials







0) kcalmol

Saccharides 119892119892 119892119905 119905119892 119892119892 119892119905 119905119892 119892119892 119892119905 119905119892


141 579


12870 049



2O with TSP as a










0) kcalmol




4and 2S













0was calculated to




1ring conformation


4conformation with


1conformation


0and 4C

1


4








0and 1C

4but not in

the 4C1conformation










Monosaccharide 119892119892 119892119905 119905119892






















0




1









13C 0999 (0998) 0994 0995 0993 (0998) 0995 0994 (0996)119877Spearman

13C 1000 1000 1000 0893 (0929) 1000 0959 (0965)119877Pearson

1H 0980 0980 0932 0991 0939 0961119877Spearman








4 Conclusions








Acknowledgments


References

























































CeramicsJournal of







Biomaterials




Journal of


Journal of





CoatingsJournal of

Advances in








MaterialsJournal of


Nano

materials







0) kcalmol




4and 2S













0was calculated to




1ring conformation


4conformation with


1conformation


0and 4C

1


4








0and 1C

4but not in

the 4C1conformation










Monosaccharide 119892119892 119892119905 119905119892






















0




1









13C 0999 (0998) 0994 0995 0993 (0998) 0995 0994 (0996)119877Spearman

13C 1000 1000 1000 0893 (0929) 1000 0959 (0965)119877Pearson

1H 0980 0980 0932 0991 0939 0961119877Spearman








4 Conclusions








Acknowledgments


References

























































CeramicsJournal of







Biomaterials




Journal of


Journal of





CoatingsJournal of

Advances in








MaterialsJournal of


Nano

materials





0was calculated to




1ring conformation


4conformation with


1conformation


0and 4C

1


4








0and 1C

4but not in

the 4C1conformation










Monosaccharide 119892119892 119892119905 119905119892






















0




1









13C 0999 (0998) 0994 0995 0993 (0998) 0995 0994 (0996)119877Spearman

13C 1000 1000 1000 0893 (0929) 1000 0959 (0965)119877Pearson

1H 0980 0980 0932 0991 0939 0961119877Spearman








4 Conclusions








Acknowledgments


References

























































CeramicsJournal of







Biomaterials




Journal of


Journal of





CoatingsJournal of

Advances in








MaterialsJournal of


Nano

materials





Monosaccharide 119892119892 119892119905 119905119892






















0




1









13C 0999 (0998) 0994 0995 0993 (0998) 0995 0994 (0996)119877Spearman

13C 1000 1000 1000 0893 (0929) 1000 0959 (0965)119877Pearson

1H 0980 0980 0932 0991 0939 0961119877Spearman








4 Conclusions








Acknowledgments


References

























































CeramicsJournal of







Biomaterials




Journal of


Journal of





CoatingsJournal of

Advances in








MaterialsJournal of


Nano

materials












0




1









13C 0999 (0998) 0994 0995 0993 (0998) 0995 0994 (0996)119877Spearman

13C 1000 1000 1000 0893 (0929) 1000 0959 (0965)119877Pearson

1H 0980 0980 0932 0991 0939 0961119877Spearman








4 Conclusions








Acknowledgments


References

























































CeramicsJournal of







Biomaterials




Journal of


Journal of





CoatingsJournal of

Advances in








MaterialsJournal of


Nano

materials






13C 0999 (0998) 0994 0995 0993 (0998) 0995 0994 (0996)119877Spearman

13C 1000 1000 1000 0893 (0929) 1000 0959 (0965)119877Pearson

1H 0980 0980 0932 0991 0939 0961119877Spearman








4 Conclusions








Acknowledgments


References

























































CeramicsJournal of







Biomaterials




Journal of


Journal of





CoatingsJournal of

Advances in








MaterialsJournal of


Nano

materials









Acknowledgments


References

























































CeramicsJournal of







Biomaterials




Journal of


Journal of





CoatingsJournal of

Advances in








MaterialsJournal of


Nano

materials







































CeramicsJournal of







Biomaterials




Journal of


Journal of





CoatingsJournal of

Advances in








MaterialsJournal of


Nano

materials



research article glycosaminoglycan monosaccharide blocks...

Documents