Modification of Proteins In Vitro by Physiological …Paul A. Voziyan# , Raja G. Khalifah‡,...
Transcript of Modification of Proteins In Vitro by Physiological …Paul A. Voziyan# , Raja G. Khalifah‡,...
Modification of Proteins In Vitro by Physiological Levels of Glucose: Pyridoxamine Inhibits Conversion of Amadori Intermediate to Advanced Glycation End-products
Through Binding of Redox Metal Ions
Paul A. Voziyan#§, Raja G. Khalifah‡, Christophe Thibaudeau£, Alaattin Yildiz#, Jaison Jacob¶, Anthony S. Serianni£, and Billy G. Hudson#§
From the Departments of #Medicine and ¶Biochemistry, Vanderbilt University Medical
Center, Nashville, TN 37232, £Department of Chemistry and Biochemistry, University of
Notre Dame, Notre Dame, IN 46556, and ‡BioStratum, Inc., Durham, NC 27703
Running title: Pyridoxamine inhibition of AGE pathways
§ To whom correspondence should be addressed: Division of Nephrology, Vanderbilt
University Medical Center, S-3223 MCN, 1161 21st Avenue South, Nashville,
Tennessee 37232-2372. Tel.: 615-322-2089 Fax: 615-343-7156; E-mail:
[email protected] or [email protected]
* This work was supported by NIDDKD Research Grants DK065138 and P60DK20593,
and by a Research Grant from BioStratum, Inc.
1Abbreviations: : AGE, advanced glycation end-product; AG, aminoguanidine; ALE,
advanced lipoxidation end-product; BSA, bovine serum albumin; CML, Nε-
(carboxymethyl)lysine; DTPA, diethylenetriaminepentaacetic acid; FL, fructosyllysine;
GO, glyoxal; GLA, glycolaldehyde; MGO, methylglyoxal; MOLD, methylglyoxal-lysine
dimer; PM, pyridoxamine; RNase, bovine pancreatic ribonuclease A.
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
2
Hyperglycemic conditions of diabetes accelerate protein modifications by glucose
leading to the accumulation of advanced glycation end-products (AGEs). The role of
AGEs in the development of serious complications of diabetes and other diseases
emphasizes the necessity for understanding the mechanisms of their formation and
inhibition. We have investigated the conversion of protein-Amadori intermediate to
protein-AGE and the mechanism of its inhibition by pyridoxamine (PM), a potent AGE
inhibitor that has been shown to prevent diabetic complications in animal models.
During incubation of proteins with physiological diabetic concentration of glucose, PM
prevented the degradation of the protein glycation intermediate identified as
fructosyllysine (Amadori) by 13C NMR using [2-13C]-enriched glucose. Subsequent
removal of glucose and PM led to conversion of protein-Amadori to AGE Nε-
carboxymethyllysine (CML). We utilized this inhibition of post-Amadori reactions by PM
to isolate protein-Amadori intermediate and to study the inhibitory effect of PM on its
degradation to protein-CML. We first tested the hypothesis that PM blocks Amadori-to-
CML conversion by interfering with the catalytic role of redox metal ions that are
required for this glycoxidative reaction. Support for this hypothesis was obtained by
examining structural analogs of PM in which its known bidentate metal ion binding sites
were modified and by determining the effect of endogenous metal ions on PM inhibition.
We also tested the alternative hypothesis that the inhibitory mechanism involves
formation of covalent adducts between PM and protein-Amadori. However, our 13C NMR
studies demonstrated that PM does not react with the Amadori. Because the
mechanism of interference with redox metal catalysis in post-Amadori AGE formation is
operative under the conditions closely mimicking diabetic state, it may contribute
significantly to PM efficacy in preventing diabetic complications in vivo. Inhibition of
protein-Amadori degradation by PM also provides a simple procedure for the isolation of
protein-Amadori intermediate, prepared at physiological levels of glucose for relevancy,
to study both the biological effects and the chemistry of post-Amadori pathways of AGE
formation.
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
3
Chemical modifications of circulating, cellular and matrix proteins by glucose are
thought to be a major factor in pathogenesis of diabetes, atherosclerosis, and
neurodegenerative diseases, as well as in the process of ageing (1-4). These
modifications derive, in part, from glycation reactions, i.e. the reversible condensation of
the aldehyde group of glucose with a protein amino group, forming a Schiff base
followed by an essentially irreversible rearrangement to an Amadori intermediate
(Fig.1). This intermediate undergoes cycles of condensations with additional amines,
dehydrations, and oxidative fragmentations to yield heterogeneous chemical
compounds collectively referred to as advanced glycation end-products (AGEs)1.
Formation of chemically stable AGEs in this Maillard reaction can permanently alter
protein structure and function (for review, see ref. 5-7). They range from CML (the
pathway is shown in Fig.1) to more complex structures such as pentosidine (a lysine-
arginine crosslink) and pyralline (8-10). More AGEs have been identified in the recent
years, among them inter-lysine cross-links crosslines (11) and vesperlysines (12).
Some Maillard products are antigenic, and it has been shown that anti-AGE sera
predominantly recognize CML attached to protein (13). A number of pathways of AGE
formation include oxidative steps catalyzed by redox-active transition metal ions, as
exemplified by the conversion of protein-Amadori intermediate to CML (Fig.1).
CML is a major AGE modification associated with different human pathologic
states. It can be derived either directly from the glycated protein (Fig. 1), or from protein
modification by small reactive carbonyl and dicarbonyl compounds that are either
metabolites or autoxidation products of glucose, ascorbate, Schiff bases, Amadori
intermediates or polyunsaturated lipids (14-16). CML-modified proteins have been found
in plasma, renal tissues, retinas, and collagen of diabetic patients (17-22). CML is the
predominant AGE in intracellular neurofibrillary deposits in patients with Alzheimer
disease (23) and in macrophage-derived foam cells in human atherosclerotic plaques
(24). Its concentration in human tissues increases significantly with age (25). CML-
modified protein has also been reported to be a ligand for RAGE, an AGE receptor that
mediates cellular pro-inflammatory responses (26).
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
4
We earlier reported a novel way to isolate ribose-Amadori intermediate using
high concentration of ribose (27, 28). This ribose-derived intermediate, although not
physiologically relevant, provided an experimental strategy to elucidate mechanisms of
AGE formation and to search for inhibitors that block it. This led to discovery that
pyridoxamine, a natural intermediate of vitamin B6 metabolism, is an efficient in vitro
inhibitor of conversion of the Amadori intermediate to AGEs, particularly CML (28-30).
Under in vivo conditions, pyridoxamine inhibited the accumulation of CML modifications
in skin collagen and in retina, and prevented the development of early renal disease and
retinopathy in the streptozotocin rat and db/db mouse models of diabetes (31-33). In
non-diabetic Zucker obese (fa/fa) rats, the protection of PM against renal and vascular
pathology was also accompanied by the inhibition of CML formation (34).
Pyridoxamine can form moderately stable complexes with a number of transition
metal ions (35, 36). Recently, Baynes et al. reported that PM and other AGE inhibitors
can inhibit copper-dependent oxidation of ascorbic acid and proposed that metal
chelation may play a major role in the inhibition of glycoxidation reactions (37).
However, it remains to be demonstrated whether or not the redox metal ion requirement
and its inhibition by ligand binding are identical in these two reactions. It also has not
been established whether other mechanisms of inhibition, such as covalent adduct
formation of PM with the Amadori, could significantly contribute to AGE inhibition. In the
present work, we demonstrate for the first time that PM can inhibit degradation of a
glucose-derived protein-Amadori intermediate and have utilized this inhibition to isolate
the glucose-derived Amadori at physiological glucose concentrations. The isolation of
this intermediate has enabled the study of the mechanism of PM inhibition of post-
Amadori pathway leading to CML formation, including demonstration by 13C NMR that
no adducts are formed between PM and Amadori. Our overall results support the
hypothesis that PM inhibits post-Amadori glycoxidation reactions by binding to required
redox metal ions and interfering with their catalytic role in these reactions. Since this
PM inhibition occurred in vitro under physiological diabetic concentrations of glucose,
we propose that this PM mechanism contributes to inhibition of AGE formation by PM
observed in diabetic animal models (31-33).
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
5
Experimental procedures Materials. D-ribose, pyridoxamine dihydrochloride, 1-deoxy-1-morpholino-D-
fructose, 3-hydroxypyridine, 4-aminomethylpyridine, and bovine serum albumin were
purchased from Sigma-Aldrich. 5-Nitro-2-hydroxybenzylamine was synthesized by
ChemSyn Laboratories (Lenexa, KS), and its structure and purity were confirmed by
proton NMR and mass spectrometry. D-Glucose was purchased from Gibco BRL.
RNase A was obtained from Worthington Biochemical Co. D-[2-13C]Glucose (99 atom-
% 13C) was purchased from Omicron Biochemicals.
Preparation of glucose-derived and ribose-derived protein-Amadori. For the
preparation of protein-Amadori intermediate under physiologically relevant conditions,
BSA or RNase (20 mg/ml) was incubated with either 5 mM or 30 mM glucose and 20
mM pyridoxamine in 200 mM Na-phosphate pH 7.5 buffer, containing 0.02% sodium
azide, at 37 °C for 80 days. In some cases, protein-Amadori intermediate was prepared
using 100 mM glucose and the conditions described above. Accumulation of BSA-
Amadori was followed using the nitroblue tetrazolium method described below. When
the concentration of Amadori intermediate reached a maximum constant value, glucose
and pyridoxamine were removed by dialysis against 100 volumes of the same Na-
phosphate buffer with 6 buffer changes over a 24 h period. The buffer was equilibrated
at 4°C and dialysis was carried out at this temperature. The complete removal of PM
was confirmed by the absence of PM fluorescence (Ex=324 nm, Em=393 nm).
Ribose-Amadori modified BSA was prepared as described earlier (29). Briefly,
protein was incubated in the presence of 500 mM ribose in 200 mM Na-phosphate pH
7.5 buffer at 37°C for 24 h. Sodium azide (0.02%) was added to the incubations to
prevent bacterial growth. At the end of the incubation, free and Schiff-base ribose was
removed by dialysis at 4°C.
Detection of protein-CML using ELISA. Modifications of protein lysine residues
to CML were measured by ELISA using polyclonal anti-AGE antibody R618 as
described earlier (28-30). Protein-CML was the antigenic epitope in our ELISA
measurements (38).
Quantitative measurements of protein-Amadori. The formation of protein-
Amadori was analyzed using a Sigma Fructosamine kit. The assay is based on the
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
6
ability of Amadori intermediates to reduce nitroblue tetrazolium leading to the formation
of formazan that can be monitored spectrophotometrically at 550 nm. We modified the
Sigma protocol by filtering the working solution prior to use in the assay. This
modification significantly reduced the background signal and improved the
reproducibility of the measurements. Aliquots (40 µl) of protein-Amadori (5 mg/ml) were
mixed with 0.7 ml of Sigma Fructosamine assay working solution and incubated at 37°C
for 10 min. The absorbance of each sample was measured and incubation at 37°C was
resumed. After 10 min, a second absorbance measurement was taken. Amadori
intermediate was quantified by comparing the differences in absorbance between two
measurements for each sample with those of a 1-deoxy-1-morpholino-D-fructose
standard. The protein concentration was determined using a second derivative analysis
of the protein absorbance spectrum to minimize contributions from glycation-related
spectral components (39).
Determination of pKa. The apparent pKa values for 5-nitro-2-hydroxybenzylamine
were determined by spectrophotometric titration at 22°C using an Agilent 8453 UV-
visible spectrophotometer. The value of pKa1 (phenolic group), obtained by least
squares curve fitting of absorbance titration data, was 6.12±0.01. These data indicate
that at physiological pH, the phenolic group of 5-nitro-2-hydroxybenzylamine is present
mostly in the deprotonated state, a condition favoring complex formation between PM
and metal ions (35).
Solution 13C NMR. Amadori-RNase was prepared using D-[2-13C]glucose under
experimental conditions described earlier. After the removal of free and Schiff-base [2-13C]glucose and PM by dialysis at 4 ºC, the buffer was replaced by 50 mM Na-
phosphate buffer prepared with D2O, pD 7.5, using a Centricon-3 centrifugation
concentrator. NMR experiments were performed at 21 ºC on a Varian UnityPlus 600
MHz spectrometer. FID transients (20,000) were collected over spectral widths of 225
ppm (Fig. 6A, B) or 300 ppm (Fig. 6C-E). Larger spectral widths were used in Fig. 6C-E
to determine whether the adducts between PM and acyclic [2-13C]-labeled Amadori
intermediate may have formed upon incubation with 20 mM PM for 8 days at 37ºC. No
signals were detected at δ > 180 ppm in Fig. 6D, E compared to Fig. 6C. In all NMR
experiments, the recycle time was 4.7 s, which was judged sufficient to ensure a return
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
7
to magnetization equilibrium between pulses. All NMR experiments were performed
under waltz-16 decoupling conditions. The FIDs were zero-filled (final digital resolution
0.2 Hz/ data point), apodized with an exponential window function (line broadening = 15
Hz) and Fourier transformed to give the spectra shown in Fig. 6. Reprocessing the
FIDs with a Gaussian-to-Lorentzian window function showed that the apparent singlets
at ~ 97.7 ppm and ~ 95.5 ppm contained significant fine structure. At least 3 resonances
were observed between 97.6 and 97.8 ppm, and at least 6 signals were observed
between 95.3 and 95.6 ppm. Presumably this multiplicity is due to the small effect of
protein site environments on Amadori chemical shifts.
Absorbance measurements. Single-wavelength absorbance measurements,
absorbance spectra, and second derivative analyses were performed on a Hewlett
Packard 8452A diode array spectrophotometer equipped with a Peltier temperature
control unit.
Results
Reversible inhibition of CML formation by pyridoxamine. We initially described a
method to prepare Amadori intermediate in the AGE pathways using high
concentrations of ribose (27). This ribose-derived intermediate provided an
experimental strategy to search for inhibitors that block AGE formation and led to the
discovery of PM inhibition of post-Amadori reactions (28-30). In the present work, we
have used PM to trap and identify glucose-derived protein-Amadori intermediate. As
shown in Fig. 2A, PM completely inhibited formation of immunoreactive BSA-CML
during the initial 35 days of incubation of BSA with 200 mM glucose. When PM and
glucose were removed, the formation of BSA-CML from the putative intermediate was
observed (Fig. 2A, triangles). After 35 days of incubation with glucose, significant
accumulation of fructosyllysine (Amadori) intermediate was detected only in the sample
containing PM (Fig. 2B, bar graph). The structural identity of protein Amadori
intermediate was confirmed using 13C NMR as described vide infra in Fig. 6. Upon the
removal of PM and glucose, BSA-Amadori degraded (Fig. 2B, inset) and converted to
BSA-CML (Fig. 2A, triangles). It is noteworthy that CML is the major but not the only
AGE product of degradation of Amadori intermediate. This complexity of post-Amadori
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
8
reactions may account for the apparent differences in kinetics of Amadori degradation
and formation of Amadori-derived CML (Fig. 2).
The data in Fig. 2 present two important findings. 1) PM inhibits the formation of
glucose-derived protein-CML by inhibiting degradation of protein-bound Amadori
intermediate; 2) this PM property provides a strategy to specifically follow the formation
and degradation of protein-Amadori intermediate.
We utilized the isolation of the glucose Amadori by PM to determine the kinetics
of formation and degradation of BSA-Amadori under physiologically relevant
concentrations of glucose found in the plasma of diabetic patients (>5 mM and up to 30
mM). As shown in Fig. 3A, incubation of BSA with 30 mM glucose at 37°C in the
presence of PM resulted in significant increase in the rate of accumulation of BSA-
Amadori intermediate compared to that at 5 mM glucose. The kinetics was
characterized by t1/2 ~ 9 days, consistent with the notion that relatively short-lived
proteins (t1/2 ~ days) can be modified by glucose in vivo (40, 41). The degree of
modification of protein lysine residues in the albumin-Amadori prepared with diabetic
concentration of glucose was about 9% (Fig. 3A), or ~5 lysine residues per molecule out
of total 58 Lys. This result is in good agreement with preferential modification of 4 lysine
residues found in albumin purified from diabetic patients (42).
The removal of unreacted glucose and PM by extensive dialysis at 4°C followed
by incubation again at 37°C led to the disappearance of ~90% of protein-Amadori
intermediate (FL) after 80 days (Fig. 3A). This process was accompanied by formation
of BSA-CML. Different concentrations of freshly added PM inhibited this CML formation
(Fig. 3B). This inhibition of post-Amadori protein-CML formation is consistent with the
results of our earlier experiments using ribose-derived protein-Amadori intermediate
(28, 30). At 1 mM PM (a 2.5-fold molar excess of PM over Amadori moiety), BSA-CML
formation was completely inhibited (Fig. 3B). Even at the substoichiometric
concentration of PM (0.1 mM), a significant inhibition was still evident (Fig. 3B).
Importantly, this PM concentration is similar to a steady-state pyridoxamine level in
plasma of PM-treated animals (31). The data in Fig. 3 establish that PM can inhibit
CML formation under physiologically relevant solution conditions. Since PM prevented
the degradation of protein-Amadori intermediate, the target of PM action must be post-
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
9
Amadori steps in the CML pathway that lead to irreversible oxidation (Fig. 1). We set
out to establish which of these steps is targeted by PM.
Role of metal ions in inhibition of the post-Amadori CML pathway by PM.
Pyridoxamine is known to form complexes with transition metal ions such as Fe3+ and
Cu2+ and to also inhibit copper-dependent oxidation of ascorbic acid (35-37). We
checked if PM interferes with the oxidative steps in the CML pathway through binding
catalytic metal ions (Fig. 1). The redox metal ion catalysts required for CML formation
occur naturally in the Na-phosphate buffer used in our experiments (43). Electron pairs
on oxygen and nitrogen of the aminomethyl and phenol moieties of PM participate in
bidentate coordination with the metal ion (Fig. 4). If pyridoxamine-metal complex
formation is critical for the inhibition of the conversion of protein-Amadori intermediate to
protein-CML, then modification of these PM functional groups would abolish inhibition.
We utilized ribose-derived BSA-Amadori, a model that exhibits a more rapid rate of
Amadori-to-CML conversion due to the higher abundance of reactive acyclic form of
ribose-Amadori compared to glucose-Amadori (28, 30). We tested three PM structural
analogs to determine their inhibitory potential. 3-Hydroxypyridine and 4-
aminomethylpyridine lack one of the two moieties critical for bidentate complex
formation with the metal ion (Fig. 4A and B). These PM analogs showed no inhibition,
even at very high concentrations (Fig. 4A and B). On the other hand, 5-nitro-2-
hydroxybenzylamine, a compound that has both 3-hydroxy and 4-aminomethyl groups,
showed strong inhibition similar to that reported previously for PM (28-30), even though
its other aromatic ring differs from that of PM (Fig. 4C). These results suggest that PM
inhibits the post-Amadori CML pathway through complex formation with catalytic metal
ions. This conclusion is further supported by the effects of addition of CuCl2 to
incubations of protein-Amadori and PM (Fig. 5). In these experiments, metal ion
contaminants naturally present in phosphate buffer were not removed prior to
incubation. As shown in Fig. 5, the increase in metal concentration significantly
diminished the inhibitory effect of PM on protein-CML formation in a concentration-
dependent fashion, consistent with the importance of metal ion complexation in the PM
inhibition of post-Amadori reactions.
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
10
Lack of reactivity of PM with protein-Amadori intermediate. We demonstrated
recently that PM forms relatively stable five-ring adduct structures with carbonyl groups
of low molecular weight compounds such as glycolaldehyde (38). Since the protein-
bound Amadori moiety, a ketoamine, contains a carbonyl group (Fig. 1), we checked
whether the amino group of PM may form a stable adduct with an acyclic Amadori
carbonyl. In such AGE inhibition mechanism, originally hypothesized for
aminoguanidine (44) but later refuted (45), the adduct formation would inhibit the
conversion of the reactive form of the protein-Amadori intermediate to protein-CML. The
formation of a stable adduct with PM, whether covalent or reversible, would necessarily
shift the tautomeric equilibrium towards the acyclic form of the Amadori intermediate,
thus depleting the cyclic forms (Fig. 1).
This possibility was investigated using 13C NMR and RNase as the model protein,
since resonance assignments for the Amadori moieties have been established
previously (46). Like with BSA, RNase-Amadori intermediate was forming in the
presence of 30 mM glucose and PM, t1/2~22 d (data not shown). For the 13C NMR
experiments the RNase-Amadori intermediate was prepared with 30 mM D-[2-13C]glucose and also with 100 mM D-[2-13C]glucose to achieve a higher degree of lysine
modification and improve the signal-to-noise ratio. The 13C NMR spectra of RNase-[2-13C]-Amadori (Fig. 6A, C) were in good agreement with previously reported resonance
assignments for RNase-[2-13C]-Amadori in the anomeric carbon region (46).
Characteristic resonances (Fig. 6C) arising from the prevailing cyclic β-pyranose forms
(at 95.5, 96.1 and 97.7 ppm) of Amadori intermediates on the ε-amino group of lysine-
41, on ε-amino groups of remaining lysines, and on the N-terminal α-amino group were
observed. Corresponding signals arising from the minor α- and β-furanose forms of
Amadori intermediates on similar amino groups were also detected (101.9, 102.8
(broad) and 104.1 ppm for α; 98.9 (ε-amino group) and 101.4 ppm (ε-amino group of
Lys-41) for β). The signal at 95.8 ppm previously assigned (46) to the α-pyranose
Amadori on ε-amino groups may overlap with that at 95.5 ppm assigned to β-pyranose
Amadori on similar groups. Neither the unmodified RNase nor PM contributed to the
anomeric region of 13C NMR spectrum (data not shown). In RNase-Amadori prepared
under physiological diabetic conditions (30 mM glucose), the observed ratio of modified
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
11
ε-amino to α-amino groups of ~1 appears significantly smaller than reported in RNase-
Amadori prepared with 250 mM glucose (46).
In the presence of either 3 mM or 20 mM PM, no significant change in the
intensities of resonances corresponding to cyclic Amadori forms was observed even
after prolonged incubation (8 days). Importantly, in the experiments shown in Fig. 6B,
6D and 6E, the same PM-to-Amadori molar ratio was used at which PM inhibited
conversion of protein-Amadori to protein-CML (Fig. 3B, data for 0.1 mM and 1mM PM).
If the PM inhibitory effect was due to Amadori-PM adduct formation, it would cause a
shift of the C2 Amadori signals out of the anomeric region of the 13C NMR spectrum,
thus decreasing their intensities. The observed lack of intensity changes and the
absence of additional 13C NMR signals suggests that Amadori-PM adducts, if formed,
are produced in minor amounts. The content of these adducts would be substantially
less than 5% of total protein-Amadori as determined by the sensitivity of 13C NMR
experiments (Fig. 6). Therefore, the formation of such adducts cannot significantly
contribute to PM inhibition of Amadori-to-CML conversion.
Discussion Glucose is a ubiquitous natural metabolite present in all human tissues. In its
aldehyde form it reacts non-enzymatically with nucleophilic groups on proteins,
preferentially with ε-amino groups of lysines and N-terminal α-amino groups. An early
product of this reaction, protein-Amadori, is converted to a variety of stable advanced
glycation end-products or AGEs. Although this reaction is very slow due to the low
abundance of the reactive aldehyde form of glucose at equilibrium (0.003%), the
gradual accumulation of stable AGEs alters protein conformation and function. This
protein modification by glucose is significantly accelerated under hyperglycemic
conditions of diabetes and is considered one of the major factors in the development of
diabetic complications.
Our previous work showed that pyridoxamine, unlike other AGE inhibitors such
as aminoguanidine, blocks the conversion of ribose-derived Amadori intermediate to
protein-CML (28-30). In the present work, we identified the nature of the protein-bound
intermediate formed from glucose in the presence of PM and investigated the
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
12
mechanism of inhibition of post-Amadori protein-CML formation by PM. In the presence
of PM and diabetic levels of glucose (30 mM), the CML precursor was identified as
protein-Amadori intermediate using 13C NMR (Fig. 6). When glucose and PM were
removed, the protein-bound Amadori moiety underwent oxidative degradation and
formed CML, indicating that PM acted on one of the post-Amadori steps of AGE
formation (Fig. 3). The absence of reversible or irreversible adducts between protein-
Amadori intermediate and PM was deduced also using 13C NMR (Fig. 6). The slow
physiologically relevant kinetics of degradation, however, presented an experimental
challenge. Therefore, once the inhibitory effects of PM were established under the
physiological solution conditions, we utilized the experimental system based on ribose-
derived Amadori that we introduced earlier, which allows for significantly higher rates of
post-Amadori reactions (28-30).
We utilized the ribose-derived Amadori to demonstrate that pyridoxamine inhibits
the conversion of protein-Amadori intermediate to protein-CML primarily by binding with
catalytic metal ions and blocking the oxidative steps in the pathway (Fig. 1). This
conclusion is based on several experimental observations: pyridoxamine did not react
with the carbonyl group of protein-Amadori (Fig. 4); the removal of either one of the PM
pyridine ring substituents that participate in bidentate metal ion coordination caused a
complete loss of PM inhibition of post-Amadori CML formation (Fig. 5); the inhibitory
effect of PM can be diminished by increasing concentration of metal ions (Fig. 6).
Importantly, this inhibition occurred at PM concentrations similar to those reported in
preclinical studies of PM efficacy (31) and with the level of protein modification similar to
that found in diabetic patients (Fig. 3).
The inhibition of degradation of important protein glycation intermediate by PM,
which can be reversed by removal of PM, provides an experimental tool to study the
formation and the degradation of protein-Amadori intermediate. Blocking of the post-
Amadori reactions enables the rates and levels of formation of protein-Amadori
intermediate to be easily measurable, even at low physiologically relevant levels of
glucose. Upon the removal of PM and glucose, post-Amadori AGE pathways could be
investigated without the interference from glucose or Schiff-base oxidation products that
may contribute to AGE formation (47, 48). Often, high concentrations of glucose (up to
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
13
1 M) are used in protein glycation reactions in vitro. These reaction conditions lead to
high levels of protein modifications by glucose and enhance additional modifications of
protein by reactive products of glucose degradation. As an illustration, the amount of
albumin-Amadori formed at 200 mM glucose was about 7-fold greater than that formed
at more physiological 30 mM glucose (Fig. 2 and 3). The level of Amadori and AGE
moieties formed at high non-physiological concentration of glucose is likely to have
different impact on structure, function and cellular effects of the proteins compared to
more subtle modifications occurred at physiologically relevant glucose concentrations.
In our experiments, the formation of albumin-Amadori at diabetic glucose concentration
(30 mM) was consistent with elevated levels of glucose-modified proteins in diabetes
(40, 41), while the degree of modification of lysine residues was in good agreement with
that found in albumin purified from diabetic patients (42). Thus the inhibition by PM
provides a simple method for isolation of protein-Amadori intermediates, prepared at
physiological levels of glucose, for studies of its biological effects in cell culture or in
vitro experiments. Under in vivo conditions, Cu2+ and Fe3+ are the most important metal ion
catalysts of oxidative reactions. In blood plasma and in cellular cytoplasm virtually all of
these ions are sequestered in specific metal transporters and other metalloproteins (49,
50). However, copper metalloproteins such as ceruloplasmin and Zn,Cu-superoxide
dismutase and the iron metalloproteins such as hemoglobin and myoglobin have been
shown to undergo significant conformational change and even fragmentation during
glycation reactions, causing the release of bound metal (51-53). Pyridoxamine can form
moderately stable complexes with a number of transition metal ions but has a
preference for Cu2+ and Fe3+ (36, 54). By forming such complexes, PM can reduce
catalytic activity of metal ions released from damaged proteins.
Our results demonstrate that metal binding appears to be critical for PM inhibition
of post-Amadori AGE formation. However, the observation that aminoguanidine also
binds catalytic metal ions (37) but does not inhibit post-Amadori AGE formation (28)
indicates that there are other factors required for efficient inhibition of post-Amadori
reactions by PM. For example, it has been suggested that metal ions could also be
bound by CML clusters on glycated proteins or by native proteins, such as serum
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
14
albumin. Such bound metal ions may retain a significant redox activity for catalysis of
oxidative reactions (55-58). Uniquely, PM may form ternary complexes with such
protein-bound metal ions, making them significantly less active as oxidation catalysts.
Previously, we and others have demonstrated that PM can prevent protein
modifications by chemically trapping free low molecular weight reactive carbonyl
products of glucose and lipid degradation (38, 59-61, Fig. 7). In the present study, we
determined that PM inhibition of post-Amadori CML formation (28-30) is due primarily to
its interference with metal ion catalysis (Fig. 7). It is noteworthy that metal ion
complexation may also contribute to PM inhibition of AGEs derived from free low
molecular weight carbonyl compounds (Fig. 7). Some of these reactions, such as the
formation of CML from glycolaldehyde, require oxidative steps and can be catalyzed by
redox metal ions (14). In this case, complexation of metal ions could be complementary
to the PM carbonyl trapping mechanism as suggested by our previous work (38).
It is clear that the effective inhibition of protein AGE modifications has an
important therapeutic potential. In recent reports, PM alleviated renal pathology in the
streptozotocin rat and spontaneous db/db mouse models of diabetes (31-33), and it
showed successful progress through safety and efficacy clinical trials for diabetic
nephropathy (62). It is possible that different inhibitory mechanisms such as
scavenging of toxic carbonyl metabolites from carbohydrates (38, 59) and lipids (60,
61), and interfering with metal ion catalysis (37 and this study) may operate in vivo. Supporting this notion, PM treatment of diabetic animals has resulted in lower levels of
protein-CML, as well as MOLD, a dicarbonyl-derived protein crosslink (31, 32, 59). The
relative importance of these mechanisms is difficult to estimate because pyridoxamine
action in vivo will be influenced by factors such as local tissue and intracellular
concentrations of redox metal ions, reactive carbonyls, and PM itself. It may also be
affected by competition from endogenous carbonyl and metal scavengers. However,
the greater efficacy of PM, compared to carbonyl scavengers, in diabetic animal studies
(31) suggests that the inhibition of degradation of protein-Amadori via interference with
redox metal ion catalysis may play a more important role in vivo.
The pathogenicity of protein-Amadori intermediate itself remains an open
question. Damaging effects of early protein glycation intermediates in diabetic animals
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
15
and cell culture systems have been reported (63). However, the transient nature of
such intermediates makes it difficult to attribute these effects to specific chemical
structures. The observed effects may reflect the conversion of these early
intermediates to the chemically stable AGEs or may be mediated by the byproducts of
this conversion, such as superoxide. Interestingly, the level of the circulating protein-
Amadori intermediates remained unchanged in diabetic animals treated with PM (31),
even though their conversion to AGEs was blocked. These data suggest the existence
of mechanisms that confer a rapid catabolism of protein-Amadori intermediate in vivo.
Regardless of whether this intermediate is pathogenic itself or simply a precursor to
pathogenic AGEs, the rate of its accumulation at diabetic glucose concentration
suggests that proteins with relatively short lifetimes (days) may undergo modifications in
diabetes (Fig. 3).
The combination of different inhibitory mechanisms may place PM at an
advantage over the compounds functioning primarily as either scavengers of metal ions
or carbonyl trapping agents. For instance, PM reactivity towards dicarbonyl compounds
is significantly lower than that of aminoguanidine (P. A. Voziyan and B. G. Hudson,
unpublished data), but PM exhibits greater efficacy in animal models (31). PM also
exhibits relatively weak metal binding, with a stability constant for Cu2+ several orders of
magnitude lower than that of strong chelators such as EDTA or DTPA (50). Hence, PM
is a potent inhibitor of formation of protein-AGEs that is less likely to interfere with
metabolic pathways that are not the intended targets.
Asknowledgements – We thank Ms. Parvin Todd for excellent technical assistance.
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
16
References: 1. Brownlee, M. (1995) Annu-Rev-Med. 46, 223-234
2. Colaco, C. A. and Harrington, C. R. (1994) Neuroreport. 5, 859-861
3. Thorpe, S. R. and Baynes, J. W. (1996) Drugs Aging. 9, 69-77
4. Acosta, J., Hettinga, J., Fluckiger, R., Krumrei,N., Goldfine, A., Angarita, L., and
Halperin, J. (2000) Proc Natl Acad Sci U S A. 97, 5450-5455.
5. Ahmed, M.U., Thorpe, S.R., and Baynes, J.W. (1986) J Biol Chem. 261, 4889-4894
6. Sell, D.R. and Monnier, V.M. (1989) J Biol Chem. 264, 21597-21602
7. Hayase, F, Nagaraj, R.H., Miyata, S., Njoroge, F.G., and Monnier, V.M. (1989) J Biol
Chem. 264, 3758-3764.
8. Ienaga, K., Nakamura, K., Hochi, T., Nakazawa, Y., Fukunaga, Y., Kakita, H., and
Nakano, K. (1995) Contrib Nephrol. 112, 42-51
9. Nakamura, K., Nakazawa, Y., and Ienaga, K. (1997) Biochem Biophys Res Commun. 232,
227-230
10. Reddy, S., Bichler, J., Wells-Knecht, K.J., Thorpe, S.R., and Baynes, J.W. (1995)
Biochemistry 34, 10872-10878
11. Baynes, J.W., and Thorpe, S. R. (1999) Diabetes. 48, 1-9
12. Monnier, V. M. (2001) J. Clin. Invest. 107, 799-801
13. Ikeda, K., Higashi, T., Sano, H., Jinnouchi, Y., Yoshida, M., Araki, T., Ueda, S., and
Horiuchi, S. (1996) Biochemistry 35, 8075-8083
14. Glomb, M. A. Monnier, V. M. (1995) J. Biol. Chem. 270, 10017-10026
15. Dunn, J.A., Ahmed, M.U., Murtiashaw, M.H., Richardson, J.M., Walla, M.D., Thorpe,
S.R., and Baynes, J.W. (1990) Biochemistry 29, 10964-10970
16. Requena, J.R., Fu, M.X., Ahmed, M.U., Jenkins, A.J., Lyons, T.J., and Thorpe, S.R.
(1996) Nephrol Dial Transplant. 11 Suppl 5, 48-53
17. Misselwitz, J., Franke, S., Kauf, E., John, U., and Stein, G. (2002) Pediatr. Nephrol.
17, 316-321
18. Wendt, T., Tanji, N., Guo, J., Hudson, B.I., Bierhaus, A., Ramasamy, R., Arnold, B.,
Nawroth, P.P., Yan, S.F., D'Agati, V., and Schmidt, A.M. (2003) J Am Soc Nephrol.
14, 1383-1395.
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
17
19. Uesugi, N., Sakata, N., Horiuchi, S., Nagai, R., Takeya, M., Meng, J., Saito, T., and
Takebayashi, S. (2001) Am J Kidney Dis. 38, 1016-1025
20. Monnier, VM; Bautista, O; Kenny, D; Sell, DR; Fogarty, J; Dahms, W; Cleary, PA;
Lachin, J; Genuth, S (1999) Diabetes 48, 870-880
21. Dyer, D.G., Dunn, J.A., Thorpe, S.R., Bailie, K.E., Lyons, T.J., McCance, D.R., and
Baynes, J.W. (1993) J.Clin.Invest. 91, 2463-2469
22. Mott, J.D., Khalifah, R.G., Nagase, H., Shield, C.F., Hudson, J.K., and Hudson B.G.
(1997) Kidney Int. 52, 1302-1312
23. Castellani, R.J., Harris, P.L., Sayre, L.M., Fujii, J., Taniguchi, N., Vitek, M.P., Founds,
H., Atwood, C.S., Perry,G., and Smith, M.A. (2001) Free Radic Biol Med. 31, 175-
180
24. Imanaga, Y., Sakata, N., Takebayashi, S., Matsunaga, A., Sasaki, J., Arakawa, K.,
Nagai, R., Horiuchi, S., Itabe, H., and Takano, T. (2000) Atherosclerosis 150, 343-
355
25. Verzijl, N., DeGroot, J., Oldehinkel, E., Bank, R.A., Thorpe, S.R., Baynes, J.W.,
Bayliss, M.T., Bijlsma, J.W., Lafeber, F.P., and Tekoppele, J.M. (2000) Biochem J.
350, 381-387
26. Kislinger, T., Fu, C., Huber, B., Qu, W., Taguchi, A., Yan, S.D., Hofmann, M., Yan,
S.F., Pischetsrieder, M., Stern, D., and Schmidt, A.M. (1999) J. Biol. Chem. 274,
31740-31749
27. Khalifah, R.G., Todd, P., Booth, A. A., Yang, S. X., Mott, J.D., and Hudson, B.G
(1996) Biochemistry. 35, 4645-4654
28. Khalifah, R.G., Baynes, J. W., and Hudson, B.G. (1999) Biochem. Biophys.Res.
Commun. 257, 251-258
29. Booth, A.A., Khalifah, R.G., Todd, P., and Hudson, B. G. (1997) J. Biol. Chem. 272,
5430-5437
30. Booth, A.A., Khalifah, R.G.,and Hudson, B.G. (1996) Biochem. Biophys.
Res.Commun. 220, 113-119
31. Degenhardt, T.P., Alderson, N.L., Arrington, D.D., Beattie, R.J., Basgen, J.M., Steffes,
M.W., Thorpe, S.R., and Baynes, J.W. (2002) Kidney Int. 61, 939-950
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
18
32. Stitt, A., Gardiner, T.A., Alderson, N.L., Canning, P., Frizzell, N., Duffy, N., Boyle, C.,
Januszewski, A.S., Chachich, M., Baynes, J.W., and Thorpe SR. (2002) Diabetes. 51,
2826-2832
33. Zheng, F., Leclercq, B., Elliot, S.J., Striker, L.J., and Striker, G.E. (2002). J. Am Soc
Nephrol, 13, 534A
34. Alderson, N.L., Chachich, M.E., Yousseff, N.N., Beattie, R.J., Nachigal, M., Thorpe, S.R.,
and Baynes, J.W. (2003) Kidney Int. 63, 2123-2133
35. Thompson, D. M., Balenovich, W., Hornich, L.H.M., and Richardson, M. F. (1980).
Inorg. Chim. Acta 46, 199-203
36. Gustafson, R. L. and Martell, A. E. (1957) Arch. Biochem. Biophys. 68, 485-498
37. Price, D.L., Rhett, P.M., Thorpe, S.R., and Baynes, J.W. (2001) J. Biol. Chem. 276,
48967-48972
38. Voziyan, P. A., Metz, T. O., Baynes, J. W., and Hudson, B.G. (2002) J. Biol. Chem.
277, 3397-3403
39. Butler, W.L. (1979) Methods Enzymol. 56, 505-515
40. McFarland, K.F., Catalano, E.W., Day, J.F., Thorpe, S.R., and Baynes, J.W. (1979)
Diabetes 28, 1011-1014
41. Bunn, H.F., Gabbay, K.H., and Gallop, P.M. (1978) Science, 4337, 21-27 42. Iberg, N. and Fluckiger, R. (1986) J Biol Chem. 261, 13542-13545
43. Buettner, G.R. (1988) J.Biochem.Biophys.Methods. 16, 27-40
44. Brownlee, M., Vlassara, H., Kooney, A., Ulrich, P., and Cerami, A. (1986) Science
232, 1629-1632
45. Chen, H.-J. and Cerami, A (1993) J. Carbohydr. Chem. 12, 731-742
46. Neglia, I.C., Cohen, H.J., Garber, A.R., Thorpe, S.R., Baynes, J.W. (1985) J Biol Chem.
260, 5406-5410
47. Wolff, S.P., and Dean, R.T. (1987) Biochem. J. 245, 243-250
48. Hayashi, T. and Namiki, M. (1986) in Amino-Carbonyl Reactions in Food and
Biological Systems (Fujimaki, M., Namiki, M., and Kato, H., eds.) pp. 29-35.
49. Evans, P.J., Bomford, A., Halliwell, B. (1989) Free Radic. Res. Commun. 7, 55-62
50. Rae, T.D., Schmidt, P.J., Pufahl, R.A., Culotta, V.C., O'Halloran, T.V. (1999) Science
284, 805-508
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
19
51. Takata, I., Kawamura, N., Myint, T., Miyazava, N., Suzuki, K., Maruyiama, N., Mino,
M., and Taniguchi, N. (1996) Biochem. Biophys. Res. Commun. 219, 243-248
52. Islam, K.N., Takahashi, M., Higashiyama, S., Myint, T., Uozumi, N. (1995) J.
Biochem. 118, 1054-1060
53. Cussimanio, B.L., Booth, A.A., Todd, P., Hudson, B.G., and Khalifah, R.G. (2003)
Biophys. Chem., 105, 739-752
54. Data for Biochemical Research. Compiled by Dawson, R. M. C,. Elliott, D. C., Elliott,
W. H., and Jones, K. M. Clarendon Press, Oxford, 1986. p. 412
55. Hawkins, C.L., and Davies, M.J. (1997) Biocim. Biophys. Acta 1360, 84-96
56. Mukhopadhyay, Ch. K., Ehrenwald, E., and Fox, P.L. (1996) J. Biol. Chem. 271,
14773-14778
57. Marx, G., and Chevion, M. (1986) Biochem. J. 236, 397-400
58. Saxena,A.K., Saxena,P., Wu, X., Obrenovich, M., Weiss, M.F., and Monnier, V.M.
(1999) Biochem.Biophys.Res.Commun. 260, 332-338
59. Nagaraj, R.H., Sarkar, P., Mally, A., Biemel, K.M., Lederer, M.O., and Padayatti, P.S.
(2002) Arch. Biochem. Biophys. 402, 110-119
60. Onorato, J.M., Jenkins, A.J., Thorpe, S.R., and Baynes, J.W. (2000) J Biol Chem.
275, 21177-21184
61. Metz, T. O., Alderson, N. L., Chachich, M. E., Thorpe, S. R., and Baynes, J. W. (2003)
J. Biol. Chem., Aug 15, Epub ahead of print 62. Degenhardt, T. P., Khalifah, R. G., Klaich, G. M., and Schotzinger, R. J. (2002) J. Am.
Soc. Nephrol. 13, 652A
63.Chen, S., Cohen, M. P., and Ziadeh, F. N. (2000) Kidney Int. 58, S40-S44
64. Smith PR, Thornalley PJ. (1992) Eur J Biochem. 210, 729-739
65. Thornalley, P., Wolff, S., Crabbe, J., Stern, A. (1984) Biochim. Biophys. Acta 797,
276-287.
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
20
Figure legends
Figure 1. Pathway of formation of glucose-derived protein-Amadori and protein-
CML (from ref. 5, 47, and 64). Reversible condensation of the aldehyde group of
glucose with a protein amino group results in formation of a Schiff base followed by
essentially irreversible rearrangement to an Amadori intermediate. The Amadori
intermediate enolizes to the 2,3-enediol in a reaction catalyzed by phosphate anions
(65). The 2,3-enediol undergoes spontaneous autoxidation involving the formation of
superoxide, catalyzed by transition metal ions; hydrogen peroxide, formed by
superoxide dismutation, regenerates the catalytic metal oxidation state. The putative
dicarbonyl product of 2,3-enediol autoxidation undergoes further oxidative degradation,
producing Nε-carboxymethyllysine and D-erythronic acid.
Figure 2. Reversible inhibition of the formation of glucose-derived BSA-CML by
pyridoxamine. (A) BSA (5 mg/ml) was incubated with 200 mM glucose and 20 mM
pyridoxamine in 200 mM Na-phosphate buffer containing 0.02% sodium azide, pH 7.5,
at 37°C for 35 days. Glucose and pyridoxamine were removed by dialysis at 4°C; the
complete removal of PM was confirmed by the absence of a fluorescence signal at
Ex=324 nm, Em=393 nm. After dialysis, the protein solution was again incubated at
37°C in the same buffer for up to 120 days (triangles). An aliquot from this sample was
incubated continuously without the removal of PM or glucose (squares). Control sample
did not contain PM (circles). Modifications of protein lysine residues to CML were
measured by ELISA as described under Experimental procedures. Each data point
represents an average of two measurements. (B) After 35 days of incubation, the
content of fructosyllysine in the samples was determined using the nitroblue tetrazolium
method described under Experimental procedures. Inset: after the complete removal of
glucose and PM, the protein solution was again incubated at 37°C in the same buffer for
up to 120 days followed by the determination of fructosyllysine. Each data point
represents an average of two measurements.
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
21
Figure 3. Kinetics of formation and degradation of BSA-Amadori at physiological
levels of glucose. (A) BSA (20 mg/ml) was incubated with either 5 mM (open circles) or
30 mM glucose (filled circles) and 20 mM pyridoxamine in 200 mM Na-phosphate buffer
containing 0.02% sodium azide, pH 7.5, at 37°C. Accumulation of BSA-Amadori
(fructosyllysine, FL) was followed using the nitroblue tetrazolium method described
under Experimental procedures with the baseline readings (BSA alone incubated under
the same conditions) subtracted. For the incubations with 30 mM glucose: when the
concentration of Amadori reached a maximum constant value, glucose and
pyridoxamine were removed by dialysis at 4°C; the complete removal of PM was
confirmed by the absence of a fluorescence signal at Ex=324 nm, Em=393 nm. After
dialysis, the protein solution was again incubated at 37°C in the same buffer and
degradation of Amadori-BSA was followed for up to 160 days. (B) BSA-Amadori was
prepared with 30 mM glucose as described above. After dialysis, BSA-Amadori (5
mg/ml) was again incubated at 37°C in the same buffer either without PM (circles) or
with different concentrations of freshly added PM: 0.1 mM (diamonds), 1 mM (squares)
or 15 mM (triangles). Inverted triangles represent the data from incubations of
unmodified BSA. Modifications of protein lysine residues to CML were measured by
ELISA as described under Experimental procedures. Each data point represents an
average of two measurements.
Figure 4. Inhibition of post-Amadori CML formation by structural analogs of PM.
Ribose-derived BSA-Amadori intermediate was prepared as described under
Experimental procedures. After the removal of sugar by extensive dialysis at 4°C,
incubation at 37°C was resumed in the presence of the indicated concentrations of 3-
hydroxypyridine (A), 4-aminomethylpyridine (B), and 5-nitro-2-hydroxybenzylamine (C).
CML was measured using ELISA; each experimental point represents an average of
two measurements. Note the differences in the concentration range used in (A) and (B)
compared to that in (C).
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
22
Figure 5. Inhibition of post-Amadori CML formation by pyridoxamine at different
metal concentrations. For the preparation of protein-Amadori intermediate, BSA (10
mg/ml) was incubated with 0.5 M ribose in 200 mM Na-phosphate buffer containing
0.02% sodium azide, pH 7.5, at 37°C for 24 h. After the removal of sugar by an
extensive dialysis at 4°C, incubation at 37°C was resumed in the presence of the
indicated concentrations of PM in buffer alone (circles), buffer containing 10 µM CuCl2
(triangles) or buffer containing 100 µM CuCl2 (diamonds). After 48 h, BSA-CML was
measured by ELISA; each symbol represents an average of two measurements.
Figure 6. 13C NMR spectra of RNase-[2-13C]-Amadori with and without PM.
Amadori-RNase was prepared using either 30 mM D-[2-13C]glucose (A and B) or 100
mM D-[2-13C]glucose (C, D and E) and experimental conditions described earlier. After
removal of the free [2-13C]glucose and PM by dialysis, the buffer was replaced by 50
mM Na-phosphate buffer (pD 7.5), prepared with D2O. NMR experiments were
performed on a Varian UnityPlus 600 MHz NMR spectrometer at 21ºC, collecting
20,000 transients per spectrum (see Experimental Procedures). (A) and (C) Spectra of
RNase-[2-13C]-Amadori without PM using two samples at 4.1 mM (A; Sample 1) and
3.8 mM (C; Sample 2) protein concentration. (B) Spectrum of Sample 1 after the
addition of 3 mM PM and incubation for 24 h (final protein concentration = 4.0 mM). (D)
and (E) Spectra of Sample 2 after incubation with 20 mM PM for 1 day (D) or 8 days
(E). RNase concentration was 3.1 mM in D and E. Abbreviations pyr and fur,
respectively, designate the pyranose and the furanose forms of the protein-Amadori
intermediate.
Figure 7. A model for inhibition of AGE/ALE formation by PM. The numbers
represent steps in AGE pathways inhibited by PM. Step 1. PM prevents conversion of
protein-Amadori to protein-CML via complexation of catalytic metal ions (present work).
Step 2. PM inhibits protein modifications by scavenging free carbonyl products of
glucose and lipid degradation (38, 59-61). Step 3. PM may also inhibit metal-catalyzed
oxidative steps in protein modifications by low molecular weight carbonyl compounds
(ref. 38 and present work).
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
C=OH-C-OH
--H
-HO-C-H
-
glucose aldehyde
Schiff base Amadori intermediate
Mn+ M(n-1)+
O2
2,3-enediol
C=OH2C-NH-protein-
OH
-
Nε -carboxymethyllysine (CML)
erythronic acid
H2N-protein
C3H7O3
OH
-C=O
-
C3H7O3
C=H-C-OH
--H
-HO-C-H
-
C3H7O3
N-protein H2C-NH-proteinC=O
--HO-C-H
-
C3H7O3
C-OHH2C-NH-protein-
-C-OH
=
C3H7O3
C=OH2C-NH-protein-
-C=O
-
C3H7O2
Mn+
H2O2
M(n-1)+
1-amino-1-deoxy-fructofuranose form of protein-Amadori
1-amino-1-deoxy-fructopyranose form of protein-Amadori
O
CH2NH
OH
protein
OHOCH2 CH2NH
OH
protein
Voziyan et al., Fig.1
OHOH OH
HOOH
H2O2
OH.
by guest on March 17, 2020 http://www.jbc.org/ Downloaded from
Voziyan et al., Fig.2
Time (d)0 20 40 60 80 100 120
BSA
-CM
L(E
LISA
read
ings
at 4
10 n
m)
0.5
1.0
1.5
2.0
FL (A
bs a
t 550
nm
)
0.0
0.1
0.2
0.3
0.4
A
B
Time (d)
40 60 80 100 120
FL (m
ol/m
ol L
ys)
0.2
0.4
0.6
BSA aloneafter 35 d
BSA/glucose after 35 d
BSA/glucose+PM after 35 d
BSA/glucose
BSA/glucose+PM
Glucose and PM removed
by guest on March 17, 2020 http://www.jbc.org/ Downloaded from
Voziyan et al., Fig.3
Incubation time (d)0 20 40 60 80 100 120 140 160
BSA
-CM
L(E
LISA
read
ings
at 4
10 n
m)
0.1
0.2
0.3
Incubation time (d)0 20 40 60 80 100 120 140 160
FL (m
ol/m
ol L
ys)
0.00
0.02
0.04
0.06
0.08
0.10 30 mM Glucose and 20 mM PM removed
30 mM Glucose and 20 mM PM removed
A
B
by guest on March 17, 2020 http://www.jbc.org/ Downloaded from
Time (days)0 1 2 3 4 5 6
ELIS
A R
eadi
ng (4
10 n
m)
0.0
0.1
0.2
0.3
0.4
0.5
0.6 3 mM NHBA 1 mM NHBA 0.5 mM NHBA 0.1 mM NHBA none
O
CH2
NH2
O
H
N2
O
CH2
NH2
O
H
N2
0 2 4 6 8 10 12
0.0
0.5
1.0
1.5
2.0 none3 mM 3 - HP15 mM 3 - HP50 mM 3 - HP
ELIS
A R
eadi
ng (4
10 n
m)
Time (days)
N
OH
N
OH
0 2 4 6 8 10
0.0
0.5
1.0
1.5
2.0none3 mM 4- AMP15 mM 4 - AMP50 mM 4 -AMP
Time (days)
N
CH2
NH2
N
CH2
NH2
ELIS
A R
eadi
ng (4
10 n
m)
A
B
3-hydroxypyridine
4-aminomethylpyridine
5-nitro-2-hydroxybenzylamine
C
H2NH2C
O
O
NH2
H
N
C 2
H3C
OH
M2+
CH3
N
OH
Pyridoxamine metal complex (from ref. 35 and 36)
Voziyan et al., Fig. 4
by guest on March 17, 2020 http://www.jbc.org/ Downloaded from
Concentration of PM (mM)0 2 4 6 8 10
Nor
mal
ized
ELI
SA re
adin
gs
0.2
0.4
0.6
0.8
1.0
Voziyan et al., Fig.5
100 µM CuCl2
10 µM CuCl2
No CuCl2 added
by guest on March 17, 2020 http://www.jbc.org/ Downloaded from
Voziyan et al., Fig.6
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from
Reducing sugars
Protein-Amadori adducts
Protein
Protein
Reactive carbonyl compounds
AGE/ALE-modified proteins
PM-carbonyl adducts
Mn+ / [
O 2]
M n+/ [O2 ]
PM
(Mn+)PM2
PM
(Mn+)PM2
Voziyan et al., Fig.7
13
2 PM
Lipids
by guest on March 17, 2020 http://www.jbc.org/ Downloaded from
Anthony S. Serianni and Billy G. HudsonPaul A. Voziyan, Raja G. Khalifah, Christophe Thibaudeau, Alaattin Yildiz, Jaison Jacob,
through binding of redox metal ionsinhibits conversion of amadori intermediate to advanced glycation end-products Modification of proteins In vitro by physiological levels of glucose: Pyridoxamine
published online September 15, 2003J. Biol. Chem.
10.1074/jbc.M307155200Access the most updated version of this article at doi:
Alerts:
When a correction for this article is posted•
When this article is cited•
to choose from all of JBC's e-mail alertsClick here
by guest on March 17, 2020
http://ww
w.jbc.org/
Dow
nloaded from