lable at ScienceDirect

Placenta 32, Supplement B, Trophoblast Research, Vol. 25 (2011) S159eS164

Contents lists avai

Placenta

journal homepage: www.elsevier .com/locate/placenta

Review: Differential placental macrovascular and microvascular endothelialdysfunction in gestational diabetes

L. Sobrevia a,*, F. Abarzúa b, J.K. Nien c, C. Salomón a, F. Westermeier a, C. Puebla a, F. Cifuentes a,d,E. Guzmán-Gutiérrez a, A. Leiva a, P. Casanello a

aCellular and Molecular Physiology Laboratory (CMPL) & Perinatology Research Laboratory (PRL), Division of Obstetrics and Gynecology, Medical Research Centre (CIM),School of Medicine, Faculty of Medicine, Pontificia Universidad Católica de Chile, P.O. Box 114-D, Santiago, ChilebDivision of Obstetrics and Gynecology, School of Medicine, Pontificia Universidad Católica de Chile, Santiago, ChilecDepartment of Obstetrics and Gynecology, Clínica Dávila, Santiago, ChiledBiomedical Department, Faculty of Health Science, Universidad de Antofagasta, Antofagasta, Chile

a r t i c l e i n f o

Article history:Accepted 9 December 2010

Keywords:EndotheliumPlacentaMembrane transportGestational diabetes

* Corresponding author. Tel.: þ56 2 3548116; fax: þE-mail address: sobrevia@med.puc.cl (L. Sobrevia)

0143-4004/$ e see front matter � 2011 Published bydoi:10.1016/j.placenta.2010.12.011

a b s t r a c t

Human endothelial dysfunction is a common feature in many diseases of pregnancy, such as gestationaldiabetes (GD). Metabolic changes include abnormal synthesis of nitric oxide (NO) and abnormalmembrane transport of L-arginine and adenosine in primary cultures of human umbilical vein (HUVEC,macrovascular) and placental microvillus (hPMEC, microvascular) endothelial cells. These alterations areassociated with modifications in the expression and activity of endothelial (eNOS) and inducible (iNOS)NO synthases, respectively, an effect that is maintained at least up to passage 5 in culture. HUVEC andhPMEC exhibit expression and activity of the human cationic amino acid transporter 1 (hCAT-1), equi-librative nucleoside transporters 1 (hENT1) and hENT2, as well as the corresponding SLC7A1, SLC29A1 andSLC29A2 gene promoter activities. Altered gene expression results from increased NO level, proteinkinase C, mitogen-activated protein kinases, and hCHOP-C/EBPa transcription factor activation. ReducedENT-mediated adenosine transport in GD is associated with stimulation of the L-arginine/NO pathway,and mainly due to reduced expression and activity of hENT1. In addition, hENT2 activity seems able torestore the reduced adenosine transport in GD. Additionally, insulin exerts a differential modulation ofendothelial cells from macrocirculation compared with microcirculation, possibly due to expression ofdifferent insulin receptor isoforms. It is suggested that a common functional characteristic leading tochanges in the bioavailability of adenosine and metabolism of L-arginine is evidenced by human fetalmicro and macrovascular endothelium in GD.

� 2011 Published by IFPA and Elsevier Ltd.

1. Introduction

The endothelium is a specialized tissue related to several physi-ological functions, including blood flow modulation duringhomeostasis and in pathological conditions [1,2]. The humanplacenta is the key organ for fetal development mostly due to itsabundant vascularization. Since the placenta lacks innervation [3],locally released molecules play key roles in maintaining a propermother-to-fetus and fetus-to-mother flux of nutrients includingoxygen (O2), D-glucose, amino acids, and other less well character-ized metabolites, such as nucleosides [1,2,4]. According to location,the endothelial cells of different types of vessels aremorphologically

56 2 6321924..

IFPA and Elsevier Ltd.

and functionally different [5,6]. Many pathological conditionsduring pregnancy are related to damage of the placental macro-vasculature and microvasculature and they show differentialresponse to local modulators, involving signaling mechanisms thatcould result from the disease itself or be a potential adaptiveresponse to an abnormal intrauterine environment [1,2]. An im-balance or loss of essential endothelial functions, including angio-genesis, modulation of vascular tone and the ability to act asa physical and metabolic barrier, is suggestive of abnormal cellfunction, hereafter referred to as ‘endothelial dysfunction’.

The human placental macrovascular endothelium synthesizesand releases vasodilators, such as nitric oxide (NO), with high effi-ciency [1,2,4,7,8]. This is feasible due to efficientmechanisms to takeup and metabolize the cationic, semi-essential amino acid L-argi-nine. Interestingly, the human placental microvascular endothe-lium, which is much less characterized in its potential to participate

L. Sobrevia et al. / Placenta 32, Supplement B, Trophoblast Research, Vol. 25 (2011) S159eS164S160

in these phenomena, also synthesizes NO and takes up L-arginineand adenosine [9,10]; however, the intrinsicmechanisms seem tobedifferent to those of the macrovascular milieu and subject todifferent modulation [10e12]. Cellular mechanisms characterizedin macrovascular endothelium are not necessarily similar orappropriate to extrapolate to the microvascular endothelium andvice-versa [1,2,4,10]. For comparison, Table 1 summarizes andcompares some of reported characteristics of the human placentalmacrovascular and microvascular endothelium. Interestingly, itseems that macrovasculature is in general less reactive comparedwith endothelium from the microvasculature. For example,expression of homeobox genes such asHLX1, TLX1 and TLX2 is higherin placental microvascular than in macrovascular endothelium

Table 1Characteristics of macrovascular and microvascular endothelial cells from humanplacenta.

Characteristic Macrovascular Microvascular References

Celltype

Finding Celltype

Finding

PROKR1 expressiona HUVEC þ HPEC þ [6]PROKR2 expressiona HUVEC þ HPEC þ [6]EG-VEGF stimulated

proliferationHUVEC e HPEC þ [6]

EG-VEGF stimulatedmigration

HUVEC e HPEC þ [6]

EG-VEGF stimulatedsurvival

HUVEC þ HPEC þ [6]

Homeobox genesexpressionDLX3 HUVEC þ PLEC þ [13]DLX4 HUVEC þ PLEC þ [13]MSX2 HUVEC þ PLEC þ [13]GAX HUVEC þ PLEC þ [13]HLX1 HUVEC þ PLEC þþþþ [13]TLX1 HUVEC þ PLEC/HPEC þþþþ [14]TLX2 HUVEC þ PLEC/HPEC þþþþ [14]PHOX1 HUVEC þ PLEC/HPEC þþ [14]MEIS2 HUVEC þþ PLEC/HPEC þ [14]TGIF HUVEC þþ PLEC/HPEC þ [14]

aLDL internalization HUVEC þ PLECb þ [5]VEGF stimulated

proliferationHUVEC þ PLECb þ [5]

FGF2-stimulatedproliferation

HUVEC þ PLECb þ [5]

PIGF1-stimulatedproliferation

HUVEC þ PLECb þþþþþþþ [5]

PIGF2-stimulatedproliferation

HUVEC þ PLECb þþþþþþþ [5]

PAL-Eexpressiona LPV þþ SPV þ [50]QBEND10

expressionaLPV þþ SPV þ [50]

1F10 expressiona LPV þ SPV þ [50]HM15/3 expressiona LPV þþ SPV þ [50]Transferrin receptor

expressionaLPV þþ SPV þ [50]

IgC receptorexpressiona

LPV þþ SPV þ [50]

HUVEC, human umbilical vein endothelial cells; HPEC, cultured human placentalmicrovascular endothelial cells; PLEC, uncultured (freshly isolated) placentalmicrovascular endothelial cells; PROKR1, endocrine gland derived vascular endo-thelial growth factor receptor 1; PROKR2, endocrine gland derived vascular endo-thelial growth factor receptor 2; EG-VEGF, endocrine gland derived vascularendothelial growth factor; PAL-E, endothelial cell marker; QBEND10, Hematopoieticprogenitor cell antigen CD34; 1F10, Differentiated endothelial cells marker; HM15/3, endothelial cell marker; LPV, large placental vessels; SPV, small placental vessels;DLX3, DLX4, MSX2, GAX, HLX1, TLX1, TLX2, PHOX1, MEIS2, TGIF homeobox genesdetails in corresponding references; NECA, 50-N-ethyl-carboxamidoadenosine; TNF,tumor necrosis factor; aLDL, acetylated low density lipoprotein; VEGF, vascularendothelial growth factor 121 and 165 (in quoted study); FGF2, fibroblast growthfactor 2; PIGF1, placenta growth factor 1; PIGF2, placenta growth factor 2,þ denotespresence or function, e denotes absence or not function.

a Immunohistochemistry.b Cultured PLEC.

[13,14] and the proliferative response to growth factors, includingplacental growth factor (PGF), is greater in microvascular comparedto macrovascular endothelium, while the response to vascularendothelial growth factor (VEGF) is similar in these cell types [5]. Itis expected that, as in other vascular beds [6], placental endotheliafrom micro and macrovasculature exhibit differential responses tostimuli such as hormones (e.g., insulin, VEGF) or vasoactive mole-cules (e.g., NO, adenosine), a phenomenon that indeed could bealtered in diseases of pregnancy, such as gestational diabetes (GD).These differences could be crucial, for example, in placental angio-genesis, where a broad array of factors is involved, such as VEGF,PGF, circulating soluble fms-like tyrosine kinase-1 (sFlt-1), anendogenous antiangiogenic protein synthesized in placenta thatmay neutralizes proangiogenic effect of VEGF and PGF, and adeno-sine [15]. Since expression and activity of these proteins aredifferent in the macro compared with the microcirculation of thehuman fetoplacental unit [2,10,15], it is to be expected that differentmechanisms leading to or repressing angiogenesis could be occur-ring in these vascular beds. In addition, environmental, genetic, andimmunologic factors could be involved in this potentially differen-tial response of the human fetoplacental vasculature to agentsinvolved in vasculogenesis [16]. Evenwhen altered responses of thehuman placental endothelium (micro and macrovascular) havebeen established to, for example, membrane receptor expressionand activity (e.g., serotonin and insulin receptors) [2,11,17], junc-tional molecules (e.g., cadherin, catenin, occludin) [18], or signalingmechanisms (e.g., vasorelaxation, membrane transporters-medi-ated transport, differential responses to insulin) [19,20], nothing isreported regarding the intrinsic mechanisms leading to thesephenomena in GD [2].

Gestational diabetes is a syndrome characterized by glucoseintolerance with onset or first recognition during pregnancy [21].The clinicalmanifestations of GD have been attributedmainly to theconditions of hyperglycemia, hyperlipidaemia, hyperinsulinaemia,and fetoplacental endothelial dysfunction [22,23]. It hasbeen shownthat GD-associated alterations result froma change in the amount ofD-glucose available to the fetus due tomaternal or placental changes(e.g., transplacental transport of D-glucose) or hormone-induceddysfunction (e.g., insulin signaling) leading to abnormal growth ofthe fetus and perinatal complications [1,2,23,24]. Severalmarkers ofendothelial dysfunction have been reported in GD, includingincreased levels of endothelium-derived reactive oxygen (ROS) andnitrative (RNS) derived species,whereNADPHoxidase and xanthineoxidase, among other molecules, play key roles in the fetal andmaternal circulation [25]; increased levels in maternal plasma ofsymmetric dimethylarginine (ADMA) and high-sensitive C-reactiveprotein [26]; and altered expression of junctional adhesion mole-cules [18]. Table 2 shows a summary of the reported GD-associatedchanges in the function of human placenta endothelial cells frommacrovasculature compared with microvasculature. This reviewfocuses on the ability of macrovascular and microvascular endo-thelium from human placenta to take up and metabolize L-arginineand adenosine, controlling their broad biological effects.

2. Vascular reactivity in gestational diabetes

Altered placental vascular reactivity is characteristic of GD and ismainly due to fetoplacental endothelial dysfunction [1,2,25,27,28].Patients with diabetes mellitus exhibit reduced endothelium-dependent vasodilatation due to diminished NO synthesis by theendothelium or lower NO bioavailability resulting from in activa-tion by reactive oxygen species (ROS) [7,29]; however, increased NOsynthesis has been reported in human placental veins and arteries[19] and in primary cultures of HUVEC [20,31] from GD pregnancy.Vascular dysfunction resulting from this syndrome may also be

Table 2Effect of gestational diabetes on the human placental endothelial function.

Molecule/function Cell type/tissue Effect of GD References

SERT proteina Fetal capillaryendothelium

Decreased [17]

5HT-2A receptorproteina

Fetal capillaryendothelium

Decreased [17]

SERT mRNA Fetal capillaryendothelium

Decreased [17]

5HT-2A receptormRNA

Fetal capillaryendothelium

Decreased [17]

Extracellular adenosineconcentration

HUVEC Increased [31]

Extracellular adenosineconcentration

hPMEC Increased [12]

hENT1 protein HUVEC Decreased [20, 31]hENT2 protein hPMEC Decreased [12]hENT1 mRNA HUVEC Decreased [31]hENT1 mRNAb HUVEC Decreased [20]hENT2 mRNAb hPMEC Decreased [20]SLC29A1 promoter

expressionHUVEC Decreased [20]

Placenta insulin receptorsabundancea

Fetalendothelium

Increased [11]

IR-A mRNAb HUVEC Increased [51]Insulin effect on

IR-A mRNAbHUVEC Decreased [51]

IR-A mRNAb hPMEC Decreased [12]IR-B mRNAb hPMEC Increased [12]Insulin effect on

IR-A mRNAbhPMEC Increased [12]

Insulin effect onIR-B mRNAb

hPMEC Decreased [12]

hENTs-mediated transport HUVEC Decreased [52]hENTs-mediated transport hPMEC Decreased [12]hENT1-mediated transport HUVEC Decreased [20, 31]hENT1-mediated transport hPMEC Decreased [12]hCATs-mediated transport HUVEC Increased [31]hCAT-1 mRNA HUVEC Increased [31]NOS activity HUVEC Increased [20, 31]eNOS protein HUVEC Increased [20, 31]eNOS mRNA HUVEC Increased [20, 31]Cadherina Placental vessels Decreased [18]beCatenina Placental vessels Decreased [18]Occludina Placental vessels Decreased [18]Zonula occludens-1a Placental vessels Decreased [18]Junctional

phosphotyrosineaPlacental vessels Increased [18]

Hypoxia-inducedvascular relaxationc

Placental arteries Increased [53]

Hypoxia-inducedvascular relaxationc

Placental veins Increased [53]

Endothelial integrityd Umbilical artery Decreased [54]Vascular tone Umbilical vein Decreased [54]

SERT, serotonin receptor; 5HT-2A, 2A serotonin receptor; hCATs, human cationicamino acid transporters; hCAT-1, hCAT isoform 1; hENTs, human equilibrativenucleoside transporters; hENT1, hENT isoform 1; hENT2, hENT isoform 2; IR-A,insulin receptor A, IR-B, insulin receptor B; NOS, nitric oxide synthase; eNOS,endothelial NOS; HUVEC, human umbilical vein endothelial cells; hPMEC, humanplacental microvascular endothelial cells.

a Immunohistochemistry.b Quantitative RT-PCR expressed in number of mRNA copies.c Hypoxia reached exposing preparations to 95% nitrogen/5% carbon dioxide.d Analysis by light microscopy.

L. Sobrevia et al. / Placenta 32, Supplement B, Trophoblast Research, Vol. 25 (2011) S159eS164 S161

a consequence of a functional dissociation between NO synthesisand uptake of L-arginine (i.e., the L-arginine/NO signaling pathway)[32,33] and NO bioavailability to the endothelium and vascularsmooth muscle in the human placental circulation [1,2,31,36].

2.1. The L-arginine/NO signaling pathway

Nitric oxide is synthesized from L-arginine in a metabolic reac-tion leading to equimolar formation of L-citrulline andNO, requiringactivity of endothelial NO synthase in endothelial cells (eNOS or

Type III) [32]. L-Arginine transport is mediated by systems yþ,y;þL, b0,þ and B0,þ [1,4,7,30,31,33,34]. System yþ is a family ofproteins known as the cationic amino acid transporter (CAT) family,comprising CAT-1, CAT-2A, CAT-2B, CAT-3 and CAT-4, and encodedby SLC7A1, 2, 3 and 4 genes, respectively [35]. CAT-1 is ubiquitouslyexpressed, while CAT-2A and CAT-3 are constitutively expressed inliver and brain, respectively, and CAT-2B is induced in a variety ofcell types in response to bacterial endotoxins and pro-inflammatorycytokines. No transport activity has yet been described for CAT-4[1,4,7,35,36]. Only human CAT-1 (hCAT-1) and hCAT-2B have beenfunctionally characterized and protein and mRNA detected inprimary cultures of HUVEC [1,4,7,30,33,36].

2.2. SLC7A1 expression

Currently, there is no information regarding regulation of humanSLC7A1 expression [7]. This gene consists of an open reading framewith 11 exons and 10 introns, with two additional untranslatableexons (�2 and �1) towards the 50-end [37]. The promoter regionlacks a TATA box, has multiple specific protein 1 (Sp1) transcriptionfactor binding sites and has an extensive 30-untranslated (30UTR)region [38]. Although there are no studies describing the tran-scriptional activity of the SLC7A1 promoter, a polymorphism in the30UTR identified in somepatients suffering fromhypertension couldindicate a region potentially involved in regulation of hCAT-1expression [39]. It has been proposed that a fragment �650 bpupstream the ATG of SLC7A1 promoter cloned fromprimary culturesof HUVEC contains several consensus sequences for transcriptionfactors that may be critical for the regulation of its expression byinsulin and/or D-glucose [7]. Transcriptional activity of the SLC7A1promoter region is increased by insulin due to higher Sp1 activity [7]and in response to elevated extracellular D-glucose, NO, or TGF-b1[40,43]. However, nothing has been reported regarding modulationof SLC7A1 promoter activity in GD including in endothelium fromthe human fetoplacental unit [2,7].

3. Placental vascular dysfunction in gestational diabetes

Synthesis of NO and L-arginine transport activity are increased inprimary cultures of HUVEC from GD [1,4,7,20,31,36]. Increased L-arginine transport is complemented by a higher number of hCAT-1mRNA copies compared with cells from normal pregnancies [31].Interestingly, since PKC inhibition increases L-arginine transport,and because p42 and p44 mitogen-activated protein kinase (p42/44mapk) activity is higher in response to NO and PKC, the mecha-nisms by which L-arginine transport is activated in HUVEC from GDseem to depend on these signaling molecules. Unfortunately,nothing is yet available regarding the potential effects of GD in cellsignaling in macrovascular or microvascular endothelium of theplacenta [1,4,7,20]. There are a few reports in the literature showingthat endothelium from placental macrovasculature exhibit alter-ations in its capacity to take up and release adenosine in response toseveral stimuli, including elevated D-glucose [42], hypoxia [41],insulin [42], TGF-b1 [40] or pathologies such asGD [31] (see Table 2).However, there is no information regarding the mechanisms bywhich human placental circulation at the microvascular bed ismodulated in GD in terms of adenosine and other vasoactivemolecules [2]. It has been shown that primary cultures of hPMECfrom normal pregnancies express L-arginine transporters, whichapparently drive NO synthesis via the inducible NOS isoform (iNOS)[15]. Interestingly, even though preliminary results show alteredplasma levels of adenosine in the umbilical venous blood of babiesfrom GD [12], information regarding cellular and molecular mech-anisms responsible for these changes in the human fetoplacentalendothelium (and in the endothelium in general) is limited [1,2,12].

L. Sobrevia et al. / Placenta 32, Supplement B, Trophoblast Research, Vol. 25 (2011) S159eS164S162

4. Adenosine as a modulator of the L-arginine/NO pathway

Adenosine modulates L-arginine transport and NO synthesis inHUVEC, i.e., ‘ALANO’ (Adenosine/L-Arginine/Nitric Oxide) signalingpathway [1,2,20,31,36,43]. Extracellular adenosine is increased inGD and activates L-arginine transport and NO synthesis in HUVECthrough A2A adenosine receptor activation, triggering a sequence ofsignaling reactions where PKC activation precedes eNOS and p42/44mapk phosphorylation [2,31,36]. Thus, a relationship betweenexpression and activity of L-arginine and adenosine membranetransporters and receptors in HUVEC fromGD has been established.In addition, GD is associated with activation of transcription factorsrepressing promoter activity of SLC29A1 (coding for human equili-brative nucleoside transporter 1, hENT1, see Section 5 below)resulting in reduced transcript and protein abundance withsubsequently reduced adenosine uptake. Since GD is also associatedwith increased expression of SLC7A1, leading to increased uptake ofL-arginine [7], it is suggested that a similar signaling pathway maybe involved in this response of endothelial cells.

5. Adenosine transport

Two families of nucleoside transporters have been identified,i.e., the equilibrative nucleoside transporters (ENTs) and theconcentrative nucleoside transporters (CNTs) [1,2,36,44]. Fourmembers of the ENT family of solute carriers (SLC29A genes) havebeen cloned, i.e., hENT1, hENT2, hENT3 and hENT4 [1,2,36,44].Adenosine transport in HUVEC is mainly (w80%) mediated byhENT1, with the remaining transport via hENT2 [1,2,20,31,41,42].These proteins are also expressed in hPMEC with, in contrast,a similar contribution to total adenosine transport [10,15]. The rolesof hENT3 (a lysosomal transporter) and hENT4 (activated atpHw6.5 and described in HUVEC) as nucleoside transporters in

Fig. 1. Adenosine and L-arginine transport in HUVEC and hPMEC from gestational diabetesadenosine either in macrovascular as in microvascular endothelial cells from the human placuptake of adenosine via the human equilibrative nucleoside transporter 1 (hENT1) with notin parallel to hENT1 in endothelial cells from the microvasculature of the human placenta treduced expression of SLC29A1 (for hENT1), leading to lower hENT1 mRNA and proteinmicrovasculature may result from reduced expression of SLC29A2 (for hENT2), and hENT2L-arginine transport via human cationic amino acid transporter 1 (hCAT-1), a phenomenonsynthesis of nitric oxide (NO). Based in data reported in references [1,2,7,10,12,15,20,31,36,5

endothelial cells is uncertain [1,2,45]. CNTs have not been describedin endothelial cells [1,2,44]; however, human CNT1 mRNA andCNT1-like transport activity have been recently suggested to playa role in the differentiation process of human endothelial progen-itor cells (hEPCs) from adult subjects [46].

GD is associated with reduced adenosine uptake due to reducedhENT1 maximal transport capacity [1,2]. Alternatively, a reducednumber of nucleoside binding sites per endothelial cell (w50%) hasbeen estimated in HUVEC from GD [31,36] compared with cellsfrom normal pregnancies. Since adenosine uptake efficiency (i.e.,adenosine molecules transported per nucleoside transporter perunit of time) is reported to be unaltered in HUVEC from GD [31,36],a reduced hENT1 expression is also likely in this cell type. It hasbeen reported that hPMEC express functional hENT1 and hENT2,and that GD is associated with reduced total (i.e., hENT1 þ hENT2)and saturable hENT1 and hENT2-mediated adenosine transport inthis cell type. Since GD is associated with elevated extracellularadenosine concentration compared with normal pregnancies,activation of adenosine receptors could repress adenosine trans-port via hENT1 and/or hENT2 in hPMEC [10,15]. These could becrucial mechanisms in the maintenance of physiological extracel-lular adenosine concentration in the microcirculation and macro-circulation of the human placenta in GD, where an abnormallyelevated extracellular adenosine concentration has been reportedin the culture medium of HUVEC (w2 mM) compared with cellsfrom normal pregnancies (w50e500 nM) and with the concen-tration of this nucleoside in human umbilical venous blood fromnormal pregnancies (w500 nM) [47].

5.1. SLC29A1 and SLC29A2 regulation

Gestational diabetes is also associated with reduced hENT1protein abundance in HUVEC, a phenomenon not due to increased

. Gestational diabetes is associated with increased ([) extracellular concentrations ofental vasculature. In the macrovasculature this phenomenon result from a reduced (Y)significant alterations in hENT2 transport activity. However, hENT2 seems to contributeo this phenomenon. Reduced hENT1-mediated transport in both cell types result fromlevels. In addition, reduced hENT2-mediated adenosine transport in cells from themRNA and protein levels. All these phenomena seem to be paralleled by increasedthat seems associated with increased expression of SLC7A1 (for hCAT-1) and higher

1].

L. Sobrevia et al. / Placenta 32, Supplement B, Trophoblast Research, Vol. 25 (2011) S159eS164 S163

hENT1 protein degradation [20]. Thus, a reduced maximal velocity(Vmax) of adenosine transport could result from reduced hENT1protein availability at the plasma membrane. Supporting the latterthere is now evidence showing that hENT1-EGFP fusion protein isrelocated to intracellular compartments in HUVEC from GD [FaríasM, Sobrevia L, unpublished results]. In addition, a sequence span-ning from �2154 to �1114 bp of the SLC29A1 promoter containsconsensus sites for repressive transcription factor(s) leading todown-regulation of SLC29A1 promoter activity, such as the NO-activated hCHOPeC/EBPa protein complex [20], and diabetesmellitus [48] in other cell types. In contrast, there is no informationregarding modulation of SLC29A2 expression in endothelial cellsfrom the fetal circulation [1,2]. Two fragments of �1479 and�586 bp (from ATG) have been cloned from SLC29A2 in HUVECfrom normal pregnancies (Salomón C, Sobrevia L, unpublishedresults). These findings show that SLC29A2 promoter activity isreduced in hPMEC from GD, but it is unaltered in HUVEC from thissyndrome, and that SCL29A2 promoter activity is increased;although it is decreased by insulin in HUVEC and hPMEC, respec-tively, from normal pregnancies. This reinforces the concept ofa differential response of HUVEC compared with hPMEC even atthis level of regulation in the human placenta.

In summary, Fig. 1 shows the potential effect of GD on adenosineand L-arginine transport in macro and microvascular endotheliumof the human fetoplacental unit. The reported effects of GD onadenosine transport via hENT1 and hENT2, and L-arginine transportvia hCAT-1 in HUVEC suggest that this syndrome alters cellsignaling cascades (involving PI3K, PKC, NO and p42/44mapk) [1,2],resulting in activation of transcription factors that repress SLC29A1promoter activity to reduce hENT1 mRNA level and protein abun-dance with the subsequent reduced adenosine uptake andincreased extracellular concentration of this nucleoside. Far less isknown regarding expression and activity of hENT2 in these celltypes. It has been proposed that NO is not involved in the modu-lation of hENT2 in HUVEC, but nothing is clear regarding theimplications of the proposed signaling pathway in the promoteractivity of SLC29A2 in GD. In addition, GD is associated withincreased SLC7A1 expression leading to higher L-arginine transportvia hCAT-1, a phenomenon that could be determinant in increasingeNOS expression and activity in HUVEC from GD, or potentiallyiNOS in hPMEC. The role of NO is then shown to be as a modulatorof expression and activity of transcription factors with consensussequences in SLC29A1, SLC29A2 and a SLC7A1 in HUVEC, andpotentially in hPMEC.

6. Concluding remarks

Endothelial cells from macro and microvasculature of theplacenta differ in their ability to take up and/or metabolizesubstrates that are directly or indirectly involved in the modulationof vascular tone in the placental vascular bed. This could be theresult of differential expression of genes encoding for membranetransporters for adenosine and L-arginine, as well as NOS. Severalmechanisms are feasible, including different sorts of transcriptionfactors, and changes in the efficiency of membrane transportersand NOS. Since the plasma concentration of adenosine in humanumbilical vessels is apparently increased in patients with GD (Sal-omón C, Sobrevia L, unpublished results), it is feasible for abnormaladenosine transport and/or metabolism to be considered in thepathophysiological processes of GD and eventually included as partof the management of mother and fetus. In addition, fetal andmaternal hyperinsulinaemia in patients with GD may be due to anabnormal insulin sensitization, a phenomenon described in theplacentas of GD pregnancies [49], involving other mechanisms thanthe single increase of insulin release, making the endotheliummore

sensitive to insulin. In this review we have recalled the importanceof adenosine, L-arginine and NO (ALANO pathway) [1,2,36] inGD-associated endothelial dysfunction, yet more knowledge isnecessary to further understand the pathological processes as wellas the requirements for an accurate clinical management.

Role of funding source

Fondo Nacional de Desarrollo Científico y Tecnológico (FONDE-CYT 1110977, 1070865, 1080534); Programa de InvestigaciónInterdisciplinario (PIA) from Comisión Nacional de Investigación enCiencia y Tecnología (CONICYT)(Anillos ACT-73); Fellowship Apoyode Tesis CONICYT (AT-24090190, AT-24100210), Chile. Dirección deInvestigación (DI-1339-07) and Vicerrectoría Académica (AnillosACT-73 postdoctoral research associate at CMPL-PRL, PontificiaUniversidad Católica de Chile), Universidad de Antofagasta, Chile.E Guzmán-Gutiérrez, F Westermeier, C Puebla hold CONICYT-PhD(Chile) fellowships. C Salomón holds and A Leiva was the recipientof a Faculty of Medicine, Pontificia Universidad Católica de Chile-PhD fellowship.

Conflict of Interest Statement

The authors state they have no conflict of interest.

Acknowledgments

We thank the researchers at the Cellular and Molecular Physi-ology Laboratory (CMPL) and Perinatology Research Laboratory(PRL) of the Pontificia Universidad Católica de Chile for theircontribution in the production of the experimental data that hasbeen cited throughout the text. Authors also thank Mrs NinoskaMuñoz for excellent secretarial assistance, and the personnel of theHospital Clínico Pontificia Universidad Católica de Chile labourward for supply of placentas. E Guzmán-Gutiérrez and C Pueblaacknowledge YW Loke New Investigator Awards for communica-tion of part of the results described in this review at the 2010 IFPAmeeting in Santiago de Chile.

References

[1] Casanello P, Escudero C, Sobrevia L. Equilibrative nucleoside (ENTs) andcationic amino acid (CATs) transporters: implications in foetal endothelialdysfunction in human pregnancy diseases. Curr Vasc Pharmacol2007;5:69e84.

[2] Westermeier F, Puebla C, Vega JL, Farías M, Escudero C, Casanello P, et al.Equilibrative nucleoside transporters in fetal endothelial dysfunction in dia-betes mellitus and hyperglycaemia. Curr Vasc Pharmacol 2009;7:435e49.

[3] Fox SB, Khong TY. Lack of innervation of human umbilical cord. An immu-nohistological and histochemical study. Placenta 1990;11:59e62.

[4] Sobrevia L, Puebla C, Farías M, Casanello P. Role of equilibrative nucleosidetransporters in fetal endothelial dysfunction in gestational diabetes. In:Sobrevia L, Casanello P, editors. Membrane Transporters and Receptors inDisease. Kerala, India: Research Signpost, 2009.

[5] Lang I, Pabst MA, Hiden U, Blaschitz A, Dohr G, Hahn T, et al. Heterogeneity ofmicrovascular endothelial cells isolated from human term placenta and mac-rovascular umbilical vein endothelial cells. Eur J Cell Biol 2003;82:163e73.

[6] Brouillet S, Hoffmann P, Benharouga M, Salomon A, Schaal JP, Feige JJ, et al.Molecular characterization of EG-VEGF-mediated angiogenesis: differentialeffects on microvascular and macrovascular endothelial cells. Mol Biol Cell2010;21:2832e43.

[7] Sobrevia L, González MA. Role for insulin on L-arginine transport in fetal endo-thelial dysfunction in hyperglycaemia. Curr Vasc Pharmacol 2009;7:467e74.

[8] Myatt L. Reactive oxygen and nitrogen species and functional adaptation ofthe placenta. Placenta 2010;31:S66e9.

[9] Dye J, Lawrence L, Linge C, Leach L, Firth J, Clark P. Distinct patterns ofmicrovascular endothelial cell morphology are determined by extracellularmatrix composition. Endothelium 2004;11:151e67.

[10] Escudero C, Casanello P, Sobrevia L. Human equilibrative nucleoside trans-porters 1 and 2 may be differentially modulated by A2B adenosine receptorsin placenta microvascular endothelial cells from preeclampsia. Placenta2008;29:816e25.

L. Sobrevia et al. / Placenta 32, Supplement B, Trophoblast Research, Vol. 25 (2011) S159eS164S164

[11] Hiden U, Lang I, Ghaffari-Tabrizi N, Gauster M, Lang U, Desoye G. Insulinaction on the human placental endothelium in normal and diabetic preg-nancy. Curr Vasc Pharmacol 2009;7:460e6.

[12] Salomón C, Westermeier F, Casanello P, Sobrevia L. Differential modulation ofinsulin receptor isoforms expression and NOS activity by insulin in humanplacenta microvascular endothelial cells from gestational diabetes. Placenta2010;31. A-80. Abstract.

[13] Murthi P, SoM,GudeNM,DohertyVL, Brennecke SP, Kalionis B.Homeobox genesare differentially expressed in macrovascular human umbilical vein endothelialcells and microvascular placental endothelial cells. Placenta 2007;28:219e23.

[14] Murthi P, Hiden U, Rajaraman G, Liu H, Borg AJ, Coombes F, et al. Novelhomeobox genes are differentially expressed in placental microvascular endo-thelial cells compared with macrovascular cells. Placenta 2008;29:624e30.

[15] Escudero C, Puebla C, Westermeier F, Sobrevia L. Potential cell signallingmechanisms involved in differential placental angiogenesis in mild and severepre-eclampsia. Curr Vasc Pharmacol 2009;7:475e85.

[16] Krause B, Sobrevia L, Casanello P. Epigenetics: new concepts of oldphenomena in vascular physiology. Curr Vasc Pharmacol 2009;7:513e20.

[17] Viau M, Lafond J, Vaillancourt C. Expression of placental serotonin transporterand 5-HT 2A receptor in normal and gestational diabetes mellitus pregnan-cies. Reprod Biomed Online 2009;19:207e15.

[18] Babawale MO, Lovat S, Mayhew TM, Lammiman MJ, James DK, Leach L. Effectsof gestational diabetes on junctional adhesion molecules in human termplacental vasculature. Diabetologia 2000;43:1185e96.

[19] Figueroa R, Martinez E, Fayngersh RP, Tejani N, Mohazzab-H KM, Wolin MS.Alterations in relaxation to lactate and H2O2 in human placental vessels fromgestational diabetic pregnancies. Am J Physiol 2000;278:H706e13.

[20] Farías M, Puebla C, Westermeier F, Jo MJ, Pastor-Anglada M, Casanello P, et al.Nitric oxide reduces SLC29A1 promoter activity and adenosine transportinvolving transcription factor complex hCHOP-C/EBPa in human umbilical veinendothelial cells from gestational diabetes. Cardiovasc Res 2010;86:45e54.

[21] Metzger BE, Buchanan TA, Coustan DR, de Leiva A, Dunger DB, Hadden DR,et al. Summary and recommendations of the fifth international workshop-conference on gestational diabetes mellitus. Diab Care 2007;30:S251e60.

[22] Nold JL, Georgieff MK. Infants of diabetic mothers. Pediatr Clin North Am2004;51:619e37.

[23] Greene MF, Solomon CG. Gestational diabetes mellitus - time to treat. N EnglJ Med 2005;352:2544e6.

[24] Desoye G, Hauguel-de Mouzon S. The human placenta in gestational diabetesmellitus. The insulin and cytokine network. Diab Care 2007;30:120e6.

[25] Biri A, Onan A, Devrim E, Babacan F, Kavutcu M, Durak I. Oxidant status inmaternal and cord plasma and placental tissue in gestational diabetes.Placenta 2006;27:327e32.

[26] Akturk M, Altinova A, Mert I, Dincel A, Sargin A, Buyukkagnici U, et al.Asymmetric dimethylarginine concentrations are elevated in women withgestational diabetes. Endocrine 2010;38:134e41.

[27] Omar HA, Ramirez R, Arsich J, Tracy T, Glover D, Gibson M. Reduction of thehuman placental vascular relaxation to progesterone by gestational diabetes.J Matern Fetal Investig 1998;8:27e30.

[28] Tchirikov M, Rybakowski C, Hüneke B, Schoder V, Schröder HJ. Umbilical veinblood volume flow rate and umbilical artery pulsatility as ‘venous-arterialindex’ in the prediction of neonatal compromise. Ultrasound Obstet Gyneco2002;20:580e5.

[29] Poston L, Taylor PD. Endothelium-mediated vascular function in insulin-dependent diabetes mellitus. Clin Sci 1995;88:245e55.

[30] González M, Flores C, Pearson JD, Casanello P, Sobrevia L. Cell signalling-mediating insulin increase of mRNA expression for cationic amino acidtransporters-1 and -2 and membrane hyperpolarization in human umbilicalvein endothelial cells. Pflugers Arch 2004;448:383e94.

[31] Vásquez G, Sanhueza F, Vásquez R, González M, San Martín R, Casanello P,et al. Role of adenosine transport in gestational diabetes-induced L-argininetransport and nitric oxide synthesis in human umbilical vein endothelium.J Physiol 2004;560:111e22.

[32] Alderton WK, Cooper CE, Knowles RG. Nitric oxide synthases: structure,function and inhibition. Biochem J 2001;357:593e615.

[33] Casanello P, Sobrevia L. Intrauterine growth retardation is associated withreduced activity and expression of the cationic amino acid transport systems

yþ/hCAT-1 and yþ/hCAT-2B and lower activity of nitric oxide synthase inhuman umbilical vein endothelial cells. Circ Res 2002;91:127e34.

[34] Arancibia-Garavilla Y, Toledo F, Casanello P, Sobrevia L. Nitric oxide synthesisrequires activity of the cationic and neutral amino acid transport systemyþL in human umbilical vein endothelium. Exp Physiol 2003;88:699e710.

[35] Verrey F, Closs EI, Wagner CA, Palacin M, Endou H, Kanai Y. CATs and HATs:the SLC7 family of amino acid transporters. Pflügers Arch 2004;447:532e42.

[36] San Martín R, Sobrevia L. Gestational diabetes and the adenosine/L-arginine/nitric oxide (ALANO) pathway in human umbilical vein endothelium. Placenta2006;27:1e10.

[37] Hammermann R, Brunn G, Racké K. Analysis of the genomic organization ofthe human cationic amino acid transporters CAT-1, CAT-2 and CAT-4. AminoAcids 2001;21:211e9.

[38] Hatzoglou M, Fernandez J, Yaman I, Closs E. Regulation of cationic amino acidtransport: the story of the CAT-1 transporter. Annu Rev Nutr 2004;24:377e99.

[39] Yang Z, Venardos K, Jones E, Morris B, Chin-Dusting J, Kaye D. Identification ofa novel polymorphism in the 30UTR of the L-arginine transporter geneSLC7A1: contribution to hypertension and endothelial dysfunction. Circula-tion 2007;115:1269e74.

[40] Vega JL, Puebla C, Vásquez R, Farías M, Alarcón J, Pastor-Anglada M, et al.TGF-beta1 inhibits expression and activity of hENT1 in a nitric oxide-de-pendent manner in human umbilical vein endothelium. Cardiovasc Res2009;82:458e67.

[41] Casanello P, Torres A, Sanhueza F, González M, Farías M, Gallardo V, et al.Equilibrative nucleoside transporter 1 expression is downregulated byhypoxia in human umbilical vein endothelium. Circ Res 2005;97:16e24.

[42] Muñoz G, San Martín R, Farías M, Cea L, Vecchiola A, Casanello P, et al. Insulinrestores glucose inhibition of adenosine transport by increasing the expres-sion and activity of the equilibrative nucleoside transporter 2 in humanumbilical vein endothelium. J Cell Physiol 2006;209:826e35.

[43] Wyatt AW, Steinert JR, Wheeler-Jones CP, Morgan AJ, Sudgen D, Pearson JD,et al. Early activation of the p42/44MAPK pathway mediates adenosine-induced nitric oxide production in human endothelial cells: a novel calcium-insensitive mechanism. FASEB J 2002;16:1584e94.

[44] Baldwin SA, Beal PR, Yao SY, King AE, Cass CE, Young JD. The equilibrativenucleoside transporter family, SLC29. Pflügers Arch 2004;447:735e43.

[45] Barnes K, Dobrzynski H, Foppolo S, Beal PR, Ismat F, Scullion ER, et al.Distribution and functional characterization of equilibrative nucleosidetransporter-4, a novel cardiac adenosine transporter activated at acidic pH.Circ Res 2006;99:510e9.

[46] Guzmán-Gutiérrez E, Sandoval C, Nova E, Castillo JL, Vera JC, Lamperti L, et al.Differential expression of functional nucleoside transporters in non-differ-entiated and differentiated human endothelial progenitor cells. Placenta2010;31:928e36.

[47] Yoneyama Y, Sawa R, Suzuki S, Ishino H, Miura A, Kuwabara Y, et al. Regu-lation of plasma adenosine levels in normal pregnancy. Gynecol Obstet Invest2002a;53:71e4.

[48] Ozcan U, Cao Q, Yilmaz E, Lee AH, Iwakoshi NN, Ozdelen E, et al. Endoplasmicreticulum stress links obesity, insulin action, and type 2 diabetes. Science2004;306:457e61.

[49] Colomiere M, Permezel M, Riley C, Desoye G, Lappas M. Defective insulinsignaling in placenta from pregnancies complicated by gestational diabetesmellitus. Eur J Endocrinol 2009;160:567e78.

[50] Lang I, Hartmann M, Blaschitz A, Dohr G, Skofitsch G, Desoye G. Immunohis-tochemical evidence for the heterogeneity of maternal and fetal vascularendothelial cells in human full-term placenta. Cell Tissue Res 1993;274:211e8.

[51] Westermeier F, Salomón C, Casanello P, Sobrevia L. Insulin increased adeno-sine transport in HUVEC from gestational diabetic pregnancies involvesincreased expression and activity of hENT1. Placenta 2010;31. A78. Abstract.

[52] Sobrevia L, Jarvis SM, Yudilevich DL. Adenosine transport in cultured humanumbilical vein endothelial cells is reduced in diabetes. Am J Physiol1994;267:C39e47.

[53] Figueroa R, Omar HA, Tejani N, Wolin MS. Gestational diabetes alters humanplacental vascular responses to changes in oxygen tension. Am J ObstetGynecol 1993;168:1616e22.

[54] Singh SD. Gestational diabetes and its effect on the umbilical cord. Early HumDev 1986;14:89e98.