Fast and Accurate Quantitation of Glucans in Complex Mixtures by

(6) Two-dimensional 1H–13C NMR-based methods have rapidly gained popularity as fast quantitative tools for metabolic profiling.(7) These methods all...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/ac

Fast and Accurate Quantitation of Glucans in Complex Mixtures by Optimized Heteronuclear NMR Spectroscopy Marie Bøjstrup, Bent O. Petersen, Sophie R. Beeren, Ole Hindsgaul, and Sebastian Meier* Carlsberg Laboratory, Gamle Carlsberg Vej 10, 1799 Copenhagen V, Denmark S Supporting Information *

ABSTRACT: Nuclear magnetic resonance (NMR) spectroscopy is a widely used technique for mixture analysis, but it has shortcomings in resolving carbohydrate mixtures due to the narrow chemical shift range of glycans in general and fragments of homopolymers in particular. Here, we suggest a protocol toward fast spectroscopic glycan mixture analysis. We show that a plethora of oligosaccharides comprising only αglucopyranosyl residues can be resolved into distinct quantifiable signals with NMR experiments that are substantially faster than chromatographic runs. Conceptually, the approach fully exploits the narrow line widths of glycans (ν1/2 < 3 Hz) in the 13C spectral dimension while disregarding superfluous spectral information in compound identification and quantitation. The acetal (H1C1) groups suffice to spectroscopically resolve ∼20 different starch fragments in optimized 1H−13C NMR with a narrow 13C spectral width (3 ppm) that allows sampling the indirect 13C dimension at high resolution within 15 min. Rapid quantitations by high-resolution NMR data are achieved for glycans at concentrations as low as 10 μg/mL. For validation, comparisons were made with quantitations obtained by more time-consuming chromatographic methods and yielded coefficients of determination (R2) above 0.99.

G

weights but different chemical structures, such as isomeric linear and branched starch fragments, remains a challenge. Chromatographic methods for glycan separation and analysis in complex mixtures such as foodstuffs, feedstock, and biofluids typically require on the order of 1 h per sample.3 High-resolution NMR spectroscopy provides rich chemical information on analytes of interest in solution.6 Two-dimensional 1H−13C NMR-based methods have rapidly gained popularity as fast quantitative tools for metabolic profiling.7 These methods allow the sampling of the 13C spectral dimension with its wide chemical shift range at acceptable sensitivity due to magnetization transfer from strongly coupled sensitive nuclei (1H) to 13C, alleviating sensitivity penalties resulting from the low natural abundance and the lower magnetogyric ratio of 13C as compared to 1H. However, NMR spectroscopy has been rarely applied to the quantitative analysis of carbohydrate mixtures, owing to the small chemical shift dispersion and high degree of overlap of 1H and 13C signals in NMR spectra of carbohydrates in general and of glucan homopolymers in particular. In the current study, we show that by focusing on the acetal spectral region as a reporter of glycan structure, optimized 1 H−13C HSQC NMR can be used as a rapid quantitative profiling tool even for homopolymeric starch fragments having

lycans are chemically and functionally diverse biomolecules that play central roles as reactants and storage compounds in cellular and organismal carbon metabolism as well as in the global carbon cycle.1 Especially starch and starchderived glucans are of key interest, as they constitute the major source of food energy intake in modern societies and currently represent the most utilized carbon source in biotechnological ethanol production.2 Reliable and efficient qualitative and quantitative methods for carbohydrate analysis of plant extracts, food, and raw materials are therefore needed. These methods should combine high throughput with high resolution of different glycans to facilitate sample characterization in terms of quality, origin, health-related properties, taste, and production conditions.3 High resolution is critical as carbohydrates often occur in complex mixtures that are produced in batch processes resulting in mixtures of polysaccharides, oligosaccharides, and simple sugars.4 Principal methods for the qualitative and quantitative analysis of glycans include high-performance anion exchange chromatography (HPAEC), hydrophilic interaction liquid chromatography (HILIC), capillary electrophoresis, and mass spectrometry, as well as hyphenated combinations of these.1,3,5 These methods usually provide limited structural information and depend on the availability of standard compounds for the calibration of instrument-dependent retention times and responses. Furthermore, chemical derivatization is often required prior to analysis to covalently introduce detectable chromophores, fluorophores, or ionizable groups. Chromatographic resolution especially of sugars with similar molecular © XXXX American Chemical Society

Received: July 1, 2013 Accepted: August 18, 2013

A

dx.doi.org/10.1021/ac401980m | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

MA, USA) by eluting with H2O and an increasing amount of MeCN. The desired fractions were lyophilized to give a white solid (8.43 mg) containing a mixture of the two oligosaccharides. NMR Spectroscopy. For NMR spectroscopy, 600 μL of beer samples was lyophilized and redissolved in D2O (99.9%; Cambridge Isotope Laboratories, Andover, MA, USA). Maltooligosaccharides were weighed and placed into Eppendorf tubes and dissolved in 1000 μL of D2O. Three hundred microliters of the resultant solutions was mixed with 600 μL of D2O, prior to a second 3-fold dilution step in D2O. Calibrations were done in duplicate. Due to the chemical resolution of different species by optimized high-resolution HSQC, mixtures of glucose, maltotriose, and maltopentaose as well as a mixture of maltose and maltotetraose were used in order to minimize the number of samples requiring analysis (see Figure 3). All NMR spectra were recorded at 298 K on an 800 MHz Bruker (Fällanden, Switzerland) DRX spectrometer equipped with a TCI CryoProbe and an 18.7 T magnet (Oxford Magnet Technology, Oxford, UK) using conventional linear data sampling. 1H−13C HSQC spectra were recorded with a spectral sweep width of 3 ppm (602.786 Hz) around a 13C offset of 99.0 ppm. Due to the relatively narrow 13C1 chemical shift range in starch-derived fragments, the narrow spectral width does not result in spectral aliasing except for the reducing end 13C1α and 13 β C1 signals. The 13C T2 relaxation time of acetal carbon 13C spins in oligosaccharides is expected to be above 100 ms at ambient temperature,12 and a data matrix of 1024 × 200 complex data points was acquired in order to sample 332 ms in the 13C dimension. The 1D proton spectrum in Figure 1 was acquired by sampling 16 384 complex data points during an acquisition time of 1.57 s, while the 2D DQF-COSY spectrum was recorded with water suppression by presaturation as a data matrix of 4096 × 512 complex data points sampling 0.85 and 0.11 s in the direct and indirect dimensions, respectively. All spectra employed hard excitation and refocusing pulses only. Spectral Processing and Quantitation. Spectra were processed with extensive zero filling in both dimensions and a very mild resolution enhancement employing a Lorentzian− Gaussian transformation in the 13C dimension, as well as a time domain shifted sine bell window function in the direct dimension in Topspin 2.1 (Bruker, Fällanden, CH) and in nmrPipe.13 1H1−13C1 HSQC cross signals were automatically picked and integrated in nmrPipe. The resultant signal volumes were used for quantitation, as shown in Figures 4 and 5. All spectra were referenced relative to internal acetone with δ1H = 2.225 ppm and δ13C = 31.45 ppm. Chromatography. Reference methods for validating the NMR quantitations of maltooligosaccharides included fluorescence labeling with 2-aminobenzamide prior to UPLC analysis (Waters) and high-performance anion exchange chromatography with pulsed amperometric detection (Dionex, Sunnyvale, CA, USA).14 For UPLC analysis, glucose, maltose, maltotriose, maltotetraose, maltopentaose, and maltohexaose were dissolved in water to concentrations of 1 mg/mL. Between 25 and 200 μL of these solutions was lyophilized. Likewise, 100 μL of beer was lyophilized. To these lyophilized samples was added 100 μL of a 1 M solution of 2aminobenzamide in DMSO/AcOH (7:3), prior to adding 100 μL of a 1 M solution of NaBH3CN in DMSO/AcOH (7:3). Tightly capped samples were subsequently whirlmixed and incubated at 40 °C overnight. The sample was cooled to room temperature and diluted 1:400 in 10 mM ammonium formate

very poor chemical shift dispersion. Using sensitivity-enhanced 1 H−13C HSQC spectra7a,8 with optimized spectral width in the 13 C dimension, we distinguish at least eight linear and more than a dozen branched starch fragments in beer, when using the acetal 1H1−13C1 signal as a reporter in high-field (18.7 T) NMR experiments. The limit of quantitation of starch fragments in beer without fractionation or derivatization depends on the experiment time, but detection limits on the order of 10 μg/mL are achieved using experiments of 15 min duration. Comparison to quantitation using more time-consuming chromatographic methods gives a strong linear association with coefficients of determination (R2) above 0.99. Thus, we show that sensitivity-enhanced 1H1−13C1 HSQC spectra with optimized spectral widths yield accurate quantitations of carbohydrate mixtures in high-throughput and high-resolution assays even for analytes with strong 1H chemical shift overlap and narrow 13C chemical shift range.

■

EXPERIMENTAL SECTION Beer Samples. Commercial beer samples were purchased from local stores. The beer styles were bottom-fermented european lager and pilsner beers as well as european topfermented ale, which are expected to exhibit different glucan profiles because ale and lager beers are produced at different temperatures by specific yeast strains with different rates of maltooligosaccharide uptake.9 Chemicals. Glucose and maltooligosaccharides (maltose, maltotriose, maltotetraose, maltopentaose, maltohexaose, maltoheptaose, and maltooctaose) as well as panose and isomaltose were obtained from Sigma-Aldrich (St. Louis, MO, USA), Dextra Laboratories (Reading, UK), or Carbosynth (Compton, UK). Likewise, α,α-trehalose, α,β-trehalose, cellobiose, cellotriose, cellotetraose, erlose, galactose, galactosamine, gentianose, gentiobiose, glucosamine, N-acetyl glucosamine, isomaltotriose, kestose, kojibiose, lactulose, laminaribiose, mannose, nigerose, raffinose, rhamnose, sophorose, stachyose, sucrose, turanose, xylobiose, and xylose were obtained from Sigma-Aldrich, Dextra Laboratories, or Carbosynth. We purchased 63-α-D-glucosyl-maltotriose from Megazyme (Bray, UK). Other NMR reference spectra were obtained from the literature, such as those of nigerotriose, 6-kestose, and neokestose.10 Both 63-α-D-maltosyl-maltotriose and 63-α-D-maltosyl-maltotetraose were produced by degradation of 6-O-α-maltosyl-βcyclodextrin (50 mg, Megazyme) with granular maltogenic amylase (EC 3.2.1.133) from Bacillus stearothermophilus (12 mg, Sigma-Aldrich) in 2 mL of 10 mM sodium acetate buffer (pH 4.95) overnight.11 Subsequently, the evaporated reaction mixture was peracetylated by dissolution in 2 mL of dry pyridine and 1 mL of acetic anhydride and incubated at room temperature overnight in the presence of a catalytic amount of DBU (1,8-diazabicyclo[5.4.0]undec-7-ene). The solvent was evaporated, and the residue was purified by flash chromatography (EtOAc/heptane) at a ratio of 2:1 to 4:1. Desired fractions were concentrated to give the peracetylated products (26 mg), which were deacetylated by dissolution in MeOH/ Et3N/H2O (2:1:0.4, 3.4 mL) and incubation under stirring overnight. The solvent was evaporated, and the residue was redissolved in 2 mL of H2O. The solution was passed through an Amberlite MB-1 column (mixed-bed resin, Sigma-Aldrich) for desalting. Fractions containing the desired compounds were concentrated and purified on Seppak C18 (Waters, Milford, B

dx.doi.org/10.1021/ac401980m | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

(pH 4.5) and acetonitrile (22:78). The sample was centrifuged and directly analyzed. Five microliters of the labeled samples was injected into a Waters Acquity UPLC system equipped with a FLR detector (fluorescence detector, excitation wavelength of 320 nm and emission wavelength of 420 nm), binary solvent manager, sample manager, and column oven manager. Separation was performed using an Acquity UPLC BEH glycan 1.7 μm, 2.1 × 150 mm column with a VanGuard BEH glycan 1.7 μm, 2.1 × 5 mm precolumn at room temperature. Quantitation by high-performance anion exchange chromatography using pulsed amperometric detection was performed according to Application Note 46 of Dionex (www.dionex. com) on a CarboPac PA-100 column (Dionex) in runs of 64 min duration. The samples were filtered (Seppak C18, Waters) prior to analysis.

■

RESULTS AND DISCUSSION Design of Fast High-Resolution 2D NMR for Carbohydrate Mixture Analysis. Ongoing challenges in the spectral resolution and quantitation of homopolymeric glycan fragments are exemplified in Figure 1, using starch-derived linear and branched α-glucopyranosyl oligomers in beer as an example.15 The homonuclear scalar coupling between protons splits signals into multiplets yielding spectra with a number of 1H signals that is a multiple of the number of 1H sites, which exacerbates the spectral overlap problems (Figure 1A,B).16 1H−13C HSQC spectra have the benefit that 13C chemical shift dispersion is usually on the order of 20 times larger than the 1H chemical shift dispersion, and homonuclear couplings are largely absent due to the low natural isotopic abundance of 13C. Standard 1 H−13C HSQC spectra covering the full 13C spectral region and employing a few hundred increments in the 13C dimension to allow acquisition on the minute time scale inevitably suffer from low 13C resolution and the inability to resolve any glucan signals (Figure 1C). In order to render the rapid spectral resolution of glycans with strong 1H overlap and narrow 13C chemical shift range possible, we employed 1H−13C HSQC spectra with an extremely narrow spectral sweep width in their 13C dimension centered near the 13C1 chemical shift (Figure 2). In this manner, spectra can be acquired rapidly without trade-offs in 13 C signal resolution. As long as spectral resolution is not limited by relaxation, reducing the HSQC 13C spectral width by a given factor increases resolution in the 13C domain by that very factor.17 Due to their intrinsic high-conformational flexibility, soluble glycans can yield sharp 13C signals with line widths below 3 Hz as a result of 13C T2 relaxation times above 100 ms.12 Using a spectral sweep width of only 3 ppm and recording two transients of 200 complex points in the 13C dimension, the 13C free induction decay was sampled for 332 ms on an 800 MHz NMR spectrometer to record 1H1−13C1 HSQC experiments within 15 min duration. Accordingly, we find that commercially available maltooligosaccharides with degrees of polymerization (DP) 1−8 can be resolved in spite of overlapping 1H1 chemical shifts, due to 13C1 chemical shift variations between residues from the different maltooligosaccharides on the order of at least 0.01 ppm (Figure 2A). The resolution of commercially available linear glucose oligomers paves the way for qualitative and quantitative analysis of maltooligosaccharides in biofluids, raw materials, and foods. Figure 2B shows narrow-spectral-width 1H1−13C1 HSQC

Figure 1. One-dimensional and 2D 1H NMR spectra of starch-derived fragments in commercial lager beer. Both the acetal (1H1) region of αglucans in the 1H NMR spectrum (A) and the 1H1−1H2 spectral region in the phase-sensitive DQF-COSY spectrum (B) indicate extremely poor 1H chemical shift dispersion even at 18.7 T (800 MHz) magnetic field strength, precluding the identification and quantification of maltooligosaccharides from 1H NMR in complex mixtures. Spectra of 1H−13C HSQC with spectral widths covering all analyte 13C signals and limited increments in the 13C dimension to allow acquisition on the minute time scale inevitably suffer from low 13 C resolution and the inability to resolve any glucan signals (C). The inset in (C) demonstrates the spectral similarity of maltooligosaccharides. The spectral region highlighted in gray was monitored in optimized HSQC spectra. All spectra are referenced relative to acetone with δ1H = 2.225 and δ13C = 31.45 ppm.

spectra of three commercial beer samples of ale, lager, and pilsner-type in comparison with the corresponding spectra of the commercially available maltooligosaccharides G1−G8. Different levels of maltooligosaccharides in the different beer samples due to different production conditions and raw materials are directly evident. Quantitation of Starch Fragments in Mixtures and Comparison with Chromatographic Methods. Beyond the fingerprinting of beers, narrow-spectral-width 1H−13C HSQC spectra should be applicable in the reliable quantitation of C

dx.doi.org/10.1021/ac401980m | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Figure 2. Narrow spectral width 1H−13C HSQC spectra of commercially available maltooligosaccharide reference compounds (A) and beer samples of ale, lager, and pilsner-type (left to right; B). The spectra exploit the narrow line widths of glycans in the 13C spectral dimension while focusing on 1 H1−13C1 signals for compound identification and rapid sampling of the 13C dimension. All spectra were recorded in D2O at 298 K within 15 min using a spectral width of 3 ppm, a carrier frequency of 99.0 ppm, and a linear sampling of 200 complex data points in the indirect 13C dimension.

experiments of 15 min duration. As the NMR analysis of biotechnological samples by high-resolution 1H−13C HSQC NMR is often sensitivity-limited, sample preconcentration or hardware developments will inherently further reduce the analysis time without influencing the resolution. All spectra were acquired as conventional spectra with linear sampling schemes. Thus, sample throughput could in principle be further improved by nonconventional methodologies to optimize data acquisition and processing, such as sparse data sampling or rapid pulsing.18 Subsequently, concentrations of homopolymeric fragments of starch in mixture were determined and compared to validated methods. Integration of maltooligosaccharide, isomaltose, and panose signals in narrow-spectral-width 1H−13C HSQC spectra of beer was used to obtain molecular concentrations. This subset of glucans is routinely resolved by validated chromatographic methods, and thus HSQC-based quantitation of starch fragments in beer could be compared with quantitation by HPAEC (high-performance anion exchange chromatography) with pulsed amperometric detection and by precolumn derivatization and HILIC (hydrophilic interaction liquid chromatography) with fluorescence detection (Figure 4A).3 Comparison of NMR-determined concentrations for G1−G8, isomaltose, and panose to HPAEC-derived concentrations yielded an R2 >0.99 and a slope of 1.01, while quantitation with HILIC−UPLC yielded an R2 >0.99 and a slope of 1.02. A principal difference between chromatographic and NMR spectroscopic methods is the detection of whole molecules in the former and of atomic sites in the latter. Accordingly, several 1 H1−13C1 signals are observed for each molecular species in NMR spectra. While the presence of several 1H1−13C1 signals per molecular species inherently increases spectral crowding, the chance of finding non-overlapped reporter groups19 for quantitation increases at the same time. Quantifying beer maltooligosaccharides by different 1H1−13C1 NMR signals from the same molecular species yields correlations, as shown in Figure 4B, with an R2 = 0.996 for independent measurements. While the proportionality of cross-peak volumes to metabolite concentrations is not guaranteed in HSQC-based mixture analysis due to uneven 13C excitation as well as mismatched coherence transfer times, these shortcomings lose significance

maltooligosaccharides in mixtures because distinguishable, nonoverlapped signals were detected for all commercially available linear glucose oligomers. To examine the possibility of quantitative spectroscopic profiling of starch fragments in mixtures, standard samples of G1−G8, isomaltose, and panose were prepared by serial dilution (Figure 3A, G1−G5 shown)7a

Figure 3. Maltooligosaccharide reference compound mixtures used for quantitative calibration. Samples of maltooligosaccharide mixtures were prepared by weighing in reference compounds, subsequent 3and 9-fold dilution, narrow-spectral-width 1H1−13C1 HSQC spectroscopy (A), and integration of resultant signals to yield standard curves, as shown in (B).

and subjected to spectral runs of 15 min. Calibration curves of resultant HSQC signal volumes versus concentration can be obtained for all pure reference compounds and showed excellent linearity, as expected (exemplified for G1−G4 in Figure 3B). HSQC spectra were recorded with INEPT delays that were optimized for 1JCH = 170 Hz of the 1H113C1 group in α-glucopyranosyl units. A lower limit of quantification on the order of 10 μg/mL could be inferred for the 1H−13C HSQC D

dx.doi.org/10.1021/ac401980m | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Figure 4. Quantitation of maltooligosaccharides in lager beers by 1 H1−13C1 HSQC spectroscopy and by chromatography-based assays. Comparison to quantitation with more time-consuming chromatographic runs yielded a strong linear association with R2 above 0.99 (A). Likewise, the reproduction of NMR quantitation with different atomic sites in the same maltooligosaccharide underlines the high robustness of NMR-based maltooligosaccharide quantitation (B).

in glycan analytes with narrow chemical shift ranges and highly similar 1JC1H1 coupling constants.20 Superior Resolution of Linear and Branched Glucans with Identical Degrees of Polymerization in Optimized HSQC. Chromatographic methods are well-suited to separate glycans according to their size. But at the same time, it remains a challenge to obtain baseline-separated signals in chromatography-based glycan analysis due to the overlap of glycans with the same molecular mass, such as linear and various branched starch fragments with the same degree of polymerization. The structural information provided by NMR is particularly wellsuited to remedy these shortcomings. Figure 5 compares the chemical resolution provided by UPLC−HILIC and NMR spectroscopy. Besides improved resolution, the absence of different response factors for different glycans, the detection of chemical shifts as instrument invariable values, and a wide dynamic range are principal strongholds of NMR-based detection. The independent benefits of chromatographic separation according to size (Figure 5A) and the rich structural information encoded in the NMR chemical shifts (Figure 5B) suggest that hyphenated LC-NMR methods employing rapid HSQC-based NMR detection have great future potential for glycan mixture analysis. Figure 5 highlights the ease of distinguishing signals arising from branched and linear glucans in HSQC spectra. Glycosidic α(1−4) linkages in the vicinity of branch points have distinctly downfield 13C1 chemical shifts relative to 13C1 signals in linear α(1−4) glucans, while branch point (α(1−6)) 1H1 signals are well-separated from α(1−4) 1H1 signals. The branched starch component amylopectin contains on the order of 5% glucopyranosyl residues in α(1−6) glycosidic bonds. Hence, extensive starch degradation by endohydrolases mostly yields single branched glucans containing fewer α(1−6) glycosidic bonds than α(1−4) glycosidic bonds, thus approaching the case where most branched starch fragments yield single HSQC cross signals per molecule in the spectral region of α(1−6) glycosidic bonds. Accordingly, branch point cross signals in highly resolved HSQC spectra are particularly well-suited reporters

Figure 5. Comparison between resolution of glycans in a beer sample by chromatography (HILIC−UPLC) and by narrow-sweep-width HSQC NMR. Chromatography (A) yields overlapped signals of glycans with identical degrees of polymerization, severely even for simple sugars in complex mixtures (top). Chromatogram of standard maltooligosaccharide compounds is shown (A, bottom). The sample has undergone derivatization at the reducing end with a fluorophore (2-aminobenzamide) prior to liquid chromatography separation. (B) Narrow-spectral-width 1H1−13C1 HSQC recorded within 15 min for glycan mixture analysis in a native beer sample in D2O. Assignments were performed by comparison to standard compound spectra.

of molecular structures. Simple counting shows that at least twelve different α(1−6) attached branch-point-types could be distinguished in beer, eight of which could be assigned to molecular structures; only the availability of reference compounds currently limits the identification of branched fragments carrying several α(1−4) glucopyranosyl residues attached to branches. In addition, at least eight structural isomers of glucopyranosyl-disaccharides are easily detected and differentiated, including α,α-trehalose, α,β-trehalose, kojibiose, cellobiose, and gentiobiose, besides maltose and isomaltose (Figure 5B). The comparison of beer spectra with spectra of commercially available and in-house synthesized branched glucans for the assignment of branched starch fragments is shown in Figure 6A,B. Different spectroscopic fingerprints in different beer E

dx.doi.org/10.1021/ac401980m | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

NMR spectroscopy exploiting the high resolution along the 13C dimension. Considerably higher throughput can be achieved in this way, as compared to chromatographic methods. Twodimensional NMR-based approaches yield distinct analytical benefits in glycan mixture analysis due to the unique molecular information on analytes obtainable, albeit NMR spectroscopy as of now remains a less-sensitive detection method compared with fluorescence or pulsed amperometric detection. Signal overlap, however, often presents the bigger challenge in the analysis of food-grade carbohydrate mixtures or plant extracts than sensitivity.22 Overall, carefully devised high-resolution experiments exploiting narrow 13C signal widths in fast experiments, preferably recorded at state-of-the-art high-field NMR instruments, seem underexplored in their unique potential to support industrial and academic work streams related to the profiling of glycan composition and its dynamic change.

■

ASSOCIATED CONTENT

S Supporting Information *

Relevant information that amplifies Figure 5A,B. HSQC spectra of selected standard compounds and expanded view of the chromatogram. This material is available free of charge via the Internet at http://pubs.acs.org.

■

Figure 6. Comparison of 1H1−13C1 HSQC spectra of lager beer (light green) with spectra of commercial and in-house synthesized branched α-glucan reference compounds (A,B). Both the α(1−4) (A) and especially the α(1−6) (B) anomeric signals in branched fragments are resolved from maltooligosaccharide signals. (C) Comparison between 1 H spectral regions at 4.9−5.0 ppm of lager-type and ale-type beers shows the different degree of amylolysis in the vicinity of branch points as the result of different production conditions with the insets demonstrating the narrow line width of glucans in the absence of resolution enhancement during postprocessing.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Fax: +45-3327-4708. Tel.: +45-3327-5301. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS NMR spectra were recorded at the 800 MHz spectrometer of the Danish National Instrument Center for NMR spectroscopy of Biological Macromolecules at the Carlsberg Laboratory.

■

spectra reflect differences in production processes, ingredients, nutritional value, and organoleptic properties (Figure 6C). As an example, comparison of lager-type and ale-type beers directly reflects the different degree of amylolysis in the vicinity of branch points, resulting in different relative levels of branched dextrins (Figure 6C). As an example, a substantial amount of branched starch fragments with DP 2−5 is found in lager beer, while longer fragments with DP >5 predominate in the ale sample (δ13C ≈ 99.8 ppm). The inset in Figure 6C shows traces along the 13C dimension of the lager beer HSQC to demonstrate narrow line widths (