Patent Application
20080287375
This application relates to compounds which selectively inhibit glycosidases, methods of making the inhibitors, and uses thereof.
A wide range of cellular proteins, both nuclear and cytoplasmic, are post-translationally modified by the addition of the monosaccharide 2-acetamido-2-deoxy-β-D-glucopyranoside (βN-acetylglucosamine) which is attached via an β-glycosidic linkage[1]. This modification is generally referred to as O-linked N-acetylglucosamine or O-GlcNAc.
O-GlcNAc-modified proteins have a wide range of vital cellular functions including, for example, transcription[25], proteasomal degradation[6], and cellular signaling[7]. O-GlcNAc is also found on many structural proteins[8-10]. For example, it has been found on a number of cytoskeletal proteins, including neurofilament proteins[11, 12], synapsins[13, 14], synapsin-specific clathrin assembly protein AP-3[15], and ankyrinG[16]. O-GlcNAc modification has been found to be abundant in the brain[17, 18]. It has also been found on proteins clearly implicated in the etiology of several diseases including type II diabetes, Alzheimer's disease (AD), and cancer.
For example, it is well established that AD and a number of related tauopathies including Downs' syndrome, Pick's disease, Niemann-Pick Type C disease, and amyotrophic lateral sclerosis (ALS) are characterized, in part, by the development of neurofibrillary tangles (NFTs). These NFTs are aggregates of paired helical filaments (PHFs) and are composed of an abnormal form of the important protein “tau”. Normal tau stabilizes a key cellular network of microtubules that is essential for distributing proteins and nutrients within neurons. In AD patients, however, tau becomes hyperphosphorylated, disrupting its normal functions, forming PHFs and ultimately aggregating to form detrimental NFTs. A clear parallel between NFT levels in the brains of AD patients and the severity of dementia strongly supports a key role for tau dysfunction in AD[19, 20]. The precise causes of this hyperphosphorylation of tau remain elusive. Accordingly, considerable effort has been dedicated toward a) elucidating the molecular physiological basis of tau hyperphosphorylation,[6] and b) identifying strategies that could limit tau hyperphosphorylation in the hope that these might halt, or even reverse, the progression of Alzheimer's disease.[7, 8]. Thus far, several lines of evidence suggest that up-regulation of a number of kinases may be involved[9, 10, 21] in hypelphosphorylation of tau, although very recently, an alternative basis for this hyperphosphorylation has been advanced.[21]. In particular, it has recently emerged that phosphate levels of tau are regulated by the levels of O-GlcNAc on tau. Hyperphosphorylated tau in human AD brains has markedly lower levels of O-GlcNAc than are found in healthy human brains[2,3]. These results suggest that a malfunction in the mechanisms regulating tau O-GlcNAc levels may be vitally important in the formation of NFTs and associated neurodegeneration.
Humans have three genes encoding enzymes that cleave terminal β-N-acetyl-glucosamine residues from glycoconjugates. The first of these encodes the enzyme O-glycoprotein 2-acetamido-2-deoxy-β-D-glucopyranosidase, (O-GlcNAcase). O-GlcNAcase is a member of family 84 of glycoside hydrolases that includes enzymes from organisms as diverse as prokaryotic pathogens to humans (for the family classification of glycoside hydrolases see Coutinho, P. M. & Henrissat, B. (1999) Carbohydrate-Active Enzymes server at URL: http://afmb.cnrs-mrs.fr/CAZY/([19, 20]). O-GlcNAcase acts to hydrolyse O-GlcNAc off of serine and threonine residues of post-translationally modified proteins[1, 22, 23]. Consistent with the presence of O-GlcNAc on many intracellular proteins, the enzyme O-GlcNAcase appears to have a role in the etiology of several diseases including type II diabetes[7, 21], AD[9, 17, 24], and cancer[18]. Although O-GlcNAcase was likely isolated earlier on[11, 12], about 20 years elapsed before its biochemical role in acting to cleave O-GlcNAc from serine and threonine residues of proteins was understood[13]. More recently O-GlcNAcase has been cloned[15], partially characterized[16], and suggested to have additional activity as a histone acetyltransferase[14]. However, little was known about the catalytic mechanism of this enzyme.
The other two genes, HEXA and HEXB, encode enzymes catalyzing the hydrolytic cleavage of terminal β-N-acetylglucosamine residues from glycoconjugates. The gene products of HEXA and HEXB predominantly yield two dimeric isozymes, hexosaminidase A and hexosaminidase B, respectively. Hexosaminidase A (αβ), a heterodimeric isozyme, is composed of an α- and a β-subunit. Hexosaminidase B (1313), a homodimeric isozyme, is composed of two β-subunits. The two subunits, α- and β-, bear a high level of sequence identity. Both of these enzymes are classified as members of family 20 of glycoside hydrolases and are normally localized within lysosomes. The proper functioning of these lysosomal β-hexosaminidases is critical for human development, a fact that is underscored by the tragic genetic illnesses, Tay-Sach's and Sandhoff diseases which stem from a dysfunction in, respectively, hexosaminidase A and hexosaminidase B[25]. These enzymatic deficiencies cause an accumulation of glycolipids and glycoconjugates in the lysosomes resulting in neurological impairment and deformation. The deleterious effects of accumulation of gangliosides at the organismal level are still being uncovered[26].
As a result of the biological importance of these β-N-acetyl-glucosaminidases, small molecule inhibitors of glycosidases[27-30] have received a great deal of attention[31], both as tools for elucidating the role of these enzymes in biological processes and in developing potential therapeutic applications. The control of glycosidase function using small molecules offers several advantages over genetic knockout studies including the ability to rapidly vary doses or to entirely withdraw treatment.
However, a major challenge in developing inhibitors for blocking the function of mammalian glycosidases, including O-GlcNAcase, is the large number of functionally related enzymes present in tissues of higher eukaryotes. Accordingly, the use of non-selective inhibitors in studying the cellular and organismal physiological role of one particular enzyme is complicated because complex phenotypes arise from the concomitant inhibition of such functionally related enzymes. In the case of β-N-acetylglucosaminidases, existing compounds that act to block O-GlcNAcase function are non-specific and act potently to inhibit the lysosomal β-hexosaminidases. Heretofore no potent inhibitor was known that is selective for nucleocytoplasmic O-GlcNAcase over the lysosomal β-hexosaminidases.
A few of the better characterized inhibitors of β-N-acetyl-glucosaminidases which have been used in studies of O-GlcNAc post-translational modification within both cells and tissues are streptozotocin (STZ), 2′-methyl-α-
STZ has long been used as a diabetogenic compound because it has a particularly detrimental effect on β-islet cells[36]. STZ exerts its cytotoxic effects through both the alkylation of cellular DNA[36, 37] as well as the generation of radical species including nitric oxide[38]. The resulting DNA strand breakage promotes the activation of poly(ADP-ribose) polymerase (PARP)[39] with the net effect of depleting cellular NAD+ levels and, ultimately, leading to cell death[40, 41]. Other investigators have proposed instead that STZ toxicity is a consequence of the irreversible inhibition of O-GlcNAcase, which is highly expressed within 3-islet cells[32, 42]. This hypothesis has, however, been brought into question by two independent research groups[43, 44]. Because cellular O-GlcNAc levels on proteins increase in response to many forms of cellular stress[45] it seems possible that STZ results in increased O-GlcNAc-modification levels on proteins by inducing cellular stress rather than through any specific and direct action on O-GlcNAcase. Indeed, Hanover and coworkers have shown that STZ functions as a poor and somewhat selective inhibitor of O-GlcNAcase[46] and although it has been proposed by others that STZ acts to irreversibly inhibit O-GlcNAcase[47], there has been no clear demonstration of this mode of action.
NAG-thiazoline has been found to be a potent inhibitor of family 20 hexosaminidases,[30, 48] and more recently, the family 84 O-GlcNAcases[49]. Despite its potency, a downside to using NAG-thiazoline in a complex biological context is that it lacks selectivity and therefore perturbs multiple cellular processes.
PUGNAc is another compound that suffers from the same problem of lack of selectivity, yet has enjoyed use as an inhibitor of both human O-GlcNAcase[13, 50] and the family 20 human β-hexosaminidases[5]. This molecule, developed by Vasella and coworkers, was found to be a potent competitive inhibitor of the β-N-acetyl-glucosaminidases from Canavalia ensiformis, Mucor rouxii, and the β-hexosamimidase from bovine kidney[28].
The need has therefore arisen for improved selective inhibitors of glycosidases.
The embodiments of the inventions relate to compounds which selectively inhibit glycosidases. The invention also relates to methods of making such compounds and uses thereof.
In one embodiment of the invention, the compounds have the general chemical formula (I):
wherein R3, R5, R6, are selected from the group consisting of branched alkyl chains, unbranched allyl chains, cycloalkyl groups, aromatic groups, alcohols, ethers, amines, substituted or unsubstituted carbamates, substituted or unsubstituted ureas, esters, amides, aldehydes, carboxylic acids, and heteroatom containing derivatives thereof, wherein said esters and amides may comprise an acyl group selected from the group consisting of branched alkyl chains, unbranched allyl chains, cycloalkyl groups, aromatic groups, and heteroatom derivatives thereof; R2 and R4 are CH2, CHR1, NH, NR1, or any heteroatom, and R1 is selected from the group consisting of H, ethers, amines, branched alkyl chains, unbranched alkyl chains, cycloalkyl groups, aromatic groups, and heteroatom derivatives thereof. The invention includes pharmaceutically acceptable salts of the above compounds. In some embodiments of the invention, R2 is S, and R1 is selected from the group consisting of CH2CH3, (CH2)2CH3, (CH2)3CH3, (CH2)4CH3, CH(CH3)2 and CH2CH(CH3)2, and R4 is O. The invention also relates to prodrugs of the compounds, pharmaceutical compositions containing the compounds and a pharmaceutically acceptable carrier, and pharmaceutical compositions containing prodrugs of the compounds and a pharmaceutically acceptable carrier.
In another embodiment of the invention, the compounds have the general chemical formula (II):
wherein X1-X6 are O, NH, NR1, CHR1, CH2 or any heteroatom, R1 to R5 are selected from the group consisting of branched alkyl chains, unbranched alkyl chains, cycloalkyl groups, aromatic groups, alcohols, ethers, amines, substituted or unsubstituted carbamates, substituted or unsubstituted ureas, esters, amides, aldehydes, carboxylic acids, and heteroatom containing derivatives thereof, wherein said esters and amides may comprise an acyl group selected from the group consisting of branched alkyl chains, unbranched alkyl chains, cycloalkyl groups, aromatic groups, and heteroatom derivatives thereof, and pharmaceutically acceptable salts thereof.
In some embodiments the compounds selectively inhibit the activity of specific glycosidases over other glycosidases. In one embodiment of the invention, the glycosidases comprise glycoside hydrolases. In another embodiment of the invention, the glycoside hydrolases are family 84 glycoside hydrolases. In a specific embodiment of the invention, the glycosidase is O-GlcNAcase. The compounds selectively inhibit the activity O-GlcNAcase over β-hexosaminidase. In particular, the compounds selectively inhibit the cleavage of 2-acetamido-2-deoxy-β-D-glucopyranoside (O-GlcNAc) from proteins in this particular embodiment.
The compounds are useful in the development of animal models for studying diseases or disorders related to deficiencies in O-GlcNAcase, over-expression of O-GlcNAcase, accumulation of O-GlcNAc, depletion of O-GlcNAc, and for studying treatment of diseases and disorders related to deficiency or over-expression of O-GlcNAcase, or accumulation or depletion of O-GlcNAc. Such diseases and disorders include diabetes, neurodegenerative diseases, including Alzheimer's disease, and cancer. The compounds are also useful in the treatment of diseases and disorders responsive to glycosidase inhibition therapy. The compounds are also useful in preparing cells and tissues for stress associated with tissue damage or stress, stimulating cells, and promoting differentiation of cells. These compounds may also be useful in inhibiting specific glycosidases, for example microbial toxins that are members of family 84 glycoside hydrolases, and thereby find use as antimicrobials.
The invention also relates to methods of making the compounds. In one embodiment, the method may comprise the steps of:
The invention also relates to methods of making selective glycosidase inhibitors comprising:
In some embodiments of the invention, the method of making selective glycoside inhibitors comprises the steps of:
wherein X1-X6 are O, NH, NR1, CHR1, CH2 or any heteroatom, R1 to R5 are selected from the group consisting of branched alkyl chains, unbranched alkyl chains, cycloalkyl groups, aromatic groups, alcohols, ethers, amines, substituted or unsubstituted carbamates, substituted or unsubstituted ureas, esters, amides, aldehydes, carboxylic acids, and heteroatom containing derivatives thereof, wherein said esters and amides may comprise an acyl group selected from the group consisting of branched alkyl chains, unbranched alkyl chains, cycloalkyl groups, aromatic groups, and heteroatom derivatives thereof, and pharmaceutically acceptable salts thereof;
In some embodiments of the invention, the R1 side chain of the inhibitor can be enlarged by inserting a branched alkyl chain, an unbranched alkyl chain, a cycloalkyl group, an aromatic group, or heteroatom derivatives thereof, into the side chain. In some specific embodiments, the R1 chain is enlarged by inserting groups selected from the group consisting of CH2CH3, (CH2)2CH3, (CH2)3CH3, (CH2)4CH3, CH(CH3)2 and CH2CH(CH3)2.
In drawings which are intended to illustrate embodiments of the invention:
and β-hexosaminidase
catalyzed hydrolysis of MU-GlcNAc.
Throughout the following description, specific details are set forth in order to provide a more thorough understanding of the invention. However, the invention may be practiced without these particulars. In other instances, well known elements have not been shown or described in detail to avoid unnecessarily obscuring the invention. Accordingly, the specification and drawings are to be regarded in an illustrative, rather than a restrictive, sense.
The invention comprises compounds which selectively inhibit glycosidases. In one embodiment of the invention, the compounds have the general chemical formula (I):
wherein R3, R5, R6, are selected from the group consisting of branched alkyl chains, unbranched alkyl chains, cycloalkyl groups, aromatic groups, alcohols, ethers, amines, substituted or unsubstituted carbamates, substituted or unsubstituted ureas, esters, amides, aldehydes, carboxylic acids, and heteroatom containing derivatives thereof, wherein said esters and amides may comprise an acyl group selected from the group consisting of branched alkyl chains, unbranched alkyl chains, cycloalkyl groups, aromatic groups, and heteroatom derivatives thereof; R2 and R4 are CH2, CHR1, NH, NR1, or any heteroatom, and R1 is selected from the group consisting of H, ethers, amines, branched alkyl chains, unbranched alkyl chains, cycloalkyl groups, aromatic groups, and heteroatom derivatives thereof. The invention includes pharmaceutically acceptable salts of the above compounds. In some embodiments, R2 is S, R1 is selected from the group consisting of CH2CH3, (CH2)2CH3, (CH2)3CH3, (CH2)4CH3, CH(CH3)2 and CH2CH(CH3)2 and R4 is O. In some specific embodiments of the invention, the compounds comprise 1,2-dideoxy-2′-ethyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline, 1,2-dideoxy-2′-propyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline, 1,2-dideoxy-2′-butyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline, 1,2-dideoxy-2′-pentyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline, 1,2-dideoxy-2′-isopropyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline, 1,2-dideoxy-2′-isobutyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline.
As will be appreciated by a person skilled in the art, formula (I) above may also be represented alternatively as follows:
In another embodiment of the invention, the compounds have the general chemical formula (II):
wherein X1-X6 are O, NH, NR1, CHR1, CH2 or any heteroatom, R1 to R5 are selected from the group consisting of branched alkyl chains, unbranched alkyl chains, cycloalkyl groups, aromatic groups, alcohols, ethers, amines, substituted or unsubstituted carbamates, substituted or unsubstituted ureas, esters, amides, aldehydes, carboxylic acids, and heteroatom containing derivatives thereof, wherein said esters and amides may comprise an acyl group selected from the group consisting of branched alkyl chains, unbranched alkyl chains, cycloalkyl groups, aromatic groups, and heteroatom derivatives thereof, and pharmaceutically acceptable salts thereof.
The invention includes pharmaceutically acceptable salts of the above compounds. In some embodiments, R1 is selected from the group consisting of CH2CH3, (CH2)2CH3, (CH2)3CH3, (CH2)4CH3, CH(CH3)2 and CH2CH(CH3)2, R2, R3 and R4 are OH, X1 and X2 are 0, X3 is NH, X4 and X5 are 0, X6 is NH, and R5 is C6H6. In some specific embodiments of the invention, the compounds comprise O-(2-deoxy-2-propamido-
Throughout this application, it is contemplated that the term “compounds” refers to the compounds discussed above and includes derivatives of the compounds, including acyl-protected derivatives, and pharmaceutically acceptable salts of the compounds and the derivatives. The invention also comprises prodrugs of the compounds, pharmaceutical compositions containing the compounds and a pharmaceutically acceptable carrier, and pharmaceutical compositions containing prodrugs of the compounds and a pharmaceutically acceptable carrier.
In some embodiments of the invention the compounds selectively inhibit the activity of specific glycosidases over other glycosidases. The glycosidases may comprise glycoside hydrolases. For example, the glycoside hydrolases may be family 84 glycoside hydrolases. In a specific embodiment of the invention, the glycosidase is O-GlcNAcase. The invention comprises compounds which selectively inhibit the activity of O-GlcNAcase over β-hexosaminidase. In particular, the compounds selectively inhibit the cleavage of O-GlcNAc from proteins.
The compounds of the invention are valuable tools in studying the physiological role of O-GlcNAc at the cellular and organismal level. The compounds are useful in the development of animal models to study a disease or disorder, or for studying treatment of a disease or disorder, related to deficiency or over-expression of O-GlcNAcase or accumulation or depletion of O-GlcNAc. As an example, the compounds are useful in the development of a disease model for the development of Type I or Type II diabetes.
Type II diabetes develops when humans or animals are unable to properly regulate blood glucose levels. Tissues must be able to sense and rapidly respond to changes in glucose availability as well as to signals from other components of the endocrine system. Glucose is the key nutrient in regulating insulin synthesis and secretion from the β-cells of the pancreas. Of all glucose entering into cells, 2-5% is shunted into the hexosamine biosynthetic pathway, thereby regulating cellular concentrations of the end product of this pathway, uridine diphosphate-N-acetylglucosamine (UDP-GlcNAc).[52] UDP-GlcNAc is a substrate of the nucleocytoplasmnic enzyme O-GlcNAc transferase (OGTase),[53-56] which acts to post-translationally add GlcNAc to specific serine and threonine residues of numerous nucleocytoplasmic proteins. OGTase recognizes many of its substrates[57, 58] and binding partners[59, 60] through its tetratricopeptide repeat (TPR) domains[61, 62] As described above, O-GlcNAcase[13, 15] removes this post-translational modification to liberate proteins making the O-GlcNAc-modification a dynamic cycle occurring several times during the lifetime of a protein[63] O-GlcNAc has been found in several proteins on known phosphorylation sites,[3, 64-66] suggesting a role for O-GlcNAc in cellular signaling. Additionally, OGTase shows unusual kinetic behaviour making it exquisitely sensitive to intracellular UDP-GlcNAc substrate concentrations and therefore glucose supply.[67] All of these data point to a logical role for O-GlcNAc levels acting as a nutrient sensing mechanism. To support such a potential role in nutrient sensing it has been shown that in peripheral tissues, elevated O-GlcNAc levels result in the development of insulin resistance.[7, 21] Indeed, it has been proposed that markedly elevated O-GlcNAc levels may lead to type II diabetes[68] in accord with a recent study showing that a single nucleotide polymorphism is associated with the occurrence of Type II diabetes in a Mexican population. PUGNAc has been used in culture cells and in tissues to show that increased O-GlcNAc levels cause insulin resistance[7, 32-35]. By analogy, the compounds of the invention can also be used in similar studies and used to develop animal models to show the role of O-GlcNAc levels in the development of diabetes.
The compounds of the invention are also useful in the treatment of diseases or disorders related to deficiency or over-expression of O-GlcNAcase or accumulation or depletion of O-GlcNAc, or any disease or disorder responsive to glycosidase inhibition therapy. Such diseases and disorders include, but are not limited to, diabetes, neurodegenerative disorders, such as Alzheimer's disease (AD), and cancer. Such diseases and disorders may also include diseases or disorders related to the accumulation or deficiency in the enzyme OGTase.
For example, because of relationship between O-GlcNAc-levels and phosphorylation levels on tau as discussed above, the compounds of the invention may be used to study and treat AD and other tauopathies. Six isoforms of tau are found in the human brain. In AD patients, all six isoforms of tau are found in NFTs, and all are markedly hyperphosphorylated.[69, 70] Tau in healthy brain tissue bears only 2 or 3 phosphate groups, whereas those found in the brains of AD patients bear, on average, 8 phosphate groups[71, 72].
The presence of O-GlcNAc on tau has stimulated studies that correlate O-GlcNAc levels with tau phosphorylation levels. The recent interest in this field stems from the observation that O-GlcNAc modification has been found to occur on many proteins at amino acid residues that are also known to be phosphorylated.[65, 66, 73] Consistent with this observation, it has been found that increases in phosphorylation levels result in decreased O-GlcNAc levels and conversely, increased O-GlcNAc levels correlate with decreased phosphorylation levels[74] This reciprocal relationship between O-GlcNAc and phosphorylation has been termed the “Ying-Yang hypothesis”[75] and has gained strong biochemical support by the recent discovery that the enzyme OGTase[55] forms a functional complex with phosphatases that act to remove phosphate groups from proteins[60] Like phosphorylation, O-GlcNAc is a dynamic modification that can be removed and reinstalled several times during the lifespan of a protein. Suggestively, the gene encoding O-GlcNAcase has been mapped to a chromosomal locus that is linked to AD.[15, 76]
Very recently, it has been shown that O-GlcNAc levels of soluble tau protein from human brains affected with AD are markedly lower than those from healthy brain[17] Furthermore, PHF from diseased brain was suggested to lack completely any O-GlcNAc modification whatsoever.[17] The molecular basis of this hypoglycosylation of tau is not known, although it may stem from increased activity of kinases and/or dysfunction of one of the enzymes involved in processing O-GlcNAc. Supporting this latter view, in both PC-12 neuronal cells and in brain tissue sections from mice, a nonselective N-acetylglucosamindase inhibitor was used to increase tau O-GlcNAc levels, whereupon it was observed that phosphorylation levels decreased[17] The implication of these collective results is that by maintaining healthy O-GlcNAc levels in AD patients, such as by inhibiting the action of O-GlcNAcase, one should be able to block hyperphosphorylation of tau and all of the associated effects of tau hyperphosphorylation, including the formation of NFTs and downstream effects. However, because the proper functioning of the β-hexosaminidases is critical, any potential therapeutic intervention for the treatment of AD that blocks the action of O-GlcNAcase would have to avoid the concomitant inhibition of both hexosaminidases A and B. This selective inhibition is provided by compounds of the present invention.
The compounds of the invention are useful as selective inhibitors of other types of glycosidases. For example, some microbial toxins are members of family 84 glycoside hydrolases, and therefore, the compounds of the invention may be useful as antimicrobials.
The compounds of the invention are also useful in promoting differentiation of cells, such as promoting pluripotent cells into islet β-cells. For example, O-GlcNAc is known to regulate the function of several transcription factors and is known to be found on the transcription factor PDX-1. Modification of the transcription factor PDX-1 by modification of O-GlcNAc residues may affect activity of PDX-1, which in turn affects cell differentiation.
The compounds of the invention are also useful in preparing cells for stress. Recent studies have indicated that PUGNAc can be used in an animal model to reduce myocardial infarct size after left coronary artery occlusions[77].
The invention also relates to various methods of making the compounds. In one embodiment the method may comprise the steps of:
The invention also relates to methods of making selective glycosidase inhibitors comprising, for example, the steps of:
In some embodiments of the invention, the method of making selective glycoside inhibitors comprises the steps of:
wherein X1-X6 are O, NH, NR1, CHR1, CH2 or any heteroatom, R1 to R5 are selected from the group consisting of branched alkyl chains, unbranched alkyl chains, cycloalkyl groups, aromatic groups, alcohols, ethers, amines, substituted or unsubstituted carbamates, substituted or unsubstituted ureas, esters, amides, aldehydes, carboxylic acids, and heteroatom containing derivatives thereof, wherein said esters and amides may comprise an acyl group selected from the group consisting of branched alkyl chains, unbranched alkyl chains, cycloalkyl groups, aromatic groups, and heteroatom derivatives thereof, and pharmaceutically acceptable salts thereof;
In some embodiments of the invention, the R1 chain of the inhibitor can be enlarged by inserting a branched alkyl chain, an unbranched alkyl chain, a cycloalkyl group, an aromatic group or heteroatom derivatives thereof, into the side chain. In some specific embodiments, the R1 chain is enlarged by inserting groups selected from the group consisting of CH2CH3, (CH2)2CH3, (CH2)3CH3, (CH2)4CH3, CH(CH3)2 and CH2CH(CH3)2.
The following examples are intended to illustrate embodiments of the invention and are not intended to be construed in a limiting manner.
A logical starting point for the design of inhibitors of O-GlcNAcase takes into consideration the catalytic mechanism of O-GlcNAcase and β-hexosaminidase. Although the catalytic mechanism of action of the family 20 human β-hexosaminidase A and B have been fairly well established[78], that of the family 84 O-GlcNAcase remains unknown. The inventors therefore first elucidated the catalytic mechanism of human O-GlcNAcase and secondly, used this information in designing simple inhibitors that would be potent, cell permeable, and highly selective for O-GlcNAcase over the lysosomal β-hexosaminidases.
Three realistic mechanistic alternatives exist for O-GlcNAcase and family 84 glycoside hydrolases. The first alternative is an inverting mechanism (
The second mechanistic possibility is the canonical two step double displacement mechanism that results in retention of configuration at the anomeric center (
The third mechanistic alternative involves the nucleophilic participation of the 2-acetamido group of the substrate in place of an enzymic catalytic nucleophile (
A key difference between these mechanistic alternatives is the involvement of the 2-acetamido group of the substrate. This moiety may actively participate in catalysis as a nucleophile, as for the lysosomal β-hexosaminidases, or it may act as a bystander.
To address the role of the 2-acetamido group of the substrate, several substrate analogues were synthesized bearing differing levels of fluorine substitution on the N-acetyl group (Scheme 1). The highly electronegative fluorine substituents decrease the basicity of the carbonyl group and the expected effect of such substitutions on an enzymatic reaction using anchimeric assistance would be to decrease its rate.
All buffer salts used in this study were obtained from Sigma-Aldrich. Dry methanol and toluene were purchased from Across Organics. Dichloromethane and triethylamine were dried by distillation over CaH2 prior to use. β-Hexosaminidase was purchased from Sigma (Lot 043K3783). STZ was purchased from Sigma-Aldrich and samples were freshly dissolved just prior to assays. PUGNAc was obtained from Toronto Research Chemicals. All other reagents were purchased from Sigma-Aldrich and used without further purification. Milli-Q (18.2 mΩ cm−1) water was used to prepare all buffers. Synthetic reactions were monitored by TLC using Merck Kieselgel 60 F254 aluminum-backed sheets. Compounds were detected by charring with 10% ammonium molybdate in 2 M H2SO4 and heating. Flash chromatography under a positive pressure was performed with Merck Kieselgel 60 (230-400 mesh) using the specified eluants. 1H NMR spectra were recorded on a Varian AS500 Unity Innova spectrometer at 500 MHz (chemical shifts quoted relative to CDCl3, CD3OD or (CD3)2SO where appropriate). 19F NMR spectra were recorded on a Varian AS500 Unity Innova spectrometer at 470 MHz and are proton-coupled with CF3CO2H as a reference. 13C NMR spectra were recorded on a Varian AS500 Unity Innova spectrometer at 125 MHz (chemical shifts quoted relative to CDCl3, CD3OD, or (CD3)2SO). Elemental analyses of all compounds used in cell culture and enzyme assay were performed at the Simon Fraser University Analytical Facility.
4-Methylumbelliferyl 2-amino-2-deoxy-β-D-glucopyranoside hydrochloride (3) was prepared essentially as described by Roeser and Legler[84] and was used without further purification.
To a cooled (0° C.) solution of the hydrochloride salt 3 (0.50 g, 1.0 mmol) in a solution of dimethylformamide (DMF, 10 mL) was added triethylamine (0.3 mL, 0.21 g, 2.1 mmol) and dry pyridine (10 mL). To a stirred mixture of dry DMF (45 mL) containing dried DOWEX-50H+ resin (6g) was added sodium fluoroacetate (0.9 g). After one hour dicyclohexylcarbodiimide (DCC, 1.6 g, 7.8 mmol) and 30 mL of the fluoroacetic acid solution (6.0 mmol) was added via canula to the reaction vessel containing the hydrochloride salt 3. The resulting solution was allowed to stand for 16 h at 0° C. after which time the reaction was judged complete by TLC analysis. The solvent was partially removed in vacuo after which ethyl acetate (50 mL) and a solution of saturated sodium chloride (20 mL) were added. The organic layer was collected and the aqueous layer was extracted twice with ethyl acetate. The combined organic extracts were washed successively with water, twice with saturated sodium bicarbonate, and finally with a solution of saturated sodium chloride. The organic extracts were dried over MgSO4, filtered, and the solvent removed in vacuo to yield a light yellow syrup. The desired product was purified using flash column silica chromatography (2:1; ethyl acetate-hexanes) to yield the partially purified desired compound as an amorphous white solid (z 356 mg, 0.68 mmol, 68%) that was used in the next step without further purification.
To a cooled (0° C.) solution of the hydrochloride salt 3 (0.15 g, 0.3 mmol) in a solution of dimethylformamide (DMF, 6 mL) was added triethylamine (0.09 mL, 0.063 g, 0.62 mmol) and dry pyridine (3 mL). Dicyclohexylcarbodiimide (DCC, 0.48 g, 2.3 mmol) and difluoroacetic acid (0.12 mL, 0.18 g, 1.3 mmol) was added to the reaction mixture via syringe. The resulting solution was allowed to stand for 16 h at 0° C. after which time two drops of difluoroacetic acid were added. After a further 3.5 h at RT the reaction was judged complete by TLC analysis. The solvent was partially removed in vacuo after which ethyl acetate (50 mL) and a solution of saturated sodium chloride (20 mL) were added. The organic layer was collected and the aqueous layer was extracted twice with ethyl acetate. The combined organic extracts were washed successively with water, twice with saturated sodium bicarbonate, and finally with a solution of saturated sodium chloride. The organic extracts were dried over MgSO4, filtered, and the solvent removed in vacuo to yield a light yellow syrup. The desired product was purified using flash column silica chromatography using a gradient solvent system (1:1; hexanes-ethyl acetate) to yield the partially purified desired compound as a white amorphous solid (−0.10 mg, 0.19 mmol, 64%) that was used in the next step without further purification.
To a cooled (0° C.) solution of the hydrochloride salt 3 (0.10 g, 0.2 mmol) in a solution of dimethylformamide (DMF, 6 mL) was added triethylamine (0.06 mL, 0.42 g, 0.41 mmol). The reaction mixture was then cooled to 0° C. and trifluoroacetic anhydride (0.08 mL, 0.12 g, 5.7 mmol) was added via syringe. The resulting solution was allowed to stand for 16 h at 0° C. after which time the reaction was judged complete by TLC analysis. The reaction mixture was then diluted with ethyl acetate (20 mL) and a solution of saturated sodium chloride (40 mL) was added. The organic layer was collected and the aqueous layer was extracted twice with ethyl acetate. The combined organic extracts were washed successively with water, twice with saturated sodium bicarbonate, and finally with a solution of saturated sodium chloride. The organic extracts were dried over MgSO4, filtered, and the solvent removed in vacuo to yield a light yellow syrup. The desired product was purified using flash column silica chromatography using a gradient solvent system (1:1; hexanes-ethyl acetate) to yield the partially purified desired compound as an amorphous white solid (≈0.93 g, 0.17 mmol, 82%) that was used in the next step without further purification.
To a solution of each glycoside in dry methanol was added a spatula tip of anhydrous sodium methoxide. The resulting basic solution was stirred under nitrogen until the reaction was judged complete by TLC analysis. Dowex-50 H+ resin was added to the stirred reaction mixture until the pH of the solution became neutral. The suspension was filtered and the filter cake rinsed extensively with methanol after which the solvent from the combined filtrates was removed in vacuo. The desired deprotected glycosides were isolated by flash column silica chromatography using the following solvent systems: ethyl acetate-methanol-water (12:1:1) for the N-tri- and N-difluoroacetyl derivatives (5b and 5c) and ethyl acetate-methanol (1:1) for the N-monofluoroacetyl derivative (5a). Products were recrystallized from ethanol and diethyl ether to yield the desired products with the overall yields over two steps of 66% for the N-trifluoroacetyl derivative (5c), 37% for the N-difluoroacetyl derivative (5b), and 45% for the N-fluoroacetyl derivative (5a).
4-Methylumbelliferyl 2-deoxy-2-fluoroacetamido-β-D-glucopyranoside (MuGlcNAc-F) (5a)—1H-NMR (500 MHz, d6-DMSO) δ 7.71 (1H, d, JH5AR-H6AR=8.8, H-5AR), 7.00 (1H, JH8AR-H6AR=2.4, H-8AR), 6.96 (1H, dd, H-6AR), 6.26 (1H, d, JH3AR-CH3=1.1, H-3AR), 5.21 (1H, d, JH1-H2=8.5, H-1), 4.81 (2H, d, JH-F=47.0, CH2F), 3.86 (1H, dd, JH2-H3=9.9, H-2), 3.74 (1-H, dd, JH6-H6′=11.7, JH6-H5=1.7, H-6), 3.53-3.46 (2H, m, H-3, H-6′), 3.38 (1H, ddd, JH5-H4=9.6, JH5-H6′=6.0, H-5), 3.20 (1H, dd, JH4-H3=8.9, H-4), 2.39 (1H, d, CH3) ppm; 19F-NMR (500 MHz, d6-DMSO) −225.24 ppp (dd, JF-H=53). Analytical calculated for C18H20FNO8; C, 54.41; H, 5.07; N, 3.53; Experimental C, 54.20; H, 4.97; N, 3.59.
4-Methylumbelliferyl 2-deoxy-2-difluoroacetamido-β-D-glucopyranoside (MuGlcNAc-F2) (5b)—1H-NMR (500 MHz, d6-DMSO) δ 7.69 (1H, d, JH5AR-H6AR=8.8, H-5AR), 6.97 (1H, JH8AR-H6AR=2.4, H-8AR), 6.92 (1H, dd, H-6AR), 6.24 (1H, d, JH3AR-CH3=1.2, H-3AR), 6.21 (1H, d, JH-F=53.6 Hz, CHF2), 5.16 (1H, d, JH1-H2=8.5, H-1), 3.79 (1H, dd, JH2-H3=10.3, H-2), 3.72 (1-H, dd, JH6-H6=11.8, JH6-H5=1.9, H-6), 3.53-3.42 (2H, m, H-3, H-5, H-6′), 3.20 (1H, dd, JH4-H5=9.6 Hz, JH4-H3=8.9 Hz, H-4), 2.38 (1H, d, CH3) ppm; 19F-NMR (500 MHz, d6-DMSO) −127.32 (d, J=54 Hz); Analytical calculated for C18H19F2NO8; C, 52.05; H, 4.61; N, 3.37; Experimental C, 51.92; H, 4.62; N, 3.31.
4-Methylumbelliferyl 2-deoxy-2-trifluoroacetamido-β-D-glucopyranoside (MuGlcNAc-F3) (5c)—1H-NMR (500 MHz, d6-DMSO) δ 7.70 (1H, d, JH5AR-H6AR=8.8, H-5AR), 6.86 (1H, JH8AR-H6AR=2.3, H-8AR), 6.91 (1H, dd, H-6AR), 6.25 (1H, d, JH3AR-CH3=1.2, H-3AR), 5.16 (1H, d, JH1-H2=8.5, H-1), 3.80 (1H, dd, JH2-H3=10.2, H-2), 3.72 (1-H, dd, JH6-H6′=11.7, JH6-H5=1.8, H-6), 3.50-3.42 (2H, m, H-6′, H-3, H-5), 3.22 (1H, dd, JH4-H5=9.5 Hz, JH4-H3=8.9 Hz, H-4), 2.38 (1H, d, CH3) ppm; 19F-NMR (500 MHz, d6-DMSO) −77.29; Analytical calculated for C18H18F3NO8; C, 49.89; H, 4.19; N, 3.23; Experimental C, 49.80; H, 4.29; N, 3.11.
All assays were carried out in triplicate at 37° C. for 30 minutes using a stopped assay procedure in which the enzymatic reactions (25 μL) were quenched by the addition of a 6 fold excess (150 μL) of quenching buffer (200 mM glycine, pH 10.75). Assays were initiated by the addition, via syringe, of enzyme (3 μL), and in all cases the final pH of the resulting quenched solution was greater than 10. Time dependent assay of β-hexosaminidase and O-GlcNAcase revealed that both enzymes were stable over this period in their respective buffers; 50 mM citrate, 100 mM NaCl, 0.1% BSA, pH 4.25 and 50 mM NaH2PO4, 100 mM NaCl, 0.1% BSA, pH 6.5. The progress of the reaction at the end of thirty minutes was determined by measuring the extent of 4-methylumbelliferone liberated as determined by fluorescence measurements using a Varian CARY Eclipse Fluorescence-Spectrophotometer 96-well plate system and comparison to a standard curve of 4-methylumbelliferone under identical buffer conditions. Excitation and emission wavelengths of 368 and 450 nM were used, respectively, with 5 mm slit openings. Human placental O-hexosaminidase was purchased from Sigma-Aldrich (Lot 043K3783). The cloning and expression of O-GlcNAcase is described in the literature[85]. Both enzymes were dialyzed against PBS buffer and their concentrations determined using the Bradford assay. The concentration (μg/μl) of β-hexosaminidase and O-GlcNAcase used in assays with fluorinated substrates were as follows: for 4-methylumbelliferyl 2-acetamido-2-deoxy-β-D-glucopyranoside (MuGlcNAc) (5): 0.00077, 0.0126; MuGlcNAc-F (5a): 0.0031, 0.0189; MuGlcNAc-F2 (5b): 0.0154, 0.0756, and for MuGlcNAc-F3 (5c): 0.0154, 0.01523. In addition, β-hexosaminidase and O-GlcNAcase were used at a concentration (μg/μL) of 0.0154 and 0.0378, respectively to test the inhibitors using substrate 5 at a concentration of 0.64 mM. All inhibitors were tested at eight concentrations ranging from 5 times to ⅕th KI with the exception of the assay of inhibitor 8e with β-hexosaminidase, where such high concentrations of inhibitor could not be reached owing to the high KI value of 8e. KI values were determined by linear regression of data in Dixon plots. Where necessary, assays were carried out in triplicate and error bars are included in plots of the data.
These compounds were first tested with the lysosomal human β-hexosamindase since this enzyme is known to proceed via a mechanism involving anchimeric assistance (
aThe Taft electronic parameters (σ*) used for each N-acyl substituent were obtained from Hansch and Leo[86].
bValues were estimated by non-linear regression of the Michelis-Menten data. Note that substrate concentrations assayed matched but did not exceed KM due to limited substrate solubility.
cThese values could not be determined as saturation kinetics were not observed owing to limitations in substrate solubility.
dValues were determined by linear regression of the second order region of the Michelis-Menten plot.
This data is consistent with a mechanism involving electrophilic migration of the anomeric center and consequent oxocarbenium-ion like transition states that have been generally proposed for enzyme-catalyzed glycoside hydrolysis[80, 82]. For O-GlcNAcase, Michaelian saturation kinetics were observed for all four substrates, therefore, both kinetic parameters were determined (Table 1,
The reaction constant (ρ), which is the slope of the correlations in the Taft-like analyses, is an indication of the sensitivity of the reaction to different substituents. This constant can be considered a function of both an electronic component (ρ*, which is governed by sensitivity of the reaction to the electronic parameter of the substituents, σ*) and a steric component (δ, which is governed by sensitivity of the reaction to the steric Taft parameters of the substituent, Es) according to the following equation.
ρ=ρ*+δ (Eqn 1)
The difference between the slopes measured for lysosomal β-hexosaminidase and O-GlcNAcase may thus reflect the position of the transition state along the reaction coordinate or may indicate that the lysosomal β-hexosaminidase has a more sterically constrained active site architecture than does O-GlcNAcase. Indeed, a common misconception is that fluorine (147 pm Van der Waals radius and 138 pm C—F bond length) is often considered to have a negligible difference in size as compared to hydrogen (120 pm Van der Waals radius and 109 pm C—H bond length). Therefore, it is possible that unfavorable steric interactions between the substrate and the active site of human β-hexosaminidase may play an additional role, beyond electronics, in discriminating between the varying levels of fluorine substitution. Indeed, the recent crystal structures of human hexosaminidase B revealed a carefully structured pocket that tightly nestles the acetamido group between three tryptophan residues[78, 87]. For O-GlcNAcase, no three-dimensional structure is available. The relatively constant KM values measured for all of the substrate analogues suggest, however, that steric effects are not a major contributor and that the electronic effect of the N-acyl-fluorine substituents predominate. An earlier study on an isolated enzyme of unknown family using two para-nitrophenyl 2-acetamido-2-deoxy-β-D-glucopyanoside (pNP-GlcNAc) with either two or three fluorine substituents in the acetamido group yielded a p* of −1.41±0.1.[88] This value is even greater than that found for the lysosomal β-hexosaminidase in this study (ρ*=−1.0±0.1) suggesting that the enzyme Kosman and Jones studied from Aspergillus niger is more likely a member of family 20 of glycoside hydrolases than family 84.
As a further test of whether O-GlcNAcase uses a catalytic mechanism involving anchimeric assistance, the inhibitor NAG-thiazoline (9a) was tested with this enzyme. NAG-thiazoline, designed as a mimic of the bicyclic oxazoline intermediate, has been previously demonstrated to function as an inhibitor of family 20 hexosaminidases[30, 48]. Using pNP-GlcNAc as a substrate, NAG-thiazoline was found to be a potent inhibitor of family 84 human O-GlcNAcase and a clear pattern of competitive inhibition was observed (
With the mechanism of human O-GlcNAcase, and by extension other members of family 84 of glycoside hydrolases, established, attention was turned to using this information in the design of inhibitors that would be selective for this enzyme over the human lysosomal β-hexosaminidase. Because both β-hexosaminidase and O-GlcNAcase use a mechanism involving anchimeric assistance, NAG-thiazoline was chosen as a scaffold that could be elaborated to generate the required selectivity. Three observations provided a starting point in the design of the inhibitor. The first is that the slope of the Taft-like analysis for the lysosomal enzyme is much steeper than that measured for O-GlcNAcase thereby suggesting that the bulk of the N-acyl group may be a determinant in substrate recognition (vide supra). The second, and related, consideration is that the structure of the human lysosomal β-hexosaminidase B reveals a snug pocket into which the methyl group of the acetamido substituent is poised[78]. The third is that STZ, which bears a bulky N-acyl substituent shows some selectivity for O-GlcNAcase over β-hexosaminidase[46].
A series of seven inhibitors were prepared in which the thiazoline ring was elaborated with aliphatic chains of increasing length in the expectation that these compounds would allow the discriminative inhibition of O-GlcNAcase over lysosomal hexosaminidase. The synthesis of this panel of inhibitors is outlined in Scheme 2. This facile synthetic route enables the production of quantities of inhibitor from commercially available starting materials in three steps or from the inexpensive starting material 2-amino-2-deoxy-glucopyranose in six steps.
The general procedures are the same as for the synthesis of the substrate analogues set out above.
To a solution of the hydrochloride salt of 2-amino-2-deoxy-1,3,4,6-tetra-O-acetyl-β-D-glucopyranose (6)[92] in 1 volume of dry dichloromethane was added 2 equivalents of dry triethylamine at which time the starting material dissolved. The reaction mixture was cooled to 0° C. and 1.2 equivalents of the appropriate acyl chloride was added via syringe. The resultant mixture was stirred for approximately two hours at room temperature. When the reaction mixture was judged complete by TLC analysis, 5 volumes of ethyl acetate were added. The resulting organic phase was washed successively with water, 1 M NaOH, and saturated sodium chloride. The organic phase was dried over MgSO4, filtered, and concentrated to yield a white crystalline solid. The material was recrystallized using a mixture of ethyl acetate and hexanes to yield the desired N-acylated materials in yields ranging from 46 to 74%.
1,3,4,6-tetra-O-acetyl-2-deoxy-2-N-propyl-β-D-glucopyranose (7b)—1H NMR (500M Hz, CDCl3) δ: 5.68 (1H, d, JH1-H2=8.8 Hz, H-1), 5.42 (1H, d, JNH,H2=6.7 Hz, NH), 5.14 (1H, dd, JH3,H4=8.7 Hz, H-3), 5.12 (1H, dd, JH,H5=8.7 Hz, H-4), 4.34 (1H, ddd, JH2,H3=8.7 Hz, H-2), 4.27 (1H, dd, JH6,H6′=12.5 Hz, JH5,H6=4.6 Hz, H-6), 4.12 (1H, dd, JH5,H6′=2.1 Hz, H-6′), 3.79-3.75 (1H, m, H-5), 2.12 (3H, s, OAc), 2.10 (3H, s, OAc), 2.04 (3H, s, OAc), 2.02 (3H, s, OAc), 2.15-2.10 (2H, m, H-7), 0.90 (3H, t JH8,H7=7.4 Hz, H-8) ppm. 13C NMR (125 MHz, MeOD) δ: 174.1, 171.5, 170.9, 169.8, 169.5, 92.9, 73.2, 72.8, 67.9, 61.8, 53.1, 30.0, 21.1, 21.0, 20.9, 20.8, 10.1 ppm.
1,3,4,6-tetra-O-acetyl-2-deoxy-2-N-butyl-β-D-glucopyranose (7c)—1H NMR (500 MHz, CDCl3) δ: 5.73 (1H, d, JH1,H2=8.8 Hz, H-1), 5.58 (1H, d, JNH,H2=9.5 Hz, NH), 5.21-5.14 (2H, m, H-3/H-4), 4.38 (1H, ddd, H-2), 4.31 (1H, dd, JH6,H6′=12.5 Hz, H-6),4.16 (1H, dd, JH5,H6′=2.2 Hz, H-6′), 3.84 (1H, ddd, JH5,H6=4.7 Hz, H-5), 2.16 (3H, s, OAc), 2.13 (3H, s, OAc), 2.18-2.10 (2H, m, H-7), 2.08 (3H, s, OAc), 2.05 (3H, s, OAc), 1.64 (2H, ddd, H-8), 0.94 (3H, t, JH8,H9=7.4 Hz, H-9) ppm. 13C NMR (125 MHz, MeOD) δ: 173.1, 171.4, 170.9, 169.8, 169.5, 92.9, 73.2, 72.8, 67.9, 61.9, 53.0, 38.8, 21.1, 21.0, 20.9, 20.8, 19.2, 13.7 ppm.
1,3,4,6-tetra-O-acetyl-2-deoxy-2-N-pentyl-β-D-glucopyranose (7d)—1H NMR (500 MHz, CDCl3) δ: 5.68 (1H, d, JH1,H2=8.8 Hz, H-1), 5.50 (1H, d, JNH,H2=9.5 Hz, NH), 5.18-5.12 (2H, m, H-3/H-4), 4.36-4.30 (1H, m, H-2), 4.26 (1H, dd, JH6,H6′=12.5 Hz, H-6), 4.13 (1H, dd, JH5,H6′=2.2 Hz, H-6′), 3.82-3.77 (1H, m, JH5,H6=4.5 Hz, H-5), 2.11 (3H, s, OAc), 2.11 (2H, m, H-7), 2.09 (3H, s, OAc), 2.04 (3H, s, OAc), 2.02 (3H, s, OAc), 1.56 (2H, dd, JH7,H8=7.5 Hz, H-8), 1.52 (2H, dd, JH7′,H8=7.5 Hz, H-8), 1.26 (2H, ddq, JH8,H9=7.5 Hz, H-9), 0.94 (3H, dd, JH9,H10=7.5 Hz, H-10) ppm. 13C NMR (125 MHz, MeOD) δ: 173.4, 171.4, 170.9, 169.8, 169.5, 92.9, 73.2, 72.7, 67.9, 61.9, 53.0, 36.7, 27.9, 22.4, 21.1, 21.0, 20.9, 20.8, 13.9 ppm.
1,3,4,6-tetra-O-acetyl-2-deoxy-2-N-hexyl-β-D-glucopyranose (7e)—1H NMR (500 MHz, CDCl3) δ: 5.69 (1H, d, JH1,H2=8.8 Hz, H-1), 5.52 (1H, d, JNH,H2=9.5 Hz, NH), 5.14 (2H, m, H-3/H-4), 4.37-4.30 (1H, m, H-2), 4.26 (1H, dd, JH6,H6′=12.5 Hz, H-6), 4.15-4.08 (1H, m, H-6′), 3.82-3.77 (1H, m, JH5,H6=4.7 Hz, H-5), 2.11 (3H, s, OAc), 2.09 (3H, s, OAc), 2.04 (3H, s, OAc), 2.02 (3H, s, OAc), 1.58-1.52 (2H, m, H-7), 1.33-1.08 (6H, m, H-8, H-9, H-10), 0.87 (3H, t, JH11,H10=7.1 Hz, H-11) ppm. 13C NMR (125 MHz, MeOD) δ: 173.4, 171.4, 170.9, 169.8, 169.5, 92.9, 73.2, 72.8, 67.9, 61.9, 60.6, 53.0, 36.9, 31.4, 25.5, 22.5, 21.1, 21.0, 20.9, 20.8, 14.1 ppm.
1,3,4,6-tetra-O-acetyl-2-deoxy-2-N-isobutyl-β-D-glucopyranose (7f)—1H NMR (500 MHz, CDCl3) δ: 5.70 (1H, d, JH1,H2=8.8 Hz, H-1), 5.58 (1H, d, JNH,H2=9.6 Hz, NH), 5.18-5.12 (2H, m, H-3/H-4), 4.36-4.30 (1H, m, H-2), 4.27 (1H, dd, JH6,H6′=12.5 Hz, H-6), 4.12 (1H, dd, JH5,H6′=2.2 Hz, H-6′), 3.84-3.87 (1H, m, JH5,H6=4.8 Hz, H-5), 2.10 (3H, s, OAc), 2.08 (3H, s, OAc), 2.28 (1H, m, H-7), 2.04 (3H, s, OAc), 2.03 (3H, s, OAc), 1.08 (6H, t, JH8,H7=2.8 Hz, H-8) ppm. 13C NMR (125 MHz, MeOD) δ: 177.2, 171.5, 170.9, 169.8, 169.5, 93.0, 73.3, 72.7, 67.9, 61.9, 52.9, 36.0, 21.0, 21.0, 20.8, 20.8, 19.6 ppm.
1,3,4,6-tetra-O-acetyl-2-deoxy-2-N-isopentyl-β-D-glucopyranose (7g)—1H NMR (500 MHz, CDCl3) δ: 5.67 (1H, d, JH1,H2=8.8 Hz, H-1), 5.58 (1H, d, JNH,H2=9.5 Hz, NH), 5.18-5.10 (2H, m, H-3/H-4), 4.38-4.32 (1H, m, H-2), 4.27 (1H, dd, JH6,H6′=12.5 Hz, H-6), 4.12 (1H, 1, JH5,H6′=2.2 Hz, H-6′), 3.82-3.78 (1H, m, JH5,H6=4.6 Hz, H-5), 2.10 (3H, s, OAc), 2.08 (3H, s, OAc), 2.06-2.01 (2H, m, H-7), 2.04 (3H, s, OAc), 2.03 (3H, s, OAc), 1.98 (1H, ddd, H-8), 0.94 (6H, d, JH8,H9=6.5 Hz, H-9) ppm. 13C NMR (125 MHz, MeOD) δ: 172.6, 171.4, 170.9, 169.8, 169.5, 92.9, 73.2, 72.7, 67.9, 61.8, 52.8, 46.2, 26.3, 22.4, 22.3, 21.1, 21.0, 20.9, 20.8 ppm.
To a solution of the appropriate 1,3,4,6-tetra-O-acetyl-2-N-acyl-2-deoxy-α-D-glucopyranose (7a-g) in anhydrous toluene was added Lawesson's Reagent (0.6 equivalents) and the reaction mixture was refluxed for 2 h after which time the reaction was judged to be complete by TLC analysis. The solution was cooled to room temperature and the solvent was removed in vacuo. The residue was dissolved in toluene and the desired material was isolated by flash column silica chromatography using a solvent system of hexanes and ethyl acetate in ratios ranging from 4:1 to 1:2 as appropriate. Products were isolated in yields ranging from 62 to 83%. 3,4,6-tri-O-acetyl-1,2-dideoxy-2′-ethyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline (8a) has been previously prepared using similar reaction conditions as described above.[30] All spectral characterization agreed with the literature values.
3,4,6-tri-1-acetyl-1,2-dideoxy-2′-ethyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline (8b)—1H NMR (500 MHz, CDCl3) δ: 6.28 (1H, d, JH1,H2=7.4 Hz, H-1), 5.60 (1H, dd, JH3,H4=3.3 Hz, H-3), 4.97 (1H, d, JH4,H5=9.3 Hz, H-4), 4.56-4.52 (1H, m, H-2), 4.18-4.10 (2H, m, H-6/H-6′), 3.58 (1H, ddd, JH5,H6=3.2 Hz, H-5), 2.74-2.67 (1H, m, H-7), 2.14 (3H, s, OAc), 2.11 (3H, s, OAc), 2.09 (3H, s, OAc), 1.29 (3H, t, JH8,H7=7.60 Hz, H-8) ppm.
3,4,6-tri-O-acetyl-1,2-dideoxy-2′-propyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline (8c)—1H NMR (500 MHz, CDCl3) δ: 6.26 (1H, d, JH1,H2=7.2 Hz, H-1), 5.58 (1H, dd, JH3,H4=3.3 Hz, H-3), 4.96 (1H, d, JH4,H5=9.2 Hz, H-4), 4.54-4.50 (1H, m, H-2), 4.16-4.08 (2H, m, H-6/H-6′), 3.58 (1H, ddd, JH5,H6=3.3 Hz, H-5), 2.70-2.58 (2H, m, H-7), 2.14 (3H, s, OAc), 2.08 (3H, s, OAc), 2.07 (3H, s, OAc), 1.76-1.69 (2H, m, H-8), 1.00 (3H, t, JH9,H8=7.4 Hz, H-9) ppm.
3,4,6-tri-O-acetyl-1,2-dideoxy-2′-butyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline (8d)—1H NMR (500 MHz, CDCl3) δ: 6.21 (1H, d, JH1,H2=7.2 Hz, H-1), 5.57 (1H, dd, JH3,H4=3.3 Hz, H-3), 4.94 (1H, d, JH4,H5=9.4 Hz, H-4), 4.48-4.44 (1 H, m, H-2), 4.12-4.07 (2H, m, H-6/H-6′), 3.53 (1H, ddd, JH5,H6=3.0 Hz, H-5), 2.60-2.57 (2H, m, H-7), 2.12 (3H, s, OAc), 2.07 (6H, s, OAc), 1.67-1.63 (2H, m, H-8), 1.40 (2H, ddd, JH9,H8=7.3 Hz, H-9), 0.92 (3H, t, JH10,H9=7.4 Hz, H-10) ppm.
3,4,6-tri-O-acetyl-1,2-dideoxy-2′-pentyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline (8e)—1H NMR (500 MHz, CDCl3) δ: 6.24 (1H, d, JH1,H2=7.2 Hz, H-1), 5.60 (1H, s, H-3), 4.96 (1H, d, JH4,H5=9.4 Hz, H-4), 4.52-4.49 (1H, m, H-2), 4.14-4.10 (2H, m, JH6,H6′=12.2 Hz H-6/H-6′), 3.56 (1H, ddd, JH5,H6=3.1 Hz, JH5,H6′=5.6 Hz, H-5), 2.65-2.60 (2H, m, H-7), 2.14 (3H, s, OAc), 2.09 (3H, s, OAc), 2.08 (3H, s, OAc), 1.71-1.68 (2H, m, H-8), 1.38-1.33 (4H, m, H-9/H-10), 0.91 (3H, t, JH11,H10=6.9 Hz, H-11) ppm.
3,4,6-tri-O-acetyl-1,2-dideoxy-2′-isopropyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline (8f)—1H NMR (500 MHz, CDCl3) δ: 6.27 (1H, d, JH1,H2=7.2 Hz, H-1), 5.60 (1H, m, H-3), 4.97 (1H, d, JH4,H5=9.3 Hz, H-4), 4.55-4.49 (1H, m, H-2), 4.18-4.11 (2H, m, H-6/H-6′), 3.60 (1H, ddd, JH5,H6=3.1 Hz, H-5), 2.55 (2H, s, H-7), 2.15 (3H, s, OAc), 2.09 (3H, s, OAc), 2.08 (3H, s, OAc), 1.07 (6H, t, JH8,H7=6.6 Hz, H-8) ppm.
3,4,6-tri-O-acetyl-1,2-dideoxy-2′-isobutyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline (8g)—1H NMR (500 MHz, CDCl3) δ: 6.26 (1H, d, JH1,H2=7.2 Hz, H-1), 5.60 (1H, dd, JH3,H4=3.3 Hz, H-3), 4.97 (1H, d, JH4,H5=9.3 Hz, H-4), 4.56-4.50 (1H, m, H-2), 4.16-4.10 (2H, m, H-6/H-6′), 3.62-3.58 (1H, m, JH5,H6=3.1 Hz, H-5), 2.57-2.52 (2H, m, H-7), 2.15 (3H, s, OAc), 2.09 (3H, s, OAc), 2.08 (3H, s, OAc), 1.28-1.23 (2H, m, H-8), 1.02 (6H, t, JH9,H8=6.6 Hz, H-9) ppm.
To a solution of the appropriate protected thiazoline (8a-g) in dry methanol was added a spatula tip of anhydrous sodium methoxide. The basic solution was stirred until the reaction was judged complete by TLC analysis (typically 2 hours). A solution of glacial acetic acid in methanol (1:20) was added dropwise to the reaction mixture until the pH of the solution was found to be neutral. The solvent was then removed in vacuo and the desired materials (9a-g) was isolated by flash column silica chromatography using a solvent system of ethyl acetate and methanol in ratios ranging from 2:1 to 6:1 as appropriate. Products were isolated in yields ranging from 86 to 99%. 1,2-dideoxy-2′-ethyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline (9a) has been previously prepared using similar reaction conditions as described above[30]. All spectral characterization agreed with the literature values as did the elemental analysis of the sample used in these assays. Analytical calculated for C8H13O4NS; C, 43.82; H, 5.98; N, 6.39; Experimental C, 43.45; H, 6.23; N, 6.18.
1,2-dideoxy-2′-ethyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline (9b)—1H NMR (500 MHz, MeOD) δ: 6.31 (1H, d, JH1,H2=7.0 Hz, H-1), 4.29-4.26 (1H, m, JH2,H3=4.32 Hz, H-2), 4.09 (1H, dd, JH3,H4=4.32 Hz, H-3), 3.70 (1H, dd, JH6,H6′=12.07 Hz, H-6), 3.57 (1H, dd, H-6′) 3.52 (1H, dd, JH4,H5=9.10 Hz, H-4), 3.32-3.29 (1H, m, JH5,H6=6.29 Hz, JH5,H6′=2.21 Hz, H-5), 2.52 (2H, dd, JH7,H8=7.59 Hz, H-7), 1.18 (3H, t, H-8) ppm. 13C NMR (125 MHz, MeOD) δ: 175.1, 89.4, 80.2, 75.7, 74.5, 71.4, 62.5, 28.9, 11.1 ppm; Analytical calculated for C9H15O4NS; C, 46.34; H, 6.48; N, 6.00; Experimental C, 45.95; H, 6.33; N, 5.93.
1,2-dideoxy-2′-propyl-α-D-glucopyranoso-[2,1-d]-2′-thiazoline (9c)—1H NMR (500 MHz, MeOD) δ: 6.32 (1H, d, JH1,H2=7.0 Hz, H-1), 4.28 (1H, m, JH2,H3=4.4 Hz, H-2), 4.09 (1H, dd, JH3,H4=4.4 Hz, H-3), 3.72 (1H, dd, JH6,H6′=12.0 Hz, H-6), 3.59 (1H, dd, H-6′), 3.53 (1H, dd, JH4,H5=9.1 Hz, H-4), 3.33 (1H, m, JH5,H6=6.3 Hz, JH5,H6′=2.5 Hz, H-5), 2.51-2.48 (1H, m, JH7,H8=7.3 Hz, JH7,H7′=14.9 Hz, H-7), 2.49-2.47 (1H, m, H-7′), 1.68-1.65 (2H, ml JH8,H9=7.4 Hz, H-8) 0.98 (3H, dd, H-9); Analytical calculated for C10H17O4NS; C, 48.57; H, 6.93; N, 5.66; Experimental C, 48.32; H, 6.77; N, 5.45.
1,2-dideoxy-2′-butyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline (9d)—1H NMR (500 MHz, MeOD) δ: 6.31 (1H, d, JH1,H2=7.0 Hz, H-1), 4.30-4.28 (1H, m, JH2,H3=4.4 Hz, H-2), 4.09 (1H, dd, JH3,H4=4.4 Hz, H-3), 3.71 (1H, dd, JH6,H6′=12.0 Hz, H-6), 3.59 (1H, dd, H-6′), 3.53 (1H, dd, JH4,H5=9.1 Hz, H-4), 3.35-3.30 (1H, m, JH5,H6=6.3 Hz, JH5,H6′=2.5 Hz, H-5), 2.58-2.53 (1H, m, JH7,H8=7.8 Hz, JH7,H7′=14.8 Hz, H-7) 2.53-2.50 (1H, m, JH7′,H8=7.6 Hz, H-7′) 1.60 (2H, ddd, JH8,H9=14.9 Hz, H-8) 1.37 (2H, ddd, JH9,H10=7.4, H-9) 0.92 (3H, t, H-10) ppm. 13C NMR (125 MHz, MeOD) δ: 175.5, 89.4, 79.1, 75.2, 74.2, 71.5, 62.7, 36.9, 20.4, 15.2 ppm; Analytical calculated for C11H19O4NS; C, 50.55; H, 7.33; N, 5.36; Experimental C, 50.68; H, 7.12; N, 5.13.
1,2-dideoxy-2′-pentyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline (9e)—1H NMR (500 MHz, MeOD) δ: 6.20 (1H, d, JH1,H2=6.9 Hz, H-1), 4.29-4.27 (1H, m, JH2,H3=6.0 Hz, H-2), 4.09 (1H, dd, JH3,H4=4.4 Hz, H-3), 3.72 (1H, dd, JH6,H6=12.0 Hz, H-6), 3.59 (1H, dd, H-6′), 3.53 (1H, dd, JH4,H5=9.1 Hz, H-4), 3.35-3.31 (1H, m, JH5,H6=6.3, JH5,H6′=2.4 Hz, H-5), 2.52 (2H, ddd, JH7,H8=6.2 Hz, H-7), 1.37-1.32 (2H, m, H-8), 1.35-1.30 (2H, m, H-9), 1.66-1.63 (2H, m, JH10,H11=7.1 Hz, H-10), 0.90 (3H, t, H-11) ppm. 13C NMR (125 MHz, MeOD) δ: 174.8, 89.3, 79.2, 75.5, 73.4, 71.5, 62.2, 35.4, 30.3, 28.8, 22.1, 13.1 ppm; Analytical calculated for C12H21O4NS; C, 52.34; H, 7.69; N, 5.09; Experimental C, 52.48; H, 7.67; N, 4.40.
1,2-dideoxy-2′-isopropyl-α-D-glucopyranoso-[2,1-d]-Δ2′-thiazoline (9f)—1H NMR (500 MHz, MeOD) δ: 6.31 (1H, d, JH1,H2=7.0 Hz, H-1), 4.29 (1H, m, JH2,H3=4.4 Hz, H-2), 4.11 (1H, dd, JH3,H4=4.4 Hz, H-3), 3.60 (1H, dd, JH6,H6′=12.0 Hz, H-6), 3.72 (1H, dd, H-6′), 3.55 (1H, dd, JH4,H5=9.1 Hz, H-4), 3.35-3.32 (1H, m, JH5,H6=6.4 Hz, JH5,H6′=2.4 Hz, H-5), 2.87-2.83 (2H, m, JH7,H8=6.9 Hz, H-7), 1.24 (3H, d, H-8), 1.21 (3H, d, H-8′) ppm. 13C NMR (125 MHz, MeOD) δ: 175.2, 89.5, 79.7, 75.6, 72.3, 70.8, 62.3, 35.1, 30.3, 22.5, 13.8 ppm; Analytical calculated for C10H17O4NS; C, 48.57; H, 6.93; N, 5.66; Experimental C, 48.40; H, 6.70; N, 5.33.
1,2-dideoxy-2′-isobutyl-α-D-glucopyranoso-[2,1-d]-42′-triazoline D (9g)—1H NMR (500 MHz, MeOD) δ: 6.38 (1H, d, JH1,H2=7.1 Hz, H-1), 4.28-4.24 (1H, m, JH2,H3=6.0 Hz, H-2), 4.06 (1H, dd, JH3,H4=6.0 Hz, H-3), 3.71 (1H, dd, JH6,H6′=12.0 Hz, H-6), 3.58 (1H, dd, H-6′), 3.58 (1H, dd, JH4,H5=9.2 Hz, H-4), 3.35-3.30 (1H, m, JH5,H6=6.3 Hz, JH5,H6′=2.4 Hz, H-5), 2.46-2.40 (1H, m, JH7,H8=7.3 Hz, JH7,H7′=14.1 Hz, H-7) 2.37-2.33 (1H, m, JH7′,H8=7.3 Hz, H-7′) 2.00 (2H, ddd, JH8,H9=6.7 Hz, H-8) 0.97 (3H, d, H-9) ppm; Analytical calculated for C, 11H19O4NS; C, 50.55; H, 7.33; N, 5.36; Experimental C, 50.68; H, 7.12; N, 5.13.
All assays were carried out in triplicate at 37° C. for 30 minutes using a stopped assay procedure in which the enzymatic reactions (25 μL) were quenched by the addition of a 6 fold excess (150 μL) of quenching buffer (200 mM glycine, pH 10.75). Assays were initiated by the addition, via syringe, of enzyme (3 μL), and in all cases the final pH of the resulting quenched solution was greater than 10. Time dependent assay of β-hexosaminidase and O-GlcNAcase revealed that both enzymes were stable over this period in their respective buffers; 50 mM citrate, 100 mM NaCl, 0.1% BSA, pH 4.25 and 50 mM NaH2PO4, 100 mM NaCl, 0.1% BSA, pH 6.5. The progress of the reaction at the end of thirty minutes was determined by the measuring the extent of 4-methylumbelliferone liberated as determined by fluorescence measurements using a Varian CARY Eclipse Fluorescence-Spectrophotometer 96-well plate system and comparison to a standard curve of 4-methylumbelliferone under identical buffer conditions. Excitation and emission wavelengths of 368 and 450 nM were used, respectively, with 5 mm slit openings. The possible time dependent inactivation of O-GlcNAcase was assayed by incubating 10 mM STZ with 0.016 mg/mL β-GlcNAcase in the presence of 50 mM NaH2PO4, 100 mM NaCl, 1% BSA, 5 mM β-mercaptoethanol, pH 6.5 or 0.036 mg/mL β-hexosaminidase in the presence of 50 mM citrate, 100 mM NaCl, 0.1% BSA, pH 4.25. At several time intervals the residual enzyme activity contained in the inactivation mixture was assayed. Assays were carried out as described above for each enzyme except that reactions were initiated by the addition of an aliquot of the reaction mixture into the assay mixture containing 5.7 mM MU-GlcNAc and the appropriate buffer for each enzyme. The stability of STZ was first tested by following its decomposition in deuterated water by NMR. The half-life of STZ in aqueous solution at room temperature was well greater than 6 hours as determined by following its decomposition by NMR at room temperature. Human placental β-hexosaminidase was purchased from Sigma-Aldrich (Lot 043K3783). The cloning and expression of O-GlcNAcase is described in the literature[85]. Both enzymes were dialyzed against PBS buffer and their concentrations determined using the Bradford assay. The concentration (μg/μl) of β-hexosaminidase and O-GlcNAcase used in assays with fluorinated substrates were as follows: for 4-methylumbelliferyl 2-acetamido-2-deoxy-β-D-glucopyranoside (MuGlcNAc) (5): 0.00077, 0.0126; MuGlcNAc-F (5a): 0.0031, 0.0189; MuGlcNAc-F2 (5b): 0.0154, 0.0756, and for MuGlcNAc-F3 (5c): 0.0154, 0.01523. In addition, β-hexosaminidase and O-GlcNAcase were used at a concentration (μg/μL) of 0.0154 and 0.0378, respectively to test the inhibitors using substrate 5 at a concentration of 0.64 mM. All inhibitors were tested at eight concentrations ranging from 5 times to ⅕th KI with the exception of the assay of inhibitor 8e with β-hexosaminidase, where a such high concentrations of inhibitor could not reached owing to the high KI value of 8e. KI values were determined by linear regression of data in Dixon plots. Where necessary, assays were carried out in triplicate and error bars are included in plots of the data.
Analysis of the inhibition of human β-hexosaminidase reveals that increasing the chain length resulted in a marked decrease in the potency of these inhibitors (
Existing inhibitors of O-GlcNAcase were also tested for their inhibitory properties and selectivities. The KI values for STZ (Table 2) with both O-GlcNAcase (KI=1.5 mM) and β-hexosaminidase (KI=47 mM) were determined and the value measured for O-GlcNAcase was found to be consistent with previous determinations of the IC50 value that were in the range of 1 to 2.5 mM[42, 46]. The selectivity of STZ for O-GlcNAcase over β-hexosaminidase is surprisingly modest (31-fold) given the bulk of the N-acyl group of this compound. Perhaps the thiazoline compounds demonstrate greater selectivity than STZ by virtue of the fact that they may emulate a transition state or tightly bound intermediate. The possible STZ induced irreversible inactivation of O-GlcNAcase and β-hexosaminidase was also investigated. Irreversible inhibitors result in the time dependent loss of enzyme activity as the inactivator modifies the protein. To first check the stability of STZ in aqueous solution and the purity of the commercially available material, the time dependent decomposition of STZ freshly dissolved in deuterated water was monitored by NMR. STZ decomposed over time with a half-life clearly greater than 6 hours (see
These data also support the view, vide supra, that the steric bulk of the N-acyl group of the fluorine-substituted substrates does not contribute greatly to the slope of the Taft-like analysis (ρ=−0.42) measured with O-GlcNAcase. The slope (ρ=−1.0) measured with lysosomal β-hexosaminidase, however, appears likely to be a composite of both electronic and steric effects according to equation 1. The sensitivity of the reaction to the electronic effect of the fluorine substituents (ρ*) may therefore be the same for both enzymes but the significant sensitivity of the lysosomal β-hexosaminidase catalyzed reaction to the steric effect of the substrates (δ) results in an apparently steeper slope for β-hexosaminidase than that measured for O-GlcNAcase. Together, the Taft-like linear free energy analysis and the selective inhibition data suggest that the active site of O-GlcNAcase has considerably more space in the region surrounding the 2-acetamido group of the substrate than does lysosomal β-hexosaminidase.
Based on the above observations, modifications in the pendant N-acyl chain of PUGNAc were made to yield a second embodiment of potent and selective inhibitors for O-GlcNAcase based on a different inhibitor scaffold. Such inhibitors may have different pharmacokinetic properties and be valuable tools in dissecting the role of the O-GlcNAc post-translational modification at the cellular and organismal level.
The synthesis of a series of six inhibitors using PUGNAc as a scaffold is outlined in Scheme 3.
Starting from the readily accessible hydrochloride 10, the Boc group was introduced to protect the amine moiety to give the tetraacetate 11[94] (Scheme 3). With the tetraacetate 11 in hand, selective de-O-acetylation using (NH4)2CO3 afforded the hemiacetal 12 in excellent yield. Treating this hemiacetal 12 with NH2OH.HCl yields the crude E and Z oximes 13 in good yield.
The next reaction, an oxidative ring closure, has been reported by Mohan and Vasella to be quite temperamental (Scheme 4).
When oxidising the oxime 14 using 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) and N-chlorosuccinimide (NCS) careful control of the temperature of the reaction mixture during the addition of the NCS was critical to avoid formation of the undesired 1,4-lactone oxime 15.[95] Consistent with these observations several trial reactions with 14 show that the temperature is indeed critical to successfully produce the 1,5-lactone oxime 16 over the undesired 1,4-lactone oxime 15. Indeed, at slightly elevated temperatures (−20° C.) only the 1,4-lactone oxime 15 is obtained.
Referring back to Scheme 3, treating the oxime 13 with DBU/NCS under the literature conditions[95] provided a mixture of the desired 1,5-lactone 17 and the unwanted 1,4-lactone oxime 18 in a 4:1 ratio. It was found, however, that having NCS already in solution with 13 and only then adding DBU to the mixture furnished exclusively 17, albeit in a somewhat lower yield. One complication encountered here is that succinimide, a by-product of these reactions, is difficult to separate from 17. Furthermore, spectral analysis of 17 is complicated due to broadening of the signals obtained in the 1H n.m.r. spectrum.
Nonetheless, treating impure hydroximolactone 17 with phenyl isocyanate yields pure carbamate 19 in good yield after purification by flash chromatography. Indeed, the Boc protecting group is smoothly removed using anhydrous trifluoroacetic acid in dichloromethane. Treating the resulting crude amine salt 20 with the appropriate acyl chloride yields amides 21a-g in good overall yields. Detailed analyses of 21a and 21b support the identities of this entire K-acylated series of compounds. Subsequent de-O-acetylation of 21a with saturated ammonia in methanol affords PUGNAc in good yield. 22a was obtained from 21b in a similar fashion. To highlight the ease of these conversions, the common intermediate, 20 can be treated with a range of acyl chlorides to provide 21c-g as crude products. Immediate de-O-acetylation of these crude intermediates readily furnishes the triols 22b-f; in good yield.
All solvents were dried prior to use. Synthetic reactions were monitored by TLC using Merck Kieselgel 60 F254 aluminium-backed sheets. Compounds were detected by charring with a 10% concentrated sulfuric acid in ethanol solution and heating. Flash chromatography under a positive pressure was performed with Merck Kieselgel 60 (230-400 mesh) using specified eluants. 1H and 13C NMR spectra were recorded on a Bruker AMX400 at 400 MHz (100 MHz for 13C) or a Varian AS500 Unity Innova spectrometer at 500 MHz (125 MHz for 13C) (chemical shifts quoted relative to CDCl3 or CD3OD where appropriate). Elemental analyses of all synthesized compounds used in enzyme assays were performed at the Simon Fraser University or the University of British Columbia Analytical Facility.
Hydroxylamine hydrochloride (3.2 g, 46 mmol) was added to the hemiacetal 12[93] (12 g, 31 mmol) and pyridine (6.3 mL, 77 mmol) in MeOH (200 mL) and the resulting solution stirred at reflux (2 h). The solution was concentrated and co-evaporated with toluene (2×20 mL). The residue was taken up in EtOAc and washed with water (2×50 mL), brine (50 mL), dried (MgSO4), filtered and concentrated to give the presumed oxime 13 (9.5 g). The residue was used without further purification.
(a) 1,8-Diazabicyclo[5.4.0]undec-7-ene (0.98 mL, 6.6 mmol) was added to the crude oxime 13 (2.5 g, 5.9 mmol) in CH2Cl2 (60 mL) at −45° C. and the mixture stirred (5 min). N-chlorosuccinimide (0.87 g, 6.5 mmol) was then added to the solution in such a way that the temperature did not go above −40° C. and the resulting mixture was allowed to stir for 30 minutes at this temperature and then was allowed to warm to room temperature over two hours. The mixture was quenched with water and diluted with EtOAc (100 mL). The organic layer was separated and washed with water (2×50 mL), brine (1×50 mL), dried (MgSO4) filtered and concentrated. Flash chromatography of the resultant residue (EtOAc/hexanes 2:3) gave the title compound 17 as a colourless oil (1.7 g, 68%). The 1H and 13C NMR spectra appeared to show the compound of interest but was contaminated with the 1,4-lactone oxime 18 and succinimide.
(b)) 1,8-Diazabicyclo[5.4.0]undec-7-ene (3.0 mL, 20 mmol) was added to the oxime 13 (7.7 g, 18 mmol) and N-chlorosuccinimide (2.7 g, 20 mmol) in CH2Cl2 (110 mL) at 45° C. in such a way that the temperature did not go above −40° C. and the resulting mixture was allowed to stir for 30 minutes at this temperature and then was allowed to warm to room temperature over two hours. The mixture subsequently treated as in (a) to give the title compound 17 (4 g, 52%†). Rf 0.16 (EtOAc/hexane 1:1); δH (500 MHz, CD3OD): 9.12 (br s, 1H, NOH), 5.33-5.18 (m, 3H, H3, H4, NH), 4.60-4.52 (m, 1H, H2), 4.33 (dd, J=3.5, 12.5, 1H, H6), 4.33 (dd, J=2.0, 1H, H6), 4.19 (ddd, J=9.5, 1H, H5), 2.08 (s, 3H, CH3), 2.03 (s, 3H, CH3), 2.02 (s, 3H, CH3), 1.40 (s, 9H, C(CH3)3). δC (125 MHz, CD3OD): 171.5, 170.5, 170.3, 169.0 (4C, C═O), 150.7 (C1), 76.5, 72.4, 67.3 (3C, C3, C4, C5), 61.5 (C6), 50.8 (C2), 29.6 (C(CH3)3), 28.2 (C(CH3)3), 20.7, 20.6, 20.5 (3C, CH3).
Phenyl isocyanate (0.5 mL, 3.7 mmol) was added to the lactone 17 (1.3 g, 3.1 mmol) and Et3N (1.3 mL, 9.3 mmol) in THF (50 mL) and the solution stirred (r.t., 3 h). Concentration followed by flash chromatography of the resultant residue (EtOAc/hexanes 1:4) yielded the carbamate 19 as a colourless oil (1.2 g, 71%). Rf 0.65 (EtOAc/hexane 1:1); OH (500 MHz, CDCl3): 7.83 (br s, 1H, PhNH), 7.42 (m, 2H, Ar), 7.32 (m, 2H, Ar), 7.11 (m, 1H, Ar), 5.38-5.30 (m, 2H, H3, H4), 5.18 (br s, 1H, NH), 4.62 (m, 1H, H2), 4.45-4.40 (m, 2H, H5, H6), 4.31 (dd, J=3.5, 13.5, 1H, H6), 2.12 (s, 3H, CH3), 2.09 (s, 3H, CH3), 2.07 (s, 3H, CH3), 1.44 (s, 9H, C(CH3)3). δC (125 MHz, CDCl3): 171.1, 170.3, 170.0, 169.1 (4C, C═O), 155.0 (CONHPh), 151.4 (C1), 136.8, 129.1, 124.2, 119.4 (4C, Ar), 77.2, 71.0, 67.2 (3C, C3, C4, C5), 61.2 (C6), 51.1 (C2), 30.6 (C(CH3)3), 28.2 (C(CH3)3), (20.7, 20.6, 20.5 (3C, CH3); Anal. calcd for C24H31N3O11: C, 53.63; H, 5.81; N, 7.82. Found: C, 53.72; H, 5.78; N, 7.78%.
Trifluoroacetic acid (13 mmol) was added to the carbamate 19 (1 mmol) in CH2Cl2 (10 mL) at 0° C. and the solution stirred (2 h). Pyridine (200 mmol) was then slowly added to the solution and the resulting mixture left to stand (0° C., 10 min). The appropriate acyl chloride (3 mmol) was then added at 0° C. and the solution allowed to stand at 4° C. overnight. Concentration of the mixture gave a yellowish residue which was dissolved in EtOAc (30 mL) and washed with (2×20 mL), brine (1×20 mL), dried (MgSO4) filtered and concentrated. For the presumed intermediate tri-O-acetates 21c-g these were carried through without further purification. Flash chromatography of the residues presumably 21a and 21b (EtOAc/hexanes 1:1) gave the desired acyl derivatives 21a and 21b in yields of 48% and 42% respectively.
O-2-acetamido-3,4,6-tri-O-acetyl-2-deoxy-
O-(3,4,6-Tri-O-acetyl-2-deoxy-2-propamido-
A saturated solution of ammonia in MeOH (2 mL) was added to the carbamate (0.3 mmol) in MeOH (10 mL) and the solution left to stand (r.t., 2 h). Concentration followed by flash chromatography of the residue (MeOH/EtOAc 3:97) gave the desired triols PUGNAc, 22a-f in yields ranging from 21% to 32%.
O-(2-Acetamido-2-deoxy-D-glucopyranosylidene)amino N-Phenylcarbamate (PUGNAc)—(32%) Gave 1H and 13C NMR spectra consistent with that found in the literature.[96] Rf 0.15 (MeOH/EtOAc 1:19)
O-2-deoxy-2-propamido-
O-(2-deoxy-2-butamido-
O-2-deoxy-2-valeramido-
O-(2-deoxy-2-hexamido-
O-(2-deoxy-2-isobutamido-
O-(2-deoxy-2-isovaleramido-
All assays were carried out in triplicate at 37° C. for 30 minutes by using a stopped assay procedure in which the enzymatic reactions (50 μL) were quenched by the addition of a 4-fold excess (200 μL) of quenching buffer (200 mM glycine, pH 10.75). Assays were initiated by the careful addition, via pipette, of enzyme (5 μL), and in all cases the final pH of the resulting quenched solution was greater than 10. Time-dependent assay of O-GlcNAcase and β-hexosaminidase revealed that enzymes were stable in their respective buffers over the period of the assay: 50 mM NaH2PO4, 100 mM NaCl, 0.1% BSA, pH 6.5 and 50 mM citrate, 100 mM NaCl, 0.1% BSA, pH 4.25. The progress of the reaction at the end of 30 minutes was determined by measuring the extent of 4-nitrophenol liberated as determined by UV measurements at 400 nm using a 96-well plate (Sarstedt) and 96-well plate reader (Molecular Devices). Human placental β-hexosaminidase was purchased from Sigma (lot043K3783), and O-GlcNAcase was overexpressed and purified freshly, prior to use.[85] O-GlcNAcase and β-hexosaminidase were used in the inhibition assays at a concentration (μg/μL) of 0.0406 and 0.012, respectively using substrate pNP-GlcNAc at a concentration of 0.5 mM. All inhibitors were tested at seven concentrations ranging from 3 times to ⅓ KI, with the exception of the assay of inhibitor 22d with β-hexosaminidase, where such high concentrations of inhibitor could not be reached because of its high KI value. KI values were determined by linear regression of data from Dixon plots.
As stated above, PUGNAc is a potent competitive inhibitor of both O-GlcNAcase[13, 50] and β-hexosaminidase[28, 51]. For the human enzymes the respective KI values are 46 nM and 36 nM.[13, 49] The selectivity of inhibitors 22a-f were evaluated and compared to PUGNAc. Using pNP-GlcNAc as a substrate it was found that these compounds 22a-f are inhibitors of both human O-GlcNAcase and human β-hexosamimidase (Table 3).
Analysis of the inhibition of human β-hexosamimidase reveals that increasing the chain length of the N-acyl group results in a marked decrease in the potency of these inhibitors (Table 3). The inclusion of only one methylene group (compound 22a) results in a 33-fold increase in the KI value for β-hexosaminidase as compared to the parent PUGNAc. Additional increases in chain length lead to still greater increases in KI values. Furthermore, O-GlcNAcase tolerates increases in chain length better than β-hexosaminidase (Table 3).
For these two human enzymes, O-GlcNAcase and β-hexosaminidase, direct comparison of the KI values of the previously prepared NAG-thiazoline derivatives, 9b-g,[49] reveals that the modified thiazolines are more selective and potent inhibitors than the corresponding PUGNAc-based compounds, 22a-f. These observations point to the importance of the position of the acyl group of the inhibitor and potentially, the hybridisation of the anomeric carbon. The thiazolines, NAG-thiazoline and 9b-g, with a sp3 hybridised anomeric carbon, comprise a bicyclic scaffold that restricts movement of the acyl chain as compared to the acyl chain on the corresponding PUGNAc derivatives, PUGNAc and 22a-f, that have a sp2 hybridised anomeric carbon. As well, the precise positioning of the side chain within the active sites must vary between these two sets of compounds. Together, these two factors must contribute to both the overall somewhat poorer inhibition and lesser selectivity of the PUGNAc-derived compounds compared to the thiazoline derivatives. Consistent with these observations is that streptozotocin (STZ), a poor inhibitor of O-GlcNAcase (KI=1.5 mM),[46, 47, 49] also has a bulky, freely rotating acyl chain and shows moderate selectivity for O-GlcNAcase over β-hexosaminidase.[46, 47, 49]
These differences between the thiazoline-based compounds, NAG-thiazoline and 9b-g, and the PUGNAc derivatives, PUGNAc and 22a-f, may stem from the former compounds emulating the presumed bicyclic-like transition state involved in the catalytic mechanism of O-GlcNAcase[49] and β-hexosaminidase[30, 48, 83, 97]. In contrast, the position of the N-acyl group of the PUGNAc analogues and STZ may resemble that of the natural substrate N-acetyl-
All of the above prepared compounds were inhibitors of both human O-GlcNAcase and human β-hexosaminidase. These compounds, however, exploit differences in active site architectures between these two enzymes, which results in them being selective for O-GlcNAcase. Despite the lower selectivity of the PUGNAc analogues as compared to the thiazoline derivatives, these compounds may have different pharmacokinetic properties owing to their different scaffold. Accordingly, both the thiazoline and PUGNAc derivatives may prove to be valuable tools for dissecting the role of the O-GlcNAc post-translational modification at the cellular and organismal level. Indeed, elaboration of other glycosidase inhibitors has yielded clinically useful compounds targeting entirely different enzymes.[98] As well, using the strategy outlined here, systematic elaboration of other β-N-acetyl-glucosaminidase inhibitor scaffolds[99, 100] may also yield selective inhibitors for human O-GlcNAcase.
Having demonstrated the selectivity of these compounds in vitro the use of the compounds in living cells was evaluated.
COS-7 cells were cultured in DMEM medium (Invitrogen) supplemented with 5-10% FBS (Invitrogen). Aliquots of inhibitors (50 μL of a stock in 95% ethanol) were delivered onto tissue culture plates and the ethanol was evaporated. The cells were incubated at 37° C. for 40 hours at which time they reached approximately 80% confluence. The time dependent accumulation of O-GlcNAc-modified proteins in response to treatment with 50 μM of compound 9a, 9c or 9g in cells was studied as follows. COS-7 cells were cultured to 25% confluence in 5% FBS and an aliquot (100 μL) of inhibitor dissolved in media and filter sterilized was added to each plate to yield a final concentration of 50 μM of inhibitor. COS-7 cells (2×10 cm plates) were harvested at the appropriate times by scraping and were pooled by centrifugation (200×g, 10 min). Cells were washed once with PBS, pH 7.0 (10 mL) and pelleted (200×g, 10 min). The cells could be frozen at −80° C. at this point. Control cultures without inhibitors were treated in the same manner.
COS-7 cells were cultured in the presence of inhibitors 9a, 9c, or 9g as described above to approximately 90% of confluence. A culture of control cells was treated in the same manner as follows but the cultures contained no inhibitor. Cells were harvested as described above. Frozen cells were thawed at 4° C., and cold lysis buffer (1 mL of 50 mM Tris, pH 8.0 containing 150 mM NaCl, 1 mM EDTA, 1 mM PMSF, 1% NP-40, 0.5% sodium deoxycholate, and 1 mM of inhibitor 90) was added. After 10 minutes at 4° C. the solution was centrifuged at 14,000 rpm in an Eppendorf 5415C microcentrifuge and the supernatant was collected. SDS/PAGE loading buffer was added to an aliquot (15 μL) of each sample, and after heating at 96° C. aliquots were loaded onto 10% or 12% Tris-HCl polyacrylamide gels. After electrophoresis, the samples were electroblotted to nitrocellulose membrane (0.45 μm, Bio-Rad). Transfer was verified by visual inspection of the transfer of prestained markers (Dual Colour Precision Plus Protein Standard—Biorad). The membrane was blocked by using 5% BSA (fraction V, Sigma) in PBS (blocking buffer A for samples probed with mouse anti-O-GlcNAc monoclonal IgM antibody (MAb CTD 110.6—Covance)) or 5% low-fat dry powdered milk (blocking buffer B for samples probed with anti-β-actin), pH 7.4, containing 0.1% Tween 20 for 1 h at room temperature or overnight at 4° C. The blocking solution was decanted, and a solution of blocking buffer A containing MAb CTD 110.6 (1:2500 of the stock) or blocking buffer B containing mouse monoclonal anti-β-actin IgG (Clone AC-40—Sigma) was added (1:1000 dilution) as appropriate. The membrane was incubated at room temperature for 1 h or overnight at 4° C. after which the blocking buffer was decanted and the membrane was rinsed with PBS, pH 7.4, containing 0.1% Tween 20 (wash buffer). Membranes were then rinsed for 2×5 min and 3×15 min with wash buffer. For immunological detection of O-GlcNAc, the membrane was incubated in blocking buffer A for 1 hour at RT and, after washing, the membrane was incubated with a secondary goat anti-mouse-IgM-HRP-conjugate (1:2500, Santa Cruz Biotech) for one hour at RT or 4° C. overnight in blocking solution. For detection of β-actin levels, the membrane was incubated with a secondary goat anti-mouse-IgG-HRP conjugate (1:100000, Sigma) for one hour at RT or 4° C. overnight in blocking solution B. Membranes were washed and detection of membrane bound goat anti-mouse-IgG-HRP conjugate was accomplished by chemiluminescent detection using the SuperSignal West Pico Chemiluminescent Detection Kit (Pierce) and film (Kodak Biomax MR).
COS-7 cells incubated in plates with 50 μM of inhibitor 9a, 9c, or 9g revealed no abnormalities in proliferation rate or morphology as compared to control cells (data not shown). Cellular levels of O-GlcNAc-modified proteins within cells cultured for 40 hours in the presence of inhibitors 9a, 9c or 9g, or in their absence was carried out using the O-GlcNAc directed monoclonal antibody (64) mAbCTD110.6. Marked increases in cellular levels of O-GlcNAc-modified proteins within the cells were observed as compared to the control (
Eight give-week old healthy Sprague-Dawley rats (Charles River) were caged in pairs. The animals were allowed to acclimatize to their new surroundings at the Simon Fraser University Animal Care Facilities (SFU ACF) for one week. At 3:00 pm, four rats were given a tail vein injection of 400 μl of inhibitor 9c dissolved in PBS buffer (pH 7.4) at a dose of 50 mg kg−1. The remaining four rats were given a tail vein injection of 400 μl of PBS as a control. All solutions were pre-sterilized through a 0.2 μm filter (Millipore) and injected with 0.5 mL syringes containing a 28 gauge×0.5″ needle (Terumo) as bolus over 10 seconds. Food was withdrawn at approximately 11:00 pm. At 8:30 am the following morning, the rats were given a second injection. After three hours the rats were anesthetized with isoflurane and a 100 μL blood sample was taken from the jugular vein to measure the fasting blood glucose levels. An intravenous glucose tolerance test (IVGTT) was then performed. A 0.5 mL injection of a 50% w/v solution of glucose dissolved in PBS (pre-sterilized) was administered via the tail vein over 30 seconds. At 10, 20, 30, 40, 60, and 90 minutes after the injection of glucose, a 100 μL sample of blood was taken from the jugular vein. Immediately following blood sampling, a small aliquot of the blood was used to measure the blood glucose concentration using a glucometer (Accu-Check Advantage, Roche) and the remaining blood was stored on ice for 20 minutes and spun down to isolate the serum. Following completion of the IVGTT, the rats were sacrificed with 1:1 mixture of CO2/O2. Tissue samples (brain, muscle, liver, spleen, pancreas, fat) were immediately harvested and flash frozen with liquid nitrogen and stored at −80° C. To homogenize the tissues to obtain a cell extract, tissue were kept frozen while they were ground with a mortal and pestle into a fine powder. 100 mg of the powder was then homogenized in 1 mL of cold lysis buffer (50 mM Tris, 0.5% sodium deoxycholate, 0.1% SDS, 1% nonidet P-40, 1 mM EDTA, 1 mM PMSF, and 1 mM butyl-NAG-thiazoline) using two 10 second pulse on a Jenke and Kunkel Ultra-Turrax tissue homogenizer tissue homogenizer. Tissues were then spun down at 13,000 rpm in a microcentrifuge (eppendrof) to remove the cell debris. The soluble cell extract was then analyzed for levels O-GlcNAc modified proteins using Western blot analysis as described above for the cell studies.
Rats injected with varying doses of inhibitor 9c via the tail vein showed that intravenous administration of inhibitor 9c affects cellular levels of O-GlcNAc modified proteins in various tissue types (
From the results, it appears that the inhibitor is only inhibiting O-GlcNAcase in a semi dose-dependent manner. This shows that 50 mg/kg of inhibitor 9c is enough to elicit the maximal increase in O-GlcNAc modified proteins. As well, these results show that the inhibitor is able to gain access to a wide variety of tissues in less than 24 hours and retains an effect for a minimum of 6 hours. It should also be noted that rats used in this study that were treated with the inhibitor did not appear to show any discomforts compared to the control rat.
Having shown that the inhibitor 9c is capable of inhibiting O-GlcNAcase in rats, attention was turned to addressing whether or not increased levels of O-GlcNAc modified proteins alters the ability of the rats to maintain glucose homeostasis. An intravenous glucose tolerance test (IVGTT) was used as a means of measuring the ability of the animal to respond to a challenge of 1g kg-1 of glucose. The results clearly indicate that short term exposure of the inhibitor (<24 hours) does not affect the glucose clearance rate in rats (
This result is surprising, given that it has been shown previously that elevated levels of O-GlcNAc modified proteins in cultured adipocytes cells results in insulin resistance. This result is surprising, given that it has been shown previously that elevated levels of O-GlcNAc modified proteins in cultured adipocytes cells results in insulin resistance. The inventors hypothesize that the differences between these results stem from one of two possibilities. First, the results obtained from a whole organism is more physiologically relevant than experiments conducted using cultured cells, and therefore insulin resistance does not develop in peripheral tissues. The second possibility is that the inhibitor 9c actually affects the pancreatic β-cells in a way that causes them to secrete more insulin. This effect would overcome insulin insensitivity in the peripheral tissues, possibly resulting in normal glucose clearance as was observed. Support for this second hypothesis comes from the fact that a number of proteins involved in producing and transporting insulin within the β-cells are themselves modified by O-GlcNAc, such as two of the transcription factors (pdx-1 and sp1) that stimulate insulin production.
To obtain information for how rapid the inhibitor is cleared from tissues, five 10-week old Sprague-Dawley rats were given a tail vein injection of the inhibitor (50 mg kg−1). The animals were sacrificed sequentially 3, 7, 24, 27, and 32 hours following the injection. Two additional rats were used as controls. One rat was sacrificed immediately before the others received their injections in order to obtain information about the normal levels of O-GlcNAc modified proteins. Also, one rat was injected with sterilized PBS (pH 7.4) along with the other rats and sacrificed at the end of the experiment (32 hours). Tissue collection, homogenization, and western blot analyses were done in the same matter as described above.
Rats injected with inhibitor 9c showed that intravenous administration of inhibitor 9c affects cellular levels of O-GlcNAc modified proteins in various tissue types as early as three hours following injection (
To determine if the inhibitor 9c is orally available, it was incorporated into the chow (Lab Diet 5001 Rodent Diet, PMI Nutrition International, LLC). To make the rat chow, 600 g of the ground chow was mixed with 355 ml of water and 5 μL of the inhibitor dissolved in ethanol. Small pieces were then prepared with a pasta machine (Pasta Express, Creative) and dehydrated overnight at 37° C. (Snackmaster Dehydrator, American Harvest). Five sets of chow were made with the following amounts of inhibitor: 0 mg kg−1 days−1, 100 or 1000 mg kg−1 days−1 of the deprotected (polar) inhibitor, and 100 or 1000 mg kg−1 day−1 of the protected (non-polar) inhibitor. These numbers are based on data obtained from previous studies that show that a 6-week old rat eats approximately 25g food per day. Ten five-week old healthy Sprague-Dawley rats (two per set of food) were then allowed to feed on the rat chow for three days. The animals were then sacrificed and tissues were collected, homogenized, and analyzed as described above.
Rats were fed two different doses of protected (non-polar) and deprotected (polar) forms of inhibitor 9c for three days. Western blot analyses showed that oral administration of both forms of inhibitor 9c affects cellular levels of O-GlcNAc modified proteins in various tissue types (
As will be apparent to those skilled in the art in the light of the foregoing disclosure, many alterations and modifications are possible in the practice of this invention without departing from the spirit or scope thereof.
This application claims the benefit of the filing date of U.S. provisional patent application No. 60/656,878 filed 1 Mar. 2005 which is hereby incorporated by reference.
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/CA2006/000300 | 3/1/2006 | WO | 00 | 5/27/2008 |
| Number | Date | Country | |
|---|---|---|---|
| 60656878 | Mar 2005 | US |
