The apelin receptor APJ: journey from an orphan to a multifaceted regulator of homeostasis

23 Pages·2013·2.44 MB·English

Review A-MO’CARROLLandothers APJandregulationof 219:1 R13–R35
homeostasis
The apelin receptor APJ: journey
from an orphan to a multifaceted
regulator of homeostasis
Correspondence
Anne-MarieO’Carroll,StephenJLolait,LouiseEHarrisandGeorgeRPope shouldbeaddressed
toA-MO’Carroll
HenryWellcomeLaboratoriesforIntegrativeNeuroscienceandEndocrinology,SchoolofClinicalSciences,
Email
UniversityofBristol,DorothyHodgkinBuilding,WhitsonStreet,BristolBS13NY,UK
[email protected]
Abstract
Theapelinreceptor(APJ;genesymbolAPLNR)isamemberoftheGprotein-coupledreceptor KeyWords
genefamily.NeuralgeneexpressionpatternsofAPJ,anditscognateligandapelin,inthe " APJ
brainimplicatetheapelinergicsystemintheregulationofanumberofphysiological " apelin
processes.APJandapelinarehighlyexpressedinthehypothalamo–neurohypophysialsystem, " Gprotein-coupledreceptor
y
og whichregulatesfluidhomeostasis,inthehypothalamic–pituitary–adrenalaxis,which " homeostasis
ol
n controlstheneuroendocrineresponsetostress,andintheforebrainandlowerbrainstem
ri
oc regions,whichareinvolvedincardiovascularfunction.Recently,apelin,synthesisedand
d
n secretedbyadipocytes,hasbeendescribedasabeneficialadipokinerelatedtoobesity,and
E
of thereisgrowingawarenessofapotentialroleforapelinandAPJinglucoseandenergy
nal metabolism.Inthisreviewweprovideacomprehensiveoverviewofthestructure,expression
r
u
o patternandregulationofapelinanditsreceptor,aswellasthemainsecondmessengersand
J
signallingproteinsactivatedbyapelin.Wealsohighlightthephysiologicalandpathological
rolesthatsupportthissystemasanoveltherapeutictargetforpharmacologicalintervention
intreatingconditionsrelatedtoalteredwaterbalance,stress-induceddisorderssuchas
anxietyanddepression,andcardiovascularandmetabolicdisorders.
JournalofEndocrinology
(2013)219,R13–R35
Introduction
G protein-coupled receptors (GPCRs) are activated by a cellularandtissuedistributionsoftheorphanGPCRsand
plethora of molecules including neuropeptides, poly- occasionallyusing‘reversepharmacology’,whereorphan
peptide hormones and non-peptides such as biogenic GPCRs have been used to isolate novel endogenous
amines, lipids, nucleotides and ions. They are classically substances.Thehumanapelinreceptor(APJ,genesymbol
composed of seven membrane-spanning domains and APLNR; O’Dowd et al. 1993) is one such GPCR whose
constitute one of the largest and most diverse gene endogenousligand,apelin,hasbeendescribed(Tatemoto
families in the mammalian genome (Ostrom & Insel etal.1998).BothAPJandapelinhavebeenimplicatedas
2004).SomenovelGPCRsdonothaveobviousendogen- the key mediators of physiological responses to multiple
ousligandsandaretermedorphanreceptors,anumberof homeostaticperturbations,includingcardiovascularcon-
whichappeartobeconstitutivelyactive(Jonesetal.2007, trol, water balance, hypothalamic–pituitary–adrenal
Tanakaetal.2007).Thecognateligandsforsomeofthese (HPA)axisregulationandmetabolichomeostasis.Homeo-
orphan GPCRs have been identified, often based on the staticstabilityiscriticalinmammalianorganisms,andour
http://joe.endocrinology-journals.org (cid:1)2013SocietyforEndocrinology PublishedbyBioscientificaLtd.
DOI:10.1530/JOE-13-0227 PrintedinGreatBritain
Downloaded from Bioscientifica.com at 03/06/2023 01:09:58AM
via free access
Review A-MO’CARROLLandothers APJandregulationof 219:1 R14
homeostasis
knowledge as to how this vital function is regulated and with consensus sites for phosphorylation by protein
how this mechanism can go wrong in pathological kinase A (PKA), palmitoylation and glycosylation
conditionsisstilllimited. (O’Dowd et al. 1993). The N-terminal glycosylation of
GPCRs has been implicated in receptor expression,
stability,correctfoldingofthenascentproteinandligand
The apelin receptor, APJ
binding (Wheatley & Hawtin 1999). Furthermore, the
APJ was first identified as an orphan GPCR, with closest palmitoylation of the C-terminal tail has been reported
identitytotheangiotensinII(AngII)receptor,typeAT to play a role in membrane association and, combined
1a
(O’Dowd et al. 1993). In the ensuing years, the receptor with receptor phosphorylation, these fatty acid modi-
was deorphanised when its cognate ligand, apelin, was fications can influence the internalisation, dimerisation
isolated from bovine stomach extracts (Tatemoto et al. and ligand binding of a GPCR (Huynh et al. 2009).
1998).Recently,theapelinergicsystemhasbeenshownto Structural studies on APJ have determined that amino
becriticallyinvolvedinmultiplehomeostaticprocesses. acids in both the N-terminal (e.g. Asp23 and Glu20) and
C-terminal portions of the receptor are required for
internalisation(Zhouetal.2003a,c,Masrietal.2006).
APJgeneandproteinstructures
ThegeneencodingAPJisintronlessandistermedAPLNRin
Generegulation
humansandAplnrintheratandmouse.APLNRencodesa
380-amino acid protein and is located on chromosome The regulation of APJ gene expression has not been
11q12(O’Dowdetal.1993).Inthemouseandratthegenes extensively characterised to date. At the transcriptional
encode377-aminoacidproteins(Devicetal.1999,Hosoya level, the region with the highest rat APJ gene promoter
y
og et al. 2000, O’Carroll et al. 2000) and are present at the activity is found between K966 and K165bp (O’Carroll
nol chromosomallocations2E1and3q24respectively.Human etal.2006).Electrophoreticmobilityshift,super-shiftand
ri
oc APJ shares 92% amino acid sequence homology with competition assays have indicated that the promoter is
d
n mouseAPJ(Devicetal.1999)and90%homologywithrat under complex regulation by Sp1, oestrogen receptor,
E
of APJ,andthereis96%homologybetweenratandmouseAPJ glucocorticoid receptor and CCAAT enhancer-binding
nal (O’Carroll et al. 2000), indicating a strong evolutionary protein g (C/EBPg (CEBPG)) transcription factors, with
r
ou conservationofthegene.Furthermore,APJorthologuesare Sp1being implicatedas amajor regulatorof rat APJgene
J
presentinanumberofspeciesincludingtheAfricanclawed promoteractivity(O’Carrolletal.2006).
frog,rhesusmacaque,cowandzebrafish(Devicetal.1996, Anumberofsingle-nucleotidepolymorphisms(SNPs)
Margulies et al. 2001, Tucker et al. 2007, Schilffarth et al. have been reportedfor APLNR. A large study of SNPs has
2009) – the latter has w50% amino acid homology with found an association of a SNP (rs9943582 (G/A)) in the
thatofhumanAPJ.ThepromoterregionoftheratAPJgene 50 flanking region Sp1-binding site of APLNR with
is TATA less, but contains a potential CAAT box at bp susceptibility to brain infarction (Hata et al. 2007), while
positionK1257relativetotheinitiatingATGandanumber the 50UTR G212A variant of APLNR has been reported to
ofactivatorprotein1andspecificityprotein1(Sp1)motifs be associated with slower heart failure progression in
(O’Carrolletal.2006).Therearetwotranscriptionalstart idiopathic dilated cardiomyopathy (Sarzani et al. 2007).
sitesatK247andK210bp(O’Carrolletal.2006).Although Additionally, two APLNR SNPs, rs7119375 (G/A) and
thestructure/functionofthehumanAPJgenepromoteris rs10501367 (G/A), in the Han Chinese population (Niu
notknown,twotranscriptvariants(oneofwhichencodes et al. 2010) and the G212A polymorphism in Italian
the full-length APJ protein and the other which may be patients (Falcone et al. 2012) may be associated with
non-coding)havebeenannotatedintheNationalCenter hypertension,withtwofurtherSNPs(rs2282623(C/T)and
forBiotechnologyInformation(NCBI)database.ThecDNA rs746886(C/T))beingreportedtobeassociatedwithblood
sequencetransposedonthegenesequence(NCBIreference pressure (BP) responses to dietary sodium interventions
sequence NC_000011.9) reveals no introns interrupting (Zhaoetal.2010).
theprotein-codingsequence,similartothatfoundinthe APJ gene is up-regulated in response to acute and
ratandmouse.TheredonotappeartobeanyAPJsubtypes repeated stress (O’Carroll et al. 2003), changes that are
ingenedatabasessuchasGenBankandEnsembl. likely to be glucocorticoid dependent. There is also
The protein structure of APJ is typical of a GPCR, evidence that the endogenous ligand, apelin, regulates
containing seven hydrophobic transmembrane domains, the expression of APJ within the gastrointestinal tract
http://joe.endocrinology-journals.org (cid:1)2013SocietyforEndocrinology PublishedbyBioscientificaLtd.
DOI:10.1530/JOE-13-0227 PrintedinGreatBritain
Downloaded from Bioscientifica.com at 03/06/2023 01:09:58AM
via free access
Review A-MO’CARROLLandothers APJandregulationof 219:1 R15
homeostasis
(Wangetal.2009),whilerecentlytheexpressionofAPJhas dimeric protein, as a consequence of disulphide bridges
been shown to be up-regulated in adipose tissues by formedbetweencysteineresidues(Leeetal.2005).
insulin(Drayetal.2010). Thereareseveralmatureformsoftheapelinpeptide.
AsthesequenceofthepeptidepurifiedbyTatemotoetal.
(1998)correspondedtothe36C-terminalaminoacidsof
Apelin
the preproapelin protein, it was predicted that apelin-36
APJ remained an orphan receptor until 1998 when wouldconstituteamatureformofthepeptide.Addition-
Tatemotoetal.(1998)identifieda36-aminoacidpeptide ally, as the C-terminal portion of preproapelin also
termedapelin,forAPJendogenousligand. contained lysine (Lys, K) and arginine (Arg, R) residues,
andgiventheirpotentialassitesforproteolyticcleavage,
the existence of apelin-17 and apelin-13 peptides was
Apelingeneandproteinstructures
predicted, along with a pyroglutamylated form of
Thegeneencodinghumanapelin,termedAPLN,islocated apelin-13 ((Pyr1)apelin-13) (Fig. 1). These mature forms
on chromosome Xq25–26.1 and possesses one intron of apelin lack cysteine residues and are probably only
within its open reading frame of w6kb. In the rat and present in monomeric form. The likely secondary
mouse, the genes are termed Apln and are located at structures of apelin-36 and apelin-13 have been
chromosomallocationsXq35andXA3.2respectively.The determined in aqueous solution, indicating that both
corepromoterregionsofthesegeneshavebeenidentified possess an unordered structure (Fan et al. 2003). The
as K207/K1 and K100/C74bp in rats and humans amino acid sequence homology of the mature apelin-36
respectively (Wang et al. 2006). Similar to APJ, a CAAT peptide is more conserved between species than that of
box,butnoTATAbox,sequenceispresentintheratand preproapelin, with 86–100% homology between bovine,
y
og humanpromoterregions(Wangetal.2006).Furthermore, human,ratandmouseaminoacidsequences,whilethe23
ol
n ratandhumanpreproapelincDNAsdonothaveaclassical C-terminal amino acids have 100% homology between
ri
oc Kozakconsensussequence(Kozak1996) surrounding the species (Habata et al. 1999), suggesting an important
d
n initiatingmethioninecodon(Leeetal.2000). physiologicalrole.
E
of Human and bovine APLN cDNA sequences encode a Although APJ does not bind Ang II (O’Dowd et al.
nal 77-amino acid preproprotein (preproapelin) (Tatemoto 1993), apelin-13 shares a limited homology (four amino
r
u
o et al. 1998) containing a hydrophobic rich N-terminal acids) with the vasoconstrictive peptide (Lee et al. 2000).
J
region, likely to be a secretory signal sequence (for a Moreover, Ang I-converting enzyme 2 (ACE2), which
review, see Rapoport (2007)). Bovine, human, rat and catalyses the C-terminal dipeptide cleavage of Ang I to
mouse preproapelin precursors (Habata et al. 1999) have AngII,orAngIItoAng1–7(Tipnisetal.2000),alsoactson
76–95%homologyandappeartoexistendogenouslyasa apelin-13 with a high catalytic efficiency, removing the
A
pGlu Arg Pro Arg Leu Ser His Lys Gly Pro Met Pro Phe
B
Gln Arg Pro Arg Leu Ser His Lys Gly Pro Met Pro Phe
C
Lys Phe Arg Arg Gln Arg Pro Arg Leu Ser His Lys Gly Pro Met Pro Phe
D
Lys Phe Arg Arg Gln Arg Pro Arg Leu Ser His Lys Gly Pro Met Pro Phe
Arg
Arg
Gly
Gly Gln Trp Ala Gly Pro Gly Thr Arg Ser Thr Arg Pro Lys Val Leu
Figure1
Aminoacidsequenceofmatureratapelinisoforms.Aminoacidsequencesof(A)(Pyr1)apelin-13,(B)apelin-13,(C)apelin-17and(D)apelin-36.Blackcircled
residuesindicatethoseidenticalbetweenhuman,bovine,ratandmouse.
http://joe.endocrinology-journals.org (cid:1)2013SocietyforEndocrinology PublishedbyBioscientificaLtd.
DOI:10.1530/JOE-13-0227 PrintedinGreatBritain
Downloaded from Bioscientifica.com at 03/06/2023 01:09:58AM
via free access
Review A-MO’CARROLLandothers APJandregulationof 219:1 R16
homeostasis
C-terminal phenylalanine (Phe, F) residue (Vickers et al. expression of apelin in 3T3-L1 adipocytes via the p38
2002). However, this cleavage may not inactivate the MAPK pathway (Jiang et al. 2013). Furthermore, studies
peptide, as the apelin isoform K16P, which lacks the on white adipocytes have found that the peroxisome
terminal Phe, while ineffective at inducing receptor proliferator-activated receptor g co-activator 1a also
internalisation or regulating blood pressure (BP) (effects up-regulates the expression of apelin, possibly indicating
associated with the full peptide), still binds to APJ and that apelin plays a role in energy metabolism
inhibitsforskolin-stimulatedcAMPproduction(ElMessari (Mazzucotelli et al. 2008), while recently in diabetic rats,
etal.2004). ghrelin has been reported to reduce apelin mRNA
synthesisandreleaseintothelumen(Coskunetal.2013).
Generegulation
APJ distribution
The regulation of apelin gene expression is mediated by
several effectors, with the involvement of a number of Although it is clear that APJ and apelin mRNAs and
transcriptionfactors.ASNPstudyhasreportedaprobable proteinsarewidelydistributedintheCNSandperipheral
role for Sp1 in the regulation of the expression of APLN tissues, whether the levels of mRNAs present in most of
(Hata et al. 2007). Additionally, the cytokine tumour the regions of the brain and tissues are functionally
necrosis factor-a (TNFa) has been reported to induce the relevantisnotyetknown.
expression of apelin via phosphatidylinositol 3-kinase
(PI3K), c-Jun N-terminal kinase (JNK) and MEK1/2 in
Humandistribution
adipocytes (Daviaud et al. 2006). Furthermore, in studies
using lipopolysaccharide (LPS) and cytokines to elicit an EarlystudiesoftheexpressionofAPJmRNAbynorthern
y
og immune response in rodents, the expression of apelin blot and quantitative PCR (qPCR) analyses have revealed
nol mRNAhasbeenreportedtobeup-regulated,involvingthe strongest signals in the human caudate nucleus, corpus
ri
oc Jak/Statpathway,whilestudiesusingchromatinimmuno- callosum, hippocampus, substantia nigra, subthalamic
d
n precipitation (ChIP) have revealed the binding of Stat3 nucleus, medulla and spinal cord (Matsumoto et al.
E
of (Han et al. 2008). Mutagenesis, electrophoretic mobility 1996,Edingeretal.1998,Medhurstetal.2003).Recently,
nal shift assays and ChIP have also been used in vitro to the expression of APJ mRNA has also been demonstrated
r
ou determine whether upstream stimulatory factors 1 and 2 in the human cortex and hippocampus using a sensitive
J
(USF1andUSF2)areinvolvedintheexpressionofapelin GPCR gene array profiling method – interestingly, APJ
inthebreast,whileinvitroChIPanalyseshaveshownthat transcripts have also been detected in human bone
endogenous USF up-regulates the expression of apelin in marrowstromalcelllines(Hansenetal.2007).Transcrip-
the lactating rat breast (Wang et al. 2006). Putative tomicanalysisofmultiplebrainregionsofhumandonors
hypoxia response elements (HREs) are present in the hasrevealedawidespreadcentralexpressionofAPJmRNA
apelin gene promoter and intron sequence of various with high levels in samples including the hippocampus
species in silico (Cox et al. 2006), and hypoxia has (e.g. CA4 region), habenular nuclei, paraventricular
subsequently been found to up-regulate the expression nucleus (PVN) of the thalamus, supraoptic nucleus
ofapelinincardiacmyocytes(Ronkainenetal.2007),and (SON) of the hypothalamus and various hindbrain
hypoxia-induciblefactor1a(HIF1a)hasbeenreportedto structures (see Allen Brain Atlas: www.brain-map.org).
induce the expression of apelin in adipocytes (Glassford The salient feature of these studies is that APJ has been
et al. 2007). Additional studies have determined that reported to have a widespread central distribution;
hypoxia-inducible apelin up-regulation is mediated by although the function of APJ in the majority of brain
HIF1aataHREwithintheAPLNintron(conservedinrat regions is unknown, foremost among those regions
and mouse apelin genes) between C813 and C826bp probablyimportantfromafunctionalperspectiveinclude
(Eyriesetal.2008),indicatingthatapelinmayplayarole thePVNandSONofthehypothalamus.
inthehomeostaticresponsetolowoxygenlevels.Insulin Intheperiphery,theexpressionofhumanAPJmRNA
hasalsobeenshowntoincreasetheexpressionofapelinin wasoriginallyreportedtobestrongestinthespleen,with
humanandmouseadipocytes,viaPI3K,proteinkinaseC less expression being reported for the small intestine,
(PKC), mitogen-activated and ERK kinase (MAPK) 1 colonicmucosaandovary(Edingeretal.1998).Abroader
(Boucher et al. 2005) and HIF1a (Glassford et al. 2007), qPCR study has also reported strongest expression in the
while aldosterone has been shown to decrease the spleen,withhighlevelsalsobeingreportedtobepresent
http://joe.endocrinology-journals.org (cid:1)2013SocietyforEndocrinology PublishedbyBioscientificaLtd.
DOI:10.1530/JOE-13-0227 PrintedinGreatBritain
Downloaded from Bioscientifica.com at 03/06/2023 01:09:58AM
via free access
Review A-MO’CARROLLandothers APJandregulationof 219:1 R17
homeostasis
intheplacentaandweakerlevelsinthelung,stomachand In addition, (Pyr1)apelin-13-binding sites are present in
intestine (Medhurst et al. 2003). (Pyr1)apelin-13-binding the lung and heart and, to a lesser degree, in the kidney
sitescanbefoundwithinthemediaandintimallayersof cortex (Katugampola et al. 2001). Structures showing the
musculararteriesandlargeelasticarteriesandveins,while strongestexpressionofAPJandapelingenesintheratare
in the lung, apelin-binding sites have a predominantly showninFig.2.
vascular localisation (Katugampola et al. 2001). Further-
more, APJ distribution in cardiovascular tissues, as
Mousedistribution
demonstratedbyimmunohistochemistry(IHC),indicates
APJtobepresent inventricular cardiomyocytes, vascular Negligible levels of APJ expression can be found in the
smooth muscle cells (VSMCs) and intramyocardial endo- whole brain, cerebellum, hypothalamus, hippocampus
thelialcells(Kleinzetal.2005). and olfactory bulb (Medhurst et al. 2003, Regard et al.
2008);however,recently,adetailedISHHcharacterisation
of APJ distribution in the mouse has revealed a very
Ratdistribution
restricted localisation in the CNS, with strong hybrid-
The expression of APJ mRNA in the rat CNS has been isation specifically in the PVN and SON and also in the
mapped by a variety of techniques including northern anterior pituitary, with marginally lower levels in the
blotting,insituhybridisationhistochemistry(ISHH),IHC, posterior pituitary (Pope et al. 2012). This suggests a
receptor autoradiography and qPCR. At a detailed ana- speciesdifferenceinthecentralandpituitarydistributions
tomical resolution using ISHH, discrete but significant of APJ mRNA between mouse and rat. While the
expressionofAPJmRNAcanbefoundthroughouttherat significance of this is not known, it may reflect a more
brain(forcomprehensivedetails,seeDeMotaetal.(2000), extensive role for apelin in mouse pituitary function.
y
og Hosoya et al. (2000), Lee et al. (2000), O’Carroll et al. Additionally, strong hybridisation can be found in the
nol (2000),Medhurstetal.(2003)andXia&Krukoff(2003)), lung, heart, adrenal cortex, renal medulla, ovary and
ri
oc particularly in the PVN and SON where APJ mRNA is uterus. These findings confirm the findings of previous
d
n present in arginine–vasopressin (VP)-expressing cells qPCRstudies,withlowerlevelsalsobeingreportedforthe
E
of (Reaux et al. 2001, O’Carroll & Lolait 2003). In many of thyroid, kidney, spleen, pancreas, skeletal muscle and
nal these regions of thebrain it has been confirmed that the adiposetissues (Medhurstetal. 2003, Regardetal. 2008).
r
ou APJ gene is translated into immunoreactive protein – High levels of APJ-binding sites can be found in the
J
interestinglyAPJimmunoreactivitycanbefoundinboth anterior pituitary, while lower levels can be observed in
neuronalandglialcellpopulations(Medhurstetal.2003). the posterior pituitary, PVN and SON. In the periphery,
(Pyr1)apelin-13-bindingsitescanbefoundinthemolecu- strong receptor binding can be observed in those tissues
lar layer of the cerebellum, the basal surface of the that exhibit strong ISHH signals, indicating a good
hypothalamic diencephalon and the PVN (Katugampola correlation between receptor transcription and trans-
etal.2001,Hazelletal.2012).Inthepituitary,conflicting lation (Pope et al. 2012). Recently, in the mouse embryo
patternsofAPJmRNAexpressionhavebeenreported,with (E9.5–10.5) cardiovascular system, APJ mRNA has been
labelling of the anterior and intermediate lobes being shown to predominate in the endothelial layers of
describedinonestudy(DeMotaetal.2000),asopposedto arteries and veins and in the endocardial layer of the
a moderately strong signal in the anterior lobe alone in heart (Kang et al. 2013), while APJ protein has been
another investigation (O’Carroll et al. 2000). In contrast, shown to be present in mouse hepatocytes and hepatic
APJ immunoreactivity has been found in the nerve tissue (Chu et al. 2013).
terminalsoftheratposteriorpituitary(Tobinetal.2008).
RT-PCRstudieshavereportedhighlevelsofAPJinthe
lung and heart, with lower levels being reported for the Apelin distribution
placenta, thyroid gland, skeletal muscle, costal cartilage,
Humandistribution
ovary, uterus and adipose tissues (Hosoya et al. 2000,
Medhurst et al. 2003). More detailed ISHH studies have Preproapelin transcripts are present in the human CNS,
shown prominent cellular labelling of rat APJ mRNA in with highest levels being present in the thalamus and
the parenchyma of the lung, heart, a subpopulation of frontal cortex and lower levels in the hypothalamus,
glomeruli in the kidney and in the thecal cell layer midbrain,caudate,hippocampusandbasalforebrain(Lee
and corpora lutea in the ovary (O’Carroll et al. 2000). et al. 2000). A strong signal can also be observed in the
http://joe.endocrinology-journals.org (cid:1)2013SocietyforEndocrinology PublishedbyBioscientificaLtd.
DOI:10.1530/JOE-13-0227 PrintedinGreatBritain
Downloaded from Bioscientifica.com at 03/06/2023 01:09:58AM
via free access
Review A-MO’CARROLLandothers APJandregulationof 219:1 R18
homeostasis
CNS main sites of expression
PVN: apelin and APJ
SON: apelin and APJ
Pituitary: apelin and APJ
Endothelial cells: apelin and APJ
Osteoblasts: apelin and APJ
Thymus: –
Heart: apelin and APJ
Lung: apelin and APJ
Liver: apelin
Adrenal gland: apelin and APJ
Adipocytes: apelin and APJ
Kidney: apelin and APJ
Stomach: apelin and APJ
Spleen: –
Pancreas: APJ
Ovary: apelin and APJ
Intestine: apelin and APJ
Uterus: apelin and APJ
Bladder: APJ
Mammary gland: apelin and APJ
Colon: apelin and APJ
Testis: apelin
Figure2
y ApelinandAPJgeneexpressioninrattissues.Geneexpressionofapelin/APJ rattissueswherebothapelinand/orAPJgeneandimmunoreactive
g
o intherat(seetextfordetails).Therehavebeenfewerstudiesdemonstrating protein/bindingsiteshavebeenfound(andmaybefunctionallyrelevant)
ol
n theexpressionofapelinand/orAPJproteinintherat(orotherspecies includethebrain,pituitary,lung,heart,gastrointestinaltract,liverand
cri includinghumans)ordeterminingwhetherapelinandAPJarelocalisedin kidney(seetextandreferencesHus-Cithareletal.(2008),Wangetal.(2009),
o
d differentcellpopulationsorco-expressedwithinagiventissue.Examplesof Zengetal.(2009)andPiairoetal.(2011)fordetails).
n
E
f
o
spinal cord and pituitary gland, with lower levels being directedagainstapelin-17.Furtherstudiesfocusingonthe
al
n
r observed in many regions of the brain, including PVN, SON, median eminence and dorsolateral accessory
u
o
J the amygdala, corpus callosum and substantia nigra magnocellular nucleus have shown strong labelling for
(Medhurstetal.2003). apelinandco-localisationwithVP(Brailoiuetal.2002,De
In human peripheral tissues, strong levels of apelin Mota et al. 2004, Reaux-Le Goazigo et al. 2004) and with
expressioncanbefoundintheplacenta,withlowerlevels oxytocin (OT; Brailoiu et al. 2002) in the SON and PVN.
beingfoundintheheart,lungandkidney(Medhurstetal. Apelin is also co-localised with adrenocorticotrophin
2003). Immunoreactive apelin is present in the vascular (ACTH) in corticotrophs and to a lesser extent with
endothelial cells of large conduit vessels, such as the growthhormoneinsomatotropesintheanteriorpituitary
coronary artery and saphenous vein, blood vessels of the (Reaux-LeGoazigoetal.2007).
kidney and adrenal gland, and vascular and endocardial In rat peripheral tissues, apelin levels are high in the
endothelial cells of the atria and ventricles (Kleinz & mammarygland,ovary,heartandadiposetissues(Habata
Davenport2004). et al. 1999), as well as in the kidney, adrenal gland,
intestine, skeletal muscle, vas deferens, testis and uterus
(Leeetal.2000,O’Carrolletal.2000,Kawamataetal.2001,
Ratdistribution
Medhurst et al. 2003). The expression of apelin mRNA is
Apelin mRNA has a more abundant and widespread increased in the mammary gland during pregnancy and
expression than that of APJ mRNA within the rat CNS. lactation, reaching a maximum level around parturition,
High levels of preproapelin transcripts can be found in andthepresenceofapelininrat,bovineandhumanmilk
several regions including the cerebral cortex, claustrum, hasbeenreported(Habataetal.1999).
anterior and posterior cingulate, retrosplenial area and Thereappeartobediscrepanciesinthedistributionof
thalamicnuclei(seeLeeetal. (2000)).Reauxetal. (2002) rat preproapelin mRNA and that of apelin immunoreac-
have described an extensive study of apelin protein tivity. High levels of preproapelin mRNA have been
distribution in the rat brain, using primary antibodies reported to be present in the hippocampus and cerebral
http://joe.endocrinology-journals.org (cid:1)2013SocietyforEndocrinology PublishedbyBioscientificaLtd.
DOI:10.1530/JOE-13-0227 PrintedinGreatBritain
Downloaded from Bioscientifica.com at 03/06/2023 01:09:58AM
via free access
Review A-MO’CARROLLandothers APJandregulationof 219:1 R19
homeostasis
cortex where no apelin immunoreactivity is detected, apelin-17 and apelin-13 and a pyroglutamylated isoform
whereas in some cerebral regions, e.g. the thalamus, the (Pyr1)apelin-13 (Tatemoto et al. 1998). These isoforms
bed nucleus of the stria terminalis and the median were predicted from the potential basic amino acid
eminence, the opposite has been observed (Lee et al. cleavage sites present in the primary structure of pre-
2000,Reauxetal.2002).Thismaybeduetotheinabilityof proapelin (Habata et al. 1999). There appear to be three
theantibodyusedinthesestudiestodetecttheendogen- active apelin isoforms in bovine milk and five in
ousapelinisoformsintheseregionsorbecausethelevels bovine colostrum, although the precise isoforms were
of preproapelin mRNA in these regions are too low to not identified in the study of Habata et al. (1999).
allow detection by ISHH. Additionally, there are other Subsequently, a study carried out using gel filtration
regions, e.g. olfactory regions, piriform and entorhinal chromatography revealed the presence of apelin-36
cortex and dentate gyrus, where APJ mRNA is expressed, and (Pyr1)apelin-13 in bovine colostrum (Hosoya et al.
but where there is no apelin immunostaining. This 2000).Apelin-36appearstobethemostprevalentisoform
localisation may imply that APJ synthesised in these cell in the lung, testis and uterus, while apelin-36 and
bodiesistransportedtoaxonterminalsinotherregionsof (Pyr1)apelin-13 predominate in the mammary gland
thebrainwheretheligandisexpressedandwhereAPJmay (Kawamata etal. 2001), and asingle short form of apelin
befunctional. Conversely, APJmRNA is not expressed in corresponding to (Pyr1)apelin-13 is present in the hypo-
the rat testis where moderate levels of apelin mRNA are thalamus (De Mota et al. 2004). The latter is also
present. It is possible that the low levels of (possibly the predominant form in rat whole brain and plasma
rapidly turning-over) APJ mRNA are below the detection (DeMotaetal.2004)andinhumancardiactissue(Maguire
threshold of ISHH or that, assuming that apelin of et al. 2009), while apelin-17 is present at a lower level
testicular origin is not redundant, the peptide is binding in the rat hypothalamus and plasma (De Mota et al.
y
og to a unknown, perhaps related receptor. We cannot 2004). In human plasma, all three isoforms, apelin-13,
nol excludethepossibility thattesticularAPJisalsodevelop- (Pyr1)apelin-13 and apelin-17, are present (Reaux et al.
ri
oc mentally regulated since the expression of APJ mRNA 2002, Azizi etal.2008). InChinesehamster ovary (CHO)
d
n appears to be higher in infant peripheral tissues than in cells engineered to express the cloned human APJ,
E
of adultperipheraltissues(Hosoyaetal.2000). apelin-13 binds with high affinity to and associates with
nal APJmoreefficientlythanapelin-36andrapidlydissociates
r
ou from APJ (Hosoya et al. 2000). On the other hand,
J Mousedistribution
apelin-36 has a higher affinity for APJ in this cell line
A single qPCR distribution study has been conducted on (K Z6.3pM (apelin-36) vs K Z22.3pM (apelin-13)
d d
the expression of preproapelin in the mouse where the (Hosoya et al. 2000)) and it is more difficult to dissociate
highest signal hasbeenfoundinthewhole brain, witha itfromthereceptor(Kawamataetal.2001).
moderatesignalintheheart,kidneyandlung,andalow The C-terminal region of the apelin peptide may
signal in the testis, spleen, ovary and muscle (Medhurst beresponsibleforitsoverallbiologicalactivity.N-terminal
etal.2003). deletionsofapelin-17revealthatthe12C-terminalamino
acidsmaybethecorerequirementsfortheinternalisation
and biological potency of APJ (El Messari et al. 2004).
APJ signalling
Apelin-17 induces the internalisation of APJ, which
APJ binds numerous apelin isoforms and signalsthrough decreases with every N-terminal deletion to apelin-12,
various G proteins to a variety of signalling pathways to whilethedeletionoftheterminalFaminoacidresultsina
culminateindifferentpatternsofactivationanddesensi- peptide that no longer internalises APJ or affects arterial
tisation that may be tissue- and cell type-specific. BP. The N-terminal residues within the RPRL motif
Recently, APJ has also been reported to heterodimerise (residues 2–5) of apelin-13 are critical for functional
with other GPCRs and to signal in the absence of an potency (Medhurst et al. 2003), and the C-terminal
endogenousligand. sequence KGPM (residues 8–11) is important for binding
activity and for internalisation (Fan et al. 2003). In
contrast, the five N-terminal and two C-terminal amino
Endogenousapelinisoforms
acids of apelin-17 are not required for binding of the
The initial synthesis of apelin isoforms included the peptide to APJ or activation of receptor signalling (e.g.
mature apelin-36, as well as the C-terminal fragments of cAMP production) (El Messari et al. 2004). Although this
http://joe.endocrinology-journals.org (cid:1)2013SocietyforEndocrinology PublishedbyBioscientificaLtd.
DOI:10.1530/JOE-13-0227 PrintedinGreatBritain
Downloaded from Bioscientifica.com at 03/06/2023 01:09:58AM
via free access
Review A-MO’CARROLLandothers APJandregulationof 219:1 R20
homeostasis
may indicate a possible dissociation between the confor- in the identification of a competitive antagonist at APJ
mational states of the receptor responsible for receptor (Macalusoetal.2011).
signalling and internalisation, it is also possible that Morerecently,small-moleculeagonistsandantagonists
differentligandisoformsmayinducedifferentialreceptor of APJ have been identified – the non-peptidic E339-3D6
trafficking and signalling. These studies provide infor- is a partial agonist of APJ in the inhibition of cAMP
mationonthestructuralimportanceofkeyapelinresidues production, possesses the ability to act as a vasorelaxant
critical for efficient binding, activity and internalisation, of the rat aorta ex vivo, and is a full agonist in terms of
whichhaveprovedsignificantinthedesignandsynthesis APJinternalisation(Iturriozetal.2010);theCXCR4small-
ofapelinanalogues. molecule ALX40-4C (N-a-acetyl-nona-D-arginine amide),
whichinhibitsCXCR4interactionswithhumanimmuno-
deficiency virus (HIV)-1, acts as an antagonist of APJ
Apelin-13(F13A):anAPJantagonist
internalisation(Zhouetal.2003b)andML221,akojicacid-
The first structural studies on apelin activity involved based small molecule, antagonises apelin-13-mediated
the replacement of the C-terminal F residue of (Pyr1)- activation of APJ in the inhibition of cAMP production
apelin-13withalanine(Ala,A),ananaloguetermedF13A. (Maloneyetal.2012).Theseagonistsandantagonistsmay
Unlike (Pyr1)apelin-13, F13A is ineffective at inhibiting providenewtoolstoexplorethefunctionoftheapelinergic
forskolin-stimulated cAMP accumulation in CHO cells system and deliver vital information that could lead to
transfected with rat APJ (De Mota et al. 2000) and anta- potentialpharmaceuticaltherapiestargetedatAPJ.
gonisesapelin-13-induceddecreasesinBP(Leeetal.2005).
Additionally, F13A exhibits an approximately eightfold
Receptoroligomers
lower potency than (Pyr1)apelin-13 in intracellular
y
og calcium mobilisation; an approximately threefold lower Like many other members of the GPCR superfamily,
nol inhibition of cAMP accumulation; and between 2- and APJ may heterodimerise with other GPCRs to modulate
ri
oc 14-fold lower receptor binding efficiency for human APJ establishedsignaltransductionpathwaysinculturedcells.
d
n expressed inavariety of celllines (Medhurstetal. 2003). Using co-immunoprecipitation and fluorescence reso-
E
of However, for human APJ in vitro, F13A exhibits binding, nance energy transfer studies, APJ was first reported to
nal calcium mobilisation and internalisation responses com- dimerise with the Ang II receptor AT1, which results in
r
ou parabletothoseof(Pyr1)apelin-13(Fanetal.2003),while aninhibitoryeffectofapelinonAngIIsignallingandon
J
inhumancardiactissue,F13Acompetesforbindingwith an Ang II-mediated model of atherosclerosis (Chun et al.
[125I]-(Pyr1)apelin-13 in the left ventricle and effectively 2008). A further study has established that APJ hetero-
constricts endothelium-denuded saphenous vein (Pitkin dimeriseswiththek-opioidreceptor(KOR;Lietal.2012).
et al. 2009). Therefore, although this mutant peptide has Treatmentwithapelin-13ortheKORliganddynorphinA
been reported to act as an antagonist, it may in fact act induceshigherlevelsofERK1/2activationinHEK293cells
asacompetitiveagonistatAPJ. stably transfected with APJ and KOR than in HEK293
cells transfected with either APJ or KOR alone. This
activation is mediated by increased PKC and decreased
Cyclicandotherapelinanalogues
PKA activities and results in an increase in cell prolifer-
The use of cyclic analogues is a proven method of ation (Li et al. 2012). The possible heterodimerisation of
studyingconformationalconfigurationastheseanalogues the native APJ with other co-expressed GPCRs or other
restrictthestructuralbackboneofthepeptideandconfine signalling proteins in vivo may have possible important
their secondary structure. In a study carried out to consequences for understanding APJ function and in the
understand the molecular features of apelin required for rationaldesignofAPJtherapeutics.
a signalling response from APJ, three cyclic analogues of
apelin-12 (C1, C3 and C4) have proved to be novel APJ
Mechanicalstretch
agonists, but are less potent than (Pyr1)apelin-13 in the
inhibitionofcAMPaccumulationandinthephosphoryl- APJ is expressed in cardiomyocytes of human and rat
ationofproteinkinaseB(Akt)andERK1/2(Hamadaetal. hearts (Kleinz et al. 2005), and apelin and APJ have been
2008), while the use of a bivalent ligand approach suggested to have roles in cardiac pathophysiology (see
incorporating a b-turn within the RPRL region of apelin, below). Apelin and APJ mRNA levels are reduced in
which is critical in APJ ligand recognition, has resulted neonatalratventricularmyocytessubjectedtomechanical
http://joe.endocrinology-journals.org (cid:1)2013SocietyforEndocrinology PublishedbyBioscientificaLtd.
DOI:10.1530/JOE-13-0227 PrintedinGreatBritain
Downloaded from Bioscientifica.com at 03/06/2023 01:09:58AM
via free access
Review A-MO’CARROLLandothers APJandregulationof 219:1 R21
homeostasis
stretch, while apelin gene expression has been shown to class II histone deacetylases HDAC4 and HDAC5 and
be reduced in two in vivo models of chronic ventricular the activation of the transcription factor MEF2, in an
pressure overload (Szokodi et al. 2002). Recently, apelin-independentmanner(Kangetal.2013).Therefore,
however,ithasbeenshownthatAPJpromptsmyocardial these studies suggest that apelin signalling may exhibit
hypertrophy in response to mechanical stretch by an ‘functionalselectivity’or‘biasedsignalling’.
apelin-independent, b-arrestin-dependent mechanism. Apelin,signallingthroughAPJ,cantriggernumerous
This stretch-mediated hypertrophy is diminished intracellular signalling cascades whose final targets are
by apelin treatment (Scimia et al. 2012). Furthermore, often transcription factors. It is not yet clear through
the signalling response induced by stretch is pertussis which transcription factors or other cellular effectors
toxin (PTX)-insensitive and G protein-independent, many of the actions of apelin are transduced – what is
unlike the response observed with apelin. Additionally, clear is that apelin signals through a diverse set of
stretch diminishes apelin signalling by decreasing intermediaries. An overview of the signalling pathways
Gproteinactivationandincreasingb-arrestinrecruitment potentiallyrelevanttoAPJsignallingisshowninFig.3.
(Scimia et al. 2012). Thus, it appears that both apelin
andstretchactivateAPJtoaffectcardiachypertrophy.
APJsignallingtoERK1/2
Both apelin-13 and apelin-36 activate the phosphoryl-
Gprotein-couplingofAPJ
ation of ERK1/2 in CHO cells stably expressing mouse
APJ was originally hypothesised to couple to G based APJ (Masri et al. 2006). The activation of ERK1/2 is time-
i/o
on initial experiments showing that forskolin-stimulated and dose-dependent and is mediated via a PTX-sensitive
cAMP production is suppressed by apelin-13 (Tatemoto G-protein, yet independent of the bg-complex, in a
y i
og et al. 1998). This coupling hypothesis was strengthened Ras-independent and a PKC- and MEK-dependent path-
nol by the inability of (Pyr1)apelin-13 and apelin-36 to way(Masrietal.2002).Similarstudiesusingexogenously
ocri generate Ca2C mobilisation or to release arachidonic transfectedhumanAPJinHEK293cellshavedescribedthe
d
n acid metabolites into CHO cells stably expressing activationofERK1/2viaGa byapelin,withnoactivation
E i2
of human APJ (Habata et al. 1999). However, both these of p38 MAPK (Bai et al. 2008). Studies on hippocampal
nal analogues increase intracellular Ca2C levels in NT2N cultures expressing APJ, and on mouse hearts, have also
r
ou neurones (Choe et al. 2000) and in HEK293 cells stably shown that apelin mediates the activation of ERK1/2
J
expressing human APJ (Zhou et al. 2003a). The coupling (O’Donnell et al. 2007, Simpkin et al. 2007). However,
of APJ to G has firmly been established by studies apelindoesnotactivateERK1/2inhumanosteoblastsand
i/o
demonstrating PTX abrogation of apelin-13- and apelin- dose dependently decreases the phosphorylation of
36-induced actions in assays measuring extracellular ERK1/2 in mouse cortical neurones, cells that endogen-
acidification rates (Hosoya et al. 2000) and phosphoryl- ouslyexpressAPJ(Xieetal.2006,Zengetal.2010).
ation of ERK and p70S6 kinase (Masri et al. 2002, 2004). The stimulation of some GPCRs by agonists leads
Mouse APJ couples preferentially to Ga and Ga , but to the activation of metalloprotease isoenzymes,
i1 i2
not to Ga , in inhibition of adenylate cyclase and members of the a disintegrin and metalloproteinase
i3
phosphorylation of ERK1/2 (Masri et al. 2006). Similarly, family of peptidase proteins, to produce ligands that
human APJ activates ERK1/2 through a Ga -dependent activate the epidermal growth factor receptor (EGFR)
i2
pathway (Bai et al. 2008). Moreover, the activation of and subsequently ERK1/2. The GPCR and EGF transacti-
ERK1/2 by apelin is mediated via PKC in HEK293 cells vation pathways are cell specific and depend on various
expressingmouseAPJ,indicativeofcouplingtoeitherG parameters such as the G protein type, receptor type
o
or G . However, the positive inotropic effect of apelin and cellular network (Liebmann 2011). In VSMCs, the
q/11
in rats in vivo is only partially abrogated by PTX and activation of ERK1/2 by Ang II via AT is partially
I
by PKC inhibitors. These effects suggest possible PTX- dependent on the transactivation of EGFR (Eguchi et al.
sensitiveand-insensitivesignallingpathwaystobelinked 1998); however, this pathway is not activated by
to this receptor and indicate that some of the actions of (Pyr1)apelin-13 in HEK293 cells expressing rat or
APJ could be mediated by G and/or G coupling mouse APJ, suggesting that APJ does not activate
i/o q/11
(Szokodi et al. 2002). Recently, in human umbilical vein ERK1/2 via the transactivation of EGFR in these cells
endothelialcells(HUVECs),APJhasbeenshowntoactivate (A-M O’Carroll, S Tilve & GR Pope, 2010, unpublished
Ga , resulting in the cytoplasmic translocation of observations).
13
http://joe.endocrinology-journals.org (cid:1)2013SocietyforEndocrinology PublishedbyBioscientificaLtd.
DOI:10.1530/JOE-13-0227 PrintedinGreatBritain
Downloaded from Bioscientifica.com at 03/06/2023 01:09:58AM
via free access
Review A-MO’CARROLLandothers APJandregulationof 219:1 R22
homeostasis
APJ APJ
Extracellular space
γ γ
AC Cytoplasm
Gαq/11 β PLCβ PIP2 Gαi/o β
DAG ATP cAMP
IP
3
PKC P13K
PKA
Ras Akt
Raf-1
mTOR
MEK1/2
JNK ERK1/2 P70S6K eNOS
y
g
o
ol
n
ri
c
o
d
n
E SP-1 c-Jun c-Fos
f Nucleus
o
al
n
r
u
o
J Figure3
OverviewofAPJsignallingpathways.SchematicdiagramofAPJsignalling ofPKC,butstillviaRasandtheMAPKcascade.G stimulatestheMAPK
i/o
pathways.CouplingtoG stimulatesPLC-bsignalling,includingthe cascadeviaPKC,anditcanalsoactivatephosphoinositide3-kinase(PI3K)
q/11
hydrolysisofphosphatidylinositol4,5-biphosphate(PIP2)toIP3anddiacyl withthesubsequentactivationofAktandmammaliantargetofrapamycin
glycerol(DAG).DAGsubsequentlyactivatesPKC,whichisanactivatorof (mTOR),leadingtotheactivationofbothp70S6Kandendothelialnitric
thesmallG-protein,Ras.Rastheneitheractivatesacascadeleadingtothe oxidesynthase(eNOS).Furthermore,G signallinginhibitsadenylate
i/o
activationofJNK,andthetranscriptionfactorsSP1andc-JunortheMAPK cyclase(AC)activity.Incontrast,GsactivatesAC,increasingcAMPsynthesis
cascadeofRaf-1,MAPK-/ERKkinase(MEK1/2)andERK1/2.ERK1/2havea fromATP,leadingtotheactivationofproteinkinaseA(PKA).Thinblack
varietyofsubstratesincludingnumeroustranscriptionfactors(e.g.c-Jun arrowsindicateactivationpathwaysandtheredbluntedarrowindicates
andc-fos)andotherkinases(e.g.p70S6K).Gq/11alsosignalsindependently inhibition.
suggestinganeuroprotectiverole(O’Donnelletal.2007),
ApelinandthePI3K/Aktpathway
while in mouse cortical neurones, the neuroprotective
Thephosphorylation,andthusactivation,ofAkthasbeen action of apelin is blocked by the PI3K inhibitor
shown to be a downstream effector of apelin signalling; wortmannin, implicating the PI3K/Akt pathway in this
thiswasfirstshowntooccurviaaPTX-sensitiveG-protein process(Zengetal.2010).APJsignallingviaPI3K/Aktisalso
and PKC (Masri et al. 2004). Apelin-13- and apelin-36- involvedinproliferationandanti-apoptoticactionsinrat
mediated Akt phosphorylation in CHO cells expressing andhumanVSMCsrespectively(Cuietal.2010,Liuetal.
mouseAPJtakesplaceviacouplingtoGai1orGai2.Apelin 2010). Studies carried out in vivo have shown the
also activates Akt in HUVECs (Masri et al. 2006) and in involvement of apelin in cardioprotection via Akt-
osteoblasts, where it has proliferative and anti-apoptotic mediated signalling (Simpkin et al. 2007); however, the
effects(Xieetal.2006,2007,Tangetal.2007).Additionally, protectiveeffectofapelininischaemia/reperfusionmaybe
apelinactivatesAktinrathippocampalneuronalcultures, independentofAktsignalling(Kleinz&Baxter2008).
http://joe.endocrinology-journals.org (cid:1)2013SocietyforEndocrinology PublishedbyBioscientificaLtd.
DOI:10.1530/JOE-13-0227 PrintedinGreatBritain
Downloaded from Bioscientifica.com at 03/06/2023 01:09:58AM
via free access