DDR
DRUG DEVELOPMENT RESEARCH 67:448–455 (2006)
Research Article
A New Inhibitor for Fructosamine 3-Kinase (Amadorase) Henry J. Tsai,1 Shan-Yen Chou, 1 Frank Kappler,2 Michael L. Schwartz,2 and Annette M. Tobia2 1 Pharmaceutical R&D Program, Development Center for Biotechnology, Hsi-Chih City, Taiwan 2 Dynamis Therapeutics, Inc., Elkins Park, Pennsylvania
Strategy, Management and Health Policy Enabling Technology, Genomics, Proteomics
Preclinical Research
Preclinical Development Toxicology, Formulation Drug Delivery, Pharmacokinetics
Clinical Development Phases I-III Regulatory, Quality, Manufacturing
Postmarketing Phase IV
ABSTRACT Non-enzymatic glycation is a Maillard reaction that occurs spontaneously between glucose and proteins and results in the accumulation of advanced glycation end products (AGEs), which are believed to be one of the causes for age-related tissue damage. Fructosamine 3-kinase (FN3K) is an enzyme involved in the protein deglycation process by fructosyl phosphorylation, returning amino acid residues to their pristine state with 3-deoxyglucosone (3-DG) and inorganic phosphate as by-products. 3-DG is a very reactive electrophile that is even more potent at generating AGEs than glucose; therefore, inhibitors specific to FN3K have been suggested as a means to suppress 3-DG formation. Dyn-12 is a substrate mimicking inhibitor with an IC50 value of approximately 6 mM. We report a new hydrophobic competitive inhibitor with respect to fructolysine that has an IC50 of 1.7 mM. Drug Dev. Res. 67:448–455, c 2006 Wiley-Liss, Inc. 2006. Key words: advanced glycation end products; 3-deoxyglucosone; fructosamine 3-kinase; glycation; inhibitor; protein deglycation
INTRODUCTION
Glucose is an aldose that can potentially conjugate with an exposed amine (e.g., the e-amino group of a surface lysine residue) through Schiff base condensation and Amadori rearrangement to form various fructosamine derivatives (e.g., fructolysine) [Szwergold et al., 2001, Delpierre et al., 2002]. The overall reaction is a non-enzymatic Maillard reaction and referred to as glycation, which is different from an enzyme catalyzed glycosylation [Delpierre et al., 2000b]. Under physiological conditions, protein glycation occurs spontaneously and accumulates in erythrocytes or eye lenses due to their slow or non-existent protein turnover [Delpierre et al., 2002]. Once a protein is glycated, it was thought that the sugar adduct would eventually become advanced glycated end products (AGEs) [Thornalley et al., 1999]. However, recently a fructosamine 3-kinase (FN3K; E.C. 2.7.1.) was isolated from erythrocytes
c
2006 Wiley-Liss, Inc.
and its cDNA cloned by two different laboratories [Szwergold et al., 2001, Delpierre et al., 2000a] who have demonstrated that FN3K facilitates the deglycation process of hemoglobin by phosphorylating the 3-hydroxyl group of fructolysine, forming fructolysine 3-phosphate [Delpierre et al., 2002, Szwergold and Abbreviations: AGE, advanced glycation end products; 3-DG, 3-deoxyglucosone; DMF, 1-deoxy-1-morpholinofructose; 3-DF, 3-deoxyfructose; FN3K, fructosamine 3-kinase. Grant sponsor: Ministry of Economic Affairs, Taiwan, ROC. Correspondence to: Henry J. Tsai or S-Y. Chou, Pharma-
ceutical R&D Program, Development Center for Biotechnology, 101 Ln 169, Kang-Ning St., Hsi-Chih City, Taipei County 221, Taiwan. E-mail:
[email protected] or
[email protected] Received 28 January 2006; Accepted 15 May 2006 Published online in Wiley InterScience (www.interscience.wiley. com). DOI: 10.1002/ddr.20105
FRUCTOSAMINE 3-KINASE INHIBITOR
Beisswenger, 2003]. Because of the close proximity to the 2-keto group, the 3-phosphate group is labile and hydrolyzes spontaneously through b-elimination [Szwergold et al., 2001]. The hydrolysis of the 3-phosphate group triggers the breakage of Schiff base imine bond and results in the regeneration of lysine in its pristine state with 3-dexoyglucosone (3-DG) as the by-product [Szwergold et al., 2001, Delpierre et al., 2002, Szwergold and Beisswenger, 2003, Delpierre and Van Schaftingen, 2003], as depicted in Figure 1. 3-DG is a 2-oxoaldehyde (dicarbonyl hexose) with high reactivity to amine or guanidine derivatives; it is, therefore, a glycating agent even more potent than glucose or fructolysine to form AGEs, e.g., imidazolone, pyrraline [Beisswenger et al., 2003]. Though 3-DG can be metabolized by intracellular aldehyde reductase to 3-deoxyfructose (3-DF) or oxidized by aldoketo dehydrogenase to 2-keto-3-deoxygluconic acid [Szwergold et al., 2001, Beisswenger et al., 2003, Wu and Monnier, 2003], the concentrations of 3-DG and imidazolone were 6.7- and 3.7-fold, respectively, higher in erythrocytes of diabetic hemodialysis patients suggesting that 3-DG and AGEs can accumulate under chronic hyperglycemic conditions [Tsukushi et al., 1999]. Since the protein deglycation pathway is predominated by FN3K, it has been suggested by Brown et al. [2003] that it may be possible, using FN3K specific inhibitors, to reduce the formation of 3-DG. 1-Deoxy-1-morpholinofructose (DMF) was originally employed as an FN3K inhibitor (Ki 5 100 mM) for
H N
demonstrating the enzymatic process of hemoglobin deglycation, but DMF is, in fact, a substrate (Km 5 100 mM) for FN3K [Delpierre et al., 2000b, 2002; Delpierre and Van Schaftingen, 2003]. As a substrate, DMF can be phosphorylated at the 3-hydroxyl position by FN3K, generating DMF-3phosphate as the product [Szwergold et al., 2001, Delpierre et al., 2000b, 2002]. Just like fructolysine 3-phosphate, DMF-3-phsophate may undergo spontaneous b-elimination generating 3-DG as a by-product, and, therefore, DMF is not suitable for a pharmaceutical purpose. Dyn-12 is a non-substrate inhibitor, which mimics fructolysine [Kappler et al., 2001] with an IC50 value of approximately 6 mM. We now report a new hydrophobic inhibitor that is competitive with respect to substrate fructolysine and has an IC50 of 1.7 mM. MATERIALS AND METHODS
General chemicals were purchased from SigmaAldrich (St. Louis, MO). Other specialized kits were purchased from respective suppliers as indicated in the text. cDNA Construct and Expression The 0.9-kb cDNA encoding mouse FN3K was subcloned to BamHI and PstI sites in pFastBac HT of Invitrogen (Carlsbad, CA). The vector in this construct adds an N-terminal poly (6x) His tag to the protein and is designed for baculovirus/insect cell expression.
OH
O
H N
FN3K
Lysine OH
OH
O
Lysine
OH OH
449
ATP
ADP
OH OPO3 OH
3-Phosphofructolysine
Fructolysine
O
OH
O
OH H
+ Lysine + PO43-
OH
3-Deoxyglucosone (3-DG) Fig. 1. Protein deglycation pathway facilitated by fructosamine 3-kinase (FN3K). The deglycation process is accelerated by phosphorylation on the 3-hydroxyl group of fructosyl moiety that is catalyzed by fructosamine 3-kinase. Spontaneous hydrolysis of 3-phosphoryl group on 3-phosphofructolysine triggers hydrolysis of the Schiff base bond, returning pristine lysine residues with 3-deoxyglucosone and inorganic phosphate as by-products.
Drug Dev. Res. DOI 10.1002/ddr
450
TSAI ET AL.
The DNA was amplified by DH10Bac and made into baculovirus by using a Bac-to-Bac kit from Invitrogen. The initial virus was amplified by diluting 1:100 (v/v) with Sf-9 cells (1 106cells/ml) and cultured for 10 h at 271C. This virus batch was thus amplified 2 more rounds to generate a 3rd passage virus. The His-tag FN3K fusion protein was expressed in Sf-9 cells (1 106cells/ml), which were infected with 3rd passage baculovirus (1:10 dilution) and cultured at 271C for 72 h after infection. After 72 h, Sf-9 cell culture was spun down and stored at 301C until purification Purification of Recombinant Fusion Protein The recombinant mouse FN3K was purified with a B-Per 6xHis Fusion Protein Purification Kit (Pierce, Rockford, IL). The purification procedure followed the protocol described in the kit manual. Briefly, a 250-ml culture volume equivalent of Sf-9 cell pellet was resuspended in 10 ml B-Per buffer. Cells were lysed by one freeze-thaw ( 801C/room temperature) cycle. Lysis was confirmed by visual inspection under a microscope and the cell lysate was centrifuged at 3,800g (Beckman S4180 rotor, 4800 rpm) for 30 min at 41C. The supernatant was then poured onto a Ni-resin column provided in the kit, which was pre-rinsed with 10 ml B-PER buffer. The loaded Ni-column was washed with 12 ml Wash Buffer 1 followed by 18 ml Wash Buffer 2, and then eluted with 6 ml Elution Buffer, and fractionated into 1 ml each. The majority of FN3K was eluted in the 2nd and 3rd fractions. The protein concentration in each fraction was determined by the Bradford reagent from BioRad with BSA as standard. Eluted fractions with high protein content were then pooled and dialyzed against a buffer containing 20 mM Tricine, pH 8, 5 mM DTT, 100 mM NaCl, and 0.5 mM EDTA overnight. The dialyzed solution was aliquoted and stored at 801C after being flash frozen in liquid nitrogen. Measurement of FN3K Activity The assay condition for FN3K inhibitor screening was optimized for sensitivity with substrate concentrations near respective Km values so that the reaction rate was more sensitive to compound inhibition. The assay was carried out at pH 7.0 in the presence of 50 mM HEPES, 0.25 mM ATP (0.2 mCi), 2.25 mM MgCl2, 0.5 mM fructolysine, 1 mM DTT, and 1.6 mg (0.4 mM) of FN3K. The reaction was carried out in 100 ml at 371C for 10 min and quenched by an equal volume of 25 mM EDTA. An aliquot of the reaction mix (85 ml) was then blotted on an inch-square P81 filter paper, air dried, and rinsed 5 times in deioned water to wash Drug Dev. Res. DOI 10.1002/ddr
away g-P33-ATP. The P81 filter paper was then submerged in 5 ml of scintillation fluid and counted in a b-particle liquid scintillation counter. When determining Michaelis constants, Km, respective substrate concentrations varied over 1 order and encompassed the Km. The IC50 value was defined as the concentration at which the compound produces 50% inhibition as compared to the non-inhibited control. Compound Syntheses Dyn-12 was prepared by reductive amination of 3-O-methylglucose a-N-Cbz lysine using sodium cyanoborohydride as reducing agent, followed by Cbz deprotection, and the structure was confirmed by comparing 1H NMR and LC/MS spectroscopy with those of an original sample provided by Dynamis Co. D-400100 was prepared as an unidentified diastereomeric mixture by heating 2-(1-benzyloxy- butyl)oxirane with 3-phenylpropylamine at 100–1101C. The oxirane intermediate was prepared by epoxidation of 1-hexen-3-ol using m-chloroperbenzoic acid and followed by O-benzylation. The structure of D-400100 was verified by 1H NMR and LC/MS spectroscopy: 1H NMR (CDCl3) 7.27 (m, 5H), 7.17 (m, 5H), 4.51, 4.61 (each m, 2H, diastereomeric ratio 4:1), 3.72 (m, 1H), 3.38, 3.50 (each m, 1H), 2.63–2.74 (m, 6H), 1.80 (m, 2H), 1.40–1.60 (m, 4H), 0.93 (t, J 5 7.1 Hz, 3H). LC/MS 342.5 (M111). D-400540 was prepared by reductive amination of a chiral 4-methoxy-5-pentyl-tetrahydrofuran-2,3-diol with n-nonyl amine using sodium cyanoborohydride as reducing agent. The tetrahydrofuran-2,3-diol intermediate was prepared from 3-O-methyl-diacetone D-glucose (i.e., 3-O-methyl-2,3,5,6-diketal D-glucose) by a five-step sequence: (1) selective hydrolysis of the 5,6-ketal moiety to a 5,6-diol; (2) oxidative cleavage of the 5,6-diol by sodium periodate to an aldehyde; (3) Wittig reaction on the aldehyde with butylidenetriphenyl- l5-phosphane to generate an olefinic double bond; (4) hydrogenation of the olefinic double bond; and (5) hydrolysis of the 1,2- ketal moiety on the furan ring. The structure of D-400540 was verified by 1H NMR and LC/MS spectroscopy. 1H NMR (CDCl3) 3.92 (m, 1H), 3.75 (m, 1H), 3.53 (s, 3H), 3.10 (s, 1H), 2.86 (m, 2H), 2.72 (m, 2H), 1.55, 1.26 (each m, 22H), 0.91, 0.88 (each t, J 5 7.0 Hz, 6H). LC/MS 332.5 (M111). D-400634 was prepared by reductive amination of the aforementioned chiral 4-methoxy-5-pentyl-tetrahydrofuran-2,3-diol with a-N-Cbz lysine using sodium cyanoborohydride as a reducing agent, followed by Cbz deprotection. The structure was verified by 1H NMR and LC/MS spectroscopy. Characteristic peaks appeared at around 3.6 and 0.9 ppm for OCH3 and
FRUCTOSAMINE 3-KINASE INHIBITOR OH OH
H N
HO2C NH2
H N
OH H3C
O
451
OH
OH OH
H3 C
O
D-400540
DYN-12 OH
H N
O
D-400100
OH OH
H N
HO2C NH2
H3C
O
D-400634
H N
OH OH OH H3 C
O
OH
D-400637
Fig. 2. Structures of Dyn-12 and FN3K inhibitor analogues. Dyn-12 (3-methylsorbitolysine) is a direct mimic of FN3K substrate, fructolysine. D-400540 retains the stereospecificity of C2–4 on the polyol moiety but the carbon chains were replaced by alkyl groups with 3 additional carbons on each end. D-400100 has a phenylpropanyl group in place of the amino caproic acid group and a phenylmethoxyl group on the C3 position. D-400634 retains the amino caproic acid group but the polyol side chain is identical to that of D-400540. D-400637 retains the sorbitol group but the amino caproic acid group is replaced by a 9 carbon alkyl group.
terminal CH3 groups, respectively. LC/MS revealed a base peak at 322.0 for M111. D-400637 was prepared by reductive amination of 3-O-methylglucose with n-nonyl amine using sodium cyanoborohydride as a reducing agent. The structure was verified by 1H NMR and LC/MS spectroscopy. Beside the characteristic peaks for OCH3 and terminal CH3 groups, the spectroscopy was also correlated with the spectrum of Dyn-12 as well as D-400540, respectively. LC/MS revealed a base peak at 308.0 for M111. Structures of these compounds are depicted in Figure 2. RESULTS
Purification of Recombinant Mouse FN3K Recombinant mouse FN3K was purified from the Sf-9 cell lysate by a Ni-resin column to near homogeneity in one pass (Table 1; Fig. 3). The majority of the purified enzyme was eluted in the 2nd and 3rd fractions. The purification increased specific activity by nearly 2,000-fold (5.952/0.003); the higher total FN3K activity in the purified fraction than in the crude extract suggests there is some interfering factor in the crude extract (Table 1). The specific activity determined was under subsaturating levels of substrates and the
velocity is estimated to be approximately 20% Vmax. Therefore, the specific activity of 5.95 mU/mg observed in the purified FN3K is on a par with the Vmax of recombinant mouse FN3K reported by Delpierre et al. [2000a]. Purified human FN3K from erythrocytes has a reported MW of about 35 kDa. In our hands, the recombinant mouse FN3K showed a molecular weight of about 40 kDa on the SDS-PA gel (Fig. 3). The entire sequence of FN3K has been reconfirmed; therefore, the higher molecular weight is probably due to its high isoelectric point (calculated pI 5 8.4) and the additional peptide sequence gathered from the polyHis tag expression vector. Enzymatic Activity Assay Conditions and Inhibition Pattern The optimal FN3K activity occurred near pH 6.0, but to mimic physiological conditions better, pH 7.0 was chosen as the assay condition (Fig. 4). The small activity peak at pH 9.0 was also observed in insulin-like growth factor-1 (IGF-1) receptor tyrosine kinase domain, but the actual reason is not clear. The purified recombinant mouse FN3K possesses Michaelis-Menten behavior with apparent Michaelis constants being 0.73 mM for fructolysine, and 0.23 mM Drug Dev. Res. DOI 10.1002/ddr
452
TSAI ET AL.
TABLE 1. Recombinant Mouse FN3K Purification Tablea
Fraction
Volume (ml)
Protein concentration (mg/ml)
Protein (mg)
FN3K activity (mU/ml)
Total FN3K activity (mU)
Specific activityb (mU/mg)
Extr Supc Flowthru Wash Elute
11.1 10.5 21 4.2
4.6 5.3 0.7 0.25
51.06 55.65 14.7 1.05
0.014 0.016 0.011 1.488
0.150 0.171 0.224 6.250
0.003 0.003 0.015 5.952
a
The recombinant mouse FN3K was purified with a Ni-resin column as described in Materials and Methods. Protein concentrations were determined by the Bradford method with BSA as standard. b Enzymatic activity was determined with 0.5 mM fructolysine and 0.25 mM ATP, which should be approximately 20% of theoretical Vmax. c Supernatant of Sf-9 cell lysate extract.
pH Dependency of FN3K Activity 3500
CPM incorporated
3000 2500 2000 1500 1000 500 0
3
5
7 pH
9
11
Fig. 4. pH dependency of FN3K activity. FN3K activities (5 ml, 0.8 mg) were assayed under 50 mM of buffers at various pH as indicated, but otherwise assayed under standard conditions as stated in Materials and Methods. Buffers used are pH 4 and 5, acetic acid (pKa 5 4.76); pH 6, MES (pKa 5 6.1); pH 7, HEPES (pKa 5 7.5); pH8, Tricine (pKa 5 8.1); pH9 and 10, glycine (pKa2 5 9.8). Fig. 3. SDS-PAGE of purification fractions from the Ni-resin column. Various amounts of protein from each fraction were loaded in each well on a 12% SDS-PA gel due to their drastic difference in protein concentrations. From left to right, Lane 1: BioRad low range protein marker (molecular weight from the top: 118, 85, 47, 36, 26, 20 kDa); Lane 2: Supernatant of cell extract (5 mg); Lane 3: Flowthrough (5 mg); Lane 4: Wash faction (0.3 mg); Lane 5: Elute (0.75 mg).
for ATP.Mg. For better inhibitory response, the standard enzymatic activity assay was carried out at pH 7.0 with 0.5 mM fructolysine and 0.25 mM ATP.Mg. Under such conditions, the reaction is linear within the first 12 min or up to 2.5 mg of FN3K (data not shown). These conditions were employed throughout protein purification and subsequent compound evaluation. Since compound D-400540 is a substrate mimetic of fructolysine, it would be expected to be a competitive inhibitor with respect to fructolysine. Our inhibition study suggests D-400540 is indeed competitive with respect to fructolysine with a Ki of 5.3 mM (Fig. 5). Drug Dev. Res. DOI 10.1002/ddr
Inhibitor Potency Since the lead compound, Dyn-12, consisted of a lysine and a sorbitol, the structure is fairly hydrophilic and may not be ideal for pharmaceutical purposes. We elected to keep the secondary amine group, and to replace the lysine or sorbitol moiety with more hydrophobic acyl or aryl substitutes. The resulting derivatives include amino-diol, e.g., D-400100, and amino-triol compounds, e.g., D-400540, D-400634, and D-400637 (see Compound Syntheses in the Material and Methods). The IC50 for Dyn-12 is greater than 3 mM, but extrapolation predicts approximately 6 mM, while the IC50 for compound D-400100 and D-400540 is 2 and 1.7 mM, respectively. Because pilot studies showed D-400100 had poor solubility and strong tissue toxicity (data not shown), its derivatives were not further investigated. D-400540 is a stereospecific compound
FRUCTOSAMINE 3-KINASE INHIBITOR
2.50
Dyn-12. The water solubility of Dyn-12 is 576 g/L at 251C, which is very high but not surprising given its relationship to fructolysine, whereas D-400540 has minimal solubility at 251C, which is to be expected because its alkyl side chain replacement at both ends should make it fairy hydrophobic.
2.00
DISCUSSION
1.50
Protein deglycation catalyzed by FN3K is a process of restoring proteins to their original structure and function prior to the Maillard modification [Delpierre et al., 2002; Szwergold and Beisswenger, 2003; Delpierre and Van Schaftingen, 2003]. Glycated proteins may be recovered at the expense of ATP, but the sugar by-product, 3-DG, may actually pose even more of a problem for the body to manage. 3-DG is a 2oxoaldehyde, thousands of fold more reactive than glucose, and as a result, toxic to the body [Beisswenger et al., 2003]. Though 3-DG is rapidly excreted in the urine [Kato et al., 1990] or metabolized by redox reactions to 3-DF or 2-keto-3-deoxygluconic acid [Szwergold et al., 2001; Beisswenger et al., 2003; Wu and Monnier, 2003; Kato et al., 1990], the accumulation of its derivatives (e.g., imidazolone, pyrraline) suggests that 3-DG reacts rapidly with other amine or guanidine species to form AGEs [Beisswenger et al., 2003]. Aminoguanidine is a highly reactive nucleophile that can chemically, i.e., non-enzymatically, neutralize 2-oxoaldehyde (e.g., 3-DG, methylglyoxal) and has been used to prevent AGE formation [Thornalley, 2003]. Because the reaction takes place in the solvent environment instead of a confined reaction pocket of an enzyme, aminoguanidine reacts non-specifically on other metabolites or enzymes like nitric oxide synthase or semicarbazide sensitive amine oxidase. Therefore, the clinical trial with aminoguanidine was terminated due to safety concerns [Thornalley, 2003]. Once glycated, some protein residues remain glycated until the protein is turned over, because of inaccessibility or low affinity to FN3K, e.g., the N-terminal Val b-1 and Lys a-61 of hemoglobin [Delpierrre et al., 2004]. Given the choice between the loss of function due to fructose adduct or the toxicity of 3-DG, the former does seem more tolerable and so too the concept of developing FN3K inhibitor for preventing diabetes related tissue damages. Dyn-12 is a fructolysine-mimicking but non-reactive inhibitor to FN3K [Kappler et al., 2001]. Its lysine and sorbitol moieties ensure specificity to FN3K but make it fairly hydrophilic, which is not ideal for pharmacological purposes. In comparison, D-400540 retains the Schiff base imine and stereospecific hydroxyl groups on C2–4 of fructolysine, but replaces the remainder with alkyl chains. From the present studies, those
4.50 4.00 3.50 3.00 1/V
453
1.00 0.50 0.00 0
0.5
1
1.5
2.5
2
1/S Fig. 5. Lineweaver-Burk plot of compound D-400540 inhibition against substrate fructolysine. Assay conditions are standard as stated in Materials and Methods. The inhibition pattern is competitive with Ki estimated to be 5.3 mM.
TABLE 2. FN3K Activity Inhibition by Stereospecific Inhibitorsa
D-400540 D-400634 D-400637
% inhibitionb
nc
78.874.4 3.173.1 9.572.5
5 2 2
a
FN3K activities were assayed under standard conditions as described in Materials and Methods, in the absence or presence of 3 mM respective compound inhibitors. Values represent Mean7S.E. for D-400540 and Mean71/2 difference for D-400634 and D-400637. b As a percentage of the control that received blank DMSO instead. c n 5 number of independent determinations.
of which those three hydroxyl groups have the same orientations as those of Dyn-12 or fructolysine on the corresponding positions. Interestingly, D-400634, which has the same stereo-specific hydroxyl groups as D-400540 but retains the lysine moiety on the amino acid part, showed weaker inhibition than D-400540 (Table 2), suggesting the amino carboxyl moiety has a negative impact on the inhibitor potency. Since FN3K is a sugar kinase, we suspected it may recognize the sugar moiety during the substratebinding event. Therefore, we elected to keep the sorbitol moiety but replace the lysine residue with an alkyl chain identical to that on D-400540. The resulting compound D-400637 was not as potent as D-400540 (Table 2), leaving D-400540 as the best FN3K inhibitor studied (Fig. 6). Water Solubility Since water solubility of a compound will affect bioavailability and other pharmacological parameters, we compared water solubility of D-400540 against
Drug Dev. Res. DOI 10.1002/ddr
454
TSAI ET AL.
% Inhibition by Compounds 100.0
% inhibition
80.0 60.0 40.0 20.0 0.0
0
1
2 mM D-400540
3
4
Dyn-12
Fig. 6. Inhibition curves of Dyn-12 and D-400540. Enzymatic activity of FN3K was assayed under standard substrate concentrations with inhibitor concentrations as indicated on the X-axis. Each curve is an average of 5 separate determinations. Error bars represent standard errors.
stereospecificity features are important for efficacy, but, interestingly, the addition of 5- and 6-hydroxyl groups makes no positive contribution to efficacy (D400637). This stereospecificity is in accord with the report of Delplanque et al. [2004] that the FN3K recognizes its substrate based on the stereo positions of those hydroxyl groups on C1-3. Similarly, the replacement of the 9 carbon alkyl chain with an a-amino caproic acid (D-400634) makes a negative impact on its efficacy, suggesting the addition of a carboxyl group poses a negative contribution to FN3K binding. Presumably, FN3K prefers multiple positive charges at its substrate site, since the addition of each positive charge lowers the substrate Km by almost 1 order of magnitude [Szwergold et al., 2001], suggesting positive charges or amines will favor the substrate or a competitive inhibitor binding at the reaction site. Therefore, addition of amine groups (positive charges) may enhance the inhibitor potency in vitro. ACKNOWLEDGMENT
Sponsored by the Ministry of Economic Affairs, Taiwan, ROC.
Delpierre G, Van Schaftingen E. 2003. Fructosamine 3-kinase, an enzyme involved in protein deglycation. Biochem Soc Trans 31: 1354–1357. Delpierre G, Rider MH, Collard F, Stroobant V, Vanstapel F, Santos H, Van Schaftingen E. 2000a. Identification, cloning, and heterologous expression of a mammalian fructosamine-3-kinase. Diabetes 49:1627–1634. Delpierre G, Vanstapel F, Stroobant V, Van Schaftingen E. 2000b. Conversion of a synthetic fructosamine into its 3-phospho derivative in human erythrocytes. Biochem J 352:835–839. Delpierre G, Collard F, Fortpied J, Van Schaftingen E. 2002. Fructosamine 3-kinase is involved in an intracellular deglycation pathway in human erythrocytes. Biochem J 365:801–808. Delpierrre G, Vertommen D, Communi D, Rider MH, Van Schaftingen E. 2004. Identification of fructosamine residues deglycated by fructosamine-3-kinase in human hemoglobin. J Biol Chem 279:27613–27620. Delplanque J, Delpierre G, Opperdoes FR, Van Schaftingen E. 2004. Tissue distribution and evolution of fructosamine 3-kinase and fructosamine 3-kinase-related protein. J Biol Chem 279: 46606–46613. Kappler F, Schwartz ML, Su B, Tobia AM, Brown T. 2001. DYN 12, a small molecule inhibitor of the enzyme amadorase, lowers plasma 3-deoxyglucosone levels in diabetic rats. Diabetes Technol Ther 3:609–616.
REFERENCES
Kato H, van Chuyen N, Shinoda T, Sekiya F, Hayase F. 1990. Metabolism of 3-deoxyglucosone, an intermediate compound in the Maillard reaction, administered orally or intravenously to rats. Biochim Biophys Acta 1035:71–76.
Beisswenger PJ, Howell SK, Nelson RG, Mauer M, Szwergold BS. 2003. Alpha-oxoaldehyde metabolism and diabetic complications. Biochem Soc Trans 31:1358–1363.
Szwergold BS, Beisswenger PJ. 2003. Enzymatic deglycation: a new paradigm or an epiphenomenon? Biochem Soc Trans 31: 1428–1432.
Brown TR, Su B, Brown KA, Schwartz MA, Tobia AM, Kappler F. 2003. Modulation of in vivo 3-deoxyglucosone levels. Biochem Soc Trans 31:1433–1437.
Szwergold BS, Howell S, Beisswenger PJ. 2001. Human fructosamine-3-kinase: purification, sequencing, substrate specificity, and evidence of activity in vivo. Diabetes 50:2139–2147.
Drug Dev. Res. DOI 10.1002/ddr
FRUCTOSAMINE 3-KINASE INHIBITOR Thornalley PJ. 2003. Use of aminoguanidine (Pimagedine) to prevent the formation of advanced glycation endproducts. Arch Biochem Biophys 419:31–40. Thornalley PJ, Langborg A, Minhas HS. 1999. Formation of glyoxal, methylglyoxal and 3-deoxyglucosone in the glycation of proteins by glucose. Biochem J 344:109–116.
455
Tsukushi S, Katsuzaki T, Aoyama I, Takayama F, Miyazaki T, Shimokata K, Niwa T. 1999. Increased erythrocyte 3-DG and AGEs in diabetic hemodialysis patients: role of the polyol pathway. Kidney Int 55:1970–1976. Wu X, Monnier VM. 2003. Enzymatic deglycation of proteins. Arch Biochem Biophys 419:16–24.
Drug Dev. Res. DOI 10.1002/ddr
