Denis Blache,1 Emmanuel Bourdon,1 Pauline Salloignon,2 Géraldine Lucchi,2 Patrick Ducoroy,2

Jean-Michel Petit,1,3 Bruno Verges,1,3 and Laurent Lagrost1

Glycated Albumin With Loss ofFatty Acid Binding CapacityContributes to EnhancedArachidonate Oxygenationand Platelet Hyperactivity:Relevance in Patients WithType 2 DiabetesDiabetes 2015;64:960–972 | DOI: 10.2337/db14-0879

High plasma concentrations of nonesterified fatty acids(NEFAs), transported bound to serum albumin, are asso-ciated with type 2 diabetes (T2D). The effects of albuminon platelet function were investigated in vitro. Modifica-tions of albumin, such as those due to glycoxidation,were found in patients with T2D, and the consequencesof these modifications on biological mechanisms relatedto NEFA handling were investigated. Mass spectrometryprofiles of albumin from patients with T2D differed fromthose from healthy control subjects. Diabetic albuminshowed impaired NEFA binding capacity, and bothstructural and functional alterations could be reproducedin vitro by incubating native albumin with glucose andmethylglyoxal. Platelets incubated with albumin isolatedfrom patients with T2D aggregated approximately twiceas much as platelets incubated with albumin isolatedfrom healthy control subjects. Accordingly, plateletsincubated with modified albumin produced significantlyhigher amounts of arachidonate metabolites than didplatelets incubated with control albumin. We concludedthat higher amounts of free arachidonate are madeavailable for the generation of active metabolites in

platelets when the NEFA binding capacity of albumin isblunted by glycoxidation. This newly described mecha-nism, in addition to hypoalbuminemia, may contribute toplatelet hyperactivity and increased thrombosis, knownto occur in patients with T2D.

Many epidemiological studies have established an inverserelationship between the level of serum albumin and therisk of death (1,2). After adjustment for usual risk factors,this association was also found to hold not only for liverdysfunction (3) but also for cardiovascular disease andstroke (4–6). Albumin is involved in several biologicalmechanisms, including the regulation of oncotic pressure,and the binding and transport of a number of hydropho-bic or amphipathic molecules (mainly nonesterified fattyacids [NEFAs]). In addition, albumin displays potent anti-oxidant and free radical scavenging activities throughthe redox cycling of its free thiol (Cys34) (7) and its abilityto bind metal ions (2,8,9). High levels of NEFAs are as-sociated with type 2 diabetes (T2D) and metabolic syn-drome (10,11), and they are known to mediate ventricular

1Centre de Recherche INSERM UMR 866, Lipides, Nutrition, Cancer, Faculté deMédecine, Université de Bourgogne, Dijon, France2Clinical and Innovation Plateforme de Protéomique, Centre Hospitalier Universi-taire, Dijon, France3Service Endocrinologie, Diabétologie et Maladies Métaboliques, Centre HospitalierUniversitaire, Dijon, France

Corresponding author: Denis Blache, dblache@u-bourgogne.fr.

Received 5 June 2014 and accepted 16 August 2014.

This article contains Supplementary Data online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db14-0879/-/DC1.

E.B. is currently affiliated with Groupe d’Etude sur l’Inflammation Chronique etl’Obésité, Plateforme Cyclotron Réunion Océan Indien (CYROI) Université de LaRéunion, St. Denis, France.

© 2015 by the American Diabetes Association. Readers may use this article aslong as the work is properly cited, the use is educational and not for profit, andthe work is not altered.

See accompanying article, p. 703.

960 Diabetes Volume 64, March 2015

PATHOPHYSIO

LOGY

arrhythmia and sudden death (10,12). Hypoalbuminemiaresulting from either very rare familial deficiency or in-creased clearance in kidney disease has been associatedboth with abnormally elevated NEFA concentrations in se-rum and with high cardiovascular risk (13), suggesting aputative role of serum albumin in NEFA metabolism. Lowplasma concentrations of albumin are known to have adverseeffects, and high levels of glucose and free radicals may im-pair the biological properties of serum albumin through theformation of harmful adducts (14). The nonenzymatic co-valent attachment of glucose molecules to protein and thesubsequent free radical–mediated oxidation give rise to ad-vanced glycoxidation end (AGE) products, of which methyl-glyoxal (MGO) is considered a key component through theAmadori rearrangement (15). Several reports indicate thatAGE products are able at least in vitro to induce activationof various cells including platelets (16,17). In addition toclassical glycemic indices, several recent reports indicatedthat glycated albumin is a relevant and practical biomarkerfor the progression of diabetic atherosclerosis (18–20).

Although most circulating NEFAs are produced by theadipose tissue during lipolysis, phospholipases also con-tribute to NEFA production during the platelet activationprocess. Blood platelets are not only essential for primaryhemostasis but also play a central role in the pathophys-iology of atherothrombosis. Upon adhesion and activationby various stimuli (such as thrombin, ADP, and prostanoids),platelets undergo shape changes, express receptors, aggre-gate, and release their cytosolic granule content (21). Clin-ical and epidemiological studies provide strong evidencefor an association between platelet hyperactivity and ath-erothrombosis (22). Numerous investigations have reportedthat risk factors for cardiovascular diseases such as to-bacco smoking, hyperlipidemia, hypertension, and diabe-tes are associated with enhanced adhesion, as well asspontaneous and inducible platelet activation (23,24).

Arachidonate can bind to albumin, which can thenmodulate its conversion into active oxidized metabolites.However, current knowledge concerning the influence ofmodified versus native albumin (N-Alb) on blood plate-lets, in particular on their ability to release arachidonatemetabolites, is scarce. In the first part of the currentstudy, we investigated abnormalities in the structure andNEFA-binding property of albumin in plasma of T2Dpatients compared with healthy control subjects. In thesecond part, we set up an experimental protocol toreproduce in vitro the harmful derivatives of albuminthat actually occur in the plasma of T2D patients. Finally,we demonstrated that unmodified albumin can exerta negative regulatory effect on platelet adhesion andaggregation. When glycoxidation of albumin occurs, itsfatty acid binding capacity diminishes, and fewer arach-idonate molecules are sequestered. Thus, more arachido-nate is available for conversion into eicosanoids, which arepotent platelet activators. This newly described mecha-nism might contribute to the elevated thromboembolicrisk in patients with T2D.

RESEARCH DESIGN AND METHODS

SubjectsControl subjects were healthy volunteers (n = 20) with nohistory of hyperlipidemia, coronary artery disease, or di-abetes. Patients (n = 20) had T2D. The protocol was ap-proved by the local medical ethics committee, and allparticipants gave written informed consent. The clinicaland biological parameters of the subjects are described inTable 1. For the patients with T2D, 7 were received in-sulin treatment, 11 with sulfamide, and 11 with metfor-min. Five patients received a treatment with a fibrate, and14 were treated with a statin for hyperlipidemia. Onepatient was administrated an antiplatelet agent. As shownby creatinine clearance, there was not any patient in a se-vere chronic renal insufficiency (clearance ,30 mL/min),and three had a moderate chronic renal insufficiency(clearance between 30 and 60 mL/min).

Blood SamplingFasting blood samples were collected and specificallyhandled for each test. Plasma was obtained from citrate-containing glass tubes and separated by centrifugation(15 min, 4°C, 3,000g). Routine biochemistry analyses(plasma lipids, lipoproteins, and apolipoproteins) wereperformed according to standard protocols (25,26). Plasmaalbumin concentrations were measured by a standard pro-cedure using bromocresol green with human albumin asthe standard (27,28).

Albumin Purification and AnalysesBlue Sepharose (Pharmacia Fine Chemicals) was used foralbumin isolation by gel purification according to a pre-viously published procedure (29). Briefly, the plasma(1.350 mL) was mixed with the gel (2.5 mL) equilibratedwith starting buffer (50 mmol/L Tris/HCl [pH 7.4] and0.5 mol/L NaCl). After centrifugation (4,000g, 5 min,4°C), the supernatant was discarded. The gel was washedthree times with starting buffer under the same condi-tions. Gel-bound albumin was eluted (0.650 mL) with50 mmol/L Tris/HCl (pH 7.4) containing 1.5 mol/L NaCl.

Proteomics and Mass Spectrometry Analysis

Two-Dimensional AnalysesNative BSA and human serum albumin (HSA), MGO- andglucose-treated albumin, and isolated albumin fractionsfrom control subjects or T2D patients and protein pIstandards (SERVA) were diluted in 125 mL hydration buffer(8 mol/L urea, 4% 3-[(3-Cholamidopropyl)dimethylammonio]-1-propanesulfonate [CHAPS], 20 mmol/L dithiothreitol,0.2% Bio-Lyte 3/10). After overnight hydration of 7-cm-long ReadyStrip, pH 4–7 (Bio-Rad), at 50 V in a Protean IEFcell (Bio-Rad), isoelectrofocalization was conducted for 12kV-h. Either the strips were stained with Coomassie BrilliantBlue G-250 or proteins were subjected to second-dimensionSDS-PAGE (NuPAGE 4–12% bis-Tris gels; Invitrogen). Blotswere analyzed with Image Laboratory software on a Chemi-Doc device (Bio-Rad). Coomassie-stained spots of interestwere manually excised for samples from six control subjects

diabetes.diabetesjournals.org Blache and Associates 961

and six T2D patients. The gel pieces were then washed twicewith 0.1 mol/L NH4HCO3 and 100% acetonitrile (ACN) for10 min. Reduction/alkylation was achieved by incubating theexcised spots successively in 10 mmol/L Tris(2-carboxyethyl)phosphine/0.1 mol/L NH4HCO3 for 30 min at 37°C andin 55 mmol/L iodoacetamide/0.1 mol/L NH4HCO3 for20 min. Peptide fragments were obtained after digestionwith a solution of 40 mmol/L NH4HCO3 and 10% ACNcontaining 5 ng/mL trypsin (Promega) for 3 h at 37°C. Theresulting peptides were acidified with 0.1% formic acid.Extraction from the polyacrylamide gel piece was per-formed by two successive incubations for 5 min in ACNwith sonication. Digests were concentrated by evaporationuntil 10 mL. The concentrate (0.5 mL) was deposited ontoa Ground Steel target (Bruker Daltonics, Bremen, Ger-many), mixed with 1 mL matrix solution (3.5 mg/mL cyano-4-hydrocycinnamic acid [CHCA] in ACN 50% and TFA0.2%). Analysis was conducted using a MALDI-TOF/TOFultrafleXtreme (Bruker Daltonics) in the automatic modeoperating at 1,000 Hz in the reflectron mode. Mass cali-bration was done using the peptide calibration standardsfrom Bruker Daltonics. Proteins were identified withinSwiss-Prot, restricting the taxonomy to Homo sapiens. Me-thionine oxidation and carbamidomethyl modification of

cysteines were accepted as variable and global modifica-tions, respectively. The mass deviation tolerance was 30ppm in mass spectrometry, and only one miscleavage wassuggested.

Antioxidant StatusThe antioxidant defense of the subjects was examinedusing a test based on in vitro free radical–induced bloodhemolysis (27,30). Fasting blood samples were drawn us-ing 10% (v/v) buffered sodium citrate as the anticoagu-lant. Hemolysis was started by adding 52.4 mmol/L2,29-azo-bis(2-amidinopropane) HCl and was assayed bymonitoring absorbance at 450 nm (Dynatech MR5000)of diluted blood from the subjects. Results are expressedas 50% of maximal hemolysis time (HT50) (in min) andreferred to as the susceptibility of blood to free radicals.In humans and other animals in which oxidative stresshas been well documented, HT50 was shown to be rep-resentative of the total defenses against free radicals(31).

Plasma lipid peroxidation was measured as thiobarbi-turic acid reactive substances (TBARS) in the presence of2,6-di-tertbutyl-p-cresol and expressed as malondialdehydeequivalents as previously described (30).

Table 1—Clinical and oxidative parameters of control subjects and T2D patients

Control subjects T2D patients

n 20 20

Age (years) 36.8 6 12.3 56.9 6 12.9***

Sex (male/female) 8/12 7/13

BMI (kg/m2) 24.3 6 2.9 29.7 6 6.8**

Glycemia (mmol/L) 5.38 6 0.90 9.82 6 2.76**

HbA1c (%) ,6 9.3 6 1.4

Plasma cholesterol (mmol/L) 5.29 6 0.89 5.38 6 0.79

Plasma TG (mmol/L) 0.89 6 0.49 1.73 6 0.71***

Plasma NEFAs (mmol/L) 302.4 6 104.1 396.9 6 120.1*

LDL cholesterol (mmol/L) 3.24 6 0.71 3.30 6 0.71

HDL cholesterol (mmol/L) 1.49 6 0.57 0.90 6 0.48**

Fibrinogen (g/L) 2.85 6 0.30 3.21 6 0.33**

Albuminemia (g/L) 49.4 6 3.3 40.1 6 6.1***

Plasma TBARS (pmol/mL) 1.63 6 0.78 2.68 6 1.11**

HT50% (min)a 133.7 6 16.4 115.2 6 18.1**

Plasma thiols (mmol/L) 5.70 6 0.75 4.10 6 0.55**

Albumin thiols (mmol/g prot.) 4.21 6 0.83 2.89 6 0.92**

Albumin antiox (min)b 335 6 52 288 6 35**

OxLDL (mU/L) 59.2 6 16.1 74.8 6 18.4*

Glc-LDL (mg/g prot.) 0.81 6 0.31 1.11 6 0.58**

Glc-HSA (% HSA) 4.7 6 0.8 19.6 6 3.1***

Data are means6 SD. Glc-LDL, glycated LDL, Glc-HSA, glycated HSA; OxLDL, oxidized LDL; prot., protein; TG, triglyceride. Data werecompared using an unpaired t test: *P , 0.05, **P , 0.01, and ***P , 0.001. an = 14 for HT50 test for both control subjects and T2Dpatients; balbumin antioxidant activity (albumin antiox) was assayed as detailed in RESEARCHDESIGNANDMETHODS: the results are expressed(in minutes) as the time to obtain 50% of free radical–induced hemolysis in the presence of albumin isolated from the plasma of controlsubjects and T2D patients.

962 Albumin and Platelet Function in Diabetes Diabetes Volume 64, March 2015

Platelet Preparation, Aggregation, and AdhesionAs we needed a robust tool to test the effects of variousalbumin preparations on platelet function, we chose touse rat platelets because, due to their well-known lowerreactivity, they present a much higher reproducibility interms of activity and signaling compared with those fromhumans (32).

AnimalsMale white Sprague-Dawley rats (230–250 g), 7 weeks ofage, purchased from Charles River (L’Arbresle, France),were used as blood donors. They were housed in an air-conditioned room that was kept at constant temperatureand hygrometry and on a 12-h light-dark cycle. Water andfood were supplied ad libitum. All of the experiments in-volving animals were approved by the ethics committeeon the Use of Laboratory Animals of Université de Bourgogne(protocol no. 3508), and they were performed in accordancewith institutional guidelines.

Platelet PreparationBlood was drawn as in previous studies (24,33,34) fromthe jugular vein of overnight fasted animals into plasticsyringes with 1 v anticoagulant (38 mmol/L citric acid, 75mmol/L sodium citrate, and 136 mmol/L glucose) for 4 vblood. The platelet-rich plasma was obtained after centri-fugation (210g, 10 min). The platelets, which were isolatedby low-speed centrifugation, were counted with a thrombo-coulter (Coultronics) and resuspended in Ca2+-free Tyrodebuffer, pH 7.4.

Platelet AggregationAggregation experiments were performed as in previousstudies (24,33,34). Briefly, platelet suspension (3–4 3108/mL) was stimulated with agonists in a computerizedfour-channel laser-light coagulo-aggregometer (Servibio)under steering (1,100 rpm, 37°C). The final thrombin,ADP, calcium ionophore A23187, and arachidonate con-centrations were chosen so as to induce 30–40% of themaximal aggregation (100% = Tyrode buffer) and wereusually 0.08 units/mL, 0.6 mmol/L, 0.1 mmol/L, and 50mmol/L, respectively. The final results are expressed as thepercentage of control values obtained without albumin.

Platelet AdhesionThe platelet adhesion test was carried out using themodified colorimetric method of Bellavite et al. (35) andEriksson and Whiss (36). Briefly, after coating of a micro-titer plate with collagen I (Sigma), native BSA (N-BSA), ormodified BSA and thorough washing, diluted platelet-richplasma was added to the wells. After three washes withPBS, adherent platelets were measured by the reactionbetween p-nitrophenyl-diphosphate and the intracellularenzyme acid phosphatase liberated by lysis with 0.2% Tri-ton3100. Absorbance readings were performed at 405 nmafter background subtraction (630 nm) with a computer-ized microplate reader (Dynatech MR5000). Control valueswere obtained without coating (plastic alone). The

results are expressed as percentages with the followingformula: (sample 2 blank)/(total 2 blank) 3 100.

Platelet Biochemistry

Platelet Labeling and Extraction of ArachidonateMetabolitesPhospholipids of washed rat platelets were labeled as pre-viously described (33) by incubation with [1-14C]arachidonicacid (AA) (60 mCi/mmol; Amersham) for 60 min at 37°C.After temperature stabilization, the platelets (2 3 109

platelets/mL) were stimulated at 37°C for 2 min withthrombin (final concentration 0.2 units/mL; Sigma). Theresults for metabolites were expressed as dpm per 108

platelets. Platelet synthesis of thromboxane A2 was alsodetermined by enzyme immunoassay of its stable hydro-lysis product, thromboxane B2 (Cayman Chemical kits), aspreviously described (33). The results were expressed aspicomoles of thromboxane B2 (TxB2) per 10

8 platelets for5 min stimulation.

Calcium Uptake StudiesCalcium uptake by platelets was measured using radio-calcium 45CaCl2 (Amersham), following published meth-ods (33,34). Briefly, after the calcium concentration wasadjusted to 0.3 mmol/L, radiocalcium was added to theplatelet suspension (5 3 108/mL), 1 min before incuba-tion with thrombin (0.08 IU/mL), and centrifuged(9,000g, 30 s) through a layer of separating oil. The plate-let pellet was then treated with 0.5 mL 0.5% Triton X-100and transferred to counting vials with scintillation fluid.The results of the thrombin-induced calcium uptake wereexpressed as picomoles of Ca2+ per 109 platelets and com-pared with appropriate controls.

Platelet Production of Reactive Oxygen SpeciesTo measure intraplatelet reactive oxygen species (ROS), weused dichlorofluroscin diacetate (DCFH-DA) in the reducednonfluorescent form, which is able to penetrate platelets asdescribed in previous works (37). The results were expressedas picomoles of dichlorofluorescein (DCF) per 107 plateletsusing a calibration curve constructed with pure DCF.

Modifications of AlbuminAs the sulfhydryl (SH) group in most commercial BSA andHSA preparations is in the oxidized form (38), fatty acid–free BSA and HSA obtained from Sigma were first reducedby a 4-h treatment with 4.5 mmol/L dithiotreitol andpreserved from light. After dialysis against 20 mmol/LPBS, pH 7.4, to remove excess dithiotreitol, stock solu-tions of BSA (N-BSA) and native HSA (N-HSA) werestored for no longer than 2 weeks at 4°C under argon.This reduction treatment allowed the partial recovery offree Cys34 from albumin as assessed by the thiol value,which increases from ~0.13 to 0.72 SH/mol albumin (seebelow). These reduced BSA or HSA preparations wereused as the starting material for all of our studies. Pro-teins were measured using the bicinchoninic acid tech-nique (Pierce) (39).

diabetes.diabetesjournals.org Blache and Associates 963

Glycated BSA (G25-BSA) was obtained as previouslydetailed by Bourdon et al. (14). Briefly, BSA was incubatedwith 25 mmol/L glucose in sterile conditions for 30 daysat 37°C under argon. Excess glucose was removed by ex-tensive dialyses against PBS, pH 7.4. BSA was also mod-ified with 2 mmol/L MGO for 24 h at 37°C. Excess MGOwas removed by extensive dialyses against PBS to obtainMGO-modified BSA (MGO-BSA). Similar modificationswere performed with HSA to get G25-HSA and MGO-HSA.

Measurement of Thiol GroupsThiol groups of N-HSA or modified HSA were measuredaccording to the Ellman assay (40) using 5,59-dithiobis, 2-nitrobenzoic acid. The free thiol concentration was calcu-lated from absorbance data obtained at 410 nm using themolar extinction coefficient e = 13,600 L 3 mol21 3 cm21

and expressed as SH per mole of BSA.

Measurement of Primary Amino GroupsFree amino groups of native or modified BSA/HSAessentially from lysine were measured by means of theirreactivity with 2,4,6-trinitrobenzenesulfonic acid (41). Datafrom absorbance at 340 nm were calculated with the molarextinction coefficient e = 14,500 L 3 mol21 3 cm21, andLys-NH2 was thus expressed as a percentage of controlvalues obtained with N-BSA and N-HSA.

Fluorescence Studies of Albumin PreparationsAs BSA contains two tryptophan residues (Trp213 andTrp134) and HSA a single tryptophan (Trp214), we wereable to evaluate the effects of albumin modifications bymonitoring the molecular conformation change by assay-ing its intrinsic fluorescence. Fluorescence measurementsof albumin preparations were monitored using a spectro-fluorometer (LS 50B; PerkinElmer) at excitation and emis-sion wavelengths of 293 nm and 340 nm, respectively, asin previous studies (14). Quenching of fluorescence wasalso measured as previously described using acrylamide infinal concentrations up to 40 mmol/L. The results of trip-licate assays were expressed as mean values of I0/I (Stern-Volmer constant), where I0 and I represent the fluorescencebefore and after addition of the quencher, respectively.

Molecular conformation changes of albumin prepara-tions were also investigated by measuring the fluores-cence of 8-anilinonaphthalene-1-sulfonic acid (ANSA),which evaluates variations in the probe’s accessibility tohydrophobic protein sites, in accordance with previousstudies (42). After incubation for 10 min of 0.5 mg/mLnative or modified albumin with 200 mmol/L ANSA in0.15 mol/L NaCl, fluorescence was measured at excitationand emission wavelengths of 385 and 465 nm, respec-tively. The results are expressed as arbitrary units permg protein and changes normalized to N-BSA or N-HSA.

Interactions Between Oleic Acid and AlbuminInteractions between oleic acid and native and modifiedalbumin were evaluated by measuring the maximalabsorbance in the 250–300 nm range 10 min after the

addition of increasing concentrations of sodium oleate(up to 20 mmol/L) solutions (pH 9) according to themethodology of Fang et al. (43). The fitting equationwas Y = (Bmax 3 X)/(Kd + X), where X was the oleateconcentration in micromoles per liter, Bmax the maximalbinding as absorbance changes (at 292 nm), and Kd theapparent oleate concentration in micromoles per liter re-quired to reach half of the maximal binding.

Curve Fitting and Statistical AnalysesData are expressed as means 6 SD from at least threeexperiments performed in triplicate. Linear and nonlinearcurve fitting as well as the statistical comparisons wereevaluated using one-way ANOVA followed by the Bonferroniposttest performed with Prism software (GraphPad Soft-ware, Inc., San Diego, CA). ANCOVA was used to examinethe effect of T2D on plasma albumin levels when age wascontrolled. Statistical analyses were performed with STATAsoftware, version 11 (StataCorp).

Biostatistics for ProteomicsThirty spectra were available from the MALDI-TOF/TOFanalysis, but six of these did not meet biological criteriaand were removed from the analysis. The following stepswere taken to process the raw data: denoising, baselinesubtraction, and normalization. Peaks were identified andquantified by their intensities. Statistical analysis wasthus performed on 24 spectra (8 healthy and 16 diabeticspectra, respectively, representing healthy and diabeticpatients) for 231 variables. Data were preprocessed andanalyzed using R software (44). Principal componentsanalysis (PCA) was implemented to visualize the principalsources of variability in the data. The aim of PCA is toreduce the dimensionality of the data set by finding newcomponents that maximize variance of the data. The newvariables, called principal components, are linear combi-nations of all the original variables. A sparse partial leastsquares (SPLS) regression (45) was thus implemented onthe data to select peaks that mostly contribute to discrim-inate between both groups. This method is derived fromPLS (46). The aim of the SPLS is to reduce the dimen-sionality of the data set by finding new components thatmaximize the discrimination between diagnostic groups.SPLS simultaneously makes it possible to select variablesthat mostly contribute to the discrimination, thus facili-tating the biological interpretation; 20 peaks were therebyselected by SPLS as those separating at best samples fromcontrol subjects and samples from patients with diabetes.It was shown that these 20 peaks were sufficient to opti-mize predictions.

RESULTS

Patients With T2D Display Dysfunctional AlbuminTable 1 shows the clinical and lipid parameters of 20patients with T2D and 20 matched control subjects. Fast-ing glycemia was significantly higher in T2D with a meanHbA1c value of 9.3% compared with values ,6% in con-trol plasma. Triglyceridemia, plasma NEFAs, age, and BMI

964 Albumin and Platelet Function in Diabetes Diabetes Volume 64, March 2015

were significantly higher in T2D patients than in controlsubjects (P , 0.001, P , 0.05, P , 0.001, and P , 0.01,respectively). Although total cholesterol and LDL choles-terol did not differ significantly between the two groups,HDL cholesterol levels were significantly lower in T2Dpatients (P , 0.01). Plasma fibrinogen levels were higher(12%) and albumin values were lower (18%) in T2Dpatients than in control subjects, with threefold higherGlc-HSA in the former group (Table 1). This could notbe ascribable to an impairment of the renal status, ascreatinine values were normal (92.3 6 29.9 mmol/L), aswas the MDRD marker (64.2 6 26.8 mL/min). GlycatedLDL, TBARS, and oxidized LDL levels were significantlyhigher in T2D than in control subjects (P , 0.05 in allcases). Total free radical defense was measured on wholeblood samples, and the HT50 parameter (i.e., the 50%maximal hemolysis time) was found to be significantlyshorter in T2D than in control subjects (P , 0.01). Theseresults are in favor of a general impairment of the redoxstatus in T2D patients. This is in line with the reducedplasma thiol level in T2D patients compared with controlsubjects (P , 0.01) (Table 1).

Albumin was isolated from the plasma of controlsubjects and T2D patients. HSA fractions obtained forsix control subjects and six T2D patients were subjectedto two-dimensional SDS-PAGE, and the correspondingCoomassie-stained spots were analyzed by proteomics.After digestion with trypsin and extraction from gelpieces, peptide fragments were identified and quantitatedafter MALDI-TOF/TOF analysis. PCA enabled to showa clear separation between the two groups (Supplemen-tary Fig. 1). SPLS (45) aimed to make a regression oncomponents retaining only peaks contributing the mostto discrimination. This method was implemented on thedata, and 20 peaks were sufficient to separate the twogroups (Fig. 1A). A detailed comparative analysis of themost important peaks of trypsin-treated HSA, highlightedby SPLS, indicated that intensities differed between thetwo groups, suggesting differently modified residues (es-pecially Lys) (Supplementary Fig. 2). From these data, itcan be concluded that albumin fractions isolated from theplasma of control subjects and T2D patients have differ-ent structural characteristics.

HSA fractions were also specifically assayed for theirthiol content and antioxidant properties using a freeradical–induced hemolysis test. We found a significantlylower level of total plasma thiols (228%, P , 0.01), es-pecially with a lower level of albumin-associated free thi-ols (229%, P , 0.01), as well as shorter HT50 times(230%, P , 0.01) in T2D patients than in control sub-jects (Table 1). These data illustrate the reduced antioxi-dant capacity of T2D albumin compared with controlalbumin. The fatty acid binding capacity of albuminfrom control subjects and T2D patients was evaluatedthrough incubations with increasing concentrations ofoleic acid. As shown in Fig. 1B, the Bmax of isolated albu-min was significantly lower in T2D than in control

subjects (232%, P , 0.001). The half-saturation value(apparent Sat50%) of albumin was reached with a lowerconcentration of oleic acid for T2D patients than forcontrol subjects (4.61 6 1.05 vs. 6.60 6 1.08 mmol/L,respectively; P , 0.001).

Taken together, these data come in support of reducedsaturation of T2D albumin in the presence of oleic acidand suggest that the capacity of T2D albumin to bindunesterified fatty acids is blunted. This finding comes in

Figure 1—A: Plot of the first two SPLS components based on the20 peaks that best separated both groups. Peaks were obtained byMALDI-TOF/TOF of isolated albumins from control (Ctl-Alb) andT2D (Diab-Alb) plasma. B: Oleate binding to isolated albuminsfrom Ctl-Alb and Diab-Alb plasma. Albumin was isolated fromplasma, and oleate binding was conducted as detailed in RESEARCH

DESIGN AND METHODS. The gradual increase in Abs292 after addition ofincreasing concentrations of oleate to albumin preparations (finalalbumin concentrations: 1.5 mg/mL) was fitted with a sigmoidaldose response. The fitting equation was a four-parameter logisticequation: Y = bottom + {Bmax 2 bottom/[1 + 10^(LogSat50% 2 X)]3hillslope}, where X is the logarithm of oleate concentration in micro-moles per liter and Y starts at bottom and goes to Bmax with a sig-moidal shape to reach Bmax, which half saturates for Sat50%. *P< 0.05.

diabetes.diabetesjournals.org Blache and Associates 965

addition to the lower albumin levels in T2D patients thanin control subjects (Table 1).

Albumin Glycated In Vitro Mimics the Main Features ofAlbumin From T2D PatientsN-HSA was treated in vitro with either 25 mmol/L glucose(G25-HSA) or 2 mmol/L MGO (MGO-HSA). The experi-mental conditions were identical to those previouslydescribed and specifically set up to reproduce in vivomodifications resulting from hyperglycemia (14). Asexpected from earlier data for glycation (14) and as illus-trated in Fig. 2, changes in Trp emission fluorescencespectra were observed and plots of quenching experi-ments were drawn. The almost linear nature of the plotsindicated that the quenching process followed a collision-type mechanism. The Stern-Volmer constants were calcu-lated and showed significant differences between nativeand modified albumin (32.96 6 1.25, 25.37 6 1.26, and19.84 6 1.09 Ksv values for N-HSA, G25-HSA, and MGO-HSA, respectively, P , 0.01). As shown on spectra and Ksvplots (Fig. 2), albumin preparations from T2D patients (Diab-Alb) were found to behave similarly to MGO-HSA and G25-HSA, while, in contrast, albumin from control subjects didnot differ from N-Alb.

We further aimed to characterize the structural mod-ifications induced in albumin under our experimentalconditions, as well as the putative changes in the properties

of albumin. By using 5,59-dithiobis, 2-nitrobenzoic acid,we found that fewer thiol groups were available in G25-BSA and MGO-BSA than in N-BSA (259 and 231%,respectively, P , 0.01) (Table 2). In addition, therewere significantly fewer free Lys-NH2 groups in BSA in-cubated with 25 mmol/L glucose and BSA treated with2 mmol/L MGO than in N-Alb (229 and –33%, respec-tively, P , 0.01) (Table 2). As revealed by PAGE, ourconditions did not result in extensive damage to the pro-tein, with no detectable fragmentation or aggregation (datanot shown). ANSA fluorescence data indicated that alter-ations in albumin conformation led to major decreases inthe accessibility of hydrophobic domains. These observa-tions were similar for MGO- and G25-BSA (Fig. 3A).

As it is well known that albumin binds and transportsNEFAs, our results concerning ANSA binding promptedus to specifically investigate the interactions of fattyacids with regard to albumin modifications. After theincubation of distinct fatty acid–free albumin prepara-tions with increasing amounts of oleic acid, maximal ab-sorbance was measured in the 280–300 nm rangeaccording to previous works (43). Under our experimen-tal conditions, we found a gradual increase in the absor-bance at ;292 nm, which could be fitted according toa binding hyperbole (Fig. 3B). The apparent Kd valuecalculated with N-BSA was 1.6 6 0.4 mmol/L, whereasit was up to 4.5 times higher with modified albumin (P ,

Figure 2—Fluorescence emission spectra and acrylamide fluorescence quenching of native, modified albumin preparations and thealbumin fraction from control and diabetic subjects. A: Fluorescence emission spectra of N-HSA and selected representative samplesof albumin fractions isolated from one representative control subject (Ctl-Alb) and from one representative patient with diabetes (Diab-Alb).B: Stern-Volmer plots of acrylamide quenching of the tryptophan fluorescence of N-HSA and modified HSA (G25-HSA and MGO-HSA).Selected plots for acrylamide quenching experiments of samples of albumin fractions isolated from Ctl-Alb and from Diab-Alb are alsoillustrated. C: Fluorescence emission spectra of native and modified BSA preparations (G25-BSA and MGO-BSA). D: Stern-Volmer plots ofacrylamide quenching of the tryptophan fluorescence of N-BSA and modified BSA preparations. Note: The excitation wavelength was 293nm. I0 and I are the fluorescence intensities with and without quencher, respectively. SD bars are not shown for reasons of clarity.

966 Albumin and Platelet Function in Diabetes Diabetes Volume 64, March 2015

0.001) (Fig. 3B). These results indicate that, because ofthe modifications, albumin lost part of its affinity forNEFAs, since oleate binding is considered a good repre-sentative of other fatty acids (47). Based on curvesobtained in Fig. 3B and compared with N-Alb, we foundthat albumin modified with G25 and MGO had a signifi-cantly decreased Bmax for oleate.

Modification of BSA and Release of ActiveArachidonate DerivativesArachidonate-labeled platelets were used to monitor theeffects of the arachidonate pathway on thrombin stimu-lation in the presence of increasing concentrations ofN-BSA or modified BSA. We chose to use rat plateletsbecause of their known easy handling that confer a betterreproducibility than human platelets, especially for long-term incubation (32,33,48). Under our experimental con-ditions, most of the labeling material was recovered in theplatelet phospholipids, with no difference in the relativedistribution of the phospholipid fractions whether albu-min was added or not (not shown). The thrombin-inducedmobilization of radiolabeled AA preincorporated intophospholipids and the subsequent formation of labeledcyclooxygenase (COX) and lipoxygenase (LOX) productswere measured. After only 2 min of stimulation, labeledAA was released from the phospholipids, and the mainmetabolites (i.e., the stable derivatives of thromboxaneA2, TxB2, 12-hydroxy-heptadecanoic acid [12-HHT], and12-hydroxy-eicosatetraenoic acid [12-HETE]) were detectedin platelet extracts. The production of active metabo-lites of the COX pathway (HHT and TxB2) and of theLOX pathway (HETE) was significantly higher with mod-ified albumin (G25-BSA and MGO-BSA) than with N-BSA(Fig. 4).

These observations suggest that platelet-releasedarachidonate is sequestered by N-Alb and thus becomesless available for conversion into active metabolites byCOX and LOX enzymes. In contrast, with glycoxidizedalbumin, higher amounts of phospholipid-derived arach-idonate are available for conversion into oxidized derivatives,and concordant observations were made with G25-BSA andMGO-BSA.

Albumin Modification and Platelet ActivityFirst, the appropriate conditions to investigate the effectof albumin on platelet aggregation were set up. The re-sults illustrated in Fig. 5A clearly indicate that increasedconcentrations of N-BSA led to an increased inhibitoryeffect of albumin on platelet aggregation (as induced byvarious agonists such as thrombin, ADP, calcium ionophoreA23187, and arachidonate). In addition, the spontaneousadhesion of platelets to the microplates was found to begradually inhibited when the microplates were coated withincreasing amounts of N-BSA (Fig. 5B). In contrast, and inagreement with earlier observations (35), coating with col-lagen resulted in platelet adhesion as uncoated wells (Fig.5B). To investigate further the inhibitory properties of al-bumin on platelet signaling, we also analyzed TxB2 releaseby means of an ELISA and calcium fluxes in the presenceof radiocalcium. Thrombin-induced TxB2 was signifi-cantly reduced and in a concentration-dependent mannerin the presence of increasing concentrations of N-BSA(Fig. 5C1). The basal calcium uptake, measured without

Figure 3—Binding of the ANSA fluorescent probe and oleate tonative and modified albumin. A: Experiments with ANSA were car-ried out as detailed in RESEARCH DESIGN ANDMETHODS. The fluorescencemeasured at excitation and emission wavelengths of 385 and 465nm, respectively, is expressed as arbitrary unit per milligram proteinand changes normalized to N-BSA. B: Effects of increasing con-centrations of oleate on N-BSA and modified BSA on ultravioletabsorbance. The gradual increase in the absorbance at 292 nmafter the addition of increasing concentrations of oleate to BSApreparations was fitted with a binding hyperbole. The fitting equa-tion was Y = (Bmax 3 X)/(Kd + X), where X was oleate concentrationin micromoles per liter, Bmax the maximal binding as absorbancechanges, and Kd the oleate concentration in micromoles required toreach half-maximal binding. For reasons of clarity, the SD bars arenot shown. Statistical significance was calculated with ANOVA fordata comparisons vs. N-BSA, ***P < 0.001.

Table 2—Thiols and free Lys-amino groups in native andmodified BSA preparations

BSApreparations Thiols (per mol BSA) Lys-NH2 groups (%)

N-Alb 0.717 6 0.080 100

G25-Alb 0.426 6 0.009** 71.4 6 7.3**

MGO-Alb 0.226 6 0.041** 66.8 6 6.9**

Effects of glucose and MGO on thiol and Lys-NH2 groups ofBSA. Results are means 6 SD of 3–5 experiments. Treatmentprotocols were as detailed in RESEARCHDESIGN ANDMETHODS. Meanswere multiple compared with N-Alb using ANOVA with theDunnett posttest for unpaired samples. G25-Alb, BSA incubatedwith 25 mmol/L glucose; MGO-Alb, BSA treated with 2 mmol/LMGO. **P , 0.01.

diabetes.diabetesjournals.org Blache and Associates 967

thrombin, was not different whether N-BSA was added ornot (data not shown). In contrast, the thrombin-inducedcalcium influx was significantly reduced with increasingconcentrations of N-BSA (Fig. 5C2). To analyze whetherN-BSA influenced platelet oxidative metabolism, plateletswere probed with DCFH-DA. N-BSA significantly decreasedarachidonate-mediated DCF production in a dose-dependentway (Fig. 5C3).

Again, albumin was modified by incubation with 25mmol/L glucose (G25-BSA) and MGO (MGO-BSA), andpreparations were tested for their effects on plateletfunction in comparison with N-BSA at identical finalconcentrations (2 mg/mL). The results for thrombin-induced platelet aggregation and platelet adhesion areshown in Fig. 6. We found that G25-BSA and MGO-BSAsignificantly lost their inhibitory capacity on platelet ag-gregation compared with N-BSA (Fig. 6A). Similarly, theinhibitory effect on platelet adhesion was greatly impairedwhen wells were coated with modified albumin (Fig. 6B).In addition, platelet signaling (Fig. 6C and D) and ROSproduction (Fig. 6E) stimulated with thrombin were alsoimpaired after incubation with modified albumin prepa-rations as shown by calcium uptake and DCF level.

Finally, platelets were incubated with albumin frac-tions isolated from control subjects and patients withT2D, and they were stimulated with thrombin. As forN-HSA, which was used in control samples, albumin iso-lated from healthy control subjects was able to markedlyinhibit the thrombin-induced aggregation of platelets(Fig. 6F). Compared with albumin isolated from healthycontrol subjects, albumin from patients with T2D diabetesdisplayed an impaired capacity to block platelet aggrega-tion (44.0% in T2D patients vs. 80.9% in healthy controlsubjects). These results are in line with observations madeabove with modified BSA.

DISCUSSION

In the present work, albumin abnormalities resultingfrom glycation and glycoxidation were characterized. Theywere found to produce significant impairment in thebinding capacity of NEFAs, as found in patients with T2D,thus contributing significantly to platelet hyperactivity inthis population.

Isolated albumin fractions from T2D patients were shownhere to display reduced NEFA binding capacity. Whenalbumin was incubated in vitro with increasing concentra-tions of oleic acid, the saturation curves obtained for T2Dalbumin were significantly different from those for albuminfrom control volunteers, as characterized by reduced halfsaturation and decreased Bmax of NEFAs. Importantly, thisimpaired ability to bind NEFAs may lead to several biologicalconsequences. First, relative amounts of unbound NEFAsmay increase, thus raising their potential adverse effectsthrough interactions with lipoproteins and cells. Second,NEFA binding and albumin oxidation of the thiol groupare known to be intimately linked, and the more NEFAsbound to albumin, the more Cys34 was oxidized (49). Incomparison with albumin from control subjects, that fromT2D patients displayed reduced levels of free thiols and Lys-NH2 and decreased antioxidant activity (14,42). The presentdata are in line with an increase in the overall oxidativestatus of patients with diabetes as assessed by increasedplasma TBARS and oxidized LDL (50,51). Significant struc-tural differences were found between isolated albumin ofcontrol subjects and that of T2D patients as assessed byproteomics. Furthermore, conformation analyses of albuminfrom T2D patients carried out with an ANSA fluorescentprobe or Trp fluorescence revealed similarities with albuminmodified in vitro by glycation and MGO. Most importantly,the modifications of N-Alb obtained here in vitro led todrastic decreases in the accessibility of hydrophobic domainsthat constitute NEFA-binding sites. These data indicate thatthe abnormalities of T2D albumin can be reproduced in vitro.Although the current study involved reconstituted experi-mental media, both albumin and NEFAS were used in rela-tive proportions that were similar to those found in vivo (inboth cases, approximately one-twentieth of normal biologicalconcentrations). In this respect, it could be considered ofpotential clinical relevance.

As interactions between NEFAs and platelets are ofmajor importance in hemostasis, the effects of native andmodified albumin on platelet function were compared.NEFAs, including AA, are well-known platelet stimulators(52). When platelets are activated, large amounts of AAare released from membrane phospholipids through thelipolytic activity of phospholipase A2, and AA is the initialsubstrate for the subsequent biosynthesis of prostanoids,hydoxyacids, and thromboxanes produced by LOX andCOX (21,22). In the present work, we showed that plate-lets treated with Glc-Alb and MGO-Alb released signifi-cantly higher amounts of AA metabolites (HHT, HETE,and TxB2) than did platelets incubated with N-Alb or

Figure 4—Effects of native and modified BSA on arachidonate me-tabolism in stimulated platelets. Washed platelets from rats wereprelabeled with AA, and the platelets were then stimulated withthrombin (0.2 units/mL) for 2 min. AA metabolites were extractedand separated by TLC, which resolves phospholipids (PL), TxB2,12-HHT, and 12-HETE. Values are expressed as dpm per 108

platelets of labeling recovered in AA-containing lipid and metabo-lites. Statistical significance was calculated with ANOVA for datacomparisons vs. either control, *P < 0.05 and ***P < 0.001, or BSA#P < 0.05, ##P < 0.01, and ###P < 0.001.

968 Albumin and Platelet Function in Diabetes Diabetes Volume 64, March 2015

those incubated in the absence of albumin. Albumin iso-lated from T2D patients presented a higher thrombin-induced aggregation than the albumin fraction of control

subjects. Concomitantly, platelet adhesion and thrombin-induced aggregation were blocked by N-Alb but to a muchlower extent by Glc-Alb and MGO-Alb. These findings

Figure 5—Effect of increasing concentrations of albumin on thrombin-induced platelet aggregation and on platelet adhesion. A: Foraggregation, the agonist was added to washed rat platelets 2 min after incubation with the indicated BSA concentration (in milligramsper milliliter), and aggregation was recorded as indicated in RESEARCH DESIGN ANDMETHODS. A1, thrombin (0.08 units/mL); A2, ADP (0.6 mmol/L);A3, ionophore A23187 (0.1 mmol/L); and A4, sodium arachidonate (50 mmol/L). Each value corresponds to means and SD of at least threedifferent assays. **P< 0.01. B: For platelet adhesion, adhesion of resting platelets to microplate wells uncoated (Ctl) or coated with 2 mg/mLcollagen (Coll) and N-BSA (1–2 mg/mL) was assessed as indicated in RESEARCH DESIGN AND METHODS. After washing away of nonadherentplatelets, the acid phosphatase activity of lysed adherent platelets was measured. Results (mean 6 SD, n = 3 of triplicate values) areexpressed as percentages of total incubated platelets. The significance of the results was determined by ANOVA followed by Dunnettmultiple comparison test: ***P< 0.001 comparison with uncoated or collagen coating. C: Effects of BSA on platelet signaling. C1: Thrombin-induced platelet TxB2 synthesis. TxB2 released after 2-min stimulation with thrombin was measured in supernatants with an enzymeimmunoassay. C2: Thrombin-induced calcium uptake. After 1 min, the platelets preincubated with radiocalcium were thrombin stimulatedfor 2 min. The platelet pellets obtained after centrifugation (see RESEARCH DESIGN AND METHODS) were counted for radioactivity. C3: Thrombin-stimulated ROS production. After loading with DCFH-DA (see RESEARCH DESIGN AND METHODS), the platelets were stimulated with thrombin andfluorescence was measured in Triton-lysed platelets (lexc504/lem526). The significance of the results, expressed as picomoles of DCF per107 platelets and mean 6 SD, n = 3–7, was determined by ANOVA followed by Tukey multiple comparison test: *P < 0.05 and **P < 0.01with control conditions (Ctl). Plts, platelets.

diabetes.diabetesjournals.org Blache and Associates 969

support the hypothesis that the NEFA-binding propertyof albumin is essential to sequester platelet-derivedNEFAs. This mechanism occurs to a significantly lowerextent when the albumin has undergone glycoxidation.Beyond its role in the systemic, intravascular transportof adipocyte-derived NEFAs toward the liver (53), albu-min also seems to be a key component in minimizing theavailable amounts of the AA precursor in the vicinity ofplatelets. This provides a new explanation for the knownhyperreactivity of washed platelets, namely, plateletsresuspended in buffer in the absence of plasma compo-nents (54). Given the impaired redox status that occurs ina context of hyperglycemia (55), the conditions requiredto obtain such modifications in the albumin moleculeare likely to be encountered in T2D. In support of thisview, we found in the current study that T2D patients

cumulated the two conditions that are potentially harmfulin terms of NEFA handling, i.e., concomitant reductionsin both the concentration and the NEFA binding capacityof albumin. A number of pathological states, includinglipid disorders, nephrotic syndrome, malnutrition, obe-sity, and metabolic syndrome in addition to diabetes, areassociated with high thrombotic risk (22). The current studystrongly suggests that, as well as hypoalbuminemia, which isknown to occur in these pathological states (56,57), an ab-normal reduction in the NEFA binding capacity of albuminmight well contribute significantly to the disorders.

A high level of circulating NEFAs is predictive of suddendeath, stroke, and ischemic heart disease as reported bya large number of population studies (10,58,59). The nu-merous deleterious effects such as cardiac arrhythmia andproinflammatory and prothrombotic states have been the

Figure 6—Effects of native and modified BSA and isolated albumin on platelet aggregation, adhesion, or signaling. A: Thrombin-inducedplatelet aggregation was measured in the presence of 2 mg/mL N-BSA or modified BSA under the same conditions as Fig. 5A. B: Adhesionof resting platelets to microplate wells coated with various BSA preparations (2 mg/mL) under the same conditions as Fig. 5B. Results(mean6 SD) are expressed as percentages of total incubated platelets for platelet adhesion. The significance of the results was determinedby ANOVA followed by Bonferroni multiple comparison test: ***P < 0.001 comparison with N-BSA. C–E: Effects of N-BSA and modifiedBSA on thrombin-stimulated platelet signaling. C: Thrombin-induced platelet TxB2 synthesis. D: Thrombin-induced calcium uptake. E: ROSgeneration measured as DCF. The significance of the results (A–E, expressed as mean6 SD, n = 3–7) was determined by ANOVA followedby Tukey multiple comparison test: ***P < 0.001. F: Thrombin-induced platelet aggregation measured in the presence of 2 mg/mL albuminisolated from 12 control or T2D subjects (Ctl-Alb and Diab-Alb, respectively) and compared with N-HSA. Statistical significance wascalculated with ANOVA followed by Bonferroni multiple comparison test: ***P < 0.001. Plts, platelets.

970 Albumin and Platelet Function in Diabetes Diabetes Volume 64, March 2015

subject of intense research, which has shown that some ofthe deleterious effects could be directly attributed to ele-vated plasma NEFAs (12,60). However, though the effectsof high levels of total plasma NEFAs have been studied(10,11,61), the putative deterioration in the NEFA bindingcapacity of albumin was not addressed in a systematic way.In previous studies, it was reported that oxidative stressand glycation caused conformational changes in albuminstructure (14,29,62) with possible effects in vitro on plateletfunction (17,63,64). The novelty of the current study isthe demonstration that glycation probably impairs thebinding properties of albumin of patients with T2D.When the NEFA binding capacity of albumin is bluntedby glycoxidation, higher amounts of free arachidonateare available for the generation of prothrombotic oxidizedmetabolites in platelets, which become hyperactive.

LimitationsThe current study presents some limitations, in particularthe enrollment of a relatively low number of patients withT2D from a single center. In addition, the patients haddiabetes that was poorly controlled, with high HbA1c lev-els, thus limiting the conclusions of the study to a transienttime in the life of most patients with T2D. At this time,one cannot assess whether the observed changes also ap-ply to patients with type 1 diabetes and whether they aredue to insulin resistance, glycemia, BMI, or age. The pres-ent work constitutes a first line of investigation and war-rants additional exploration in larger and specificallydesigned clinical trials to confirm that higher amounts offree arachidonate are made available for the generation ofactive metabolites in platelets when the NEFA bindingcapacity of albumin is blunted by glycoxidation.

Acknowledgments. The authors thank Philip Bastable (Research Unit,Dijon Centre Hospitalier Universitaire) for manuscript editing and D. Souyhel andN. Loreau for their help in this study.Funding. This work was supported in part by INSERM, Nouvelle SociétéFrançaise d’Athérosclérose, Conseil Régional de Bourgogne, Université deBourgogne, and a French government grant managed by the French NationalResearch Agency under the program Investissements d’Avenir (ANR-11-LABX-0021-LipSTIC).Duality of Interest. No potential conflicts of interest relevant to this articlewere reported.Author Contributions. D.B. researched data and drafted, reviewed, andedited the manuscript. E.B. researched data and reviewed and edited the man-uscript. P.S., G.L., and P.D. researched data for proteomics and statistics andreviewed and edited the manuscript. J.-M.P. and B.V. took part in providingclinical data and recruitment and reviewed and edited the manuscript. L.L.contributed to data interpretation and reviewed and edited the manuscript.D.B. is the guarantor of this work and, as such, had full access to all the datain the study and takes responsibility for the integrity of the data and the accuracyof the data analysis.

References1. Goldwasser P, Feldman J. Association of serum albumin and mortality risk.J Clin Epidemiol 1997;50:693–7032. Bourdon E, Blache D. The importance of proteins in defense against oxi-dation. Antioxid Redox Signal 2001;3:293–311

3. Jalan R, Bernardi M. Effective albumin concentration and cirrhosis mortality:from concept to reality. J Hepatol 2013;59:918–9204. Luoma PV, Näyhä S, Sikkilä K, Hassi J. High serum alpha-tocopherol, al-bumin, selenium and cholesterol, and low mortality from coronary heart diseasein northern Finland. J Intern Med 1995;237:49–545. Phillips A, Shaper AG, Whincup PH. Association between serum albuminand mortality from cardiovascular disease, cancer, and other causes. Lancet1989;2:1434–14366. Dziedzic T, Slowik A, Szczudlik A. Serum albumin level as a predictor ofischemic stroke outcome. Stroke 2004;35:e156–e1587. Nagumo K, Tanaka M, Chuang VT, et al. Cys34-cysteinylated human serumalbumin is a sensitive plasma marker in oxidative stress-related chronic dis-eases. PLoS ONE 2014;9:e852168. Quinlan GJ, Martin GS, Evans TW. Albumin: biochemical properties andtherapeutic potential. Hepatology 2005;41:1211–12199. Zheng H, Liu J, Liu Y, Klaassen CD. Hepatocytes from metallothionein-I andII knock-out mice are sensitive to cadmium- and tert-butylhydroperoxide-inducedcytotoxicity. Toxicol Lett 1996;87:139–14510. Jouven X, Charles MA, Desnos M, Ducimetière P. Circulating nonesterifiedfatty acid level as a predictive risk factor for sudden death in the population.Circulation 2001;104:756–76111. Boden G. Role of fatty acids in the pathogenesis of insulin resistance andNIDDM. Diabetes 1997;46:3–1012. Oliver MF, Opie LH. Effects of glucose and fatty acids on myocardial is-chaemia and arrhythmias. Lancet 1994;343:155–15813. Ordoñez JD, Hiatt RA, Killebrew EJ, Fireman BH. The increased risk ofcoronary heart disease associated with nephrotic syndrome. Kidney Int 1993;44:638–64214. Bourdon E, Loreau N, Blache D. Glucose and free radicals impair the anti-oxidant properties of serum albumin. FASEB J 1999;13:233–24415. Huebschmann AG, Regensteiner JG, Vlassara H, Reusch JE. Diabetes andadvanced glycoxidation end products. Diabetes Care 2006;29:1420–143216. Gawlowski T, Stratmann B, Stirban AO, Negrean M, Tschoepe D. AGEs andmethylglyoxal induce apoptosis and expression of Mac-1 on neutrophils resultingin platelet-neutrophil aggregation. Thromb Res 2007;121:117–12617. Gawlowski T, Stratmann B, Ruetter R, et al. Advanced glycation endproducts strongly activate platelets. Eur J Nutr 2009;48:475–48118. Cohen MP. Clinical, pathophysiological and structure/function consequencesof modification of albumin by Amadori-glucose adducts. Biochim Biophys Acta2013;1830:5480–548519. Song SO, Kim KJ, Lee BW, Kang ES, Cha BS, Lee HC. Serum glycatedalbumin predicts the progression of carotid arterial atherosclerosis. Atheroscle-rosis 2012;225:450–45520. Ahmed N, Thornalley PJ. Advanced glycation endproducts: what is theirrelevance to diabetic complications? Diabetes Obes Metab 2007;9:233–24521. Blache D. Structure and function of blood platelets. Arch Int Physiol BiochimBiophys 1992;100:A17–A24 [in French]22. Davì G, Patrono C. Platelet activation and atherothrombosis. N Engl J Med2007;357:2482–249423. Blache D, Bouthillier D, Davignon J. Acute influence of smoking on plateletbehaviour, endothelium and plasma lipids and normalization by aspirin. Ath-erosclerosis 1992;93:179–18824. Durand P, Prost M, Blache D. Pro-thrombotic effects of a folic acid de-ficient diet in rat platelets and macrophages related to elevated homocysteineand decreased n-3 polyunsaturated fatty acids. Atherosclerosis 1996;121:231–24325. Schneider M, Vergès B, Klein A, et al. Alterations in plasma vitamin Edistribution in type 2 diabetic patients with elevated plasma phospholipid transferprotein activity. Diabetes 2004;53:2633–263926. Zeller M, Masson D, Farnier M, et al. High serum cholesteryl ester transferrates and small high-density lipoproteins are associated with young age in pa-tients with acute myocardial infarction. J Am Coll Cardiol 2007;50:1948–1955

diabetes.diabetesjournals.org Blache and Associates 971

27. Blache D, Devaux S, Joubert O, et al. Long-term moderate magnesium-deficient diet shows relationships between blood pressure, inflammation andoxidant stress defense in aging rats. Free Radic Biol Med 2006;41:277–28428. Doumas BT, Watson WA, Biggs HG. Albumin standards and the measure-ment of serum albumin with bromcresol green. Clin Chim Acta 1971;31:87–9629. Faure P, Wiernsperger N, Polge C, Favier A, Halimi S. Impairment of theantioxidant properties of serum albumin in patients with diabetes: protectiveeffects of metformin. Clin Sci (Lond) 2008;114:251–25630. Blache D, Lussier-Cacan S, Gagnon J, et al. Effect of exercise training on invitro LDL oxidation and free radical-induced hemolysis: the HERITAGE FamilyStudy. Antioxid Redox Signal 2007;9:123–13031. Girodon F, Blache D, Monget AL, et al. Effect of a two-year supplementationwith low doses of antioxidant vitamins and/or minerals in elderly subjects onlevels of nutrients and antioxidant defense parameters. J Am Coll Nutr 1997;16:357–36532. Sudo T, Ito H, Kimura Y. Characterization of platelet aggregation in wholeblood of laboratory animals by a screen filtration pressure method. Platelets2003;14:239–24633. Blache D, Durand P, Prost M, Loreau N. (+)-Catechin inhibits platelet hy-peractivity induced by an acute iron load in vivo. Free Radic Biol Med 2002;33:1670–168034. Polette A, Blache D. Effect of vitamin E on acute iron load-potentiatedaggregation, secretion, calcium uptake and thromboxane biosynthesis in ratplatelets. Atherosclerosis 1992;96:171–17935. Bellavite P, Andrioli G, Guzzo P, et al. A colorimetric method for the mea-surement of platelet adhesion in microtiter plates. Anal Biochem 1994;216:444–45036. Eriksson AC, Whiss PA. Measurement of adhesion of human platelets inplasma to protein surfaces in microplates. J Pharmacol Toxicol Methods 2005;52:356–36537. Blache D, Gesquière L, Loreau N, Durand P. Oxidant stress: the role ofnutrients in cell-lipoprotein interactions. Proc Nutr Soc 1999;58:559–56338. Bar-Or D, Bar-Or R, Rael LT, Gardner DK, Slone DS, Craun ML. Hetero-geneity and oxidation status of commercial human albumin preparations inclinical use. Crit Care Med 2005;33:1638–164139. Smith PK, Krohn RI, Hermanson GT, et al. Measurement of protein usingbicinchoninic acid. Anal Biochem 1985;150:76–8540. Ellman GL. Tissue sulfhydryl groups. Arch Biochem Biophys 1959;82:70–7741. Steinbrecher UP, Witztum JL, Parthasarathy S, Steinberg D. Decrease inreactive amino groups during oxidation or endothelial cell modification of LDL.Correlation with changes in receptor-mediated catabolism. Arteriosclerosis 1987;7:135–14342. Bourdon E, Loreau N, Lagrost L, Blache D. Differential effects of cysteineand methionine residues in the antioxidant activity of human serum albumin. FreeRadic Res 2005;39:15–2043. Fang Y, Tong GC, Means GE. Structural changes accompanying humanserum albumin’s binding of fatty acids are concerted. Biochim Biophys Acta2006;1764:285–29144. R Core Team. R: a language for statistical computing. R foundation forStatistical Computing, Vienna, Austria, 2013

45. Lê Cao KA, Rossouw D, Robert-Granié C, Besse P. A sparse PLS for variable

selection when integrating omics data. Stat Appl Genet Mol Biol 2008;7:3546. Wold H. Estimation of principal components and related models by iterarive

least squares. In Multivariate Analysis. Krishnaiah P, Ed. New York, Academic

Press, 1966, p. 391–42047. Spector AA. Fatty acid binding to plasma albumin. J Lipid Res 1975;16:

165–17948. Durand P, Prost M, Blache D. Folic acid deficiency enhances oral contraceptive-

induced platelet hyperactivity. Arterioscler Thromb Vasc Biol 1997;17:1939–194649. Gryzunov YA, Arroyo A, Vigne JL, et al. Binding of fatty acids facilitates

oxidation of cysteine-34 and converts copper-albumin complexes from anti-

oxidants to prooxidants. Arch Biochem Biophys 2003;413:53–6650. Davì G, Falco A, Patrono C. Lipid peroxidation in diabetes mellitus. Antioxid

Redox Signal 2005;7:256–26851. Stocker R, Keaney JF Jr. Role of oxidative modifications in atherosclerosis.

Physiol Rev 2004;84:1381–147852. Hoak JC. Stearic acid, clotting, and thrombosis. Am J Clin Nutr 1994;60

(Suppl.):1050S–1053S53. Lafontan M. Advances in adipose tissue metabolism. Int J Obes (Lond)

2008;32(Suppl. 7):S39–S5154. Doweiko JP, Bistrian BR. The effect of glycosylated albumin on platelet

aggregation. JPEN J Parenter Enteral Nutr 1994;18:516–52055. Brownlee M, Cerami A, Vlassara H. Advanced glycosylation end products in

tissue and the biochemical basis of diabetic complications. N Engl J Med 1988;

318:1315–132156. Braschi S, Lagrost L, Florentin E, et al. Increased cholesteryl ester transfer

activity in plasma from analbuminemic patients. Arterioscler Thromb Vasc Biol1996;16:441–44957. Braschi S, Masson D, Rostoker G, et al. Role of lipoprotein-bound NEFAs in

enhancing the specific activity of plasma CETP in the nephrotic syndrome. Ar-

terioscler Thromb Vasc Biol 1997;17:2559–256758. Pirro M, Mauriège P, Tchernof A, et al. Plasma free fatty acid levels and the

risk of ischemic heart disease in men: prospective results from the Québec

Cardiovascular Study. Atherosclerosis 2002;160:377–38459. Carlsson M, Wessman Y, Almgren P, Groop L. High levels of nonesterified

fatty acids are associated with increased familial risk of cardiovascular disease.

Arterioscler Thromb Vasc Biol 2000;20:1588–159460. Makiguchi M, Kawaguchi H, Tamura M, Yasuda H. Effect of palmitic acidand fatty acid binding protein on ventricular fibrillation threshold in the perfused

rat heart. Cardiovasc Drugs Ther 1991;5:753–76161. Pilz S, März W. Free fatty acids as a cardiovascular risk factor. Clin Chem

Lab Med 2008;46:429–43462. Oettl K, Stauber RE. Physiological and pathological changes in the redox

state of human serum albumin critically influence its binding properties. Br J

Pharmacol 2007;151:580–59063. Rubenstein DA, Yin W. Glycated albumin modulates platelet susceptibility to

flow induced activation and aggregation. Platelets 2009;20:206–21564. Yamazaki E, Inagaki M, Kurita O, Inoue T. Kinetics of fatty acid binding

ability of glycated human serum albumin. J Biosci 2005;30:475–481

972 Albumin and Platelet Function in Diabetes Diabetes Volume 64, March 2015