Convert Chemstation D Files Endo

The alkylation method employed in this study converts amines and organic acids into volatile esters and carbamates, allowing them to be analysed by GC–MS. Hence, the hydroxyl. The internal standard solution d4–alanine was added to the pellet following the addition of extraction solution. A total of 6. Individual raw data files were converted to the vendor neutral mzML format with msconvert [37] and processed with the Trans-Proteomic Pipeline Version 4.8. Sequence determination was performed using Comet software [39] to search the Swissprot D. Rerio database, with an abbreviated FASTA.

The UAG termination codon is generally recognized as the least efficient and least frequently used of the three universal stop codons. This is substantiated by numerous studies in an array of organisms. We present here evidence of a translational readthrough of a mutant nonsense UAG codon in the transcript from the cysteine sulfinic acid decarboxylase ( csad) gene (ENSDARG8) in zebrafish. The csad gene encodes the terminal enzyme in the taurine biosynthetic pathway. Taurine is a critical amino acid for all animals, playing several essential roles throughout the body, including modulation of the immune system. The sa9430 zebrafish strain (ZDB-ALT-1) has a point mutation leading to a premature stop codon (UAG) 20 amino acids 5’ of the normal stop codon, UGA.

Data from immunoblotting, enzyme activity assays, and mass spectrometry provide evidence that the mutant is making a CSAD protein identical to that of the wild-type () in terms of size, activity, and amino acid sequence. UAG readthrough has been described in several species, but this is the first presentation of a case in fish. Also presented are the first data substantiating the ability of a fish CSAD to utilize cysteic acid, an alternative to the standard substrate cysteine sulfinic acid, to produce taurine. Introduction To assess a possible amelioratory role for taurine in dietary induced inflammation, our laboratory seeks to procure a strain of zebrafish deficient in endogenous taurine synthesis.

This will enable clear observation of the effects of graded dietary supplementation. Taurine (2-aminoethanesulfonic acid) is a critical amino acid for animals and must be synthesized de novo or obtained through the diet.

For taurine synthesizers, the cysteine sulfinic acid decarboxylase (CSAD) enzyme () catalyzes the terminal reaction in the primary biosynthetic pathway (). Essential roles for taurine include osmoregulation, bile salt conjugation, and protection from oxidative stress []. Species lacking sufficient endogenous synthesis require dietary supplementation. This is particularly apparent in carnivores who tend to have low levels of CSAD or none at all. Strictly carnivorous cats have low levels of CSAD but require high levels of taurine as well as methionine, which along with cysteine is a precursor of taurine synthesis [,].

In fish, deficiencies in taurine production have become increasingly apparent as aquaculture feeds exchange greater proportions of fish meal for plant protein sources containing no taurine. The supplementation requirements of several commercially relevant species have been described, including by our laboratory for the strict marine carnivore, cobia ( Rachycentron canadum). Taurine synthesizers also benefit in terms of growth from supplementation in the feed, with 1.5% being the standard amount in commercial formulations. FDA (United States Food and Drug Administration) recently approved taurine supplementation of feed for farmed fish intended for human consumption []. Cysteine sulfinic acid decarboxylase (CSAD) is the terminal enzyme in the taurine biosynthetic pathway.

(Modified from Vitvitsky et al []). Knocking out the csad gene in mice leads to an 83% reduction in plasma levels of taurine, with resulting mortality within 24 h of birth by the third generation of inbred homozygotes []. Residual production of taurine may be via the cysteamine pathway (). Cats, as insufficient synthesizers of taurine, exhibit retinal degeneration and hepatic lipidosis when fed a diet without taurine supplementation [,]. Taurine-deficient fish display similar liver maladies, largely expected since fish conjugate bile salts to taurine [,]. CSAD mRNA is found in the earliest stages of zebrafish embryonic development, including in the yolk at the gastrula stage, and in both the yolk and notochord in the somite and pharyngula stages.

In the pharygula stage, 24–48 hpf (hours post fertilization), CSAD expression is evident in several additional tissues including the liver, brain, and optic cup [,]. Knockdown of CSAD mRNA or TauT (taurine transporter) mRNA during early embryonic development results in cardiomyopathy and mortality in zebrafish [,].

Taurine is also an immunomodulator, mitigating inflammation by scavenging free radicals and inhibiting the production of several potent inflammatory agents including tumor necrosis factor-α (TNF-α), interleukins, prostaglandins, superoxide anion, and nitric oxide [,]. The sa9430 mutant zebrafish strain (Sanger, Zebrafish Mutation Project, []) was acquired as a potential knockout for csad suitable for our dietary studies. This strain has a C→T point mutation in the single copy of the csad gene on chromosome 23. The mutation alters a codon that would normally be translated to glutamine to a UAG stop codon 20 amino acids prior to the normal UGA stop codon. No phenotype has been described for this mutant, but commercial feeds typically contain adequate dietary taurine and the lack of de novo synthesis may be masked. Zebrafish deficient in taurine synthesis that do not receive dietary supplementation are expected to exhibit pathologies similar to those seen in the knockdown studies, including mortality. Since our fish were reared and maintained on a commercial taurine-supplemented diet, as well as taurine-containing live artemia, gross changes in phenotype were not necessarily expected.

However, we did not anticipate the results presented here from immunoblotting, enzyme activity assays, and mass spectrometry which show that the sa9430 mutant is producing wild-type CSAD protein. The CSAD of some animals can utilize cysteic acid as a substrate for taurine synthesis, including cats and rats [,]. In this pathway, there is the direct conversion of cysteic acid to taurine without a hypotaurine intermediate.

Data presented here reveal the ability of zebrafish CSAD to utilize cysteic acid as an alternative to cysteine sulfinic acid as a substrate for taurine synthesis. Standard Feeds for Zebrafish Contain Taurine At 5–10 dpf (days post fertilization) larvae are fed paramecia until their mouths are sufficiently large to consume artemia or the smallest commercial pelleted feeds. Our analysis by HPLC detects no taurine in paramecia, so during this feeding period, the larvae are dependent upon maternal contribution of taurine and csad mRNA as well as their own de novo synthesis []. At approximately 10 dpf, zebrafish begin consuming live artemia, which are grown without taurine enrichment in our facility.

According to LC-MS analysis, artemia naturally contain 0.033% ± 0.003% taurine. At this stage of development, zebrafish larvae can also begin consuming the smallest pelleted feeds. These formulations, commonly fed to zebrafish in our facility, were found to contain approximately 1.5% taurine. Zebrafish Synthesize Sufficient Taurine for Homeostasis Juvenile zebrafish were fed diets containing either zero or 4% taurine for 8 weeks. Survival and growth for zebrafish on both diets were equivalent. Whole body taurine values at the conclusion of the feeding trial were 1.37 ± 0.03% for fish on the zero taurine diet and 2.04 ± 0.28% for fish on the 4% taurine diet. These differences are statistically significant.

Expression of taurine biosynthetic genes CSAD, CDO (cysteine dioxygenase), and ADO (2-aminoethanthiol dioxygenase), as well as TauT, were confirmed by RT-PCR []. Nucleotides That Border the UAG May Increase Readthrough Frequency The sa9430 zebrafish strain has an adenine in the −1 position and a pyrimidine in the +4 position relative to the nonsense UAG stop codon (). Sequence information supplied with the strain was confirmed by PCR amplification and sequencing of a 364-bp (base pair) region of DNA containing the C→T point mutation.

If the same trends hold true for zebrafish as for other studied organisms, these particular nucleotides flanking the UAG could increase frequency of readthrough [,]. UAG can be translated as glutamine during readthrough, which would result in a wild-type protein [,,,].

Sa9430 CSAD Is the Same Size as the Wild-Type CSAD Liver proteins from wild-type and sa9430 zebrafish were extracted, genotypes confirmed, and immunoblotting performed. For genotyping, a 364-base region was sequenced which includes the site of the C→T point mutation in the sa9430 mutant.

As shown in, protein sizes are the same for detectable CSAD from both wild-type and sa9430. Immunoblotting of heterozygotes resulted in a single band of the same size (data not shown). The binding site for the anti-CSAD antibody is predicted to be present in a truncated CSAD in addition to the wild-type as shown in the protein models in.

Wild-type CSAD and sa9430 CSAD are an estimated 58.1 kDa and 60.0 kDa, respectively, using molecular weight standards for reference for analysis using the Image Lab software (Bio-Rad). These are likely equivalent values within the error range of the program. Though the gradient gel type used is sufficient to visualize a size difference of 20 amino acids (2.5 kDa), we confirmed the absence of a smaller band by running the gel 30% longer (data not shown). In addition, to be sure a concentrated level of protein was not masking a smaller band, the 1:10 dilutions for each strain were analyzed (lanes 2 and 4 of ).

The absence of a visible truncated protein suggests that if any is being produced it is below detectable levels or being degraded. Wild-Type and sa9430 CSAD Enzymes Catalyze the Conversion of Cysteic Acid to Taurine at Comparable Rates For analysis of enzyme activity, liver proteins were extracted from wild-type and F2 generation sa9430 homozygotes. Based on computer modeling suggesting that the active site would be preserved even in a truncated protein and immunoblotting data indicating a full-length protein in the sa9430 mutant, we anticipated some level of functional enzymatic activity. As shown in, the wild-type and sa9430 CSAD convert cysteic acid to taurine at statistically indistinguishable rates. Mass Spectrometry of Tryptic Peptides Indicates sa9430 Is Producing A Wild-Type CSAD Protein Total protein from liver extracts was immunopurified using an anti-CSAD antibody and prepared for mass spectrometry (MS). Peptides detected by tandem MS were aligned to the published CSAD sequence 482 amino acids in length () as well as the predicted X1 isoform 544 amino acids in length (). The sa9430 zebrafish strain is producing full-length CSAD protein, with sequence coverage of 542 amino acids for both wild-type and sa9430 ().

The nonsense UAG in sa9430 is being translated to glutamine, the same amino acid as in the corresponding position in the wild-type. N-terminal sequence coverage matches the predicted X1 isoform. Discussion In mammals, the UAG codon has been described as notoriously “leaky”, and the nucleotides preceding and following the stop codon influence the frequency of readthrough. Readthrough frequency is increased in mammals when an adenine is in the −1 position, the nucleotide preceding the UAG stop codon. A pyrimidine in the +4 position, which corresponds to the nucleotide directly following the UAG, is associated with a greater chance of readthrough as compared to a purine []. The sa9430 zebrafish strain has both an adenine in the −1 position and a pyrimidine (cytosine) in the +4 position.

Stop codon efficiency and frequency are correlated. In many organisms, UAG is not only the least efficient in terminating translation, it is also the least frequently occurring codon []. In lower eukaryotes such as Candida and Drosophila, UAGA is the most efficient and frequent sequence, while UAGC is the least efficient and frequent sequence []. UAGC is the sequence found in the sa9430 zebrafish strain (). A study on stop codon frequency in blunt snout bream ( Megalobrama amblycephala) reported that UAG has a relative synonymous codon usage (RSCU) value of 0.59 as compared to 1.11 for UAA and 1.29 for UGA []. RCSU value is a measure of codon bias, and the score of.

Zebrafish Strains and Maintenance Wild-type and sa9430 zebrafish were maintained at the zebrafish facility of the Aquaculture Research Center at the Institute of Marine and Environmental Technology (Institutional Animal Care and Use Committee at University of Maryland, Baltimore #0315011, approved April 2015). Fish were maintained on a 14 h light, 10 h dark cycle at 28.5 °C.

Larvae were fed a combination of paramecia, artemia, and GEMMA Micro 75 (Skretting). At the appropriate size, fish transitioned to GEMMA Micro 150, then 300, with occasional additions of artemia, particularly in the week preceding mating. Heterozygous embryos of csad sa9430 Tb (Sanger, Zebrafish Mutation Project) were raised and naturally bred to obtain homozygotes with wild-type csad and homozygotes with sa9430 mutant csad. Genotyping For confirmation of genotype, caudal fin clips were obtained and immersed in genomic DNA extraction buffer (50 mM KCl, 10 mM Tris-HCl (pH = 8.0), 150 mM MgCl 2, 0.3% Tween-20, and 0.3% NP40 in sterile MilliQ water), boiled at 95–100 °C for 15 min, cooled on ice, and digested with proteinase K at 55 °C for 1–3 h. After digestion, lysates were boiled at 95–100 °C for 15 min to inactivate proteinase K and centrifuged for 3 min at 12,000× g [].

The resulting genomic DNA was amplified by PCR using zebrafish csad-specific primers: •. Reverse: 5’ GATGCCAATCGTTTGACCAGT 3’ Sequencing of the 364-base-pair PCR product was performed in a 10 µL reaction volume consisting of 40–150 ng PCR product, 3 pmol of primer, 0.5 µL Big Dye v3.1 sequencing mix and 1.5 µL 5X sequencing buffer (Applied Biosystems, ThermoFisher Scientific, Waltham, MA, USA). Cycling parameters were 95 °C for 5 min, followed by 50 cycles at 95 °C for 15 s, 50 °C for 15 s, 60 °C for 4 min. The sequencing product was purified by adding 60 µL 100% isopropanol and 30 µL H 2O, mixing thoroughly, incubating at room temperature for 30 min, and centrifuging at 2000× g for 30 min. The supernatant was decanted and 100 µL 70% isopropanol was added to wash the DNA, followed by another centrifugation at 2000× g for 14 min and decanting of the supernatant.

Labeled products were air dried for 20–30 min before addition of 10 µL HI-DI formamide from Applied Biosystems. The mixture was heated at 95 °C for 2 min and then immediately put on ice. The denatured product was sequenced using an Applied Biosystems 3130XL Genetic Analyzer (ThermoFisher Scientific, Waltham, MA, USA) and compared with the published sequences for the wild-type (ENSDARG8) and sa9430 (ZBD-ALT-1) strains using the Sequencher program (Version 5.0.1, Gene Codes, Ann Arbor, MI, USA). Liver Protein Isolation Zebrafish were euthanized by rapid cooling followed by decapitation. Livers were isolated from wild-type and sa9430 zebrafish and frozen at −80 °C. Liver tissue was homogenized in buffer containing 60 mM Potassium phosphate (pH 7.4), 5 mM DTT, 50 mM sucrose, and 0.5 µM pyridoxal-5’–phosphate (PLP) at a volume of approximately 1:15 weight (mg): lysis buffer volume (μL) []. Homogenization was achieved by vortexing and pipetting up and down with a P200 micropipettor.

Samples were centrifuged for 5 min at 1500× g and supernatant containing the protein fraction collected. For the activity assay, the supernatant was dialyzed in a Slide-a-Lyzer Mini Dialysis Unit (ThermoFisher Scientific, Waltham, MA, USA) for 2 h at 4 °C in the homogenization buffer according to manufacturer’s instructions. Total protein concentration was determined using the Qubit Assay (ThermoFisher Scientific, Waltham, MA, USA). Immunoblotting Protein extracts were prepared as described in, combined with standard SDS-PAGE sample buffer, heated for 3 minutes at 95 degrees C, and centrifuged for one minute at 10,000× g. For the immunoblot shown in, lanes 1–4 of a 4%–12% Bis-Tris protein gel (NuPAGE Novex, (ThermoFisher Scientific, Waltham, MA, USA)) were loaded with 3.7 μg, 0.37 μg, 2.6 μg, and 0.26 μg protein, respectively, and run according to the manufacturer’s protocol for 32 min using MOPS running buffer.

Proteins were transferred to a PVDF membrane in the Trans-Blot Turbo Transfer System (Bio-Rad, Hercules, CA, USA). Immunoblotting was performed in the iBind Western System (ThermoFisher Scientific, Waltham, MA, USA) using rabbit anti-zebrafish CSAD antibody at a dilution of 1:1000 (#6862, provided by Plant Sensory Systems, LLC, Halethorpe, MD, USA), as the primary antibody, and goat anti-rabbit IgG H&L HRP conjugate at a dilution of 1:2000 (Bio-Rad, Hercules, CA, USA) as the secondary antibody. A chemiluminescent signal was generated with addition of Clarity Western ECL substrate and imaged in a ChemiDoc Touch Imaging System (Bio-Rad, Hercules, CA, USA). CSAD Activity Assay For the activity assay, reaction mixtures were prepared containing 85 μL buffer (same as the homogenization buffer in ) and 5 μL total protein from liver (4.4 μg wild-type or 7.1 μg sa9430, ). The CSAD enzyme and its PLP cofactor in the buffer were allowed to incubate at 23 °C for 15 min before proceeding with the assay at the same temperature.

The time points for the assay were 0, 30, and 90 min. Beginning with the 90-min time point, 10 μL 50 mM cysteic acid (substrate for CSAD) was added to the reaction mixtures.

Immediately following the addition of the substrate at the 0-time point, all reactions were stopped with addition of 100 μL (volume equal to total reaction volume) ice cold ethanol with 5% acetic acid. Reaction mixtures were centrifuged at 350× g for 10 min 100 μL supernatant was transferred to a clean microcentrifuge tube and dried by evaporation at 70 °C. Amino Acid Analysis by HPLC Dry samples were suspended in 0.1N HCl and filtered through 0.45 micron filters (EMD Millipore, Billerica, MA, USA). 5 μl of the filtered extracts was derivatized according to the AccQTag Ultra Derivitization Kit protocol (Waters Corporation, Milford, MA, USA). Amino acids were analyzed using an Agilent 1260 Infinity High Performance Liquid Chromatography System equipped with ChemStation (Agilent Technologies, Santa Clara, CA, USA) by injecting 5 μL of the derivatization mix onto an AccQTag Amino Acid Analysis C18 (Waters, Milford, MA, USA) 4.0 μm 3.9 × 150 mm column heated at 37 °C.

Amino acids were eluted at 1.0 mLmin −1 flow with a mix of 10-fold diluted AccQTag Ultra Eluent (C; Waters Corporation, Milford, MA, USA), ultra-pure water (A) and acetonitrile (B) according to the following gradient: initial, 98.0% C/2.0% B; 2.0 min, 97.5% C/2.5% B; 25.0 min, 95.0% C/5.0% B; 30.5 min, 94.9% C/5.1% B; 33.0 min, 91.0% C/9.0% B; 38 min, 40.0% A/60.0% B; 43 min, 98.0% C/2.0% B. Derivatized amino acids were detected at 260 nm using a photo diode array detector. Amount of amino acids was expressed in g per g of dry weight of sample (% DW) making reference to AABA signal, external calibration curve of standard hydrolysate amino acids and dry weight of samples.

Preparation of Embryos for Amino Acid Analysis by HPLC Fifty 1-hpf F2 embryos were collected from each of two matings of wild-type fish as well as 50 from two matings of homozygous sa9430 fish. Lyophilized embryos were extracted in 70% ethanol containing 0.154 mM d-norleucine for evaluation of extraction efficiency. Samples were sonicated for 60 min at 25 °C in a bath sonicator (Branson 1200, Emerson, Danbury, CT, USA) followed by centrifugation at 350× g for 10 min (IEC, ThermoFisher Scientific, Waltham, MA, USA). The ethanol fraction was retained and dried. These samples were then analyzed by HPLC as described in. Preparation of Paramecia and Artemia for Taurine Analysis Samples of paramecia and artemia used to feed zebrafish in our facility were centrifuged at 350× g for 10 minutes (IEC, ThermoFisher Scientific, Waltham, MA, USA), and the pellet was retained and lyophilized. Samples were resuspended in 70% methanol and sonicated for 60 min at 25 °C in a bath sonicator (Branson 1200).

Following centrifugation at 350× g for 10 min (IEC), the methanol fraction was retained and dried. These samples were then prepared for taurine analysis by HPLC () or LC-MS [], respectively. Taurine levels in artemia were analyzed by Aaron Watson, Ph.D. Mass Spectrometry Liver extracts were prepared as described in.

100 μL total liver protein extracts contained 74. Modem Zte Mf 1805 more. 7 μg and 52.2 μg from wild-type and sa9430 fish, respectively. These extracts were each incubated with 5 μg rabbit anti-zebrafish CSAD (provided by Plant Sensory Systems, LLC, Halethorpe, MD, USA) with shaking at 1400 rpm for 1 h at 4 °C.

5 μg goat anti-rabbit—biotin (ThermoFisher Scientific, Waltham, MA, USA) was added to each tube, followed by shaking at 1400 rpm for 1 h at 4 °C. 30 μL streptavidin beads from the SMART Digest Immunoaffinity Kit (ThermoFisher Scientific, Waltham, MA, USA) were added followed by shaking at 1400 rpm at 4 °C for one hour followed by an additional hour at 23 °C. The wash protocol outlined in the kit was followed. Digestion occurred at 70 °C for one hour. The supernatant was retained and lyophilized. Peptides were analyzed by LC-MS/MS using a Dionex UltiMate 3000 Rapid Separation nanoLC and a linear ion trap—Orbitrap hybrid mass spectrometer (ThermoFisher Scientific, Waltham, MA, USA).

Approximately 1 μg of peptide samples was loaded onto the trap column, which was 150 μm × 3 cm in-house packed with 3 um C18 beads. The analytical column was a 75 um × 10.5 cm PicoChip column packed with 1.9 um C18 beads (New Objectives). The flow rate was kept at 300nL/min.

Solvent A was 0.1% FA in water and Solvent B was 0.1% FA in ACN. The peptide was separated on a 90-min analytical gradient from 5% ACN/0.1% FA to 40% ACN/0.1% FA. The mass spectrometer was operated in data-dependent mode. The source voltage was 2.10 kV and the capillary temperature was 275 °C.

MS 1 scans were acquired from 400–2000 m/ z at 60,000 resolving power and automatic gain control (AGC) set to 1 × 10 6. The top ten most abundant precursor ions in each MS 1 scan were selected for fragmentation. Precursors were selected with an isolation width of 1 Da and fragmented by collision-induced dissociation (CID) at 35% normalized collision energy in the ion trap. Previously selected ions were dynamically excluded from re-selection for 60 s. The MS 2 AGC was set to 3 × 10 5. Apex triggering was enabled for the peptide analysis on the Q Exactive HF (ThermoFisher Scientific, Waltham, MA, USA).

All setup was same as above except that the top15 most abundant precursor ions in each MS1 scan were selected for fragmentation. Precursors were selected with an isolation width of 2 Da and fragmented by Higher-energy collisional dissociation (HCD) at 30% normalized collision energy in the HCD cell. Individual raw data files were converted to the vendor neutral mzML format with msconvert [] and processed with the Trans-Proteomic Pipeline Version 4.8. Sequence determination was performed using Comet software [] to search the Swissprot D. Rerio database, with an abbreviated FASTA containing CSAD sequence 482 amino acids in length () or the predicted X1 isoform 544 amino acids in length (). Methionine oxidation and carbamidomethylation of cysteine were allowed as variable and fixed modifications, respectively. The enzyme specificity parameter was set to trypsin allowing for 1 missed cleavage site per peptide.

MS1 precursor ion mass tolerance and MS2 product ion mass tolerance parameters were set to 10 ppm and 0.4 Da, respectively. Peptide spectra matched to theoretical spectra calculated from the database were validated using Peptide Prophet and peptides were assembled into protein groups with Protein Prophet. Sequences were analyzed using MacVector 12.7.5. This work was supported by a Maryland Industrial Partnerships Grant (Project #5504) in conjunction with Plant Sensory Systems, LLC. We thank Michelle Price, for assistance with HPLC.

Proteomics services were performed by the Northwestern Proteomics Core Facility, generously supported by NCI CCSG P30 CA060553 awarded to the Robert H Lurie Comprehensive Cancer Center and the National Resource for Translational and Developmental Proteomics supported by P41 GM108569. We thank Benjamin Oyler for analysis of mass spectrometry data.

This is contribution #5360 from UMCES and #17-206 from IMET.

And * • Institute of Food, Nutrition and Health, ETH Zurich, Zurich, Switzerland The oxidation of cereal (1→3,1→4)-β-D-glucan can influence the health promoting and technological properties of this linear, soluble homopolysaccharide by introduction of new functional groups or chain scission. Apart from deliberate oxidative modifications, oxidation of β-glucan can already occur during processing and storage, which is mediated by hydroxyl radicals (HO •) formed by the Fenton reaction.

We present four complementary sample preparation strategies to investigate oat and barley β-glucan oxidation products by hydrophilic interaction ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS), employing selective enzymatic digestion, graphitized carbon solid phase extraction (SPE), and functional group labeling techniques. The combination of these methods allows for detection of both lytic (C1, C3/4, C5) and non-lytic (C2, C4/3, C6) oxidation products resulting from HO •-attack at different glucose-carbons. By treating oxidized β-glucan with lichenase and β-glucosidase, only oxidized parts of the polymer remained in oligomeric form, which could be separated by SPE from the vast majority of non-oxidized glucose units. This allowed for the detection of oligomers with mid-chain glucuronic acids (C6) and carbonyls, as well as carbonyls at the non-reducing end from lytic C3/C4 oxidation. Neutral reducing ends were detected by reductive amination with anthranilic acid/amide as labeled glucose and cross-ring cleaved units (arabinose, erythrose) after enzyme treatment and SPE. New acidic chain termini were observed by carbodiimide-mediated amidation of carboxylic acids as anilides of gluconic, arabinonic, and erythronic acids. Hence, a full characterization of all types of oxidation products was possible by combining complementary sample preparation strategies.

Differences in fine structure depending on source (oat vs. Barley) translates to the ratio of observed oxidized oligomers, with in-depth analysis corroborating a random HO •-attack on glucose units irrespective of glycosidic linkage and neighborhood. The method was demonstrated to be (1) sufficiently sensitive to allow for the analysis of oxidation products also from a mild ascorbate-driven Fenton reaction, and (2) to be specific for cereal β-glucan even in the presence of other co-oxidized polysaccharides. This opens doors to applications in food processing to assess potential oxidations and provides the detailed structural basis to understand the effect oxidized functional groups have on β-glucan's health promoting and technological properties.

Introduction Cereal mixed-linkage (1→3,1→4)-β- D-glucan (BG) is a soluble dietary fiber with great potential for functional foods due to its well-established health-promoting properties such as blood cholesterol lowering and blood glucose regulation (; ). BG is mainly (≥90%) comprised of cellotriosyl and cellotetraosyl units linked by β-(1→3) glycosidic bonds, forming a linear homo-polysaccharide of β- D-glucopyranose (see Figure ). The cellotriosyl/cellotetraosyl ratio is characteristic for the source of BG, with oat (OBG; 1.7–2.4) having smaller ratios than barley (BBG; 2.7–3.6), and is typically determined by hydrolysis with lichenase and ion-exchange chromatography-pulsed amperometric detection. The endo-enzyme lichenase (EC 3.2.1.73) selectively cleaves the β-(1→4)-linkages of β-(1→3)-linked glucose units, releasing gluco-oligomers (Glc n) with β-(1→3)-linked reducing end units (abbreviated as G 1- 4G 1- 3 G and G 1- 4G 1- 4G 1- 3 G for degrees of polymerization (DP) of n = 3 and 4, respectively; see Figure ).

The DP3/DP4 fine structure differences have an impact on BG's physico-chemical properties, for example, BBG having a higher propensity to form gels than OBG (;; ). (A) Chemical structure of the glucopyranose (Glc) repeating unit of cereal β-D-glucan (BG) with numbered carbons and indicated glycosidic linkage proportions, together with the symbolic representation of the polysaccharide and its cellulosic β-(1→4) regions. In addition, the fine structure analysis by selective hydrolysis with lichenase is shown, with blue arrows indicating the specific β-(1→4)-linkages susceptible to cleavage by the endo-enzyme and the resulting characteristic ratios of formed oligosaccharides DP3 and DP4. (B) Workflow of the cereal BG oxidation study (from oat and barley) with the two oxidation conditions (a) and (b) and the four complementary sample preparation strategies I–IV.

Blue ○ = G, glucose unit; blue ○–OH = G, reducing end glucose unit; oxBG; oxidized β-glucan; AH 2, ascorbic acid; C=O, carbonyl group; CO 2H, carboxylic acid group; SPE, solid phase extraction; UPLC-MS/MS, ultraperformance liquid chromatography tandem mass spectrometry. During processing and storage, BG can be degraded enzymatically or chemically to lower-molecular-weight products with diminished viscosity () and health benefits (). Thereby, the chemically-induced oxidative degradation of BG has been shown to occur in the presence of substances commonly found in foodstuff when in contact to atmospheric oxygen (O 2), namely traces of transition metal (Fe or Cu) and a reducing agent such as ascorbic acid (AH 2) (;,). For example, during thermal or high pressure treatment of BG solutions, new carbonyl groups of up to ~10 μmol/g BG (or ~2 C=O per 1,000 repeating units) were detected (). The reactive oxygen species responsible for the loss in viscosity/ molecular weight ( M w) was identified by indirect spin trapping and electron spin resonance (ESR) spectroscopy as hydroxyl radical (HO •; ).

The catalytic cycle of HO • production is thought to be induced by the pro-oxidant activity of AH 2, which reduces intrinsic iron and dissolved O 2 to produce Fe 2+ and hydrogen peroxide (H 2O 2; ), the two substrates for the reaction (). Apart from threatening the molecular integrity of BG and its health benefits (;; ), deliberate incorporation of new functional groups through oxidation has also been reported to change BG's technological () and health promoting properties (), providing great potential to influence BG functionality. Hence, cereal BG oxidation has been the focus of numerous studies employing various analytical techniques to investigate the radical reaction (ESR), to characterize the products (NMR/FT-IR), and monosaccharide composition (HPAEC-PAD), to quantify carbonyl or carboxylate groups (titration, fluorescent labeling), and to determine the change in bulk properties (rheology, M w distribution, aggregation;,,;,,;, ).

On a molecular level, indirect detection of oxo-products from HO •-mediated degradation of barley BG was reported by, employing reductive tritium labeling followed by enzymatic or acid catalyzed hydrolysis and analysis by paper- and thin-layer chromatography. They confirmed the formation of new reducing ends and mid-chain oxo-groups, allowing to some extent the localization of the oxidative changes on monosaccharide units, but providing little information about their original connectivity prior to hydrolysis and no information about carboxylic acid products. In other fields of polysaccharide degradation, a recent notable accomplishment was the detection of HO •-attack in ripening fruit with a sensitive method developed by for pectin oxidation using labeling of carbonyl groups by reductive amination, enzymatic digestion and electrophoresis/HPLC with fluorescent detection ().

For a deeper understanding on a molecular level, we have recently studied BG oxidation with constitutionally isomeric oligosaccharide model compounds by means of hydrophilic interaction ultra-performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) with high resolution detection (). The direct study of polysaccharide oxidation by MS techniques, on the other hand, is impeded by the large M w and comparatively low numbers of oxidation sites along the polymer chain. In the literature, various strategies have been used to study the oxidation of other polysaccharides by MS or LC-MS, such as enzymatic treatments to remove non-oxidized portions as in the lytic polysaccharide monooxygenase (LPMO)-induced degradation of cellulose () and the HO •-mediated oxidation of starch (), functional group labeling with laminaran oxidation for higher sensitivity/ LC-separation (), and analysis of released oligomers (directly or after carbonyl reduction/ tagging) with hyaluronan () and pneumococcal type-3 polysaccharide (). Typically, harsh oxidative conditions were applied in those studies to accumulate enough products, presumably leading to secondary oxidation.

However, detection of products from Fenton-induced oxidation of cereal BG polymer specifically, and from mild oxidation of any polysaccharide in general (94% dry weight basis; 495 kg/mol; Lot 90501b) and high viscosity oat β-glucan (OBG; >94% dry weight basis; 361 kg/mol; Lot 80608c) were purchased from Megazyme (Ireland) and as 1% aqueous solutions, purified by extensive dialysis against EDTA (1 mM) and water, followed by precipitation with two volumes of ethanol, centrifugation, lyophilization, and milling. Stock solutions of 1.0% (m/v) OBG and BBG were prepared by dissolving the purified samples in water under stirring and heating in a boiling water bath for 2 h with occasional vigorous shaking. D-Laminaribiose (G 1- 3 G), 3- O-β-cellobiosyl- D-glucose (G 1- 4G 1- 3 G), and 3- O-β-cellotriosyl- D-glucose (G 1- 4G 1- 4G 1- 3 G) were purchased from Megazyme (Ireland) with a purity of >95%, as well as six cereal β-glucan molecular weight standards, lichenase (from Bacillus subtilis; EC 3.2.1.73; GH family 16) and β-glucosidase (from Aspergillus niger; EC 3.2.1.21; GH family 3).

All other chemicals were of analytical purity and used without further purification unless otherwise noted. Starting from β-(1→4)-linked glucosyl repeating unit (middle), the detectable oxidation products from hydroxyl radical (HO •)-attack at any of the C1–6 glucose carbons after reacting with oxygen (O 2) are represented as structures, depicted symbols, abbreviations, and full names.

Depending on site of attack, the resulting product is formed with or without chain scission, classified as lytic (plane arrows) or non-lytic oxidation (hashed arrows), respectively. Arabinose can by formed both directly under lytic cross-ring cleavage (C1-C2), or as secondary product of gluconic acid (from C1-oxidation) by the indicated, so-called Ruff-degradation (bottom right). *Lytic oxidations under cross-ring cleavage and loss of one or two carbons (e.g., formic acid, glycolic acid). Part 1—Methodology SPE Strategy I: Released Oligomers The first sample preparation strategy entailed running the reaction mixture directly through a non-porous graphitized carbon SPE followed by a washing step with water. As confirmed with control SPE experiments of native BBG solutions spiked with a “glucose ladder” (Glc n, n = 1–6) and TLC, neutral monosaccharides as well as polymeric materials were not retained (see Figure S1 for details). While neutral monosaccharides are known to be washed off with water, the lack of retention for the polysaccharide on graphitized carbon is unexpected in light of adsorption strength being a function of degree of polymerization (DP), at least for dextrans of DP ≤ 24 (). Observed that sodium dodecyl sulfate was retained, unless it formed micelles under certain conditions that could be washed off with water from the carbon solid phase.

One may speculate that BG as a hydrocolloid forms supramolecular aggregates that behave similarly. Hence, carbon SPE allowed for isolation of both neutral and acidic oligosaccharides released through harsh oxidation of BG (100 mM H 2O 2), which accumulated on the solid phase as expected and could be detected after elution, giving rise to similarly complex UPLC-MS patterns as observed in our previous oligo-BG oxidation study (). The most prominent detected neutral products as evident in Figure were gluco-oligosaccharides (Glc n in blue) and their respective counterparts with one oxidized hydroxyl group (oxo-Glc n in red, referring to gluco-oligomers with the carbonyl group located on any of the units/ possible glucose carbons). Isomeric mixtures of mixed-linkage oligomers were detected with random position of the β-(1→3)-linkage along the chain as suggested by MS/MS analysis (Figure S2), in accordance with a random chain scission by HO • irrespective of the linkage type. Acidic products, on the other hand, exhibited as expected low retention under the basic chromatographic conditions (; ), and could be classified into two groups: (1) oligomers with oxidized end groups, mostly gluconic acids Glc ( n−1)Glc1A from lytic C1-oxidation, and (2) oligomers from C6-oxidation, Glc ( n−1)GlcA (see Figure for MS spectrum). Products carrying carboxyl groups at both C1+C6 positions could also be detected [GlcAGlc (n−2)Glc1A], as well as the respective cross-ring cleavage product arabinose (Ara) at the reducing end (GlcAGlc (n−2)Ara; for overview of main oxidation product structures, see Figure ).

(A) UPLC-MS base peak chromatogram in the negative ionization mode using a BEH amide column and ACN/H 2O gradient (0.1% NH 3) for analysis of the released oligosaccharides from barley β-glucan (BBG) oxidation (harsh conditions) extracted by SPE and concentrated (sample preparation strategy I). Note the 6x zoom for region 6–12 min. Peaks are labeled with their respective base peak m/z and with n = number of monosaccharide units in blue for gluco-oligosaccharides Glc n and in red for oxo-Glc n species (with letter qualifier if same n occurs more than once, e.g., 3a, 3b). (B) MS spectrum of the acidic products obtained from the indicated chromatogram region in (A). Signals are labeled with m/z, with the glucose-carbon location of the carboxylic acid (C1, C6, or C1+C6), and with type of reducing end if cross-ring cleaved to arabinose (Ara).

Glc, glucose; oxo-Glc n, oligomer Glc n oxidized anywhere along the chain (new C=O); Glc 3Ara/Ery, glucotriose linked to arabinose/erythrose end group; Glc1A, gluconic acid group (C1 oxidized); GlcA, glucuronic acid group (C6 oxidized). It is the nature of random polymer chain oxidations that harsh degradation conditions are required to produce appreciable amounts of small oligomers ( n = 2–8).

Due to the low concentration of the small fraction of products that have a suitable size for UPLC-MS analysis, the sensitivity of this sample preparation strategy is intrinsically low. Indeed, no BG oxidation products could be observed in the mild oxidation (250 μM AH 2) with this direct SPE method, presumably due to the low degree of oxidation resulting in polymeric products that were too large to be retained by SPE. In fact, estimation of the relative amount of released and SPE collected BG material by phenol-sulfuric acid assay () indicates that the oligomers with n = 2–8 make up. Depiction of mixed-linkage BG and examples of possible oxidation products formed through attack of HO • at different glucose carbons, followed by sample preparation strategy II. Lichenase selectively hydrolyzed as expected the β-(1→4)-linkage of the β-(1→3)-linked glucose units. Digestion by β-glucosidase, an exo-enzyme, was found to successively cleave off glucose units, leaving oxidized oligomers behind, whereas native regions largely hydrolyzed to glucose.

Graphitized carbon SPE removed monosaccharides from the reaction mixture (acidic monomers only partially removed). Downstream (reducing) end of oligomers is on the right side (symbolic differentiation as ○–OH omitted for clarity). Figure makes it clear that the major oxidized oligomer products detected after enzymatic digestion and SPE should in theory have two features in common, irrespective of the type of product: (i) The β-(1→3)-linked reducing end unit, and (ii) The location of the oxidation-site at the non-reducing end. The former allows differentiation of mixed-linkage BG oxidation from e.g. Cellulose, which only has β-(1→4)-linkages. The resulting UPLC-MS chromatograms of the prepared sample and MS/MS of the detected BBG products are shown in Figure using basic (0.1% NH 3) and buffered (60 mM NH 4HCO 2) eluent systems for the neutral oxo-Glc n and the acidic products, respectively.

UPLC-MS/MS results in the negative mode using a BEH amide column and ACN/H 2O gradient for analysis of barley β-glucan (BBG) oxidation products (harsh conditions) after enzymatic digestion and SPE (strategy II). (A) Analysis of the neutral products using basic 0.1% NH 3 eluent additive (oxo-product peaks with carbonyl group at the non-reducing end oxoGlcGlc ( n−1) in red) and (B) their respective MS/MS spectra. The main oxo-Glc 5 species with C=O at a different position is also included (bottom spectrum). (C) Analysis of acidic products using a buffered 60 mM ammonium formate eluent additive (pH ~ 8) with glucuronic acid bearing oligomers GlcAGlc ( n−1) in red. (D) Respective MS/MS spectra of acidic products.

The MS/MS fragments are named according to the nomenclature of (cleaved linkage: C i fragments; cross-ring cleavage: A i fragments; i = 1, 2, n). *oxo-products with oxidized carbonyl group not at the non-reducing end; **Disaccharide signal from catalase material; ***Buffer salt/solvent peak. Acidic products As previously reported, the acidic products are poorly retained on the BEH amide column with the basic eluent, allowing the differentiation of acidic vs. Neutral products ().

The observed C6-oxidation products GlcAGlc ( n−1) with n = 2–5 could be separated according to size by UPLC with buffered eluent, having the expected (i) β-(1→3)-linked reducing end unit and (ii) C6-oxidation at the non-reducing end (see Figure ). This could be confirmed by negative mode MS/MS on the basis of fragmentation mechanisms proposed in our previous model oligo-BG oxidation study (nomenclature according to ).

In short, fragmentation occurs largely unidirectional in negative mode from reducing end (C n fragment) to non-reducing end (C 1 fragment), and cross-ring cleavage fragments (e.g., 0,2A i & 2,5A i; i = 1, 2, n) are observed for β-(1→4)-, but not for β-(1→3)-linkages (). Note that split peaks of m/ z 517 ( n = 3) and m/ z 679 ( n = 4) are due to anomers of the reducing end that do not coalesce under buffered eluent conditions. For the disaccharide GlcAGlc of m/ z 355 ([M-H] −), however, two peaks could be observed. One is consistent with the expected GlcAβ(1→3)Glc product ( n = 2a), whereas the other was assigned to Glcβ(1→3)GlcA ( n = 2b in Figures ) by MS/MS. In addition, gluconic acid end groups from lytic C1-oxidation were also not fully hydrolyzed by β-glucosidase, as m/ z 357 (eluting at ~6 min) could be identified as Glcβ(1→3)Glc1A by MS/MS and comparison with a prepared standard ().

In contrast, no β-(1→4)-linked disaccharides were detected. Apparently, some disaccharides with a modified glucose unit (Glc1A, GlcA) as the β-(1→3)-linked downstream (reducing) end unit are resistant to the used β-glucosidase from Aspergillus niger on the time scales used for the treatment (max.

1 day) despite the native non-reducing glucose unit, whereas the respective species with β-(1→4)-linkage seem to behave as expected and are fully hydrolyzed to the monomers (Figure ). It is the combination of β-(1→3)-linkage plus modification of the 3-O-glucosylated unit that leads to resistance, and not the β-(1→3)-linkage alone, as native isomeric β-(1→3, 1→4)-gluco- and cello-oligomers are known to be equally good substrates for this β-glucosidase regardless of the β-(1→3)-linkage being located at the reducing end, the non-reducing end, or no β-(1→3)-linkage at all [only β-(1→4)] (). Information on oxidized end units from C1-oxidation is thus partially conserved in the case of β-(1→3)-linkage (in theory ~30%) as the detected disaccharide, whereas information on β-(1→4)-linked Glc1A is mostly lost due to their observed complete hydrolysis and poor retention of the resulting Glc1A monosaccharide during SPE fractionation. Non-reducing end oxo-products Elution behavior of oxidation products with a carbonyl group at the non-reducing end oxoGlcGlc ( n−1) (red peaks in Figure ) is in accordance with the made observations in an earlier study of BG oligomer oxidation, namely eluting later than their respective native oligomer (). MS/MS confirms for all four oligomers ( n = 2–5) of this type again the expected β-(1→3)-linked reducing end unit and location of the carbonyl group at the non-reducing end (Figure ).

It cannot, however, determine the exact location of the carbonyl on the glucose unit. Those observed oxo-oligomers might thus be a mixture of products resulting from cleavage at C3/4, and from a non-lytic oxidation of a hydroxyl group (e.g., at C2 as shown in Figure ). For the latter, the subsequent removal of native glucose units from the non-reducing end by β-glucosidase is responsible for the final location of the oxidized unit as the non-reducing end (Figure; final oligomer product on the bottom right), just like for lytic C4-oxidation (bottom left).

However, from comparison of the relative amounts of the main species m/ z 501 ( n = 3) and m/ z 663 ( n = 4) before and after β-glucosidase (not shown), we suspect that most of the detected oxoGlcGlc ( n−1) signals after β-glucosidase originate directly from lytic C3/C4 oxidation (colored in red in Figure ). Mid-chain and reducing end oxo-products MS/MS for the rest of the additional isobaric peaks for n = 2–5 eluting earlier than their oxoGlcGlc ( n−1) counterpart (see Figure, m/ z 663 and 825 peaks labeled with *) revealed the carbonyl to be located on a mid-chain unit or at the reducing end, making some of them non-lytic oxo-products (see Figure S3 for MS/MS). For n = 2 ( m/ z 339* at ~5.3 min), one can imagine this oxo-product to be a Glcβ(1→3) oxoGlc species, e.g. From lytic C5-oxidation, consistent with being resistant to β-glucosidase due to the oxidized β-(1→3)-linked reducing end as was the case for GlcGlcA (Figure ). For n = 5, the isobaric oxo-Glc n peak even happens to be the main isomer (oxo- Glc 5 * peak at ~9.7 min).

MS/MS of the main oxo-Glc 5 product is consistent with the carbonyl group at a β-(1→3)-linked unit next to the reducing end, as it has the same type of fragmentation patterns (loss of H 2O and CH 2O) as determined to be diagnostic for cereal BG oxidation in our previous BG model compound study (Figure, bottom MS/MS; ). Thereby, certain oxidation sites might hinder lichenase from hydrolyzing according to its general activity, explaining why products such as oxo- Glc 5 * were not cleaved to the oxo-Glc 4 species (lichenase resistant site marked with blue arrow in proposed oxo- Glc 5 * structure in Figure ). Precedence for lichenase' inability to hydrolyze certain β-(1→4)-linkages despite their location next to a β-(1→3)-linked unit exists, namely if located at the non-reducing end (such as G 1- 3 G 1 - 4 G 1- 4G 1- 4G 1- 3 G, surprisingly resistant linkage in bold; ). However, it is unclear why the non-reducing end portion of several native glucose units were not removed by β-glucosidase in substrates such as oxo- Glc 5 *. This is subject to further investigation.

In addition, the mild oxidation (250 μM AH 2) led to different proportions of oxo-Glc n products, with oxoGlcGlc ( n−1) (C=O at non-reducing end) not being the predominant product for n = 3, 4 as is the case under the harsh degradation conditions (100 mM H 2O 2; see Figure S4 for extracted ion chromatograms). We speculate that some oxo-Glc products are more susceptible to secondary oxidation than others. A likely candidate would be 6-oxo-glucose units containing species, whose C6 aldehyde group might easily oxidize to the respective C6 carboxylate to form glucuronic acid (GlcA) under the harsh conditions, which would explain the diminished number of isobaric oxo-Glc n peaks compared with the mild oxidation. Nonetheless, the unexpected resistance toward the enzyme of some oligomers with mid-chain & reducing end carbonyl location, namely oxo-Glc n products labeled with * in Figure, can be attributed mostly to non-lytic oxidation products. Hence, products from lytic C3/C4 action can be easily distinguished from the non-lytic action by the retention time and MS/MS of the detected oxo-Glc n species after lichenase/β-glucosidase digestion and SPE. Additionally, ionization in the positive mode allowed differentiation of isobaric oxo-species due to different adduct formation preferences (e.g., [M+Na+H 2O] + for oxoGlcGlc ( n−1) vs.

[M+Na] + for other oxo-Glc n) depending on C=O-location (see Figure S5). C=O Labeling Strategy III: (New) Reducing Ends The principle of the carbonyl labeling strategy III is shown in Figure. Reductive amination tags the carbonyls formed through oxidation as well as the reducing ends, which are naturally present but also newly formed during the degradation (e.g., by lytic C3/C4 oxidation, see Figure ). This results in the same types of oligomer products after enzyme treatment as in strategy II.

The main difference lies in the conserved information of the reducing ends that get lost without labeling, as they cannot be distinguished from glucose originating from non-oxidized parts of the polymer released by β-glucosidase if untagged. As labeled glucose, on the other hand, they are retained on SPE thanks to their bigger size, higher hydrophobicity, or added charge. Carbonyl labeling by reductive amination with NaBH 3CN was conducted using two labels under different conditions, namely harsher conditions for labeling with 2-aminobenzamide (2-AB; 80°C with AcOH additive), and milder conditions for anthranilic acid (2-AA; 65°C).

This allowed for the comparison of product profile proportions to validate the results and determine the occurrence of potential side reactions. (A) Depiction of carbonyl labeling strategy III by reductive amination with anthranilic acid (2-AA; X = OH) or 2-aminobenzamide (2-AB; X = NH 2) with enzyme digestion and SPE steps analogous to strategy II (see Figure ). Negative mode UPLC-MS base peak chromatograms (BPI; basic eluent) resulting from oxidation of barley β-glucan (BBG) under a) harsh and b) mild conditions after the C=O sample preparation are shown for (B) SPE fraction 2 containing labeled reducing termini, and (C) SPE fraction 1 containing labeled oxo-products with extracted ion chromatograms (XIC) of 2AB-(oxo-Glc n). BPI of (C) were obtained after BPI-subtraction of the control sample (0.6% BBG) subjected to the same sample preparation, and the insert under b) shows the 36x zoomed XIC of m/ z 281. AH 2, ascorbic acid; Glc, glucose; Ara, arabinose; Ery, erythrose; Ar, aromatic ring; 5oxoGlc-2AB o, presumed cyclic product from 5-oxo-reducing ends (see Figure ); oxoTet-2AB o = “T,” presumed cyclic product of L- threo-tetrodialdose ( m/ z 221); 2AB-(oxo-Glc n), labeled oxo-Glc n at oxidized C=O group; HexGlc ( n−1), oligosaccharide with one sugar unit being an undefined hexose (side products, presumably from direct reduction of oxo-Glc n; see Figure S9). Detected reducing ends Both amines 2-AB and 2-AA gave labeled products with neutral and charged character, respectively, owing to their different substitution on the phenyl ring (CONH 2 vs.

CO 2H), leading to differing behavior in elution. As for the neutral and acidic products from the enzyme digested oxidized BG (strategy II), neutral 2-AB products were best analyzed with the basic 0.1% NH 3 eluent, and acidic 2-AA products with the buffered eluent ideal for carboxylic acids (60 mM NH 4HCO 2, pH~8). Both labeling procedures ran to full conversion with no alditol side products, and gave the same type of products in essentially the same relative proportions.

Thereby, SPE fraction 2 contained most products with labeled Glc and Glc 2 being the most prominent signals, followed by cross-ring cleavage products Ara and GlcAra, and faint Ery and GlcEry signals (Figure ). These labeled neutral mono- and disaccharides represent actual carbonyls from reducing ends. As for the enzyme digestion products without labels, where Glcβ(1→3)Glc1A and Glcβ(1→3)GlcA turned out to be resistant to β-glucosidase (Figure ), the labeled glucobiose was identified to be β-(1→3)-linked by comparison of MS/MS and retention time with labeled standards (see Figure S6). The mild oxidation (250 μM AH 2) gave similar signal strength of labeled reducing end Glc-2AB as the harsh oxidation (100 mM H 2O 2).

The possibility of hydrolysis occurring during the reductive amination as an explanation for the similar Glc-2AB signals in the harsh and mild oxidation could be excluded, as the corresponding 2-AA species from the milder labeling-condition resulted in the same relative proportions for Glc-2AA from AH 2 vs. Additionally, control experiments with oligomers as substrate, where no smaller labeled oligomers and hence no hydrolysis could be detected under identical conditions (not shown). Interestingly, however, the two oxidation conditions exhibited different proportions regarding the more diagnostic cross-ring cleavage products such as Ara- and Ery-2AB (Figure ).

This contrasts results from, who also observed arabinose from OBG and BBG oxidation with H 2O 2 (10–70 mM) by HPAEC-PAD after hydrolysis with lichenase/β-glucosidase, but not for samples oxidized with AH 2 (10–70 mM). This underlines the sensitivity of our method with 2-AB labeling and UPLC-MS analysis that allowed detection of cross-ring cleavage products even from a very mild oxidation. Lytic C5-oxidation Apart from the described products predominantly detected in SPE fraction 2, fraction 1 contained one major product with m/ z 281.11 (in the case of 2-AB), which is smaller by 18 Da = H 2O compared to Glc-2AB ( m/ z 299.13; see Figure ). Two peaks of m/ z 443.17 were also found, which are 162 Da = one anhydroglucose unit larger than m/ z 281.11. These products were also detected in fraction 1 with 2-AA as the label ( m/ z 282.10 and 444.15). The two species could be assigned to labeled 5oxoGlc and Glcβ(1→3) 5oxoGlc, if instead of tagging the two carbonyls (reducing end (C1) and C5) with two amino-labels, only one label does the job for both positions, with epimers from the reduction explaining the double peaks for the disaccharide (for a rationalization of epimer proportions with regard to linkage type, see Figure S7).

This is possible, as the spatial distance of carbonyls in this δ-keto-aldehyde allows for a unique mode of labeling by reductive amination: a second, intramolecular reductive amination step resulting in cyclization to form a piperidine derivative (see Figure ). With intramolecular reactions being often faster than intermolecular ones, especially when forming 6-membered rings, no double-tagged species 2AB- 5oxoGlc-2AB could be detected. Lytic C5-oxidation forming a new terminus 5oxoGlc under loss of the non-reducing end-portion of the β-glucan polymer, and its reductive amination with 2-aminobenzamide (2-AB) (C=O labeling strategy III).

The shown mechanism explains the cyclic end products 5oxoGlc-2AB o and Glcβ(1→3) 5oxoGlc-2AB o detected in SPE fraction 1), which involves two reductive amination steps to form the piperidine derivative (first intermolecular tagging of the reducing end carbonyl (C1), then intramolecular with the C5-ketone). Glc, Glucose; [O], oxidation (Fenton-induced); [H], reduction (hydride from NaBH 3CN). Precedence for such a double reductive amination exists for 5oxoGlc and other dicarbonyl sugars as reported by, who exploited the ring-formation for the synthesis of aza-sugar building blocks.

Since no other constellation of carbonyls in primary oxidation products with 6 carbons are expected to result in such a cyclization, and dehydration side products with the same m/ z could be excluded, these products in fraction 1 could be ascribed to lytic C5-oxidation (see Figures S8A,B for MS/MS). We also found a peak of m/ z 221.09 that could be assigned to labeled L- threo-tetrodialdose ( oxoTet-2AB o in Figure ), which was also observed by in their Glc irradiation study under O 2.

It is like 5oxoGlc a product induced by HO •-mediated H-atom abstraction on C5, but undergoing an alternative pathway with additional cleavage of the C4–C5 bond, resulting in the loss of 2 carbons (for the reaction scheme, see Figure ). The successful detection of lytic C5-oxidation products is fortunate, as 5oxoGlc is a clear marker for BG oxidation, in contrast to reducing ends that are already present before oxidation. Interestingly, lytic C5-oxidation has been largely overlooked as a pathway for direct cleavage of glycosidic linkages in HO •-mediated polysaccharide oxidation under formation of 5oxoGlc reducing ends (; ). This recent lack of reporting on C5-oxidation is especially puzzling in light of HO •-attack being somewhat favored at the C5-position as determined by EPR studies () and pulsed γ-radiolysis of cellobiose ().

Proposed for the intermediate C5-oxonium ion (R 2C=O +R′; formed after HO •-attack at C5 and O 2/ O 2 • - addition/elimination), to undergo a tautomerisation that leads to labile Glc1A esters, instead of simply liberating 5oxoGlc through hydrolysis. However, such a direct, thermal [1,3]-hydride shift to Glc1A-esters is mechanistically not possible ().

We, on the other hand, have first hypothesized the occurrence of 5oxoGlc species as the direct result of lytic C5-oxidation under Fenton-conditions in our BG model compound oxidation study (), but were also not able to unambiguously identify it by direct-injection UPLC-MS/MS due to complex isobaric product mixtures and sensitivity issues. In this study, however, with the oxidation of polymeric BG, C=O labeling, and enzymatic treatments that largely eliminate interferences and simplify product profiles by focusing only on carbonyls, we present for the first time evidence for a direct lytic C5-oxidation in BG under formation of 5-oxo-reducing ends. Non-reducing end and mid-chain oxo-products The other expected type of products is labeled oxoGlcGlc ( n−1) or oxo-Glc n in general.

They could be found in SPE fraction 1 as complex mixtures [see extracted ion chromatograms (XIC) of m/ z 459, 621, and 783 in Figure ]. As for the labeled cyclic 5oxoGlc, each oxo-product should lead to two epimers after the reductive amination due to the intermediate imin that can be reduced from two different faces (with the exception of C6-aldehyde, which leads to only one product). Thereby, each C=O location is expected to lead to different proportions of those epimers due to the respective steric situations (1:1 to up to 95% selectivity; ). Naturally, this complicates the product mixtures, which explains the observed higher number of peaks for these labeled oxo-Glc n products than the respective non-labeled species from strategy II (Figure ). While the complexity of these isomeric product mixtures was expected, MS/MS did not suffice to unambiguously assign the peaks to specific products due to their similar fragmentation spectra (see Figures S8C,D for XIC and average MS/MS spectra). Next to the acidic products also observed as in strategy II (e.g., m/ z 357, 517) and the mentioned labeled oxo-species, broad peaks with masses isobaric to native glucosyl-oligomers were detected in SPE fraction 1 ( m/ z 503, 665; see Figure ). They are much more prominent than the labeled oxo-Glc n products discussed above.

It is possible that these oxo-products are partially lost by a side reaction of the reductive amination, forming the prominent unlabeled oligomers. The more sterically hindered keto-groups (compared to the reducing end aldehyde/hemiacetal) could slow down the initial imin-formation step, making a competing direct reduction to glucosyl epimers with NaBH 3CN possible, which is known to reduce carbonyls at pH 3–4 as well ().

Detection of signals isobaric to Glc n in elevated levels compared to the control, but with different retention times than the native BG oligomers from lichenase digestion, speaks for this hypothesis (see Figure S9). Recently, attempted to circumvent such a side reaction when labeling 4-oxo-Glc bearing oligomers produced from LPMO oxidation of cellulose by using NaBH(OAc) 3 as reducing agent.

However, they also observed direct reduction of the carbonyl. As an alternative, the authors investigated the direct non-reductive labeling to form the Schiff base imin selectively at the reducing end. However, the low degree of conversion to the imin and the reversibility of the labeling reaction impeded an effective quantification. Other solutions to optimize the labeling of non-reducing end/ mid-chain oxo-products are part of future studies. Nevertheless, the described method is suitable for labeling of the reducing end (including 5oxoGlc) with full conversion, conserving the information of neutral termini in BG oxidation. CO 2H Labeling Strategy IV: Lytic and Non-lytic Acidic Products Carboxylic acid labeling was accomplished by amidation via carbodiimide activation of CO 2H using aniline (PhNH 2) both as label and as pH buffer and was based on a procedure by for energy metabolism analysis (see Figure ).

Depiction of (A) carboxylic acid labeling strategy IV by EDC-mediated amidation of oxidized β-glucan with aniline (PhNH 2) and enzyme digestion/SPE steps analogous to strategy II (see Figure ). (B) The resulting base peak ion chromatograms (BPI) from oxidation of barley β-glucan (BBG) under (a) harsh and (b) mild conditions after the CO 2H sample preparation (basic eluent). The peaks are labeled with their base peak ion m/ z and corresponding structure (chemical structures on the right-hand side). The insert under (b) shows 100x zoom of overlaid extracted ion chromatograms of m/ z 270 and 432.

Glc, glucose; Glc1A, gluconic acid; GlcA, glucuronic acid; Ara1A, arabinonic acid; Ery1A, erythronic acid; AH 2, ascorbic acid; Ph, phenyl; -NHPh or PhNH-, aniline amide (anilide) of acid species; EDC, ethyl-3-(3-dimethylaminopropyl)carbodiimide. Detected products as anilides The expected oxidized gluconic acid end units from lytic C1 oxidation of BG degradation was the major product using the CO 2H labeling sample preparation strategy IV (see Figure ). They were detected as Glc1A-anilides (Glc1A-NHPh, amide of aniline) after enzyme treatment/SPE, conserving information about lytic C1-oxidation that is otherwise (partially) lost in the enzymatic treatment without carboxylate tagging (strategy II), as the necessary SPE step only retains monosaccharides if labeled. In addition, gluconic acid was also detected as Glcβ(1→3)Glc1A-NHPh amide, apparently also resistant to β-glucosidase as the free acid GlcGlc1A itself (see Figure S10 for MS/MS spectra).

Other oxidized termini that were conserved through labeling were the cross-ring cleavage products, which were observed in minor amounts as arabinonic and erythronic amides (Glc)Ara1A- and GlcEry1A-NHPh, respectively. Interestingly, the peak of Ara1A-NHPh was much smaller compared to Glc1A-NHPh (Figure ) than the respective peak of C=O labeled Ara-2AB compared to Glc-2AB termini (strategy III; Figure ).

In addition, glucuronic acid products PhNH-GlcAGlc ( n−1) with n = 1–3 were also detected, representing non-lytic C6 oxidation. However, in contrast to strategy II, the labeled uronic acid species with n >3 were only detected in traces, possibly due to the labeled GlcA unit rendering some linkages in their proximity not susceptible to lichenase anymore, resulting in larger oligomers ( n >8) that are not eluted from SPE and/or the UPLC-column under the chosen conditions.

Hence, strategy IV is most useful for retaining the otherwise lost information on acidic termini. CO 2H-labeling efficiency The reaction of carboxylates with PhNH 2 did not lead to complete tagging, which was also observed for test reactions with (Glc)Glc1A and GlcA standards (not shown). This seems to be intrinsic to carbodiimide activated carboxylic acids in aqueous polysaccharide solutions, with, for example, typical degrees of substitutions of 0.1–0.3 for alginates (). A reason might lie in lactone formation after activation of the carboxylate instead of reaction with the amine (e.g., Glc1A + EDC→→ Glc1A-δ-lactone). Those lactones are not readily hydrolyzed back to the carboxylate at pH 4.5, and apparently do not (fully) open up with PhNH 2 to form the amide. However, known side reactions such as rearrangement of EDC-activated carboxylates to stable N-acyl ureas () were not observed.

Attempts for optimization, e.g., by using a different amine such as 2-AB, addition of N-hydroxysuccinimide (NHS), changing pH, testing alternative coupling reagents (e.g., PyBOP) and anhydrous conditions (in DMF) were not successful and lead to lower yields in test reactions with acid standards (not shown). Stronghold Torrent Pirate Bay News. Nevertheless, under our reaction conditions, the labeling of carboxylates in BG proved to be robust with constant conversions, therefore allowing comparison of different samples. Comparison, Scope, and Limitations of the Four Strategies The presented four complementary sample preparation strategies (I–IV) used to access different information about BG oxidation are summarized in Table, revealing similarities and differences of the various approaches (see Figure S11 for the extended version with product structures, symbols, names, and abbreviations; and Table S1 for lost vs. Preserved information). The comparison makes it clear that SPE strategy I provides limited information and low sensitivity since it is confined to a narrow window of product size ( n = 2–8. Information on certain termini and their formation, namely reducing ends (Glc, Ara, Ery), lytic C1- and C5-oxidation (Glc1A & 5oxoGlc), though, are mostly lost due to hydrolysis to monomers and their lack of retention on SPE. Strategies III and IV take care of these neutral and acidic termini by C=O and CO 2H labeling, respectively, while increasing sensitivity and making in principle an additional mode of detection possible, namely by UV/fluorescence.

This has potential as it would allow in the future the quantification of certain products by taking advantage of the equal response factor regardless of tagged molecule. Due to side reactions for C=O labeling strategy III and presumably incomplete hydrolysis of CO 2H labeled species by lichenase in strategy IV, strategy II is still best suited to assess lytic C3/C4 and non-lytic C6 oxidation [ via oxoGlcGlc ( n−1) & GlcAGlc ( n−1), respectively]. However, due to the extensive washing/precipitation steps, none of the sample preparation strategies in their current form capture small oxidation fragments that are not covalently bound to the polymer (C n aldehydes/acids with number of carbons n.

Comparison of UPLC-MS base-peak ion (BPI) chromatograms of polysaccharide oxidation under harsh conditions (100 mM H 2O 2, 50 μM FeSO 4) after lichenase and β-glucosidase treatment/SPE for (A) only barley β-glucan (BBG) (0.6%; as in Figure ), (B) BBG and corn starch (each 0.6%), and (C) only corn starch (0.6%). (D) UPLC-MS of malto-oligosaccharide standards up to maltohexaose ( n = 6).

Glc, glucose; Glc1A, gluconic acid; GlcA, glucuronic acid; oxoGlc, oxo-glucose; oxo-Glc n, gluco-oligomer with oxidized carbonyl group anywhere along the chain; oxoGlcGlc ( n−1), gluco-oligomer with oxidized carbonyl at non-reducing end; L+βG/SPE, after lichenase and β-glucosidase treatment, followed by fractionation by SPE; *, oxo-Glc n with oxidized C=O not at the non-reducing end; α, malto-oligosaccharides [Glc n with all α-(1→4)]. Since no standards are commercially available, the respective ratios in Table refer to signal and not molar ratios, which does not give absolute values but still allows comparing the ratios with each other (Table, oat/barley).

The results were mixed, with oxoGlcGlc ( n−1) species having ( n = 3)/( n = 4) ratios for OBG and BBG that were equal, whereas the GlcAGlc ( n−1) ratios with 2.0 and 2.9, respectively, differed depending on BG source. Analyzing the possibilities of formation systematically for these two types of products revealed the reason behind their different product ratio pattern. OxoGlcGlc (n-1) ratios The oxo-products originate mainly from lytic C3/C4 cleavage, meaning they can lose glucose units that are part of the cellotriosyl/cellotetraosyl fine structure, such that a resulting oxoGlcGlc ( n−1) species with n = 3 not necessarily originates from a cellotriosyl substructure, but also from DP ≥ 4. Assuming equal probability of oxidative cleavage for each glycosidic linkage regardless of position, a cellotriosyl subunit can be oxidized to give—after lichenase and β-glucosidase— oxoGlcGlc ( n−1) with n = 2, 3, and 4 in a 1:1:1 ratio, whereas oxidation of a cellotetraosyl unit leads to oxo-products with n = 2, 3, 4, and 5 in a 1:1:1:1 ratio (see Figure ). The original DP3/DP4 ratio is thus not reflected in those lytic oxo-products, as both cellotriosyl and cellotetraosyl subunits can produce both oxo-Glc 3 and oxo-Glc 4 equally, explaining the identical ratios for OBG and BBG. Influence of fine structure differences (DP3/DP4) of oat and barley β-glucan on the ratios of oligomeric oxidation products.

Explanation for the resulting oligomeric oxidation product ratios ( n = 3)/( n = 4) from (A) lytic C3/C4-oxidation and (B) non-lytic C6-oxidation, both after lichenase and β-glucosidase treatment/SPE (strategy II). OxoGlcGlc ( n−1), gluco-oligomer with oxidized carbonyl at non-reducing end; GlcAGlc ( n−1), oligomer with glucuronic acid (C6-oxidized) unit at non-reducing end (unless for n = 2, where also at reducing end); BBG, barley β-glucan; OBG, oat β-glucan; DP3/DP4, molar ratio of cellotriosyl to cellotetraosyl units in β-glucan (defines fine structure). GlcAGlc (n-1) ratios The ratios of glucuronic acid species, on the other hand, are according to the native DP3/DP4 ratio (BBG>OBG; Table ). Since introduction of a carboxylic acid at C6 is a non-lytic process, the possibilities of formation are different from the lytic oxo-products above.

Here, a cellotriosyl subunit can lead to GlcAGlc ( n−1) with n = 2a, 2b, and 3 after enzymatic treatment, while for cellotetraosyl, glucuronic acids with n = 2a, 2b, 3, and 4 can be formed [2a and 2b species refer to β-(1→3)-linked GlcAGlc and GlcGlcA, respectively; see Figures ]. Hence, GlcAGlc ( n−1) with n = 4 cannot be formed from cellotriosyl units, whereas the n = 3 species can be formed from both cellotriosyl and cellotetraosyl units, making the C6-oxidation products depend on the DP3/DP4 ratio of the original starting material (see Figure ). Β-(1→4)/β-(1→3) vs. Monosaccharide/disaccharide termini As for the linkage type ratios, the resistance toward β-glucosidase hydrolysis of β-(1→3)-linked disaccharide products with labeled end units is used as an advantage for additional information on the randomness of the degradation. Thereby, aldonic acid CO 2H labeling provides information about the selectivity on the lytic C1-oxidation process, and reducing end C=O labeling about the lytic C3/C4-oxidation process of the previously neighboring sugar unit (primary mode for releasing new reducing ends; Table, bottom row).

The ratio of labeled mono- to disaccharide should directly reflect the β-(1→4)/β-(1→3) linkage ratio of the BG material, unless one linkage type is preferably cleaved, skewing the ratio toward one side or the other. The observed ratios of labeled mono- to disaccharide are 7 and 4% higher for OBG in the C=O- and CO 2H-labeling, respectively, and in good agreement with the calculated ~5% higher β-(1→4)/β-(1→3) ratios for the used BG materials from oat vs. Barley (see Table ). Implications on HO •-attack preferences As discussed above, no significant differences could be found between the calculated theoretical DP3/DP4 ratios (based on the assumption of equal attack) and the experimental results regarding the lytic C3/C4- and non-lytic C6-oxidation oligo-products, as well as the linkage-ratio reflected in Glc1A and Glc reducing termini (labeled mono- vs. This implies equal probability of attack on each unit along the mixed-linkage chain for lytic as well as non-lytic processes within the accuracy of the measurements.

Yet, estimated minor differences of ≤ 10% in preferential attack depending on the repeating unit's neighborhood cannot be excluded, as they might be masked by the experimental uncertainty and the averaging nature of some of these ratios. Nevertheless, the results corroborate earlier findings from a study using constitutionally isomeric BG model compounds (Glc 4) with and without β-(1→3)-linkage, exhibiting no significant differences in their half-lives for the oxidative degradation (). This is contrary to previous claims of preferred β-(1→3)-cleavage in a BBG oxidation study, where the linkage types were assessed by 13C-NMR analysis during the degradation (). However, the degree of oxidation was much higher in the NMR study as evident by the M w of BG being too low to be detectable by dynamic light scattering already after 2 h (85°C, 100 mM H 2O 2). The reported faster disappearance of the C3-glycosidic linkage signal could therefore be the result of late stages in the degradation that are rarely reached under physiological or food processing conditions. Hence, we conclude that the kinetic differences in the cleavage of β-(1→3)- vs.

Β-(1→4)-glycosidic bonds induced by the Fenton reaction average out to be either non-existent or too small to have any significant effect on the formation of products from lytic oxidation at C1 or C3/C4. Relationship of BG Molar Mass with Content of Termini from Lytic Oxidations The number average molar masses ( M n) of the OBG and BBG starting materials as well as their values after 24 h oxidation under the two oxidation conditions (a) 100 mM H 2O 2 and (b) 250 μM AH 2 were determined by SEC. They are listed in Table, showing that the harsher H 2O 2 conditions led to a ~20x reduction of M n, about 8x more drastic regarding loss in polymer length than the mild AH 2 conditions. Using the molar masses, the theoretical reducing end contents can be calculated for each reaction (proportional to 1/ M n, assuming all ends are in the form of Glc), which were set relative to the reducing end group (REG) content of the mild condition (AH 2). This can then be compared with the relative UPLC-MS signals of termini detected with enzymes/SPE strategy II for lytic C3/C4-oxidation, C=O labeling strategy III for (oxo-)reducing ends, and CO 2H labeling strategy IV for aldonic acid termini (see Table ). Comparison of reaction conditions If oxidation under the harsh (H 2O 2) and mild (AH 2) conditions only differ in the extent of oxidation, but have the same product profiles, meaning that both are based on the same mechanism with the same probabilities of initial attack regarding the different glucose carbons, then the relative amounts of detected products should be similar for each oxidation type.

This was indeed the case for the termini oxoGlcGlc ( n−1), Glc1A, and 5oxoGlc from lytic C3/C4-, C1-, and C5-oxidation, respectively (H 2O 2/AH 2 ratios in the range of 8.3–12:1 for OBG; Table ). The relative values also roughly matched the expected theoretical content of termini calculated from 1/ M n (7.7:1). The experimental ratio for Glc reducing ends as observed by C=O labeling, however, is with 1.2:1 much lower than expected. The same tendencies with detected Glc reducing ends. EFSA Panel on dietetic products, nutrition and allergies (NDA).

Scientific opinion on the substantiation of health claims related to beta-glucans from oats and barley and maintenance of normal blood LDL-cholesterol concentrations (ID 1236, 1299), increase in satiety leading to a reduction in energy intake (ID 851, 852), reduction of post-prandial glycaemic responses (ID 821, 824), and “digestive function” (ID 850) pursuant to Article 13(1) of Regulation (EC) No 1924/2006. Doi: 10.2903/j.efsa.2011.2207.