Biomacromolecules 2003, 4, 1362-1371
1362
Structure and Properties of a Bacterial Polysaccharide Named Fucogel O. Guetta, K. Mazeau, R. Auzely, M. Milas, and M. Rinaudo* Centre de Recherche sur les Macromole´ cules Ve´ ge´ tales, CNRS, associated with Universite´ Joseph Fourier, BP 53, 38041 Grenoble Cedex 9, France Received April 14, 2003; Revised Manuscript Received June 24, 2003
The chemical structure of a polysaccharide named Fucogel was characterized and the position of acetylation was identified by NMR. A conformational analysis was performed on this 3-sugar repeating unit. From this, the persistence length, characterizing the stiffness of the polysaccharide, was determined and the role of the presence of acetyl group, reducing the stiffness, was pointed out. The helical conformations were also predicted, one of these being in agreement with X-ray data obtained on a similar polysaccharide. Experimental characterization of the native and deacetylated polysaccharides was developed. SEC experiments allowed us to determine the molar mass and the persistence length on the deacetylated polysaccharide. The value is in good agreement with that predicted from the molecular modeling. Microcalorimetry, rheology, and fluorescence spectroscopy demonstrated respectively that no helical conformation exists in solution but that loose interchain interactions due to the acetyl substituents exist in dilute solutions. Introduction Fucogel, a highly viscous polysaccharide developed by BioEurope and marketed by SOLABIA, is produced by fermentation of an original strain of Klebsiella pneumoniae I-1507.1 Fucogel success in the cosmetic industry is based on its original psychosensorial qualities (a very soft touch), hydrating and self-emulsifying properties. Recent investigations demonstrated that Fucogel is effective as antiadhesive against Staph. epidermidis compared to several other oligo or polysaccharides.1 In the literature, two other polysaccharides, produced by different Klebsiella pneumoniae strains, are mentioned having similar structure. Klebsiella K-type 632 produces a polysaccharide with the repeating unit consisting of 3)-R-L-Fucp-(1 f 3)-R-D-Galp-(1f3)-R-D-GalpA-(1f Acetyl substituent was indicated at a low yield (0 3.6
torsion angles are rotated in 10° increment over the whole angular range, and the resulting conformations are relaxed (with the exception of Φ and Ψ) using the block diagonal method with a termination criterion of n0.000 03 kcal/mol (n, number of atoms). The (Φ, Ψ) adiabatic maps of the disaccharide look very classical and, consequently, are not shown in this paper. These maps will be presented in graphical format in the CERMAV data bank at http://www.cermav.cnrs.fr. The accessible areas (with relative energy lower than 10 kcal/ mol) of the potential energy surfaces are all located for Φ around the gauche orientations. The exo-anomeric effect occurs in all saccharidic structures, because of the presence of the hemiacetal function. From calculations on model compounds together with experimental observations, the following general empirical rules about the conformational preferences of hemiacetals have emerged. The conformational potential of the aglycone of the glycosidic C-1-O-1 bond (torsion angle φ) is strongly influenced by electrostatic interactions between the dipoles along the axes of the sp3type lone pairs orbitals on the oxygen atoms. The most stable conformer is that having on oxygen atoms (O-5 and O-1) the minimum number of syn-axial lone-pair interactions that can cause dipolar repulsion. For the two disaccharides whose nonreducing end is a R-D in the 4C1 chair form (R-D-Galp(1 f 3)-R-D-GalpA, and R-D-GalpA-(1 f 3)-R-L-Fucp), the +60° orientation is preferred because it has no such interactions whereas the two others ones, at +180° and -60°, have a single syn-axial lone-pair orbital interaction. The nonreducing end of the remaining dimer is a R-L configuration in the 1C4 chair form: R-L-Fucp-(1 f 3)-R-D-Galp. Hereagain, only one orientation of the aglycone (-60°) have no lone pair orbital interaction whereas the two other conformers (+60° and +180°) have one such interaction. Our results are in good agreement with the above-mentioned rule concerning the φ conformational preferences that qualitatively explain the exo-anomeric effect: the low energy axis is centered at Φ values of +60° for (R-D-Galp-(1 f 3)-R-D-GalpA, and R-D-GalpA-(1 f 3)-R-L-Fucp dimers and at -60° for R-L-Fucp-(1 f 3)-R-D-Galp. On the contrary to the Φ angle, the rotational profile of the torsion angle Ψ depends mainly on steric constraints. Accessible values of
hydrogen bond
O5(Fuc)‚‚‚O2(Gal) O5(Gal)‚‚‚O2(GalA) O5(Gal)‚‚‚O4(GalA)
O5(GalA)‚‚‚O2(Fuc) O5(Gal)‚‚‚O2(GalAac)
O5(GalAac)‚‚‚O2(Fuc)
this angle (with energy lower than the 10 kcal/mol limit) cover the whole angular domain. The classical three staggered orientations correspond to energy minima. All of the computed maps exhibit two distinct low energy regions. The main domain encompasses two wells which corresponds to A and B minima. They are separated by a low energy barrier; thus, conformational interconversions between these two areas are expected easy. An additional domain occurs, and it corresponds to the C minimum. It is separated from the main domain by a larger energy barrier. The main geometrical and energetic characteristics of the three favored conformations for the five disaccharides are given in Table 2. In this table are also reported the interresidue hydrogen bonds, based on the computed oxygen to oxygen distances. Analysis of the calculated data indicate that the minimum minimorum for each disaccharide is not stabilized by hydrogen bonds which indicates that such interactions are not essential for the stabilization of the structure. Comparing the results for acetylated andnon acetylated disaccharides, it comes out that the influence of the acetyl group on the energy maps is not graphically significant; some small differences can however be identified by inspection of Table 2. In particular hydrogen bonds appear when the polysaccharide is deacetylated. Disordered Conformations. To evaluate the average dimensions of Fucogel polymer chains in solution, in θ conditions, large samples of chains in random conformations were generated using the Monte Carlo technique as implemented in the METROPOL program.16 The detailed description of the procedure is reported elsewhere.16 The conformations were selected using the Metropolis algorithm17 and distributed according to Boltzmann statistics. The approach uses the results of the adiabatic map established for the disaccharide subunits and does not take into account the interactions between residues in longer segments; calculations thus refer to a nonperturbed state. Statistical samples of 3000 polymer chains, each containing 1500 glycosyl residues, were constructed for Fucogel (mean error deviation not exceeding 3%) at different temperatures from 298 to 373 K. The extension and relative stiffness of the chain are described by the characteristic ratio (C∞) and the related
1368
Biomacromolecules, Vol. 4, No. 5, 2003
Figure 6. Persistence length, Lp (Å), calculated as a function of the number of glycosyl residues (×), of the acetylated ([) and deacetylated (9) Fucogel chain, at 298 K.
parameter Lp (persistence length).18 These parameters were evaluated as follows: C∞ ) limCx ) lim〈r2(x)〉/x Lo with r(x) ) xf∞
xf∞
d(O-1(1)-O-1(x)) The angular brackets indicate the average of the square endto-end distance r2 over all of the conformations of polymer chains, Lo is the bond length connecting two adjacent glycosidic oxygen atoms, and x is the number of glycosyl residues. The persistence length, Lp(x), directly related to Cx, is defined as the projection of the end-to-end distance vector r on the first bond of the chain. It represents the maximum contour length over which a correlation between monomeric units persists. The asymptotic values of Cx and Lp(x) when x increases, corresponding to a range of x where they become independent of the degree of polymerization (here x > 500), and equals C∞ and Lp, respectively. This intrinsic persistence length (Lp) reflected the stiffness of the chain and can be directly compared to the experimental parameters in θ conditions from the determination of the radius of gyration and molar mass obtained by SEC experiments. Using this procedure, average persistence lengths of 64 and 88 Å are predicted for the acetylated and the deacetylated polysaccharides, respectively (Figure 6). These values are characteristic of semirigid polysaccharides; other polysaccharides were described previously, and their persistence length was found in the range of 80-120 Å for hyaluronan, chitosan, and galactomannan, respectively.19-21 The absence of acetyl groups induces an increase of the rigidity of chains that may be due to the presence of additional hydrogen bonds. The variations of the persistence length (Lp) at different temperatures are shown in Figure 7. Increasing the temperature regularly decreases Lp and C∞. Regular Helices. Several possible stable single regular helical forms of acetylated and deacetylated Fucogel were established using the program POLYS.22 The helices are characterized by the two helical parameters n and h, where n is the number of repeating units (trisaccharide unit) per turn of the helix and h is the projection of one repeating unit on the helical axis. The chirality of the helix is described by a sign attributed to n: a positive value of n corresponds to a right-handed helix and a negative value to a left-handed
Guetta et al.
Figure 7. Persistence length (Lp) predicted by molecular modeling as a function of temperature.
helix. The study was performed on 10 repeating units. The different single helical forms were extrapolated from the combination of all of the lowest energy conformations of the different disaccharides. A total of 27 helices were generated, and unreasonably high energy helices were rejected. In the POLYS procedure, sets of (Φ,Ψ) values, correlated to the local minima, are applied on each glycosidic linkages to construct the polymer chain in an helical form. Then, POLYS optimizes the glycosidic bond torsion angles values in order to obtain the nearest integral n-fold helical structure. The integral helices are then constructed again using those optimized (Φ,Ψ) values. Finally, all of these helical structures were minimized (with ) 4 to simulate the solid state). Side and top views of the different structures predicted for the acetylated Fucogel in the same time as the helical parameters and relative energy are given in Figure 8. Six extended stable helices ranging from the right-handed 2-fold symmetry to right-handed 4-fold symmetry are predicted. When the polysaccharide is deacetylated, no significant differences were detected. In the case of the deacetylated polysaccharide, the helical parameters obtained for the more stable helice (n ) 2, h ) 12.6 nm) are in good agreement with those determined by Elloway et al.15 by X-ray diffraction (n ) 2, h ) 11.85 nm) on the K-63 deacetylated polysaccharide having the same repeating unit. 3. Solution Properties. As predicted by molecular modeling the presence of the acetyl group decreases the local stiffness of the molecules. Nevertheless, because of their hydrophobic character, they may play a role on the behavior in aqueous solution such as loose interactions. In the following, we will compare the behavior of the acetylated and the deacetylated Fucogel in dilute and semidilute solutions. Characterization of Polysaccharides. The weight average molecular weights were determined by steric exclusion chromatography. Figure 9 shows aggregates for the acetylated Fucogel that disappeared with the deacetylation. In this last case, the weight-average molecular weight is about 106g/mol. Intrinsic viscosities and Huggins constant, determined in the Newtonian regime, are given in the Table 3. The value
Biomacromolecules, Vol. 4, No. 5, 2003 1369
A Bacterial Polysaccharide Named Fucogel
Figure 8. Side (a) and top (b) views of the different stable helices proposed for acetylated Fucogel with their different helical parameters. Table 3. Characteristic Parameters for Dilute Solutions of Fucogel, at 25 °C, in 0.1 M NaCla Fucogel
[η] (cm3/g)
k′
Mw (g/mol)
Lp expt. (nm)
Lp calcd (nm)
acetylated deacetylated
1670 1500
0.55 0.42
106
10.2
6.4 8.8
a The experimental and calculated persistence lengths of acetylated and deacetylated fucogel are also given.
Figure 9. SEC chromatograms for acetylated and deacetylated Fucogel. T ) 30 °C, eluent: 0.1 M NaNO3.
of k′, higher for the acetylated than for the deacetylated sample, confirms the presence of aggregates. Conformation in Solution. It was established from X-ray diffraction23 and by the present modeling that deacetylated K-63 polysaccharide adopts an extended 2-fold helix. For this reason, we used microcalorimetry and optical rotary power to evidence the existence or the absence of a conformational transition between an ordered state (helix) to a disordered one (random coil) for the acetylated or the
deacetylated Fucogel in aqueous solution. The study was realized from 5 to 85 °C, for two concentrations of polymer (10 and 20 g/L) in water and 0.4 M NaCl supposed to favor the stabilization of the helical structure if it exists in solution. The signals obtained from these techniques give flat lines indicating no conformational transition for the polysaccharide in solution, whatever the conditions. Determination of Persistence Length. The polymer chain rigidity is characterized by the persistence length. For polyelectrolytes, electrostatic repulsions between the charges can strongly influence this parameter. The total persistence length, LT, can be written as the sum of two contributions: LT ) Lp + Le where Lp is the intrinsic persistence length due to the rigidity of the corresponding uncharged chain and Le is the electrostatic persistence length arising from the repulsion between ionic sites, introduced by Odijk.24,25 Le is
1370
Biomacromolecules, Vol. 4, No. 5, 2003
Guetta et al.
Figure 10. Reduced viscosity of acetylated ([) and deacetylated (]) Fucogel in 0.1 M NaCl, at 25 °C, as a function of the overlap parameter C[η].
given by the following relation when λ, the charge parameter, is lower than 1: Le )
0.3375λ2 λ2 ) 4κ2Q (Cp + Cs)
where λ is equal to 0.46 in our case, κ-1 is the DebyeHu¨ckel screening length related to the concentration of the counterions, and Q is the Bjerrum length (7.13 Å, in water at 25 °C). Benoit and Doty26 have expressed the radius of gyration RG as a function of the contour length, L, and the total persistence length, in θ conditions:
( ( ))
Lp4 LLp 2Lp3 L 2 - Lp + -2 2 1 〈R G〉θ ) 3 L L L p 2
For L . Lp: 〈R2G〉θ )
LLp 3
For polyelectrolyte, at a finite salt concentration (Cs), excluded volume effects due to electrostatic repulsions are very important. So an expansion parameter, Rs,el, was introduced by Odijk and Houwaart27 and Fixman and Skolnick,28 in addition to the electrostatic contribution in the total persistence length (LT ) Lp + Le). Then, it comes R2G )
LLT R 2 3 s,el
An average persistence length can be obtained experimentally from determination of the radius of gyration RG and molecular weight, determined by steric exclusion chromatography. The total contour length was determined from the molar mass and the chemical structure of the polysaccharides, considering a molar mass for the repeating unit of 506 and 548 g/mol for the deacetylated and acetyled polysaccharides, respectively. In the experimental conditions (Cs ) 0.1 M
NaNO3), we can considered that Le , Lp and Rs,el ≈ 1. The average error deviation is about 4%; the values are given in Table 3. The experimental value for the deacetylated polysaccharide is in good agreement with the calculated value obtained by molecular modeling. The existence of aggregates shown in Figure 9 for the acetylated polysaccharide does not permit to determine experimentally Lp. Despite the intrinsic limitations of the theoretical calculations, the agreement obtained between theoretical and experimental values, in θ conditions, suggests that the conformational behavior of the polysaccharide chain may be properly simulated using the reasonable approximations adopted. Concentration Dependence. Figure 10 shows a decrease of the reduced viscosity when the polysaccharide is deacetylated. This could be due to molecular interactions, occurring in all of the range of polysaccharide concentrations but more efficient for Cp > 1 g/L, induced by the presence of acetyl groups and favored by the rigidity of the polysaccharides. The existence of hydrophobic interchain interactions could explain this phenomena. To evidence their contribution, fluorescence spectroscopy experiments were realized in the presence pyrene as a probe for hydrophobic environment. On the emission spectrum of pyrene, the ratio of the intensity of the first (374 nm) and the third (384 nm) peaks is very sensitive to the polarity of the medium (I1/I3 is higher in polar media).29 The study was realized for different concentrations of polymer from 0.025 to 10 g/L, in 0.1 M NaCl at 25 °C. At low concentrations (Cp < 0.1 g/L), the ratio for the acetylated and the deacetylated polysaccharides are close to that obtain for free pyrene in NaCl 0.1 M, which means that there are no hydrophobic domains. For higher concentrations (Cp > 0.1 g/L), we can observe a decrease of the polarity parameter for both polysaccharides, but it starts first for the acetylated polysaccharide; it indicates existence of associations having a hydrophobic character for the acetylated polymer. This result confirms the data from viscosity and
A Bacterial Polysaccharide Named Fucogel
Biomacromolecules, Vol. 4, No. 5, 2003 1371
producing an abnormal increase of the viscosity in dilute and semidilute solution. Acknowledgment. The authors thank Solabia Cy (France) for their financial help. References and Notes
Figure 11. Polarity parameter I1/I3 of pyrene in aqueous solutions of acetylated ([) and deacetylated (]) Fucogel in 0.1 M NaCl at 25 °C, as a function of polymer concentration. (-) Reference representing pyrene in water.
SEC measurements showing interchain interaction in the presence of acetyl groups. On the opposite, the behavior obtained with the deacetylated Fucogel superposes with that obtained for other semirigid water soluble polysaccharides.30 Conclusion In this study, the structure and the physicochemical properties in a solution of a new Klebsiella polysaccharide were investigated. The structure of the repeating unit of this polysaccharide, an acetylated charged linear trisaccharide, was confirmed and completed by a full NMR experiment (1D and 2D) allowing the determination of the configuration of the sugars and the location of the acetyl group. A conformational analysis was then realized by molecular modeling. Disordered chains were generated, and the influence of acetyl groups on structural parameters such as persistence length was studied. The acetylation induces a lowering of the rigidity of the chain, with a persistence length decreasing from 88 to 64 Å at 25 °C. It is also predicted that Lp decreases when the temperature increases. Then, the possible existence of stable helices was studied. For the deactivated polysaccharide, the most favorable one has helical parameters very close to those observed by X-ray diffraction and described in the literature on a polysaccharide having a similar structure. The rheological properties of Fucogel were then studied in 0.1 M NaCl at 25 °C. Aggregates were observed by SEC and fluorescence experiments for the acetylated polysaccharide; these aggregates are due to lose interchain interactions
(1) Paul, F. M. B.; Perry, D. F.; Monsan, P. F., US Patent 9,623,057, 1996. (2) Joseleau, J. P.; Marais, M. F. Carbohydr. Res. 1979, 77, 183-190. (3) Johansson, A.; Jansson, P. E.; Wildmalm, G. Carbohydr. Res. 1994, 253, 317-322. (4) Hakomori, S. J. Biochem. (Tokyo) 1964, 55, 205-208. (5) Wagner, G. J. Magn. Reson. 1983, 55, 151-156. (6) Bax, A.; Drobny, G. J. Magn. Reson. 1985, 61, 306-320. (7) Palmer, A. G., III.; Cavanagh, J.; Wright, P. E.; Rance, M. J. Magn. Reson. 1991, 93, 151-170. (8) Kay, L. E.; Keifer, P.; Saarinen, T. J. Am. Chem. Soc. 1992, 114, 10663-10665. (9) Schleucher, J.; Schwendiger, M.; Sattler, M.; Schmidt, P.; Schedletzky, O.; Glaser, S. J.; Sorensen, O. W.; Griesinger, C. J. Biomol. NMR 1994, 4, 301-306. (10) Taylor, R. L.; Conrad, H. E. Biochemistry 1972, 12, 1383-1388. (11) Bock, K.; Thogersen, H. Annual Reports on NMR Spectroscopy, Volume 13; Webb, G. A., Ed.; Acadedmic Press: New York, 1982; pp 1-57. (12) Baumann, H.; Jansson, P.-E.; Kenne, L. J. Chem. Soc., Perkin Trans. 1 1988, 209-217. (13) Allinger, N. L.; Yuh, Y. H.; Lii, J. H. J. Am. Chem. Soc. 1989, 111, 8551-8566. (14) Allinger, N. L.; Rahman, M.; Lii, J. H. J. Am. Chem. Soc. 1990, 112, 8293-8307. (15) Allinger, N. L.; Zhu, Z. Q. S.; Chen, K. J. Am. Chem. Soc. 1992, 114, 6120-6133. (16) Boutherin, B.; Mazeau, K.; Tvaroska, I. Carbohydr. Polym. 1997, 31, 1-12. (17) Metropolis, N.; Rosenbluth, A. W.; Rosenbluth, M. M.; Teller, A. H.; Teller, E. J. Chem. Phys. 1953, 21, 1087-1092. (18) Lapasin, R., Pricl, S., Eds.; Rheology of industrial polysaccharides, theory and applications; Chapman and Hall: Glasgow, 1995. (19) Haxaire, K.; Braccini, I.; Milas, M.; Rinaudo, M.; Perez, S. Glycobiology 2000, 10, 587-594. (20) Mazeau, K.; Perez, S.; Rinaudo, M. J. Carbohydr. Chem. 2000, 19, 1269-1284. (21) Petkowicz, C. L. O.; Rinaudo, M.; Milas, M.; Mazeau, K.; Bresolin, T.; Reicher, F.; Ganter, J. L. M. S. Food Hydrocoll. 1999, 13, 263266. (22) Engelsen, S. B.; Gos, S.; Mackie, W.; Pe´rez, S. Biopolymers 1996, 39, 417-433. (23) Elloway, H. F.; Isaac, D. H.; Atkins, E. D. T. ACS Symp. Ser. 1980, 141, 429-458. (24) Odijk, T. Polymer 1978, 19, 989-990. (25) Odijk, T. Biopolymers 1979, 18, 3111-3113. (26) Benoit, H.; Doty, P. J. Phys. Chem. 1953, 57, 958-963. (27) Odijk, T.; Houwaart, A. C. J. Polym. Sci., Polym. Phys. Ed. 1978, 16, 627-639. (28) Fixman, M.; Skolnick, J. Macromolecules 1978, 11, 863-867. (29) Kalyanasundaram, K.; Thomas, J. K. J. Am. Chem. Soc. 1977, 99, 2039-2044. (30) Milas, M.; Rinaudo, M. Polysaccharides. Structural diVersity and functional Versatility, 2nd ed.; Dumitriu, S., Ed.; Marcel Dekker: New York, 2003, in preparation.
BM030033H
