Inhibition of Nitrobenzylthioinosine-Sensitive AdenosineTransport by Elevated D-Glucose Involves Activation of P2Y2

Purinoceptors in Human Umbilical Vein Endothelial Cells

Jorge Parodi, Carlos Flores, Claudio Aguayo, M. Isolde Rudolph, Paola Casanello, Luis Sobrevia

Abstract—Chronic incubation with elevated D-glucose reduces adenosine transport in endothelial cells. In this study,exposure of human umbilical vein endothelial cells to 25 mmol/L D-glucose or 100 �mol/L ATP, ATP-�-S, or UTP, butnot ADP or �,�-methylene ATP, reduced adenosine transport with no change in transport affinity. Inhibition of transportby D-glucose, ATP, and ATP-�-S was associated with reduced maximal binding, with no changes in the apparentdissociation constant for nitrobenzylthioinosine (NBMPR). A significant reduction (�60�10%, P�0.05; n�6) in thenumber of human equilibrative NBMPR-sensitive nucleoside transporters (hENT1s) per cell (1.8�0.1�106 in 5 mmol/LD-glucose) and in hENT1 mRNA levels was observed in cells exposed to D-glucose or ATP-�-S. Incubation withelevated D-glucose, but not with D-mannitol, increased the ATP release by 3�0.2-fold . The effects of D-glucose andnucleotides on the number and activity of hENT1 and hENT1 mRNA were blocked by reactive blue 2 (nonspecific P2Y

purinoceptor antagonist), suramin (G�s protein inhibitor), or hexokinase but not by pyridoxal phosphate-6-azophenyl-2�,4�-disulfonic acid (nonselective P2 purinoceptor antagonist). Our findings demonstrate that inhibition of adenosinetransport via hENT1 in endothelial cells cultured in 25 mmol/L D-glucose could be due to stimulation of P2Y2

purinoceptors by ATP, which is released from these cells in response to D-glucose. This could be a mechanism to explainin part the vasodilatation observed in the early stages of diabetes mellitus or in response to D-glucose infusion. (Circ Res.2002;90:570-577.)

Key Words: endothelium � adenosine � nitric oxide � glucose � purinoceptors

Removal of extracellular adenosine is an essential step inthe modulation of several of the biological actions of this

endogenous nucleoside.1–4 Plasma and tissue levels of aden-osine are regulated by an efficient membrane transportmediated by the Na�-independent, nitrobenzylthioinosine(NBMPR)-sensitive equilibrative nucleoside transporter (sys-tem es or ENT1)3,4 in human vascular endothelium5,6 andsmooth muscle.7,8 Human ENT1 (hENT1) expression in Rajicells (a human B-lymphocyte cell line) is dependent on NOlevels and the activity of protein kinase C (PKC).9 Incubationof human umbilical vein endothelial cells (HUVECs) with25 mmol/L D-glucose for 24 hours has been reported toreduce the NBMPR-sensitive adenosine transport associatedwith increased protein levels and the activity of endothelialNO synthase, intracellular Ca2�, PKC, and mitogen-activatedprotein kinases p42/p44mapk.6,10 Thus, hENT1 adenosine trans-porters could be expressed and modulated in HUVECs.

It has been reported that ATP inhibits dipyridamole-sensitive adenosine transport in human pulmonary artery

endothelium.11 ATP also induces activation of PKC in endo-thelium from human umbilical vein,12 bovine pulmonaryartery,13 and porcine aorta.14,15 Activation of P2Y1 and P2Y2

purinoceptors with ATP induced the phosphorylation ofp42mapk in the human endothelial cell line EAhy 92616 andp42/p44mapk in bovine aortic endothelium.17 Therefore, thecellular effects of elevated D-glucose and activation of P2Y

purinoceptors could involve common signal transductionpathways in human endothelium.

We have investigated the involvement of P2Y purinoceptors inthe effect of elevated D-glucose on NBMPR-sensitive adenosinetransport in cultures of HUVECs. We established that endothe-lial cells express the hENT1 isoform of nucleoside transportersand that incubation with 25 mmol/L D-glucose leads to inhibitionof adenosine transport by a mechanism that involves the activa-tion of P2Y2 purinoceptors. In addition, elevated D-glucosediminished hENT1 mRNA levels, an effect mimicked by ATPand blocked by P2Y antagonists. A preliminary account of thepresent study has been reported.18

Original received July 27, 2001; revision received January 29, 2002; accepted January 29, 2002.From the Cellular and Molecular Physiology Laboratory, Department of Physiology (J.P., C.F., C.A., M.I.R., P.C., L.S.), the Department of

Pharmacology (M.I.R.), Faculty of Biological Sciences, and the Department of Obstetrics and Gynecology (P.C.), Faculty of Medicine, University ofConcepción, Concepción, Chile.

Presented in part at The Physiological Society meeting, King’s College London, UK, December 18–20, 2000, and published in abstract form [J Physiol(Lond). 2001;531:36P].

Correspondence to Dr L. Sobrevia, Cellular and Molecular Physiology Laboratory (CMPL), Department of Physiology, Faculty of Biological Sciences,University of Concepción, PO Box 160-C, Concepción, Chile. E-mail lsobrev@udec.cl

© 2002 American Heart Association, Inc.

Circulation Research is available at http://www.circresaha.org DOI: 10.1161/01.RES.0000012582.11979.8B

570

Materials and Methods

Cell CultureHUVECs were isolated from full-term normal pregnancies. Informedwritten consent was given from the hospital for the use of theumbilical cords. Cells isolated by collagenase (0.25 mg/mL) diges-tion were cultured (37°C, 5% CO2) in medium 199 (M199) contain-ing 5 mmol/L D-glucose, 20% bovine sera, 3.2 mmol/L L-glutamine,and 100 U/mL penicillin-streptomycin as described.5 Twenty-fourhours before an experiment, the incubation medium was changed toserum-free M199.

Adenosine TransportAdenosine transport (4 �Ci/mL) was measured as described.5,6 Cellswere rinsed with warmed (37°C) Krebs solution containing(mmol/L) NaCl 131, KCl 5.6, NaHCO3 25, NaH2PO4 1, D-glucose 5,HEPES 20, CaCl2 2.5, and MgCl2 1 (pH 7.4), containing 100 �mol/LL-arginine. Triplicate monolayer wells were then preincubated (30minutes, 22°C) in Krebs solution or in Krebs solution containing theadenosine transport inhibitor NBMPR (10 �mol/L).

Endothelial cells were preexposed for 2, 4, 10, or 60 minutes and12, 18, or 24 hours to M199 containing 5 mmol/L D-glucose,25 mmol/L D-glucose or L-glucose, or 5 mmol/L D-glucose plus20 mmol/L D-mannitol as osmotic control.6,19 The kinetics ofadenosine transport was measured in cells incubated with increasingconcentrations of adenosine (0 to 500 �mol/L, 5 seconds, 22°C) inKrebs solution. Tracer uptake was terminated by rinsing the mono-layers (3 times) with 200 �L ice-cold Krebs solution containing 10�mol/L NBMPR, and cell radioactivity was determined by liquidscintillation counting.6,8

Adenosine transport was also determined in cells exposed to theP2Y antagonists reactive blue 2 (RB2, 0.1 to 100 nmol/L, 5 minutesor 24 hours), pyridoxal phosphate-6-azophenyl-2�,4�-disulfonic acid(PPADS, 0.1 to 100 nmol/L, 5 minutes or 24 hours),20,21 or the G�s

protein inhibitor 8-(3-benzamido-4-methylbenzamido)-naphthalene-1,3,4-trisulfonic acid (suramin, 100 �mol/L, 15 minutes or 24hours).22 Cells were then exposed to ATP (0.1 to 100 �mol/L, 2minutes), which is a nucleotide hydrolyzed by ectonucleotidases inhuman endothelium,5 ATP-�-S (0.1 to 100 �mol/L, 2 minutes or 24hours), which is a nonhydrolyzable analogue of ATP,23 ADP (0.1 to100 �mol/L, 2 minutes), UTP (0.1 to 100 �mol/L, 2 minutes), or�,�-methylene ATP dilithium (�,�-MeATP, 0.1 to 100 �mol/L, 2minutes), which is a nonselective P2X purinoceptor agonist, in theabsence or presence of RB2, PPADS, or suramin. The effects ofD-glucose and ATP were also assayed in cells preincubated (10minutes or 24 hours) with 10 U/mL hexokinase.24

NBMPR Binding[3H]NBMPR equilibrium binding studies were performed in cellspreincubated in Krebs solution or in Krebs solution containing 10�mol/L NBMPR. Cells were then exposed (30 minutes, 22°C) to[3H]NBMPR in the presence of 5 or 25 mmol/L D-glucose. Specificbinding was defined as the difference in the binding in the presenceand absence of 10 �mol/L NBMPR.5,6

Measurement of Extracellular ATPExtracellular ATP was determined in M199 from cells cultured in 5or 25 mmol/L D-glucose or in 5 mmol/L D-glucose plus 20 mmol/LD-mannitol for 2, 4, 10, or 60 minutes and 12, 18, or 24 hours byluminometry.25 Aliquots of 200 �L were collected at the beginning(time 0) and after indicated periods of time and stored at �20°C for16 to 17 hours. Aliquots of 100 �L were mixed with 100 �Lluciferase reagent (pH 7.7), and the reaction was processed with theATP bioluminescence assay kit CLS II (Roche). Bioluminescence ofsamples and standards was monitored at 562 nm (10 seconds, 22°C)in a luminometer (Lumat LB 9501, Berthold). Detection limit was 1fmol ATP per sample.

Detection of hENT1Cells cultured in M199 containing 5 or 25 mmol/L D-glucose for 24hours were rinsed with PBS, and mRNA was extracted by using theDynabeads technique (Dynal). The mRNA was reversed-transcribedinto cDNA by using oligo(dT18) plus random hexamers and Moloneymurine leukemia virus reverse transcriptase (Promega) for 1 hour at37°C. Polymerase chain reactions (PCRs) were performed in a totalvolume of 20 �L containing 2 �L of 10� PCR buffer, 2 mmol/LMgCl2, 2 U Taq DNA polymerase (GIBCO Life Technologies), andsequence-specific oligonucleotide primers (0.5 �mol/L) for humanENT1. Samples were incubated for 3 minutes at 97°C, followed by5 cycles of 30 seconds at 94°C, 4 minutes at 67°C, 5 cycles of 30seconds at 94°C, 4 minutes at 65°C, 35 cycles of 45 seconds at 94°C,6 minutes at 63°C, and a final extension for 7 minutes at 61°C.�-Actin primers were used as housekeepers.

Oligonucleotide primers were for hENT1 (sense) 5�-CATGAT-CTGCGCTATTGCCAGTGG-3�, hENT1 (antisense) 5�-AACCA-GGCATCGTGCTCGAAGACCA-3�, �-actin (sense) 5�-AACCGC-GAGAAGATGACCCAGATCATCTTT-3�, and �-actin (antisense)5�-AGCAGCCGTGGCCATCTCTTGCTCGAAGTC-3�. Expectedsize products were 617 bp for hENT1 and 350 bp for �-actin.

MaterialsNewborn and fetal calf serum and agarose were from GIBCO LifeTechnologies. Collagenase type II (Clostridium histolyticum) wasfrom Boehringer-Mannheim. Bradford protein reagent was from

Figure 1. Involvement of P2 purinoceptors in adenosine trans-port in HUVECs. A, Overall transport of adenosine (10 �mol/L,20 seconds, 22°C) was determined in passage-2 cells culturedfor 24 hours in 5 or 25 mmol/L D-glucose in the absence orpresence of RB2, PPADS, or suramin. B, Adenosine transportwas determined in cells cultured in 5 mmol/L D-glucose andincubated with ATP-�-S (24 hours) or ATP (2 minutes), underthe same conditions as in panel A. Values are mean�SEM(n�6). *P�0.05 vs all other values.

Parodi et al Inhibition of Adenosine Transport by Glucose 571

Bio-Rad Laboratories. D-Glucose, D-mannitol, hexokinase, andethidium bromide were from Sigma Chemical Co. [2,8,5�-3H]Aden-osine (60 Ci/mmol) and D-[1-14C]mannitol (49.3 mCi/mmol) werefrom NEN. [3H]NBMPR (80 mCi/mmol) was from MoraveckBiochemicals. Agonists and antagonists were from RBI ResearchBiochemical International.

Statistical AnalysisValues are mean�SEM, and n indicates different umbilical veinendothelial cell cultures with 3 to 6 replicate measurements perexperiment. Statistical analyses were carried out on raw data byusing the Peritz F multiple means comparison test.26 A Student t testwas applied for unpaired data, and a value of P�0.05 was consideredstatistically significant.

ResultsEffect of D-Glucose on Adenosine TransportWe have reported that adenosine transport is inhibited by 10nmol/L NBMPR or after incubation with 25 mmol/LD-glucose.5,6 In the present study, inhibition of NBMPR-sensitive adenosine (10 �mol/L) transport induced by25 mmol/L D-glucose was blocked after incubation of thecells with RB2 or suramin but not PPADS (Figure 1A).Adenosine transport was also inhibited by ATP-�-S or UTP;this effect was blocked by RB2 and suramin (Figure 1B).Inhibition of adenosine transport by ATP, ATP-�-S, or UTPin cells cultured in 5 mmol/L D-glucose was concentrationdependent (Figure 2A), with similar apparent Ki values (Table1). Neither ADP nor �,�-MeATP changed adenosine trans-port in HUVECs. Adenosine transport in 25 mmol/LD-glucose was unaltered by nucleotides (Figure 2B). Prein-cubation of the cells with hexokinase blocked (P�0.05, n�4)the inhibitory effect of 2-minute exposure (45�5 pmol/106

cells per second) or 24-hour exposure (37�6 pmol/106 cellsper second) to 25 mmol/L D-glucose or 2-minute exposure to100 �mol/L ATP (41�3 pmol/106 cells per second) on 10�mol/L adenosine transport.

Inhibition of adenosine transport by D-glucose, ATP-�-S,or UTP (24 hours) was associated with reduced Vmax forsaturable transport, with negligible changes in apparent Km

(Table 1). Cells incubated for 2 minutes with D-glucose orATP exhibited a reduced adenosine transport that was alsoassociated with lower Vmax (245�56 or 225�34 pmol/106

cells per second for D-glucose or ATP, respectively), with nosignificant changes in apparent Km (112�34 or 109�13�mol/L for D-glucose or ATP, respectively). Cell incubationwith RB2, but not with PPADS (not shown), restored thereduced Vmax for adenosine transport induced by 2-minuteincubation with D-glucose (574�63 pmol/106 cells per sec-ond, Km 107�44 �mol/L) or ATP (633�76 pmol/106 cellsper second, Km 118�51 �mol/L) or 24-hour incubation withelevated D-glucose (Figure 3A) or ATP-�-S (Figure 3B) tovalues in cells cultured in 5 mmol/L D-glucose (Vmax 641�29pmol/106 cells per second, Km 90�11 �mol/L). RB2 orPPADS had no significant effect on adenosine transportkinetics in cells in 5 mmol/L D-glucose (Table 1).

Effect of D-Glucose on NBMPR BindingTo determine whether the effects of D-glucose or ATP-�-S onVmax for adenosine transport were due to changes in thenumber of available adenosine transport sites, [3H]NBMPR

equilibrium binding was determined.5 Table 2 shows thatD-glucose or ATP-�-S (24 hours) reduced the maximalbinding (Bmax) of [3H]NBMPR by 58�12%, with no signifi-cant changes in the Kd. The effects of D-glucose and ATP-�-Son Bmax were blocked by RB2 but not by PPADS. Scatchardplots of specific binding data were lineal (not shown),indicating a single population of high-affinity NBMPR bind-ing sites in cells cultured in 5 or 25 mmol/L D-glucose, in theabsence or presence of ATP-�-S and/or RB2. Similar resultswere obtained in cells exposed for 2 minutes to elevatedD-glucose (Bmax 1.1�0.2 pmol/106 cells, Kd 0.17�0.02nmol/L) or ATP (Bmax 0.9�0.3 pmol/106 cells, Kd 0.22�0.03nmol/L) compared with values in 5 mmol/L D-glucose (Bmax

3.1�0.2 pmol/106 cells, Kd 0.21�0.02 nmol/L). RB2 blockedthe effect of 2 minutes of D-glucose (Bmax 2.9�0.4 pmol/106

cells, Kd 0.18�0.02 nmol/L) or ATP (Bmax 3.3�0.6 pmol/106

cells, Kd 0.20�0.02 nmol/L) on NBMPR binding.

Figure 2. Effect of different nucleotides on adenosine transportin HUVECs. Adenosine transport (10 �mol/L, 20 seconds, 22°C)was determined in cells cultured for 24 hours in M199 contain-ing 5 mmol/L (A) or 25 mmol/L (B) D-glucose in the absence orpresence of ATP-�-S or UTP. Cells were also exposed for 2minutes to ATP, ADP, or �,�-MeATP. Adenosine transport in theabsence of nucleotides (100% transport) was 32�5 and 12�5pmol/106 cells per second for 5 and 25 mmol/L D-glucose,respectively. Values are mean�SEM (n�8). Some error bars(�7.5% of measured transport) and connecting lines weredeleted for clarity.

572 Circulation Research March 22, 2002

Time-Course Effect of D-Glucose on AdenosineTransport and ATP ReleaseATP release from cells cultured in M199 containing5 mmol/L D-glucose was increased by 25 mmol/L D-glucosefor different time periods (Figure 4A). The effect ofD-glucose was not due to osmotic changes, inasmuch as cellsincubated with equimolar concentrations of D-mannitol (ie,5 mmol/L D-glucose�20 mmol/L D-mannitol) exhibited ATPrelease similar to that of cells in 5 mmol/L D-glucose. ATPrelease in cells exposed to hexokinase for 2 minutes or 24hours was marginal. D-Glucose–induced ATP release wasparalleled by reduced adenosine transport, an effect blockedby hexokinase (Figure 4B) and RB2 but not by PPADS (notshown).

Effect of D-Glucose and ATP-�-S on hENT1mRNA LevelsCompared with incubation of the cells in 5 mmol/LD-glucose, incubation of the cells in 25 mmol/L D-glucose for24 hours reduced the hENT1 mRNA level (Figure 5). Theeffect of D-glucose was inhibited by RB2 but not by PPADS.

RB2 and PPADS alone did not significantly alter hENT1mRNA in cells in 5 mmol/L D-glucose. Similarly, when cellswere incubated with ATP-�-S, hENT1 mRNA was signifi-cantly reduced, an effect blocked by RB2 but not by PPADS(Figure 6). The hENT1 mRNA level was unchanged in cellsexposed for 2 to 60 minutes to elevated D-glucose, ATP, orATP-�-S (not shown).

DiscussionThe present study has established that HUVECs express thehENT1 transporter isoform and that inhibition of adenosinetransport and of NBMPR binding by elevated D-glucose isassociated with the activation of P2Y2 purinoceptors.D-Glucose increased ATP release, and ATP, ATP-�-S, orUTP, but not ADP or �,�-MeATP, mimicked the inhibitoryeffects of D-glucose on adenosine transport and NBMPRbinding. D-Glucose and ATP-�-S also reduced the number ofNBMPR-sensitive adenosine transporters and hENT1 mRNAlevels; this effect was blocked by P2Y purinoceptorantagonists.

TABLE 1. Effect of D-Glucose and Nucleotides on the Kinetic Parameters ofAdenosine Transport in HUVECs

Km, �mol/LVmax, pmol � (106

Cells)�1 � s�1 Ki, �mol/L

5 mmol/L D-glucose

Control 90�11 641�29 ND

ATP 98�45 156�21* 0.35�0.06

ATP-�-S 128�41 211�26* 0.42�0.09

UTP 127�39 198�64* 0.41�0.05

ADP 101�29 598�54 No inhibition

RB2 102�34 598�45 ND

PPADS 95�45 624�47 ND

ATP-�-S�RB2 108�30 660�70† ND

ATP-�-S�PPADS 118�50 271�31* ND

25 mmol/L D-glucose

Control 127�44 227�30* ND

ATP 131�26 254�49* No inhibition

ATP-�-S 145�41 237�61* No inhibition

UTP 112�21 199�32* No inhibition

ADP 125�19 187�44* No inhibition

RB2 86�26 554�59‡ ND

PPADS 132�31 199�58* ND

ATP-�-S�RB2 95�32 559�61‡ ND

ATP-�-S�PPADS 112�14 199�34* ND

ND indicates not determined. Values are mean�SEM (n�8). Saturable adenosine transport wasdetermined in cells cultured for 24 hours in M199 containing 5 or 25 mmol/L D-glucose in theabsence or presence of 100 �mol/L ATP-�-S, 100 �mol/L UTP, 100 nmol/L RB2, or 100 nmol/LPPADS. The effect of 100 �mol/L ATP or 100 �mol/L ADP on transport was assayed by incubationof cells for 2 minutes with these nucleotides. For inhibition studies, adenosine transport wasdetermined in cells exposed to increasing concentrations (0 to 100 �mol/L) of nucleotides. Theapparent inhibition constants (Ki) were calculated by using the expression Ki�IC50/(1�[Ado]/Km),where Km is the apparent Km value for adenosine transport, [Ado] is adenosine concentration(10 �mol/L), and IC50 is the half-maximal inhibitory concentration of the inhibitors.5

*P�0.05 vs control in 5 mmol/L D-glucose; †P�0.05 vs ATP-�-S in 5 mmol/L D-glucose; and‡P�0.05 vs control in 25 mmol/L D-glucose.

Parodi et al Inhibition of Adenosine Transport by Glucose 573

Adenosine transport was inhibited after the incubation ofendothelial cells with 25 mmol/L D-glucose, confirming ourprevious observations in this cell type.6 The inhibition ofadenosine transport induced by D-glucose was blocked by thenoncompetitive nonspecific P2Y purinoceptor antagonistRB227,28 and by the G�s protein inhibitor suramin29,30 but wasunaffected by the nonselective P2 purinoceptor antagonistPPADS, suggesting the involvement of P2 purinoceptors inthe effects of D-glucose. This could be due to ATP releasedfrom HUVECs in response to D-glucose, inasmuch as hex-okinase, an ATP-degrading enzyme,24 blocked the effect ofD-glucose, and a 3-fold increase in the extracellular ATP levelwas detected in cells cultured in 25 mmol/L D-glucosecompared with 5 mmol/L D-glucose (�35 pmol/mL). BasalATP release from HUVECs is within the range of concen-trations reported for this cell type (�40 pmol/mL).25 In-creased extracellular ATP derived from freshly dissociated orcultured endothelial cells has been shown to be a rapid

response of cells to shear stress.25,31 Elevated D-glucose is astress condition associated with metabolic alterations invascular endothelium,2,32,33 which could explain our findingsof a higher extracellular ATP level.

Involvement of P2Y2 Purinoceptors in the Effect ofD-Glucose on Adenosine TransportHUVECs express at least 4 isoforms of P2Y purinergicreceptors, ie, P2Y1, P2Y2, P2Y4, and P2Y6,34,35 which exhibitdifferent sensitivities for nucleotides and have been shown tomediate several cellular responses.20,21,36 P2Y2 and P2Y4 puri-noceptors are stimulated by ATP and UTP but are insensitiveto ADP; P2Y1 purinoceptors are stimulated by ATP and ADPbut not by UTP; and P2Y6 purinoceptors are stimulated byADP but are insensitive to ATP or UTP.21,36 Thus, theinhibition of adenosine transport by high D-glucose, ATP,ATP-�-S, or UTP could result from the activation of P2Y2 orP2Y4 purinoceptors in HUVECs. In addition, P2Y2, but not P2Y1,purinoceptors are stimulated by UTP; both purinoceptors areinhibited by RB220; and P2Y4 purinoceptors are insensitive toinhibition by suramin.22 Thus, P2Y2 purinoceptors (the formerP2U receptors)37 could be responsible for the inhibitory effectof D-glucose on adenosine transport in human endothelium.Because �,�-MeATP, a general P2X purinoceptor agonist,20,21

does not alter adenosine transport, it is suggested that thesepurinoceptors are not involved in the effect of elevatedD-glucose on adenosine transport.

Figure 3. Involvement of P2 purinoceptors in the effect of ele-vated D-glucose on kinetics of adenosine transport in HUVECs.A, Initial rates of adenosine transport (20 seconds, 22°C) weremeasured in cells cultured for 24 hours in M199 containing 5 or25 mmol/L D-glucose in the absence or presence of RB2 (100nmol/L). B, Adenosine transport was measured in cells culturedin M199 containing 5 mmol/L D-glucose in the absence (control)or presence of ATP-�-S (100 �mol/L) or ATP-�-S and RB2 (100nmol/L). Values are mean�SEM (n�6).

TABLE 2. Effect of D-Glucose and ATP-�-S on the KineticParameters of NBMPR Binding in HUVECs

Kd, nmol/L Bmax, pmol/106 Cells

5 mmol/L D-glucose

Control 0.21�0.02 3.1�0.2

RB2 0.19�0.03 2.9�0.3

PPADS 0.22�0.01 2.9�0.4

ATP-�-S 0.28�0.04 0.8�0.2*

ATP-�-S�RB2 0.18�0.03 2.7�0.2†

ATP-�-S�PPADS 0.19�0.04 1.1�0.3*

25 mmol/L D-glucose

Control 0.19�0.04 1.3�0.3*

RB2 0.18�0.02 3.5�0.5‡

PPADS 0.21�0.01 0.9�0.3*

ATP-�-S 0.16�0.04 1.4�0.1*

ATP-�-S�RB2 0.19�0.03 3.6�0.5‡

ATP-�-S�PPADS 0.22�0.01 1.1�0.3*

Values are mean�SEM (n�6). Endothelial cells were cultured for 24 hoursin M199 containing 5 or 25 mmol/L D-glucose in the absence or presence of100 �mol/L ATP-�-S, 100 nmol/L RB2, or 100 nmol/L PPADS. Cells were thenwashed and preincubated in Krebs buffer for 15 minutes in the absence orpresence of 10 �mol/L NBMPR. Monolayers were then incubated with[3H]NBMPR for 30 minutes at 22°C in Krebs buffer. Specific cell-associatedradioactivity was defined as the difference between total binding and bindingin the presence of 10 �mol/L NBMPR.

*P�0.05 vs control in 5 mmol/L D-glucose; †P�0.05 vs ATP-�-S in5 mmol/L D-glucose; and ‡P�0.05 vs control in 25 mmol/L D-glucose.

574 Circulation Research March 22, 2002

Effect of D-Glucose on the Number ofAdenosine TransportersAs reported, inhibition of adenosine transport by elevatedD-glucose was associated with a reduced Vmax.6 The effect ofD-glucose was mimicked by ATP, ATP-�-S, and UTP andblocked by RB2. These results were similar to changesinduced by D-glucose, ATP, and ATP-�-S in NBMPR-binding kinetics. The adenosine transport inhibitor NBMPRbinds specifically to ENT1 (system es) transporters but is nottransported itself; therefore, it can be used to estimate thesurface density of ENT1 transporters in intact cells.5,38,39

Thus, the inhibition of adenosine transport by elevatedD-glucose and adenine or uridine nucleotides could be due to

the reduced number rather than the activity of an existingpool of NBMPR-sensitive nucleoside transporters in theplasma membrane of HUVECs. This conjecture is supportedby the finding that the number of adenosine transporters percell (1.8�0.1�06 transporters/cell) was significantly reducedby 25 mmol/L D-glucose (0.7�0.2�06 transporters/cell,P�0.05; n�8) or 100 �mol/L ATP-�-S (0.5�0.1�06 trans-porters/cell, P�0.04; n�12). However, the D-glucose– orATP-�-S–induced reduction in adenosine transport is not dueto changes in the turnover number (ie, Vmax/number oftransporters per cell)5,8 for adenosine (356�30 versus324�45 or 439�75 adenosine molecules/transporter persecond for 5 mmol/L versus 25 mmol/L D-glucose or 100�mol/L ATP-�-S, respectively). These results are similar toprevious reports showing a reduced number of adenosinemembrane transporters without altering its turnover rate inhuman vascular endothelium5 or smooth muscle cells7 ob-tained from gestational diabetic pregnancies or in vascularsmooth muscle cells exposed to human insulin.8

Parallel experiments demonstrated a reduced hENT1mRNA level in cells incubated with elevated D-glucose orATP-�-S for 24 hours. However, as expected, acute incuba-tion of cells with elevated D-glucose or ATP (2 minutes) didnot change hENT1 mRNA levels. Thus, possible explana-tions for a reduced number of hENT1 transporters are a lowertranscription due to long exposure to D-glucose or an in-creased turnover rate of hENT1 transporters as described inother cell types.1–3 The latter is supported by the finding of areduced number of hENT1 transporters available at theplasma membrane after a brief (2-minute) exposure to ele-vated D-glucose (0.7�0.1�106 transporters/cell, P�0.05;n�6) or ATP (0.5�0.2�106 transporters/cell, P�0.05; n�6).Reduction in the number of adenosine transporters andhENT1 mRNA by D-glucose, ATP, and ATP-�-S wasblocked by RB2 but was unaltered by PPADS, indicating thatactivation of P2Y purinoceptors leads to a lower uptake ofadenosine by reducing hENT1 expression. hENT1 has beencolocalized with A1 nucleoside receptors in the human centralnervous system,4,40,41 suggesting a role of the hENT1-mediated transport process in the control of adenosine-mediated biological actions.2,42,43 Thus, expression of hENT1transporters could be crucial in human pathological tissues inwhich high levels of D-glucose or adenosine nucleotidescould modulate endothelial cell function, such as in diabetesmellitus.2

The present results demonstrate that elevated D-glucoseinduced a reduction in adenosine transport in human umbil-ical vein endothelium by a mechanism that involves activa-tion of P2Y purinoceptors (possibly the P2Y2 subtype). ATPmay mediate the effect of elevated D-glucose, inasmuch asextracellular levels of this nucleotide are elevated in25 mmol/L D-glucose, and ATP (and ATP-�-S) mimicked theeffects of D-glucose on adenosine transport and expression ofhENT1. Thus, ATP could be playing an autocrine role inresponse to elevated D-glucose in HUVECs. The presentstudy is the first report to demonstrate modulation of hENT1expression and activity in human endothelium since thecloning of this transporter from human tissue.3,39,42 Removalof extracellular adenosine is a key mechanism in the reduc-

Figure 4. Time-course effect of elevated D-glucose on ATPrelease and adenosine transport in HUVECs. A, Cells were cul-tured for different periods of time in M199 containing 5 or25 mmol/L D-glucose, 5 mmol/L D-glucose�20 mmol/LD-mannitol, or 25 mmol/L D-glucose�10 U/mL hexokinase. Ali-quots (100 �L) of M199 collected at the beginning (time 0) or atindicated incubation periods were mixed with 100 �L luciferasereagent, and ATP bioluminescence was monitored at 562 nm for10 seconds at 22°C. B, Overall transport of 10 �mol/L adeno-sine (20 seconds, 22°C) was measured in M199 containing5 mmol/L D-glucose (time 0) or M199 containing 25 mmol/LD-glucose in the absence or presence of hexokinase (10 U/mL)for the indicated incubation periods. Values are mean�SEM(n�12).

Parodi et al Inhibition of Adenosine Transport by Glucose 575

tion of extracellular levels of this nucleoside, modulating itsbiological actions on vascular cells.1–4 Adenosine has beenshown to mediate vasodilatation via adenosine receptors byincreasing NO synthesis from endothelial cells.43,44 Thus, areduced removal of extracellular adenosine by the endotheli-um under pathological conditions in which plasma D-glucose

is increased (such as in uncontrolled diabetes) could, in part,explain the early generalized vasodilatation observed inpatients affected by this syndrome.2,32,33,45

AcknowledgmentsThis study was supported by Fondo Nacional de Ciencia y Tec-nología (FONDECYT 1000354 and 7000354) and Dirección deInvestigación, University of Concepción (DIUC 201.084.003-1.0),Concepción, Chile, and The Wellcome Trust (UK). J. Parodi holdsan MSc fellowship and P. Casanello holds a PhD fellowship fromBeca Docente University of Concepción. C. Aguayo holds a CONI-CYT (Chile) PhD fellowship. We thank Dr J. Villegas (UniversidadLa Frontera, Chile) for contributing the ATP measurements. We alsothank the midwives of Hospital Regional, Concepción, Chile, laborwards for the supply of umbilical cords, Susana Rojas for technicalassistance, and Isabel Jara for secretarial assistance.

References1. Griffith DA, Jarvis SM. Nucleoside and nucleobase transport systems of

mammalian cells. Biochim Biophys Acta. 1996;1286:153–181.2. Sobrevia L, Mann GE. Dysfunction of the nitric oxide signalling pathway

in diabetes and hyperglycaemia. Exp Physiol. 1997;82:1–30.3. Baldwin S, Mackey J, Cass C, Young J. Nucleoside transporters:

molecular biology and implications for therapeutic development. MolMed Today. 1999;53:216–224.

4. Jennings LL, Hao C, Cabrita MA, Vickers MF, Baldwin SA, Young JD,Cass CE. Distinct regional distribution of human equilibrative nucleosidetransporter proteins 1 and 2 (hENT1 and hENT2) in the central nervoussystem. Neuropharmacology. 2001;40:722–731.

5. Sobrevia L, Jarvis SM, Yudilevich DL. Adenosine transport in culturedhuman umbilical vein endothelial cells is reduced in diabetes. Am JPhysiol. 1994;267:C39–C47.

6. Montecinos VP, Aguayo C, Flores C, Wyatt AW, Pearson JD, Mann GE,Sobrevia L. Regulation of adenosine transport by D-glucose in humanfetal endothelial cells: involvement of nitric oxide, protein kinase C andmitogen-activated protein kinase. J Physiol (Lond). 2000;529:777–790.

7. Aguayo C, Sobrevia L. Nitric oxide, cGMP and cAMP modulate nitro-benzylthioinosine sensitive adenosine transport in human umbilical arterysmooth muscle cells from gestational diabetes. Exp Physiol. 2000;85:399–409.

Figure 5. Effect of elevated D-glucose onhENT1 mRNA levels in HUVECs.RT-PCR was performed for mRNAextracted from cells cultured for 24hours in M199 containing 5 or25 mmol/L D-glucose in the absence orpresence of RB2 or PPADS. The mRNAwas reversed-transcribed into cDNA (1hour, 37°C), and PCRs were performedby using sequence-specific oligonucleo-tide primers (0.5 �mol/L) for hENT1 (sizeproduct 617 bp), with �-actin (size prod-uct 350 bp) used as housekeeper. Dataare representative of 5 different cellcultures.

Figure 6. Effect of ATP-�-S on hENT1 mRNA levels in HUVECs.RT-PCR was performed for mRNA extracted from cells culturedfor 24 hours in M199 containing 5 mmol/L D-glucose and ATP-�-S in the absence or presence of RB2 or PPADS. Reversetranscription and PCR for hENT1 (size product 617 bp) wereperformed as described in the Figure 5 legend. �-Actin (sizeproduct 350 bp) was used as housekeeper. Data are represen-tative of 5 different cell cultures.

576 Circulation Research March 22, 2002

8. Aguayo C, Flores C, Parodi J, Rojas R, Mann GE, Pearson JD, SobreviaL. Modulation of adenosine transport by insulin in human umbilical arterysmooth muscle cells from normal or gestational diabetic pregnancies.J Physiol (Lond). 2001;534:243–254.

9. Soler C, Felipe A, Mata JF, Casado FJ, Celada A, Pastor-Anglada M.Regulation of nucleoside transport by lipopolysaccharide, phorbol esters,and tumor necrosis factor-� in human B-lymphocytes. J Biol Chem.1998;273:26939–26945.

10. Sobrevia L, Nadal A, Yudilevich DL, Mann GE. Activation of L-argininetransport (system y�) and nitric oxide synthase by elevated glucose andinsulin in human endothelial cells. J Physiol (Lond). 1996;490:775–781.

11. Dawicki DD, Chatterjee D, Wyche J, Rounds S. Extracellular ATP andadenosine cause apoptosis of pulmonary artery endothelial cells. Am JPhysiol. 1997;273:L485–L494.

12. Ethier MF, Dobson JG Jr. Adenosine stimulation of DNA synthesis inhuman endothelial cells. Am J Physiol. 1997;272:H1470–H1479.

13. Chen BC, Lin WW. PKC�I mediates the inhibition of P2Y receptor-induced inositol phosphate formation in endothelial cells. Br JPharmacol. 1999;27:1908–1914.

14. Castro AF, Amorena C, Müller A, Ottaviano G, Tellez-Inon MT, TaquiniAC. Extracellular ATP and bradykinin increase cGMP in vascular endo-thelial cells via activation of PKC. Am J Physiol. 1998;275:C113–C119.

15. Noll T, Holschermann H, Koprek K, Gunduz D, Haberbosch W,Tillmanns H, Piper HM. ATP reduces macromolecule permeability ofendothelial monolayers despite increasing [Ca2�]i. Am J Physiol. 1999;276:H1892–H18901.

16. Graham A, McLees A, Kennedy C, Gould GW, Plevin R. Stimulation bythe nucleotides, ATP and UTP of mitogen-activated protein kinase inEAhy 926 endothelial cells. Br J Pharmacol. 1996;117:1341–1347.

17. Patel J, Abeles A, Block ER. Nitric oxide exposure and sulfhydrylmodulation alter L-arginine transport in cultured pulmonary artery endo-thelial cells. Free Radic Biol Med. 1996;20:629–637.

18. Parodi J, Roa J, Rudolph I, Flores C, Aguayo C, Guijarro MV, SobreviaL. Modulation of L-arginine and adenosine transport by elevatedD-glucose is mediated by activation of P2Y purinoceptors involving acti-vation of p44/p42mapkin human fetal vein endothelium. J Physiol (Lond).2001;531:36P. Abstract.

19. Sobrevia L, Cesare P, Yudilevich DL, Mann GE. Diabetes-induced acti-vation of system y� and nitric oxide synthase in human endothelial cells:association with membrane hyperpolarization. J Physiol (Lond). 1995;489:183–192.

20. Ralevic V, Burnstock G. Receptors for purines and pyrimidinesPharmacol Rev. 1998;50:413–492.

21. Boarder M, Hourani S. The regulation of vascular function by P2

receptors: multiple sites and multiple receptors. Trends Physiol Sci. 1998;19:99–107.

22. Freissmuth M, Boehm S, Beindl W, Nickel P, Ijzerman AP, HoheneggerM, Nanoff C. Suramin analogues as subtype selective G protein inhibi-tors. Mol Pharmacol. 996;49:602–611.

23. Han J, Kim N, Kim E, Ho WK, Earm YE. Modulation of ATP-sensitivepotassium channels by cGMP-dependent protein kinase in rabbit ventric-ular myocytes. J Biol Chem. 2001;276:22140–22147.

24. Nicholas RA, Watt WC, Lazarowski ER, Li Q, Harden TK. Uridinenucleotide selectivity of three phospholipase C-activating P2 receptors:identification of a UDP-selective, a UTP-selective, and an ATP- andUTP-specific receptor. Mol Pharmacol. 1996;50:224–229.

25. Bodin P, Burnstock G. Increased release of ATP from endothelial cellsduring acute inflammation. Inflamm Res. 1998;47:351–354.

26. Harper JF. Peritz’ F test: BASIC program of a robust multiple comparisontest for statistical analysis of all differences among group means. ComputBiol Med. 1984;14:437–445.

27. Burnstock G, Warland JJ. P2-purinoceptors of two subtypes in the rabbitmesenteric artery: reactive blue 2 selectively inhibits responses mediatedvia the P2Y- but not the P2X-purinoceptor. Br J Pharmacol. 1987;90:383–391.

28. Chen BC, Lee C-M, Lin WW. Inhibition of ecto-ATPase by PPADS,suramin and reactive blue 2 in endothelial cells, C6 glioma cells and RAW264.7 macrophages. Br J Pharmacol. 1996;119:1628–1634.

29. Dunn PM, Blakeley AG. Suramin. a reversible P2-purinoceptor antagonistin the mouse vas deferens. Br J Pharmacol. 1988;93:243–245.

30. Usune S, Katsuragi T, Furukawa T. Effects of PPADS and suramin oncontractions and cytoplasmic Ca2�changes evoked by AP4A, ATP and�,�-methylene ATP in guinea-pig urinary bladder. Br J Pharmacol.1996;117:698–702.

31. Bodin P, Bailey D, Burnstock G. Increased flow-induced ATP releasefrom isolated vascular endothelial cells but not smooth muscle cells. Br JPharmacol. 1991;103:1203–1205.

32. Pieper GM. Review of alterations in endothelial nitric oxide production indiabetes. Hypertension. 1998;31:1047–1060.

33. De Vriese AS, Verbeuren TJ, Van de Voorde J, Lameire NH, Vanhoutte PM.Endothelial dysfunction in diabetes. Br J Pharmacol. 2000;130:963–974.

34. Jin J, Dasari R, Sistare FD, Kunapuli S. Distribution of P2Y receptorsubtypes on haematopoietic cells. Br J Pharmacol. 1998;123:789–794.

35. Kunapuli SP, Daniel JL. P2 receptor subtypes in the cardiovascularsystem. Biochem J. 1998;336:513–523.

36. Vassort G. Adenosine 5�-triphosphate: a P2-purinergic agonist in themyocardium. Physiol Rev. 2001;81:767–806.

37. Lustig KD, Shiau AK, Brake AJ, Julius D. Expression cloning of an ATPreceptor from mouse neuroblastoma cells. Proc Natl Acad Sci U S A.1993;90:5113–5117.

38. Paterson AR, Kolassa N, Cass CE. Transport of nucleoside drugs inanimal cells. Pharmacol Ther. 1981;12:515–536.

39. Hyde RJ, Cass CE, Young JD, Baldwin SA. The ENT family of eukaryotenucleoside and nucleobase transporters: recent advances in the investi-gation of structure/function relationships and the identification of novelisoforms. Mol Membr Biol. 2001;18:53–63.

40. Pennycooke M, Chaudary N, Shuralyova I, Zhang Y, Coe IR. Differentialexpression of human nucleoside transporters in normal and tumor tissue.Biochem Biophys Res Commun. 2001;280:951–959.

41. Anderson CM, Xiong W, Geiger JD, Young JD, Cass CE, Baldwin SA,Parkinson FE. Distribution of equilibrative, nitrobenzylthioinosine-sensitive nucleoside transporters (ENT1) in brain. J Neurochem. 1999;73:867–873.

42. Griffiths M, Beaumont N, Yao SY, Sundaram M, Boumah CE, Davies A,Kwong FY, Coe I, Cass CE, Young JD, Baldwin SA. Cloning of a humannucleoside transporter implicated in the cellular uptake of adenosine andchemotherapeutic drugs. Nat Med. 1997;3:89–93.

43. Sobrevia L, Yudilevich DL, Mann GE. Activation of A2-purinoceptors byadenosine stimulates L-arginine transport (system y�) and nitric oxidesynthesis in human fetal endothelial cells. J Physiol (Lond). 1997;499:135–140.

44. Sexl V, Mancusi G, Höller C, Gloria-Maercker E, Schutz W, FreissmuthM. Stimulation of the mitogen-activated protein kinase via the A2a-aden-osine receptor in primary human endothelial cells. J Biol Chem. 1997;272:5792–5799.

45. Hayoz D, Ziegler T, Brunner HR, Ruiz J. Diabetes mellitus and vascularlesions. Metabolism. 1998;47:16–19.

Parodi et al Inhibition of Adenosine Transport by Glucose 577