Methods of the site-selective solid phase synthesis of peptide-derived Amadori products

Two procedures of glycated peptides’ synthesis have been developed. The first method involves reductive alkylation of the ε-amino groups of lysine with 2,3:4,5-di-O-isopropylidene-β-d-arabino-hexos-2-ulo-2,6-pyranose in the presence of sodium cyanoborohydride on solid support. The second one uses a new fully protected lysine derivative, which is a building block designed for direct introduction of the glycated lysine moiety into a peptide, according to the standard solid phase synthesis protocol. The applicability of the proposed methods for the synthesis of peptide-derived Amadori products is discussed. The structure of the synthesized glycated peptides was confirmed by high-resolution mass spectrometry and enzymatic hydrolysis. Circular dichroism studies, performed in water solution, revealed that the formation of the Amadori rearrangement product in the lysine side chain does not influence significantly the conformational preferences of the peptides studied. However, when the solvent was changed to trifluoroethanol, the glycated peptides preferred β-turn conformation.     

Enzymatic repair of Amadori products

Protein deglycation, a new form of protein repair, involves several enzymes. Fructosamine-3-kinase (FN3K), an enzyme found in mammals and birds, phosphorylates fructosamines on the third carbon of their sugar moiety, making them unstable and causing them to detach from proteins. This enzyme acts particularly well on fructose-epsilon-lysine, both in free form and in the accessible regions of proteins. Mice deficient in FN3K accumulate protein-bound fructosamines and free fructoselysine, indicating that the deglycation mechanism initiated by FN3K is operative in vivo. Mammals and birds also have an enzyme designated ‘FN3K-related protein’ (FN3KRP), which shares ≈65% sequence identity with FN3K. Unlike FN3K, FN3KRP does not phosphorylate fructosamines, but acts on ribulosamines and erythrulosamines. As with FN3K, the third carbon is phosphorylated and this leads to destabilization of the ketoamines. Experiments with intact erythrocytes indicate that FN3KRP is also a protein-repair enzyme. Its physiological substrates are most likely formed from ribose 5-phosphate and erythrose 4-phosphate, which give rise to ketoamine 5- or 4-phosphates. The latter are dephosphorylated by ‘low-molecular-weight protein-tyrosine-phosphatase-A’ (LMW-PTP-A) before FN3KRP transfers a phosphate on the third carbon. The specificity of FN3K homologues present in plants and bacteria is similar to that of mammalian FN3KRP, suggesting that deglycation of ribulosamines and/or erythrulosamines is an ancient mechanism. Mammalian cells contain also a phosphatase acting on fructosamine 6-phosphates, which result from the reaction of proteins with glucose 6-phosphate.     

Synthesis of isotopically labeled delta2-isopentenylpyrophosphate

The high yield chemical conversion of acetate to Δ²-isopentenylpyrophosphate is described. This procedure, plus the availability of high specific activity isotopic labels in acetate, allows the efficient synthesis of large quantities of highly labeled Δ²-isopentenylpyrophosphate.     

Resonance Raman studies of isotopically labeled chloroperoxidase

Chloroperoxidase (CPO) and cytochrome P450cam have been shown by several techniques to have similar active site properties. Recent resonance Raman investigations using isotopically enriched 34S-labeled samples have demonstrated thiolate ligation in the P450cam system. We report here on a number of parallel studies involving CPO. On the basis of isotopic labeling (34S, 13CO), we assign the Fe-S and Fe-CO stretching frequencies of CPO at 347 (-vFe-S) and 488 cm-1 (-vFe-CO). The differences of the -vFe-S and -vFe-CO in CPO and P450cam may suggest subtle differences in the thiolate binding in the two systems.     

Preparative-scale isolation of isotopically labeled amino acids

Over 10 g of individual ²H, ¹⁵N-labeled amino acids was resolved and recovered on a laboratory-scale ion-exchange system from a crude bacterial protein hydrolyzate derived from 20 g of lyophilized cells. The 17 amino acids (cystine was not isolated) were recovered containing less than 1.0% of other contaminating amino acids except for proline (4.0%). The aromatic and basic amino acids were isolated on a dual-column carrier displacement system (390-ml resin bed volume) while most of the neutral and acidic amino acids were separated on a pyrazolium chloride elution system (560-ml resin bed volume). The two remaining overlapping pairs were resolved on small carrier displacement columns. In addition, the overlapping fractions from adjacent peaks of the pyrazolium chloride elution system represent only 3.5% (0.37 g) of the total sample.     

An efficient and convenient synthesis of deuterium-labelled seminolipid isotopomers and their ESI-MS characterization

Seminolipids 1a and 1b and galactosylalkylacylglycerols 2a and 2b, labelled with deuterium on the alkyl or acyl chain, respectively, were obtained isotopically and chemically pure through a straightforward synthesis from protected glycidyl galactoside 3 in an overall 22% yield. The identity and purity of compounds was ascertained by NMR spectroscopy and ESI mass spectrometry analysis. These labelled compounds are important as internal standards for quantification of these lipids by mass spectrometry, and they could also be used in metabolic studies in in vitro and even in vivo systems. Extension of the procedure could provide a route for the preparation of isotopomers of other compounds of the same general class.     

Preparation and storage of isotopically labeled reduced nicotinamide adenine dinucleotide

A method for obtaining highly purified NADH in a dry, solid, and stable form is described. The method involves improvements of the ion-exchange and reversed-phase chromatographic procedures of C. J. Newton and S. M. Faynor, and D. B. Northrop (Anal. Biochem., 1983, 132, 50-53). The necessary time to prepare pure NADH has been reduced to a few hours. The final product, obtained by drying the nucleotide from absolute ethanol, shows no detectable decomposition either during the drying procedure or during storage under nitrogen gas at -20 degrees C for several months. Using dry product prepared from fixed volumes of ethanolic solution, standardized solutions of known amounts of the highly purified and stored NADH can be obtained in a few seconds.     

Biosynthesis of isotopically labeled gramicidins and tyrocidins by Bacillus brevis

The three-dimensional structure of bilayer-associated gramicidin A is available from a structural data base. This and related peptides are, therefore, ideal model compounds to use during the implementation and development of new NMR techniques for the structural investigations of membrane proteins. As these methods rely on the isotopic labelling of single, selected or all sites, we have, investigated and optimised biochemical protocols using different strains of the Gram-positive bacterium Bacillus brevis. With newly developed schemes for isotopic labelling large amounts of gramicidin and tyrocidin enriched with stable isotopes such as ¹⁵N or ¹⁵N/¹³C have been obtained at low cost. A variety of analytical and spectroscopic techniques, including HPLC, mass spectrometry and NMR spectroscopy are used to characterise the resulting products.     

The metabolic fate of isotopically labeled trimethylamine-N-oxide (TMAO) in humans

Trimethylamine-N-oxide (TMAO) is associated with chronic disease risk. However, little is known about the metabolic fate of dietary TMAO. This study sought to quantitatively elucidate the metabolic fate of orally consumed TMAO in humans. As part of a crossover feeding study, healthy young men (n=40) consumed 50-mg deuterium-labeled methyl d9-TMAO (d9-TMAO), and enrichments of TMAO and its derivatives were measured in blood for 6 h, urine and stool, as well as skeletal muscle in a subset of men (n=6). Plasma d9-TMAO was detected as early as 15 min, increased until 1 h and remained elevated through the 6-h period. TMAO exhibited an estimated turnover time of 5.3 h, and ~96% of the dose was eliminated in urine by 24 h, mainly as d9-TMAO. No d9-TMAO was detected in feces. Notably, d9-TMAO and d9-trimethylamine were detected in skeletal muscle (n=6) at 6 h, and the enrichment ratio of d9-TMAO to d9-trimethylamine was influenced by a genetic variant in flavin-containing monooxygenase isoform 3 (FMO3 G472A). These results suggest that the absorption of orally consumed TMAO is near complete and does not require processing by gut microbes. TMAO exhibits fast turnover in the circulation with the majority being eliminated in urine within 24 h. A small portion of the dose, however, is taken up by extrahepatic tissue in a manner that appears to be under the influence of FMO3 G472A polymorphism. This trial was registered at clinicaltrials.gov as NCT02558673.     

Microalgae as a source of stable isotopically labeled compounds

Microalgae have the ability to convert inorganic compounds into organic compounds. When they are cultured in the presence of stable (non-radioactive) isotopes (i.e.¹³CO₂,¹⁵NO ₃ ^– ,²H₂O) their biomass becomes labeled with the stable isotopes, and a variety of stable isotopically-labeled compounds can be extracted and purified from that biomass.Two applications for stable isotopically-labeled compounds are as cell culture nutrients and as breath test diagnostics. Bacteria that are cultured with labeled nutrients will produce bacterial products that are labeled with stable isotopes. The presence of these isotopes in the bacterial products, along with recent developments in NMR technology, greatly reduces the time and effort required to determine the three-dimensional structure of macromolecules and the interaction of proteins with ligands. As breath test diagnostics, compounds labeled with¹³C are used to measure the metabolism of particular organs and thus diagnose various disease conditions. These tests are based on the principle that a particular compound is metabolized primarily by a single organ, and when that compound is labeled with¹³C, the appearance of¹³CO₂ in exhaled breath provides information about the metabolic activity of the target organ. Tests of this type are simple to perform, non-invasive, and less expensive than many conventional diagnostic procedures.The commercialization of stable isotopically labeled compounds requires that these compounds be produced in a cost-effective manner. Our approach is to identify microalgal overproducers of the desired compounds, maximize the product content of those organisms, and purify the resulting products.     

Isolation and quantitation of isotopically labeled amino acids from biological samples

To face the problem of simultaneous isolation and quantitation of isotopically labeled amino acids in biological samples, two semi-preparative chromatographic methods were developed. One method was especially designed to isolate radioactively labeled amino acids for which we used derivatization with the fluorophore o-phtaaldialdehyde (OPA), which is known to be easy and reliable. Isolation of amino acids labeled with stable isotopes required another approach as we wanted to use isotope ratio mass spectroscopy (IRMS), which can only be performed on pure, non-derivatized amino acids. Becuase the OPA probe cannot be removed after isolation of the derivative, we used 9-fluorenylmethylchloroformate (FMOC) instead. This probe is linked to an amino acid via a peptide bond which can easily be broken byb gas-phase acid hydrolysis (103% recovery after 5 h at 150°C: S.D = 3.5%, n = 14). Run time (injection to injection) was 60 min for the OPA method and 75 min for the FMOC method. Both fluorescence and UV absorbance detection can be employed. The coefficient of variation (C.V.) for peak area measurement was below 2% for most OPA amino acids and below 3% for most FMOC amino acids. At maximum, a total of 1000 μl could be injcted, representing approximately 200 μl of deproteinized plasma. The methods were linear up to injection of 0.5 μmol of all amino acids (OPA: r²=0.995−0.999; FMOC: r²=0.992−0.999). The C.V. of the IRMS measurement within the range which can be isolated maximally in one chromatographic run (50–500 nmol), was less than 3% above 100 mmol, indicating that chromatographic isolation fulfils the needs of the IRMS determination. The resulting methods are suitable for the isolation and quantitation of micromolar amounts of labeled amino acids from biological samples.     

Solid-phase synthesis of glucose-derived Amadori peptides.

Nonenzymatic glycosylation or glycation of amino groups in peptides and proteins by D-glucose is a universal reaction with important implications for the pathogenesis of many diseases including diabetes mellitus. Here a general approach is reported to synthesize site specifically glucose-derived N-glycated peptides. Therefore, model peptides H-AKASASFL-NH(2), H-AKASADFL-NH(2), H-ASKASKFL-NH(2), and H-AKDSASFL-NH(2) were synthesized on solid phase by Fmoc chemistry using Fmoc-Lys(4-methyltrityl)-OH in positions 2 or 3 to be glycated. After completion of the synthesis, the acid labile 4-methyltrityl-group was cleaved with 1% TFA in DCM and the free amino groups were glycated by the Lobry de Bruyn reaction using 2,3:4,5-di-O-isopropylidene-aldehydo-beta-D-arabino-hexos-2-ulo-2,6-pyranose on solid phase. After TFA treatment, the crude peptides were obtained in high yields and purities above 80%. Minor by-products were well separated on reversed-phase HPLC.     

The Amadori rearrangement as key reaction for the synthesis of neoglycoconjugates

The Amadori rearrangement was introduced as a key step for the conjugation of carbohydrate moieties with suitable amines such as aliphatic amines and amino acid derivatives. The rearrangement products were further transformed into the corresponding 1-N,2-O cyclic carbamates employing triphosgene to obtain anomerically stable glycoconjugates. The reaction conditions were probed on a model substrate, 3,5-di-O-benzyl-alpha,beta-d-glucofuranose and further applied to d-glycero-d-gulo-heptose, which gave 'd-gluco-conjugates' in the alpha-anomeric form exclusively in high isolated yields.     

Plasma appearance of labeled beta-carotene, lutein, and retinol in humans after consumption of isotopically labeled kale

The bioavailability of carotenoids from kale was investigated by labeling nutrients in kale with 13C, feeding the kale to seven adult volunteers, and analyzing serial plasma samples for labeled lutein, beta-carotene, and retinol. Ingested doses of labeled carotenoids were 34 micromol for beta-carotene and 33 micromol for lutein. Peak plasma concentrations, areas under the plasma concentration-time curves (AUCs), and percentages of dose recovered at peak plasma concentrations were calculated. Average peak plasma concentrations were 0.38, 0.068, and 0.079 microM for [13C]lutein, [13C]beta-carotene, and [13C]retinol, respectively. Average AUC values (over 28 days) were 42.8, 13.6, 13.2 microM h for [13C]lutein, [13C]beta-carotene, and [13C]retinol, respectively. Percentages of dose recovered at peak plasma concentrations were 3.6, 0.7, and 0.7% for [13C]lutein, [13C]beta-carotene, and [13C]retinol, respectively. A positive relationship was observed between baseline plasma retinol levels and [13C]retinol plasma response. It is possible that this relationship was mediated either through some aspect of beta-carotene absorption or via the common pathways of metabolism for postdose and endogenous retinoid.