R

T

EA(

a

ARRAA

KVPEPVEEA

C

h10

Pharmacological Research 113 (2016) 125–145

Contents lists available at ScienceDirect

Pharmacological Research

j ourna l h om epage: w ww.elsev ier .com/ locate /yphrs

eview

he paracrine control of vascular motion. A historical perspective

duardo Nava (M.D., Ph.D.) ∗, Silvia Llorens (Ph.D.)rea of Physiology, Department of Medical Sciences, University of Castilla-La Mancha, School of Medicine and Regional Centre for Biomedical ResearchCRIB), Albacete, Spain

r t i c l e i n f o

rticle history:eceived 24 May 2016eceived in revised form 13 July 2016ccepted 1 August 2016vailable online 12 August 2016

eywords:asomotionrostacyclinndothelium-derived NOVATasoactive adipokineDHFDCFdventitia

a b s t r a c t

During the last quarter of the past century, the leading role the endocrine and nervous systems had onthe regulation of vasomotion, shifted towards a more paracrine-based regulation. This begun with therecognition of endothelial cells as active players of vascular control, when the vessel’s intimal layer wasidentified as the main source of prostacyclin and was followed by the discovery of an endothelium-derivedsmooth muscle cell relaxing factor (EDRF). The new position acquired by endothelial cells promptedthe discovery of other endothelium-derived regulatory products: vasoconstrictors, generally known asEDCFs, endothelin, and other vasodilators with hyperpolarizing properties (EDHFs). While this researchwas taking place, a quest for the discovery of the nature of EDRF carried back to a research line com-menced a decade earlier: the recently found intracellular messenger cGMP and nitrovasodilators. Bothwere smooth muscle relaxants and appeared to interact in a hormonal fashion. Prejudice against anunconventional gaseous molecule delayed the acceptance that EDRF was nitric oxide (NO). When thishappened, a new era of research that exceeded the vascular field commenced. The discovery of the path-way for NO synthesis from L-arginine involved the clever assembling of numerous unrelated observationsof different areas of knowledge. The last ten years of research on the paracrine regulation of the vascu-

lar wall has shifted to perivascular fat (PVAT), which is beginning to be regarded as the fourth layer ofthe vascular wall. Starting with the discovery of an adipose-derived relaxing substance (ADRF), the rolethat different adipokines have on the paracrine control of vasomotion is now filling the research activityof many vascular pharmacology labs, and surprising interactions between the endothelium, PVAT andsmooth muscle are being unveiled.

© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-NDlicense (http://creativecommons.org/licenses/by-nc-nd/4.0/).

ontents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1262. Early approaches to the paracrine role of the intima . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1263. Prostacyclin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1264. Endothelium-derived relaxing factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1275. Endothelium-derived constrictor factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1286. Endothelium-derived hyperpolarizing factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1297. Search for the chemical nature of EDRF. Nitrovasodilators and cyclic GMP. The link EDRF-NO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1318. The NO pathway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1329. Adventitia and periadventitial adipose tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13410. The paracrine role of the adventitia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13511. The paracrine role of PVAT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13512. Identity and mechanism of action of PVAT’s paracrine substances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13613. Interactions between PVAT, smooth muscle and endothelium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

14. Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

∗ Corresponding author at: Faculty of Medicine, University of Castilla-La Mancha, 0200E-mail address: eduardo.nava@uclm.es (E. Nava).

ttp://dx.doi.org/10.1016/j.phrs.2016.08.003043-6618/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article

/).

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140

6 Albacete, Spain.

under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.

1 ogical

1

tagacttpcmotvptftaeiteemobicwmb

ptcsapbaraipqlttbstrd

2

ieeyl

26 E. Nava, S. Llorens / Pharmacol

. Introduction

During much part of last century, vasomotion, i.e.: change inhe vessel’s diameter, was thought to be mostly regulated by theutonomous nervous system and hormones released by specificlands. An alternative to this was proposed during the late fortiesnd early fifties by Jimenez-Díaz and colleagues who, by means ofross-circulation experiments, demonstrated the endocrine role ofhe vessel wall [1,2]. However, no hormonal function was assignedo specific vascular structures until much later. The concept ofaracrine secretion was introduced by this time because it wasonvenient for distinguishing locally-acting gastrointestinal hor-ones from systemically-acting hormones [3] but the term was not

f general use in physiology as it is today. The first substance ableo dilate vascular smooth muscle synthesized and released by theessels themselves was described in the mid-seventies [4]. This wasrostacyclin [5]. With the turn of the decade, the endothelial cell ofhe vascular intimal layer was specifically pinpointed as essentialor the relaxation of adjacent smooth muscle cells [6]. The concep-ion of an endothelial cell releasing substances that preferentiallyffect nearby smooth muscle cells, i.e.: paracrine regulation, wasxtended during the eighties to a variety of substances, includ-ng prostacyclin and nitric oxide (NO), which is probably the onehat has received more attention. With this panorama in sight, thendothelium was proposed as a real endocrine organ [7]. Whilendothelial cells had gained a prominent role as regulators of vaso-otion and various other functions, other vascular cells timidly

btained the qualification of paracrine regulatory cells. The first toreak the ice was not, as one would initially think, the anatomically

mmediate adventitial fibroblast, but rather, the outermost perivas-ular fat tissue [8]. This opening work [8] had to wait until 2002,hen the burst of perivascular fat research took place [9]. In theeanwhile, new roles for adventitial cells as a source of substances

egan to be assigned [10].In this review we will search for the seminal advances and

ioneering work that yielded the knowledge we have today onhe major paracrine elements controlling vascular smooth muscleontraction or relaxation. These are, the endothelium-derived sub-tances: NO, prostacyclin, contracting and hyperpolarizing factorsnd the, relatively newer, substances released by the adventitia anderivascular adipose tissue. A chronological order of reporting haseen given priority to a more academic order. Occasionally, this canppear odd at reading. For example, for the endothelium-derivedelaxing factor (EDRF)-NO topic, NO is not explained right awayfter EDRF because nearly a decade of contracting and hyperpolar-zing factor research took place in between. In some occasions, thereliminary work necessary for a subsequent discovery took placeuite a long time and with, apparently, little connection with the

atter. In many instances, various groups published so closely inime to the report traditionally recognized as the first and principalhat it is easy to conceive that work aimed at the same target waseing carried out in parallel. Often, an idea is more or less vaguelyuggested in preliminary works and definitely demonstrated byhe same or other authors in subsequent reports. In the presenteview we have tried to quote every report related or involved in aiscovery and not solely that generally recognized as the first one.

. Early approaches to the paracrine role of the intima

The existence of endothelial cells was first reported by von Reck-nghausen in 1860. The German pathologist managed to observe the

ndothelial lining of the vessel’s inner wall by staining the edges ofndothelial cells with silver nitrate [11]. From then on and for manyears, the endothelium was regarded, more or less, as a tapestryacking any functional role. For example, at the sight of some of

Research 113 (2016) 125–145

the first electron microscope plates, endothelial cells were literallydescribed as “. . .to form little more than a sheet of nucleated cel-lophane” [12]. Nevertheless, early electron microscopist, Rhodin,intuitively suggested that “. . .humoral transmitter substances [. . .]are picked up by the luminal cell membrane of the endothelial cell andpassed on to the smooth muscle cells. . .”. This seemed to be a preludeof some kind of regulatory function [13]. Furthermore, some phar-macologists reported shortly later intriguing findings that were notconsistent with a passive notion of endothelial function. Bevan &Duckles, for instance, realized that noradrenaline bound to glassmicrobeads present in the vascular lumen and, therefore, unableto reach the smooth muscle wall, did elicit a contractile effect [14].They reasoned that the only way this could happen was if endothe-lial cells had surface receptors to the amine and endothelial cellshad a mechanism of spread towards the underlining muscle cells.They raised the possibility that circulating drugs initiate changesin the tone of vascular smooth muscle by an action on the inti-mal endothelial cells [14] by means of receptors on these cells thatwould be the first to detect blood-borne vasoactive substances [15].In those years, the existence of receptors to a variety of vasoac-tive agents on endothelial cell membranes and linked to cAMP andcGMP was already known [16]. Thus, the scenario for discovery wasready.

Another suggestive finding, both for the putative paracrine rolesof intimal and adventitial cells, was the observation that the effectsof contractile amines was very different when applied via theintimal side or through the adventitial side [17,18]. Vessels weresystematically more sensitive when exposed through the intima.Among others, one of the explanations they considered was thatthe different vascular layers the drug had to cross, intima andadventitia, carried out some sort of active role. Unfortunately, thissuspected active role was tested only in the adventitia, by specif-ically removing it [19], but not in the intimal endothelial cells, asFurchgott did very shortly later [6]. They missed an opportunity foran important discovery.

3. Prostacyclin

Prostacyclin, together with thromboxane A2 (TXA2) and vari-ous prostaglandins are synthesized from arachidonic acid, a fattyacid derived from membrane phospholipids. Endoperoxides areintermediate molecules between arachidonic acid and the variousprostanoids. Enzymes cyclooxygenase I and II catalyze the forma-tion of these endoperoxides. Prostacyclin was discovered in 1976by the group led by Vane [4]. Vane’s interest in prostaglandins datesback to the late sixties studying anaphylactic responses of perfusedlungs from sensitized guinea pigs and the different substancesthereby released, such as histamine, slow-reacting substance inanaphylaxis (SRS-A) and also prostaglandin-like substances [20].One of these was the so-called, “rabbit aorta contracting substance”or RCS, because it caused a strong contraction of strips of rabbitaorta. This was later identified as TXA2 [21], a powerful proaggre-gatory and vasoconstritive mediator, and the synthesizing enzyme,“thromboxane synthetase”, was isolated from the “microsomal”fraction of platelets [22]. Eventually, the group began to study othertissues in an attempt to find alternative sources of TXA2 gener-ation. John Vane explains how they came across an unexpectedfinding [23]: “It was Moncada’s suggestion that we should look intothe biosynthetic system of vascular tissue, since vascular endotheliumand platelets might share some structural features. Indeed, after severalweeks of work we found that microsomal fractions of pig aorta [. . .]

did not generate classical prostaglandins”. The prostaglandin foundin the vascular tissue, definitely was not TXA2, but a substancewith opposite properties: an anti-aggregating and vasodilator sub-stance. It did not take too long until the chemical structure of the

E. Nava, S. Llorens / Pharmacological

Fig. 1. Differential generation of prostacyclin by the layers of the vessel wall.The first to attribute a functional role to the endothelium were early prostacyclinresearchers showing that endothelial cells are the most important, if not the onlysource of this vasoactive eicosanoid [26,27]. The intimal layer, which accounts for5Ta

n[awrtatsttfsvtrlrcpopf[me

4

dmppreiFamid

% of the calculated weight, produced about 40% of the total vessel’s prostacyclin.he other layers contribute to the 60% prostacyclin left, but represent 95% of thertery’s weight. Modified from Moncada et al. [26].

ew prostanoid was identified and renamed prostacyclin (PGI2)24]. It was also shown that the actions of prostacyclin are medi-ted by the adenylate cyclase/cyclic AMP pathway [25] and that,ith a half-life long enough to reach distant targets, it could be

egarded as a circulating hormone as well [5]. The first report onhe newly discovered prostacyclin, still called PGX [4], was remark-bly mature in that today’s well established concepts of physiologyextbooks on the role of platelets and endothelial cells in hemosta-is were already stated in this seminal paper. Especially relevant forhe topic of the present review is that this paper also put forwardhe notion of a paracrine control of the tone of the vessel wall: “Theact that PGX is locally generated in the arterial wall and can relaxmooth muscle also suggests that it plays a part in the local control ofascular tone”. Shortly after this publication, the differential forma-ion of prostacyclin by the different layers of the vessel wall waseported by this group [26] showing this is highest in the endothe-ial layer, thus providing the endothelium for the first time a leadingole in the local control of vascular tone (Fig. 1). Still, and maybeonfounded by the fact the media and adventitia also released somerostacyclin [26], the notion of an endothelially-centered controlf vascular tone, could not be established. Ironically, another groupublishing a little bit later was unable to detect prostacyclin releaserom smooth muscle or any other cell studied but endothelial cells27]. If Vane’s group had failed to find prostacyclin synthesis in the

edia, possibly the leadership of the endothelial cell had becomearlier.

. Endothelium-derived relaxing factor

In the immediate years following Vane’s publications, the pro-uction of prostacyclin by the intimal or other vascular layers wasostly observed from the point of view of the anti-aggregating

roperties of the new prostaglandin able to counterbalancelatelets’ TXA2. From this angle, an endothelial damage wasegarded as a hemostatic rather than a vasomotor threat. Thus,xperimentation on vascular pharmacology was still carried outgnoring the endothelium. The first to change this scenario wereurchgott and Zawadszki with the publication in 1980 of “The oblig-

tory role of endothelial cells in the relaxation of arterial smoothuscle by acetylcholine” [6]. This paper was very remarkable in that

n a single seminal report, providing unequivocal experimental evi-ence, demonstrated that acetylcholine required the presence of

Research 113 (2016) 125–145 127

an intact endothelium to exert vasodilatory effects on the adjacentsmooth muscle (Fig. 2). Also that, via receptors on the endothelialcells, acetylcholine stimulates the release of a transferable sub-stance which causes relaxation of the vessel wall. For these reasons,this work became the major milestone for vascular pharmacologyand physiology research and was the most cited original paper inthis field (6569 citations). Interestingly, Furchgott admits that thefinding was quite accidental [28]: “. . .in the very first experimentof this planned series in May 1978, my technician did not follow mydirections correctly: [. . .] he tested carbachol for its contracting activ-ity before rather than after (as prescribed in the directions) washout ofa previous test dose of norepinephrine. The response to carbachol wasnot a contraction, but was a partial relaxation [. . .]. Acetylcholine wasthen tested. It too produced relaxation of the aortic preparation pre-contracted with norepinephrine. This was the first time that I had everobserved relaxation of rabbit aorta in response to muscarinic agonistsover the many years that I had been using this blood vessel. . .”. Theywere expecting acetylcholine to contract, but by mistake the tech-nician added the cholinergic agonist on top of an artery that shouldnot have been precontracted. To Furchgott’s surprise, a relaxationtook place. Indeed, acetylcholine had shown vascular contractileresponses in vitro since ever tested [29]. Still, Furchgott was con-scious and that, for unknown reasons, vasodilator responses toacetylcholine had been occasionally detected [30] and, more impor-tantly, that acetylcholine in vivo causes vasodilation [31]. The keyto understand how Furchgott suspected that the answer was in theendothelial cells relies on the fact that in that set of experiments,they used intact aortic rings instead of spiral strips. Normally, hecut the arterial strips in a helical fashion and then he measuredthe force attained by hooking the edges to the myograph devise.To facilitate the cutting procedure he (and most pharmacologists)systematically introduced a wire through the lumen of the arteries,unaware that the endothelium was being damaged or removed. Infact, still today, introducing a rod or similar is a conventional proce-dure to deendothelialize a vessel. This was their regular procedure,but not the day the technician erred, in which an intact ring wastested. The original recording of the serendipitous finding can befound in ref. 28. A preliminary report on these findings, prior to the1980 paper, was an abstract from a 1979 meeting [32] in which,surprisingly, he revealed too many clues of the discovery.

Very shortly after Furchgott & Zawadszki’s paper, a couple ofgroups corroborated the endothelium findings. De Mey & Van-houtte described the role of the intima in various responsesincluding cholinergic, purinergic and thrombin relaxations [33,34].This last paper started the listing of substances that elicit vasodi-lation via the endothelium and initiated the speculations on thenature of the endothelium-to-smooth muscle signals causing relax-ation. Another report published nearly at the same time was thatof Altura and Chand who were concerned about the unmatchedin vitro versus in vivo effects of some agonists, in this case, kinins[35]. A year later, Furchgott, also interested in the endothelialmechanisms of bradykinin, reached similar conclusions as thoseof Altura and Chand on the endothelium-dependency of the kinin’saction [36]. In this paper, Furchgott introduced for the first timethe term “endothelium-derived relaxing factor” and its acronym“EDRF” to refer to a substance released from endothelial cellsand able to relax the muscle. Substance P was also added to thegrowing list of hormones, which relax via endothelium. Anothernovelty of this paper was the first report of EDRF in human arter-ies. Furchgott’s group also provided the first preliminary insights

on the calcium-dependency of EDRF by examining the effects of thecalcium ionophore A23187 on the endothelium-dependent relax-ations [37], an issue that was later investigated in more depth [38].

128 E. Nava, S. Llorens / Pharmacological Research 113 (2016) 125–145

Fig. 2. Seminal experiments performed by Furchgott and Zawadzki led, beyond any doubt, to the conclusion that endothelial cells, upon acetylcholine stimulation, releasein a paracrine manner a substance that reaches smooth muscle cells to relax them (EDRF). Arteries were cut along and attached to a myograph. A, B and C represent variousvascular arrangements and D is the so-called “sandwich” vascular preparation that definitely proved that intact artery 1 supplied with a relaxing factor deendothelializeda teriesp to myt as add

5

vwdatiiissrtcfsoctesttvV

ciserA

rtery 2 (smooth muscle 2). Author’s micrographs of intact and deendothelialized arreparations. Note alignment of endothelial and smooth muscle cells with respect

o acetylcholine in rubbed strips have been omitted for simplicity. Acetylcholine w

. Endothelium-derived constrictor factor

The realization of the existence of endothelium-dependentasoconstriction took place early after Furchgott’s publication andas somehow linked to the discovery of EDRF. Possibly the mainifference was that the finding of EDRF was serendipitous in naturend the uncovering of the existence of the constrictor counterpartook part under a specific rationale: the attempt to clarify whyn some instances acetylcholine relaxed a vessel, and contractedn others. Vanhoutte and De Mey were studying the mechanismsnvolved in the unexplained relaxations to acetylcholine in certainet-ups. One of these was anoxia, which they knew from previoustudies that is able to augment catecholamine-induced contractileesponses. Just before the publication of Furchgott’s 1980 paper,hey found that anoxia abolished the relaxing responses to acetyl-holine [39]. Once knowing about the role of endothelial cellsrom Furchgott’s studies, they took up the topic and published ahort report which, in addition to confirming the obligatory rolesf the endothelium, they also indicated that the augmented cate-holamine contractile responses caused by subjecting the vesselso hypoxic conditions depended on the preservation of a functionalndothelium [40]. The notion that the endothelium contributes, notolely to smooth muscle relaxation, but in some instances, to con-ractile responses was matured in following studies [41] and finally,he first evidence that anoxia causes the release of a diffusibleasoconstrictor substance from endothelial cells, was published byanhoutte and Rubanyi in 1985 [42].

By the same time, Rubanyi together with Hickey and otherolleagues were also working on a bioassay system aimed todentify the chemical nature of EDRF. They worked in a bioassayet-up in which the donors of the relaxing factor were cultured

ndothelial cells and the detectors were luminally rubbed arterialings challenged with the medium from the endothelial culture.gainst expectations, what they found were not relaxations but,

stained with silver nitrate have been used to exemplify the different myographicalograph orientation. Pre-contraction with noradrenaline and eventual contractionsed in cumulative doses from 10−8 to 10−5 M.

instead, that the acceptor rings consistently contracted. They pur-sued the studies trying to uncover the nature of the transferablesubstance and found that the contracting agent was compatiblewith a polypeptide [43]. This was the prelude of endothelin, thename Yanagisawa and his colleagues decided to give to the pep-tidic vasoconstrictor present in the endothelial culture mediumonce purified and sequenced [44]. Before this, the peptidic natureof the vasoconstrictor reported by the Hickey group was confirmedby Gillespie et al. who introduced the acronym EDCF in the liter-ature to distinguish it from EDRF [45]. In a nice review publisheda few years ago, Rubanyi admits that the peptidic substance theydiscovered was nearly named “endotensin”, but unfortunately, thiswas discarded in the reviewing process of the manuscript [46] andit ended up being called endothelin [44].

In parallel to the endothelin work, Vanhoutte kept work-ing on the lines initiated with hypoxia. He did not believe thatendothelin was the only answer to the endothelium-dependentcontractions [47]. One of the lines started with a 1985 papershowing that endothelial cyclooxygenase generates vasoconstric-tor prostaglandins [48], which were specifically ascribed to TXA2in a following paper by Altiere and coworkers [49]. Shortlyafterwards, Lüscher published his classic paper for endothelial dys-function showing that in the aorta of hypertensive rats, and due toan excess in vasoconstrictor prostaglandins, acetylcholine causesendothelium-dependent contractions rather than relaxations [50].

During the following years Vanhoutte’s group was engagedin a quest to pinpoint the contractile prostaglandin. Also, theyembarked in a second line of research that introduced reac-tive oxygen species in the pathway of prostaglandins and theendothelium-dependent contractions. This line started rather cat-egorical, proposing superoxide anion as EDCF [51]. This was

not the first report on reactive oxygen species, prostanoids andvasoconstriction. It was already known that endothelial cells gen-erate superoxide [52] and a few years before, Tate and others

E. Nava, S. Llorens / Pharmacological Research 113 (2016) 125–145 129

Fig. 3. Depiction of endothelium-derived paracrine vasoactive factors commented and referenced in the text. Agonist (A), receptor (R), cytochrome P-450 epoxygenases(P450), cyclooxygenases (COX), endothelial NO synthase (eNOS), endothelin-converting enzyme (ECE), epoxyeicosatrienoic acids (EETs), thromboxane A2 (TXA2), prostacyclin(PGI2), TXA2 receptor (TP), PGI2 receptor (IP), calcium-dependent K+ channel (KCa), smooth muscle cell (SMC), endothelial cell (EC).Reference matching (first author followed by year abbreviation): a Hecker ‘94; Campbell ‘96 [78,79], b Svensson ‘75 [21], c Moncada ‘76 [4], d Miller ‘85; Altiere ‘86 [48,49],e Gluais ‘05; Rapoport ‘96 [59,62], f Parkington ‘93 [75], g Tateson ‘77 [25], h Katusic ‘89; Tate ‘84; Cosentino ‘94; Tang ‘07; Martinez-Revelles ‘13 [51,53–56], i Beny ‘91;M [149A Edwarm

rmlesriaabstdetcrmgrtlrsstoiaEr

f

atoba ‘00 [76,77], j Tare ‘90 [74], k Zawadzki ‘80; Singer ‘82 [37,38], l Palmer ‘89rnold ‘77, [106,107], o Katsuki ‘77b*; Kukovetz ‘79 [110,112], p Gordon ‘83 [67], q

uscle, which was tested before vascular smooth muscle.

eported that oxygen metabolites cause vasoconstriction of pul-onary arteries via TXA2 [53]. After the start-up of the superoxide

ine, numerous publications trying to elucidate the link betweenndothelial activation, oxygen species generation, prostaglandinynthesis and finally smooth muscle contraction, followed. Mosteports, such as that of Cosentino and coworkers [54] coincidedn various aspects: a) the endothelial cell, when activated withcetylcholine would generate oxygen radicals which would in turnctivate a cyclooxygenase that oversynthesizes all prostaglandins.) Neither TXA2 nor PGI2 seemed to be the prostaglandins respon-ible. c) Most papers concurred in that TXA2 receptor blockade onhe smooth muscle cell reverted the contractile process and finally,) the whole process, although present in healthy vessels, is exac-rbated in hypertension. The problem of the sequence of eventsaking place from the initial acetylcholine activation with an intra-ellular calcium increase to the final contraction of the vessel wasesumed a few years ago using state-of-the-art methodology. Byeans of intracellular imaging, it was demonstrated that cyclooxy-

enase activation occurs upstream of the generation of oxygenadicals [55]. One thing that was surprising in this report is thathe traditional sequence of events (oxygen radical generation fol-owed by cyclooxygenase activation) was inverted. A more recenteport, elegantly resolves the debate on the order of events byhowing the existence of a feedback loop between reactive oxygenpecies and cyclooxygenase-2. Under this scheme, one modulateshe other and vice versa [56] (Fig. 3). As stated before, the researchn prostaglandins as EDCFs continued in parallel with the superox-de topic, and often both coincided in the same paper. For instance,

1990 paper insisted in that oxygen-derived free radicals might beDCFs and would cause contraction by stimulation of TXA /PGH

2 2eceptors [57]. The latter was also found by Kato and others [58].

A breakthrough in the EDCF topic took place with the publicationeaturing “The Janus face of prostacyclin” [59]. This referred to the

], m Furchgott ‘80; Palmer ‘87; Ignarro ‘87; Khan ‘87 [6,126–128], n Katsuki ‘77a;ds ‘98 [68], r Chen ‘88 [70], s Hickey ‘85; Yanagisawa ‘88 [43,44]. *Tracheal smooth

double role of this, traditionally regarded vasodilator prostaglandin[23,60] which was then proposed as one of the EDCFs togetherwith other prostaglandins. This was not the first time this vasodila-tory prostaglandin had been attributed vasoconstrictive properties[61] or even proposed as the chemical underlying EDCF [62].What the Janus face paper introduced to the scientific commu-nity was the notion that all tested prostaglandins, including PGI2,elicit endothelium-dependent contractions by converging on TXA2receptors on the smooth muscle cell (Fig. 3). This happens inhealthy, but most especially in hypertensive arteries and explainswhy many papers that appeared during the previous decade hadpersistently found that cyclooxygenase inhibition or TXA2 recep-tor blockade, but not TXA2 synthase inhibition, curtailed theendothelium-dependent contractions [50,57,58,63].

6. Endothelium-derived hyperpolarizing factor

During the progress of EDRF/EDCF investigation, someresearchers realized that two facts did not fit in theendothelium—smooth muscle axis that had already been estab-lished. One was the observation that endothelium-dependentrelaxation relied on three rather than two systems. Thus, EDRF andprostacyclin could not explain all the relaxing phenomena [64].The second fact was the finding that the vasodilation elicited byacetylcholine could be detected, not only on the adjacent smoothmuscle cells, but also along the arterial wall at a distance thatexceeded the range of a mere paracrine effect [65]. To sort out thisproblem, the possibility that the endothelium causes electricalchanges in the nearby smooth muscle cells was considered. The

relationship between endothelial activity and myocyte hyper-porization/relaxation was instinctively connected to the existenceof released endothelium-derived hyperpolarizing substances,probably because of the enormous influence the EDRF milestone

1 ogical

htevekterficttKSfamtTatttivodahcpdosawvhuFa[

taoban(tEatpte[ccppeao

30 E. Nava, S. Llorens / Pharmacol

ad on the scientific community [6]. Before Furchgott’s discoveryhere was a confusion of reports in which acetylcholine causedither depolarization or hyperpolarization associated with eitherasoconstriction or vasodilation of smooth muscle cells. Severalxamples were referenced by Bolton [66]. In the light of the recentlynown role of endothelial cells, Gordon and Martin studied in 1983he effects of vasodilators on calcium-activated K+ efflux fromndothelial cells and its relationship with endothelium-dependentelaxation [67]. They found an increase in this efflux, which is therst step to induce a hyperpolarization, at least of the endothelialell (Fig. 3). Although they did not assess membrane potentials,hey did suggest that this cationic efflux might have some rela-ionship with the relaxations. This was the prelude of the idea of+ as the EDHF that still took some years to be considered [68].hortly after the 1983 study, Bolton and colleagues went a stepurther showing that the inhibitory mechanical effect of carbachol,

stable cholinergic agonist similar to acetylcholine, on smoothuscle was associated to hyperpolarization of these cells. This

urned into depolarization when the endothelium was removed.hey were the first to mention the possibility of the existence of

factor released by endothelial cells causing hyperpolarization ofhe vessel wall [66]. With all this knowledge in mind, and that ofheir forthcoming publications, Weston and others introduced tohe scientific public the term endothelium-derived hyperpolar-zing factor or its acronym, EDHF [69]. This took place in 1988, aery productive year for EDHF. For example, this year Chen andthers demonstrated the existence of a hyperpolarizing substance,istinct from EDRF, released upon acetylcholine stimulation. Using

K+ efflux marker, concluded that this substance causes myocyteyperpolarization by opening K+ channels on the smooth muscleell [70] (Fig. 3). Next, Olesen and others also checked membraneotentials, but not in the myocyte but in endothelial cells, andemonstrated the existence of acetylcholine-activated (receptor-perated) channels on these cells. They showed that cholinergictimulation causes hyperpolarization of the endothelial membranend suggested that this endothelial hyperpolarization somehowould trigger the characteristic events of endothelium-dependent

asodilator responses [71]. The real link between endothelialyperpolarization and myocyte hyperpolarization could not benderstood until much later [68]. Finally, during this year of 1988,élétou and Vanhoutte bioassayed the hyperpolarizing substancesnd provided the first evidence of the diffusible properties of EDHF72].

From then on, a quest aimed to find out the chemical nature ofhe EDHF(s) begun. By this time, EDRF had been already identifieds NO, so this was the first proposed molecule. NO was initially ruledut [73], but later reconsidered by Tare and others proving that NOlockers reduce acetylcholine-elicited hyperpolarization in somerteries [74]. Sometime later, these researchers demonstrated thatot only NO, but also prostacyclin, exhibit hyperpolarizing abilitiesFig. 3). Their proposal was that both of them contribute to the elec-rical phenomenon but without precluding the existence of a majorDHF [75]. Other candidates on the line were hydrogen peroxidend epoxyeicosatrienoic acids (EETs). Initial evidence showed thathe oxygen-derived species, hydrogen peroxide, exhibits hyper-olarizing properties on myocytes but failed to pass the catalaseest (forced conversion of hydrogen peroxide into water by thisnzyme should inhibit endothelium-dependent hyperpolarization)76]. However, Matoba and colleagues, nearly a decade later, suc-eeded in this aspect and presented hydrogen peroxide as a firmandidate for EDHF in a sequence of papers, the first of which wasublished in 2000 [77] (Fig. 3). Another possible EDHF was pro-

osed in 1994. Hecker and others, on the basis of myographicalvidences, suggested that the cytochrome P450 derivatives, EETsct as EDHFs [78]. A subsequent report that included the assessmentf membrane potential changes confirmed Hecker’s proposal [79]

Research 113 (2016) 125–145

(Fig. 3). A large number of possible substances with endothelium-dependent hyperpolarizing properties has been suggested over theyears and the whole coverage exceeds this revision. There are goodreviews on this specific topic, for example, ref. [80].

In 1998 Edwards and coworkers [68] picked up the thread of thework on endothelial K+ efflux initiated by Gordon and Martin [67]and followed by Olesen’s analyses of endothelial membrane poten-tials [71]. They showed that hyperpolarization of smooth musclecells might be caused by endothelium-derived K+ and performeda complete study of the endothelial channel source of this cationand the sites of action on the myocytes. Specifically, they confirmedthe existence of a K+ efflux from endothelial cells upon choliner-gic activation via endothelial calcium-dependent K+ channels. Theyalso demonstrated that this ionized potassium present in the extra-cellular space acts as the EDHF by activating inwardly rectifying K+

channels and a Na+/K+-ATPase of the smooth muscle cell mem-brane. In Edward’s model, acetylcholine stimulation involves notonly, the hyperpolarization of smooth muscle, but also the hyper-polarization of the endothelium (Fig. 3). This simple idea of ionizedpotassium being an EDHF strongly reminds the well known text-book concept of local blood flow regulation, in which K+ along withother metabolites cause local vasodilation. Indeed, the notion thatpotassium in low concentrations is a vasodilator can be dated backas early as 1938 [81].

With all the mentioned EDHF possibilities, the problem of thethird vasodilating factor we started with [64] had been addressed.However, the “remote” acetylcholine vascular relaxation describedby Segal and Duling [65], did not have a satisfactory answer ifone counts only with the concept of hyperpolarization of adjacentcells. The only apparent way to solve the obstacle was to con-template the possibility that this change in membrane potential,either of the endothelial cell or the myocyte, could be electro-tonically transmitted to neighbouring cells. This is only possibleif cell membranes join together in low resistance discrete loca-tions called gap junctions. In the case of endothelial to myocytejunctions, the structural basis involves the approach of smoothmuscle cell membranes towards those of endothelial cells. Thiswas first described by Thoma in the early twenties who claimedto have observed cytoplasmic protrusions passing through fenes-trae of elastic lamina which apparently contacted smooth musclecells [82]. The advent of electron microscopy imaging in the fiftiesconfirmed these pioneering observations [83] and a decade laterRhodin, with improved electron microscopy images observed ingreat detail these protrusions which he named “myoendothelialjunctions”. He also suggested that myoendothelial junctions mayserve as a rapid communicating pathway between circulating hor-mones and vascular smooth muscle via endothelial cells: “. . .Itwould therefore be quite advantageous if some endothelial cells couldserve as receptors [of humoral transmitter substances] which initiatethe depolarization of the smooth muscle membrane. . .”. Thus, Rhodinanticipated a functional role for these junctions [13]. Gap junctionsare present in numerous cell types and have been detected betweenadjacent endothelial cells [84] and adjacent arterial smooth musclecells [85]. The mere approach of two membranes does not imply theexistence of a low resistance electrotonic connection. Gap junctionsexhibit a characteristic pentalaminar pattern in electron micro-scope plates. Thus, bridges connecting endothelium and myocytescan only qualify as an authentic myoendothelial gap junction whenthis pattern is observed. In the case of myoendothelial connections,his was first described by Spagnoli and coworkers [86].

The ultrastructure of gap junctions involves the docking of twomembrane channels, each formed by six proteins. These were first

characterized in 1974 by Goodenough who resolved naming then“connexins” [87], inspired by the name “nexus”, formerly given togap junctions before this term was generally utilized [88]. A vari-ety of connexins have been described over the years. Blood vessels

ogical

([cati

tScSwbpnhdrstIrwseee

7N

wAscbamenhlnTc

n[ucM(noWmIncw[m

E. Nava, S. Llorens / Pharmacol

intima and/or media) specifically express connexins 43 [89], 4090], 37 [91] and 45 [92] (chronologically ordered as described,itations refer to the first description published). In 2006, Haddocknd colleagues localized, by means of high resolution immunohis-ochemistry, connexins 37 and 40 to myoendothelial gap junctionsn brain small vessels [93].

The electrotonical transmition of hyperpolarization from cello cell along the arterial wall was explored in profound detail byegal and Duling’s group during the years following their acetyl-holine report [65]. The whole picture was beautifully reviewed byegal in recent times [94]. Therefore, the problem of transmissionas solved. Moreover, it was an incentive for some to go further

y suggesting that, with functional myoendothelial gap junctionsresent, perhaps the existence of a diffusible factor (EDHF) was notecessary after all. For example, Chaytor and coworkers stronglyighlighted the role of these membrane structures in endothelium-ependent relaxations [95]. Actually, these junction-transmittedelaxations have been referred to as “EDHF-type” [96], for theirtrong resemblance to those mediated by the released factor, andhe term EDH, rather than EDHF, is currently recommended [97].n any case, these forms of endothelium-dependent relaxations,eliant on myoendothelial gap junctions are not, stricto sensu,ithin the scope of the title “paracrine control” of this review. Fig. 4

ummarizes the electrical, paracrinal and mechanical phenom-na using tracings from different sources [68,98] and consideringndothelial K+ efflux as an example of an EDHF released fromndothelial cells upon acetylcholine stimulation.

. Search for the chemical nature of EDRF.itrovasodilators and cyclic GMP. The link EDRF-NO

Most of the period dedicated to EDCF and EDHF researchas also devoted to try to unveil the chemical nature of EDRF.ctually, Furchgott’s 1980 report included also a systematictudy of most of the possible candidates for EDRF (ruling outyclooxygenase-dependent products and specifically prostacyclin,radykinin, adenosine, cAMP and cGMP) and part of the arachidoniccid/lipoxygenase line mentioned above [6]. The research on theolecule responsible for endothelium-dependent relaxations was

specially focused on the possibility that EDRF could be a lipoxyge-ase product of arachidonic acid [99–102]. By the mid-eighties thisypothesis had definitely reached a dead-end. There was another

ine of evidence that, looking at it from the perspective we haveow, appears rather obvious, but it took years to be considered.his was the line of nitrovasodilators and guanylate cyclase thatommenced a decade earlier.

In the mid-seventies various groups were working on cyclicucleotides, especially the recently discovered cyclic GMP (cGMP)103]. In the process of characterization of the soluble and partic-late fractions of guanylate cyclase (the enzyme that synthesizesGMP from GTP) a number of chemicals were used. Among these,urad’s group found that nitrogen compounds, such as azide

NaN3) or nitrite (NO2Na) activated guanylate cyclase [103]. Otheritrocompounds were also reported to activate the formationf cGMP, such as nitroglycerin [104] and nitrosamines [105].hat attracted Murad’s attention was that these nitrogen-bearingolecules acted on guanylate cyclase in a hormone-like fashion.

n the course of this research, his group enlarged the variety ofitrogen compounds that activated guanylate cyclase [106] and dis-

overed that all these compounds yielded nitric oxide (NO) [106]hich strongly activated guanylate cyclase and generated cGMP

106,107]. Some of these nitrocompounds were drugs known forore than a century for their anti-anginal [108] and vasodilatory

Research 113 (2016) 125–145 131

properties [109]. Actually, and in parallel with the biochemicalresearch, the abilities of these compounds to relax various smoothmuscles and its relationship with cGMP was also being investigated[110–112]. Indeed, it was then when the term “nitrovasodilator”was introduced by Murad himself to refer to the NO-generatingdrugs that elicited vasodilation [113]. By the end of the decade, thebiochemistry of guanylate cyclase activation by NO and nitroso-compounds and its connection with smooth muscle relaxation wasquite well characterized. However, the link of NO with hormones,sought by Murad and the biological significance of the relationshipbetween NO and guanylate cyclase was still an absolute enigma asadmitted by himself [114].

During the search for the chemical identity of EDRF over theearly eighties, biochemical and physiological coincidences betweenthe properties of NO and EDRF were being continuously reported.Even before the existence of EDRF was known, it was noticed thatcarbachol, an agonist similar to acetylcholine, increased the tis-sue levels of cGMP [104]. Once in the context of EDRF research, itwas reported that this transferable factor, as stimulated by acetyl-choline, is capable of increasing the release of cGMP from arterialstrips [115]. Indeed, Murad forwarded the hypothesis that theeffects of EDRF are mediated by cGMP [116]. A year later, Ignarroreported that the effects of both, nitrovasodilators and EDRF onvascular relaxation and cGMP accumulation was antagonized bythe inhibitor of guanylate cyclase, methylene blue [117]. Shortlylater, the same was published for hemoglobin [118]. In the sub-sequent year, three suggestive papers shed light on the reasonwhy the lipoxygenase hypothesis for EDRF was being so confound-ing. Two of them demonstrated that superoxide anions destroyEDRF [119,120]. The other, directly ascribed the effects of lipoxy-genase inhibitors to their redox properties rather than to anyspecific action on EDRF [121]. Coincidences between the biolog-ical actions of nitrovasodilators and those of EDRF, such as thatboth inhibit platelet aggregation [122,123], were being reported.Perhaps the most peculiar coincidence was Rapoport’s finding thatboth, nitrovasodilators and EDRF, induce desensitization associatedwith a decrease in the formation of cGMP [124]. More precisely,EDRF (as stimulated by acetylcholine) elicited lesser relaxing effectsafter exposure to a nitrovasodilator. This meant that EDRF wasdesensitized by the nitrovasodilator and thus, necessarily held acommon underlying mechanism.

It is remarkable that with all this mass of evidence, the scientificcommunity was so reluctant to admit that EDRF was a nitrova-sodilator. An example of this reluctance is that even the nucleotidecGMP was proposed as a candidate for EDRF’s chemical nature[115]. Still, in 1983 Furchgott wrote: “. . .EDRF, after diffusing fromthe endothelial cells [. . .], stimulates the muscle guanylate cyclase[. . .]. This speculation assumes a causal relationship between increasesin cGMP levels and relaxation in vascular smooth muscle. Evidenceconsistent with such a causal relationship [. . .] by nitric oxide [. . .]has recently been presented [. . .]. However, [. . .], the case for a causalrelationship has not yet been proven. . . ” [125]. The reason why ittook so long to perform the experiments that proved this relation-ship was probably that never before a molecule of such a smallmolecular weight had been regarded as a biological mediator, andeven less when the putative mediator, NO, is a toxic gas as well.

Palmer and coworkers first reported that NO accounts for thebiological activity of EDRF in 1987 [126]. Shortly later, a paperby Ignarro [127] and a bookchapter by Khan and Furchgott [128],showing fundamentally the same, appeared in the press. Actually,Ignarro’s group published numerous papers on the chemical natureof EDRF as NO nearly at the same time [127,129–131], thus indicat-

ing that the three groups were concomitantly working in the sametopic (Fig. 3).

132 E. Nava, S. Llorens / Pharmacological Research 113 (2016) 125–145

Fig. 4. Electrical, paracrine and mechanical phenomena occurring upon acetylcholine administration on endothelial and smooth muscle cell membrane potentials (Vmb), K+

concentration assessed in the myo-endothelial area and arteriolar diameter. Tracings have been placed in the order of progress of events. Endothelial and smooth musclec ed froo of an

( ce art

8

[osbwsCbtoiteosddnwN[

sNrsboaa(hup

ell membrane potentials and arteriolar diameter are synchronous tracings modifin muscle feed arteries. K+ concentration, used in this case to illustrate an examplemodified). Acetylcholine was administered in a regular manner on hepatic resistan

. The NO pathway

The publication of the NO identity of EDRF by Moncada’s team126] came from a group that had not worked before in the fieldf nitrovasodilators but with prostaglandins, as stated in previousections. Moncada’s group took advantage of their expertise in theioassay cascade preparation [4], which was powerfully reinforcedith endothelial cells cultured on beads. They had to develop new

pecific methodology to assess NO release from endothelial cells.uriously, this methodology came from the use of nitrite sensorsased on chemiluminescence adapted from the motor industry. Allhis finally allowed proving that the pharmacological propertiesf EDRF and NO are indistinguishable. This was nicely reviewedn more recent years [60]. Despite the group was a newcomer inhe nitrovasodilator field, rapidly took the lead, especially unrav-lling the biochemical pathway of NO synthesis and the originf NO as synthesized from L-arginine. The discovery served as atrong incentive to boost the research in the NO field as seen by theramatic rise in publications over the following years (Fig. 5). Evi-ence was sufficient to recognize NO as the endothelial endogenousitrovasodilator, in the sense Murad gave to this term [113]. This,hen specifically referring to the vessel wall [132] but, in general,O was regarded as a ubiquitous mediator of cell communication

133].Especially relevant to NO exploration was the research on the

ynthesis of NO from L-arginine that led to the establishment of theO pathway. This discovery had a long history behind of various

eports on scientific topics unrelated with vascular biology and inome instances unrelated between each other. It just needed some-ody to put them together and link them with EDRF. The sequencef reports that led to the l-arginine-NO pathway started up with

topic initially unrelated with this pathway: nitrate, nitrosaminesnd cancer. There was a long lasting debate on whether nitrate

−

NO3 ) is produced endogenously by mammalian cells and if so,ow. Measurements performed early last century indicated thatrinary nitrate had to be synthesized within the body albeit noathway was known for the synthesis of this anion [134]. The

m Emerson and Segal [88]. Acetylcholine was administered by microiontophoresisEDHF, was measured in the myo-endothelial space by Edwards and coworkers [68]eries.

topic was revisited in the late seventies by scientists concernedabout the possibility that the exposure to nitrate could yield tothe fearsome compound, nitrosamine, clearly related to stom-ach cancer. Initially it was hypothesized that this endogenouslyformed nitrate was the result of the metabolism of the intesti-nal flora [135], but experiments on germ-free rats showed that,definitely, mammals were able to make themselves nitrate [136].The following year, a clinical paper appeared reporting the exis-tence of children with methaemoglobinemia, a condition caused,among other factors, by high environmental exposure to nitritesor nitrates [137]. These children exhibited a very high concen-tration of plasma nitrate and severe methaemoglobinemia. Theseobservations would have nothing to do with the present topic onthe NO pathway except for the fact that those children had neverbeen exposed to water or food-containing nitrates. What all thoseinfants had in common was that all of them suffered an acute diar-rhoea. The possibility that a bacterial gastroenteritis was behindthe appearance of this increased endogenously-synthesized nitrateconcentration prompted Tannenbaum’s group to check this pointin animals challenged with bacterial endotoxin. The result was adramatic rise in nitrate production [138]. Because this appeared tobe an immune reaction, macrophages were the first candidate cellsto be examined. Stuehr and Marletta performed the experiment inthese cells and found a sharp rise in nitrate release taking placewhen cultured macrophages were treated with endotoxin [139].

At the same time another group was carrying out a line ofresearch, which was also related with the immune system. Hibbs,the senior of this group, had been interested in the interaction ofmacrophages with tumour cells and the biochemistry of the acti-vated macrophage mechanisms of cytotoxicity. It had already beenobserved a long time before that arginine inhibited the growthof tumours [140]. Aware of this, Hibbs, during the course of thebiochemical studies, observed that macrophages failed to become

active upon immunological stimulation if no amino acids werepresent, but fully able to attack neoplastic cells if l-arginine, butnot the dextrorotary form, was present in the medium [141].Intrigued with this finding, the group went on studying in depth

E. Nava, S. Llorens / Pharmacological Research 113 (2016) 125–145 133

of the

tmtaunspnr

lENpttetttvwctcc[b

pedatltoshat

Fig. 5. Pubmed impact of the main topics

he biochemistry involved. In this initial report, they found thatethylated analogues of l-arginine inhibited macrophage induc-

ion. Further research on the arginine issue, and already knowingbout the nitrate release from macrophages upon endotoxin stim-lation [139], led the group to the proposal that the origin of thatitrate could be l-arginine [142]. In this paper [142], Hibbs demon-trated that l-arginine was oxidized to nitrite and l-citrulline, androposed a preliminary biochemical pathway, which he recog-ized, had never been seen before in living cells [142]. Still, theseesults did not link l-arginine with NO yet.

A piece of evidence that led to the association between nitrate,-arginine and EDRF was disclosed because after the discovery ofDRF as NO, many groups initiated a research campaign looking forO synthesis, not only in the endothelium, which was the startingoint, but also elsewhere in different organs and systems. One ofhese was the central nervous system. Inevitably, this search ledo a report published some years before by Deguchi and cowork-rs on the activation of guanylate cyclase in brain cells [143]. Byhe year this research was being carried out, Deguchi was awarehat nitroso-compounds could activate guanylate cyclase, but whathey were trying to find was an endogenous substance able to acti-ate this enzyme. They had a strong background in fine chemistry,hich they utilized. The result was that they were able to specifi-

ally purify L-arginine from the brain extract. They even comparedhe activation pattern of guanylate cyclase with that of nitroso-ompounds and noticed the similarities. They also pointed out thehemical resemblance of L-arginine and some nitroso-compounds143]. They did not know about NO or any synthesizing pathway,ut the data they had was remarkably close to it.

It can be seen that all the reports we have mentioned wereublished over a very long period of time, they were appar-ntly unrelated between each other and published in journalsevoted to different topics. Specifically selecting the right paperst a time when computer search engines were practically inexis-ent and putting them together in the right order appears todayike looking for a needle in a haystack. However, this seemedo be no handicap and it did not take long after the discoveryf EDRF’s chemical nature to identify L-arginine as the precur-

or of NO. The first report came from Moncada’s lab [144] whoad focused part of his research in the brain NO synthase [145]nd noticed Deguchi’s report. At the same time, they had a longime expertise in the immunology/inflammation area [20,23] that

present review over the last forty years.

helped joining together the brain guanylate cyclase report withthe different findings published earlier on endogenous nitrates andl-arginine dependency of macrophage activation. Still, and onceagain, the same idea was pursued independently by other groupswho published shortly after. One group confirming l-arginine asthe precursor of EDRF [146] and another demonstrating that NOis the intermediate molecule between L-arginine and nitrate [147]previously described in macrophages [142]. Fig. 6 summarizes thethree main lines of research that yielded to the first step of thel-arginine to NO biochemical pathway.

Quickly, l-arginine methylated analogues, already used byHibbs [141], were tested on endothelium-dependent relaxations toinhibit NO synthesis [148] and the first partial characterization ofthe enzyme, which was later named endothelial NO synthase, wasperformed [149] (Fig. 3). A major implication for research of theavailability of these analogues was the possibility to leap from thesingle vessel [148] to a whole organ [150] and finally to the entireindividual [151]. These two publications proved that NO, basallyreleased, has a permanent vasodilatory action that sets perfusionpressure [150] and even the entire blood pressure [151] to a physi-ological point that is very well below the pressure that would existwithout NO. These observations led to the notion of the vascularsystem as a conductance, rather than a resistance system and areclearly explained in a review published later on by Moncada [60].

All these advances ignited a rise in NO research that peaked in1992 (Fig. 5). NO synthases were found in numerous tissues besidesthe vessel wall (for review see ref. [152]) and the 1991 revisionpublished in Pharmacological Reviews [153] became the highestcited paper ever in vascular pharmacology (12096 citations in arecent search and followed by Furchgott’s 1980 paper with 6569citations). The following year, NO was awarded the “Molecule ofthe Year 1992” [154], and finally, the 1998 Nobel Prize in Physiologyor Medicine was awarded for the discoveries concerning NO as asignalling molecule in the cardiovascular system.

After 1992, research in the NO field was vertiginous. To mentionjust a couple of examples on how the newly incorporated biochem-ical route of l-arginine affected the map of biochemistry known bythat time, it is worth making an allusion to endothelial arginase and

endogenous arginine derivatives. Before the l-arginine-NO path-way was incorporated to the textbooks, the hepatic urea cycle wasthe major known destiny of arginine via the arginase reaction.It was eventually noticed that endothelial cells possess arginase

134 E. Nava, S. Llorens / Pharmacological Research 113 (2016) 125–145

F ine rec

arrttrnobwNTditse

vcwufpdbicwsseh

ig. 6. Three unrelated lines of research (nitrovasodilator, nitrate/nitrite and arginoncurrent publications are cited in publishing order.

ctivity and produce urea [155]. It was suggested [156], and moreecently proven [157], that the role of endothelial arginase is toegulate the substrate provision for endothelial NO synthase and,hus, NO-dependent relaxation. Another remarkable aftereffect ofhe newly acquired knowledge in arginine biochemistry was theecovery of very old reports on the presence of endogenous argi-ine derivatives in the urine [158]. This was of enormous interestnce known that l-arginine analogues inhibit NO synthesis [148]ecause it represented the possibility that NO synthase inhibitorsere, not only drugs, but physiological actors of the l-arginine-O pathway, the so-called endogenous inhibitors of NO synthase.he first of these, asymmetrical dimethylarginine (ADMA), wasetected by Vallance and colleagues in human urine [159] and later

n human endothelial cells [160]. These two natural controllers ofhe NO pathway, endothelial arginase and ADMA, have been exten-ively studied from a cardiovascular disease point of view thatxceeds the limits of our review.

At this time point, the 3 major endothelium-derived paracrineasoactive substances (EDRFs, EDCFs and EDHFs) were perfectlyatalogued (Fig. 3). One aspect that tempted vascular physiologistsas to compare each of them in different vascular beds, most partic-larly, in conductance versus resistance vessels. Furchgott verifiedrom the very beggining of his research that EDRF release takeslace, not only in large conductance vessels where he made theiscovery [6], but also in arteries small enough to control regionallood flow [161]. Once NO synthase inhibitors became available,

t did not take long to realize that these inhibitors were more effi-ient in large compared to smaller arteries [162], something thatas ascribed to a differential role of NO versus EDHF in large and

mall arteries, where the importance of EDHF increases as the ves-el’s calibre diminishes [163]. This notion confirmed the earlierlectrophysiologist’s finding that microvascular endothelial cellsyperpolarize upon agonist stimulation more efficiently than the

search) led to the discovery of the L-arginine to NO pathway in 1988. Same year

macrovascular do [164]. In practice, however, both NO and EDHFhave demonstrated to be key elements of the circulatory controlin vivo [151,165].

9. Adventitia and periadventitial adipose tissue

It can be noticed that, so far, the leading character in theparacrine control of the vessel’s function has been exclusivelyattributed to the endothelium, with NO at the front as the canon-ical vasodilator. The abluminal layer, the adventitia, had beengenerally overlooked. This scenario is now changing. While thetextbook paradigm holds that the vascular wall consists of tuni-cae intima, media and adventitia, a forth layer, the tunica adiposahas been recently claimed [166,167] on the basis of the clear influ-ence that periadventitial adipose tissue has on vascular function.Thus, perivascular adipose tissue (PVAT), up to now regarded as“extravascular”, is being increasingly considered an “intravascu-lar” structure. While some talk about four layers [166–168], othersprefer to consider the existence of three, and divide the thirdone in “adventitial compacta” and “adventitial fat” or PVAT [169].The “compacta” is the classical adventitia and is made mostlyby fibroblasts, but also macrophages, dendritic cells, progenitorcells, vasa vasorum endothelial cells, pericytes and nerve end-ings [170]. The tunica adiposa or PVAT is mostly formed by whiteor brown adipocytes (depending on the depot), but also by theso-called stromal vascular fraction, which includes preadipocytes,fibroblasts, mesenchymal stem cells, lymphocytes, macrophages,nerve endings and vasa vasorum endothelial cells [171]. As for thenerve endings, these are loose varicose sympathetic or parasympa-

thetic autonomous fibres. The two may innervate smooth musclecells but adipocytes are mainly sympathetically innervated [172].Whether adipocytes receive parasympathetic innervation is a mat-ter of intense debate [173,174]. To avoid confusions between both

ogical

ltt

cditrmratr

1

tttiarva[iarsodaSiofimaeb

oiaclfer[ahtt

1

ahmpt

adipose depot [196–199] and this confers differences in the abilityto control the underlying vessel [200]. This reminds very much thewell know heterogeneity the endothelium exhibits in the kind of

1 A dysfunctional PVAT has hypertrophied, lipotoxic adipocytes, a condition gen-erally occurring in visceral adipose tissue upon a lipid overflow due to the positive

E. Nava, S. Llorens / Pharmacol

ayers of the adventitia, in this review we will simply name “adven-itia” to the tunica adventitia stricto sensu and “PVAT” to the newunica adiposa or outermost limits of the vessel.

The reason why the adventitia has received little pharmacologi-al attention as compared to the endothelium relies on the technicalifficulties to remove it properly [175] so as to investigate its roles

n a similar fashion as it was done with the endothelium. In con-rast, PVAT is fairly easy to remove and, paradoxically, this is theeason why it took so long to notice its vasoactive roles. In the largeajority of myographic studies perivascular fat is removed on a

outine basis. This practice is based on the assumption that PVAT is extraneous tissue and, worse, may impede the diffusion of vasoac-ive substances to the “authentic” vascular tissue, as explained byesearchers in the PVAT field [176].

0. The paracrine role of the adventitia

If the endothelium was once considered a “tapestry”, the adven-itia has been generally regarded as a mere connective support ofhe vessel with no active role to be sought in. Possibly the first studyo introduce an active role on vascular function was the discoveryn 1995 of a powerful superoxide-generating system within thedventitia [10]. However, prior to this year, few scattered studiesevealed some abilities of adventitial cells with regard to variousasoactive substances such as prostacyclin [26], histamine [177],ngiotensin converting enzyme [178], traces of angiotensinogen179] or serotonin [180]. Shortly before the adventitial superox-de preliminary finding [10], a suggestive study was published on

possible functional activity of adventitia, in this case as a bar-ier specifically arranged against endothelial NO [181]. Pagano’study on adventitial superoxide generation was followed by a seriesf reports that increasingly clarified this topic: the NADPH oxi-ase, responsible for superoxide generation, was pinpointed to thedventitial fibroblasts [182] and the functional barrier described byteinhorn and others was imputed to this superoxide for its role innactivating NO [183]. In 1999, it was stated that adventitial super-xide does play a paracrine role [184]. Therefore, the adventitialbroblast began to qualify as a paracrine cell, at least in what vaso-otion is concerned. In the meanwhile, it was reported that the

dventitia is able to generate NO, but only when challenged withndotoxin [185,186]. It must be remarked that fibroblasts seem toe unable to produce NO without prior induction [187].

The technical difficulties to obtain vascular preparations devoidf adventitial cells have been overcome by various means. Ones a curious set-up in which the artery ring is pharmacologicallyssessed inside-out with the adventitia facings outwards [183] orombining physical removal with collagenase treatment [188]. Thisatter method permitted to observe that the adventitial layer allowsor stronger smooth muscle contractions (Fig. 7D) and smallerndothelium-dependent relaxations, thus clearly establishing aole for the adventitial layer in the regulation of the vascular wall188]. The impact the adventitia has on endothelial function wasgain ascribed to superoxide’s action [189] and more recently toydrogen peroxide [190]. Additionally, it has been demonstratedhat fibroblasts produce endothelin-1, thus including this peptideo the short list of adventitial paracrine vasoactive substances [191].

1. The paracrine role of PVAT

Before today’s interest in PVAT became commonplace,dipocytes were already known to synthesize and release

ormones related to energy homeostasis and, together withacrophages, operate as immune cells. This, due to their ability to

roduce cytokines, in this case, adipocytokines, which are largelyhe same ones as those produced by genuine immune cells [192].

Research 113 (2016) 125–145 135

The novelty actually was the suggestive possibility that adipocytesacted as paracrine cells to the vessel wall by releasing transfer-able substances with vasomotor abilities. Obviously, this is onlypossible when adipose tissue is close to the artery, a circumstancethat only happens in the case of adventitial fat, that is, PVAT. Inthis sense, adipocytes would operate in conjunction with intimalendothelial cells. This idea was materialized by Löhn and colleagueswho first demonstrated that the adipose tissue surrounding theadventitia is able to release a transferable factor that affects vascu-lar tone. In 2002 they published the existence of a diffusible factorderived from PVAT, which received the evocative name of ADRF, theacronym of “adventitium-derived relaxing factor” [9]. In a follow-ing paper of the same group, they decided that the “A” should standfor adipocyte instead of adventitium [193] and from then on, bothnames have been used indistinctly, with the common acronym ofADRF. In these first reports, they demonstrated that dose-responsecurves to various vasoconstrictors (phenylephrine, angiotensin IIand serotonin) were systematically smaller when the arterial seg-ment was tested without dissecting off the perivascular fat (Fig. 7C).They also showed that Krebs buffer collected from an artery incu-bated also in these conditions, i.e. with PVAT, was able to elicit arelaxing response when added to a segment fully dissected andprecontracted with one of the vasoconstrictors [9,193].

In the following lines we will review the historic progress ofthis idea, which has yielded to a new vision of the vessel, bothfunctional (perivascular adipocyte working in conjunction with theendothelial cell) and morphological (the recent inclusion of PVATas the fourth layer of the vessel’s histology) [166–169]. Even patho-physiological: a “perivascular adipose tissue dysfunction” has beenproposed [194].1 Furthermore, the ultrasonographic measurementof the adventitia/media thickness (which includes PVAT) has beenproposed as an alternative to the traditional intima/media thick-ness to predict cardiovascular risk [195].

Although a great deal of enthusiasm has been placed on PVATas an alternative paracrine source of the vessel wall, before goingon, it is important to emphazise that there are major differencesbetween the endothelial tissue and PVAT. Aside from the obviousfact that endothelial cells are directly subjected to the influence ofthe blood stream, the first dissimilarity between both tissues is thatintimal endothelial cells are (or should be) always present in what-ever kind of vessel and in the entire circulation. In contrast, PVATis absent in numerous vessels. Large conduit arteries and veins aregenerally surrounded by the adipose tissue they run through, butthis is normally left behind once they penetrate into an organ. Cere-bral blood vessels typically lack PVAT. Additionally, if we considerthat the whole pulmonary and systemic capillaries (except thosethat irrigate fat depots) are completely devoid of adipose tissue,this relegates PVAT to a relatively small proportion of the circula-tory system. A second difference is that, while PVAT is comprisedby numerous cell types (see classification in ref. 171), the tunicaintima is, in physiological conditions, made exclusively by endothe-lial cells and some subendothelial fibroblasts. On the other hand,both tissues have things in common. For example, PVAT has largephenotypic differences depending on the specific location of a given

energy balance of obesity. These visceral adipocytes exhibit an abnormal pattern ofadipokine secretion. The extrapolation of this condition to the adventitial adipocytesmakes PVAT a dysfunctional one in that there is an excess in adipocytokine secretionand an undesirable release of vasoactive adipokines both with unhealthy conse-quences for vascular function.

136 E. Nava, S. Llorens / Pharmacological Research 113 (2016) 125–145

Fig. 7. Contractile responses obtained with noradrenaline or phenylephrine by Soltis and Cassis (A and B), Löhn and colleagues (C) and Somoza and others (D) in aortic (A, Band C) and carotid segments (D) showing: 1) PVAT’s nerve endings noradrenaline recaption effect (A and B), 2) the anticontractile effect of PVAT (C) and the procontractileeffect of adventitia. Observe that Soltis and Cassis did not find any effect of PVAT on phenylephrine contractions, an observation attributed to the lack of recaption by PVAT’snerve endings (B). This recaption does take place when the vessel is challenged with noradrenaline (A), but not with phenylephrine (B). Löhn and colleagues were able tod henyt ll as c(

sppat

opmAcao(dmtcitootwdt

emonstrate the release of an anticontractile substance from PVAT in response to punica adventitia exhibits procontractile properties in the same artery [188] as weD). + = tissue present. − = tissue removed.

ubstances released depending on the vascular region [41]. Anotherroperty shared with endothelial cells is the ability to synthesizerostaglandins and NO. General adipose tissue releases NO [201]nd prostaglandins [202]. PVAT in particular, also releases thesewo substances [203,204].

The first description of the modulatory action of perivascular fatn vascular contractility was reported eleven years before Löhn’sublication, by Soltis and Cassis [8]. This work has been sometimesisinterpreted as the first postulator of a prorelaxing role of PVAT.ctually, what these researchers describe is PVAT as the physicalarrier of sympathetic recaption by these nerve endings, whichre present both in the adventitia and in PVAT. Soltis and Cassisbserved an abolishment of noradrenaline-mediated contractionselectric or tyramine stimulation) in the absence of PVAT and aecrease in the sensitivity to noradrenaline itself when aortic seg-ents have an intact PVAT (Fig. 7A). They elegantly demonstrated

hat this is caused by the uptake and removal of this particularatecholamine by adipose tissue and suggested that the nerve end-ngs within PVAT recapt and eliminate noradrenaline present inhe synaptic gap. This, understandably, ends up in a buffered effectf this effector, but they never stated that PVAT releases any sortf relaxing or anticontractile factor. In fact, they did demonstrate

hat the experiment worked only with noradrenaline, which under-ent recaption, but not with phenylephrine or potassium, whichid not (Fig. 7B). It must be noted, however, that in addition tohese observations, Soltis’s paper anticipated in nearly two decades

lephrine stimulation (C). In sharp contrast with the tunica adiposa or PVAT (C), thearotid artery (D). Graphs reconstructed from ref. [8] (A and B), [9] (C), and [187]

the possibility that angiotensin II of PVAT’s adipocytic origin influ-ences vascular tone [205]. This idea was also proposed in a previouswork by the same group showing the presence of high levels ofangiotensinogen mRNA in aortic PVAT [179].

12. Identity and mechanism of action of PVAT’s paracrinesubstances

Adipose tissue has progressively gained the category ofendocrine gland in a similar fashion as the endothelium once had[7,206]. The substances synthesized and released by adipocytes,collectively known as adipokines [207], exceed the hundred innumber [208]. Considering that adipocytes represent >60% of adi-pose tissue [209] and the enormous hormonal potential these cellspossess, it is quite reasonable to assume that some adipokines,when released from adipocytes anatomically close to the vessel(within PVAT), reach the tunica media, or even the endotheliumand the blood stream, and affect vascular function. This assumptioninspired Yudkin and others to coin the term “vasocrine”, referringspecifically to the paracrine communication between PVAT andthe artery it surrounds [210]. Amongst the catalogued adipokines,

many exhibit vasoactive properties to some degree besides theirmain function. This text does not attempt to fully describe thevascular-related findings of each adipokine. Excellent reviews onthis topic have been published, for example ref. 211. Still, we can

ogical

pi

a

b

de

wmea

a

m[alqttdsopAwidct

oolewudasGrr[ib[lth(sc[bcIir

E. Nava, S. Llorens / Pharmacol

erform a rough classification of adipokines from the vascular phys-ologist’s point of view, as follows:

) Vessel-targeted substances with uncertain chemical identity specif-ically studied in PVAT.

) Adipokines implicated in energy homeostasis.c) Classic vasoactive hormones with adipocytic origin.) Adipocytokines or pro-inflammatory cytokines.) Reactive oxygen species.

Vasoactive adipokines can be sorted as well consideringhether the vasomotor action takes place directly on smoothuscle cells o via activation (or inhibition) of the release of

ndothelium-derived vasoactive substances. This aspect will belso taken into account.

) Vessel-targeted substances with uncertain chemical identity specif-ically studied in PVAT

The existence of these substances is known through experi-ents in which a physiological solution containing undissected

9] or dissected [212] PVAT of a particular artery is transferred tonother artery devoid of PVAT that, in response, elicits a vasodi-ation. This kind of experiment proves that the adipokines inuestion are released by PVAT and are transferable. Presumablyhese adipokines are specifically released by PVAT and specificallyargeted to the PVAT’s surrounding artery, but whether other fatepots release adipokines with the same properties has never beentudied. The chemical identity of the adipokine(s) is unclear butften proves to relax the vessel by hyperpolarization after openingotassium channels of different kinds on the smooth muscle cell.s explained before, when they were discovered, these substancesere collectively named adventitium- or adipocyte-derived relax-

ng factors (ADRFs) [9,193]. The presence of ADRF is frequentlyemonstrated in experiments in which a vessel is subjected to aoncentration-response curve to a vasoconstrictor (Fig. 7C) and,hus, they are sometimes refered to as anti-contractile substances.

The initial reports on the discovery of ADRF already pointedut that the anti-contractile transferable substance is able topen K+

ATP channels and this action is dependent on extracel-ullar calcium, tyrosine kinase and protein kinase A, while beingndothelium- and perivascular nerve-independent [9,213]. Thisas demonstrated in rat aorta. When other vascular beds werender study, different channels were found to be involved. K+

Velayed rectifier channels were proposed in the rat mesentericrtery [193], while K+

Ca channels were the target of the very firsttudy on ADRF carried out in human arteries. This was performed byao and colleagues [214]. Curiously, this group, when working on

at mesenteric arteries found that the presence of PVAT augmented,ather than diminish, contractions caused by electrical stimulation215]. They argued that this is a specific phenomenon taking placen perivascular nerve endings within PVAT and this action woulde mediated by superoxide anion produced by NAD(P)H oxidase215]. In a review written later, Gao proposed to call perivascu-ar adipocyte-derived constricting factor (PVCF) to substances ofhis kind so as to distinguish them from ADRF. More recently, thisas also been refered to as “adipose-derived contracting factor”ADCF) [216]. Gao suggested that not only superoxide, but otherubstances could act as PVCFs, thus PVAT would elicit a dual, anti-ontractile and pro-contractile action on the vessel it surrounds217]. Later research demonstrating that the substance releasedy adipocytes upon electrical stimulation is angiotensin II [205],

onfirmed Gao’s suggestion of a PVCF different from superoxide.n the mean time, Gao kept working on the influence of PVATn other vascular beds. His group reported that rat aortic PVATeleases a transferable factor that operates under two mechanisms.

Research 113 (2016) 125–145 137

One mechanism would be endothelium-independent but, unlikeADRF reported in 2002 which depended on potassium channelopening, they found it would be (or depended on) H2O2. Anothermechanism would cause the opening of potassium channels but,in contrast to the initial report on rat aortic PVAT, this mechanismis endothelium-dependent and, in this case opens K+

Ca, not K+ATP,

channels [218].Although the existence of transferable factors from PVAT is

indisputable, opinions on the identity and mechanism of actionare heterogeneous. Löhn and Dubrovska initially discarded NO,prostaglandins and other mediators by using specific inhibitors[9,213]. Possibly the first attempt to pinpoint a particular substancewas that of Gao and coworkers with H2O2 [218]. However, they didnot propose a firm candidate until two years later showing that theidentity of the factor responsible of the endothelium-dependenteffects they detected is angiotensin 1–7 [219]. By the same year,another proposal was published: the gas H2S, via K+

ATP channels[220]. Later, another group asevered that the fatty acid, palmitateis the ADRF acting through K+

V channels [221]. Finally, adiponectin,which had been already ruled out as an ADRF [222], came again intoscene through two 2013 publications showing that adiponectinis an ADRF causing anticontractile effects by opening K+

Ca chan-nels [223,224]. Whatever the nature of ADRF, potassium channelopening has been repeatedly shown as a mechanism underlyingthe action of this factor. The first to positively confirm that theanti-contractile effect is accompanied by a hyperpolarization ofthe smooth muscle cell membrane were Dubrovska and colleagues[213]. Years later, Weston and others proposed the acronym ofadipocyte-derived hyperpolarizing factors (ADHFs) to refer to theanticontractile substances that relax the vessel’s smooth musclecell by means of membrane hyperpolarization [224].

b) Adipokines implicated in energy homeostasis

These are proteins involved in energy balance, insulin resis-tance, diabetes or obesity which have shown influences onvascular function as well. This group includes leptin, adiponectin,chemerin, omentin, apelin, resistin and visfatin. Leptin is the majoradipocyte-derived hormone involved in the control of food-intakehomeostasis. Leptin was also the first adipokine reported to influ-ence vascular function. The discovery that leptin receptors arepresent on endothelial cells [225] made the vascular wall a poten-tial target of this hormone. This was originally noticed by Frühbeckwho suggested that leptin induces the release of NO [226]. Theparacrine actions of leptin on the vessel wall are not easy to estab-lish on the basis of what has been published since reports onthe vascular actions of this adipokine are conflicting. The ear-liest reports show that leptin elicits NO-dependent [227] and−independent [228] vasodilation, which is difficult to reconcilewith the epidemiological data repeatedly showing that leptin islinked to the cardiovascular malfunctions of obesity (For reviewsee ref. [211]). Studies carried out specifically with PVAT havedemonstrated that PVAT-derived leptin clearly causes endothe-lial dysfunction in the obese animal (not the control counterparts)[229]. However, numerous studies suggest the possibility that lep-tin could be an endothelium-dependent ADRF [230–232]. In thissense, leptin has been considered a paradoxical substance in that itcauses NO-dependent vasodilation and, at the same time, its verypresence impairs endothelium-dependent relaxations [233].

Adiponectin is well known by cardiovascular epidemiologists asthe adipokine that opposes leptin’s actions [211]. An excess of thisone and a shortage of adiponectin has been associated with the vas-

cular undesirable effects of obesity [211]. Adiponectin has shownendothelial NO stimulating activity [234,235] and NO-dependentvasodilating actions [236]. Such is the vasodilatory reputationof adiponectin that it has been proposed as the unidentified

138 E. Nava, S. Llorens / Pharmacological Research 113 (2016) 125–145

Table 1Major events of perivascular adipose tissue (PVAT) research and discovery arranged in chronological order.

Event Year References

First report on smooth muscle responsiveness and the presence of PVAT 1991 [8]First report of the existence of an anticontractile transferable substance released from PVAT and coining of the term: ADRF 2002 [9]Demonstration that ADRF hyperpolarizes smooth muscle cells 2004 [193]First report on PVAT in obesity and incorporation of the acronym “PVAT” in the literature. 2005 [176]First report of a PVAT-derived transferable relaxing substance in human vessels 2005 [214]First report of the existence of a procontractile transferable substance released from PVAT 2006 [215]Incorporation of the term PVCF (perivascular contractile factor) 2007 [217]

a adip

scppfbtbed

c

kptaptaerisN[fEbrPsu[at[pmitsfoasTnm

m

Histological recognition of PVAT as the fourth layer of the vascular wall: the tunicThe term “PVAT dysfunction” is introduced in the literature.Incorporation of the term ADHF

ubstance behind ADRF [223,224]. In sharp contrast to adiponectin,hemerin, has been shown to be a PVAT-derived adipokine withro-contractile [237] as well as endothelium dysfunctioning [238]roperties, operating possibly via NAD(P)H oxidase [239]. Vis-atin of PVAT origin has shown no effects on contractility [240]ut, in non-PVAT set ups, visfatin elicited strong endothelial func-ion impairments [241]. Finally, omentin and resistin have noteen tested in PVAT but it has been reported that resistin impairsndothelial function [242] and omentin causes mild endothelium-ependent relaxations [243].

) Classic vasoactive hormones with adipocytic origin

These are vasoactive endocrine or paracrine hormones generallynown for being released from glands or tissues different from adi-ose tissue (e.g.: angiotensin II or NO) but were later demonstratedo have an adipocytic origin as well. The discovery that adipocytesre able to synthesize and release classic vasoactive hormonesaralleled the progressively enlarging list of adipokines. To date,raditional vasoactive substances that happen to be adipokinest the same time are: NO, prostacyclin, prostaglandin E2, TXA2,ndothelin-1, angiotensin II, aldosterone and noradrenaline. Firsteports on the synthesis and release of the vasoactive substancen question, either carried out using adipocytes from adipose tis-ue (AT) other than PVAT, or using PVAT itself, are the following:O from AT [201] and NO from PVAT [203]. Prostacyclin from AT

244] and from PVAT [204]. Prostaglandin E2 from AT [202] androm PVAT [245,246]. TXA2 from AT [247] and from PVAT [216,245].ndothelin-1 from AT [248] and from PVAT [245]. Angiotensin II,oth from AT and from PVAT [249]. Aldosterone from AT [250] (noeport on PVAT). Finally, noradrenaline from AT [251] and fromVAT [252]. It is remarkable that some of the reports are really old,uch as the adipocytic synthesis of prostaglandins [202], but notntil recently these were recognized as PVAT-derived adipokines204,216,245]. Also longstanding is the recognition that adipocytesre able to synthesize angiotensinogen [253] as well as the suspi-ion that adipose tissue would be able to produce angiotensin II8]. The fact that PVAT releases the same substances that were sup-osedly produced by the endothelium has numerous implications,ostly studied in the context of obesity, and exceed the capac-

ty of the present review. As mere examples, notice the relevancehat a PVAT of the visceral fat depot would have for the obese per-on when proven able to compete with the vascular endotheliumor the release of major vasoconstrictors like endothelin-1 [248]r TXA2 [216,245]. Or the fact that adipocytes within PVAT releasesngiotensin II towards the surrounding vessel, thus adding up to theame vasoconstrictor released from adrenal glands or elsewhere.he same can be said about aldosterone. Indeed, it is now recog-

ized that the adipocyte is provided with the complete synthesizingachinery of the renin-angiotensin-aldosterone system [249,250].Another remarkable fact of PVAT, although only one report in

ice aortae exists, is that it is able to produce sufficient NO to relax

osa 2007 [167]2007 [194]2012 [224]

an acetylcholine-stimulated deendothelialized artery [254]. If thisis confirmed, it would imply that PVAT possesses much more thana casual ability to affect the vessel’s function by substances thatotherwise all adipocytes produce. PVAT would possess a role in thecontrol of the vascular tone nearly as specific as that of the endothe-lium. That adipocytes are metabolically responsive to acetylcholineis known since the sixties [202]. It would be interesting to inves-tigate the pathways linking muscarinic receptors and NO synthaseand whether other classical endothelium-dependent vasodilatorscan stimulate PVAT to release NO. In relation to all this, it has beenproposed that PVAT-derived NO would play a transitory adaptiverole in the early stages of obesity [255].

d) Adipocytokines or pro-inflammatory cytokines

Adipocytokines are cytokines synthesized and released byadipocytes [207]. These substances are normally produced ininflammatory conditions, induce the formation of NO by theinducible form of NOS [256] but impair endothelium-dependentrelaxations [257]. Adipocytes within PVAT are also able to producethese substances. In fact, the PVAT fat store displays a particu-larly proinflammatory phenotype compared to the subcutaneousdepot, and releases more adipocytokines [196]. This is true ina normal situation [196], but is exacerbated in injurious condi-tions [258]. Therefore, and using the title of the first report oncytokines and endothelial function [257], PVAT must have “an anti-EDRF effect”. This intuitive idea has been proven in experimentsperformed in arteries studied with or without PVAT [259] andspecifically related to tumor necrosis factor-� [260]. It must beemphasized, however, that not every report confirms that PVAT isa detereous tissue merely for having a basal adipocytokine releas-ing ability. For example, Rittig and others stated that PVAT causesno endothelial dysfunction [261] and various studies indicate thatendothelium-dependent relaxations are not altered in vessels witha preserved PVAT as compared to vessels without PVAT [245,262].Rather, endothelial dysfunction is detected when this experiment isperformed in obese animals [245,262]. The obese, in turn, exhibit aloss in anticontractile function that can be improved with cytokineantagonists [263].

e) Reactive oxygen species

Following a similar scheme as with other adipocyte-derivedsubstances, production of reactive oxygen species was originallydescribed in adipocytes many years before the concept of PVATexisted. In this case, it was H2O2 released as a byproduct ofadipocyte’s glucose metabolism [264]. The first to link PVAT withfree radicals were, as previously stated, Gao and coworkers with

H2O2 in the role of a vasodilator [218] and later, superoxide actingas a vasoconstrictor [215]. A recent study emphasizes the role ofdismutation of procontractile mitochondrial superoxide into anti-contractile H2O2, thus reconciling the conflict [265]. The increase in

E. Nava, S. Llorens / Pharmacological Research 113 (2016) 125–145 139

Table 2Substances released by PVAT, which have shown to relax, contract or cause endothelial dysfunction of the surrounded vessel.

PVAT-derived relaxing adipokines References PVAT-derived contracting adipokines or adipokines producing endothelial dysfunction References

Unidentified [9,193] Superoxide anion [215]Angiotensin 1-7 [219] Visfatin [240]c

H2S [220] Angiotensin II [205]Palmitate [221] Leptin [229]Prostacyclin [204] PGE2 [245]Leptin [231,232]a TXA2 [216,245]Adiponectin [223,224] Endothelin-1 [245]d

PGE2 [246] Chemerin [237,238]NO [255]b Noradrenaline [252]

TNF� [266]e

References for vasoactive hormones do not necessarily coincide with those quoted in section 12c (Classic vasoactive hormones with adipocytic origin), which refer to firstreports on the ability of PVAT to produce the substance but with no further attempt to study its vasomotor function.

a Only Dahl salt-sensitive rats.ot lea

his ad

sfat[sbsp

dwtaidc

FdrsR2

b Gil-Ortega and coworkers do report functional PVAT-derived NO but in obese, nc PVAT-derived visfatin causes no vasomotor effects but vessels incubated with td Only obese animals and pharmacologically, not analytically, tested.e Especially in obese animals.

uperoxide production by PVAT has been linked to the vascular mal-unctions of obesity such as failure of anti-contractile activity [263]nd endothelial dysfunction [262]. Most authors coincide in thathe superoxide-generating enzyme in PVAT is NAD(P)H oxidase215,266]. Many obesity papers tend to report together the exces-ive production of adipocytokines and oxygen radicals [262,263],ut only Virdis and colleagues demonstrate a cause-effect relation-hip between tumor necrosis factor-� and reactive oxygen speciesroduction in PVAT [266].

Table 1 shows the principal milestones of PVAT research andiscovery and Table 2 summarizes the substances released by PVAThich have shown to relax, contract or cause endothelial dysfunc-

ion of the surrounded vessel. For the sake of caution, substances

re refered to as relaxing or contracting adipokines, rather than call-ng them ADRF or PVCF because the quoted reporting paper oftenoes not do so, or simply states that the substance under test is aandidate for ADRF or PVCF to be considered.

ig. 8. Interactions PVAT adipocyte-smooth muscle cell-endothelial cell exhibit a patterate involves adiponectin, leptin, angiotensin 1–7 (AT1-7), reactive oxygen species (ROS) aeceptor (NPR-A), adiponectin receptor (AdR), calcium-dependent K+ channel (KCa), leptmooth muscle cell (SMC), endothelial cell (EC).eference matching (first author followed by year): a Withers 2014a [270], b Withers 2014b012 [231], g Lee 2009 [219], h Gao 2006 [215], i Virdis 2015 [266], j Antonopoulos 2015

n animals.ipokine suffer severe endothelial dysfunction [241].

13. Interactions between PVAT, smooth muscle andendothelium

The classic vision of the paracrine control of the vessel wallimplicated an inside-to-outside signaling via prostacyclin, NO,EDHF or EDCFs. The irruption of PVAT and its ability to signalinwards has complicated the scenario. A possible crosstalk betweenendothelial cells and adipocytes was actually proposed very muchearlier than PVAT was described [267]. In addition, a review waspublished in the early nineties defending the idea that adipocytesare the source of arachidonic acid for endothelial prostaglandinsynthesis and, therefore, that endothelial prostaglandin releaserequires the cooperation between both cell types [268]. Since

the surge of these early notions, our knowledge on the coop-eration between the different vascular cells has grown notably.Fig. 8, depicts the interactions described so far between the 3cell types. Most of them have been previously described in the

n of “outside-to-inside-to-outside” communication. The trio crosstalk reported tond TNF�. Noradrenaline (NA), 4-hydroxynonenal (4HNE), type A natriuretic peptidein receptor (OB-R), Mas-related G protein-coupled receptor (MasR), oxidase (ox),

[269], c Lynch 2013 [223], d Weston 2013 [224], e Payne 2010 [229], f Galvez-Prieto [271], k Deng 2010 [273].

1 ogical

crdaarNbPofictPc

1

fivceicpls“pfwNtarlPeAtpeeioWrPdrENcc

A

d(

40 E. Nava, S. Llorens / Pharmacol

orresponding adipokine section. Interesting complex interactionsecently proposed are: 1) axis comprised by sympathetic nora-renaline – adipocyte �3 receptor – PVAT adiponectin – endothelialdiponectin receptor – NO – smooth muscle cell relaxation [269]; 2)xis comprised by blood atrial natriuretic peptide – adipocyte ANPeceptor – PVAT adiponectin – endothelial adiponectin receptor –O – smooth muscle cell relaxation [270]; and 3) axis comprisedy smooth muscle cell-derived ROS – blood 4-hydroxynonenal –VAT adiponectin – endothelial (and smooth muscle cell) NAD(P)Hxidase inhibition – NO – smooth muscle cell relaxation [271]. Anal potential route of cell communication is that of endothelialells towards PVAT adipocytes. A recent report on this route inves-igated the possible action of endothelium-derived NO on adjacentVAT compared to that on the media layer. However, it failed toonfirm any endothelium – PVAT axis [272].

4. Perspectives

The present review intended to provide a wide vision on theootsteps walked to reach the point we are now on vascular phys-ology research. As the reader has seen, advances took place in aery heterogeneous manner, serendipitous in some occasions andlearly targeted in others. Some times, ancient unconnected knowl-dge was retrieved and used in a different context to that in whicht was originally published. In other occasions, researchers werelose to a discovery but just missed it. Attempting to provide futureerspectives is a difficult task. Fig. 5 shows that research in vascu-

ar pharmacology has followed various waves in which a certainubstance or substances were on top. Now we are clearly in thePVAT wave” and the nearest perspectives will likely concern thehysiology and pathophysiology of PVAT-derived substances. Theollowing are some questions raised in the course of this reviewhich may be useful for forthcoming research: Is PVAT-derivedO able to compete with endothelium-derived NO in the con-

rol of basal vascular tone, obviously, in vascular beds that carrydipose tissue around? Alternatively, is PVAT-derived NO able toeplace endothelial NO in dysfunctional conditions? Is endothe-ial dysfunction of a certain artery necessarily accompanied byVAT dysfunction of this particular vessel or, rather, a functionalndothelium or PVAT can replace the dysfunctional counterpart?re PVAT-derived prostaglandins and (or) ADRF (which in prac-

ice is an ADHF) able to compete with the endothelial counterparts,rostacyclin, thromboxane A2 and EDHF? In general terms, to whatxtent PVAT exerts a regulation in vasomotion as compared tondothelial cells? Is PVAT-derived NO able to relax deendothelial-zed arteries upon classic endothelial stimulators like acetylcholiner bradykinin or should we look for PVAT-specific NO stimulators?hat are the similarities and differences between the biology and

egulation of endothelium-derived NO and that of NO released fromVAT? Is ADRF exclusively released from PVAT, or other fat depotso so, and thus, ADRF is an unspecific adipokine? Does the generalule stating that the lesser the caliber or a vessel, the higher the ratioDHF/NO apply for PVAT as well regarding ADHF versus adipocyticO? Are there endothelium-derived substances targeted to adja-ent PVAT or is smooth muscle the exclusive target of endothelialells?

cknowledgments

This work was supported by Ayudas para la Financiacióne Actividades de Investigación Dirigidas a Grupos de la UCLMGI20163512). I am indebted to Emeritus Prof. P. Sánchez García.

Research 113 (2016) 125–145

References

[1] C. Jimenez-Diaz, Le role de la paroi arterielle dans la regulationneurochimique de la pression arterielle, Schweiz. Med. Wochenschr. 78(1948) 920–923.

[2] C. Jimenez Diaz, P. Barreda, A.F. Molina, R. Alcala, Role of arterial wallsecretion in the regulation of blood pressure, Circulation 9 (1954) 903–907.

[3] F. Feyrter, [The theory of the peripheral endocrine (paracrine) glands], ActaNeuroveg. (Wien) 4 (1952) 409–424.

[4] S. Moncada, R.J. Gryglewski, S. Bunting, J.R. Vane, A lipid peroxide inhibitsthe enzyme in blood vessel microsomes that generates from prostaglandinendoperoxides the substance (prostaglandin X) which prevents plateletaggregation, Prostaglandins 12 (1976) 715–737.

[5] S. Moncada, R. Korbut, S. Bunting, J.R. Vane, Prostacyclin is a circulatinghormone, Nature 273 (1978) 767–768.

[6] R.F. Furchgott, J.V. Zawadzki, The obligatory role of endothelial cells in therelaxation of arterial smooth muscle by acetylcholine, Nature 288 (1980)373–376.

[7] J.R. Vane, R.J. Gryglewski, R.M. Botting, The endothelial cell as a metabolicand endocrine organ, Trends Pharmacol. Sci. 8 (1987) 491–496.

[8] E.E. Soltis, L.A. Cassis, Influence of perivascular adipose tissue on rat aorticsmooth muscle responsiveness, Clin. Exp. Hypertens. A 13 (1991) 277–296.

[9] M. Lohn, G. Dubrovska, B. Lauterbach, F.C. Luft, M. Gollasch, A.M. Sharma,Periadventitial fat releases a vascular relaxing factor, FASEB J. 16 (2002)1057–1063.

[10] P.J. Pagano, Y. Ito, K. Tornheim, P.M. Gallop, A.I. Tauber, R.A. Cohen, AnNADPH oxidase superoxide-generating system in the rabbit aorta, Am. J.Physiol. 268 (1995) H2274–2280.

[11] V. von Reckinghausen, Eine Methode, mikroskopische hohle und solideGebilde von einander zu unterscheiden, Virchows Arch. 19 (1860) 451.

[12] Florey, The endothelial cell, Br. Med. J. 2 (1966) 487–490.[13] J.A. Rhodin, The ultrastructure of mammalian arterioles and precapillary

sphincters, J. Ultrastruct. Res. 18 (1967) 181–223.[14] J.A. Bevan, S.P. Duckles, Evidence for alpha-adrenergic receptors on intimal

endothelium, Blood Vessels 12 (1975) 307–310.[15] R. Pascual, J.A. Bevan, Evidence that changes in vascular tone may be

initiated from the intima of the rabbit aorta, in: J.A. Bevan, T. Godfraind, R.A.Maxwell, H. Majewski, P.M. Vanhoutte (Eds.), Vascular NeuroeffectorMechanisms, Raven Press, New York, 1980, pp. 300–304.

[16] V. Buonassisi, J.C. Venter, Hormone and neurotransmitter receptors in anestablished vascular endothelial cell line, Proc. Natl. Acad. Sci. U. S. A. 73(1976) 1612–1616.

[17] I.S. De la Lande, D. Frewin, J.G. Waterson, The influence of sympatheticinnervation on vascular sensitivity to noradrenaline, Br. J. Pharmacol.Chemother. 31 (1967) 82–93.

[18] T.A. McCalden, J.A. Bevan, Asymmetry of the contractile response of therabbit ear artery to exogenous amines, Am. J. Physiol. 238 (1980)H618–H624.

[19] R. Pascual, J.A. Bevan, Variation in contractile response characteristics ofrabbit aorta strips with surface of drug entry, J. Pharmacol. Exp. Ther. 210(1979) 51–55.

[20] P.J. Piper, J.R. Vane, Release of additional factors in anaphylaxis and itsantagonism by anti-inflammatory drugs, Nature 223 (1969) 29–35.

[21] J. Svensson, M. Hamberg, B. Samuelsson, Prostaglandin endoperoxides IX:characterization of rabbit aorta contracting substance (RCS) from guinea piglung and human platelets, Acta Physiol. Scand. 94 (1975) 222–228.

[22] P. Needleman, S. Moncada, S. Bunting, J.R. Vane, M. Hamberg, B. Samuelsson,Identification of an enzyme in platelet microsomes which generatesthromboxane A2 from prostaglandin endoperoxides, Nature 261 (1976)558–560.

[23] J.R. Vane, Adventures and excursions in bioassay: the stepping stones toprostacylin, Biosci. Rep. 24 (2004) 254–279.

[24] N. Whittaker, S. Bunting, J. Salmon, S. Moncada, J.R. Vane, R.A. Johnson, et al.,The chemical structure of prostaglandin X (prostacyclin), Prostaglandins 12(1976) 915–928.

[25] J.E. Tateson, S. Moncada, J.R. Vane, Effects of prostacyclin (PGX) on cyclicAMP concentrations in human platelets, Prostaglandins 13 (1977) 389–397.

[26] S. Moncada, A.G. Herman, E.A. Higgs, J.R. Vane, Differential formation ofprostacyclin (PGX or PGI2) by layers of the arterial wall: an explanation forthe anti-thrombotic properties of vascular endothelium, Thromb. Res. 11(1977) 323–344.

[27] D.E. MacIntyre, J.D. Pearson, J.L. Gordon, Localisation and stimulation ofprostacyclin production in vascular cells, Nature 271 (1978) 549–551.

[28] R.F. Furchgott, Endothelium-derived relaxing factor: discovery, earlystudies, and identification as nitric oxide, Biosci. Rep. 19 (1999) 235–251.

[29] H.H. Dale, The action of certain esters and ethers of choline, and theirrelation to muscarine, J. Pharmacol. Exp. Ther. 6 (1914) 147–190.

[30] R.W. Jelliffe, Dilator and constrictor effects of acetylcholine on isolatedrabbit aortic chains, J. Pharmacol. Exp. Ther. 135 (1962) 349–353.

[31] B. Ablad, S. Mellander, Comparative effects of hydralazine, sodium nitriteand acetylcholine on resistance and capacitance blood vessels and capillary

filtration in skeletal muscle in the cat, Acta Physiol. Scand. 58 (1963)319–329.

ogical

E. Nava, S. Llorens / Pharmacol

[32] R.F. Furchgott, D. Davidson, C.I. Lin, Conditions which determine whethermuscarinic agonists contract or relax rabbit aortic rings and strips, BloodVessels 16 (1979) 213–214.

[33] J.G. De Mey, P.M. Vanhoutte, Role of the intima in cholinergic and purinergicrelaxation of isolated canine femoral arteries, J. Physiol. 316 (1981) 347–355.

[34] J.G. De Mey, P.M. Vanhoutte, Role of the intima in the relaxation of thecanine femoral artery caused by thrombin, Arch. Int. Pharmacodyn. Ther.250 (1981) 314–315.

[35] B.M. Altura, N. Chand, Bradykinin-induced relaxation of renal andpulmonary arteries is dependent upon intact endothelial cells, Br. J.Pharmacol. 74 (1981) 10–11.

[36] P.D. Cherry, R.F. Furchgott, J.V. Zawadzki, D. Jothianandan, Role ofendothelial cells in relaxation of isolated arteries by bradykinin, Proc. Natl.Acad. Sci. U. S. A. 79 (1982) 2106–2110.

[37] J.V. Zawadzki, P.D. Cherry, R.F. Furchgott, Comparison ofendothelium-dependent relaxation of rabbit aorta by A23187 and byacetylcholine (abstr.), Pharmacologist 22 (1980) 271.

[38] H.A. Singer, M.J. Peach, Calcium- and endothelial-mediated vascular smoothmuscle relaxation in rabbit aorta, Hypertension 4 (1982) 19–25.

[39] J.G. De Mey, P.M. Vanhoutte, Interaction between Na+,K+ exchanges and thedirect inhibitory effect of acetylcholine on canine femoral arteries, Circ. Res.46 (1980) 826–836.

[40] J.G. De Mey, P.M. Vanhoutte, Contribution of the endothelium to theresponse to anoxia in the canine femoral artery, Arch. Int. Pharmacodyn.Ther. 253 (1981) 325–326.

[41] J.G. De Mey, P.M. Vanhoutte, Heterogeneous behavior of the canine arterialand venous wall. Importance of the endothelium, Circ. Res. 51 (1982)439–447.

[42] G.M. Rubanyi, P.M. Vanhoutte, Hypoxia releases a vasoconstrictor substancefrom the canine vascular endothelium, J. Physiol. 364 (1985) 45–56.

[43] K.A. Hickey, G. Rubanyi, R.J. Paul, R.F. Highsmith, Characterization of acoronary vasoconstrictor produced by cultured endothelial cells, Am. J.Physiol. 248 (1985) C550–C556.

[44] M. Yanagisawa, H. Kurihara, S. Kimura, Y. Tomobe, M. Kobayashi, Y. Mitsui,et al., A novel potent vasoconstrictor peptide produced by vascularendothelial cells, Nature 332 (1988) 411–415.

[45] M.N. Gillespie, J.O. Owasoyo, I.F. McMurtry, R.F. O’Brien, Sustained coronaryvasoconstriction provoked by a peptidergic substance released fromendothelial cells in culture, J. Pharmacol. Exp. Ther. 236 (1986) 339–343.

[46] G.M. Rubanyi, The discovery of endothelin: the power of bioassay and therole of serendipity in the discovery of endothelium-derived vasocativesubstances, Pharmacol. Res. 63 (2011) 448–454.

[47] Vanhoutte P.M., Auch-Schwelk W., Boulanger C., Janssen P.A., Katusic Z.S.,Komori K., et al., Does endothelin-1 mediate endothelium-dependentcontractions during anoxia? J Cardiovasc Pharmacol 1989,13 Suppl5:S124-128; discussion S142.

[48] V.M. Miller, P.M. Vanhoutte, Endothelium-dependent contractions toarachidonic acid are mediated by products of cyclooxygenase, Am. J. Physiol.248 (1985) H432–437.

[49] R.J. Altiere, J.A. Kiritsy-Roy, J.D. Catravas, Acetylcholine-inducedcontractions in isolated rabbit pulmonary arteries: role of thromboxane A2,J. Pharmacol. Exp. Ther. 236 (1986) 535–541.

[50] T.F. Luscher, P.M. Vanhoutte, Endothelium-dependent contractions toacetylcholine in the aorta of the spontaneously hypertensive rat,Hypertension 8 (1986) 344–348.

[51] Z.S. Katusic, P.M. Vanhoutte, Superoxide anion is an endothelium-derivedcontracting factor, Am. J. Physiol. 257 (1989) H33–37.

[52] G.M. Rosen, B.A. Freeman, Detection of superoxide generated by endothelialcells, Proc. Natl. Acad. Sci. U. S. A. 81 (1984) 7269–7273.

[53] R.M. Tate, H.G. Morris, W.R. Schroeder, J.E. Repine, Oxygen metabolitesstimulate thromboxane production and vasoconstriction in isolatedsaline-perfused rabbit lungs, J. Clin. Invest. 74 (1984) 608–613.

[54] F. Cosentino, J.C. Sill, Z.S. Katusic, Role of superoxide anions in the mediationof endothelium-dependent contractions, Hypertension 23 (1994) 229–235.

[55] E.H. Tang, F.P. Leung, Y. Huang, M. Feletou, K.F. So, R.Y. Man, et al., Calciumand reactive oxygen species increase in endothelial cells in response toreleasers of endothelium-derived contracting factor, Br. J. Pharmacol. 151(2007) 15–23.

[56] S. Martinez-Revelles, M.S. Avendano, A.B. Garcia-Redondo, Y. Alvarez, A.Aguado, J.V. Perez-Giron, et al., Reciprocal relationship between reactiveoxygen species and cyclooxygenase-2 and vascular dysfunction inhypertension, Antioxid. Redox Signal. 18 (2013) 51–65.

[57] W. Auch-Schwelk, Z.S. Katusic, P.M. Vanhoutte, Thromboxane A2 receptorantagonists inhibit endothelium-dependent contractions, Hypertension 15(1990) 699–703.

[58] T. Kato, Y. Iwama, K. Okumura, H. Hashimoto, T. Ito, T. Satake, ProstaglandinH2 may be the endothelium-derived contracting factor released byacetylcholine in the aorta of the rat, Hypertension 15 (1990) 475–481.

[59] P. Gluais, M. Lonchampt, J.D. Morrow, P.M. Vanhoutte, M. Feletou,Acetylcholine-induced endothelium-dependent contractions in the SHRaorta: the Janus face of prostacyclin, Br. J. Pharmacol. 146 (2005) 834–845.

[60] S. Moncada, Adventures in vascular biology: a tale of two mediators, Philos.Trans. R Soc. Lond. B Biol. Sci. 361 (2006) 735–759.

[61] J.V. Levy, Prostacyclin-induced contraction of isolated aortic strips fromnormal and spontaneously hypertensive rats (SHR), Prostaglandins 19(1980) 517–525.

Research 113 (2016) 125–145 141

[62] R.M. Rapoport, S.P. Williams, Role of prostaglandins inacetylcholine-induced contraction of aorta from spontaneouslyhypertensive and Wistar-Kyoto rats, Hypertension 28 (1996) 64–75.

[63] Z.S. Katusic, J. Schugel, F. Cosentino, P.M. Vanhoutte,Endothelium-dependent contractions to oxygen-derived free radicals in thecanine basilar artery, Am. J. Physiol. 264 (1993) H859–864.

[64] J.G. De Mey, M. Claeys, P.M. Vanhoutte, Endothelium-dependent inhibitoryeffects of acetylcholine, adenosine triphosphate, thrombin and arachidonicacid in the canine femoral artery, J. Pharmacol. Exp. Ther. 222 (1982)166–173.

[65] S.S. Segal, B.R. Duling, Flow control among microvessels coordinated byintercellular conduction, Science 234 (1986) 868–870.

[66] T.B. Bolton, R.J. Lang, T. Takewaki, Mechanisms of action of noradrenalineand carbachol on smooth muscle of guinea-pig anterior mesenteric artery, J.Physiol. 351 (1984) 549–572.

[67] J.L. Gordon, W. Martin, Endothelium-dependent relaxation of the pig aorta:relationship to stimulation of 86Rb efflux from isolated endothelial cells, Br.J. Pharmacol. 79 (1983) 531–541.

[68] G. Edwards, K.A. Dora, M.J. Gardener, C.J. Garland, A.H. Weston, K+ is anendothelium-derived hyperpolarizing factor in rat arteries, Nature 396(1998) 269–272.

[69] Weston A.H.T., S.G., Southerton, J.S., Bray, K.M., Newgreen, D.T., McHarg,A.D., Potassium channel-opening drugs in smooth muscle. In: Vascularneuroeffector mechanisms: receptors, ion channels, second messengers andendogenous mediators. Edited by Bevan JA MH, Maxwell RA, Story DF.Oxford : IRL Press; 1988. pp. 193–200.

[70] G. Chen, H. Suzuki, A.H. Weston, Acetylcholine releasesendothelium-derived hyperpolarizing factor and EDRF from rat bloodvessels, Br. J. Pharmacol. 95 (1988) 1165–1174.

[71] S.P. Olesen, P.F. Davies, D.E. Clapham, Muscarinic-activated K+ current inbovine aortic endothelial cells, Circ. Res. 62 (1988) 1059–1064.

[72] M. Feletou, P.M. Vanhoutte, Endothelium-dependent hyperpolarization ofcanine coronary smooth muscle, Br. J. Pharmacol. 93 (1988) 515–524.

[73] A.H. Huang, R. Busse, E. Bassenge, Endothelium-dependenthyperpolarization of smooth muscle cells in rabbit femoral arteries is notmediated by EDRF (nitric oxide), Naunyn. Schmiedebergs Arch. Pharmacol.338 (1988) 438–442.

[74] M. Tare, H.C. Parkington, H.A. Coleman, T.O. Neild, G.J. Dusting,Hyperpolarization and relaxation of arterial smooth muscle caused by nitricoxide derived from the endothelium, Nature 346 (1990) 69–71.

[75] H.C. Parkington, M. Tare, M.A. Tonta, H.A. Coleman, Stretch revealed threecomponents in the hyperpolarization of guinea-pig coronary artery inresponse to acetylcholine, J. Physiol. 465 (1993) 459–476.

[76] J.L. Beny, P.Y. von der Weid, Hydrogen peroxide: an endogenous smoothmuscle cell hyperpolarizing factor, Biochem. Biophys. Res. Commun. 176(1991) 378–384.

[77] T. Matoba, H. Shimokawa, M. Nakashima, Y. Hirakawa, Y. Mukai, K. Hirano,et al., Hydrogen peroxide is an endothelium-derived hyperpolarizing factorin mice, J. Clin. Invest. 106 (2000) 1521–1530.

[78] M. Hecker, A.T. Bara, J. Bauersachs, R. Busse, Characterization ofendothelium-derived hyperpolarizing factor as a cytochrome P450-derivedarachidonic acid metabolite in mammals, J. Physiol. 481 (Pt 2) (1994)407–414.

[79] W.B. Campbell, D. Gebremedhin, P.F. Pratt, D.R. Harder, Identification ofepoxyeicosatrienoic acids as endothelium-derived hyperpolarizing factors,Circ. Res. 78 (1996) 415–423.

[80] G. Edwards, M. Feletou, A.H. Weston, Endothelium-derived hyperpolarisingfactors and associated pathways: a synopsis, Pflugers Arch. 459 (2010)863–879.

[81] L.N. Katz, E. Lindner, The action of excess Na, Ca and K on the coronaryvessels, Am. J. Physiol. 124 (1938) 155–160.

[82] R. Thoma, Uber die Intima der Arterien, Virchows Arch. Path. Anat. Physiol.230 (1921) 1–45.

[83] D.H. Moore, H. Ruska, The fine structure of capillaries and small arteries, J.Biophys. Biochem. Cytol. 3 (1957) 457–462.

[84] I. Huttner, M. Boutet, R.H. More, Gap junctions in arterial endothelium, J.Cell Biol. 57 (1973) 247–252.

[85] T. Iwayama, Nexuses between areas of the surface membrane of the samearterial smooth muscle cell, J. Cell Biol. 49 (1971) 521–525.

[86] L.G. Spagnoli, S. Villaschi, L. Neri, G. Palmieri, Gap junctions inmyo-endothelial bridges of rabbit carotid arteries, Experientia 38 (1982)124–125.

[87] D.A. Goodenough, Bulk isolation of mouse hepatocyte gap junctions.Characterization of the principal protein, connexin, J. Cell Biol 61 (1974)557–563.

[88] M.M. Dewey, L. Barr, Intercellular connection between smooth muscle cells:the nexus, Science 137 (1962) 670–672.

[89] D.M. Larson, C.C. Haudenschild, E.C. Beyer, Gap junction messenger RNAexpression by vascular wall cells, Circ. Res. 66 (1990) 1074–1080.

[90] E.C. Beyer, K.E. Reed, E.M. Westphale, H.L. Kanter, D.M. Larson, Molecularcloning and expression of rat connexin40, a gap junction protein expressed

in vascular smooth muscle, J. Membr. Biol. 127 (1992) 69–76.

[91] K.E. Reed, E.M. Westphale, D.M. Larson, H.Z. Wang, R.D. Veenstra, E.C. Beyer,Molecular cloning and functional expression of human connexin37, anendothelial cell gap junction protein, J. Clin. Invest. 91 (1993) 997–1004.

1 ogical

42 E. Nava, S. Llorens / Pharmacol

[92] S. Alcolea, M. Theveniau-Ruissy, T. Jarry-Guichard, I. Marics, E. Tzouanacou,J.P. Chauvin, et al., Downregulation of connexin 45 gene products duringmouse heart development, Circ. Res. 84 (1999) 1365–1379.

[93] R.E. Haddock, T.H. Grayson, T.D. Brackenbury, K.R. Meaney, C.B. Neylon, S.L.Sandow, et al., Endothelial coordination of cerebral vasomotion viamyoendothelial gap junctions containing connexins 37 and 40, Am. J.Physiol. Heart Circ. Physiol. 291 (2006) H2047–2056.

[94] S.S. Segal, Integration and modulation of intercellular signaling underlyingblood flow control, J. Vasc. Res. 52 (2015) 136–157.

[95] A.T. Chaytor, W.H. Evans, T.M. Griffith, Central role of heterocellular gapjunctional communication in endothelium-dependent relaxations of rabbitarteries, J. Physiol. 508 (Pt 2) (1998) 561–573.

[96] T.M. Griffith, H.J. Taylor, A.M.P. Cyclic, mediates EDHF-type relaxations ofrabbit jugular vein, Biochem. Biophys. Res. Commun. 263 (1999) 52–57.

[97] M. Feletou, P.M. Vanhoutte, Endothelium-dependent hyperpolarization: nolonger an f-word!, J. Cardiovasc. Pharmacol. 61 (2013) 91–92.

[98] G.G. Emerson, S.S. Segal, Electrical coupling between endothelial cells andsmooth muscle cells in hamster feed arteries: role in vasomotor control,Circ. Res. 87 (2000) 474–479.

[99] U. Forstermann, B. Neufang, The endothelium-dependent vasodilator effectof acetylcholine: characterization of the endothelial relaxing factor withinhibitors of arachidonic acid metabolism, Eur. J. Pharmacol. 103 (1984)65–70.

[100] T.M. Griffith, D.H. Edwards, M.J. Lewis, A.C. Newby, A.H. Henderson, Thenature of endothelium-derived vascular relaxant factor, Nature 308 (1984)645–647.

[101] H.A. Singer, M.J. Peach, Endothelium-dependent relaxation of rabbit aorta:II. Inhibition of relaxation stimulated by methacholine and A23187 withantagonists of arachidonic acid metabolism, J. Pharmacol. Exp. Ther. 226(1983) 796–801.

[102] H.A. Singer, M.J. Peach, Endothelium-dependent relaxation of rabbit aorta. I.Relaxation stimulated by arachidonic acid, J. Pharmacol. Exp. Ther. 226(1983) 790–795.

[103] H. Kimura, C.K. Mittal, F. Murad, Activation of guanylate cyclase from ratliver and other tissues by sodium azide, J. Biol. Chem. 250 (1975) 8016–8022.

[104] J. Diamond, K.S. Blisard, Effects of stimulant and relaxant drugs on tensionand cyclic nucleotide levels in canine femoral artery, Mol. Pharmacol. 12(1976) 668–692.

[105] F.R. DeRubertis, P.A. Craven, Calcium-independent modulation of cyclicGMP and activation of guanylate cyclase by nitrosamines, Science 193(1976) 897–899.

[106] S. Katsuki, W. Arnold, C. Mittal, F. Murad, Stimulation of guanylate cyclaseby sodium nitroprusside, nitroglycerin and nitric oxide in various tissuepreparations and comparison to the effects of sodium azide andhydroxylamine, J. Cyclic Nucleotide Res. 3 (1977) 23–35.

[107] W.P. Arnold, C.K. Mittal, S. Katsuki, F. Murad, Nitric oxide activatesguanylate cyclase and increases guanosine 3’:5’-cyclic monophosphatelevels in various tissue preparations, Proc. Natl. Acad. Sci. U. S. A. 74 (1977)3203–3207.

[108] T.L. Brunton, On the use of nitrite of amyl in angina pectoris, Lancet 90(1867) 97–98.

[109] T.L. Brunton, The action of nitrate of amyl on the circulation, J. Anat. Physiol.5 (1870) 92–101.

[110] S. Katsuki, F. Murad, Regulation of adenosine cyclic 3’,5’-monophosphateand guanosine cyclic 3’,5’-monophosphate levels and contractility in bovinetracheal smooth muscle, Mol. Pharmacol. 13 (1977) 330–341.

[111] C.A. Gruetter, D.Y. Gruetter, J.E. Lyon, P.J. Kadowitz, L.J. Ignarro, Relationshipbetween cyclic guanosine 3’:5’-monophosphate formation and relaxation ofcoronary arterial smooth muscle by glyceryl trinitrate, nitroprusside, nitriteand nitric oxide: effects of methylene blue and methemoglobin, J.Pharmacol. Exp. Ther. 219 (1981) 181–186.

[112] W.R. Kukovetz, S. Holzmann, A. Wurm, G. Poch, Evidence for cyclicGMP-mediated relaxant effects of nitro-compounds in coronary smoothmuscle, Naunyn. Schmiedebergs Arch. Pharmacol. 310 (1979) 129–138.

[113] R.M. Rapoport, M.B. Draznin, F. Murad, Sodium nitroprusside-inducedprotein phosphorylation in intact rat aorta is mimicked by 8-bromo cyclicGMP, Proc. Natl. Acad. Sci. U. S. A. 79 (1982) 6470–6474.

[114] F. Murad, C.K. Mittal, W.P. Arnold, S. Katsuki, H. Kimura, Guanylate cyclase:activation by azide, nitro compounds, nitric oxide, and hydroxyl radical andinhibition by hemoglobin and myoglobin, Adv. Cyclic Nucleotide Res. 9(1978) 145–158.

[115] S. Holzmann, Endothelium-induced relaxation by acetylcholine associatedwith larger rises in cyclic GMP in coronary arterial strips, J. Cyclic NucleotideRes. 8 (1982) 409–419.

[116] R.M. Rapoport, F. Murad, Agonist-induced endothelium-dependentrelaxation in rat thoracic aorta may be mediated through cGMP, Circ. Res. 52(1983) 352–357.

[117] L.J. Ignarro, T.M. Burke, K.S. Wood, M.S. Wolin, P.J. Kadowitz, Associationbetween cyclic GMP accumulation and acetylcholine-elicited relaxation ofbovine intrapulmonary artery, J. Pharmacol. Exp. Ther. 228 (1984) 682–690.

[118] W. Martin, G.M. Villani, D. Jothianandan, R.F. Furchgott, Selective blockade

of endothelium-dependent and glyceryl trinitrate-induced relaxation byhemoglobin and by methylene blue in the rabbit aorta, J. Pharmacol. Exp.Ther. 232 (1985) 708–716.

Research 113 (2016) 125–145

[119] R.J. Gryglewski, R.M. Palmer, S. Moncada, Superoxide anion is involved inthe breakdown of endothelium-derived vascular relaxing factor, Nature 320(1986) 454–456.

[120] G.M. Rubanyi, P.M. Vanhoutte, Superoxide anions and hyperoxia inactivateendothelium-derived relaxing factor, Am J. Physiol 250 (1986) H822–827.

[121] S. Moncada, R.M. Palmer, R.J. Gryglewski, Mechanism of action of someinhibitors of endothelium-derived relaxing factor, Proc. Natl. Acad. Sci. U. S.A. 83 (1986) 9164–9168.

[122] H. Azuma, M. Ishikawa, S. Sekizaki, Endothelium-dependent inhibition ofplatelet aggregation, Br. J. Pharmacol. 88 (1986) 411–415.

[123] M.W. Radomski, R.M. Palmer, S. Moncada, Comparative pharmacology ofendothelium-derived relaxing factor, nitric oxide and prostacyclin inplatelets, Br. J. Pharmacol. 92 (1987) 181–187.

[124] R.M. Rapoport, S.A. Waldman, R. Ginsburg, C.R. Molina, F. Murad, Effects ofglyceryl trinitrate on endothelium-dependent and −independent relaxationand cyclic GMP levels in rat aorta and human coronary artery, J. Cardiovasc.Pharmacol. 10 (1987) 82–89.

[125] R.F. Furchgott, Role of endothelium in responses of vascular smooth muscle,Circ. Res. 53 (1983) 557–573.

[126] R.M. Palmer, A.G. Ferrige, S. Moncada, Nitric oxide release accounts for thebiological activity of endothelium-derived relaxing factor, Nature 327(1987) 524–526.

[127] L.J. Ignarro, G.M. Buga, K.S. Wood, R.E. Byrns, G. Chaudhuri,Endothelium-derived relaxing factor produced and released from artery andvein is nitric oxide, Proc. Natl. Acad. Sci. U. S. A. 84 (1987) 9265–9269.

[128] M.T. Khan, R.F. Furchgott, Additional evidence that endothelium-derivedrelaxing factor is nitric oxide, in: M.J. Rand, C. Raper (Eds.), Pharmacology,Elsevier, Amsterdam, 1987, pp. 341–344.

[129] L.J. Ignarro, R.E. Byrns, G.M. Buga, K.S. Wood, Endothelium-derived relaxingfactor from pulmonary artery and vein possesses pharmacologic andchemical properties identical to those of nitric oxide radical, Circ. Res. 61(1987) 866–879.

[130] L.J. Ignarro, G.M. Buga, R.E. Byrns, K.S. Wood, G. Chaudhuri,Endothelium-derived relaxing factor and nitric oxide possess identicalpharmacologic properties as relaxants of bovine arterial and venous smoothmuscle, J. Pharmacol. Exp. Ther. 246 (1988) 218–226.

[131] L.J. Ignarro, R.E. Byrns, G.M. Buga, K.S. Wood, G. Chaudhuri, Pharmacologicalevidence that endothelium-derived relaxing factor is nitric oxide: use ofpyrogallol and superoxide dismutase to study endothelium-dependent andnitric oxide-elicited vascular smooth muscle relaxation, J. Pharmacol. Exp.Ther. 244 (1988) 181–189.

[132] S. Moncada, R.M. Palmer, E.A. Higgs, The discovery of nitric oxide as theendogenous nitrovasodilator, Hypertension 12 (1988) 365–372.

[133] S. Moncada, R.M. Palmer, E.A. Higgs, The biological significance of nitricoxide formation from L-arginine, Biochem. Soc. Trans. 17 (1989) 642–644.

[134] H.H. Mitchell, H.A. Shonle, H.S. Grindley, The origin of the nitrates in theurine, J. Biol. Chem. 24 (1916) 461–490.

[135] S.R. Tannenbaum, D. Fett, V.R. Young, P.D. Land, W.R. Bruce, Nitrite andnitrate are formed by endogenous synthesis in the human intestine, Science200 (1978) 1487–1489.

[136] L.C. Green, S.R. Tannenbaum, P. Goldman, Nitrate synthesis in the germfreeand conventional rat, Science 212 (1981) 56–58.

[137] E. Hegesh, J. Shiloah, Blood nitrates and infantile methemoglobinemia, Clin.Chim. Acta 125 (1982) 107–115.

[138] D.A. Wagner, V.R. Young, S.R. Tannenbaum, Mammalian nitratebiosynthesis: incorporation of 15NH3 into nitrate is enhanced by endotoxintreatment, Proc. Natl. Acad. Sci. U. S. A. 80 (1983) 4518–4521.

[139] D.J. Stuehr, M.A. Marletta, Mammalian nitrate biosynthesis: mousemacrophages produce nitrite and nitrate in response to Escherichia colilipopolysaccharide, Proc. Natl. Acad. Sci. U. S. A. 82 (1985) 7738–7742.

[140] H.H. Beard, The effect of parenteral injection of synthetic amino acids uponthe appearance, growth and disappearance of the Emge sarcoma in rats,Arch. Biochem. 1 (1943) 177–186.

[141] J.B. Hibbs Jr., Z. Vavrin, R.R. Taintor, l-arginine is required for expression ofthe activated macrophage effector mechanism causing selective metabolicinhibition in target cells, J. Immunol. 138 (1987) 550–565.

[142] J.B. Hibbs Jr., R.R. Taintor, Z. Vavrin, Macrophage cytotoxicity: role forL-arginine deiminase and imino nitrogen oxidation to nitrite, Science 235(1987) 473–476.

[143] T. Deguchi, M. Yoshioka, L-Arginine identified as an endogenous activatorfor soluble guanylate cyclase from neuroblastoma cells, J. Biol. Chem. 257(1982) 10147–10151.

[144] R.M. Palmer, D.S. Ashton, S. Moncada, Vascular endothelial cells synthesizenitric oxide from l-arginine, Nature 333 (1988) 664–666.

[145] R.G. Knowles, M. Palacios, R.M. Palmer, S. Moncada, Formation of nitricoxide from L-arginine in the central nervous system: a transductionmechanism for stimulation of the soluble guanylate cyclase, Proc. Natl.Acad. Sci. U. S. A. 86 (1989) 5159–5162.

[146] I. Sakuma, D.J. Stuehr, S.S. Gross, C. Nathan, R. Levi, Identification of arginineas a precursor of endothelium-derived relaxing factor, Proc. Natl. Acad. Sci.U. S. A. 85 (1988) 8664–8667.

[147] M.A. Marletta, P.S. Yoon, R. Iyengar, C.D. Leaf, J.S. Wishnok, Macrophageoxidation of L-arginine to nitrite and nitrate: nitric oxide is an intermediate,Biochemistry 27 (1988) 8706–8711.

[148] R.M. Palmer, D.D. Rees, D.S. Ashton, S. Moncada, L-arginine is thephysiological precursor for the formation of nitric oxide in

ogical

[

[

[

[

[

[[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

E. Nava, S. Llorens / Pharmacol

endothelium-dependent relaxation, Biochem. Biophys. Res. Commun. 153(1988) 1251–1256.

149] R.M. Palmer, S. Moncada, A novel citrulline-forming enzyme implicated inthe formation of nitric oxide by vascular endothelial cells, Biochem. Biophys.Res. Commun. 158 (1989) 348–352.

150] J.L. Amezcua, R.M. Palmer, B.M. de Souza, S. Moncada, Nitric oxidesynthesized from L-arginine regulates vascular tone in the coronarycirculation of the rabbit, Br. J. Pharmacol. 97 (1989) 1119–1124.

151] D.D. Rees, R.M. Palmer, S. Moncada, Role of endothelium-derived nitricoxide in the regulation of blood pressure, Proc. Natl. Acad. Sci. U. S. A. 86(1989) 3375–3378.

152] S. Moncada, The 1991 ulf von euler lecture: the l-arginine: nitric oxidepathway, Acta Physiol. Scand. 145 (1992) 201–227.

153] S. Moncada, R.M. Palmer, E.A. Higgs, Nitric oxide: physiology,pathophysiology, and pharmacology, Pharmacol. Rev. 43 (1991) 109–142.

154] D.E. Koshland Jr., The molecule of the year, Science 258 (1992) 1861.155] G. Wu, C.J. Meininger, Impaired arginine metabolism and NO synthesis in

coronary endothelial cells of the spontaneously diabetic BB rat, Am. J.Physiol. 269 (1995) H1312–1318.

156] G.M. Buga, R. Singh, S. Pervin, N.E. Rogers, D.A. Schmitz, C.P. Jenkinson, et al.,Arginase activity in endothelial cells: inhibition by NG-hydroxy-L-arginineduring high-output NO production, Am. J. Physiol. 271 (1996) H1988–1998.

157] C. Zhang, T.W. Hein, W. Wang, C.I. Chang, L. Kuo, Constitutive expression ofarginase in microvascular endothelial cells counteracts nitricoxide-mediated vasodilatory function, FASEB J. 15 (2001) 1264–1266.

158] Y. Kakimoto, S. Akazawa, Isolation and identification of N-G,N-G- andN-G,N’-G-dimethyl-arginine, N-epsilon-mono-, di-, and trimethyllysine, andglucosylgalactosyl- and galactosyl-delta-hydroxylysine from human urine, J.Biol. Chem. 245 (1970) 5751–5758.

159] P. Vallance, A. Leone, A. Calver, J. Collier, S. Moncada, Accumulation of anendogenous inhibitor of nitric oxide synthesis in chronic renal failure,Lancet 339 (1992) 572–575.

160] R.J. MacAllister, S.A. Fickling, G.S. Whitley, P. Vallance, Metabolism ofmethylarginines by human vasculature; implications for the regulation ofnitric oxide synthesis, Br. J. Pharmacol. 112 (1994) 43–48.

161] R.F. Furchgott, M.H. Carvalho, M.T. Khan, K. Matsunaga, Evidence forendothelium-dependent vasodilation of resistance vessels by acetylcholine,Blood Vessels 24 (1987) 145–149.

162] T. Nagao, S. Illiano, P.M. Vanhoutte, Heterogeneous distribution ofendothelium-dependent relaxations resistant to NG-nitro-L-arginine in rats,Am. J. Physiol. 263 (1992) H1090–1094.

163] J.J. Hwa, L. Ghibaudi, P. Williams, M. Chatterjee, Comparison ofacetylcholine-dependent relaxation in large and small arteries of ratmesenteric vascular bed, Am. J. Physiol. 266 (1994) H952–958.

164] G. Mehrke, U. Pohl, J. Daut, Effects of vasoactive agonists on the membranepotential of cultured bovine aortic and guinea-pig coronary endothelium, J.Physiol. 439 (1991) 277–299.

165] R.S. Scotland, M. Madhani, S. Chauhan, S. Moncada, J. Andresen, H. Nilsson,et al., Investigation of vascular responses in endothelial nitric oxidesynthase/cyclooxygenase-1 double-knockout mice: key role forendothelium-derived hyperpolarizing factor in the regulation of bloodpressure in vivo, Circulation 111 (2005) 796–803.

166] G.N. Chaldakov, J. Beltowsky, P.I. Ghenev, M. Fiore, P. Panayotov, G. Rancic,et al., Adipoparacrinology?vascular periadventitial adipose tissue (tunicaadiposa) as an example, Cell Biol. Int. 36 (2012) 327–330.

167] G.N. Chaldakov, A.B. Tonchev, I.S. Stankulov, P.I. Ghenev, M. Fiore, L. Aloe,et al., Periadventitial adipose tissue (tunica adiposa): enemy or friendaround, Arch. Pathol. Lab. Med. 131 (2007) 1766, author reply 1766–1767.

168] N.K. Brown, Z. Zhou, J. Zhang, R. Zeng, J. Wu, D.T. Eitzman, et al., Perivascularadipose tissue in vascular function and disease: a review of current researchand animal models, Arterioscler. Thromb. Vasc. Biol. 34 (2014) 1621–1630.

169] C.Y. Miao, Z.Y. Li, The role of perivascular adipose tissue in vascular smoothmuscle cell growth, Br. J. Pharmacol. 165 (2012) 643–658.

170] K.R. Stenmark, M.E. Yeager, K.C. El Kasmi, E. Nozik-Grayck, E.V.Gerasimovskaya, M. Li, et al., The adventitia: essential regulator of vascularwall structure and function, Annu. Rev. Physiol. 75 (2013) 23–47.

171] T. Szasz, G.F. Bomfim, R.C. Webb, The influence of perivascular adiposetissue on vascular homeostasis, Vasc. Health Risk Manag. 9 (2013) 105–116.

172] M. Lafontan, M. Berlan, Fat cell adrenergic receptors and the control of whiteand brown fat cell function, J. Lipid Res. 34 (1993) 1057–1091.

173] A. Giordano, C.K. Song, R.R. Bowers, J.C. Ehlen, A. Frontini, S. Cinti, et al.,Reply to Kreier and Buijs: no sympathy for the claim of parasympatheticinnervation of white adipose tissue, Am. J. Physiol. Regul. Integr. Comp.Physiol. 2913 (2007) R550–552.

174] F. Kreier, R.M. Buijs, Evidence for parasympathetic innervation of whiteadipose tissue, clearing up some vagaries, Am. J. Physiol. Regul. Integr.Comp. Physiol. 293 (2007) R548–R549, author reply R550-542, discussionR553-544.

175] M.A. Kemler, W.F. Kolkman, P.J. Slootweg, M. Kon, Adventitial stripping doesnot strip the adventitia, Plast. Reconstr. Surg. 99 (1997) 1626–1631.

176] Y.J. Gao, A.C. Holloway, Z.H. Zeng, G.E. Lim, J.J. Petrik, W.G. Foster, et al.,

Prenatal exposure to nicotine causes postnatal obesity and alteredperivascular adipose tissue function, Obes. Res. 13 (2005) 687–692.

177] A. Foldes, R.J. Head, I.S. de la Lande, Endogenous histamine in rabbit thoracicaorta, Aust. J. Exp. Biol. Med. Sci. 55 (1977) 89–102.

Research 113 (2016) 125–145 143

[178] S.K. Wilson, D.R. Lynch, S.H. Snyder, Angiotensin-converting enzyme labeledwith [3H]captopril: tissue localizations and changes in different models ofhypertension in the rat, J. Clin. Invest. 80 (1987) 841–851.

[179] L.A. Cassis, K.R. Lynch, M.J. Peach, Localization of angiotensinogenmessenger RNA in rat aorta, Circ. Res. 62 (1988) 1259–1262.

[180] Y. Kameda, Distribution of serotonin-immunoreactive cells around arteriesarising from the common carotid artery in the chicken, Anat. Rec. 227(1990) 87–96.

[181] R.H. Steinhorn, F.C. Morin 3rd., J.A. Russell, The adventitia may be a barrierspecific to nitric oxide in rabbit pulmonary artery, J. Clin. Invest. 94 (1994)1883–1888.

[182] P.J. Pagano, J.K. Clark, M.E. Cifuentes-Pagano, S.M. Clark, G.M. Callis, M.T.Quinn, Localization of a constitutively active, phagocyte-like NADPHoxidase in rabbit aortic adventitia: enhancement by angiotensin II, Proc.Natl. Acad. Sci. U. S. A. 94 (1997) 14483–14488.

[183] H.D. Wang, P.J. Pagano, Y. Du, A.J. Cayatte, M.T. Quinn, P. Brecher, et al.,Superoxide anion from the adventitia of the rat thoracic aorta inactivatesnitric oxide, Circ. Res. 82 (1998) 810–818.

[184] H. Di Wang, S. Hope, Y. Du, M.T. Quinn, A. Cayatte, P.J. Pagano, et al.,Paracrine role of adventitial superoxide anion in mediating spontaneoustone of the isolated rat aorta in angiotensin II-induced hypertension,Hypertension 33 (1999) 1225–1232.

[185] A.L. Kleschyov, B. Muller, C. Schott, J.C. Stoclet, Role of adventitial nitricoxide in vascular hyporeactivity induced by lipopolysaccharide in rat aorta,Br. J. Pharmacol. 124 (1998) 623–626.

[186] H. Zhang, Y. Du, R.A. Cohen, A.V. Chobanian, P. Brecher, Adventitia as asource of inducible nitric oxide synthase in the rat aorta, Am. J. Hypertens.12 (1999) 467–475.

[187] B. Somoza, M.C. Gonzalez, J.M. Gonzalez, F. Abderrahim, S.M. Arribas, M.S.Fernandez-Alfonso, Modulatory role of the adventitia on noradrenaline andangiotensin II responses role of endothelium and AT2 receptors, Cardiovasc.Res. 65 (2005) 478–486.

[188] M.C. Gonzalez, S.M. Arribas, F. Molero, M.S. Fernandez-Alfonso, Effect ofremoval of adventitia on vascular smooth muscle contraction andrelaxation, Am. J. Physiol. Heart Circ. Physiol. 280 (2001) H2876–2881.

[189] F.E. Rey, X.C. Li, O.A. Carretero, J.L. Garvin, P.J. Pagano, Perivascularsuperoxide anion contributes to impairment of endothelium-dependentrelaxation: role of gp91(phox), Circulation 106 (2002) 2497–2502.

[190] T. Cascino, G. Csanyi, I. Al Ghouleh, A.C. Montezano, R.M. Touyz, M.J.Haurani, et al., Adventitia-derived hydrogen peroxide impairs relaxation ofthe rat carotid artery via smooth muscle cell p38 mitogen-activated proteinkinase, Antioxid. Redox Signal. 15 (2011) 1507–1515.

[191] S.J. An, R. Boyd, Y. Wang, X. Qiu, H.D. Wang, Endothelin-1 expression invascular adventitial fibroblasts, Am. J. Physiol. Heart Circ. Physiol. 290(2006) H700–H708.

[192] G.S. Hotamisligil, N.S. Shargill, B.M. Spiegelman, Adipose expression oftumor necrosis factor-alpha: direct role in obesity-linked insulin resistance,Science 259 (1993) 87–91.

[193] S. Verlohren, G. Dubrovska, S.Y. Tsang, K. Essin, F.C. Luft, Y. Huang, et al.,Visceral periadventitial adipose tissue regulates arterial tone of mesentericarteries, Hypertension 44 (2004) 271–276.

[194] T.J. Guzik, P.J. Marvar, M. Czesnikiewicz-Guzik, R. Korbut, Perivascularadipose tissue as a messenger of the brain-vessel axis: role in vascularinflammation and dysfunction, J. Physiol. Pharmacol. 58 (2007) 591–610.

[195] M.R. Skilton, A. Serusclat, A.H. Sethu, S. Brun, S. Bernard, B. Balkau, et al.,Noninvasive measurement of carotid extra-media thickness: associationswith cardiovascular risk factors and intima-media thickness, JACCCardiovasc. Imaging 2 (2009) 176–182.

[196] T.K. Chatterjee, L.L. Stoll, G.M. Denning, A. Harrelson, A.L. Blomkalns, G.Idelman, et al., Proinflammatory phenotype of perivascular adipocytes:influence of high-fat feeding, Circ. Res. 104 (2009) 541–549.

[197] K. Rittig, J.H. Dolderer, B. Balletshofer, J. Machann, F. Schick, T. Meile, et al.,The secretion pattern of perivascular fat cells is different from that ofsubcutaneous and visceral fat cells, Diabetologia 55 (2012) 1514–1525.

[198] J. Padilla, N.T. Jenkins, V.J. Vieira-Potter, M.H. Laughlin, Divergent phenotypeof rat thoracic and abdominal perivascular adipose tissues, Am. J. Physiol.Regul. Integr. Comp. Physiol. 304 (2013) R543–552.

[199] M.K. Owen, J.N. Noblet, D.J. Sassoon, A.M. Conteh, A.G. Goodwill, J.D. Tune,Perivascular adipose tissue and coronary vascular disease, Arterioscler.Thromb. Vasc. Biol. 34 (2014) 1643–1649.

[200] M. Gil-Ortega, B. Somoza, Y. Huang, M. Gollasch, M.S. Fernandez-Alfonso,Regional differences in perivascular adipose tissue impacting vascularhomeostasis, Trends Endocrinol. Metab. 26 (2015) 367–375.

[201] C. Ribiere, A.M. Jaubert, N. Gaudiot, D. Sabourault, M.L. Marcus, J.L. Boucher,et al., White adipose tissue nitric oxide synthase: a potential source for NOproduction, Biochem. Biophys. Res. Commun. 222 (1996) 706–712.

[202] J.E. Shaw, P.W. Ramwell, Release of prostaglandin from rat epididymal fatpad on nervous and hormonal stimulation, J. Biol. Chem. 243 (1968)1498–1503.

[203] M.R. Dashwood, A. Dooley, X. Shi-Wen, D.J. Abraham, D.S. Souza, Doesperiadventitial fat-derived nitric oxide play a role in improved saphenous

vein graft patency in patients undergoing coronary artery bypass surgery? J.Vasc. Res. 44 (2007) 175–181.

[204] L. Chang, L. Villacorta, R. Li, M. Hamblin, W. Xu, C. Dou, et al., Loss ofperivascular adipose tissue on peroxisome proliferator-activatedreceptor-gamma deletion in smooth muscle cells impairs intravascular

1 ogical

44 E. Nava, S. Llorens / Pharmacol

thermoregulation and enhances atherosclerosis, Circulation 126 (2012)1067–1078.

[205] C. Lu, L.Y. Su, R.M. Lee, Y.J. Gao, Mechanisms for perivascular adiposetissue-mediated potentiation of vascular contraction to perivascularneuronal stimulation: the role of adipocyte-derived angiotensin II, Eur. J.Pharmacol. 634 (2010) 107–112.

[206] V. Mohamed-Ali, J.H. Pinkney, S.W. Coppack, Adipose tissue as an endocrineand paracrine organ, Int. J. Obes. Relat. Metab. Disord. 22 (1998) 1145–1158.

[207] P. Trayhurn, I.S. Wood, Adipokines inflammation and the pleiotropic role ofwhite adipose tissue, Br. J. Nutr. 92 (2004) 347–355.

[208] N. Halberg, I. Wernstedt-Asterholm, P.E. Scherer, The adipocyte as anendocrine cell, Endocrinol. Metab. Clin. North Am. 37 (2008) 753–768, x-xi.

[209] Frühbeck G., The adipose tissue: from a passive depot to an active endocrineorgan, In: Obesity, In: Ríos M.S. O.J., Gutiérrez-Fuentes J.A., Barcelona:Elsevier, 2010. pp. 87–106.

[210] J.S. Yudkin, E. Eringa, C.D. Stehouwer, Vasocrine signalling from perivascularfat: a mechanism linking insulin resistance to vascular disease, Lancet 365(2005) 1817–1820.

[211] H.S. Mattu, H.S. Randeva, Role of adipokines in cardiovascular disease, J.Endocrinol. 216 (2013) T17–36.

[212] M. Malinowski, M.A. Deja, K.S. Golba, T. Roleder, J. Biernat, S. Wos,Perivascular tissue of internal thoracic artery releases potent nitric oxideand prostacyclin-independent anticontractile factor, Eur. J. Cardiothorac.Surg. 33 (2008) 225–231.

[213] G. Dubrovska, S. Verlohren, F.C. Luft, M. Gollasch, Mechanisms of ADRFrelease from rat aortic adventitial adipose tissue, Am. J. Physiol. Heart Circ.Physiol. 286 (2004) H1107–1113.

[214] Y.J. Gao, Z.H. Zeng, K. Teoh, A.M. Sharma, L. Abouzahr, I. Cybulsky, et al.,Perivascular adipose tissue modulates vascular function in the humaninternal thoracic artery, J. Thorac. Cardiovasc. Surg. 130 (2005) 1130–1136.

[215] Y.J. Gao, K. Takemori, L.Y. Su, W.S. An, C. Lu, A.M. Sharma, et al., Perivascularadipose tissue promotes vasoconstriction: the role of superoxide anion,Cardiovasc. Res. 71 (2006) 363–373.

[216] M.R. Meyer, N.C. Fredette, M. Barton, E.R. Prossnitz, Regulation of vascularsmooth muscle tone by adipose-derived contracting factor, PLoS One 8(2013) e79245.

[217] Y.J. Gao, Dual modulation of vascular function by perivascular adipose tissueand its potential correlation with adiposity/lipoatrophy-related vasculardysfunction, Curr. Pharm. Des. 13 (2007) 2185–2192.

[218] Y.J. Gao, C. Lu, L.Y. Su, A.M. Sharma, R.M. Lee, Modulation of vascularfunction by perivascular adipose tissue: the role of endothelium andhydrogen peroxide, Br. J. Pharmacol. 151 (2007) 323–331.

[219] R.M. Lee, C. Lu, L.Y. Su, Y.J. Gao, Endothelium-dependent relaxation factorreleased by perivascular adipose tissue, J. Hypertens. 27 (2009) 782–790.

[220] L. Fang, J. Zhao, Y. Chen, T. Ma, G. Xu, C. Tang, et al., Hydrogen sulfide derivedfrom periadventitial adipose tissue is a vasodilator, J. Hypertens. 27 (2009)2174–2185.

[221] Y.C. Lee, H.H. Chang, C.L. Chiang, C.H. Liu, J.I. Yeh, M.F. Chen, et al., Role ofperivascular adipose tissue-derived methyl palmitate in vascular toneregulation and pathogenesis of hypertension, Circulation 124 (2011)1160–1171.

[222] G. Fesus, G. Dubrovska, K. Gorzelniak, R. Kluge, Y. Huang, F.C. Luft, et al.,Adiponectin is a novel humoral vasodilator, Cardiovasc. Res. 75 (2007)719–727.

[223] F.M. Lynch, S.B. Withers, Z. Yao, M.E. Werner, G. Edwards, A.H. Weston, et al.,Perivascular adipose tissue-derived adiponectin activates BK(Ca) channelsto induce anticontractile responses, Am. J. Physiol. Heart Circ. Physiol. 304(2013) H786–795.

[224] A.H. Weston, I. Egner, Y. Dong, E.L. Porter, A.M. Heagerty, G. Edwards,Stimulated release of a hyperpolarizing factor (ADHF) from mesentericartery perivascular adipose tissue: involvement of myocyte BKCa channelsand adiponectin, Br. J. Pharmacol. 169 (2013) 1500–1509.

[225] M.R. Sierra-Honigmann, A.K. Nath, C. Murakami, G. Garcia-Cardena, A.Papapetropoulos, W.C. Sessa, et al., Biological action of leptin as anangiogenic factor, Science 281 (1998) 1683–1686.

[226] G. Fruhbeck, Pivotal role of nitric oxide in the control of blood pressure afterleptin administration, Diabetes 48 (1999) 903–908.

[227] K. Kimura, K. Tsuda, A. Baba, T. Kawabe, S. Boh-oka, M. Ibata, et al.,Involvement of nitric oxide in endothelium-dependent arterial relaxation byleptin, Biochem. Biophys. Res. Commun. 273 (2000) 745–749.

[228] K. Nakagawa, Y. Higashi, S. Sasaki, T. Oshima, H. Matsuura, K. Chayama,Leptin causes vasodilation in humans, Hypertens. Res. 25 (2002) 161–165.

[229] G.A. Payne, L. Borbouse, S. Kumar, Z. Neeb, M. Alloosh, M. Sturek, et al.,Epicardial perivascular adipose-derived leptin exacerbates coronaryendothelial dysfunction in metabolic syndrome via a protein kinase C-betapathway, Arterioscler. Thromb. Vasc. Biol. 30 (2010) 1711–1717.

[230] M.R. Dashwood, A. Dooley, X. Shi-Wen, D.J. Abraham, M. Dreifaldt, D.S.Souza, Perivascular fat-derived leptin: a potential role in improved veingraft performance in coronary artery bypass surgery, Interact Cardiovasc.Thorac. Surg. 12 (2011) 170–173.

[231] B. Galvez-Prieto, B. Somoza, M. Gil-Ortega, C.F. Garcia-Prieto, A.I. de Las

Heras, M.C. Gonzalez, et al., Anticontractile effect of perivascular adiposetissue and leptin are reduced in hypertension, Front. Pharmacol. 3 (2012)103.

[232] F.T. Spradley, D.H. Ho, J.S. Pollock, S.S. Dahl, rats demonstrate enhancedaortic perivascular adipose tissue-mediated buffering of vasoconstriction

Research 113 (2016) 125–145

through activation of NOS in the endothelium, Am. J. Physiol. Regul. Integr.Comp. Physiol. 310 (2016) R286–296.

[233] J.D. Knudson, U.D. Dincer, C. Zhang, A.N. Swafford Jr., R. Koshida, A. Picchi,et al., Leptin receptors are expressed in coronary arteries, andhyperleptinemia causes significant coronary endothelial dysfunction, Am. J.Physiol. Heart Circ. Physiol. 289 (2005) H48–56.

[234] H. Chen, M. Montagnani, T. Funahashi, I. Shimomura, M.J. Quon, Adiponectinstimulates production of nitric oxide in vascular endothelial cells, J. Biol.Chem. 278 (2003) 45021–45026.

[235] Y. Hattori, M. Suzuki, S. Hattori, K. Kasai, Globular adiponectin upregulatesnitric oxide production in vascular endothelial cells, Diabetologia 46 (2003)1543–1549.

[236] W. Xi, H. Satoh, H. Kase, K. Suzuki, Y. Hattori, Stimulated HSP90 binding toeNOS and activation of the PI3-Akt pathway contribute to globularadiponectin-induced NO production: vasorelaxation in response to globularadiponectin, Biochem. Biophys. Res. Commun. 332 (2005) 200–205.

[237] S.W. Watts, A.M. Dorrance, M.E. Penfold, J.L. Rourke, C.J. Sinal, B. Seitz, et al.,Chemerin connects fat to arterial contraction, Arterioscler. Thromb. Vasc.Biol. 33 (2013) 1320–1328.

[238] K.B. Neves, N.S. Lobato, R.A. Lopes, F.P. Filgueira, C.Z. Zanotto, A.M. Oliveira,et al., Chemerin reduces vascular nitric oxide/cGMP signalling in rat aorta: alink to vascular dysfunction in obesity? Clin. Sci. (Lond) 127 (2014) 111–122.

[239] K.B. Neves, A. Nguyen Dinh Cat, R.A. Lopes, F.J. Rios, A. Anagnostopoulou,N.S. Lobato, et al., Chemerin regulates crosstalk between adipocytes andvascular cells through nox, Hypertension 66 (2015) 657–666.

[240] P. Wang, T.Y. Xu, Y.F. Guan, D.F. Su, G.R. Fan, C.Y. Miao, Perivascular adiposetissue-derived visfatin is a vascular smooth muscle cell growth factor: roleof nicotinamide mononucleotide, Cardiovasc. Res. 81 (2009) 370–380.

[241] S. Vallejo, T. Romacho, J. Angulo, L.A. Villalobos, E. Cercas, A. Leivas, et al.,Visfatin impairs endothelium-dependent relaxation in rat and humanmesenteric microvessels through nicotinamide phosphoribosyltransferaseactivity, PLoS One 6 (2011) e27299.

[242] P. Kougias, H. Chai, P.H. Lin, A.B. Lumsden, Q. Yao, C. Chen,Adipocyte-derived cytokine resistin causes endothelial dysfunction ofporcine coronary arteries, J. Vasc. Surg. 41 (2005) 691–698.

[243] H. Yamawaki, N. Tsubaki, M. Mukohda, M. Okada, Y. Hara, Omentin, a noveladipokine, induces vasodilation in rat isolated blood vessels, Biochem.Biophys. Res. Commun. 393 (2010) 668–672.

[244] L. Axelrod, L. Levine, Prostacyclin production by isolated adipocytes,Diabetes 30 (1981) 163–167.

[245] Y. Mendizabal, S. Llorens, E. Nava, Vasoactive effects of prostaglandins fromthe perivascular fat of mesenteric resistance arteries in WKY and SHROBrats, Life Sci. 93 (2013) 1023–1032.

[246] G. Ozen, G. Topal, I. Gomez, A. Ghorreshi, K. Boukais, C. Benyahia, et al.,Control of human vascular tone by prostanoids derived from perivascularadipose tissue, Prostaglandins Other Lipid Mediat. 107 (2013) 13–17.

[247] M.G. Farb, S. Tiwari, S. Karki, D.T. Ngo, B. Carmine, D.T. Hess, et al.,Cyclooxygenase inhibition improves endothelial vasomotor dysfunction ofvisceral adipose arterioles in human obesity, Obesity (Silver Spring) 22(2014) 349–355.

[248] V. van Harmelen, A. Eriksson, G. Astrom, K. Wahlen, E. Naslund, F. Karpe,et al., Vascular peptide endothelin-1 links fat accumulation with alterationsof visceral adipocyte lipolysis, Diabetes 57 (2008) 378–386.

[249] B. Galvez-Prieto, J. Bolbrinker, P. Stucchi, A.I. de Las Heras, B. Merino, S.Arribas, et al., Comparative expression analysis of the renin-angiotensinsystem components between white and brown perivascular adipose tissue,J. Endocrinol. 197 (2008) 55–64.

[250] A. Nguyen Dinh Cat, A.M. Briones, G.E. Callera, A. Yogi, Y. He, A.C.Montezano, et al., Adipocyte-derived factors regulate vascular smoothmuscle cells through mineralocorticoid and glucocorticoid receptors,Hypertension 58 (2011) 479–488.

[251] P. Vargovic, J. Ukropec, M. Laukova, S. Cleary, B. Manz, K. Pacak, et al.,Adipocytes as a new source of catecholamine production, FEBS Lett. 585(2011) 2279–2284.

[252] N. Ayala-Lopez, M. Martini, W.F. Jackson, E. Darios, R. Burnett, B. Seitz, et al.,Perivascular adipose tissue contains functional catecholamines, Pharmacol.Res. Perspect. 2 (2014) e00041.

[253] D.J. Campbell, J.F. Habener, Cellular localization of angiotensinogen geneexpression in brown adipose tissue and mesentery: quantification ofmessenger ribonucleic acid abundance using hybridization in situ,Endocrinology 121 (1987) 1616–1626.

[254] N. Xia, S. Horke, A. Habermeier, E.I. Closs, G. Reifenberg, A. Gericke, et al.,Uncoupling of endothelial nitric oxide synthase in perivascular adiposetissue of diet-Induced obese mice, Arterioscler. Thromb. Vasc. Biol. 36(2016) 78–85.

[255] M. Gil-Ortega, P. Stucchi, R. Guzman-Ruiz, V. Cano, S. Arribas, M.C. Gonzalez,et al., Adaptative nitric oxide overproduction in perivascular adipose tissueduring early diet-induced obesity, Endocrinology 151 (2010) 3299–3306.

[256] I.J. Amber, J.B. Hibbs Jr., R.R. Taintor, Z. Vavrin, Cytokines induce anL-arginine-dependent effector system in nonmacrophage cells, J. Leukoc.Biol. 44 (1988) 58–65.

[257] N. Aoki, M. Siegfried, A.M. Lefer, Anti-EDRF effect of tumor necrosis factor inisolated, perfused cat carotid arteries, Am. J. Physiol. 256 (1989)H1509–1512.

[258] E. Henrichot, C.E. Juge-Aubry, A. Pernin, J.C. Pache, V. Velebit, J.M. Dayer,et al., Production of chemokines by perivascular adipose tissue: a role in the

ogical

[

[

[

[

[

[

[

[

gene expression by chronic systemic inhibition of NOS in lean and obeserats, Physiol. Rep. 2 (2014) e00225.

[273] G. Deng, Y. Long, Y.-R. Yu, M.-R. Li, Adiponectin directly improvesendothelial dysfunction in obese rats through the AMPK–eNOS pathway, Int.J. Obesity 34 (2010) 165–171.

E. Nava, S. Llorens / Pharmacol

pathogenesis of atherosclerosis? Arterioscler. Thromb. Vasc. Biol. 25 (2005)2594–2599.

259] G.A. Payne, H.G. Bohlen, U.D. Dincer, L. Borbouse, J.D. Tune, Periadventitialadipose tissue impairs coronary endothelial function viaPKC-beta-dependent phosphorylation of nitric oxide synthase, Am. J.Physiol. Heart Circ. Physiol. 297 (2009) H460–465.

260] A.J. Donato, G.D. Henson, R.G. Morgan, R.A. Enz, A.E. Walker, L.A. Lesniewski,TNF-alpha impairs endothelial function in adipose tissue resistance arteriesof mice with diet-induced obesity, Am. J. Physiol. Heart Circ. Physiol. 303(2012) H672–679.

261] K. Rittig, K. Staib, J. Machann, M. Bottcher, A. Peter, F. Schick, et al.,Perivascular fatty tissue at the brachial artery is linked to insulin resistancebut not to local endothelial dysfunction, Diabetologia 51 (2008) 2093–2099.

262] J. Ketonen, J. Shi, E. Martonen, E. Mervaala, Periadventitial adipose tissuepromotes endothelial dysfunction via oxidative stress in diet-induced obeseC57Bl/6 mice, Circ. J. 74 (2010) 1479–1487.

263] A.S. Greenstein, K. Khavandi, S.B. Withers, K. Sonoyama, O. Clancy, M.Jeziorska, et al., Local inflammation and hypoxia abolish the protectiveanticontractile properties of perivascular fat in obese patients, Circulation119 (2009) 1661–1670.

264] S.P. Mukherjee, R.H. Lane, W.S. Lynn, Endogenous hydrogen peroxide andperoxidative metabolism in adipocytes in response to insulin and sulfhydrylreagents, Biochem. Pharmacol. 27 (1978) 2589–2594.

265] R.M. Costa, F.P. Filgueira, R.C. Tostes, M.H. Carvalho, E.H. Akamine, N.S.Lobato, H2O2 generated from mitochondrial electron transport chain inthoracic perivascular adipose tissue is crucial for modulation of vascular

smooth muscle contraction, Vascul. Pharmacol. (2016).

266] A. Virdis, E. Duranti, C. Rossi, U. Dell’Agnello, E. Santini, M. Anselmino, et al.,Tumour necrosis factor-alpha participates on the endothelin-1/nitric oxideimbalance in small arteries from obese patients: role of perivascular adiposetissue, Eur. Heart J. 36 (2015) 784–794.

Research 113 (2016) 125–145 145

[267] J. Parker, J. Lane, L. Axelrod, Cooperation of adipocytes and endothelial cellsrequired for catecholamine stimulation of PGI2 production by rat adiposetissue, Diabetes 38 (1989) 1123–1132.

[268] B. Richelsen, Release and effects of prostaglandins in adipose tissue,Prostaglandins Leukot. Essent. Fatty Acids 47 (1992) 171–182.

[269] S.B. Withers, C.E. Bussey, S.N. Saxton, H.M. Melrose, A.E. Watkins, A.M.Heagerty, Mechanisms of adiponectin-associated perivascular function invascular disease, Arterioscler. Thromb. Vasc. Biol. 34 (2014) 1637–1642.

[270] S.B. Withers, L. Simpson, S. Fattah, M.E. Werner, A.M. Heagerty,cGMP-dependent protein kinase (PKG) mediates the anticontractile capacityof perivascular adipose tissue, Cardiovasc. Res. 101 (2014) 130–137.

[271] A.S. Antonopoulos, M. Margaritis, P. Coutinho, C. Shirodaria, C. Psarros, L.Herdman, et al., Adiponectin as a link between type 2 diabetes and vascularNADPH oxidase activity in the human arterial wall: the regulatory role ofperivascular adipose tissue, Diabetes 64 (2015) 2207–2219.

[272] J. Padilla, N.T. Jenkins, P.K. Thorne, K.A. Lansford, N.J. Fleming, D.S. Bayless,et al., Differential regulation of adipose tissue and vascular inflammatory