Quelques enzymes sont connus pour déglyquer les kétoamines résultants de la réaction de Maillard entre des sucres et des amines primaires. Il a été démontré qu’Escherichia coli possède un opéron afin de métaboliser la fructoselysine. La fructoselysine 6-kinase, de la famille des PfkB, initie le processus de déglycation permettant l’utilisation ultérieure du glucose-6-P par la bactérie. La résolution de la structure de la FL6K par cristallographie et diffraction des rayons X a permis d’identifier son site actif en présence d’ATP, d’ADP et d’AMP-PNP. La modélisation de la fructoselysine au site actif de la kinase a permis d’identifier des résidus pouvant être importants pour sa liaison et son mécanisme enzymatique. De plus, les résultats de cinétique suggèrent que le mécanisme utilisé par la FL6K semble passer par un état ternaire de type SN2. Des modifications structurales à la FL6K pourraient permettre d’augmenter la taille des substrats afin de permettre ultimement la déglycation de protéines. / Some enzymes have been found to deglycate the products of the Maillard reaction between sugars and primary amines: ketoamines. An operon is found in Escherichia coli that allows the growth on fructoselysine media. The deglycation process is done by a kinase and a “deglycase”. The fructoselysine 6-kinase, a member of the PfkB family, phosphorylates its substrate on the sixth carbon to initiate the metabolism of fructoselysine. Here are presented x-ray crystallography structures obtained for the fructoselysine 6-kinase in its native form and bound with ATP, ADP and AMP-PNP. The active site of the kinase has been determined, and modelisation of fructoselysine allowed identification of some residues that might be important for the specific binding of the substrate and the enzymatic mechanism. Kinetic results tend to suggest a SN2 mechanism for the phosphorylation catalyzed by the enzyme. Structural modifications of the FL6K could help to increase the size of the substrates recognized by the enzyme until it binds glycated proteins.

Quelques enzymes sont connus pour déglyquer les kétoamines résultants de la réaction de Maillard entre des sucres et des amines primaires. Il a été démontré qu’Escherichia coli possède un opéron afin de métaboliser la fructoselysine. La fructoselysine 6-kinase, de la famille des PfkB, initie le processus de déglycation permettant l’utilisation ultérieure du glucose-6-P par la bactérie. La résolution de la structure de la FL6K par cristallographie et diffraction des rayons X a permis d’identifier son site actif en présence d’ATP, d’ADP et d’AMP-PNP. La modélisation de la fructoselysine au site actif de la kinase a permis d’identifier des résidus pouvant être importants pour sa liaison et son mécanisme enzymatique. De plus, les résultats de cinétique suggèrent que le mécanisme utilisé par la FL6K semble passer par un état ternaire de type SN2. Des modifications structurales à la FL6K pourraient permettre d’augmenter la taille des substrats afin de permettre ultimement la déglycation de protéines. / Some enzymes have been found to deglycate the products of the Maillard reaction between sugars and primary amines: ketoamines. An operon is found in Escherichia coli that allows the growth on fructoselysine media. The deglycation process is done by a kinase and a “deglycase”. The fructoselysine 6-kinase, a member of the PfkB family, phosphorylates its substrate on the sixth carbon to initiate the metabolism of fructoselysine. Here are presented x-ray crystallography structures obtained for the fructoselysine 6-kinase in its native form and bound with ATP, ADP and AMP-PNP. The active site of the kinase has been determined, and modelisation of fructoselysine allowed identification of some residues that might be important for the specific binding of the substrate and the enzymatic mechanism. Kinetic results tend to suggest a SN2 mechanism for the phosphorylation catalyzed by the enzyme. Structural modifications of the FL6K could help to increase the size of the substrates recognized by the enzyme until it binds glycated proteins.

Recent statistics show that approximately 415 million people worldwide have diabetes. Glycated haemoglobin (HbA1c) measurements were introduced many years ago as the gold standard tool for detecting and monitoring treatment as well as making management decisions for diabetic patients. Glycated haemoglobins are formed by the non-enzymatic glycation of haemoglobin molecules. This non-enzymatic glycation process has been strongly related to pathogenesis of chronic complications associated to diabetes. It was suggested that this glycation process may be moderated by an enzymatic deglycation process thought to involve a deglycating enzyme known as Fructosamine-3-kinase (FN3K), an enzyme that deglycates the glycated haemoglobin in erythrocytes and other glycated proteins in other tissues. FN3K acts through phosphorylation of fructosamines on the third carbon of their sugar moiety, making them unstable and consequently causing them to detach from the protein. The degree of deglycation is thought to depend on the activity of the FN3K enzyme. Moreover, variation in the activity of FN3K between individuals is hypothesised to lead to apparent differences in glycated haemoglobin levels: some individuals have high rates of deglycation so that they tend to have lower average glycaemia than actually the case, while others with low rates of deglycation appear to have higher than actual glycaemia (known as the glycation gap, G-gap). The G-gap has been reported to be associated with alteration of diabetic complications risk. The G-gap reflects the discrepancy between average glycaemia as determined from glycated haemoglobin (measured as HbA1c) and that from the determination of fructosamine. The positive G-gap is defined as a higher level of glycation of proteins than expected whereas a negative G-gap means a lower level of glycation than expected. To explore the role of FN3K in diabetes and other associated morbidities, we decided to divide our research into 3 studies. Each study was categorised according to the type and the source of samples involved. The first study explored the correlation between FN3K activity and protein level with G-gap data; it involved 148 diabetic patients who were recruited at New Cross Hospital, Wolverhampton, selected as having a consistent positive G-gap > +0.5 and a consistent negative G-gap > -0.5 over a minimum of 2 estimations. Age, gender, race and BMI were collected from patients in this study. Blood samples were also 3 collected to measure FN3K activity, protein levels, and markers of CVD in relation to G-gap. The second study involved 23 AECOPD patients who were recruited from St George’s Hospital (London) and were treated with either metformin or a placebo. Serum samples were collected from these patients for a larger study: we assayed those 23 serum samples for FN3K protein levels to explore any possible correlation between FN3K with metformin therapy in COPD patients. The third study utilised 36 human peripheral lung samples from healthy individuals, asymptomatic smokers and stable COPD patients (GOLD 2) who were recruited at The Section of Respiratory Medicine, University Hospital of Ferrara, Italy. Those samples were assessed for FN3K expression by means of immunohistochemistry to explore the difference in FN3K activity between those three categories. It was found that the intracellular activity and protein expression of the FN3K enzyme in diabetic patients negatively correlated with the values of G-gaps where FN3K activity was high in patients with negative G-gap. FN3K serum protein levels were shown to be enhanced with metformin administration in COPD diabetic patients, suggesting a protective role for FN3K enzyme against protein damaged caused by the non-enzymatic glycation of proteins. Therefore, patients with positive G-gap have lower FN3K activity than those with negative G-gap, and in turn they are more susceptible to diabetes related complications. Our data also indicate that metformin has a beneficial effect in reducing damage caused by carbonyl stress from cigarette smoking in COPD patients by the action of FN3K. Our research has demonstrated that FN3K contributes to the protein repair system which protects against damage caused by non-enzymatic glycation. The high activity for the FN3K enzyme was associated with low levels of AGEs and low carbonyl stress levels in observed among patients with diabetes and COPD. In contrast, COPD patients tend to have low FN3K-mediated protection against protein damage in comparison to the normal population. These patients tend to be at risk for developing more complications, particularly CVD complications, than normal, healthy individuals. Treatment with metformin enhances FN3K action in COPD diabetic patients, possibly as a protective enzyme against the damaged caused by the non-enzymatic glycation.

Amadori-Produkte werden spontan während der ersten Phase der Maillard-Reaktion aus reduzierenden Zuckern und Aminen wie Lysin gebildet. Sie entstehen während der Erhitzung von Lebensmitteln und in vivo. Der enzymatische Abbau solcher spontan gebildeten Produkte ist Thema dieser Arbeit. Ein Teil untersuchte die Rolle von Oligosaccharid-Amadori-Produkten während der Verdauung von Kohlenhydraten im Dünndarm. Aufgrund ihrer strukturellen Ähnlichkeit mit bekannten Glycosidase-Inhibitoren wurde eine hemmende Wirkung der Amadori-Produkte auf die Kohlenhydratverdauung vermutet. Der andere Teil beschäftigte sich mit Fructosamin-3-kinase (FN3K) und dessen verwandtem Enzym Fructosamin-3-kinase-related Protein (FN3K-RP) aus humanen Erythrocyten. Diese Ketosaminkinasen werden als Proteinreparaturenzyme betrachtet, sogar als enzymatische Verteidigung gegen Glykierung in vivo diskutiert. Durch ihre Reaktion entstehen jedoch auch hoch-reaktive 1,2-Dicarbonylverbindungen, die weitere Proteinschäden bewirken können. Noch ist nicht klar, ob die Ketosaminkinasen die pathophysiologischen Folgen der Glykierung verhindern oder fördern. In dieser Arbeit wurde die Substratspezifität von Ketosaminkinasen mit einer Reihe von Amadori-Produkten untersucht. Damit könnten Inhibitoren zur weiteren Enzymcharakterisierung oder sogar für pharmazeutische Anwendungen identifiziert werden. Außerdem wurde die Variabilität der Enzymaktivitäten von Mensch zu Mensch in einer Kohorte von 100 Probanden untersucht. Als Modell für die menschliche Kohlenhydratverdauung im Dünndarm wurden Caco-2-Zellen als Monolayer etabliert. Deren Sucrase-Isomaltase kann die alpha-glycosidische Bindung in Amadori-Produkten von Maltose und Maltotriose mit Lysin und auch in Maltulose hydrolysieren. Trotz der Aminogruppe hemmen diese Amadori-Produkte die Maltosehydrolyse nur schwach als konkurrierende Substrate. Lactulosyllysin konnte nicht durch die Lactase der Caco-2-Zellen hydrolysiert werden. Tagatosyllysin und die Heyns-Produkte Glucosyllysin und Mannosyllysin hemmten die Lactosehydrolyse schwach. Alle beobachteten Hemmeffekte sind wahrscheinlich zu schwach, um während der Verdauung in vivo bedeutsam zu sein. Für FN3K konnte Desoxypiperidinofructose als kompetitiver Inhibitor identifiziert werden (Kic 0,006 mM). FN3K zeigte nur geringe Selektivität gegenüber Amadori-Produkten verschiedener Amine, ausgenommen aromatischer Amine. FN3K-RP war in Erythrocyten wesentlich aktiver als FN3K, auch wenn die Aktivität nicht selektiv inhibiert werden konnte. Beide Enzymaktivitäten unterscheiden sich unter den 100 Probanden, mit einer Spannweite von 3 bis 12 mU/g Hämoglobin für FN3K und 60 bis 135 mU/g Hb für FN3K und FN3K-RP zusammen. Es scheint eine Verbindung zwischen der Ketosaminkinase-Aktivität in Erythrocyten und Nierenerkrankungen, familiär auftretendem Diabetes mellitus, sowie familiär aufgetretenen Herzinfarkten oder Schlaganfällen zu bestehen, wie orientierende Auswertungen zeigten. Deshalb ist eine genauere Untersuchung der physiologischen Bedeutung der Ketosaminkinasen nötig. / Amadori products are formed spontaneously from reducing sugars and amines, e.g. lysine, during the first phase of the Maillard reaction. They occur in heated food and in vivo. The thesis focuses on the enzymatic degradation of such spontaneously formed compounds. One part of this work investigated the faith and impact of oligosaccharide derived Amadori products during small intestinal carbohydrate digestion. Due to their structural similarity with known glycosidase inhibitors, an inhibitory action of Amadori products towards carbohydrate digestions was assumed. The other part dealt with fructosamine-3-kinase (FN3K) and its related protein (FN3K-RP) from human erythrocytes. Such ketosamine kinases are regarded as protein repair enzymes, maybe even an enzymatic defence against glycation in vivo. While deglycating protein bound Amadori products, however, they produce highly reactive 1,2-dicarbonyl compounds, which can lead to further protein damage. It is unclear, whether the ketosamine kinase action prevents or supports the pathophysiological effects of glycation. This work studied the substrate specifity of ketosamine kinases with a variety of Amadori products, which could result in inhibitors for further enzyme characterisation or even pharmaceutical uses. Further, the variability of both enzyme activities in a cohort of 100 subjects was examined. As a model for human small intestinal carbohydrate digestion, a Caco-2 cell monolayer was employed. Their sucrase-isomaltase is able to hydrolyse the alpha-glucosidic linkage in Amadori products of maltose and maltotriose with lysine, as well as in maltulose. Despite their amino group, those amadori products inhibited maltose hydrolysis merely weakly as competing substrates. Lactulosyl lysine on the other hand could not be hydrolysed by Caco-2 lactase. Tagatosyl lysine and the Heyns products glucosyl lysine and mannosyl lysine showed weak inhibition of lactose hydrolysis. All observed inhibitory effects are probably too weak to be of importance during carbohydrate digestion in vivo. Deoxypiperidinofructose was identified as a competitive inhibitor of FN3K (Kic 0,006 mM). FN3K acted rather non-specific towards Amadori products of different amines, except aromatic amines. FN3K-RP showed much higher activity in erythrocytes than FN3K, although its activity could not be inhibited selectively. Both enzyme activities vary among 100 subjects, with a range of 3 to 12 mU/g hemoglobin for FN3K and 60 to 135 mU/g hb for FN3K and FN3K-RP together. Relations of ketosamine kinase activity in erythrocytes with renal diseases, familial diabetes mellitus and familial cardiovascular events seem to exist. Thus, investigating the physiological impact of ketosamine kinases is necessary.

Amadori-Produkte werden spontan während der ersten Phase der Maillard-Reaktion aus reduzierenden Zuckern und Aminen wie Lysin gebildet. Sie entstehen während der Erhitzung von Lebensmitteln und in vivo. Der enzymatische Abbau solcher spontan gebildeten Produkte ist Thema dieser Arbeit. Ein Teil untersuchte die Rolle von Oligosaccharid-Amadori-Produkten während der Verdauung von Kohlenhydraten im Dünndarm. Aufgrund ihrer strukturellen Ähnlichkeit mit bekannten Glycosidase-Inhibitoren wurde eine hemmende Wirkung der Amadori-Produkte auf die Kohlenhydratverdauung vermutet. Der andere Teil beschäftigte sich mit Fructosamin-3-kinase (FN3K) und dessen verwandtem Enzym Fructosamin-3-kinase-related Protein (FN3K-RP) aus humanen Erythrocyten. Diese Ketosaminkinasen werden als Proteinreparaturenzyme betrachtet, sogar als enzymatische Verteidigung gegen Glykierung in vivo diskutiert. Durch ihre Reaktion entstehen jedoch auch hoch-reaktive 1,2-Dicarbonylverbindungen, die weitere Proteinschäden bewirken können. Noch ist nicht klar, ob die Ketosaminkinasen die pathophysiologischen Folgen der Glykierung verhindern oder fördern. In dieser Arbeit wurde die Substratspezifität von Ketosaminkinasen mit einer Reihe von Amadori-Produkten untersucht. Damit könnten Inhibitoren zur weiteren Enzymcharakterisierung oder sogar für pharmazeutische Anwendungen identifiziert werden. Außerdem wurde die Variabilität der Enzymaktivitäten von Mensch zu Mensch in einer Kohorte von 100 Probanden untersucht. Als Modell für die menschliche Kohlenhydratverdauung im Dünndarm wurden Caco-2-Zellen als Monolayer etabliert. Deren Sucrase-Isomaltase kann die alpha-glycosidische Bindung in Amadori-Produkten von Maltose und Maltotriose mit Lysin und auch in Maltulose hydrolysieren. Trotz der Aminogruppe hemmen diese Amadori-Produkte die Maltosehydrolyse nur schwach als konkurrierende Substrate. Lactulosyllysin konnte nicht durch die Lactase der Caco-2-Zellen hydrolysiert werden. Tagatosyllysin und die Heyns-Produkte Glucosyllysin und Mannosyllysin hemmten die Lactosehydrolyse schwach. Alle beobachteten Hemmeffekte sind wahrscheinlich zu schwach, um während der Verdauung in vivo bedeutsam zu sein. Für FN3K konnte Desoxypiperidinofructose als kompetitiver Inhibitor identifiziert werden (Kic 0,006 mM). FN3K zeigte nur geringe Selektivität gegenüber Amadori-Produkten verschiedener Amine, ausgenommen aromatischer Amine. FN3K-RP war in Erythrocyten wesentlich aktiver als FN3K, auch wenn die Aktivität nicht selektiv inhibiert werden konnte. Beide Enzymaktivitäten unterscheiden sich unter den 100 Probanden, mit einer Spannweite von 3 bis 12 mU/g Hämoglobin für FN3K und 60 bis 135 mU/g Hb für FN3K und FN3K-RP zusammen. Es scheint eine Verbindung zwischen der Ketosaminkinase-Aktivität in Erythrocyten und Nierenerkrankungen, familiär auftretendem Diabetes mellitus, sowie familiär aufgetretenen Herzinfarkten oder Schlaganfällen zu bestehen, wie orientierende Auswertungen zeigten. Deshalb ist eine genauere Untersuchung der physiologischen Bedeutung der Ketosaminkinasen nötig. / Amadori products are formed spontaneously from reducing sugars and amines, e.g. lysine, during the first phase of the Maillard reaction. They occur in heated food and in vivo. The thesis focuses on the enzymatic degradation of such spontaneously formed compounds. One part of this work investigated the faith and impact of oligosaccharide derived Amadori products during small intestinal carbohydrate digestion. Due to their structural similarity with known glycosidase inhibitors, an inhibitory action of Amadori products towards carbohydrate digestions was assumed. The other part dealt with fructosamine-3-kinase (FN3K) and its related protein (FN3K-RP) from human erythrocytes. Such ketosamine kinases are regarded as protein repair enzymes, maybe even an enzymatic defence against glycation in vivo. While deglycating protein bound Amadori products, however, they produce highly reactive 1,2-dicarbonyl compounds, which can lead to further protein damage. It is unclear, whether the ketosamine kinase action prevents or supports the pathophysiological effects of glycation. This work studied the substrate specifity of ketosamine kinases with a variety of Amadori products, which could result in inhibitors for further enzyme characterisation or even pharmaceutical uses. Further, the variability of both enzyme activities in a cohort of 100 subjects was examined. As a model for human small intestinal carbohydrate digestion, a Caco-2 cell monolayer was employed. Their sucrase-isomaltase is able to hydrolyse the alpha-glucosidic linkage in Amadori products of maltose and maltotriose with lysine, as well as in maltulose. Despite their amino group, those amadori products inhibited maltose hydrolysis merely weakly as competing substrates. Lactulosyl lysine on the other hand could not be hydrolysed by Caco-2 lactase. Tagatosyl lysine and the Heyns products glucosyl lysine and mannosyl lysine showed weak inhibition of lactose hydrolysis. All observed inhibitory effects are probably too weak to be of importance during carbohydrate digestion in vivo. Deoxypiperidinofructose was identified as a competitive inhibitor of FN3K (Kic 0,006 mM). FN3K acted rather non-specific towards Amadori products of different amines, except aromatic amines. FN3K-RP showed much higher activity in erythrocytes than FN3K, although its activity could not be inhibited selectively. Both enzyme activities vary among 100 subjects, with a range of 3 to 12 mU/g hemoglobin for FN3K and 60 to 135 mU/g hb for FN3K and FN3K-RP together. Relations of ketosamine kinase activity in erythrocytes with renal diseases, familial diabetes mellitus and familial cardiovascular events seem to exist. Thus, investigating the physiological impact of ketosamine kinases is necessary.