insulin-stimulated l-arginine transport requires slc7a1 gene expression and is associated with human...

Insulin-Stimulated L-ArginineTransport Requires SLC7A1 GeneExpression and Is AssociatedWith Human Umbilical VeinRelaxationMARCELO GONZALEZ,1,2 VICTORIA GALLARDO,2 NATALIA RODRIGUEZ,2

CARLOS SALOMON,1 FRANCISCO WESTERMEIER,1 ENRIQUE GUZMAN-GUTIERREZ,1

FERNANDO ABARZUA,1,3 ANDREA LEIVA,1 PAOLA CASANELLO,1 AND LUIS SOBREVIA1*1Cellular and Molecular Physiology Laboratory (CMPL) and Perinatology Research Laboratory (PRL), Division of Obstetrics

and Gynecology, School of Medicine, Faculty of Medicine, Pontificia Universidad Catolica de Chile, Santiago, Chile2Vascular Physiology Laboratory, Department of Physiology, Faculty of Biological Sciences, Universidad de Concepcion,

Concepcion, Chile3Division of Obstetrics and Gynecology, School of Medicine, Faculty of Medicine, Pontificia Universidad Catolica de Chile,

Santiago, Chile

Insulin causes endothelium-derived nitric oxide (NO)-dependent vascular relaxation, and increases L-arginine transport via cationic aminoacid transporter 1 (hCAT-1) and endothelial NO synthase (eNOS) expression and activity in human umbilical vein endothelium (HUVEC).We studied insulin effect on SLC7A1 gene (hCAT-1) expression and hCAT-transport activity role in insulin-modulated human fetalvascular reactivity. HUVECwere used for L-arginine transport and L-[3H]citrulline formation (NOS activity) assays in absence or presenceof N-ethylmaleimide (NEM) or L-lysine (L-arginine transport inhibitors). hCAT-1 protein abundance was estimated by Western blot,mRNA quantification by real time PCR, and SLC7A1 promoter activity by Luciferase activity (�1,606 and �650 bp promoter fragmentsfromATG). Specific protein 1 (Sp1), and total or phosphorylated eNOSproteinwas determined byWestern blot. Sp1 activity (at four sitesbetween �177 and �105 bp from ATG) was assayed by chromatin immunoprecipitation (ChIP) and vascular reactivity in umbilical veinrings. Insulin increased hCATs–L-arginine transport, maximal transport capacity (Vmax/Km), and hCAT-1 expression. NEM and L-lysineblocked L-arginine transport. In addition, it was trans-stimulated (�7.8-fold) by L-lysine in absence of insulin, but unaltered (�1.4-fold) inpresence of insulin. Sp1 nuclear protein abundance and binding to DNA, and SLC7A1 promoter activity was increased by insulin. Insulinincreased NO synthesis and caused endothelium-dependent vessel relaxation and reduced U46619-induced contraction, effects blockedby NEM and L-lysine, and dependent on extracellular L-arginine. We suggest that insulin induces human umbilical vein relaxation byincreasing HUVEC L-arginine transport via hCATs (likely hCAT-1) most likely requiring Sp1-activated SLC7A1 expression.J. Cell. Physiol. 226: 2916–2924, 2011. � 2011 Wiley-Liss, Inc.

The reduced ability of the endothelium to cause nitric oxide(NO)-mediated vasodilatation reflects endothelial dysfunction,a phenomenon well-correlated with increased incidence ofcardiovascular diseases exhibited by patients with hypertension(Wierzbicki et al., 2004) or diabetes mellitus (Casanello et al.,2007; Hiden et al., 2009; Sobrevia et al., 2011), and in humanumbilical vein endothelium (HUVEC) isolated frompregnancieswith pre-eclampsia, gestational diabetes or intrauterine growthrestriction (IUGR) (Casanello et al., 2007, 2009; Sobrevia andGonzalez, 2009; Westermeier et al., 2009a,b; Sobrevia et al.,2011). It is known that insulin increases leg blood flow in healthysubjects via stimulation of endothelial NO synthase (eNOS)(Steinberg and Baron, 2002). Even when this phenomenon hasbeen shown altered in type II diabetes mellitus (Steinberg andBaron, 2002) or obese (Laakso et al., 1990) patients, associatedcell-signaling mechanisms behind this effect of insulin are notreported in human endothelium (Casanello et al., 2007;Sobrevia and Gonzalez, 2009; Sobrevia et al., 2011).

Insulin also increases NO synthesis and release in primarycultures of HUVEC (Scherrer et al., 1994; Gonzalez et al.,2004). In several cell types, NO synthesis requires uptake of thesemi-essential, cationic amino acid L-arginine, a phenomenonthat involves activity of several cationic amino acid membrane

Additional Supporting Information may be found in the onlineversion of this article.

Contract grant sponsor: FondoNacional de Desarrollo Cientıfico yTecnologico (FONDECYT), Chile;Contract grant numbers: 1070865, 1080534, 11100192.Contract grant sponsor: Programa de Investigacion Asociativa(PIA) from Comision Nacional de Investigacion en Ciencia yTecnologıa (CONICYT) (Anillos);Contract grant number: ACT-73.Contract grant sponsor: Direccion de Investigacion Universidad deConcepcion (DIUC);Contract grant number: 210.033.103-1.0.Contract grant sponsor: Apoyo Realizacion de Tesis Doctoral fromCONICYT;Contract grant numbers: AT-23070213, AT-24100210.

*Correspondence to: Luis Sobrevia, Cellular and MolecularPhysiology Laboratory (CMPL), Division of Obstetrics andGynecology, School of Medicine, Pontificia Universidad Catolica deChile, P.O. Box 114-D, Santiago, Chile.E-mail: [email protected]

Received 22 November 2010; Accepted 3 January 2011

Published online in Wiley Online Library(wileyonlinelibrary.com), 1 February 2011.DOI: 10.1002/jcp.22635

ORIGINAL RESEARCH ARTICLE 2916J o u r n a l o fJ o u r n a l o f

CellularPhysiologyCellularPhysiology

� 2 0 1 1 W I L E Y - L I S S , I N C .

transport systems (Sobrevia and Gonzalez, 2009). L-Arginineuptake ismainlymediated by theNaþ-independent, high-affinitycationic amino acid transporters 1 (hCAT-1) and 2 (hCAT-2),and to a lesser extent via the very high affinity transportsystem yþL in HUVEC (Casanello et al., 2007; Sobrevia andGonzalez, 2009). Insulin also increases L-arginine transport viahCAT-1 in this cell type (Sobrevia et al., 1996; Gonzalez et al.,2004), a phenomenon proposed to be, at least in part,responsible of the increased NO synthesis in response to thishormone (Casanello et al., 2007). However, even when thesechanges in the endothelial L-arginine membrane transport inresponse to insulin are well characterized, potentialmechanisms at a transcriptional level in this phenomenon areunknown (Mann et al., 2003; Casanello et al., 2007; Sobrevia andGonzalez, 2009; Sobrevia et al., 2011).

Biological effects of insulin involve activation of severaltranscription factors, including the transcription factor specificprotein 1 (Sp1) in several cell types (Samson and Wong, 2002;Solomon et al., 2008). Since SLC7A1 gene (encoding hCAT-1)promoter lacks of TATA box (Hammermann et al., 2001;Hatzoglou et al., 2004; Sobrevia and Gonzalez, 2009), wehypothesize that insulin-increased L-arginine transport andhCAT-1 expression result from higher SLC7A1 expressioninvolving Sp1 activation in HUVEC. Our results suggest thatinsulin increased L-arginine transport mediated by hCAT-1,most likely results from increased SLC7A1 expression, whichcould require Sp1 activation in this cell type. In addition, insulincauses endothelium-dependent vasorelaxation in umbilicalveins via a phenomenon that seems to involve hCAT-liketransport activity. These results could be important for a betterunderstanding of the reported endothelial dysfunction indiseases or pregnancy where L-arginine/NO pathway in fetalendothelium is reduced, such as pre-eclampsia, IUGR or fetalhypoxia (Casanello et al., 2007).

Materials and MethodsStudy participants

Written informed consents from pregnant women frommaternityof the Hospital Clınico Universidad Catolica, Santiago, Chile, wereobtained. Investigation follows the World Medical AssociationDeclaration of Helsinki, and approved protocols by the ethicscommittee of Faculty of Medicine of the Pontificia UniversidadCatolica de Chile. An expanded Materials and Methods Section isavailable in the Supplementary Online Material.

Cell culture

Endothelial cells were isolated by collagenase [0.25mg/ml;Collagenase Type II from Clostridium histolyticum (Boehringer,Mannheim, FRG)] digestion from umbilical cord veins (HUVEC)and cultured up to confluence (Gonzalez et al., 2004; Casanelloet al., 2009). Experiments were performed in absence or presence(2–8 h) of insulin (0.001–100 nM).

L-Arginine transport

L-Arginine transport was measured in absence or presence (30minpreincubation) of 200mM N-ethylmaleimide (NEM, CATsinhibitor) (Deves et al., 1993) or 1mM L-lysine (competitive cis-inhibitor of L-arginine transport) (Deves and Boyd, 1998; Mannet al., 2003), as described (Gonzalez et al., 2004; Casanello et al.,2009). The relative contribution of insulin to saturable L-argininetransport kinetic parameters [maximal velocity (Vmax) andapparent Km] was estimated from the maximal transport capacity(Vmax/Km) values for L-arginine transport by:

1C=InsF

¼CKm

InsVmax

CVmaxInsKm

where CVmax and CKm are the kinetics parameters for L-arginine transport in control conditions, and InsVmax and

InsKm

are kinetics parameters of L-arginine transport in HUVECexposed to insulin. In trans-stimulation experiments (Gonzalezet al., 2004; Casanello et al., 2009), initial rates of L-argininetransport were increased (P< 0.05, n¼ 4) by 7.8� 0.3-fold byL-lysine in control cells, but it was not significantly (P> 0.05)altered in the presence of insulin (1.4� 0.5-fold).

L-[3H]Citrulline assay

NOS activity was determined by incubation of cells with 100mM L-arginine and 4mCi/ml L-[3H]arginine (30min, 378C) in absence orpresence of 100mMNG-nitro-L-arginine methyl ester (L-NAME).Cells were exposed to 1 nM insulin for 8 h in absence or presenceof 200mMNEM or 1mM L-lysine. Digested cells (95% formic acid)were passed through a cation ion-exchange resin Dowex-50W(50X8-200) and L-[3H]citrulline determined in H2O eluate asdescribed (Gonzalez et al., 2004).

Sub-cellular fractionation

Cells scraped into ice-cold phosphate buffer saline (PBS)supplemented with phenylmethylsulfonyl fluoride (1M) werecentrifuged (5,400 rpm) and the pellet resuspended in cold low-sodium buffer, as described (Dignam et al., 1983). Mixturewas thenpassed through a 25-gauge syringe and centrifuged (14,000 rpm) tocollect the supernatant (cytoplasmic fraction). Pellet wasresuspended in cold low-potassium buffer, passed through a 1mlmicropipette tip, incubated (48C, 20min) and centrifuged(14,000 rpm). Supernatant (nuclear fraction) was stored until use(Dignam et al., 1983).

For plasmamembrane fraction preparation, confluent cellswerewashed twice with ice-cold PBS (mM: 130 NaCl, 2.7 KCl,0.8 Na2HPO4, 1.4 KH2PO4, pH7.4) and harvested in 3-[N-morpholino]propanesulfonic acid–KOH (MOPS)–KOH buffer(20mMMOPS–KOH, 250mM sucrose, pH 7.4). Cells were quicklydisrupted in liquid nitrogen and sonicated (two cycles, 15 sec,100W, 48C). Cytosol and plasma membrane fractions wereseparated by differential centrifugation as reported (Aguayo et al.,2005).

Western blotting

Proteins (70mg) separated by polyacrylamide gel (8–10%)electrophoresis were probed with primary polyclonal rabbit anti-hCAT-1 (1:250), anti-Sp1 (1:500), anti-eNOS (1:1,500), anti-phosphorylated eNOS at serine1177 (P�eNOS, 1:250), ormonoclonal mouse anti-b-actin (1:2,000) (Santa CruzBiotechnology, Santa Cruz, CA) antibodies (Gonzalez et al., 2004;Casanello et al., 2009).

Isolation of total RNA and reverse transcription

Total RNA was isolated using the Quiagen RNAeasy kit (Quiagen,Crawley, UK). RNA quality and integrity were insured by gelvisualization and spectrophotometric analysis (OD260/280),quantified at 260 nm and precipitated to obtain 4mg/ml. Aliquots(1mg) of total RNA were reversed transcribed into cDNA asdescribed (Gonzalez et al., 2004; Casanello et al., 2009).

Quantitative RT-PCR

Experiments were performed using a LightCyclerTM rapid thermalcycler (Roche Diagnostics, Lewes, UK) in a reactionmix containing0.5mM primers, and dNTPs, Taq DNA polymerase and reactionbuffer provided in the QuantiTect SYBR Green PCR Master Mix(Quiagen) as described (Gonzalez et al., 2004; Casanello et al.,2009). HotStart Taq DNA polymerase was activated (15min,958C), and assays included a 958C denaturation (15 sec), annealing(20 sec) at 548C (hCAT-1 and 28S), and extension (10 sec) at 728C(hCAT-1 and 28S). Product melting temperature values were79.18C (hCAT-1) and 86.78C (28S). Oligonucleotide primers:

JOURNAL OF CELLULAR PHYSIOLOGY

I N S U L I N - I N C R E A S E D h C A T - 1 E X P R E S S I O N 2917

hCAT-1 (sense) 50-GAGTTAGATCCAGCAGACCA-30, hCAT-1(anti-sense) 50-TGTTCACAATTAGCCCAGAG-30, 28S (sense)50-TTGAAAATCCGGGGGAGAG-30, 28S (anti-sense) 50-ACATTGTTCCAACATGCCAG-30. The number of copies for28S rRNA was not significantly altered (P> 0.05, n¼ 4) in allexperimental conditions used in this study (not shown).

Actinomycin D effect on hCAT-1 mRNA and protein

Total RNA and protein were isolated from HUVEC exposed (8 h)to culture medium without or with 1.5mM actinomycin D(transcription inhibitor) (Puebla et al., 2008) in absence orpresence of 1 nM insulin. hCAT-1 and 28S mRNA were quantifiedby real timeRT-PCR, and hCAT-1 andb-actin proteins detected byWestern blot (see above).

hCAT1 promoter cloning

The upstream sequences �1,606 and �650 bp from thetranscription start codon of SLC7A1 gene (GenBank: AL596114)were PCR-amplified and cloned into pGL3-basic reporter system(Farıas et al., 2010). The pGL3-hCAT1 reporter constructsgenerated were pGL3-hCAT1�1,606 and pGL3-hCAT1�650.

Transient transfection

Cell suspension (3.2� 106 cells/ml) was mixed with pGL3-hCAT1reporter constructs, pGL3-Basic (empty pGL3 vector), pGL3-Control [Simian Virus 40 promoter (SV40) pGL3 vector], andinternal transfection control vector pRL-TK expressing Renillaluciferase (Puebla et al., 2008; Farıas et al., 2010). Cells wereelectroporated and transfection efficiency estimated with pEGFP-N3 vector (Clontech, Mountain View, CA) as described (Pueblaet al., 2008; Farıas et al., 2010).

Luciferase assay

Electroporated cells were lysed in 200ml passive lysis buffer(Promega), and Firefly and Renilla luciferase activity was measuredusing Dual-Luciferase1 Reporter Assay System (Promega) in aSirius luminometer (Berthold Detection System; Oak Ridge, TN)as described (Puebla et al., 2008; Farias et al., 2010).

Chromatin immunoprecipitation (ChIP) assay

ChIP assay for Sp1 response elements identified in silico in the regionbetween�177 and�105 bp fromATGof the SLC7A1 promoterwasperformed as described (Puebla et al., 2008; Farıas et al., 2010). Inbrief, fixedcells (1%paraformaldehyde)were incubatedwith125mMglycine, rinsed with ice-cold PBS, scraped and centrifuged to obtain apellet which was resuspended in cell lysis buffer and incubated(20min, 48C) as described (Puebla et al., 2008; Farias et al., 2010). Themix was then centrifuged and the obtained pellet resuspended innuclear lysis buffer containing protease inhibitors, incubated (20min,48C) and sonicated (20�, 10-sec pulse). Supernatant (nuclearfraction) was collected after centrifugation of sonicated samples(100ml aliquots of this fraction were saved as input).

Nuclear fractions (6mg chromatin) were incubatedwith proteinG-agarose beads, incubated with 20ml of salmon sperm DNA(ssDNA) solution (10mg/ml) and centrifuged. Supernatant (pre-cleared nuclear fractions) aliquots were diluted to 1ml in ChIPdilution buffer and incubated with anti-Sp1 (2mg/ml) antibody.Aliquots were transferred into a tube containing 20ml proteinG-agarose beads and incubatedwith BSA (1%) followed by a secondincubation with 20ml of ssDNA solution (10mg/ml). The mix wasfurther incubated for 4 h (DNA samples without antibody wereused as G-agarose drag-non-specific control). Theimmunocomplex (protein G-agarose beads, anti-Sp1 antibody, Sp1protein, and chromatin) was centrifuged, and pellets consecutivelyrinsed with ChIP dilution buffer, dialysis buffer, TSE-500 buffer, andLiCl detergent buffer, followed by two further rinseswith TE buffer

(Puebla et al., 2008; Farıas et al., 2010). Pellets were incubated withelution buffer and chromatin eluted from the immunocomplex bycentrifugation (supernatant fraction). Eluted (precipitated Sp1–DNA complex) and input (total sonicated DNA) DNA wasobtained by reversing the initial paraformaldehyde reaction byincubation in 330mM NaCl followed by precipitation withisopropanol/glycogen. Resulting pellet was incubated in proteinaseK buffer and the DNA purified using phenol/chloroform (1:1, v/v),precipitated with isopropanol/glycogen and then treated withRNAse A (10mg/ml). Aliquots (1ml, 1:10 dilution) of DNA (inputand precipitate) were used for PCR assays. Cycling parameters foramplifications (30 cycles)were 958C for 30 sec, 578C for 30 sec and728C for 30 sec. Oligonucleotide primers: Sp1-hCAT-1 promoter-sense 50-ATCCACCCCTCCAATCTTCT-30 and Sp1-hCAT-1promoter-anti-sense 50-GCAACCCAGATTCCAGTCTC-30(product size 249 bp).

Umbilical vein reactivity

Ring segments (2–4mm length) from the human umbilical veinswith and without endothelium were mounted in a myograph forisometric force measurements with optimal diameter adjustedfrom maximal active response to 62.5mM KCl as described (Vegaet al., 2009). Response to insulin (0.001–100 nM, 8 h) wasdetermined in 100 nM U466619-preconstricted vessels in absenceor presence of 200mM NEM or 1mM L-lysine. Vessel rings werealso incubated (378C) in Krebs for 2 or 8 h in absence or presenceof 1 nM insulin, followed by exposure to U46619 (0.1–1,000 nM).Additional assays were performed in preparations incubated inKrebs with or without 200mM L-arginine.

Statistical analysis

Values are mean� SEM, where n indicates number of different cellcultures (three to four replicates). Comparisons between two andmore groups were performed by means of Student’s unpaired t-test and analysis of variance (ANOVA), respectively. If theANOVAdemonstrated a significant interaction between variables, post hocanalyses were performed by the multiple-comparison Bonferronicorrection test. P< 0.05 was considered statistically significant.

ResultsEffect of insulin on L-arginine transport

Uptake of 100mM L-arginine was significantly increased after 4and 8 h of exposure to insulin (Fig. 1A), an effect blocked byNEM (94� 4%) confirming our previous results in this cell type(Gonzalez et al., 2004). We here extended these observationsshowing that insulin effect was concentration-dependent withsimilar (P> 0.05) half-maximal stimulatory concentrations at4 h (SC50¼ 0.24� 0.03 nM) and 8 h (SC50¼ 0.29� 0.04 nM)of incubation (Fig. 1B). Insulin also increased the Vmax, withno significant changes in the apparent Km of transport(Fig. 2A, Table 1), in a concentration dependent manner(SC50¼ 0.21� 0.06 nM; Fig. 2B). Eadie-Hofstee transformationof transport data, an approach that allows estimating potentialinvolvement of more than one transport mechanism withdifferent affinities for a common substrate, or one transportmechanism that could experience changes in its substrateaffinity, shows lineal regressions in all experimental conditions(Fig. 2C). L-Arginine (100mM) transport was also inhibited by L-lysine in a same proportion as NEM, in absence or presence ofinsulin (Fig. 2D).

Effect of insulin on L-[3H]citrulline formation

The fractions of L-[3H]citrulline formation inhibited by L-NAMEand eNOS phosphorylation at Ser1177 were increased by insulin(Fig. 2E), confirming previous results in this cell type (Gonzalezet al., 2004). Basal and insulin-increased L-[3H]citrulline


2918 G O N Z A L E Z E T A L .

formation, but not eNOS phosphorylation was blocked byNEM or L-lysine (Fig. 2F).

hCAT-1 expression

Insulin increased hCAT-1 protein abundance in aconcentration- (SC50¼ 0.37� 0.05 nM; Fig. 3A) and time- [half-maximal stimulatory time (ST50)¼ 3.13� 0.5 h] dependentmanner (Fig. 3B). Insulin (1 nM)-induced increase of hCAT-1protein abundance (2.8� 0.4-fold) reached a maximal effectafter 4 h incubation, which was maintained for further 4 h(3.2� 0.1-fold). Incubation of cells with 1 nM insulin for 8 hincreased hCAT-1 protein abundance in plasmamembrane, butdid not alter its abundance in cytosol fraction (Fig. 3C). Parallelexperiments confirmed previous results (Gonzalez et al., 2004)showing that insulin increased the mRNA number of copies forhCAT-1 (ST50¼ 2.7� 0.3 h), reaching a maximal effect at 4 h(3.3� 0.3-fold), declining by 8 h of incubation (2.2� 0.2-fold)(not shown).

Actinomycin D effect on hCAT-1 protein and mRNAexpression

Incubation of cells with actinomycin D inhibited in a similarproportion (P< 0.05) the increase induced by insulin onL-arginine uptake (91� 2%; Fig. 4A), hCAT-1 proteinabundance (96� 2%; Fig. 4B) and number of mRNA copies(79� 7%; Fig. 4C).However, actinomycinDdid not significantlyalter these parameters in absence of insulin.

SLC7A1 transcriptional activity

Exposure of cells to 1 nM insulin for 8 h showed an increase inthe SLC7A1 promoter transcriptional activity when transfectedwith pGL3-hCAT1�1,606 (1.3� 0.09-fold) and pGL3-hCAT1�650 (1.4� 0.03-fold) constructs, respectively,compared with cells incubated in absence of insulin (Fig. 5).

Sp1 protein abundance and activity

Exposure of cells to insulin increased nuclear Sp1 proteinabundance compared with total Sp1 protein abundance(Fig. 6A). Insulin also increased Sp1 binding to DNA in cellstransfected with the pGL3-hCAT1�650 construct of SLC7A1promoter (Fig. 6B).

Umbilical vein rings reactivity

Insulin induced relaxation of endothelium-intact, but not inendothelium-denuded human umbilical vein rings with a half-maximal effect (EC50)¼ 1.8� 0.2 nM insulin (Fig. 7A). Wheninsulin (1 nM) effect was assayed in preparations preincubatedwith NEM and L-lysine, insulin-induced relaxation wasabolished; however, these molecules did not significantlyalter KCl-induced vasoconstriction (Fig. 7B). Parallelexperiments show that U46619 induces vasoconstriction(EC50¼ 8.4� 0.2 nM), an effect not significantly (P> 0.05)altered following 2 h of preincubation with insulin(EC50¼ 11.2� 0.3 nM), but significantly reduced by thishormone after 8 h of preincubation (EC50¼ 20.1� 0.4 nM,P< 0.05; Fig. 7C). In addition, when vessel rings were incubatedin a L-arginine free solution, insulin did not inducevasorelaxation neither altered U46619-induced contraction(Fig. 7D).

Discussion

This study shows that insulin increases L-arginine transportvia a mechanism involving higher maximal transport capacityassociated with increased hCAT-1 expression in HUVEC.Insulin-increased L-arginine transport results from activation ofSLC7A1 promoter via amechanism thatmay involve Sp1 bindingto multiple consensus sequences between �177 and �105 bpfrom ATG. In addition, insulin also induces relaxation of humanumbilical veins that depends on endothelium and hCATs-liketransport activity. Since insulin effects are seen at a physiologicalplasma concentration, it is conceivable that SLC7A1 expressionand hCAT-1 activity are under tonic regulation by physiologicalplasma insulin in HUVEC. These findings could be determinantfor fetal insulinmodulation of endothelial-derivedNOsynthesisin human umbilical vessels from pregnancy diseases associatedwith hyperinsulinemia, such as gestational diabetes, and otherstates of insulin resistance (Sobrevia and Gonzalez, 2009;Westermeier et al., 2009a,b; Farıas et al., 2010; Sobrevia et al.,2011).

Insulin effect on overall and saturable L-arginine transportwas dependent on the concentration of this hormone leading tohigher Vmax, without significant changes in the apparent Km ofsaturable L-arginine transport. Since basal insulin concentrationin human umbilical vein blood at term from healthy pregnancieshas been reported as 0.02–0.03 nM (Westgate et al., 2006;

Fig. 1. Effect of insulin on L-arginine transport. A: Overall L-argininetransport (100mM L-arginine, 2mCi/ml L-[3H]arginine, 1min, 37-C)was measured in HUVEC monolayers in Krebs solution in absence(Control) or presence of the indicated concentrations of insulin fordifferent time periods. Experiments were also performed in thepresence of 200mM N-ethylmaleimide (NEM) in control (*), 0.1 nM(&) or 1 nM(~) insulin. B: L-Arginine transport as in (A), in absenceorpresence of the indicated concentrations of insulin for differentperiods of time. Experiments were also performed in the presence of200mM NEM for 8 h (*), 4 h (~) or 2h (&) of exposure of cells toinsulin. Values aremeansWSEM (nU 9). In (A), MP<0.04 and yP<0.05versus corresponding value in absence of insulin. In (B), MP<0.04versus without insulin.



Lindsay et al., 2007), it is feasible that basal fetal plasma insulinlevel will be tonically activating L-arginine transport in theumbilical vein endothelium. Interestingly, the latter could beimportant in gestational diabetes, since the half-maximalstimulatory effect of insulin on L-arginine transport was reachedat�0.2 nM inHUVEC, a concentration close to values reportedfor umbilical vein blood in newborns from pregnancies affectedby this syndrome (�0.07 nM) (Westgate et al., 2006; Lindsayet al., 2007). Our results show that HUVEC exposed to 0.1 nMinsulin exhibit �3-fold increase of the L-arginine transportcapacity (i.e., Vmax/Km), a finding 2-fold higher than in humantrophoblast-derived cells (Eaton and Sooranna, 1998),suggesting differential modulation of L-arginine transport byinsulin in the human feto-placenta cell types. Since physiological

concentrations of insulin (0.1–0.5 nM) selectively activateinsulin receptors (IRs), instead of insulin-like growth factor 1(IGF-I) and hybrid IR/IGF-I receptors in the endothelium(Li et al., 2005), it is likely that insulin effect was mainly dueto activation of insulin receptors in HUVEC. However, sinceHUVEC expresses at least two isoforms of insulin receptors(Westermeier et al., 2009a,b), differing by the absence (insulinreceptor A, IR-A) or presence (insulin receptor B, IR-B) of a 12amino acid sequence in the COOH-terminus of the a-subunit(Hiden et al., 2009), we cannot assign exclusive or preferentialinvolvement of IR-A or IR-B in the response to insulin.

Insulin causes vasorelaxation in the human vasculature via amechanism requiring endothelium-derived NO (Sobrevia andGonzalez, 2009), but nothing is known regarding insulin effect

Fig. 2. Insulin modulation of kinetic parameters for saturable L-arginine transport. A: Saturable L-arginine transport (0–1,000mM L-arginine,2mCi/ml L-[3H]arginine, 1min, 37-C) in HUVEC monolayers in Krebs solution in absence (Control) or presence (8 h) of the indicatedconcentrations of insulin.B:Maximal velocity (Vmax) values for L-arginine transport calculated fromdata inA(seeMaterials andMethodsSection)against insulin concentration. C: Eadie-Hofstee plot of data in (A). D: Cells were cultured in absence (Control) or presence of insulin (8 h),and L-arginine (100mM) uptake was measured after preincubation (last 30min of insulin incubation period) without (�) or with (R) 200mMN-ethylmaleimide (NEM) and/or 1mM L-lysine. E: Cells as in (D) were used to measure L-[3H]citrulline formation from L-[3H]arginine (100mML-arginine, 4mCi/ml L-[3H]arginine, 30min, 37-C) in absence or presence of 100mM NG-nitro-L-arginine methyl ester (L-NAME). Plotted datacorrespondtotheL-NAMEinhibitablefractionofL-[3H]citrulline formation(seeMaterialsandMethodsSection).F:Westernblots(representativeofotherfivedifferentcell cultures) for total (eNOS)orSer1177-phosphorylatedeNOS(P�eNOS) inwholecell extracts fromcells inabsence (�)orpresence(R)of insulin(1 nM,8 h),200mMNEMor1mML-lysine. In(B),MP<0.05versusvalues inabsenceof insulin. In(D)and(E),MP<0.04versusallother values. Values are meansWSEM (nU 5–12).



on human umbilical vein reactivity and its dependence onL-arginine transport (Muniyappa et al., 2007; Sobrevia andGonzalez, 2009; Sobrevia et al., 2011). Our results show thatinsulin causes endothelium-dependent vasorelaxation of humanumbilical vessel rings, which could depend on hCATs-liketransport activity since NEM and L-lysine, CATs-like transportinhibitors (Deves et al., 1993; Deves and Boyd, 1998; Mann

et al., 2003), inhibited L-arginine transport in a similarproportion (�94%) in HUVEC and reversed insulinvasodilatory effect.NEMand L-lysine also inhibitedNOS activityin HUVEC in absence or presence of insulin (most likely notinvolving Ser1177-dephosphorylation of eNOS), which couldresult from reduced disposal of extracellular L-arginine(Moncada and Higgs, 2006; Casanello et al., 2009; Sobrevia and

TABLE 1. Effect of insulin on L-arginine transport kinetic parameters in HUVEC

Vmax (pmol/mg protein/min) Km (mM) Vmax/Km (pmol/mg protein/min/mM) 1/C/InsF

Control 2.4� 0.3 103� 39 0.023� 0.005 —Insulin (nM)

0.01 5.0� 0.4� 147� 34 0.034� 0.005 1.5� 0.30.1 6.3� 0.6� 182� 28 0.035� 0.017� 1.5� 0.4�

1 9.2� 1.0� 204� 64 0.045� 0.009� 2.0� 0.4�

10 9.9� 0.9� 165� 48 0.060� 0.011� 2.6� 0.6�

Kinetics of saturable L-arginine transport (0–1,000mM L-arginine, 2mCi/ml L-[3H]arginine, 1min, 378C) was measured in primary cultures of HUVEC exposed (8 h) to Krebs solution without(Control) or with insulin (see Materials and Methods Section). 1/C/InsF is the relative contribution of insulin (Ins) exposure compared with absence of this hormone (C) to changes in maximaltransport capacity (Vmax/Km) of L-arginine transport (see Materials and Methods Section). Values are means� SEM (n¼ 10–12).�P< 0.05 versus Control.

Fig. 3. Effect of insulin onhCAT-1 protein abundance.A:Western blot (representative of other eight different cell cultures) for hCAT-1proteinabundanceandb-actin (internal reference) inHUVECmonolayers inabsence(0 h,control)orpresence (8h)of indicatedconcentrationsof insulin.Lower part: hCAT-1/b-actin protein ratio densitometries normalized to 1 in control. B:Western blot for hCAT-1 protein abundance andb-actinin absence (0, control) or presence of 1 nM insulin for the indicated time periods. Lower part: hCAT-1/b-actin protein ratio densitometriesnormalized to1 in control.C:Western blot for hCAT-1,NaR,KR-ATPaseb-subunit (ATPase) andb-actin protein abundance at the cytosol (C) orplasmamembrane (PM) fractions fromHUVECcultured in absence (�, control) or presence (R) of 1 nM insulin (8 h). Lowerpart: hCAT-1/b-actinprotein ratio densitometries normalized to 1 in control. D: hCAT-1/ATPase protein ratio densitometries normalized to 1 in plasmamembranefractions in control. Values aremeansWSEM (nU 8–10). In (A), MP<0.05 versus control and 0.1 nM insulin. In (B), MP<0.05 versus control and 2h.In (C), MP<0.05 versus control and insulin in cytosol fraction; yP<0.05 versus all other values. In (D), MP<0.05 versus control.



Gonzalez, 2009; Sobrevia et al., 2011). Furthermore, NEM andL-lysine blocked insulin-induced endothelium-dependentrelaxation in umbilical vein rings, suggesting that endothelial L-arginine transport could limit NO synthesis in response to thishormone. This is supported by the findings showing that insulin-induced relaxation and insulin inhibition of U46619-inducedcontraction are lost in absence of extracellular L-arginine. Thus,

changes in L-arginine transport are crucial in insulin-modulatedvascular tone in endothelial dysfunction-associated diseasessuch as hypertension (Yang and Kaye, 2009), intrauterinegrowth restriction (Casanello et al., 2007, 2009;Myatt, 2010) orpreeclampsia (Casanello et al., 2007; Myatt, 2010).

Insulin increased hCAT-1 mRNA after 2 h, but hCAT-1protein abundance and hCATs-like transport activity wasincreased after 4 h of exposure to this hormone, suggesting atemporal correlation in this phenomena. In addition, insulineffects are paralleled by higher SLC7A1 gene transcriptional

Fig. 4. Effect of actinomycin D on hCAT-1 expression. A: OverallL-arginine transport (100mM L-arginine, 2mCi/ml L-[3H]arginine,1min, 37-C) in HUVEC monolayers in Krebs solution in absence(�, control) or presence (R, 8 h) of 1 nM insulin and/or 1.5mMactinomycin D (see Materials and Methods Section). B:Western blot(representative of other eight different cell cultures) for hCAT-1protein abundance and b-actin (internal reference) in HUVECmonolayers as in (A). Lower part: hCAT-1/b-actin protein ratiodensitometries normalized to 1 in control. C: Ratio for hCAT-1mRNA/28S rRNA number of copies of hCAT-1mRNA and 28S rRNAin cells as in (A). Values are meansWSEM (nU 8–12). MP<0.05 versusall other values. yP<0.05 versus insulin.

Fig. 5. Effect of insulin in SLC7A1 (hCAT-1) promoter activity.Luciferase (Luc) reporter constructs containing serial truncations ofSLC7A1 promoter were transfected in primary cultures of HUVECincubated (8 h) in absence (&) or presence (&) of 1 nM insulin, alongwith Renilla reporter plasmid, and assayed for Firefly and Renillaluciferase activity, respectively. Results depict ratio of Firefly/Renillaluciferase activity in cells in absence or presence of insulin. Cells werealso transfected with the empty pGL3-basic vector or pGL3-controlvector (SV40 pGL3) as negative or positive controls, respectively.Values are meansWSEM (nU 8). MP<0.05 versus correspondingvalues in absence of insulin.

Fig. 6. Insulin effect on Sp1 nuclear protein abundance and activity.A:Western blot (representative of other eight different cell cultures)for nuclear and total Sp1 protein abundance (b-actin was internalloading control) in HUVEC in absence (Control) or presence (8 h)of insulin. Lower part: Sp1 nucleus/Sp1 total protein ratiodensitometries normalized to 1 in Control. B: Chromatinimmunoprecipitation assay for Sp1 binding to DNA in HUVEC. ThemRNA flanking Sp1 consensus sequences in SLC7A1 promoter region(�177 and �105bp from ATG) was amplified by RT-PCR (seeMaterials and Methods Section). Input is total DNA. Lower part: Sp1consensus sequences over Input ratio densitometries normalized to 1in Control. Values aremeansWSEM (nU 8). MP<0.05 versus Control.



activity. Thus, increased L-arginine transport may result fromincreased hCAT-1 expression in HUVEC. Since inhibition oftranscription caused a comparable reduction (�90%) inL-arginine transport, hCAT-1 protein abundance and mRNAexpression in this cell type, the possibility that L-argininetransport was higher due to insulin increased mRNA and/orprotein stability is unlikely, although this phenomenon cannotdefinitively be ruled out. In addition, since hCAT-1 proteinabundance was increased at the plasma membrane, but it wasunaltered at the cytosol fraction in HUVEC exposed to insulin,the feasibility of an internal pool of pre-existent hCAT-1translocated to the plasma membrane is unlikely. Insulin alsoincreases CAT-1 protein abundance in rat hepatocytes (Wuet al., 1994) and CAT-1 mRNA level in rat cardiomyocytes(Simmons et al., 1996), thus changes in CAT-1 expressionlimiting transport of cationic amino acids are not exclusive forHUVEC. Since Eadie-Hofstee transformation of saturabletransport data was lineal, without significant changes in theapparent Km of L-arginine transport, the possibility that insulin-increased transport was due to activation of additionalmembrane transport systems is unlikely. Our results also showthat insulin blocked U46619-induced human umbilical veinvasoconstriction only after 8 h of incubation, suggesting that atime-dependent mechanism is also associated with thereactivity of this tissue to insulin. Changes in hCAT-1 proteinabundance in HUVEC, paralleled by changes in L-argininetransport, requires 4 h of incubation with insulin, and NEMblocks insulin-induced vasodilatation in human umbilical veinrings. Thus, insulin effect requires increased expression and/or

availability of functional hCATs at the human umbilical veinendothelium.

Insulin also increases Sp1 nuclear protein abundance and itsbinding to a proximal region (�177 and �105 bp from ATG)of the SLC7A1 promoter containing at least four consensussequences for Sp1. Since increase in Sp1 protein abundancewashigher by �30% than its binding to DNA, increased SLC7A1transcriptional activity may not only result from higher nuclearSp1 protein abundance, but most likely increased activity of thistranscription factor in one or more of the identified consensussequences within this promoter region. This is furthersupported by findings showing that insulin-increasedtranscriptional activity of constructs generated for SLC7A1promoter region was similar to the increase in Sp1 binding.Interestingly, CT polymorphism in the 30-untranslated region(30-UTR) of SLC7A1, which contains Sp1 consensus sequences,in patients with essential hypertension has been reported,together with reduced SLC7A1 transcriptional activity due toreduced Sp1 activity (Yang and Kaye, 2009), supporting Sp1involvement in the response to insulin. However, since severalresponse elements to insulin have been identified in humanSLC7A1 promoter (Sobrevia and Gonzalez, 2009) as in othermammalian cells (Hatzoglou et al., 2004), activation of responseelements within the cloned promoter region may explain inpart, but not the full insulin-induced increase in hCAT-1transcript in HUVEC.

It has been reported that an enhancer element identified inthe first intron of SLC7A1may play a bifunctional role regulatingSLC7A1 transcriptional activity, that is, increasing SLC7A1

Fig. 7. Insulin effect on human umbilical vein rings reactivity. A: Endothelium-intact or endothelium-denuded human umbilical vein rings wereexposed to increasing concentrations of insulin. B: Endothelium-intact humanumbilical vein rings in absence (�, control) or presence (R) of 1 nMinsulin and/or 200mM N-ethylmaleimide (NEM) or 1mM L-lysine. C: Response of endothelium-intact human umbilical vein rings to U46619 inabsence(control)orpresenceof1 nMinsulinfortheindicatedtimeperiods.D:Endothelium-intacthumanumbilicalveinringsincubatedinmediumwith(R)orwithout(�)200mML-argininewereexposedto1nMinsulin(leftpart)and/or100nMU46619(rightpart).Relativeresponsesaregivenaspercentage fraction of the initial vessel response to KCl (see Materials andMethods Section). Values aremeanWSEM (nU 4–7). In (A), MP<0.05versus corresponding values in endothelium-intact vessels. In (B), MP<0.05 versus all other values. In (D), MP<0.03 and yP<0.05 versus all othercorresponding values.



transcriptional activity following binding of the purine-richelement binding protein A (Pur alpha) under basal conditionsand activating transcription factor 4 (ATF4) in endoplasmicreticulum stress (ERS), or decreasing SLC7A1 transcriptionalactivity by the C/EBP homologous protein 10 (CHOP) bindingin ERS in C6 rat glioma cells (Huang et al., 2009). Thus, changesin hCAT-1 protein not necessarily fit changes in SLC7A1expression, opening the possibility that a transcriptionalregulation of SLC7A1 and/or post-transcriptional regulation ofSLC7A1 transcript as well as for the protein product expressionmay occur. Thus, the findings of an insulin-increased transportactivity via hCAT-1 in HUVEC cannot exclude the possibilitythat insulin induces post-translational modifications of hCAT-1.

Our results suggest that hCAT-1 expression and activity areregulated by insulin in HUVEC, and suggest that insulin increasesexpression of SLC7A1 gene due to increased transcriptionalactivity most likely due to higher Sp1 activity. The fact that insulinalters vascular human umbilical vein reactivity in an endothelium-and L-arginine transport (possibly hCAT-1)-dependent manner,implies a physiological mechanism of relevance formodulation ofhCAT-1 expression and activity in this vessel type, particularly indiseases associated with altered L-arginine and insulinmetabolism such as hypertensionordiseasesof pregnancywherefetal vascular function is altered (Casanello et al., 2007; Sobreviaand Gonzalez, 2009;Westermeier et al., 2009a,b; Sobrevia et al.,2011).

Acknowledgments

Authors thank clinical staff from Hospital Clınico PontificiaUniversidadCatolica deChile laborward for supply of umbilicalcords. Supported by FondoNacional deDesarrollo Cientıfico yTecnologico (FONDECYT 1070865, 1080534, 11100192),Programa de Investigacion Asociativa (PIA) from ComisionNacional de Investigacion en Ciencia y Tecnologıa (CONICYT)(Anillos ACT-73) and Direccion de Investigacion Universidadde Concepcion (DIUC 210.033.103-1.0). Fellowship ApoyoRealizacion de Tesis Doctoral from CONICYT AT-23070213(MG) and AT-24100210 (FW), CONICYT-PhD (Chile)fellowships (MG, EG-G, FW) and Faculty ofMedicine, PUC-PhD(Chile) fellowship (CS).

Literature Cited

Aguayo C, Casado J, Gonzalez M, Pearson JD, Martın RS, Casanello P, Pastor-Anglada M,Sobrevia L. 2005. Equilibrative nucleoside transporter 2 is expressed in human umbilicalvein endothelium, but is not involved in the inhibition of adenosine transport induced byhyperglycaemia. Placenta 26:641–653.

Casanello P, Escudero C, Sobrevia L. 2007. Equilibrative nucleoside (ENTs) and cationicamino acid (CATs) transporters: Implications in foetal endothelial dysfunction in humanpregnancy diseases. Curr Vasc Pharmacol 5:69–84.

Casanello P, Krause B, Torres E, Gallardo V, Gonzalez M, Prieto C, Escudero C, Farias M,Sobrevia L. 2009. Reduced L-arginine transport and nitric oxide synthesis in humanumbilical vein endothelial cells from intrauterine growth restriction pregnancies is notfurther altered by hypoxia. Placenta 30:625–633.

Deves R, Boyd CA. 1998. Transporters for cationic amino acids in animal cells: Discovery,structure, and function. Physiol Rev 78:487–545.

Deves R, Angelo S, Chavez P. 1993. N-Ethylmaleimide discriminates between two lysinetransport systems in human erythrocytes. J Physiol 468:753–766.

Dignam J, Martin P, Shastry B, Roeder R. 1983. Eukaryotic gene transcription with purifiedcomponents. Methods Enzymol 101:582–598.

Eaton B, Sooranna S. 1998. In vitro modulation of L-arginine transport in trophoblast cells byglucose. Eur J Clin Invest 28:1006–1010.

Farıas M, Puebla C, Westermeier F, Jo M, Pastor-Anglada M, Casanello P, Sobrevia L. 2010.Nitric oxide reduces SLC29A1 promoter activity and adenosine transport involvingtranscription factor complex hCHOP-C/EBPa in human umbilical vein endothelial cellsfrom gestational diabetes. Cardiovasc Res 86:45–54.

Gonzalez M, Flores C, Pearson J, Casanello P, Sobrevia L. 2004. Cell signalling-mediatinginsulin increase of mRNA expression for cationic amino acid transporters-1 and -2 andmembrane hyperpolarization in human umbilical vein endothelial cells. Pflugers Arch448:383–394.

Hammermann R, BrunnG, Racke K. 2001. Analysis of the genomic organization of the humancationic amino acid transporters CAT-1, CAT-2 and CAT-4. Amino Acids 21:211–219.

Hatzoglou M, Fernandez J, Yaman I, Closs E. 2004. Regulation of cationic amino acidtransport: The story of the CAT-1 transporter. Annu Rev Nutr 24:377–399.

Hiden U, Glitzner E, Hartmann M, Desoye G. 2009. Insulin and the IGF system in the humanplacenta of normal and diabetic pregnancies. J Anat 215:60–68.

Huang C, Chiribau C, MajumderM, Chiang C,Wek R, Kelm R, Khalili K, Snider M, HatzoglouM. 2009. A bifunctional intronic element regulates the expression of the arginine/lysinetransporterCat-1 viamechanisms involving the purine-rich element binding proteinA (Puralpha). J Biol Chem 284:32312–32320.

Laakso M, Edelman S, Brechtel G, Baron A. 1990. Decreased effect of insulin to stimulateskeletal muscle blood flow in obese man. A novel mechanism for insulin resistance. J ClinInvest 85:1844–1852.

Li G, Barrett E, Wang H, Chai W, Liu Z. 2005. Insulin at physiological concentrationsselectively activates insulin but not insulin-like growth factor I (IGF-I) or insulin/IGF-I hybridreceptors in endothelial cells. Endocrinology 146:4690–4696.

Lindsay R,Westgate J, Beattie J, Pattison N, Gamble G, Mildenhall L, Breier B, Johnstone F.2007. Inverse changes in fetal insulin-like growth factor (IGF)-1 and IGF binding protein-1 in association with higher birth weight in maternal diabetes. Clin Endocrinol 66:322–328.

MannGE, YudilevichDL, Sobrevia L. 2003. Regulation of amino acid and glucose transportersin endothelial and smooth muscle cells. Physiol Rev 83:183–252.

Moncada S, Higgs EA. 2006. The discovery of nitric oxide and its role in vascular biology. Br JPharmacol 147:S193–S201.

Muniyappa R,MontagnaniM, KohK,QuonM. 2007. Cardiovascular actions of insulin. EndocrRev 28:463–491.

Myatt L. 2010. Reactive oxygen and nitrogen species and functional adaptationof the placenta.Placenta 31:S66–S69.

Puebla C, Farıas M, Gonzalez M, Vecchiola A, Aguayo C, Krause B, Casanello P, Sobrevia L.2008. High D-glucose reduces SLC29A1 promoter activity and adenosine transportinvolving specific protein 1 in human umbilical vein endothelium. J Cell Physiol 215:645–656.

Samson S, Wong N. 2002. Role of Sp1 in insulin regulation of gene expression. J MolEndocrinol 29:265–279.

Scherrer U, Randin D, Vollenweider P, Vollenweider L, Nicod P. 1994. Nitric oxide releaseaccounts for insulin’s vascular effects in humans. J Clin Invest 94:2511–2515.

Simmons W, Closs E, Cunningham J, Smith T, Kelly R. 1996. Cytokines and insulin inducecationic amino acid transporter (CAT) expression in cardiac myocytes. Regulation of L-arginine transport and no production by CAT-1, CAT-2A, and CAT-2B. J Biol Chem271:11694–11702.

Sobrevia L, Gonzalez M. 2009. A role for insulin on L-arginine transport in fetal endothelialdysfunction in hyperglycaemia. Curr Vasc Pharmacol 7:467–474.

Sobrevia L, Nadal A, Yudilevich D, Mann G. 1996. Activation of L-arginine transport (systemyþ) and nitric oxide synthase by elevated glucose and insulin in human endothelial cells. JPhysiol 490:775–781.

Sobrevia L, Abarzua F, Nien JK, Salomon C,Westermeier F, Puebla C, Cifuentes F, Guzman-Gutierrez E, Leiva A, Casanello P. 2011. Differential placental macrovascular andmicrovascular endothelial dysfunction in gestational diabetes. Placenta DOI: 10.1016/j.placenta.2010.12.011

Solomon S, Majumdar G, Martinez-Hernandez A, Raghow R. 2008. A critical role of Sp1transcription factor in regulating gene expression in response to insulin and otherhormones. Life Sci 83:305–312.

Steinberg H, Baron A. 2002. Vascular function, insulin resistance and fatty acids. Diabetologia45:623–634.

Vega J, Puebla C, Vasquez R, Farıas M, Alarcon J, Pastor-Anglada M, Krause B, Casanello P,Sobrevia L. 2009. TGF-b1 inhibits expression and activity of hENT1 in a nitric oxide-dependent manner in human umbilical vein endothelium. Cardiovasc Res 82:458–467.

Westermeier F, Puebla C, Vega J, Farias M, Escudero C, Casanello P, Sobrevia L. 2009a.Equilibrative nucleoside transporters in fetal endothelial dysfunction in diabetes mellitusand hyperglycaemia. Curr Vasc Pharmacol 7:435–449.

Westermeier F, SalomonC,CasanelloP, Sobrevia L. 2009b. Insulin receptor isoforms -11 (IR-A) and þ11 (IR-B) are expressed and could mediate insulin-increased expression andactivity of the equilibrative nucleoside transporter type 2 in human umbilical veinendothelial cells from pregnancies with gestational diabetes. Placenta 30:A87 (Abstract).

Westgate J, Lindsay R, Beattie J, Pattison N, Gamble G, Mildenhall L, Breier B, Johnstone F.2006. Hyperinsulinemia in cord blood in mothers with type 2 diabetes and gestationaldiabetes mellitus in New Zealand. Diabetes Care 29:1345–1350.

Wierzbicki A, Chowienczyk P, Cockcroft J, Brett S, Watts G, Jenkins B, Ritter J. 2004.Cardiovascular risk factors and endothelial dysfunction. Clin Sci 107:609–615.

Wu J, RobinsonD, KungH, HatzoglouM. 1994. Hormonal regulation of the gene for the typeC ecotropic retrovirus receptor in rat liver cells. J Virol 68:1615–1623.

Yang Z, Kaye D. 2009. Mechanistic insights into the link between a polymorphism of the3’UTR of the SLC7A1 gene and hypertension. Hum Mutat 30:328–333.



insulin-stimulated l-arginine transport requires slc7a1 gene expression and is associated with human...

Documents