Article pubs.acs.org/JAFC
Probing Interactions between β‑Glucan and Bile Salts at Atomic Detail by 1H−13C NMR Assays Mette Skau Mikkelsen,† Sofia Bolvig Cornali,† Morten G. Jensen,‡ Mathias Nilsson,†,§ Sophie R. Beeren,∥ and Sebastian Meier*,∥,⊥ †
Faculty of Science, Department of Food Science, University of Copenhagen, Rolighedsvej 30, DK-1958 Frederiksberg C, Denmark Carlsberg Research Center, Gamle Carlsberg Vej 10, DK-1799 Copenhagen V, Denmark § School of Chemistry, University of Manchester, Oxford Road, Manchester, M13 9PL, UK ∥ Carlsberg Laboratory, Gamle Carlsberg Vej 10, DK-1799 Copenhagen V, Denmark ⊥ Department of Chemistry, Technical University of Denmark, Kemitorvet, Building 201, DK-2800 Kongens Lyngby, Denmark ‡
S Supporting Information *
ABSTRACT: Polysaccharides are prospective hosts for the delivery and sequestration of bioactive guest molecules. Polysaccharides of dietary fiber, specifically cereal (1 → 3)(1 → 4)-β-glucans, play a role in lowering the blood plasma cholesterol level in humans. Direct host−guest interactions between β-glucans and conjugated bile salts are among the possible molecular mechanisms explaining the hypocholesterolemic effects of β-glucans. The present study shows that 1H−13C NMR assays on a time scale of minutes detect minute signal changes in both bile salts and β-glucans, thus indicating dynamic interactions between bile salts and β-glucans. Experiments are consistent with stronger interactions at pH 5.3 than at pH 6.5 in this in vitro assay. The changes in bile salt and β-glucan signals suggest a stabilization of bile salt micelles and concomitant conformational changes in βglucans. KEYWORDS: bile salt, glucan, HSQC, heteronuclear NMR
■
INTRODUCTION The health benefits of dietary fibers have increasingly attracted consumer awareness and scientific interest. The health benefits of cereal foods are to a large extent accredited to hydrosoluble dietary fibers, especially mixed-linkage (1 → 3)(1 → 4)-βglucan (henceforth referred to as β-glucan).1−4 Cereal β-glucan reportedly lowers the blood plasma cholesterol levels in humans 1 and was the first cholesterol-reducing food component to be registered by the U.S. Food and Drug Administration in 1997. Due to its cholesterol-reducing capability, β-glucan can reduce the risk of cardiovascular diseases. The cholesterollowering effect of β-glucans has been attributed to the prevention of bile salt reabsorption from the small intenstine.2,5 The molecular details of the interactions between β-glucan and bile salts in the digestive tract remain unclear, not least due to challenges in detecting binding to the homopolymeric and viscous β-glucan with molecular detail.6−11 Rapid and sensitive assays that provide detailed molecular information for both interaction partners are desirable to probe interactions between polysaccharides, their fragments, and guest molecules of any sort. Previous 13C nuclear magnetic resonance (NMR) titrations had suggested that bile salt 13C NMR signals change upon the addition of β-glucan without significantly restricting the rotational tumbling and thus the line width of the bile salt signals.9 These data hint at dynamic interactions between bile salts and dietary fiber, but no chemical shift changes were previously observed for β-glucans themselves and measurements were only practical at concentrations well above the critical micelle concentration (CMC) of bile salts. Due to the © 2014 American Chemical Society
fast exchange of bile salts between monomeric and micellar forms above the CMC, changes to the bile salt 13C NMR signals could, however, also result from increased local concentrations of bile salt through confinement or excluded volume effects. In the current study we use rapid interaction assays between bile salts and β-glucan, probing the spectral signal of both binding partners and permitting rapid binding assays even below the CMC of bile salts. High-resolution 1H−13C NMR experiments of less than 30 min duration were used to provide sufficient sensitivity and resolution for detecting weak binding events. Interactions were probed in two separate experiments: (1) by titration of β-glucan to a bile salt solution of constant concentration and (2) by the titration of bile salt to a β-glucan host of constant concentration. The combined use of twodimensional 1H−13C NMR experiments and high-field NMR instrumentation leads to a ∼50-fold acceleration as compared to previously reported 13C NMR assays of bile salt binding to βglucan.6,9 In addition, the excellent sensitivity of 1H−13C NMR experiments recorded with state of the art instrumentation allows minute chemical shift changes to be detected in the βglucan host upon titration with bile salts. The detection of chemical shift changes for both components is consistent with direct dynamic interactions between β-glucan and bile salts at physiological pH values of the small intestine.12 Interactions are Received: Revised: Accepted: Published: 11472
July 18, 2014 October 28, 2014 November 6, 2014 November 6, 2014 dx.doi.org/10.1021/jf504352w | J. Agric. Food Chem. 2014, 62, 11472−11478
Journal of Agricultural and Food Chemistry
Article
Figure 1. Schematic representation of (1→3)(1→4)-β-glucan with horizontal lines indicating β(1→4) linkages and diagonal lines indicating β(1→3) linkages.
stronger at pH 5.3 than at pH 6.5. The data are consistent with a stabilization of bile salt micelles in the presence of β-glucan due to a stronger multivalent binding of the β-glucan to bile salt micelles than to the monomers. 1H−13C NMR assays thus provide a rapid means for probing interactions between βglucan and bile salts with atomic-level resolution.
■
EXPERIMENTAL SECTION
Materials. Low and medium viscosity barley β-glucans (Figure 1) were obtained from Megazyme International Ltd. (Bray, Ireland), and these samples have previously been characterized by various physicochemical methods.13 Briefly, low viscosity β-glucan had a molecular mass of 140 kDa as determined by high performance sizeexclusion chromatography and a calibration curve using five β-glucan molar mass standard samples. Upon extensive hydrolysis with lichenase, a block structure (β-(1→4) linked glucopyranoses between β-(1→3) branches) of approximately 77% DP3, 19% DP4, and 3% DP5 (molar ratio) was derived. Medium viscosity β-glucan had a molecular mass of 200 kDa and a block structure of approximately 76% DP3, 19% DP4, and 4% DP5. Samples contained less than 1% each of protein or starch. Low viscosity barley β-glucan was chosen for the experiments monitoring β-glucan signals in order to improve spectral quality at lower sample viscosity (at high viscosity molecular tumbling rates decrease and NMR signals might become broad due to rapid relaxation). Signal appearance is less of an issue when monitoring the bile salt sample due to the higher achievable concentration, and medium viscosity barley β-glucan was used in these experiments due to the reportedly higher retention of bile salts by β-glucan of higher viscosity.14 Bile salts, glycolate (GC) and taurochenodeoxycholate (TCDC) (Figure 2), along with buffer materials, sodium phosphate and sodium acetate, were obtained from Sigma-Aldrich (St. Louis, MO, USA). Buffers prepared at 100 mM concentration were sodium phosphate buffer of pH 6.5 and sodium acetate buffer of pH 5.3. Buffers were lyophilized and redissolved in D2O (Cambridge Isotope Laboratories, Tewksbury, MA, USA). Sample Preparation. Low viscosity and medium viscosity barley β-glucan were dissolved in buffer to concentrations of 0.5, 1, 1.5, or 2% (w/v) and hydrated at 80 °C for 30 min. Stock solutions of bile salt were prepared in buffer, and bile salt was added to β-glucan samples at the concentrations indicated (3, 7, 10, 14, or 28 mM). Samples containing β-glucan and bile salts were incubated at 37 °C for 120 min under continuous shaking, and pure bile salt and β-glucan reference samples, in the absence of the potential interaction partner, were treated identically. NMR Spectroscopy. Upon incubation at 37 °C for 120 min, the samples were directly transferred to 5 mm NMR sample tubes and analyzed by sensitivity-enhanced 1H−13C heteronuclear singlequantum correlation15 (HSQC) experiments with a spectral width in the indirect (13C) dimension that was reduced to permit rapid sampling of the 13C dimension at high resolution. Experiments treating β-glucan as the host were recorded with a spectral width of 5 ppm and an offset in the 13C dimension of 80 ppm. A total of 75 complex points
Figure 2. Structure of the bile salt molecules glycocholate (GC) and taurochendeoxycholate (TCDC) including their atom numbering. were recorded in the 13C dimension to sample an indirect acquisition time of 75 ms. Experiments treating bile salt as the host were recorded with a spectral width of 20 ppm centered around a spectral offset of 32 ppm. A total of 400 complex points were recorded in the 13C dimension to sample an indirect acquisition time of 100 ms. The proton dimension was sampled with 1024 complex data points sampling 160 ms in both instances. Experiment times were 25 min for 1 H−13C HSQC experiments with β-glucan as the host and 16 min for 1 H−13C HSQC experiments with bile salt as the host. A recycle delay of 1 s was employed in each instance. For reference samples of bile salts and β-glucan, additional 1H−13C HSQC experiments were recorded sampling 60 ppm to yield the chemical shift values of the aliased spectral signals. All NMR spectra were recorded on an 800 MHz Bruker (Fällanden, Switzerland) Avance spectrometer equipped with a TCI Z-gradient CryoProbe and an 18.7 T magnet (Oxford Magnet Technology, Oxford, U.K.). Data Processing, Assignments, and Analysis. All spectra were processed in Topspin 2.1 (Bruker) with extensive zero filling in both dimensions. Projections of 1H−13C HSQC experiments onto the 13C dimension were created as projected spectra between 1H chemical shifts of 3.4 and 3.8 ppm. Spectral assignments were obtained from the literature as full chemical shift assignments have been previously reported both for bile salts and for β-glucan.16,17 In the fitting of titrations, the NMR signal changes in β-glucan upon addition of bile salt were treated as a chemical shift difference between individual βglucan signals as shown in Figure 4. This treatment provides an internal reference and renders very small chemical shift changes detectable without the need for adding potentially interfering reference compounds. Due to the plethora of equivalent potential binding sites 11473
dx.doi.org/10.1021/jf504352w | J. Agric. Food Chem. 2014, 62, 11472−11478
Journal of Agricultural and Food Chemistry
Article
Figure 3. 1H−13C HSQC spectral region of 2% (w/v) low viscosity β-glucan (140 kDa) in the presence of increasing concentrations of glycocholate in 100 mM sodium phosphate buffer of pH 6.5 (top) and 100 mM sodium acetate buffer of pH 5.3 (bottom). The two resonances are of a C4H4 in a glycosidic bond and of a C2H2 in a glucopyranosyl unit that has a glycosidic bond at O3. Chemical shift changes are highlighted by horizontal lines. in β-glucan, adsorption of bile salts to β-glucan was treated as a Langmuir adsorption isotherm according to
Δδ = Δδ0 + ΔΔδmax
c Kd + c
(1)
where Δδ is the chemical shift difference between two select β-glucan signals, Δδ0 is the given chemical shift difference in the absence of bile salt, ΔΔδmax is the maximum change in this chemical shift difference upon saturation with bile salt, c is the concentration of the bile salt, and Kd is the dissociation constant. Fitting to the Langmuir adsorption isotherm was used to qualitatively probe the effect of weakly acidic pH on interactions between β-glucan and bile salt.
■
RESULTS In the current study, we attempt to probe interactions between β-glucan (Figure 1) and bile salts (using GC and TCDC as examples, see Figure 2) by 1H−13C NMR spectroscopic experiments on the time scale of minutes. The current study is thus aimed at establishing rapid binding assays for polysaccharide−host interactions and at improving the molecular understanding of interactions between β-glucans and bile salts. In vitro binding assays were conducted using low viscosity βglucan at a concentration of 2% (w/v) as the host and adding glycocholate of varying concentrations (0, 7, 14, and 28 mM). The detection of molecular interactions by probing carbohydrate signals has remained somewhat underexplored, presumably due to the congestion of carbohydrate signals in the 1H dimension and the poor sensitivity of 13C NMR spectroscopy at natural abundance. This sensitivity restriction is ameliorated by recent generations of NMR spectrometers employing cryogenically cooled detection electronics, by rapid pulsing techniques, or by 1H homodecoupling.18,19 Highly resolved 1H−13C spectra of carbohydrates can thus be obtained even at low millimolar concentrations on time scales of minutes or even seconds, especially when using narrow sweep widths in the 13C dimension.20−22 Resultant β-glucan 1H−13C HSQC spectra provided a signal-to-noise ratio above 20 for 2% (w/v) β-glucan samples within 25 min of experiment time. This signal-to-noise ratio proves sufficient for accurate measurements of β-glucan signal positions during the titrations with bile salts (Figure 3). Titration of low viscosity β-glucan with bile salts yields systematic signal changes in the β-glucan 1H−13C HSQC spectra. Two well-separated signals showing different systematic changes were monitored to probe the interaction of βglucan with bile salts (Figures 3, 4). Differences in chemical shift were obtained by peak picking of signals in the two-
Figure 4. Chemical shift change (Δδ) between the two β-glucan resonances displayed in Figure 3 upon addition of increasing glycocholate (GC) concentrations (0, 7, 14, 28 mM) at pH 6.5 (top) and pH 5.3 (bottom). Similar chemical shift trends at pH 5.3 and 6.5 indicate systematic and titration-responsive effects on the βglucan signals upon addition of glycocholate. The chemical shift changes are consistent with a lower Kd at pH 5.3.
dimensional experiments. In order to exclude systematic effects due to instabilities of the deuterium lock signal or temperature, the anticipated small chemical shift changes in the titration were measured as the chemical shift difference (Δδ) between the two β-glucan signals. Changes in Δδ up to 15 parts per billion (3 Hz at a 13C frequency of 201 MHz) were detected in experiments of 25 min duration. Titrations were conducted at 2 different pH values, 6.5 and 5.3 (Figure 4). Similar trends are observed in the titrations conducted at different pH values, and NMR determinations are consistent with previous in vitro studies with HPLC analysis indicating that the retention of bile salts by β-glucan-rich oat extrudates is higher near pH 5 than near pH 6.5.23 1 H−13C HSQC supports the notion of interactions between bile salts and β-glucan using reasonably rapid assays. The observation of the carbohydrate signal in monitoring guest 11474
dx.doi.org/10.1021/jf504352w | J. Agric. Food Chem. 2014, 62, 11472−11478
Journal of Agricultural and Food Chemistry
Article
Figure 5. 1H−13C HSQC spectral region of glycocholate (2 mM in A and 20 mM in B) in the presence of increasing concentrations of medium viscosity β-glucan (200 kDa). Arrows indicate the change in GC signal when increasing the concentration from 2 mM to 20 mM. The addition of βglucan induces small chemical shift changes of bile salt signals with the same trend as chemical shift changes upon the addition of bile salt itself. In the displayed spectral region, H11C11 and H16C16 resonances experience 13C signal changes toward higher chemical shift, upon addition of βglucan and of bile salt, while the H9C9 resonance remains unchanged and H17C17 shows a 13C resonance change toward lower chemical shift, upon addition of β-glucan and of bile salt. These changes are more pronounced at 20 mM GC, above the critical micelle concentration of GC (7−11 mM).28
bile salt signals themselves could be a consequence of increased local bile salt concentrations through confinement or excluded volume effects, thus underlining the usefulness of detecting signal changes in the binding partner to validate interactions. Even using high-resolution spectroscopy, the chemical shift changes are at the edge of what is detectable. Hence, experiments were repeated using a different bile salt, taurochenodeoxycholate (TCDC) in order to validate the above findings. As for the titration with glycocholate, a downfield shift of the 13C4 signal in a glycosidic bond is detected upon addition of 7, 14, and 28 mM TCDC (Figure 6A). A titration of TCDC with β-glucan was conducted at a TCDC concentration of 3 mM to validate the viability of rapid 1 H−13C HSQC assays at low millimolar bile salt concentrations. Chemical shift changes are again consistent with a stabilization of bile salt aggregation in the presence of β-glucan, as the addition of β-glucan consistently induces chemical shift changes in the bile salt that resemble chemical shift changes due to an increase in bile salt concentration (Figure 6B).
absorption onto functional polysaccharides or onto fragments thereof has practical advantages. The polysaccharide signals are intrinsically more robust to inaccuracies in sample pH or concentration, for instance due to the absence of easily ionizable groups in carbohydrates. Interactions between the bile salt and β-glucan were likewise probed by titrating medium viscosity β-glucan into samples containing a constant bile salt concentration, as previously described.9 Chemical shift changes were monitored using 1H−13C HSQC experiments of 16 min duration, to yield the titration of glycocholate with β-glucan shown in Figure 5. Notably, rapid 1H−13C HSQC experiments can be conducted at concentrations near or below the critical micelle concentration of bile salts (low millimolar). Consequently, binding of β-glucan to glycocholate was monitored at 2 mM and at 20 mM bile salt concentration. Due to the aggregation into micelles, the glycocholate signals are strongly concentration dependent. Addition of up to 2% (w/v) medium viscosity β-glucan induces only small chemical shift changes at 2 mM bile salt and more pronounced chemical shift changes at 20 mM bile salt concentration. The bile salt signals are generally found to move upon β-glucan addition into the same direction that results from a concentration increase of the bile salt itself (Figure 5, Figure S1 in the Supporting Information). The presence of β-glucan therefore appears to stabilize micelle formation or lateral self-aggregation in the bile salt. Changes in
■
DISCUSSION Polysaccharides play central roles as functional components of supramolecular assemblies in biology and technology. We find that 1H−13C HSQC NMR spectroscopy provides sufficient 11475
dx.doi.org/10.1021/jf504352w | J. Agric. Food Chem. 2014, 62, 11472−11478
Journal of Agricultural and Food Chemistry
Article
Figure 6. Titrations of β-glucan with TCDC (A; top pH 6.5, bottom pH 5.3) and of TCDC with β-glucan (B) support small chemical shift changes in the β-glucan and signal changes in the bile salt consistent with a stabilization of the micellar form.
Figure 7. Model of β-glucan interaction with bile salts. The higher affinity between multivalent host and guest favors binding of β-glucan to the bile salt micelle relative to the monomer, thus stabilizing the micellar form, overall.
11476
dx.doi.org/10.1021/jf504352w | J. Agric. Food Chem. 2014, 62, 11472−11478
Journal of Agricultural and Food Chemistry
■
sensitivity to enable rapid observations of molecular interactions involving polysaccharides and guest molecules at molecular detail. Such observations are feasible for probing interactions of β-glucan and bile salts, even though the interactions are dynamic and result in chemical shift changes in viscous polysaccharide solutions on the order of only 0.01 ppm. Interactions involving carbohydrates are commonly probed by NMR spectroscopy through the detection of signals from the non-carbohydrate binding partner,9,24−26 as the carbohydrate 1H spectral region usually is crowded and overlaps with the water signal. The observation of wellseparated carbohydrate signals in aliased 1H−13C HSQC spectra enables the use of both binding partners as host and guest molecules in titrations. The detection of chemical shift changes in both binding partners then supports direct interactions with the carbohydrate. Many attempts have been made to clarify how β-glucans might interact with bile salts in the small intestine, preventing their reabsorption and leading to excess excretion via the feces. Collectively, there are three main hypotheses: (a) increase, by β-glucans, of the intestinal viscosity whereby reabsorption of bile salts may be prevented by a potential β-glucan barrier layer; (b) entrapment of bile salt micelles in β-glucan molecular networks or aggregates; and (c) binding of bile salt monomers or micelles to β-glucans at a molecular level.5,10 Here, we obtain 1 H−13C HSQC titrations of bile salts and β-glucan that are consistent with direct molecular interactions between bile salts and all used β-glucans. The interaction results in a stabilization of bile salt micelles in the presence of β-glucan, presumably due to a stronger interaction of the β-glucan with the micelle than with the bile salt monomer (Figure 7), and influences β-glucan signals, primarily at the sites of glycosidic linkages (Figure S3 in the Supporting Information). An increased interaction between bile salt micelles and β-glucan is not surprising given the repetitive structure of the β-glucan and the resultant possibility for multivalent interactions between micelle and polymer. The minute chemical shift changes observed in the β-glucan argue against a cooperative structural transition of β-glucans upon addition of the bile salt host and against highly specific binding, opposite to the formation of hydrophobic inclusion complexes between amphiphilic hosts and α(1−4)-glucans.24 Notably, the use of 1H−13C HSQC NMR at high magnetic field provides sufficient selectivity and sensitivity to examine polysaccharide− guest interactions within a time frame of minutes, making the screening of various host and guest molecules practical. We finally note that cholic acid based receptors have been devised as biomimetic receptors for carbohydrates, including βglucosides.27 These receptors exhibit multivalent interactions between the carbohydrate and the hydrophilic face of the cholic acid building blocks, thus exploiting similar interactions as those indicated in the model of Figure 7. In conclusion, this study shows that 1H−13C NMR assays on the time scale of minutes detect signal changes in both bile salts and β-glucans, thus arguing for direct dynamic interactions, without inducing cooperative conformational changes in the βglucans. Titrations are consistent with stronger interactions at pH 5.3 than at pH 6.5 in this in vitro assay. The changes in bile salt and β-glucan signals suggest a stabilization of bile salt micelles through transient, multivalent interactions with βglucans.
Article
ASSOCIATED CONTENT
S Supporting Information *
An overview figure of bile acid signal changes upon concentration increase of the bile acid itself and upon addition of β-glucan is provided. These changes are mapped onto the bile acid structure. Interaction studies using laminaribiose as a model saccharide are demonstrated, indicating transient interactions with low specificity and weak effect on molecular tumbling, but exerting some small changes in glycosidic linkage conformation. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*Technical University of Denmark, Kemitorvet, Building 201, DK-2800 Kgs. Lyngby, Denmark. E-mail: [email protected]. Funding
The authors gratefully acknowledge The Danish National Advanced Technology Foundation for financial support to the LIQFUN-project and funding by the Carlsberg Foundation Grant 2013_01_0709. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS NMR spectra were recorded with the 800 MHz spectrometer of the Danish National Instrument Center for NMR Spectroscopy of Biological Macromolecules at the Technical University of Denmark.
■
REFERENCES
(1) Davidson, M. H.; Dugan, L. D.; Burns, J. H.; Bova, J.; Story, K.; Drennan, K. B. The hypocholesterolemic effects of beta-glucan in oatmeal and oat bran. A dose-controlled study. JAMA, J. Am. Med. Assoc. 1991, 265, 1833−1839. (2) Lia, A.; Hallmans, G.; Sandberg, A. S.; Sundberg, B.; Aman, P.; Andersson, H. Oat β-glucan increases bile acid excretion and a fiberrich barley fraction increases cholesterol excretion in ileostomy subjects. Am. J. Clin. Nutr. 1995, 62, 1245−1251. (3) Truswell, A. S. Cereal grains and coronary heart disease. Eur. J. Clin. Nutr. 2002, 56, 1−14. (4) Wood, P. J. Cereal β-glucans in diet and health. J. Cereal Sci. 2007, 46, 230−238. (5) Theuwissen, E.; Mensink, R. P. Water-soluble dietary fibers and cardiovascular disease. Physiol. Behav. 2008, 94, 285−292. (6) Bowles, R. K.; Morgan, K. R.; Furneaux, R. H.; Coles, G. D. 13C CP/MAS NMR study of the interaction of bile acids with barley β-dglucan. Carbohydr. Polym. 1996, 29, 7−10. (7) Simonsen, H. T.; Nielsen, M. S.; Christensen, N. J.; Christensen, U.; La Cour, T. V.; Motawia, M. S.; Jespersen, B. P.; Engelsen, S. B.; M?ller, B. L. Molecular interactions between barley and oat betaglucans and phenolic derivatives. J. Agric. Food Chem. 2009, 57, 2056− 2064. (8) Story, J. A.; Kritchevsky, D. Comparison of the binding of various bile acids and bile salts in vitro by several types of fiber. J. Nutr. 1976, 106, 1292−1294. (9) Gunness, P.; Flanagan, B. M.; Gidley, M. J. Molecular interactions between cereal soluble dietary fibre polymers and a model bile salt deduced from 13C NMR titration. J. Cereal Sci. 2010, 52, 444−449. (10) Gunness, P.; Gidley, M. J. Mechanisms underlying the cholesterol-lowering properties of soluble dietary fibre polysaccharides. Food Funct. 2010, 1, 149−155. (11) Kim, H. J.; White, P. J. Interactional Effects of β-Glucan, Starch, and Protein in Heated Oat Slurries on Viscosity and In Vitro Bile Acid Binding. J. Agric. Food Chem. 2012, 60, 6217−6222. 11477
dx.doi.org/10.1021/jf504352w | J. Agric. Food Chem. 2014, 62, 11472−11478
Journal of Agricultural and Food Chemistry
Article
(12) Northfield, T. C.; McColl, I. Postprandial concentrations of free and conjugated bile acids down the length of the normal human small intestine. Gut 1973, 14, 513−518. (13) Mikkelsen, M. S.; Jespersen, B. M.; Larsen, F. H.; Blennow, A.; Engelsen, S. B. Molecular structure of large-scale extracted beta-glucan from barley and oat: Identification of a significantly changed block structure in a high beta-glucan barley mutant. Food Chem. 2013, 136, 130−138. (14) Hyun, J. K.; White, P. J. In vitro bile-acid binding and fermentation of high, medium, and low molecular weight β-glucan. J. Agric. Food Chem. 2010, 58, 628−634. (15) Kay, L.; Keifer, P.; Saarinen, T. Pure absorption gradient enhanced heteronuclear single quantum correlation spectroscopy with improved sensitivity. J. Am. Chem. Soc. 1992, 114, 10663−10665. (16) Ijare, O.; Somashekar, B. S.; Jadegoud, Y.; Nagana Gowda, G. A. 1 H and 13C NMR characterization and stereochemical assignments of bile acids in aqueous media. Lipids 2005, 40, 1031−1041. (17) Petersen, B. O.; Olsen, O.; Beeren, S. R.; Hindsgaul, O.; Meier, S. Monitoring pathways of beta-glucan degradation by enzyme mixtures in situ. Carbohydr. Res. 2013, 368, 47−51. (18) Schulze-Sunninghausen, D.; Becker, J.; Luy, B. Rapid heteronuclear single quantum correlation NMR spectra at natural abundance. J. Am. Chem. Soc. 2014, 136, 1242−1245. (19) Paudel, L.; Adams, R. W.; Kiraly, P.; Aguilar, J. A.; Foroozandeh, M.; Cliff, M. J.; Nilsson, M.; Sandor, P.; Waltho, J. P.; Morris, G. A. Simultaneously enhancing spectral resolution and sensitivity in heteronuclear correlation NMR spectroscopy. Angew. Chem. 2013, 52, 11616−11619. (20) Jeannerat, D. Computer optimized spectral aliasing in the indirect dimension of 1H-13C heteronuclear 2D NMR experiments. A new algorithm and examples of applications to small molecules. J. Magn. Reson. 2007, 186, 112−122. (21) Bojstrup, M.; Petersen, B. O.; Beeren, S. R.; Hindsgaul, O.; Meier, S. Fast and accurate quantitation of glucans in complex mixtures by optimized heteronuclear NMR spectroscopy. Anal. Chem. 2013, 85, 8802−8808. (22) Petersen, B. O.; Hindsgaul, O.; Meier, S. Profiling of carbohydrate mixtures at unprecedented resolution using highprecision 1H-13C chemical shift measurements and a reference library. Analyst 2014, 139, 401−406. (23) Drzikova, B.; Dongowski, G.; Gebhardt, E.; Habel, A. The composition of dietary fibre-rich extrudates from oat affects bile acid binding and fermentation in vitro. Food Chem. 2005, 90, 181−192. (24) Beeren, S. R.; Meier, S.; Hindsgaul, O. Probing helical hydrophobic binding sites in branched starch polysaccharides using NMR spectroscopy. Chem.Eur. J. 2013, 19, 16314−16320. (25) Ferrand, Y.; Crump, M. P.; Davis, A. P. A synthetic lectin analog for biomimetic disaccharide recognition. Science 2007, 318, 619−622. (26) Klein, E.; Crump, M. P.; Davis, A. P. Carbohydrate recognition in water by a tricyclic polyamide receptor. Angew. Chem. 2004, 44, 298−302. (27) Davis, A. P.; Wareham, R. S. Carbohydrate Recognition through Noncovalent Interactions: A Challenge for Biomimetic and Supramolecular Chemistry. Angew. Chem. 1999, 38, 2978−2996. (28) Gouin, S.; Zhu, X. X. Fluorescence and NMR Studies of the Effect of a Bile Acid Dimer on the Micellation of Bile Salts. Langmuir 1998, 14, 4025−4029.
11478
dx.doi.org/10.1021/jf504352w | J. Agric. Food Chem. 2014, 62, 11472−11478
Probing Interactions between β‑Glucan and Bile Salts at Atomic Detail by 1H−13C NMR Assays Mette Skau Mikkelsen,† Sofia Bolvig Cornali,† Morten G. Jensen,‡ Mathias Nilsson,†,§ Sophie R. Beeren,∥ and Sebastian Meier*,∥,⊥ †
Faculty of Science, Department of Food Science, University of Copenhagen, Rolighedsvej 30, DK-1958 Frederiksberg C, Denmark Carlsberg Research Center, Gamle Carlsberg Vej 10, DK-1799 Copenhagen V, Denmark § School of Chemistry, University of Manchester, Oxford Road, Manchester, M13 9PL, UK ∥ Carlsberg Laboratory, Gamle Carlsberg Vej 10, DK-1799 Copenhagen V, Denmark ⊥ Department of Chemistry, Technical University of Denmark, Kemitorvet, Building 201, DK-2800 Kongens Lyngby, Denmark ‡
S Supporting Information *
ABSTRACT: Polysaccharides are prospective hosts for the delivery and sequestration of bioactive guest molecules. Polysaccharides of dietary fiber, specifically cereal (1 → 3)(1 → 4)-β-glucans, play a role in lowering the blood plasma cholesterol level in humans. Direct host−guest interactions between β-glucans and conjugated bile salts are among the possible molecular mechanisms explaining the hypocholesterolemic effects of β-glucans. The present study shows that 1H−13C NMR assays on a time scale of minutes detect minute signal changes in both bile salts and β-glucans, thus indicating dynamic interactions between bile salts and β-glucans. Experiments are consistent with stronger interactions at pH 5.3 than at pH 6.5 in this in vitro assay. The changes in bile salt and β-glucan signals suggest a stabilization of bile salt micelles and concomitant conformational changes in βglucans. KEYWORDS: bile salt, glucan, HSQC, heteronuclear NMR
■
INTRODUCTION The health benefits of dietary fibers have increasingly attracted consumer awareness and scientific interest. The health benefits of cereal foods are to a large extent accredited to hydrosoluble dietary fibers, especially mixed-linkage (1 → 3)(1 → 4)-βglucan (henceforth referred to as β-glucan).1−4 Cereal β-glucan reportedly lowers the blood plasma cholesterol levels in humans 1 and was the first cholesterol-reducing food component to be registered by the U.S. Food and Drug Administration in 1997. Due to its cholesterol-reducing capability, β-glucan can reduce the risk of cardiovascular diseases. The cholesterollowering effect of β-glucans has been attributed to the prevention of bile salt reabsorption from the small intenstine.2,5 The molecular details of the interactions between β-glucan and bile salts in the digestive tract remain unclear, not least due to challenges in detecting binding to the homopolymeric and viscous β-glucan with molecular detail.6−11 Rapid and sensitive assays that provide detailed molecular information for both interaction partners are desirable to probe interactions between polysaccharides, their fragments, and guest molecules of any sort. Previous 13C nuclear magnetic resonance (NMR) titrations had suggested that bile salt 13C NMR signals change upon the addition of β-glucan without significantly restricting the rotational tumbling and thus the line width of the bile salt signals.9 These data hint at dynamic interactions between bile salts and dietary fiber, but no chemical shift changes were previously observed for β-glucans themselves and measurements were only practical at concentrations well above the critical micelle concentration (CMC) of bile salts. Due to the © 2014 American Chemical Society
fast exchange of bile salts between monomeric and micellar forms above the CMC, changes to the bile salt 13C NMR signals could, however, also result from increased local concentrations of bile salt through confinement or excluded volume effects. In the current study we use rapid interaction assays between bile salts and β-glucan, probing the spectral signal of both binding partners and permitting rapid binding assays even below the CMC of bile salts. High-resolution 1H−13C NMR experiments of less than 30 min duration were used to provide sufficient sensitivity and resolution for detecting weak binding events. Interactions were probed in two separate experiments: (1) by titration of β-glucan to a bile salt solution of constant concentration and (2) by the titration of bile salt to a β-glucan host of constant concentration. The combined use of twodimensional 1H−13C NMR experiments and high-field NMR instrumentation leads to a ∼50-fold acceleration as compared to previously reported 13C NMR assays of bile salt binding to βglucan.6,9 In addition, the excellent sensitivity of 1H−13C NMR experiments recorded with state of the art instrumentation allows minute chemical shift changes to be detected in the βglucan host upon titration with bile salts. The detection of chemical shift changes for both components is consistent with direct dynamic interactions between β-glucan and bile salts at physiological pH values of the small intestine.12 Interactions are Received: Revised: Accepted: Published: 11472
July 18, 2014 October 28, 2014 November 6, 2014 November 6, 2014 dx.doi.org/10.1021/jf504352w | J. Agric. Food Chem. 2014, 62, 11472−11478
Journal of Agricultural and Food Chemistry
Article
Figure 1. Schematic representation of (1→3)(1→4)-β-glucan with horizontal lines indicating β(1→4) linkages and diagonal lines indicating β(1→3) linkages.
stronger at pH 5.3 than at pH 6.5. The data are consistent with a stabilization of bile salt micelles in the presence of β-glucan due to a stronger multivalent binding of the β-glucan to bile salt micelles than to the monomers. 1H−13C NMR assays thus provide a rapid means for probing interactions between βglucan and bile salts with atomic-level resolution.
■
EXPERIMENTAL SECTION
Materials. Low and medium viscosity barley β-glucans (Figure 1) were obtained from Megazyme International Ltd. (Bray, Ireland), and these samples have previously been characterized by various physicochemical methods.13 Briefly, low viscosity β-glucan had a molecular mass of 140 kDa as determined by high performance sizeexclusion chromatography and a calibration curve using five β-glucan molar mass standard samples. Upon extensive hydrolysis with lichenase, a block structure (β-(1→4) linked glucopyranoses between β-(1→3) branches) of approximately 77% DP3, 19% DP4, and 3% DP5 (molar ratio) was derived. Medium viscosity β-glucan had a molecular mass of 200 kDa and a block structure of approximately 76% DP3, 19% DP4, and 4% DP5. Samples contained less than 1% each of protein or starch. Low viscosity barley β-glucan was chosen for the experiments monitoring β-glucan signals in order to improve spectral quality at lower sample viscosity (at high viscosity molecular tumbling rates decrease and NMR signals might become broad due to rapid relaxation). Signal appearance is less of an issue when monitoring the bile salt sample due to the higher achievable concentration, and medium viscosity barley β-glucan was used in these experiments due to the reportedly higher retention of bile salts by β-glucan of higher viscosity.14 Bile salts, glycolate (GC) and taurochenodeoxycholate (TCDC) (Figure 2), along with buffer materials, sodium phosphate and sodium acetate, were obtained from Sigma-Aldrich (St. Louis, MO, USA). Buffers prepared at 100 mM concentration were sodium phosphate buffer of pH 6.5 and sodium acetate buffer of pH 5.3. Buffers were lyophilized and redissolved in D2O (Cambridge Isotope Laboratories, Tewksbury, MA, USA). Sample Preparation. Low viscosity and medium viscosity barley β-glucan were dissolved in buffer to concentrations of 0.5, 1, 1.5, or 2% (w/v) and hydrated at 80 °C for 30 min. Stock solutions of bile salt were prepared in buffer, and bile salt was added to β-glucan samples at the concentrations indicated (3, 7, 10, 14, or 28 mM). Samples containing β-glucan and bile salts were incubated at 37 °C for 120 min under continuous shaking, and pure bile salt and β-glucan reference samples, in the absence of the potential interaction partner, were treated identically. NMR Spectroscopy. Upon incubation at 37 °C for 120 min, the samples were directly transferred to 5 mm NMR sample tubes and analyzed by sensitivity-enhanced 1H−13C heteronuclear singlequantum correlation15 (HSQC) experiments with a spectral width in the indirect (13C) dimension that was reduced to permit rapid sampling of the 13C dimension at high resolution. Experiments treating β-glucan as the host were recorded with a spectral width of 5 ppm and an offset in the 13C dimension of 80 ppm. A total of 75 complex points
Figure 2. Structure of the bile salt molecules glycocholate (GC) and taurochendeoxycholate (TCDC) including their atom numbering. were recorded in the 13C dimension to sample an indirect acquisition time of 75 ms. Experiments treating bile salt as the host were recorded with a spectral width of 20 ppm centered around a spectral offset of 32 ppm. A total of 400 complex points were recorded in the 13C dimension to sample an indirect acquisition time of 100 ms. The proton dimension was sampled with 1024 complex data points sampling 160 ms in both instances. Experiment times were 25 min for 1 H−13C HSQC experiments with β-glucan as the host and 16 min for 1 H−13C HSQC experiments with bile salt as the host. A recycle delay of 1 s was employed in each instance. For reference samples of bile salts and β-glucan, additional 1H−13C HSQC experiments were recorded sampling 60 ppm to yield the chemical shift values of the aliased spectral signals. All NMR spectra were recorded on an 800 MHz Bruker (Fällanden, Switzerland) Avance spectrometer equipped with a TCI Z-gradient CryoProbe and an 18.7 T magnet (Oxford Magnet Technology, Oxford, U.K.). Data Processing, Assignments, and Analysis. All spectra were processed in Topspin 2.1 (Bruker) with extensive zero filling in both dimensions. Projections of 1H−13C HSQC experiments onto the 13C dimension were created as projected spectra between 1H chemical shifts of 3.4 and 3.8 ppm. Spectral assignments were obtained from the literature as full chemical shift assignments have been previously reported both for bile salts and for β-glucan.16,17 In the fitting of titrations, the NMR signal changes in β-glucan upon addition of bile salt were treated as a chemical shift difference between individual βglucan signals as shown in Figure 4. This treatment provides an internal reference and renders very small chemical shift changes detectable without the need for adding potentially interfering reference compounds. Due to the plethora of equivalent potential binding sites 11473
dx.doi.org/10.1021/jf504352w | J. Agric. Food Chem. 2014, 62, 11472−11478
Journal of Agricultural and Food Chemistry
Article
Figure 3. 1H−13C HSQC spectral region of 2% (w/v) low viscosity β-glucan (140 kDa) in the presence of increasing concentrations of glycocholate in 100 mM sodium phosphate buffer of pH 6.5 (top) and 100 mM sodium acetate buffer of pH 5.3 (bottom). The two resonances are of a C4H4 in a glycosidic bond and of a C2H2 in a glucopyranosyl unit that has a glycosidic bond at O3. Chemical shift changes are highlighted by horizontal lines. in β-glucan, adsorption of bile salts to β-glucan was treated as a Langmuir adsorption isotherm according to
Δδ = Δδ0 + ΔΔδmax
c Kd + c
(1)
where Δδ is the chemical shift difference between two select β-glucan signals, Δδ0 is the given chemical shift difference in the absence of bile salt, ΔΔδmax is the maximum change in this chemical shift difference upon saturation with bile salt, c is the concentration of the bile salt, and Kd is the dissociation constant. Fitting to the Langmuir adsorption isotherm was used to qualitatively probe the effect of weakly acidic pH on interactions between β-glucan and bile salt.
■
RESULTS In the current study, we attempt to probe interactions between β-glucan (Figure 1) and bile salts (using GC and TCDC as examples, see Figure 2) by 1H−13C NMR spectroscopic experiments on the time scale of minutes. The current study is thus aimed at establishing rapid binding assays for polysaccharide−host interactions and at improving the molecular understanding of interactions between β-glucans and bile salts. In vitro binding assays were conducted using low viscosity βglucan at a concentration of 2% (w/v) as the host and adding glycocholate of varying concentrations (0, 7, 14, and 28 mM). The detection of molecular interactions by probing carbohydrate signals has remained somewhat underexplored, presumably due to the congestion of carbohydrate signals in the 1H dimension and the poor sensitivity of 13C NMR spectroscopy at natural abundance. This sensitivity restriction is ameliorated by recent generations of NMR spectrometers employing cryogenically cooled detection electronics, by rapid pulsing techniques, or by 1H homodecoupling.18,19 Highly resolved 1H−13C spectra of carbohydrates can thus be obtained even at low millimolar concentrations on time scales of minutes or even seconds, especially when using narrow sweep widths in the 13C dimension.20−22 Resultant β-glucan 1H−13C HSQC spectra provided a signal-to-noise ratio above 20 for 2% (w/v) β-glucan samples within 25 min of experiment time. This signal-to-noise ratio proves sufficient for accurate measurements of β-glucan signal positions during the titrations with bile salts (Figure 3). Titration of low viscosity β-glucan with bile salts yields systematic signal changes in the β-glucan 1H−13C HSQC spectra. Two well-separated signals showing different systematic changes were monitored to probe the interaction of βglucan with bile salts (Figures 3, 4). Differences in chemical shift were obtained by peak picking of signals in the two-
Figure 4. Chemical shift change (Δδ) between the two β-glucan resonances displayed in Figure 3 upon addition of increasing glycocholate (GC) concentrations (0, 7, 14, 28 mM) at pH 6.5 (top) and pH 5.3 (bottom). Similar chemical shift trends at pH 5.3 and 6.5 indicate systematic and titration-responsive effects on the βglucan signals upon addition of glycocholate. The chemical shift changes are consistent with a lower Kd at pH 5.3.
dimensional experiments. In order to exclude systematic effects due to instabilities of the deuterium lock signal or temperature, the anticipated small chemical shift changes in the titration were measured as the chemical shift difference (Δδ) between the two β-glucan signals. Changes in Δδ up to 15 parts per billion (3 Hz at a 13C frequency of 201 MHz) were detected in experiments of 25 min duration. Titrations were conducted at 2 different pH values, 6.5 and 5.3 (Figure 4). Similar trends are observed in the titrations conducted at different pH values, and NMR determinations are consistent with previous in vitro studies with HPLC analysis indicating that the retention of bile salts by β-glucan-rich oat extrudates is higher near pH 5 than near pH 6.5.23 1 H−13C HSQC supports the notion of interactions between bile salts and β-glucan using reasonably rapid assays. The observation of the carbohydrate signal in monitoring guest 11474
dx.doi.org/10.1021/jf504352w | J. Agric. Food Chem. 2014, 62, 11472−11478
Journal of Agricultural and Food Chemistry
Article
Figure 5. 1H−13C HSQC spectral region of glycocholate (2 mM in A and 20 mM in B) in the presence of increasing concentrations of medium viscosity β-glucan (200 kDa). Arrows indicate the change in GC signal when increasing the concentration from 2 mM to 20 mM. The addition of βglucan induces small chemical shift changes of bile salt signals with the same trend as chemical shift changes upon the addition of bile salt itself. In the displayed spectral region, H11C11 and H16C16 resonances experience 13C signal changes toward higher chemical shift, upon addition of βglucan and of bile salt, while the H9C9 resonance remains unchanged and H17C17 shows a 13C resonance change toward lower chemical shift, upon addition of β-glucan and of bile salt. These changes are more pronounced at 20 mM GC, above the critical micelle concentration of GC (7−11 mM).28
bile salt signals themselves could be a consequence of increased local bile salt concentrations through confinement or excluded volume effects, thus underlining the usefulness of detecting signal changes in the binding partner to validate interactions. Even using high-resolution spectroscopy, the chemical shift changes are at the edge of what is detectable. Hence, experiments were repeated using a different bile salt, taurochenodeoxycholate (TCDC) in order to validate the above findings. As for the titration with glycocholate, a downfield shift of the 13C4 signal in a glycosidic bond is detected upon addition of 7, 14, and 28 mM TCDC (Figure 6A). A titration of TCDC with β-glucan was conducted at a TCDC concentration of 3 mM to validate the viability of rapid 1 H−13C HSQC assays at low millimolar bile salt concentrations. Chemical shift changes are again consistent with a stabilization of bile salt aggregation in the presence of β-glucan, as the addition of β-glucan consistently induces chemical shift changes in the bile salt that resemble chemical shift changes due to an increase in bile salt concentration (Figure 6B).
absorption onto functional polysaccharides or onto fragments thereof has practical advantages. The polysaccharide signals are intrinsically more robust to inaccuracies in sample pH or concentration, for instance due to the absence of easily ionizable groups in carbohydrates. Interactions between the bile salt and β-glucan were likewise probed by titrating medium viscosity β-glucan into samples containing a constant bile salt concentration, as previously described.9 Chemical shift changes were monitored using 1H−13C HSQC experiments of 16 min duration, to yield the titration of glycocholate with β-glucan shown in Figure 5. Notably, rapid 1H−13C HSQC experiments can be conducted at concentrations near or below the critical micelle concentration of bile salts (low millimolar). Consequently, binding of β-glucan to glycocholate was monitored at 2 mM and at 20 mM bile salt concentration. Due to the aggregation into micelles, the glycocholate signals are strongly concentration dependent. Addition of up to 2% (w/v) medium viscosity β-glucan induces only small chemical shift changes at 2 mM bile salt and more pronounced chemical shift changes at 20 mM bile salt concentration. The bile salt signals are generally found to move upon β-glucan addition into the same direction that results from a concentration increase of the bile salt itself (Figure 5, Figure S1 in the Supporting Information). The presence of β-glucan therefore appears to stabilize micelle formation or lateral self-aggregation in the bile salt. Changes in
■
DISCUSSION Polysaccharides play central roles as functional components of supramolecular assemblies in biology and technology. We find that 1H−13C HSQC NMR spectroscopy provides sufficient 11475
dx.doi.org/10.1021/jf504352w | J. Agric. Food Chem. 2014, 62, 11472−11478
Journal of Agricultural and Food Chemistry
Article
Figure 6. Titrations of β-glucan with TCDC (A; top pH 6.5, bottom pH 5.3) and of TCDC with β-glucan (B) support small chemical shift changes in the β-glucan and signal changes in the bile salt consistent with a stabilization of the micellar form.
Figure 7. Model of β-glucan interaction with bile salts. The higher affinity between multivalent host and guest favors binding of β-glucan to the bile salt micelle relative to the monomer, thus stabilizing the micellar form, overall.
11476
dx.doi.org/10.1021/jf504352w | J. Agric. Food Chem. 2014, 62, 11472−11478
Journal of Agricultural and Food Chemistry
■
sensitivity to enable rapid observations of molecular interactions involving polysaccharides and guest molecules at molecular detail. Such observations are feasible for probing interactions of β-glucan and bile salts, even though the interactions are dynamic and result in chemical shift changes in viscous polysaccharide solutions on the order of only 0.01 ppm. Interactions involving carbohydrates are commonly probed by NMR spectroscopy through the detection of signals from the non-carbohydrate binding partner,9,24−26 as the carbohydrate 1H spectral region usually is crowded and overlaps with the water signal. The observation of wellseparated carbohydrate signals in aliased 1H−13C HSQC spectra enables the use of both binding partners as host and guest molecules in titrations. The detection of chemical shift changes in both binding partners then supports direct interactions with the carbohydrate. Many attempts have been made to clarify how β-glucans might interact with bile salts in the small intestine, preventing their reabsorption and leading to excess excretion via the feces. Collectively, there are three main hypotheses: (a) increase, by β-glucans, of the intestinal viscosity whereby reabsorption of bile salts may be prevented by a potential β-glucan barrier layer; (b) entrapment of bile salt micelles in β-glucan molecular networks or aggregates; and (c) binding of bile salt monomers or micelles to β-glucans at a molecular level.5,10 Here, we obtain 1 H−13C HSQC titrations of bile salts and β-glucan that are consistent with direct molecular interactions between bile salts and all used β-glucans. The interaction results in a stabilization of bile salt micelles in the presence of β-glucan, presumably due to a stronger interaction of the β-glucan with the micelle than with the bile salt monomer (Figure 7), and influences β-glucan signals, primarily at the sites of glycosidic linkages (Figure S3 in the Supporting Information). An increased interaction between bile salt micelles and β-glucan is not surprising given the repetitive structure of the β-glucan and the resultant possibility for multivalent interactions between micelle and polymer. The minute chemical shift changes observed in the β-glucan argue against a cooperative structural transition of β-glucans upon addition of the bile salt host and against highly specific binding, opposite to the formation of hydrophobic inclusion complexes between amphiphilic hosts and α(1−4)-glucans.24 Notably, the use of 1H−13C HSQC NMR at high magnetic field provides sufficient selectivity and sensitivity to examine polysaccharide− guest interactions within a time frame of minutes, making the screening of various host and guest molecules practical. We finally note that cholic acid based receptors have been devised as biomimetic receptors for carbohydrates, including βglucosides.27 These receptors exhibit multivalent interactions between the carbohydrate and the hydrophilic face of the cholic acid building blocks, thus exploiting similar interactions as those indicated in the model of Figure 7. In conclusion, this study shows that 1H−13C NMR assays on the time scale of minutes detect signal changes in both bile salts and β-glucans, thus arguing for direct dynamic interactions, without inducing cooperative conformational changes in the βglucans. Titrations are consistent with stronger interactions at pH 5.3 than at pH 6.5 in this in vitro assay. The changes in bile salt and β-glucan signals suggest a stabilization of bile salt micelles through transient, multivalent interactions with βglucans.
Article
ASSOCIATED CONTENT
S Supporting Information *
An overview figure of bile acid signal changes upon concentration increase of the bile acid itself and upon addition of β-glucan is provided. These changes are mapped onto the bile acid structure. Interaction studies using laminaribiose as a model saccharide are demonstrated, indicating transient interactions with low specificity and weak effect on molecular tumbling, but exerting some small changes in glycosidic linkage conformation. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*Technical University of Denmark, Kemitorvet, Building 201, DK-2800 Kgs. Lyngby, Denmark. E-mail: [email protected]. Funding
The authors gratefully acknowledge The Danish National Advanced Technology Foundation for financial support to the LIQFUN-project and funding by the Carlsberg Foundation Grant 2013_01_0709. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS NMR spectra were recorded with the 800 MHz spectrometer of the Danish National Instrument Center for NMR Spectroscopy of Biological Macromolecules at the Technical University of Denmark.
■
REFERENCES
(1) Davidson, M. H.; Dugan, L. D.; Burns, J. H.; Bova, J.; Story, K.; Drennan, K. B. The hypocholesterolemic effects of beta-glucan in oatmeal and oat bran. A dose-controlled study. JAMA, J. Am. Med. Assoc. 1991, 265, 1833−1839. (2) Lia, A.; Hallmans, G.; Sandberg, A. S.; Sundberg, B.; Aman, P.; Andersson, H. Oat β-glucan increases bile acid excretion and a fiberrich barley fraction increases cholesterol excretion in ileostomy subjects. Am. J. Clin. Nutr. 1995, 62, 1245−1251. (3) Truswell, A. S. Cereal grains and coronary heart disease. Eur. J. Clin. Nutr. 2002, 56, 1−14. (4) Wood, P. J. Cereal β-glucans in diet and health. J. Cereal Sci. 2007, 46, 230−238. (5) Theuwissen, E.; Mensink, R. P. Water-soluble dietary fibers and cardiovascular disease. Physiol. Behav. 2008, 94, 285−292. (6) Bowles, R. K.; Morgan, K. R.; Furneaux, R. H.; Coles, G. D. 13C CP/MAS NMR study of the interaction of bile acids with barley β-dglucan. Carbohydr. Polym. 1996, 29, 7−10. (7) Simonsen, H. T.; Nielsen, M. S.; Christensen, N. J.; Christensen, U.; La Cour, T. V.; Motawia, M. S.; Jespersen, B. P.; Engelsen, S. B.; M?ller, B. L. Molecular interactions between barley and oat betaglucans and phenolic derivatives. J. Agric. Food Chem. 2009, 57, 2056− 2064. (8) Story, J. A.; Kritchevsky, D. Comparison of the binding of various bile acids and bile salts in vitro by several types of fiber. J. Nutr. 1976, 106, 1292−1294. (9) Gunness, P.; Flanagan, B. M.; Gidley, M. J. Molecular interactions between cereal soluble dietary fibre polymers and a model bile salt deduced from 13C NMR titration. J. Cereal Sci. 2010, 52, 444−449. (10) Gunness, P.; Gidley, M. J. Mechanisms underlying the cholesterol-lowering properties of soluble dietary fibre polysaccharides. Food Funct. 2010, 1, 149−155. (11) Kim, H. J.; White, P. J. Interactional Effects of β-Glucan, Starch, and Protein in Heated Oat Slurries on Viscosity and In Vitro Bile Acid Binding. J. Agric. Food Chem. 2012, 60, 6217−6222. 11477
dx.doi.org/10.1021/jf504352w | J. Agric. Food Chem. 2014, 62, 11472−11478
Journal of Agricultural and Food Chemistry
Article
(12) Northfield, T. C.; McColl, I. Postprandial concentrations of free and conjugated bile acids down the length of the normal human small intestine. Gut 1973, 14, 513−518. (13) Mikkelsen, M. S.; Jespersen, B. M.; Larsen, F. H.; Blennow, A.; Engelsen, S. B. Molecular structure of large-scale extracted beta-glucan from barley and oat: Identification of a significantly changed block structure in a high beta-glucan barley mutant. Food Chem. 2013, 136, 130−138. (14) Hyun, J. K.; White, P. J. In vitro bile-acid binding and fermentation of high, medium, and low molecular weight β-glucan. J. Agric. Food Chem. 2010, 58, 628−634. (15) Kay, L.; Keifer, P.; Saarinen, T. Pure absorption gradient enhanced heteronuclear single quantum correlation spectroscopy with improved sensitivity. J. Am. Chem. Soc. 1992, 114, 10663−10665. (16) Ijare, O.; Somashekar, B. S.; Jadegoud, Y.; Nagana Gowda, G. A. 1 H and 13C NMR characterization and stereochemical assignments of bile acids in aqueous media. Lipids 2005, 40, 1031−1041. (17) Petersen, B. O.; Olsen, O.; Beeren, S. R.; Hindsgaul, O.; Meier, S. Monitoring pathways of beta-glucan degradation by enzyme mixtures in situ. Carbohydr. Res. 2013, 368, 47−51. (18) Schulze-Sunninghausen, D.; Becker, J.; Luy, B. Rapid heteronuclear single quantum correlation NMR spectra at natural abundance. J. Am. Chem. Soc. 2014, 136, 1242−1245. (19) Paudel, L.; Adams, R. W.; Kiraly, P.; Aguilar, J. A.; Foroozandeh, M.; Cliff, M. J.; Nilsson, M.; Sandor, P.; Waltho, J. P.; Morris, G. A. Simultaneously enhancing spectral resolution and sensitivity in heteronuclear correlation NMR spectroscopy. Angew. Chem. 2013, 52, 11616−11619. (20) Jeannerat, D. Computer optimized spectral aliasing in the indirect dimension of 1H-13C heteronuclear 2D NMR experiments. A new algorithm and examples of applications to small molecules. J. Magn. Reson. 2007, 186, 112−122. (21) Bojstrup, M.; Petersen, B. O.; Beeren, S. R.; Hindsgaul, O.; Meier, S. Fast and accurate quantitation of glucans in complex mixtures by optimized heteronuclear NMR spectroscopy. Anal. Chem. 2013, 85, 8802−8808. (22) Petersen, B. O.; Hindsgaul, O.; Meier, S. Profiling of carbohydrate mixtures at unprecedented resolution using highprecision 1H-13C chemical shift measurements and a reference library. Analyst 2014, 139, 401−406. (23) Drzikova, B.; Dongowski, G.; Gebhardt, E.; Habel, A. The composition of dietary fibre-rich extrudates from oat affects bile acid binding and fermentation in vitro. Food Chem. 2005, 90, 181−192. (24) Beeren, S. R.; Meier, S.; Hindsgaul, O. Probing helical hydrophobic binding sites in branched starch polysaccharides using NMR spectroscopy. Chem.Eur. J. 2013, 19, 16314−16320. (25) Ferrand, Y.; Crump, M. P.; Davis, A. P. A synthetic lectin analog for biomimetic disaccharide recognition. Science 2007, 318, 619−622. (26) Klein, E.; Crump, M. P.; Davis, A. P. Carbohydrate recognition in water by a tricyclic polyamide receptor. Angew. Chem. 2004, 44, 298−302. (27) Davis, A. P.; Wareham, R. S. Carbohydrate Recognition through Noncovalent Interactions: A Challenge for Biomimetic and Supramolecular Chemistry. Angew. Chem. 1999, 38, 2978−2996. (28) Gouin, S.; Zhu, X. X. Fluorescence and NMR Studies of the Effect of a Bile Acid Dimer on the Micellation of Bile Salts. Langmuir 1998, 14, 4025−4029.
11478
dx.doi.org/10.1021/jf504352w | J. Agric. Food Chem. 2014, 62, 11472−11478
