Glycated human serum albumin isolated from poorly controlled diabetic patients impairs cholesterol efflux from macrophages: An investigation by mass spectrometry

Article abstract of DOI:10.1255/ejms.1322

Advanced glycation end-products impair ABCA-1-mediated cholesterol efflux by eliciting inflammation, the generation of reactive oxygen species and endoplasmatic reticulum (ER) stress. The glycation level of human serum albumin (HSA) from type 1 and type 2 diabetic patients was determined by matrix-assisted laser desorption/ionization (MALDI) mass spectrometry and related to possible impairment of ER function and cellular cholesterol efflux. Comparison of the MALDI spectra from healthy and diabetic subjects allowed us to determine an increased HSA mean mass of 1297 Da for type 1 and 890 Da for type 2. These values reflect a mean condensation of at least 8 glucose units and 5 glucose units, respectively. Mouse peritoneal macrophages were treated with HSA from control, type 1 and type 2 diabetic subjects in order to measure the expression of Grp78, Grp94, protein disulfide isomerase (PDI), calreticulin (CRT), and ABCA-1. 14C-cholesterol overloaded-J774 macrophages were treated with HSA from control and diabetic subjects and further incubated with apo A-1 to determine the cholesterol efflux. Combined analyses comprising HSA from type 1 and type 2 diabetic patients were performed in cellular functional assays. In macrophages, PDI expression increased 89% and CRT 3.4 times in comparison to HSA from the control subjects. ABCA-1 protein level and apo A-I-mediated cholesterol efflux were, respectively, 50% and 60% reduced in macrophages exposed to HSA from type 1 and type 2 diabetic patients when compared to that exposed to HSA from control subjects. We provide evidence that the level of glycation that occurs in albumin in vivo damages the ER function related to the impairment in macrophage reverse cholesterol transport, and so contributes to atherosclerosis in diabetes.

Full text of DOI:10.1255/ejms.1322

G. Castilho et al., Eur. J. Mass Spectrom. 21, 233–244 (2015)
233
Received: 3 October 2014
n
Revised: 2 March 2015 n Accepted: 12 March 2015 n Publication: 29 April 2015
EUROPEAN
JOURNAL
OF
MASS
SPECTROMETRY
Special Issue Celebrating the 20th Anniversary of EJMS—European Journal of Mass Spectrometry
Glycated human serum albumin
isolated from poorly controlled diabetic
patients impairs cholesterol efflux from
macrophages: an investigation by mass
spectrometry
Gabriela Castilho,a Camila H Sartori,a Adriana Machado-Lima,a Edna R Nakandakare,a Maria Lucia C
Corrêa-Giannella,b Marco Roverso,c Simona Porcu,c Annunziata Lapolla,c Pietro Traldid and Marisa Passarellia
aLipids Laboratory (LIM 10), University of São Paulo Medical School, São Paulo, SP, Brazil
bCellular and Molecular Endocrinology Laboratory (LIM 25), University of São Paulo Medical School, São Paulo, SP, Brazil
cDepartment of Medicine, University of Padova, via Giustiniani 2, I35100 Padova, Italy
dCNR-IENI, Corso Stati Uniti 4, I35127 Padova, Italy. E-mail: pietro.traldi@adr.pd.cnr.it
Advanced glycation end-products impair ABCA-1-mediated cholesterol efflux by eliciting inflammation, the generation of reactive oxy-
gen species and endoplasmatic reticulum (ER) stress. The glycation level of human serum albumin (HSA) from type 1 and type 2 diabetic
patients was determined by matrix-assisted laser desorption/ionization (MALDI) mass spectrometry and related to possible impair-
ment of ER function and cellular cholesterol efflux. Comparison of the MALDI spectra from healthy and diabetic subjects allowed us to
determine an increased HSA mean mass of 1297Da for type 1 and 890Da for type 2. These values reflect a mean condensation of at least
8 glucose units and 5 glucose units, respectively. Mouse peritoneal macrophages were treated with HSA from control, type 1 and type 2
diabetic subjects in order to measure the expression of Grp78, Grp94, protein disulfide isomerase (PDI), calreticulin (CRT), and ABCA-1.
14C-cholesterol overloaded-J774 macrophages were treated with HSA from control and diabetic subjects and further incubated with apo
A-1 to determine the cholesterol efflux. Combined analyses comprising HSA from type 1 and type 2 diabetic patients were performed
in cellular functional assays. In macrophages, PDI expression increased 89% and CRT 3.4 times in comparison to HSA from the con-
trol subjects. ABCA-1 protein level and apo A-I-mediated cholesterol efflux were, respectively, 50% and 60% reduced in macrophages
exposed to HSA from type 1 and type 2 diabetic patients when compared to that exposed to HSA from control subjects. We provide evi-
dence that the level of glycation that occurs in albumin in vivo damages the ER function related to the impairment in macrophage reverse
cholesterol transport, and so contributes to atherosclerosis in diabetes.
Keywords: diabetes mellitus, advanced glycated albumin, endoplasmic reticulum stress, ABCA-1, atherosclerosis, reverse cholesterol
transport, MALDI-MS
Introduction
Diabetes mellitus (DM) is characterized by an increased
frequency of cardiovascular disease,1 which consequently
cannot be explained exclusively by known risk factors, such
as smoking, hypertension, and atherogenic dyslipidemia.
ISSN: 1469-0667
doi: 10.1255/ejms.1322
© IM Publications LLP 2015
All rights reserved

234
Glycated HSA from Diabetic Patients Impairs Cholesterol Efflux from Macrophages
Among the classical biochemical alterations induced by
hyperglycemia, the formation of advanced glycation end- studies on healthy subjects and on well-controlled and
glycation processes on selected proteins,15–19 extensive
products (AGEs), an increased polyol and hexosamine pathway
flux, and protein kinase C activation have been demonstrated
as relevant in diabetes.2 All of these molecular mechanisms
reflect a single hyperglycemia-induced process of superoxide
badly controlled diabetic patients were undertaken. It was
shown that a clear increase of the m/z value of the [M + H]+
species from HSA, as well as other plasma proteins (IgG and
hemoglobin), is present in the case of the patients.20–24 The
overproduction by the mitochondrial electron-transport chain. observed increase of mass allows us to evaluate the lowest
Thus, hyperglycemia and an increased oxidative stress lead to number of glucose molecule condensed on the protein caused
tissue damage via common pathways.3 Taking into considera- by the occurrence of a Maillard reaction. The number (n) of
tion the role of AGEs, a series of studies have been performed
in various cell systems, such as monocytes/macrophages and
endothelial cells, exhibiting specific cellular receptors that
bind AGE-modified proteins (RAGEs).4
condensed glucose molecules is calculated by dividing the
difference of mass between the HSA from diabetic patients
and the HSA from control subjects by 162 (the molecular
weight of the condensed glucose). However, the occurrence of
The binding of AGEs on RAGEs induces intracellular oxida- dehydration and oxidation processes led to a mass decrease
tive stress and the production of cytokines, thus leading to
of the glucose condensed moiety and consequently n can be
vascular damage.4 In this frame, we have previously demon- higher.
strated that AGEs alter macrophage lipid homeostasis by
inducing reactive oxygen species (ROS) generation,5 7-ketco-
cholesterol and total sterols accumulation,6 inflammation,7
Experimental procedures
and endoplasmic reticulum (ER) stress and unfolded protein
response.8 These conditions were related to a reduction in
ATP-binding cassette transporter A-1 (ABCA-1) content level
and also to a diminished rate of cholesterol efflux to apo A-I
that led to intracellular cholesterol accumulation.5–9
Cell culture
Mouse peritoneal macrophages were harvested after the
injection of 6mL of phosphate-buffered saline (PBS) into the
peritoneal cavity. After a soft abdominal massage, cells were
ABCA-1 is responsible for the excess cholesterol exporta- collected into sterile tubes, centrifuged, and resuspended in
tion from arterial wall macrophages to apo A-I, playing an
important role in the generation of HDL and in the prevention
of atherosclerosis.10 ER stress can be elicited by changes in
RPMI 1640 containing 10% fetal calf serum (FCS), 1% peni-
cillin–streptomycin, and 4mM l-glutamine. Cells were plated
in Petri dishes or 48-well plates. J774 macrophages were
calcium flow, overloading of protein synthesis, and distur- cultured in RPMI 1640 containing 10% FCS, 1% penicillin–
bances in N- and O-glycosylation. In addition, ER dysfunc- streptomycin and 4mM l-glutamine and maintained in a 5%
tion can also occur through oxidative stress, inflammation, CO2 incubator at 37°C. In all cell-culture experiments, the cell
and lipid accumulation, conditions also elicited by AGE. ER
resident chaperones, such as Grp78, Grp94, protein disulfide
isomerase (PDI), and calreticulin (CRT) help proper protein
folding, avoiding the aggregation of misfolded peptides
preventing their cytotoxic effects. Besides, misfolded proteins
can be targeted for degradation by the ubiquitin/proteasome
system (UPS).11 In macrophages, the ER is the main site of
free-cholesterol cytotoxicity, which is related to macrophage
apoptosis.12 In addition, ER stress markers have been found
elevated in rupture-sensitive areas of a human atherosclerotic
lesion,13 as well as in the endothelium of atherosusceptible
arterial sites.14
It is emphasized that there is no evidence in the literature
on the role played by the in vivo human serum albumin (HSA)
glycation in the induction of ER stress in macrophages and
how this relates to the ABCA-1 functionality. We thought it
was of interest to relate the glycation level of HSA drawn
from poorly controlled type 1 diabetic (DM1) patients and type
2 diabetic (DM2) patients to the ER dysfunction, in order to
investigate possible mechanisms linking HSA glycation to the
impairment in macrophage cholesterol efflux.
viability, controlled by exclusion with Trypan blue, was superior
to 98% and no cell toxicity was observed as measured by the
release of lactate dehydrogenase into the culture medium
(in vitro toxicology assay kit, Sigma Aldrich, St. Louis, MO, USA).
During the experimental treatments, cells were maintained in
Dulbecco’s modified Eagle’s medium (DMEM) containing 1%
fatty acid free albumin (FAFA) or albumin isolated from human
serum as indicated.
Patients
The control subjects were selected at the Faculty of Medical
Sciences, University of São Paulo, Brazil. Type 1 and type 2
DM patients with HbA1c > 7.5% were selected at the Clinical
Hospital of the Faculty of Medical Sciences, University of São
Paulo. All the participants signed an informed written consent
form previously approved by The Ethical Committee for
Human Research Protocols of the Clinical Hospital (protocol
#740-06). None of the participants had chronic diseases other
than DM, such as liver, renal, or thyroid disease. Subjects
diagnosed with microalbuminuria or alcohol abuse were not
included.
Matrix-assisted laser desorption/ionization (MALDI) mass
spectrometry was proved to be a valid tool to evaluate the
glycation level of HSA. After some investigations of in vitro
In the control group (n = 6), two subjects were on enalapril,
one subject was on l-thyroxine, and two were smokers. All
type 1 diabetes patients (n = 6) were on neutral protamine

G. Castilho et al., Eur. J. Mass Spectrom. 21, 233–244(2015)
235
Control and DM albumin digestion
Hagedorn (NPH) and took lispro insulin before meals. One
patient used of l-thyroxine. Among the type 2 diabetes
patients (n = 6) all used NPH insulin, four on metformin, three
on sulfonylurea, five on statins, three on enalapril, one on
fibrate, one on pioglitazone, and two on aspirin.
Blood samples were drawn from all subjects after a 12 h
fasting period for measuring plasma fructosamine and
glycemia. HbA1c was determined by high-performance
liquid chromatography in accordance with the US National
Glycohemoglobin Standardization Program, and glycated
albumin was determined by enzyme-linked immunosorbent
The enzymatic digestion was performed by incubating the HSA
solutions with a 0.2mgmL–1 trypsin solution (sample/trypsin
ratio 40.5:1), for 12h at 37°C. The tryptic digestion products
wereanalyzedbyMALDI-MSusinga-cyano-4-hydroxycinnamic
acid as matrix. An external mass calibration was obtained
using the Peptide Calibration Standard (Bruker Daltonics,
Bremen, Germany), based on the monoisotopic values of:
angiotensin II [M + H]+ (m/z 1046.54), angiotensin I [M + H]+
(m/z 1296.68), substance P [M + H]+ (m/z 1347.73), bombesin
[M + H]+ (m/z 1619.82), adrenocorticotropic hormone (ACTH)
assay (ELISA) (Kit Lucica GA-L, Asahi Kasei Pharma, Tokyo, clip 1–17 [M + H]+ (m/z 2093.07), ACTH clip 18–39 [M + H]+
Japan). The carboxymethyllysine (CML) content in HSA
samples was measured by ELISA (CML-CIRCULEX-8066,
CycLex Co, Japan).
(m/z 2465.20), and somatostatin 28 [M + H]+ (m/z 3147.47).
Determination of ER chaperones in
macrophages incubated with C- and DM-HSA
Isolation and purification of serum albumin
Mouse peritoneal macrophages were treated for 18 h with
Preservatives [2 mmol L–1 benzamidine (5 mL mL–1), 0.5%
control HSA (C-HSA) or DM-HSA at 1mgmL–1 concentration.
gentamicin plus 15 mmol L–1 chloramphenicol (20 mL mL–1), Following treatment, cells were scraped into Tris-buffered
phenylmethylsulfonyl fluoride (0.5 mmol L–1) in dimethylsul- saline containing a mix of protease inhibitors [PMFS (phenyl-
foxide (0.5mLmL–1), and aprotinin (10mLmL–1)] were added
methylsulfonyl fluoride), 0.1mM; pepstatin, 2µM; leupeptin,
to whole serum obtained from fasted controls and from 1µM; aprotinin, 1.5µM) and equal amounts of cellular protein
poorly controlled type 1 and type 2 diabetes subjects. Serum
samples were stored at –70°C until further analysis. HSA
was isolated by fast protein liquid chromatography (FPLC)
using a HiTrapBlue (GE Healthcare, Uppsala, Sweden) affinity
were applied onto a polyacrylamide gel [SDS-PAGE (sodium
dodecyl sulfate polyacrylamide gel electrophoresis]). The
expressions of Grp78, Grp94, PDI, CRT, ubiquitin, and b-actin
were determined using the following antibodies, respectively:
column, following purification by alcoholic extraction. Briefly, anti-KDEL(Stressgen,AnnArbor,MI,USA),anti-PDI(Stressgen,
200µL of the albumin sample were incubated with 10% TCA
(v/v) for 10min at room temperature. After centrifugation the
pellet was washed with absolute ethanol and centrifuged. This
procedure was repeated twice.13 Isolated HSA contained less
than 50pg of endotoxin mL–1 as determined by the Limulus
Amebocyte Lysate (Cape Cod, Falmouth, MA, USA).
Ann Arbor, MI, USA), anti-ubiquitin (ZYMED Laboratories
Carlsbad, CA, USA), anti-CRT (Stressgen, Ann Arbor, MI,
USA), and anti-b-actin (Fitzgerald Industries International
Inc., Concord, MA, USA). Tunicamycin (2 µg mL–1) was used
as a positive control for the induction of ER stress. Proteins
were detected with horseradish peroxidase-conjugated anti-
bodies and an enhanced chemiluminescence reagent (Pierce,
Rockford, IL, USA). Immunoblot band density was quantified
MALDI analysis of control and DM albumin
MALDI-MS measurements were performed using an ultra- using an ImageQuant 300 Imager (GE Healthcare), and differ-
fleXtreme MALDI-TOF/TOF (TOF, time of flight) instrument
(Bruker Daltonics, Bremen, Germany), operating in linear
positive-ion mode. Ions were formed by a Smartbeam-II laser
ences between the bands were analyzed in pixels, using the
ImageQuant TL software (GE Healthcare). The results are
expressed as arbitrary units, analyzing AGE-albumin versus
(l = 355 nm). The matrix was sinapinic acid [saturated solu- C-albumin normalized to b-actin expression.
tion in H2O/acetonitrile/0.1% trifluoroacetic acid (TFA)]. HSA
samples were diluted 1 : 10 with H2O/0.1% TFA; 5 µL of the
diluted samples were mixed with 5µL of the matrix solution J774 macrophages were pre-incubated with 8-bromoadeno-
ABCA-1 protein content
and 1µL of the resulting mixture was deposited on the stain- sine 3¢,5¢-cyclic adenosine monophosphate (8-Br-cAMP)
less steel sample holder, then waiting until dryness before
introducing it into the mass spectrometer.
External mass calibration was obtained by Protein
0.5mM (Sigma Aldrich) and treated for 8h with C- or AGE-HSA.
To assess the ABCA-1 content on the macrophage surface,
a total of 1 × 106 cells were trypsinized, washed extensively
Calibration Standard 1 and 2 (Bruker Daltonics, Bremen, with PBS, and fixed in 4% paraformaldehyde. After fixation,
Germany). The process is based on the average mass values
of insulin [M + H]+ (m/z 5734.52), cytochrome C [M + 2H]2+
(m/z 6185.05), myoglobin [M + 2H]2+ (m/z 8476.66), ubiquitin I
the cells were incubated with anti-ABCA-1 antibody (1 :250
dilution; Novus Biologicals, Inc., Littleton, CO, USA) for 1h at
room temperature, rinsed in PBS and incubated with 4µgmL–1
[M + H]+ (m/z 8565,76), cytochrome C [M + H]+ (m/z 12,360.97), Alexa Fluor 488 antibody (Invitrogen, USA). Cellular fluores-
myoglobin [M + H]+ (m/z 16,972.31), and protein A [M + 2H]2+
(m/z 22,307), trypsinogen [M + H]+ (m/z 23,982), albumin bovine
cence intensity was evaluated by flow cytometry using a FACS
Calibur and Cellquest software (BD Biosciences, San Jose,
protein A [M + 2H]2+ (m/z 33,216), protein A [M + H]+ (m/z 44,613), CA, USA). All of the conditions were corrected to basal cellular
and albumin bovine protein A [M + H]+ (m/z 66,432).
fluorescence (cells without antibody labeling).

236
Glycated HSA from Diabetic Patients Impairs Cholesterol Efflux from Macrophages
Low density lipoprotein (LDL) isolation and
acetylation
1.00
0.75
0.50
0.25
0.00
p = 0.024
LDL (d = 1.019 – 1.063gmL–1) was isolated after the sequential
preparative ultracentrifugation of fresh plasma drawn from
healthy donors and purified by discontinuous gradient ultra-
centrifugation. After dialysis, the protein concentration was
measured by the Lowry method25 and acetylation performed
according to Basu et al.,26 followed by extensive dialysis and
sterilization in a 0.22µm filter.
C
DM
Figure 1. CML content in HSA from C and DM patients. HSA
was isolated from C and DM patients by FPLC and purified by
alcoholic extraction. The CML content in HSA samples was
measured by ELISA (CML-CIRCULEX-8066, CycLex Co, Japan).
Cholesterol efflux assay
J774 macrophages were enriched with acetylated LDL
(50µgmL–1) and 0.3µCimL–1 of 14C-cholesterol (Perkin Elmer)
for 48 h. After two washes with PBS containing FAFA, cells
were maintained for 24h in DMEM containing FAFA and 8-Br- and its relationship with the ABCA-1 protein level and choles-
cAMP (Sigma), followed by 18h of incubation with 1mgmL–1
C- or DM-HSA. Macrophages were then incubated with
30µgmL–1 Apo A-I (Sigma Aldrich) for 8h to determine the
14C-cholesterol efflux, as previously described.27
terol efflux in macrophages. ABCA-1 is responsible for the
major amount of cholesterol exportation in cholesterol-over-
loaded cells, contributing to the net exportation of intracellular
lipids, and ABCA1 gene mutations or damage in its activity lead
to cholesterol accumulation in arterial wall macrophages.10
The clinical characteristics of the subjects under investi-
gation are summarized in Table 1. DM2 patients were older
Statistical analysis
Statistical analyses were performed using the GraphPad
Prism 4.0 software (GraphPad Prism, Inc., San Diego, CA, than both control subjects (C) and DM1 patients. The body
USA). One-way ANOVA with Newman Keuls post-test and
Student’s t test were used as appropriate. Summary data are
reported as the mean values standard deviation. A p-value
<0.05 was considered statistically significant.
mass index (BMI) was higher in DM2 in comparison with
DM1. Fasting glycemia, HbA1c, and glycated HSA levels were
similar between DM1 and DM2, but were higher compared
to C. Fructosamine levels were superior in DM1 compared
to C and DM2. In addition, fructosamine was higher in DM2
compared to C. Patients and control subjects had no anemia
and presented normal levels of serum albumin. CML is an
AGE2 and its amount was greater in HSA isolated from DM
Results and discussion
An alteration observed in the diabetic disease is the modifica- subjects than in HSA from control individuals (Figure 1).
tion of the reverse cholesterol transport responsible for the
First of all, to investigate the possible relationship(s) between
development of atherosclerosis. In this contest, we inves- albumin glycation level and reverse cholesterol efflux, the
tigated the role of the possibly glycated albumin present in
poorly controlled diabetic patients on ER-stress development
mean glycation level of albumin of the three different groups
(controls, DM1 patients, and DM2 patients) under investigation
Table 1. Age, body mass index (BMI), and glycemic parameters.
C
DM1
DM2
(n = 6)
(n = 6)
(n = 6)
Age (years)
28 6.8
27 1.5
26 8.0
22 0.9
61 17a
27 1.7b
226 86c
10.0 1.0d
393 73f
23 6.0d
4.6 0.5
13.1 0.7
BMI (kg m–2)
–1
Fasting glucose (mg dL )
HbA1c (%)
78 6.0
5.3 0.14
238 26
8.8 1.9
5.0 0.4
12.9 0.4
153 90c
10.2 1.36d
478 85e,f
28 6.8d
5.0 0.9
–1
Fructosamine (µmol L )
Glycated HSA (%)
–1
Serum albumin (g dL )
–1
Hemoglobin (g dL )
14.2 1.2
ap < 0.001 vs C and DM1.
bp < 0.05 vs DM1.
cp < 0.05 vs C.
dp < 0.001 vs C.
ep <0.05 vs DM2.
fp < 0.001 vs C.

G. Castilho et al., Eur. J. Mass Spectrom. 21, 233–244(2015)
237
D
Figure 2. HSA MALDI spectra. (a) Spectrum of HSA from a control subject. (b) Spectrum of HSA from a DM1 patient. (c) Spectrum of
HSA from a DM2 patient. The mass differences (1297Da, C vs DM1; 890Da, C vs DM2) indicate that in the case of the DM1 diabetic
patient at least eight glucose molecules are condensed on the protein, while in the case of the DM2 patient this value is about five. (d)
Theoretical isotopic pattern of HSA (mMass Software).
was evaluated by MALDI mass spectrometry. As reported in
determine the number of hexose molecules (nh) condensed on
the Introduction, this approach was already proved to be effec- the protein:
tive for the evaluation of the number of glucose molecules
condensed on this protein, following the Maillard reaction
pathway.28,29 In fact, considering that the condensation of a
HSAG -HSAUG
= nh
162
hexose molecule on the protein leads to a mass increase of
162Da, the difference in molecular weight of glycated albumin
(HSAG) and unglycated albumin (HSAUG) from DM patients
and healthy subjects, respectively, divided by 162 allows us to
It is emphasized that this represents the minimum
value of glucose molecules condensed on albumin. In fact,
according to the Maillard reaction pathway, both dehydration
and oxidation processes can occur on the condensed sugar

238
Glycated HSA from Diabetic Patients Impairs Cholesterol Efflux from Macrophages
moiety, leading to a decrease of the molecular weight of the
condensed glucose moiety on the protein of interest. As an
unequivocally prove that in control subjects a HSA glycation
also takes place, because of the circulating glucose; (3) finally,
example, the MALDI spectra obtained for a healthy subject, a physical phenomenon participates in the peak-enlargement
DM1 patients and DM2 patients are reported in Figure 2(a), process caused by coulombic repulsion of the ions belonging
(b), and (c), respectively. The wide and unsymmetrical peak
shape is worthy of discussion. If we compare the shape of
the theoretical isotopic cluster distribution of HSA [Figure
to the isotopic cluster along the flight pathway. However, the
clear differences observed in HSA-protonated species among
the controls and DM1 and DM2 patients must necessarily be
2(d)] with that of the control subject reported in Figure 2(a), related to the higher plasma glucose levels experienced by
at first sight it is evident that a mean increase of about 1kDa
is present for the latter and the quasi-Gaussian shape of the
theoretical isotopic cluster is not maintained. This behavior
can result from many different causes, e.g. (1) considering
that HSA is a transport protein, in an in vivo condition it can
interact, via covalent and non-covalent bonding, with different
the last two groups, as well confirmed by the metabolic data
reported in Table 1. Higher glucose levels reflect a higher yield
of the Maillard reaction and the consequent larger number
of glucose molecules condensed on HSA. To evaluate this
phenomenon the experimental m/z value of the maximum
of the HSA peak for controls can be validly consider as the
molecules present in the blood; (2) as shown by the clinical “zero point”, to be compared with the maxima of HSA peaks
data reported in Table 1, glycation processes also occur in
the case of healthy subjects, of course with a lower yield than
that present in DM1 and DM2 patients, because of the higher
glucose concentration. In this context, the data reported
in Table 1 (in particular HbA1c and fructosamine values)
of DM1 and DM2 patients. To evaluate the mean glycation
level, the m/z value of the peak maximum was considered.
In the case of the DM1 subjects a mean mass increase of
about 1297Da was observed, while for DM2 patients this value
was 890 Da, because of the condensation on the protein of
Figure 3. MALDI spectra relative to HSA trypsin digestion products. (a) Example spectrum of the HSA digestion products from a control
subject. (b) Example spectrum of the HSA digestion products from a DMI1 patient. (c) Example spectrum of the HSA digestion products
from a DM2 patient. The labeled peaks exhibit absolute intensities higher than 100 arbitrary units (a.u.).

G. Castilho et al., Eur. J. Mass Spectrom. 21, 233–244(2015)
239
at least eight glucose units for DM1 and five glucose units
for DM2. The mean value of the molecular weight of albumin
samples detected in the case of healthy subjects (n = 6) is
67,846 301Da, while in the case of DM1 (n = 6) patients the
mean value rises up to 68,544 192Da and for DM2 patients
(n = 6) the mean value is 68,547 132Da.
In order to investigate the most active sites present
in albumin and prone to glycation, albumin samples were
digested by trypsin, i.e. the enzyme usually employed in
patients are reported in Table 2. Practically, no differences are
observed between the two DM groups, while the control and
DM subjects can be easily characterized, because of the pres-
ence of specific digestion products. The exclusive peptides
obtained for the tryptic digestion of HSA from diabetic patients
might result from glycated/oxidized species, e.g. m/z 1001, as
is hypothesized in Table 2 and shown in Figure 4, or to prod-
ucts that originated from an alteration of the trypsin cleavage
sites caused by the presence of a glycated or modified residue
protein studies (an example of the results obtained for control, in the HSA sequence.
DM1 and DM2 subjects is reported in Figure 3). The sequence
coverage obtained from the analysis of the tryptic digestion
The obtained data were compared with those already avail-
able in the literature,30,31 allowing the identification of the HSA
products by MALDI-MS is between 30% and 33%. It is inter- fragments originating from the enzymatic digestion. These
esting to observe that in the case of DM patients the number
of digestion products is lower than that observed in the case
of control subjects. This is in agreement with what has already
observed, i.e. that the glycated proteins are less prone to
enzymatic digestion.30 This behavior can be rationalized by
considering that in the case of HSA from diabetic subjects, the
glycation of the e amino groups of the lysine residues does not
allow the proteolytic action of the trypsin in the corresponding
cleavage sites and consequently the relative peptides are not
produced.
data are reported in Table 3. However, it is emphasized that
the literature data were obtained under electrospray ioniza-
tion (ESI) conditions and consequently some differences are to
be expected because of the different ionization mechanisms
of the ESI and MALDI methods. Furthermore, by comparing
the molecular weight of the peptides present in the digestion
mixture of the control subjects with those of diabetic patients,
possible modifications caused by glycation processes have
been investigated, and the results summarized in Table 4. The
presence of glycated peptides that have undergone dehydra-
The most relevant differences observed among the diges- tion processes indicates that the species described above
tion products of HSA from the controls and DM1 and DM2
can be considered AGEs.28,29 This view is in agreement with
Table 2. Summary of the most important differences between digested HSA from controls and digested HSA from diabetic patients.
m/z
Absolute abundances of the exclusive
HSA digestion products from C
(mean standard deviation)
Absolute abundances of the exclusive
HSA digestion products from DM1 and DM2
(mean standard deviation)
599
–
0.147 0.075
627
–
0.18 0.014
734
0.091 0.036
–
897
0.143 0.016
–
930
–
0.128 0.041
937
0.087 0.042
–
1001
1012
1035
1155
1486
1891
2411
2440
2470
2514
2558
3058
3086
3265
3299
–
0.296 0.267
–
1.790 0.814
–
0.206 0.029
–
4.322 1.700
–
5.078 1.475
–
0.113 0.052
0.046 0.009
0.062 0.003
0.039 0.007
–
–
–
–
0.049 0.027
0.101 0.014
0.027 0.006
0.031 0.004
0.011 0.002
0.021 0.007
–
–
–
–
–

240
Glycated HSA from Diabetic Patients Impairs Cholesterol Efflux from Macrophages
Figure 4. Evidence of glycation: the peptide at m/z 875.5 is modified by glycation and dehydration processes, leading to the formation
of the peptide at m/z 1001.5. (a) Enlargement of a spectrum for digested HSA from a control subject. (b) Enlargement of a spectrum for
digested HSA from a diabetic patient.
the CML data (Figure 1). Unfortunately, attempts to perform
tandem mass spectroscopy experiments in order to confirm
the structure of these glycated peptides failed because their
abundance is too low.
It is interesting to note that, different to that observed with
in vitro HSA glycation–digestion experiments,30,31 in the present
case new glycated species are observed, reasonably as a
result of the different glycation mechanisms and the different
glycated HSA) and that purified HSA from these patients also
displayed a very similar profile in the MALDI analysis, we ran
combined analyses that comprised DM1- and DM2-HSA.
PDI expression was 89% enhanced in DM-HSA-treated
macrophages, in comparison to cells treated in the presence
of C-HSA [Figure 5(a)]. Although the immunoblot seems over-
loaded, the differences between the bands are very clear. In
addition, changes observed in the protein content of the PDI
tertiary HSA structure in an in vivo environment. Furthermore, chaperone are in agreement with that observed in another
besides the species detected at m/z 2200, ascribed to the
simple condensation of two glucose molecules, all the other
species described in Table 4 have undergone dehydration
processes, typical of the intermediate phase of the Maillard
reaction pathway (Amadori products).
However, HSA-MALDI profiles were similar between
DM1 and DM2 patients in accordance with the HbA1c levels
(10.2 1.36 for DM1 and 10.0 1.0 for DM2).
KDEL sequenced protein, CRT and with ubiquitin. In addi-
tion, the authors have previously showed that in vitro glycated
albumin—a useful tool to access the role of advanced glycation
in lipid cell metabolism—induces the expression of ER stress
markers related to the disturbance in ABCA-1-mediated
cholesterol efflux.8
Similarly, another ER stress marker, CRT, had its expres-
sion dramatically increased in macrophages incubated with
Then, we performed functional assays in order to charac- DM-HSA {3.4 times higher comparing to the C-HSA-treated
terize the role of HSA modification by glycation in macrophage
cholesterol homeostasis, which may relate to atherogenesis in
DM. Considering that both DM1 and DM2 patients presented
similar glycemic parameters (fasting glycemia, HbA1c, and
cells [Figure 5(b)]}. Ubiquitin expression was 71% higher in
DM-albumin-treated macrophages [Figure 5(c)]. We previ-
ously demonstrated that AGE-induced ER stress is related to
the impairment in macrophage reverse cholesterol transport.

G. Castilho et al., Eur. J. Mass Spectrom. 21, 233–244(2015)
241
Table 3. Comparison between the information obtained from MALDI-MS spectra from digested HSA with those reported in Lapolla et al.28
and Marotta et al.31
m/z
Position and modification
213–218
Sequence
673.3
AWAVAR
875.4
219–225
LSQRFPK
927.5
138–144
YLYEIAR
960.5
403–410
FQNALLVR
1019.7
1074.5
1128.7
1149.6
1226.6
1296.4
1311.7
1467.8
1623.8
1639.8
1651.7
1677.8
1742.8
1898.9
2721.2
210–218
AFKAWAVAR
LDELRDEGK
KQTALVELVK
LVNEVTEFAK
FKDLGEENFK
LAKTYETTLEK
HPDYSVVLLLR
RHPDYSVVLLLR
DVFLGMFLYEYAR
DVFLGMFLYEYAR
182–190
525–534
42–51
11–20
349–359
338–348
337–348
324–336
324–336 (1Met-ox)
226–240
AEFAEVSKLVTDLTK
429–432 + 565–574 + 48u
146–159
NLGK + ETCFAEEGKK + 48u
HPYFYAPELLFFAK
145–159
RHPYFYAPELLFFAK
115–137
LVRPEVDVMCTAFHDNEETFLKK
Table 4. Probable mass modifications caused by glycation processes of the peptides present in the HSA trypsin digestion product.
Unglycated peptide
Probable glycated peptide
Weight increment
Glycation process
+ C6H12O6 – 3H2O
+ C6H12O6 – 3H2O
+ C6H12O6 – 2H2O
+ C6H12O6 – H2O
+ 2C6H12O6 – 2H2O
+ C6H12O6 – 2H2O
875
1001
1440
1458
1891
2524
2614
126
126
144
162
324
144
1314
1314
1729
2200
2470
In fact, Figure 6(a) shows that DM-HSA decreased by 50%
the ABCA-1 expression in comparison to C-HSA, which was
and Grp94, present a trend of increase in macrophages treated
with DM-HSA (data not shown). Then, considering that in DM
reflected in a 60% reduction in the apo A-I-mediated choles- albumin can be modified in the arterial wall site by highly
terol efflux [Figure 6(b)].
reactive oxoaldehydes derived from hyperglycemia, oxidative
stress, and inflammation our data strengthens the contribu-
tion of AGE formed in vivo to the development of atheroscle-
rosis. Besides, AGE-HSA can reach the arterial wall compart-
ment during endothelial damage and atherosclerosis.
The reduction in the ABCA-1 protein level elicited by glyca-
tion does not seem to involve an alteration in the ABCA1 gene
transcription since no changes were observed in ABCA-1
mRNA levels upon incubation of macrophages directly with
oxoaldehydes or AGE-albumin.27,32 It is also interesting that
the effect of advanced glycation in reducing ABCA-1 cannot be
As recently published by us, in vitro produced AGE-albumin
induces the generation of ROS by the mitochondria and
the NADPH (nicotinamide adenine dinucleotide phosphate)
oxidase system.5 In addition, in vitro glycated albumin evokes
ER stress, which seems to be the mechanism linking protein
glycation, oxidative stress, and the loss of ABCA-1 protein
content, since the use of a chemical chaperone (phenyl butyric
acid) that alleviates ER stress restores both the ABCA-1
content and the cholesterol efflux.8
In this study we found that HSA that has been modified in
DM by advanced glycation indeed induces ER dysfunction— overwhelmed by incubation with cyclic AMP32 or LXR agonist
indicated by the elevated expression of PDI and CRT—that
ultimately disturbs macrophage cholesterol exportation via
ABCA-1. Other classical markers of ER stress, such as Grp78
T0901317.6
ABCA-1 is mainly regulated by posttranslational mecha-
nisms that dictate its stability in the plasma membrane and

242
Glycated HSA from Diabetic Patients Impairs Cholesterol Efflux from Macrophages
Figure 5. Expression of ER resident chaperones and ubiquitin
in macrophages treated with C- and DM-HSA. Mouse perito-
neal macrophages were treated for 18h with C- or DM-HSA
(1mgmL–1). Equal amounts of cellular protein were loaded
onto the polyacrylamide gels, and electrophoresis and
immunoblotting were performed to measure the expression of
(a) PDI (n = 7), (b) CRT (n = 12), and (c) ubiquitin (n = 12). Data
are expressed as arbitrary units corrected to b-actin levels
(representative figure from two independent experiments).
Figure 6. Expression of ABCA-1 and cholesterol efflux in
macrophages treated with C- and DM-HSA. Mouse peritoneal
macrophages were treated for 18h with C- or DM-HSA
(1mgmL–1). (a) ABCA-1 content was determined by flow
cytometry in J774 macrophages pre-incubated with 0.5mM of
AMPc (n = 6). (b) Acetylated LDL- and 14C-cholesterol-enriched
mouse peritoneal macrophages were treated for 8h with C- or
DM-HSA (1mgmL–1), and cholesterol efflux to Apo A-I
(30µgmL–1) was assessed for 8h (n = 5–10).
late endosomes. We also found in the present study a higher us to define a detail characterization of the species of interest,
expression of total ubiquitin in macrophages treated with difficult or impossible to obtain with other analytical methods.
DM-HSA, which suggests impairment in proteasomal activity.
Proteasomal inhibition has been shown to be caused by glyca-
tion, including changes of constitutional proteasomal into an
References
immunoproteasomal complex profile in macrophages.33,34 The
mechanisms that link ER stress with the reduction in ABCA-1 1. E.S. Ford, “Trends in the risk for coronary heart disease
protein level were not addressed in the present investigation
and are currently under investigation in our laboratory. They
may involve the UPS as a part of the ER-associated degrada-
tion and lysosomal authopagy, another mechanism linked
among adults with diagnosed diabetes in the US: find-
ings from the national health and nutrition examination
survey, 1999–2008”, Diabetes Care 34(6), 1337 (2011). doi:
http://dx.doi.org/10.2337/dc10-2251
to ER stress. In addition, ABCA-1 can be also regulated by 2. F. Giacco and M. Brownlee, “Oxidative stress and diabetic
surface calpains.
complications”, Circ. Res. 107(9), 1058 (2010). doi: http://
Our data indicate that the level of glycation that is attained
dx.doi.org/10.1161/CIRCRESAHA.110.223545
in HSA in both type 1 and type 2 poorly controlled DM 3. F. Piarulli, G. Sartore and A. Lapolla, “Glyco-oxidation
patients damages the ER function related to disturbances in
macrophage reverse cholesterol. In this sense an adequate
management of diabetes and glycemic level will provide a
and cardiovascular complications in type 2 diabetes:
a clinical update”, Acta Diabetol. 50(2), 101 (2013). doi:
http://dx.doi.org/10.1007/s00592-012-0412-3
reduced HSA glycation, which will reflect in cardiovascular 4. R. Ramasamy, S.F. Yan and A.M. Schmidt, “The diverse
benefit by preventing derangements in the reverse cholesterol
transport.
Finally, it is emphasized that the mass spectrometric data
either on intact proteins or on HSA digestion products allowed
ligand repertoire of the receptor for advanced glycation
endproducts and pathways to the complications of diabe-
tes”, Vascul. Pharmacol. 57(5)–(6), 160 (2012). doi: http://
dx.doi.org/10.1016/j.vph.2012.06.004

G. Castilho et al., Eur. J. Mass Spectrom. 21, 233–244(2015)
243
5. R. de Souza Pinto, G. Castilho, B.A. Paim, A. Machado-
Lima, N.M. Inada, E.R. Nakandakare, A.E. Vercesi and
M. Passarelli, “Inhibition of macrophage oxidative stress
prevents the reduction of ABCA-1 transporter induced
by advanced glycated albumin”, Lipids 47(5), 443 (2012).
doi: http://dx.doi.org/10.1007/s11745-011-3647-9
6. R.T. Iborra, A. Machado-Lima, G. Castilho, V.S. Nunes,
D.S. Abdalla, E.R. Nakandakare and M. Passarelli,
“Advanced glycation in macrophages induces intracel-
lular accumulation of 7-ketocholesterol and total sterols
by decreasing the expression of ABCA-1 and ABCG-1”,
Lipids Health Dis. 10(1), 172 (2011). doi: http://dx.doi.
org/10.1186/1476-511X-10-172
7. L.S. Okuda, G. Castilho, D.D. Rocco, E.R. Nakandakare,
S. Catanozi and M. Passarelli, “Advanced glycated
albumin impairs HDL anti-inflammatory activity and
primes macrophages for inflammatory response
that reduces reverse cholesterol transport”, Biochim.
Biophys. Acta 1821(12), 1485 (2012). doi: http://dx.doi.
org/10.1016/j.bbalip.2012.08.011
8. G. Castilho, L.S. Okuda, R.S. Pinto, R.T. Iborra, E.R.
Nakandakare, C.X. Santos, F.R. Laurindo and M.
Passarelli, “ER stress is associated with reduced
ABCA-1 protein levels in macrophages treated with
advanced glycated albumin—reversal by a chemical
chaperone”, Int. J. Biochem. Cell Biol. 44(7), 1078 (2012).
doi: http://dx.doi.org/10.1016/j.biocel.2012.03.016
9. A. Machado-Lima, R.T. Iborra, R.S. Pinto, C.H.
Sartori, E.R. Oliveira, E.R. Nakandakare, J.T. Stefano,
D. Giannella-Neto, M.L. Correa-Giannella and M.
Passarelli, “Advanced glycated albumin isolated from
poorly controlled type 1 diabetes mellitus patients
alters macrophage gene expression impairing ABCA-1-
mediated reverse cholesterol transport”, Diabetes Metab.
Res. Rev. 29(1), 66 (2013). doi: http://dx.doi.org/10.1002/
dmrr.2362
10. C. Tang and J.F. Oram, “The cell cholesterol exporter
ABCA1 as a protector from cardiovascular disease and
diabetes”, Biochim. Biophys. Acta 1791(7), 563 (2009). doi:
http://dx.doi.org/10.1016/j.bbalip.2009.03.011
11. G.S. Hotamisligil, “Endoplasmic reticulum stress
and the inflammatory basis of metabolic disease”,
Cell 140(6), 900 (2010). doi: http://dx.doi.org/10.1016/j.
cell.2010.02.034
12. B. Feng, P.M. Yao, Y. Li, C.M. Devlin, D. Zhang, H.P.
Harding, M. Sweeney, J.X. Rong, G. Kuriakose, E.A.
Fisher, A.R. Marks, D. Ron and I. Tabas, “The endoplas-
mic reticulum is the site of cholesterol-induced cytotox-
icity in macrophages”, Nat. Cell Biol. 5(9), 781 (2003). doi:
http://dx.doi.org/10.1038/ncb1035
13. M. Myoishi, H. Hao, T. Minamino, K. Watanabe, K.
Nishihira, K. Hatakeyama, Y. Asada, K. Okada, H.
Ishibashi-Ueda, G. Gabbiani, M.L. Bochaton-Piallat, N.
Mochizuki and M. Kitakaze, “Increased endoplasmic
reticulum stress in atherosclerotic plaques associ-
ated with acute coronary syndrome”, Circulation
116(11), 1226 (2007). doi: http://dx.doi.org/10.1161/
CIRCULATIONAHA.106.682054
14. M. Civelek, E. Manduchi, R.J. Riley, C.J. Stoeckert and
P.F.Davies, “Chronic endoplasmic reticulum stress acti-
vates unfolded protein response in arterial endothelium
in regions of susceptibility to atherosclerosis”, Circ.
Res. 105(5), 453 (2009). doi: http://dx.doi.org/10.1161/
CIRCRESAHA.109.203711
15. A. Lapolla, C. Gerhardinger, L. Baldo, D. Fedele, A.
Keane, R. Seraglia, S. Catinella and P. Traldi, “A study
on in vitro glycation processes by matrix-assisted laser
desorption ionization mass spectrometry”, Biochim.
Biophys. Acta 1225(1), 33 (1993). doi: http://dx.doi.
org/10.1016/0925-4439(93)90118-K
16. A. Lapolla, L. Baldo, R. Aronica, C. Gerhardinger, D.
Fedele, G. Elli, R. Seraglia, S. Catinella and P. Traldi,
“Matrix-assisted laser desorption/ionization mass spec-
trometric studies on protein glycation. 2. The reaction
of ribonuclease with hexoses”, Biol. Mass Spectrom.
23(5), 241 (1994). doi: http://dx.doi.org/10.1002/
bms.1200230502
17. A. Lapolla, D. Fedele and P. Traldi, “Studies on advanced
glycation end products by recent mass spectrometric
techniques”, Amino Acids 6(1), 65 (1994). doi: http://dx.doi.
org/10.1007/BF00808124
18. P. Traldi, A. Lapolla, R. Seraglia, S. Catinella, M.
D’Alpaos, R. Aronica and D. Fedele, “Investigations on
protein nonenzymatic glycation by a new and effective
mass spectrometric method”, Microchem. J. 54(3), 218
(1996). doi: http://dx.doi.org/10.1006/mchj.1996.0098
19. A. Lapolla, D. Fedele, L. Baldo, R. Aronica, V. Raverdino,
S. Catinella, R. Seraglia and P. Traldi, “An electrospray
investigation on in vitro glycation of ribonuclease”, Rapid
Commun. Mass Spectrom. 10(2), 178 (1996). doi: http://
dx.doi.org/10.1002/(SICI)1097-
0231(19960131)10:2<178::AID-RCM454>3.0.CO;2-O
20. A. Lapolla, D. Fedele, R. Seraglia, S. Catinella, L. Baldo,
R. Aronica and P. Traldi, “A new effective method for the
evaluation of glycated intact plasma proteins in diabetic
subjects”, Diabetologia 38(9), 1076 (1995). doi: http://
dx.doi.org/10.1007/BF00402178
21. A. Lapolla, D. Fedele, R. Aronica, M. Garbeglio,
M. D’Alpaos, R. Seraglia and P. Traldi, “The in vivo
glyco-oxidation of a- and b-globins investigated by
matrix-assisted laser desorption/ionization mass
spectrometry”, Rapid Commun. Mass Spectrom. 10(9),
1133 (1996). doi: http://dx.doi.org/10.1002/(SICI)1097-
0231(19960715)10:9<1133::AID-RCM581>3.0.CO;2-C
22. A. Lapolla, D. Fedele, M. Plebani, M. Garbeglio, R.
Seraglia, M. D’Alpaos, C.N. Arico and P. Traldi, “Direct
evaluation of glycated and glyco-oxidized globins by
matrix-assisted laser desorption/ionization mass
spectrometry”, Rapid Commun. Mass Spectrom. 13(1),
8 (1999). doi: http://dx.doi.org/10.1002/(SICI)1097-
0231(19990115)13:1<8::AID-RCM438>3.0.CO;2-H.

244
Glycated HSA from Diabetic Patients Impairs Cholesterol Efflux from Macrophages
23. A. Lapolla, D. Fedele, R. Aronica, M. Garbeglio, M.
D’Alpaos, M. Plebani, R. Seraglia and P. Traldi, “A
highly specific method for the characterization of
glycation and glyco-oxidation products of globins”,
Rapid Commun. Mass Spectrom. 11(6), 613 (1997). doi:
http://dx.doi.org/10.1002/(SICI)1097-
0231(199704)11:6<613::AID-RCM907>3.0.CO;2-2
24. A. Lapolla, D. Fedele, R. Aronica, M. Garbeglio, M.
D’Alpaos, R. Seraglia and P. Traldi, “Evaluation of
IgG glycation levels by matrix-assisted laser desorp-
tion/ionization mass spectrometry”, Rapid Commun.
Mass Spectrom. 11(12), 1342 (1997). doi: http://dx.doi.
org/10.1002/(SICI)1097-0231(199708)11:12<1342::AID-
RCM972>3.0.CO;2-T
29. A. Lapolla, D. Fedele, R. Seraglia and P. Traldi, “The role
of mass spectrometry in the study of non-enzymatic
protein glycation in diabetes: an update”, Mass Spectrom.
Rev. 25(5), 775 (2006). doi: http://dx.doi.org/10.1002/
mas.20090
30. A. Lapolla, D. Fedele, R. Reitano, N.C. Arico, R. Seraglia,
P. Traldi, E. Marotta and R. Tonani, “Enzymatic digestion
and mass spectrometry in the study of advanced glyca-
tion end products/peptides”, J. Am. Soc. Mass Spectrom.
15(4), 496 (2004). doi: http://dx.doi.org/10.1016/j.
jasms.2003.11.014
31. E. Marotta, A. Lapolla, D. Fedele, A. Senesi, R. Reitano,
M. Witt, R. Seraglia and P. Traldi, “Accurate mass meas-
urements by Fourier transform mass spectrometry in
the study of advanced glycation end products/peptides”,
J. Mass Spectrom. 38(2), 196 (2003). doi: http://dx.doi.
org/10.1002/jms.431
25. O.H. Lowry, N.J. Rosebrough, A.L. Farr and R.J. Randall,
“Protein measurement with the folin phenol reagent”,
J. Biol. Chem. 193(1), 265 (1951).
26. S.K. Basu, J.L. Goldstein, G.W. Anderson and M.S.
Brown, “Degradation of cationized low density lipo-
protein and regulation of cholesterol metabolism in
homozygous familial hypercholesterolemia fibroblasts”,
Proc. Natl Acad. Sci. USA 73(9), 3178 (1976). doi: http://
dx.doi.org/10.1073/pnas.73.9.3178
27. A.P. Machado, R.S. Pinto, Z.P. Moyses, E.R. Nakandakare,
E.C. Quintao and M. Passarelli, “Aminoguanidine
and metformin prevent the reduced rate of HDL-
mediated cell cholesterol efflux induced by formation of
advanced glycation end products”, Int. J. Biochem. Cell
Biol. 38(3), 392 (2006). doi: http://dx.doi.org/10.1016/j.
biocel.2005.09.016
32. M. Passarelli, C. Tang, T.O. McDonald, K.D. O’Brien, R.G.
Gerrity, J.W. Heinecke and J.F. Oram, “Advanced glyca-
tion end product precursors impair ABCA1-dependent
cholesterol removal from cells”, Diabetes 54(7), 2198
(2005). doi: http://dx.doi.org/10.2337/diabetes.54.7.2198
33. F. Moheimani, P.E. Morgan, D.M. van Reyk and M.J.
Davies, “Deleterious effects of reactive aldehydes and
glycated proteins on macrophage proteasomal function:
possible links between diabetes and atherosclerosis”,
Biochim. Biophys. Acta 1802(6), 561 (2010). doi: http://
dx.doi.org/10.1016/j.bbadis.2010.02.007
34. S. Grimm, C. Ott, M. Horlacher, D. Weber, A. Hohn and
T. Grune, “Advanced-glycation-end-product-induced
formation of immunoproteasomes: involvement of RAGE
and Jak2/STAT1”, Biochem. J. 448(1), 127 (2012). doi:
http://dx.doi.org/10.1042/BJ20120298
28. A. Lapolla, D. Fedele and P. Traldi, “The role of
mass spectrometry in the study of non-enzymatic
protein glycation in diabetes”, Mass Spectrom. Rev.
19(5), 279 (2000). doi: http://dx.doi.org/10.1002/1098-
2787(2000)19:5<279::AID-MAS3>3.0.CO;2-G