British journal of pharmacology 2016 volume 173 part 1

Enhanced levels of TGFβ are found in patients with heart failure Khan et al., 2014, in various animal models of cardiac remodelling and during the transition from compensated hypertrophy

Trang 2

Reviews Editor

Stephen Alexander Nottingham, UK Annette Gilchrist Senior Online EditorDowners Grove, USA

Senior Editors

Editorial Board

Amrita Ahluwalia London, UK

Michael J Curtis London, UK

James Docherty Dublin, Ireland

Mark Giembycz Calgary, Canada

Daniel Hoyer Melbourne, Australia

Paul Insel La Jolla, USA

Angelo A Izzo Naples, Italy

David MacEwan Liverpool, UK

Clare Stanford London, UK

Susan Wonnacott Bath, UK

Ruth Andrew Edinburgh, UK

Alexis Bailey Guildford, UK

Chris Bailey Bath, UK

Phillip Beart Melbourne, Australia

Tamás Bíró Budapest, Hungary

Tom Blackburn Leigh on Sea, UK

Heather Bradshaw Bloomington, USA

Keith Brain Birmingham, UK

James Alexander Brock Melbourne, Australia

Gillian Burgess Slough, UK

John Challiss Leicester, UK

Diana Chow Houston, USA

Macdonald Christie Sydney, Australia

Sandy Clanachan Edmonton, Canada

John Cryan Cork, Ireland

Anthony Davenport Cambridge, UK

Martin Diener Giessen, Germany

Peter Doris Houston, USA

Pedro D’Orléans-Juste Sherbrooke, Canada

Grant Drummond Clayton, Australia

Claire Edwards Oxford, UK

Michael Emerson London, UK

Liana Fattore Cagliari, Italy

Peter Ferdinandy Szeged, Hungary

Anthony Ford San Mateo, USA

Chris George Cardiff, UK

Jon Gibbons Reading, UK

Gary Gintant Illinois, USA

Michelle Glass Auckland, New Zealand

Jules Hancox Bristol, UK

Deborah L Hay Auckland, New Zealand

Jackie Hunter Weston, UK

Ryuji Inoue Fukuoka, Japan

Yong Ji Nanjing, China

Marcel Jiménez Barcelona, Spain

Eamonn Kelly Bristol, UK

Melanie Kelly Halifax, Canada

Terry Kenakin Durham, USA

Dave Kendall Nottingham, UK

Charles Kennedy Glasgow, UK

Simon Kennedy Glasgow, UK

Chris Langmead Welwyn Garden City, UK

Andy Lawrence Melbourne, Australia

Eliot Lilley Redhill, UK

Jon Lundberg Stockholm, Sweden

Mhairi Macrae Glasgow, UK

Karen McCloskey Belfast, UK

Barbara McDermott Belfast, UK

Alister McNeish Reading, UK

Jo De Mey Odense, Denmark

Olivier Micheau Dijon, France

Paula Moreira Coimbra, Portugal

Maria Moro Madrid, Spain

Fiona Murray San Diego, USA

Anne Negre-Salvayre Toulouse, France

Janet Nicholson Biberach an der Riss, Germany

Eliot Ohlstein Pennsylvania, USA

Saoirse O’Sullivan Nottingham, UK

Hiroshi Ozaki Tokyo, Japan

Reynold Panettieri Jr Philadelphia, USA

Andreas Papapetropoulos Athens, Greece

Clare Parish Melbourne, Australia

Adam Pawson Edinburgh, UK

Roger Phillips Bradford, UK

Michael Pugsley Jersey City, USA

Susan Pyne Strathclyde, UK

Jelena Radulovic Chicago, USA

Chris Sobey Monash, Australia

Michael Spedding Suresnes, France

Beata Sperlagh Budapest, Hungary

Shiva Sruti Pittsburgh, USA

Katarzyna Starowicz Krakow, Poland

Barbara Stefanska Quebec, Canada

Gary Stephens Reading, UK

Csaba Szabo Budapest, Hungary

Kenneth Takeda Strasbourg, France

Paolo Tammaro Oxford, UK

Anna Teti L’Aquila, Italy

Ekaterini Tiligada Athens, Greece

Jean-Pierre Valentin Macclesfield, UK

Paul Vanhoutte Hong Kong, China

Christopher Vaughan Sydney, Australia

Harald Wajant Würzburg, Germany

Julia Walker Durham, USA

Xin Wang Manchester, UK

Nina Weber Amsterdam, the Netherlands

James Whiteford London, UK

Baofeng Yang Heilongjiang, China

The British Journal of Pharmacology is a broad-based journal giving leading

international coverage of all aspects of experimental pharmacology The

Edito-rial Board represents a wide range of expertise and ensures that well-presented

work is published as promptly as possible, consistent with maintaining the

overall quality of the journal

Disclaimer

The Publisher, British Pharmacological Society and Editors cannot be held

responsible for errors or any consequences arising from the use of information

contained in this journal; the views and opinions expressed do not necessarily

reflect those of the Publisher, British Pharmacological Society and Editors

Neither does the publication of advertisements constitute any endorsement by

the Publisher, British Pharmacological Society and Editors of the products

advertised

Copyright and Copying

by any means without the prior permission in writing from the copyright holder Authorization to copy items for internal and personal use is granted by the copyright holder for libraries and other users registered with their localReproduction Rights Organisation (RRO), e.g Copyright Clearance Center (CCC),

222 Rosewood Drive, Danvers, MA 01923, USA (www.copyright.com), providedthe appropriate fee is paid directly to the RRO This consent does not extend to other kinds of copying, such as copying for general distribution for advertising

or promotional purposes, for creating new collective works, or for resale.Special requests should be addressed to: permissions@wiley.com

Trang 3

Molecular switches under

progression from cardiac

hypertrophy to heart failure

J Heger, R Schulz and G Euler

Institute of Physiology, Justus Liebig University, Giessen, Germany

Correspondence

Gerhild Euler, Institute of Physiology,Justus Liebig University, Giessen,Germany

E-mail: Gerhild.Euler@physiologie.med.uni-giessen.de

Commissioning Editor: PeterFerdinandy

we will describe examples of these molecular switches that change compensated hypertrophy to the development of heart failureand will focus on the importance of the signalling cascades of the TGFβ superfamily in this process In this context, potentialtherapeutic targets for pharmacological interventions that could attenuate the progression of heart failure will be discussed

Abbreviations

ALK, activin receptor-like kinase; AMPK, AMP kinase; ANT1, adenine nucleotide translocator 1; AP-1, activator protein 1;Hsp, heat shock protein; IGF2R, insulin-like growth factor receptor II; JDP2, jun dimerization protein 2; LNA, locked nucleicacid; LV, left ventricle; miRNA, microRNA; MPTP, mitochondrial permeability transition pore; NLRP3, nucleotide-bindingdomain and leucine-rich repeat containing PYD-3; PAH, pulmonary hypertension; RIP, receptor interacting protein; RV,right ventricle; siRNA, silencing RNA; SIRT1, sirtuin 1; SMAD, small mothers against decapentaplegic; TAC, transverse aorticconstriction; TAK1, TGFβ activated kinase 1; TGFBR1, TGFβ receptor I; TGFBR2, TGFβ receptor II; TOM, translocase of themitochondrial outer membrane; UPS, ubiquitin proteasome system; VDAC1, voltage-dependent anion channel-1

BJP British Journal of Pharmacology www.brjpharmacol.org

Trang 4

The main causes of heart failure are on the one hand chronic

pressure overload of the left ventricle (LV) resulting in

hypertension and on the other the impairment of

myocar-dial perfusion resulting in acute myocarmyocar-dial infarction or

While pressure overload creates hypertension and results

in cardiac hypertrophy, myocardial infarction primarily

results in a loss of cardiomyocytes that is compensated

for by hypertrophy of the remaining cells, the generation

of fibrosis and ventricular dilatation Thus, the

remodel-ling processes, caused by pressure overload or ischaemia

are different, but both eventually result in heart failure

This has to be considered when using different animal

models, which are induced either by chronic pressure

overload by aortic banding or by direct damage after

myo-cardial infarction

In spite of these differences, in both situations, the

organism reacts by activation of the sympathetic nervous

system and the release of local mediators (cytokines and

natriuretic peptides) to ensure a sufficient blood supply

under these conditions, thereby resulting in intra- and

intercellular remodelling processes Temporarily, this leads

to compensated hypertrophy and preserved heart

func-tion However, in the long run, progressive myocardial

dysfunction develops (Narula et al., 1996), either resulting

in around 50% of the patients having an impaired

diastolic function with preserved ejection fraction, while

the other 50% of patients develop systolic dysfunction

with reduced ejection fraction (Abbate et al., 2015) With

regard to the reduced ejection fraction, apoptotic and

necroptotic loss of cardiomyocytes, contractile

dysfunc-tion of cardiomyocytes and massive fibrosis are signiﬁcant

factors for the transition from compensated hypertrophy

to decompensation and deterioration of systolic heart

function, which will be the main focus of the current

review

Enhanced levels of TGFβ are found in patients with

heart failure (Khan et al., 2014), in various animal models

of cardiac remodelling and during the transition from

compensated hypertrophy to heart failure (Boluyt et al.,

1994; Lijnen et al., 2000; Rosenkranz, 2004) Therefore,

there is a huge drive to clarify the role of TGFβ in heart

failure progression Interestingly, TGFβ modulates nearly

all processes that are engaged in heart failure development,

that is, cardiac hypertrophy, fibrosis, apoptosis,

inﬂamma-tion and differentiainﬂamma-tion of cardiac progenitor cells In spite

of this, broad inhibition of TGFβ signalling does not only

have positive effects on heart failure progression

Adminis-tration of the TGFβ receptor I (TGFBR1) inhibitor (SM16)

after aortic banding prevented cardiac fibrosis and

attenu-ated cardiac dysfunction However, mortality rates increased

due to enhanced left ventricular dilatation and inﬂammation

(Engebretsen et al., 2014) Similar results have been found

when soluble TGFBR2 was applied after myocardial

in-farction In this case, the increase in mortality rates was

probably due to reduced inﬂammatory responses (Ikeuchi

et al., 2004) Therefore, a more target-orientated approach

pathways

TGFβ signals through binding at a heterotetrameric receptorcomplex of type II and type I receptor serine/threonine kinases.Upon TGFβ binding, TGFBR2 phosphorylates and therebyactivates type I receptor serine/threonine kinases that in turnphosphorylates and activates SMAD transcription factors De-pending on the subtype of type I receptor serine/threonine ki-nases, also known as activin receptors or activin receptor-likekinases (ALKs), different receptor-activated SMADs (R-SMADs)become activated In cardiomyocytes and also many other celltypes, TGFβ1 signalling is attributed to TGFBR1, also calledALK5, which then results in SMAD2/3 activation In endothe-lial cells TGFβ has also been shown to signal via ALK1 andSMAD1/5/8 (Goumans et al., 2002) However, this appears to

be no longer exclusive for endothelial cells; In other cells,TGFbeta1 stimulation has been found to activate SMAD1/5and SMAD2/3 as well (Wharton and Derynck, 2009) We alsoidentified both responses in cardiomyocytes After stimulation

of ventricular cardiomyocytes, from adult rats, with 1 ngml 1TGFβ1 for 2 h, enhanced phosphorylation of SMAD2/3,SMAD1/3 and SMAD1/5 was detected in Western blots (n = 5,

P < 0.05 vs unstimulated controls) (Figure 1), thereby ing that TGFβ signalling is even more complicated than origi-nally thought Activated R-SMADs form a complex withSMAD4 that translocates into the nucleus and acts as a tran-scription factor The binding specificity of SMADs to promoterscan be inﬂuenced by their association with other transcriptionfactors like activator protein 1 (AP-1) In addition to this canon-ical SMAD pathway, another prominent signalling molecule ofTGFβ is TGFβ-activated kinase 1 (TAK1) TAK1 activation is alsomediated by TGFBR2 Downstream targets of TAK1 are c-Jun, N-terminal kinase (JNK) and p38 Furthermore, via binding to its

phosphoinositide 3-kinase (PI3K) or small GTPases like Rho(reviewed by Zhang, 2009) This huge variety of TGFβ signallingpathways already implies that the effects of TGFβ in tissues will

be complex In this review, we highlight the signallingmolecules that are induced by TGFβ and modulate adverse car-diac remodelling by interfering with adrenoceptor-mediatedsignalling, mitochondrial proteins, cell death, microRNAs(miRNAs), contractile function or fibrosis (Figure 2)

The TAK1 pathway is pro-hypertrophic and prevents cell death while SMADs promote apoptotic signalling in the heart

TGFβ itself is known to be a pro-hypertrophic, pro-apoptoticand pro-fibrotic factor in the heart TAK1 and not SMADsseems to be the main mediator of TGFβ-induced hypertrophicgrowth effects TAK1 is found to be up-regulated in vivo afteraortic banding, and TAK1 overexpression promotes cardiachypertrophy in transgenic mice (Zhang et al., 2000) (Figure 3).Furthermore, in neonatal cardiomyocytes, angiotensin II-induced hypertrophic growth could be prevented by knock-down of TAK1 with silencing RNA (siRNA), but not withsiRNA against SMAD2/3 (Watkins et al., 2012) This indicatesthat SMAD signalling is not involved in angiotensin II –TGFβ1-induced hypertrophic growth In addition to its pro-hypertrophic effects, TAK1 antagonizes the apoptosis and

BJP J Heger et al.

4 British Journal of Pharmacology (2016)173 3–14

Trang 5

necroptosis induced by TNFα stimulation and prevents

ad-verse cardiac remodelling (Li et al., 2014a) Necroptosis is a

form of cell death, which combines features of necrotic and

apoptotic cell death, it is a death receptor-mediated process

which is executed via receptor activating protein (RIP)

com-plexes (Zhang et al., 2009) During TNFα stimulation, TAK1

associates with RIP1 thereby preventing RIP1 interactionwith other death signalling proteins, that is, with caspase 8

or RIP3 This results in a reduction in apoptosis andnecroptosis (Li et al., 2014a) As TAK1 is not only induced byTGFβ and TNFα but also by other cytokines (Besse et al.,2007), strong TAK1 activation may act as a pro-survival factor

in the heart

In contrast to TAK1, SMAD signalling seems to

cardiomyocytes induced by stimulation ofα-adrenoceptorswas hampered by simultaneous overexpression of SMADs(Heger et al., 2009) Hypertrophic growth of cardiomyocytesinduced by stimulation of α-adrenoceptors is mediated viathe transcription factor AP-1 (Taimor et al., 2004) Under si-multaneous SMAD4 overexpression, AP-1/SMAD complexesare formed, which may detract AP-1 from its hypertrophy-promoting target genes Indeed, a shift from hypertrophy

to the induction of apoptosis is found in

(Heger et al., 2009) Furthermore, cardiac-specific SMAD4knock-out mice displayed cardiac hypertrophy (Wang et al.,2005) This indicates that SMAD4 acts as a molecular switchfor transition from hypertrophy to apoptosis In addition,TGFβ induces apoptosis in adult cardiomyocytes via en-hancement of SMAD and AP-1 activity (Schneiders et al.,2005) Similar to these findings, inhibition of SMAD signal-ling in vivo may preserve the compensating character of hy-pertrophic growth in cardiac remodelling while preventingthe transition to apoptosis (Figure 3)

That AP-1 is a mediator of hypertrophy and apoptosis inadrenoceptor stimulated cardiomyocytes has been shown byuse of transgenic mice overexpressing the AP-1 inhibitor jun

prevented isoprenaline (ISO)-induced hypertrophy as well

as TGFβ-induced apoptosis in cardiomyocytes (Hill et al.,2013) But AP-1 is also required to preserve the contractilefunction of cardiomyocytes because AP-1 inhibition byJDP2 overexpression attenuated contractile responses in-duced byβ-adrenoceptor stimulation (Hill et al., 2013) There-fore, to prevent adverse remodelling, inhibition of SMADsignalling seems to be the better choice than inhibition ofAP-1, because this would negatively inﬂuence the contractilefunction of the heart

Modulation of β-adrenoceptor responses in the presence of TGF β

During heart failure, progressive desensitization of adrenoceptors occurs β-adrenoceptors are members of theGPCR superfamily whose stimulation results in activation

β-of PKA via AC and cAMP, which regulate different lar, sarcolemmal and myofibrillar substrates Thus, cAMPexerts the cellular effects on cardiac contractile function in-duced by activation of β-adrenoceptors However, stimula-tion of β-adrenoceptors also results in agonist-dependentdesensitization of these receptors, a phenomenon foundduring the development of heart failure This process is me-diated by the receptor adapter proteinβ-arrestin that binds

intracellu-to β-adrenoceptors This binding either results in direct

Figure 2

Overview about TGFβ inﬂuence on components of cardiac

remodel-ling in left ventricular systolic heart failure TGFβ has been shown to

promote the transition from cardiac hypertrophy to apoptosis and

to regulate mitochondrial signalling molecules, miRNA expression

and contractile function andﬁbrosis All these processes are involved

in heart failure progression

Figure 1

TGFβ signals via the SMAD2/3 and SMAD1/5 pathway in

cardiomyocytes Ventricular cardiomyocytes of adult rat were

stimu-lated with 1 ngml 1TGFβ1 for 2 h Protein extracts of these cells were

analysed by Western blots with antibodies speciﬁc against

phosphoSMAD2, phosphoSMAD1/3 or phosphoSMAD1/5

Phosphor-ylation, which is indicative of SMAD activation, was detected for all

these SMADs *P< 0.05 versus unstimulated controls, n = 5

indepen-dent culture preparations

TGFβ-guided switches to heart failure

BJP

Trang 6

inhibition ofβ-adrenoceptors, known as functional

desensi-tization, or in internalization of β-adrenoceptors that

re-duces their density (reviewed by Lymperopoulos and

Negussie, 2013) Studies inβ-arrestin1 knock-out mice

dem-onstrated a major role for β-arrestin1 in cardiac

stimulation were enhanced in these knock-out animals

β-arrestin1 prevented adverse cardiac remodelling after

myo-cardial infarction by inhibiting apoptosis and preserving

cardiac function (Bathgate-Siryk et al., 2014) Interestingly,

an increase in myocardial β-adrenoceptor density and a

β-adrenoceptor-kinase-1 were demonstrated in transgenic

TGFβ-overexpressing mice (Rosenkranz et al., 2002) And

in isolated cardiomyocytes of adult rat, TGFβ enhanced

the hypertrophic response to β-adrenoceptor stimulation

(Schlüter et al., 1995) These findings indicate that TGFβ

can preventβ-adrenoceptor desensitization in cardiomyocytes

and thereby promote pro-hypertrophic signalling Whether

this response is mediated by the down-regulation of

β-arrestin1 by TGFβ has not yet been clarified But TGFβ

may be a plausible target in order to preventβ-adrenoceptor

desensitization

So far, a connection between β-arrrestin expression

and TGFβ signalling has been shown in cardiac fibroblasts

β-Arrestins were found to be up-regulated in cardiac

fibroblasts during heart failure Overexpression ofβ-arrestin

in cardiac fibroblasts results in the uncoupling of adrenoceptors and activation of SMAD2/3, thereby promoting

β-a pro-fibrotic phenotype This mβ-ay cβ-ause enhβ-anced stiffness

of the ventricular wall and contribute to the development ofheart failure

Although TGFβ stimulation prevents the uncoupling ofβ-adrenoceptors and enhances the pro-hypertrophic signal-ling, the inotropic β-adrenoceptor-mediated response wasdiminished in TGFβ-overexpressing mice This is due to anup-regulation of mitochondrial uncoupling proteins duringβ-adrenoceptor stimulation, which results in a decreased mi-tochondrial energy production Thus, TGFβ-overexpressingmice resemble a phenotype occurring at the transition toheart failure, namely, displaying cardiomyocytes hypertro-phy and promoting apoptosis as well as mitochondrialand contractile dysfunction (Schneiders et al., 2005;Huntgeburth et al., 2011)

That these interacting pathways of ADRB-TGFβ signallingare even more complex was indicated by the findings thatGPCRs not only activate TK receptors but also alsotransactivate the serine/threonine kinase TGFBR1 in differentcell types (Burch et al., 2012) The proposed mechanism forthis transactivation is activation of integrin by GPCRs Subse-quently, integrin binding to the large latent TGFβ complexcauses a conformational change and allows TGFβ to bindand activate TGFBR2/TGFBR1, thereby resulting in SMAD

Figure 3

Influence of TGFβ-SMAD and TGFβ-TAK1 signalling on adrenoceptor-mediated pathways in LV heart failure progression Adrenoceptors (AR) ulate the expression of genes promoting hypertrophic growth via the transcription factor AP-1 Under simultaneous presence of SMAD4, the pro-hypertrophic response to adrenoceptor stimulation is shifted to a pro-apoptotic gene transcription via AP-1/SMAD complexes Also, under TGFβstimulation of cardiomyocytes, AP-1 and SMADs mediate apoptosis In addition to these effects on cardiomyocytes, activation of the TGFβ/SMADpathway or induction of SMADs viaβ-arrestins induces the transcription of fibrotic genes In contrast to the SMAD pathway, TAK1 activation stim-ulates hypertrophic growth while inhibiting cardiac necroptosis and apoptosis by interacting with RIP1 Strongβ-adrenoceptor (ADRB) activationresults inβ-adrenoceptor desensitization via β-adrenoceptor /β-arrestin complexes This process can be inhibited by TGFβ Depicted in red areswitch molecules that can modulate the response of the cell to receptor stimulation and thereby influence the outcome of this stimulation onthe remodelling process

stim-BJP J Heger et al.

Trang 7

activation (Munger et al., 1999) Whether this

cardiomyocytes has yet to be analysed

The ubiquitin system in the context of

β-adrenoceptor and TGFβ stimulation

Another focus for identification of the triggers contributing

to heart failure development or progression relies on the

anal-ysis of the proteasome, as degradation of proteins is changed

in cardiac hypertrophy The primary cellular response to

β-adrenoceptor stimulation in the heart is an increased pool

of 20S subunits with catalytic activity, while chronic

β-adrenoceptor stimulation enhanced the 26S proteasome but

decreased 20S proteasomal activity, accompanied by a

de-crease in ubiquitinated proteins (Drews et al., 2010) Elevated

26S proteasome activities were also observed in a pressure

overload model stimulating ventricular hypertrophy (Depre

et al., 2006) The switch in proteasome subpopulations,

which is facilitated by differentβ-subunits of the proteasome,

is decisive for the development of hypertrophy and depends

again on the strength ofβ-adrenoceptor activation Proteins

involved in cardiac hypertrophy are targeted by

muscle-specific ubiquitin ligase atrogin-1 for degradation (Zaglia

et al., 2014) Atrogin-1 KO hearts revealed increased apoptosis

and hypertrophy The effects were mediated by the

endosomal sorting complex required for transport III

(ESCRT-III) family protein charged multivesicular body

pro-tein 2B (CHMP2B) Thus, Zaglia et al (2014) demonstrated

the interplay between the ubiquitin proteasome system

(UPS) and autophagy and the importance of controlled

degra-dation of proteins for the control of cardiac hypertrophy and

apoptosis

UPS regulates important signalling pathways in the heart,

including MAPK, JNK and calcineurin (Portbury et al., 2012)

Huang et al (2014) suggested a proteasome-dependent

mech-anism for angiotensin II–induced apoptosis in hearts that is

accompanied by activation of insulin-like growth factor

re-ceptor II (IGF2R) signalling Heat shock transcription factor

1 (HSF1) acts as a repressor of IGF2R gene expression only if

deacetylated by sirtuin 1 However, angiotensin II and

subse-quently JNK activation mediates sirtuin 1 degradation via the

proteasome This results in an increase in the acetylation of

HSF1 that is then not able to bind to the IGF2R promoter

So, sirtuin is a negative regulator of IGF2R, thereby protecting

cardiomyocytes from apoptosis In this context, it is

remark-able that the IGF2R is required for the activation of latent

TGFβ (Dennis and Rifkin, 1991) In human umbilical-vein

endothelial cells, the association of IGF2R and the urokinase

receptor–converts plasminogen (uPAR) to active plasmin –

is essential for the activation of latent TGFβ, the release of

TGFβ and induction of apoptosis (Leksa et al., 2005) Whether

this also holds true for cardiomyocytes remains to be

evalu-ated, but we have already shown that angiotensin II induces

the release of TGFβ and SMAD-dependent apoptosis in

cardiomyocytes (Schröder et al., 2006) Not only is the

intra-cellular activity of TGFβ controlled by UPS but also the

stability and levels of TGFβ receptor complexes are mined by ubiquitination (Xu et al., 2012)

deter-In ﬂuence of TGFβ on mitochondria, energy metabolism and heart failure

Mitochondria are the power houses of the cell, generating ATPvia oxidative phosphorylation On average, 30% of thecardiomyocytes volume is filled with mitochondria (Barth

et al., 1992) One side product of the major respiratory zyme complexes is the generation of reactive oxygen species(ROS) that modifies the redox potential of the cell and is es-sential for numerous signalling pathways (Chen and Zweier,2014) Mitochondrial dysfunction occurs under pathophysi-ological conditions and involves malfunction of complexes

en-of oxidative phosphorylation, and an increase in ROS duction that leads to cell death contributing to the develop-ment of heart failure The enzymes of the respiratory chainseem to be the main site of ROS formation, but many otherenzymes contribute to ROS production in failing hearts, in-cluding monoamine oxidases and the cytosolic adaptor pro-tein p66Shc(Di Lisa et al., 2009) Cellular stress signals lead

pro-to translocation of p66Shcinto the mitochondrial brane space, where it oxidizes cytochrome c and generatesROS (Heusch, 2015) Factors that inﬂuence ROS production,therefore, critically determine the cell’s fate

intermem-A newly identified signalling molecule in the control ofmitochondrial ROS production that is under the control ofTGFβ signalling is nucleotide-binding domain and leucine-rich repeat containing PYD-3 (NLRP3), a pattern recognitionreceptor that is involved in the pathogenesis of chronic dis-eases and inﬂammation NLRP3 is expressed in the heart, lo-calized in mitochondria, and interacts with components ofthe redox system (Figure 4) Upon TGFβ stimulation of car-diac fibroblasts, NLRP3 increases mitochondrial ROS produc-tion, which supports SMAD2 phosphorylation and results inthe differentiation of cardiac fibroblasts into myofibroblasts,

an important process in adverse cardiac remodelling (Bracey

et al., 2014) The involvement of NLRP3 in cardiac fibrosishas been confirmed in an in vivo model of hypertension: an-giotensin II infusion for 28 days resulted in TGFβ-mediatedfibrosis in wild-type mice, but NLRP3-deficient mice wereprotected against this angiotensin II-induced fibrosis NLRP3,therefore, is a newly identified mitochondrial signalling fac-tor in TGFβ-induced cardiac remodelling that may promotethe transition to heart failure as it facilitates ROS-mediatedfibrosis

Increased ROS production induces the opening of the tochondrial permeability transition pore (MPTP) (Figure 4)that changes the permeability of the inner mitochondrialmembrane, leading to mitophagy, fusion/fission events andbiogenesis (Brenner and Moulin, 2012) Opening of theMPTP facilitates the release of pro-apoptotic factors fromthe mitochondria that stimulates the activation of caspasesand finally leads to cell death (Kinnally et al., 2011) The addi-tion of noradrenaline induced a concentration-dependentdecrease in mitochondrial membrane potential that was asso-ciated with a switch from compensated hypertrophy to apo-ptosis, thereby indicating that MPTP opening is involved in

mi-TGFβ-guided switches to heart failure

BJP

Trang 8

adverse remodelling (Jain et al., 2015) Inhibiting MPTP

opening by overexpression of adenine nucleotide translocase

1 (ANT1) prevented TGFβ1-induced apoptosis in ventricular

cardiomyocytes (Heger et al., 2012) and improved cardiac

function in rats with an activated renin–angiotensin system

(Walther et al., 2007) These findings highlight the

contribu-tion MPTP opening has to the adverse cardiac remodelling

in-duced by TGFβ stimulation and indicate that ANT1 is a

critical component at the inner mitochondria membrane

for regulating MPTP opening (Figure 4)

Besides modulation of MPTP opening, the B-cell lymphoma

2 (Bcl-2) family is a well-known gate keeper in

mitochondria-mediated apoptosis TGFβ can stimulate or inhibit the

expression of pro-apoptotic and anti-apoptotic Bcl-2 family

members (Grünenfelder et al., 2002) After renal artery ligation,

a model for angiotensinII/TGFβ-mediated cardiac hypertrophy,

up-regulation of the pro-apoptotic family member Bax and the

voltage-dependent anion channel-1 (VDAC1) occurred (Figure 4)

Together, they lead to permeabilization of the outer

mito-chondrial membrane, release of cytochrome c from the

intermembrane space into the cytosol, formation of the

apoptosome, activation of caspases and finally the induction

of apoptosis (Mitra et al., 2013) The small heat shock protein,

crystalline B, is able to block the pro-apoptotic action of

VDAC1, and thereby acts as a molecular key that guides

VDAC1 to be anti-apoptotic Therefore, crystalline B may

become an interesting therapeutic target for the prevention

of the transition from compensated hypertrophy to heart

failure Another heat shock protein (Hsp) with anti-apoptotic

action on the mitochondrial level is Hsp22 (Qiu et al., 2011).Overexpression of Hsp22 results in physiological hypertro-phy via up-regulation of NFκB, and binding of Hsp22 to signaltransducer and activator of transcription 3 (STAT3), which is amarker of cardiac stress responses Down-regulation of Hsp22leads to an increased remodelling of the heart and death inknock-out mice after transverse aortic constriction (TAC) bymodulating the nuclear and mitochondrial function of STAT3and STAT3-dependent genes

A further mitochondria-associated candidate, mediating aswitch to pathophysiological hypertrophy is TOM70, atranslocase of the mitochondrial outer membrane (TOM)complex that mediates the import of mitochondrialpreproteins (Figure 4) Li et al (2014b) nicely showed adown-regulation of TOM70 in pathophysiological hypertro-phy in humans as well as animal models This results in thereduced import of optical atrophy-1 (OPA1)– a protein impor-tant for mitochondrial fusion–, a reduction in complex I ac-tivity and finally in ROS production As a consequence,changes in the outer mitochondrial membrane and/or innermitochondrial membrane occurred, followed by apoptoticevents as discussed above In addition, increased TOM70levels made cardiomyocytes completely resistant to the ef-fects of various pro-hypertrophic stimuli

These findings explain the significance of the modulation ofmitochondrial pores by, for example, VDAC1, crystalline B orTOM70 for cardiac hypertrophy to progress to heart failure

At the onset of the development of heart failure, a bolic shift from fatty acid to glucose metabolism has been

meta-Figure 4

The central role of mitochondria in LV heart failure can be modulated by TGFβ Hypertrophy, ﬁbrosis and apoptosis can be controlled by chondria via generation of ROS NLRP3 is a newly identiﬁed molecule that enhances mitochondrial ROS production and that is controlled by TGFβ

mito-or angiotensin II (AngII) miR181c enhances ROS production via modulation of complex IV of the respiratmito-ory chain TOM70, acting as a repressmito-or

of mitochondrial ROS production, is found to be reduced in heart failure This reduction then provokes enhancement of ROS Enhancement ofmitochondrial uncoupling protein during stimulation ofβ-adrenoceptors by noradrenaline (NA) and TGFβ-receptor activation results in reducedenergy production and impaired contractile function Opening of the MPTP plays a central role in the induction of apoptosis Opening of this porecan be modulated by the accessory proteins VDAC and ANT1 Their expression is regulated by crystalline B, TGFβ, AngII, ROS and β-adrenoceptors(ADRB) Central molecules that modulate mitochondrial processes in heart failure are depicted in red

Trang 9

described This shift is due to the down-regulation of

en-zymes for fatty acid oxidation, whereas glycolytic enen-zymes

are up-regulated (Sack et al., 1996) This enables the heart to

increase its metabolic substrate efficiency in relation to O2

consumption However, the metabolic shift seems to be

re-lated to adverse cardiac remodelling A key regulator of

en-ergy homeostasis induced by stimulation of glycolysis and

glycogen accumulation is AMP-activated kinase (AMPK),

which is activated during cardiac remodelling (Dolinsky and

Dyck, 2006; Kolwicz and Tian, 2011) Just recently,

myostatin, a member of the TGFβ superfamily, was identiﬁed

as a repressor of AMPK (Biesemann et al., 2014) Myostatin

reduces muscle growth (skeletal or cardiac), and thus

pro-tects the heart against hypertrophy and failure, and this

function of myostatin is, in part, mediated via repression

of AMPK and the prevention of a metabolic switch towards

glycolysis

In addition to the metabolic shift, a down-regulation of

transporters for glucose (GLUT1/4) and fatty acid (CD36)

up-take into cardiomyocytes as well as a reduction of

trans-porters for pyruvate (PDH) or the carnitine shuttle (CPT1/2)

in mitochondria contribute to heart failure development, as

deletions of these transporters provoked cardiac remodelling

and or dysfunction (Bersin et al., 1994; Liao et al., 2002;

Domenighetti et al., 2010; Lai et al., 2014)

Thus, enhancing the uptake mechanisms for glucose and

fatty acids into cardiomyocytes, as well as for metabolized

substrates into mitochondria can attenuate heart failure

pro-gression Furthermore, prevention of the metabolic switch,

probably via AMPK, is a promising target for therapeutic

approaches against heart failure development

MicroRNAs in heart failure

An increased ability to regulate the processes involved in cardiacremodelling is attributed to miRNAs miRNAs are small non-coding RNAs that target the 3´-untranslated region or 5’-untranslated region of mRNA transcripts This results in the de-stabilization or translational repression of mRNAs (Bartel, 2004).Furthermore, miRNAs can regulate gene transcription by induc-ing histone modifications or DNA methylations (Hawkins andMorris, 2008) In fact, one single miRNA can affect many targetgenes generating a broad network of miRNA-controlled gene ex-pression that has a huge effect on different biological processesincluding cardiac remodelling Analysing the role of miRNAs

in heart failure development has already identified some ising new therapeutic targets

prom-The RNase III endonuclease Dicer is essential for the ing of pre-miRNA into its mature form In the adult myocar-dium, a loss of Dicer-induced biventricular enlargement isaccompanied by hypertrophic growth of cardiomyocytes, ven-tricular fibrosis and functional defects (da Costa Martins et al.,2008) A similar study by Chen et al (2008) revealed signs of di-lated cardiomyopathy and heart failure after cardiac-specific de-letion of Dicer Furthermore, they found that the level of Dicerprotein was significantly reduced in in human patients with di-lated cardiomyopathy and failing hearts These findings indi-cate that miRNAs have a major function in the control ofheart failure development and progression

process-Either an up-regulation or down-regulation of miRNAsunder pressure overload can mediate cardiac remodelling,for example, when miR25 is increased the activity of the sar-coplasmic reticulum (SR) Ca2+ATPase (SERCA2A) is reduced

Figure 5

Inﬂuence of miRNAs on LV adverse remodelling can be modulated by β-adrenoceptors (ADRBs) or TGFβ miRNAs that have been demonstrated toreverse or promote adverse cardiac remodelling are depicted Up-regulation of miR15 or miR22 prevents the induction ofﬁbrosis or apoptosisunder pressure overload (TAC) orβ-adrenoceptor stimulation (ISO) while preserving effects on moderate, compensatory hypertrophy, as whenthese miRs are inhibited adverse remodelling develops In contrast, up-regulation of miR25 or miR21 under TAC enhances adverse cardiac remod-elling, and down-regulation of miR133a under TAC preserves cardiac function, whereas the overexpression of miR133a results in the development

of adverse remodelling Black arrows indicate the responses of the cell to TAC or ISO Switch molecules in the process of adverse remodelling aredepicted in red Green arrows and symbols indicate interference of miR expression by anti-miRs or transgenic overexpression

BJP

Trang 10

and Ca-handling is impaired Anti-miR25 reverses hypertrophy,

fibrosis and heart failure progression after TAC (Wahlquist et al.,

2014) (Figure 5) miR133a is down-regulated under pressure

over-load and when this down-regulation is prevented in transgenic

mice TAC-induced fibrosis and apoptosis are attenuated, whereas

hypertrophy is not affected (Matkovich et al., 2010); hence, these

diverse processes of cardiac remodelling are differentiated

(Figure 5) This indicates that by altering the levels of miR133a,

it may be possible to stop the adverse remodelling processes

while maintaining the compensatory effects of hypertrophy

The development of moderate hypertrophy and physiological

cardiac remodelling induced by an infusion of isoprenaline is

converted to adverse remodelling in miR22 knock-out mice with

a marked enhancement of fibrosis and apoptosis that finally

leads to dilated cardiomyopathy (Huang et al., 2013) The effects

of miR22 seem to be mediated by the inhibition of histone

deacetylases, which indicates that miR22 has a role in the

epige-netic regulation of gene expression during cardiac hypertrophy

These findings indicate that duringβ-adrenoceptor stimulation

miR22 prevents the transition from compensated hypertrophy

to heart failure progression (Figure 5)

A new aspect of miRNA-controlled signalling relates to

the translocation of nuclear-encoded miR181c to

mitochon-dria (Das et al., 2014) This results in the remodelling of

mitochondrial complex IV and an enhanced production of

ROS (Figure 4) The overexpression of miR181c induces

ventricular dysfunction indicating that miRNAs can

modu-late exercise capacity directly at the mitochondrial level

Interestingly, this miR181c-induced dysfunction was not

ac-companied by cardiac hypertrophy Unfortunately, these

ob-servations were only carried out in vitro with a simulated

overexpression of miR181c and its role in cardiovascular

dis-ease in vivo has still to be proven

With regard to TGFβ signalling, the miR15 family needs to

be mentioned This family comprises six highly conserved

fam-ily members that are up-regulated in human heart failure and

can inhibit numerous components of the TGFβ signalling

path-way (Tijsen et al., 2014) Inhibition of one family member by

anti-miR15b in mice also resulted in the down-regulation of

other family members and predominantly enhanced

SMAD-signalling and cardiac fibrosis, especially after TAC (Figure 5)

Therefore, an up-regulation of miR15 members in heart failure

acts as negative feedback mechanism on TGFβ signalling in

or-der to restrict adverse remodelling However, due to the

ubiqui-tous expression of miR15 and because miR15 is also involved

in inducing apoptosis after acute myocardial infarction

(Hullinger et al., 2012), the therapeutic application of miR15

against heart failure progression should be treated with caution

as a substantial amount of additional work needs to be done to

exclude negative side effects In addition to miRs modifiying

TGFβ signalling pathways, TGFβ itself is also known as a

regula-tor of miRs In the context of heart failure progression, miR21

should be highlighted; miR21 is selectively up-regulated in

ﬁbroblasts upon TGFβ stimulation, and in the failing

myocar-dium (Thum et al., 2008; Topkara and Mann, 2011) It has been

shown to induce cardiac fibrosis by enhancing the proliferation

of fibroblasts and to stimulate endothelial mesenchymal

transi-tion during TGFβ stimulatransi-tion (Kumarswamy et al., 2012) and

also to act as a mediator of adverse cardiac remodelling after

TAC (Thum et al., 2011) (Figure 5) Just recently, cardiac

ﬁbroblasts were shown to secrete miRNA-enriched exosomes

(including miR21) Fibroblast-derived miR21 acts as a potentparacrine RNA molecule that induces cardiomyocyte hypertro-phy (Bang et al., 2014) miR155, secreted by macrophages, alsohas paracrine effects on the heart, because miR155 knockout

in macrophages prevented angiotensin II-induced or induced cardiac hypertrophy and dysfunction althoughfibrosis was still present

TAC-These findings showing that miRNAs can exert paracrineeffects implies that systemic pharmacological interference

of miRNA signalling by the use of anti-miRNAs might be ful for preventing heart failure progression Such a systemictherapeutic intervention would be easy to apply clinically.However, in cardiac pathophysiology, the systemic applica-tion of anti-miRNAs is still restricted to basic science studies

use-A promising approach in this direction has been shown byMontgomery et al (2011) using locked nucleic acid (LNA)-modified miR208a-antisense oligonucleotides Systemic de-livery of these oligos silenced miR208a-expression in theheart, thereby preventing hypertension-induced heart failure

in Dahl hypertensive rats by reducing cardiomyocyte trophy, cardiac fibrosis and improving cardiac function Be-cause cardiac miR208 overexpression in transgenic miceinduced cardiac hypertrophy, the reduction of miR208a inthe heart by LNA-antisense oligos is at least in part responsi-ble for cardioprotection in hypertensive Dahl rats However,reductions in the levels of circulating miR499 and miR208bwere also found during the treatment with LNA-modified208a antisense oligos Therefore, a combination of local andsystemic effects may contribute to the protective effects.All these studies demonstrate the enormity of the miRNAnetwork and its inﬂuence on heart failure progression Finetuning of specific miRNAs is essential for physiological hypertro-phy or decompensation, and thus has great therapeutic potentialfor the treatment of heart failure patients

hyper-Right heart failure

For many years, analysis of left ventricular systolic tion was at the centre of heart failure research, which is alsothe focus of our review However, in recent years, some prom-ising advances in the analysis of right heart failure have beenmade that should be discussed here

dysfunc-One major cause for right ventricular dysfunction is nary hypertension (PAH) Due to an increase in pulmonary vas-cular resistance, afterload on the right ventricle (RV) increases,leading to right heart failure which determines the prognosis

pulmo-of patients with PAH Therefore, it is pulmo-of utmost importance todefine new therapeutic strategies to reduce RV remodelling inorder to improve patient prognosis (Ryan et al., 2015).The RV, similar to the LV, compensates an increased workload due to hypertension by hypertrophic growth processes.However, compensatory remodelling is limited and over time,the RV decompensates, finally leading to heart failure Thereseem to be chamber-specific responses, and thus, a simple ex-tension of LV findings to the RV is not possible An interestingnew finding in this respect comes from Schreckenberg et al.(2015) who analysed the effect of chronic NO deprivation onthe remodelling processes in the LV and RV Treatment of ratswith the NOS inhibitorL-NAME resulted in moderate ventricu-lar hypertrophy without signs of dysfunction However, the

Trang 11

RV responded with dilatation and dysfunction A massive

increase in oxygen radicals due to a down-regulation of

the anti-oxidative protein, SOD, was only found in the RV,

and this was the cause of RV remodelling, whereas in the

LV, anti-oxidative enzymes were even increased under the

L-NAME treatment Thus, oxidative stress is a much greater

risk factor in the RV compared with the LV The reduction

in SOD and the development of RV dysfunction could be

attenuated by captopril, an angiotensin-converting enzyme

inhibitor with high anti-oxidative capacity Interestingly,

enhanced ROS production in the RV is responsible for

HIF1α inhibition and suppression of angiogenesis, thereby

resulting in capillary rarefication of the RV and chronic

ischaemia (Bogaard et al., 2009)

A further positive effect of therapeutics against the renin–

angiotensin system was demonstrated by Friedberg et al (2013)

Pulmonary artery banding in rabbits induced RV hypertrophy

and biventricular up-regulation of the TGFβ pathways, including

activation of SMAD3, fibrosis and apoptosis These responses

were blocked by losartan, an inhibitor of the AT1receptor In

ad-dition, TGFβ had negative effects on vascular remodelling in PAH

(Zaiman et al., 2008) Therefore, blocking this pathway has

posi-tive effects on the vasculature as well as on both ventricles and

seems to be a promising therapeutic target

Further interesting targets in RVare (i) the up-regulation of

pyruvate dehydrogenase kinase (PDK), as it is responsible for

the metabolic shift to inefficient cytosolic glyocolysis and RV

dysfunction (Piao et al., 2010) and (ii) the inhibition of

G-protein receptor kinase 2 (GRK2) to prevent the

down-regulation of β-adrenoceptors and preserve RV function in

PAH (Piao et al., 2012)

Conclusion and outlook

Cardiac remodelling is a multifactorial-induced process The

present pharmacological interventions are able to postpone

the onset of heart failure but are unable to prevent the

ongo-ing process of cardiac remodellongo-ing in hypertrophied hearts

Newly identified signalling molecules might have the

poten-tial to serve as therapeutic targets in the treatment of heart

failure Therefore, different and common pathways of left

and RVs should be considered Among the decisive signalling

molecules for LV heart failure progression are modulators of

mitochondrial pores, such as NLRP3, ANT1, VDAC1 or

TOM70 and various miRNAs; for example, miR22 speciﬁcally

impairs the progression from compensated hypertrophy to

left ventricular heart failure and miR21 has significant

potential to induce fibrosis Furthermore, prevention of the

down-regulation of mitochondrial transporters might improve

the metabolic situation in the remodelled myocardium

Other switch molecules control the adrenoceptor-mediated

signalling pathway, like SMAD4, TAK1 orβ-arrestin, thereby

modulating hypertrophic, apoptotic necroptotic and

con-tractile responses of the cell In the RV, a higher

susceptibil-ity to oxygen radical production has been found, thereby

indicating the importance of using therapeutic ROS

scaven-gers Furthermore, PDK and GRK2 have been identified as

new targets in RV

Many of these targets are regulated by TGFβ, and

ad-ministration of an ALK5 inhibitor after aortic banding

prevented cardiac fibrosis and attenuated cardiac tion However, mortality rates of animals increased due toenhanced left ventricular dilatation and inﬂammation(Engebretsen et al., 2014) Therefore, a more target-orientated approach needs to be used to inhibit the detri-mental TGFβ pathways but preserve the protective ones

dysfunc-As exemplified in this review, newly identified switchmolecules may offer novel options in the therapeutic ap-proach against heart failure development and progression

Con ﬂict of interest

None

ReferencesAbbate A, Arena R, Abouzaki N, Van Tassell BW, Canada J, Shah K et al.(2015) Heart failure with preserved ejection fraction: refocusing ondiastole Int J Cardiol 179: 430–440 doi:10.1016/j.ijcard.2014.11.106.Alexander SP, Benson HE, Faccenda E, Pawson AJ, Sharman JL,Spedding M et al (2013a) The Concise Guide to

PHARMACOLOGY 2013/14: G protein-coupled receptors Br JPharmacol 170: 1459–1581 doi:10.1111/bph.12445

Alexander SPH, Benson HE, Faccenda E, Pawson AJ, Sharman JL,Spedding M et al (2013b) The concise guide to

PHARMACOLOGY 2013/14: catalytic receptors Br J Pharmacol170: 1676–1075

Alexander SPH, Benson HE, Faccenda E, Pawson AJ, Sharman JL,Spedding M et al (2013c) The concise guide to PHARMACOLOGY2013/14: transporters Br J Pharmacol 170: 1706–1796

Alexander SPH, Benson HE, Faccenda E, Pawson AJ, Sharman JL,Spedding M et al (2013d) The concise guide to PHARMACOLOGY2013/14: enzymes Br J Pharmacol 170: 1797–1867

Bang C, Batkai S, Dangwal S, Gupta SK, Foinquinos A, Holzmann A

et al (2014) Cardiac ﬁbroblast-derived microRNA passenger enriched exosomes mediate cardiomyocyte hypertrophy J ClinInvest 124: 2136–2146 doi:10.1172/JCI70577

strand-Bartel DP (2004) MicroRNAs: genomics, biogenesis, mechanism, andfunction Cell 116: 281–297 doi:10.1016/S0092-8674(04)00045-5.Barth E, Stämmler G, Speiser B, Schaper J (1992) Ultrastructuralquantitation of mitochondria and myoﬁlaments in cardiac musclefrom 10 different animal species including man J Mol Cell Cardiol24: 669–681 doi:10.1016/0022-2828(92)93381-S

Bathgate-Siryk A, Dabul S, Pandya K, Walklett K, Rengo G, Cannavo A

et al (2014) Negative impact of β-arrestin-1 on post-myocardialinfarction heart failure via cardiac and adrenal-dependentneurohormonal mechanisms Hypertension 63: 404–412

et al (2007) TAK1-dependent signaling requires functionalinteractionwith TAB1/TAB3 J Biol Chem 282: 3918–3928

doi:10.1074/jbc.M608867200

BJP

Trang 12

Biesemann N, Mendler L, Wietelmann A, Hermann S, Schäfers M,

Krüger M et al (2014) Myostatin regulates energy homeostasis in the

heart and prevents heart failure Circ Res 115: 296–310 doi:10.1161/

CIRCRESAHA.115.304185

Boluyt MO, O’Neill L, Meredith AL, Bing OH, Brooks WW, Conrad

CH et al (1994) Alterations in cardiac gene expression during the

transition from stable hypertrophy to heart failure Circ Res 75:

23–32 doi:10.1161/01.RES.75.1.23

Bogaard HJ, Natarajan R, Henderson SC, Long CS, Kraskauskas D,

Smithson L et al (2009) Chronic pulmonary artery pressure

elevation is insufﬁcient to explain right heart failure Circulation

120: 1951–1960 doi:10.1161/CIRCULATIONAHA.109.883843

Bracey NA, Gershkovich B, Chun J, Vilaysane A, Meijndert HC,

Wright JR Jr et al (2014) Mitochondrial NLRP3 protein induces

reactive oxygen species to promote Smad protein signaling and

ﬁbrosis independent from the inﬂammasome J Biol Chem 289:

19571–19584 doi:10.1074/jbc.M114.550624

Brenner C, Moulin M (2012) Physiological roles of the permeability

transition pore Circ Res 111: 1237–1247 doi:10.1161/

CIRCRESAHA 112.265942

Burch ML, Osman N, Getachew R, Al-Aryahi S, Poronnik P, Zheng WG

et al (2012) Protein coupled receptor transactivation: extending the

paradigm to include serine/threonine kinase receptors Int J Biochem

Cell Biol 44: 722–727 doi:10.1016/j.biocel.2012.01.018

Chen JF, Murchison EP, Tang R, Callis TE, Tatsuguchi M, Deng Z et al

(2008) Targeted deletion of Dicer in the heart leads to dilated

cardiomyopathy and heart failure Proc Natl Acad Sci U S A 105:

2111–2116 doi:10.1073/pnas.0710228105

Chen Y-R, Zweier JL (2014) Cardiac mitochondria and reactive

oxygen species generation Circ Res 114: 524–537 doi:10.1161/

CIRCRESAHA 114.300559

Conner DA, Mathier MA, Mortensen RM, Christe M, Vatner SF,

Seidman CE et al (1997) beta-Arrestin1 knockout mice appear

normal but demonstrate altered cardiac responses to beta-adrenergic

stimulation Circ Res 81: 1021–1026 doi:10.1161/01.RES81.6.1021

da Costa Martins PA, Bourajjaj M, Gladka M, Kortland M, van Oort RJ,

Pinto YM et al (2008) Conditional dicer gene deletion in the postnatal

myocardium provokes spontaneous cardiac remodeling Circulation

Das S, Bedja D, Campbell N, Dunkerly B, Chenna V, Maitra A et al

(2014) miR-181c regulates the mitochondrial genome, bioenergetics,

and propensity for heart failure in vivo PLoSOne 9: e96820

doi:10.1371/journal.pone.0096820

Dennis PA, Rifkin DB (1991) Cellular activation of latent

transforming growth factor beta requires binding to the

cation-independent mannose 6-phosphate/insulin-like growth factor type II

receptor Proc Natl Acad Sci U S A 88: 580–584, PMID: 1846448

Depre C, Wang Q, Yan L, Hedhli N, Peter P, Chen L et al (2006)

Activation of the cardiac proteasome during pressure overload

promotes ventricular hypertrophy Circulation 14: 1821–1828

doi:10.1161/CIRCULATIONAHA.106.637827

Di Lisa F, Kaludercic N, Carpi A, Menabò R, Giorgio M (2009)

Mitochondria and vascular pathology Pharmacol Rep 61: 123–130,

PMID: 19307700

Dolinsky VW, Dyck JR (2006) Role of AMP-activated protein kinase in

healthy and diseased hearts Am J Physiol Heart Circ Physiol 291:

H2557–H2569 doi:10.1152/ajpheart.00329.2006

Domenighetti AA, Danes VR, Curl CL, Favaloro JM, Proietto J,

Delbridge LM (2010) Targeted GLUT-4 deﬁciency in the heart induces

cardiomyocyte hypertrophy and impaired contractility linked withCa(2+) and protonﬂux dysregulation J Mol Cell Cardiol 48: 663–672.doi:10.1016/j.yjmcc.2009.11.017

Drews O, Tsukamoto O, Liem D, Streicher J, Wang Y, Ping P (2010).Differential regulation of proteasome function in isoproterenol-induced cardiac hypertrophy Circ Res 107: 1094–1101 doi:10.1161/CIRCRESAHA.110.222364

Engebretsen KV, Skårdal K, Bjørnstad S, Marstein HS, Skrbic B,Sjaastad I et al (2014) Attenuated development of cardiacﬁbrosis

in left ventricular pressure overload by SM16, an orally active inhibitor

of ALK5 J Mol Cell Cardiol 76: 148–157 doi:10.1016/j

yjmcc.2014.08.008

Friedberg MK, Cho MY, Li J, Assad RS, Sun M, Rohailla S et al (2013).Adverse biventricular remodeling in isolated right ventricularhypertension is mediated by increased transforming growth factor-β1signaling and is abrogated by angiotensin receptor blockade Am JRespir Cell Mol Biol 49: 1019–1028 doi:10.1165/rcmb.2013-0149OC.Goumans MJ, Valdimarsdottir G, Itoh S, Rosendahl A, Sideras P, tenDijke P et al (2002) Balancing the activation state of the

endothelium via two distinct TGF-beta type I receptors EMBO J 21:

1743–1753, PMID:11927558

Grünenfelder J, Miniati DN, Murata S, Falk V, Hoyt EG, Robbins RC(2002) Up-regulation of Bcl-2 through hyperbaric pressuretransfection of TGF-beta1 ameliorates ischemia-reperfusion injury inrat cardiac allografts J Heart Lung Transplant 21: 244–250

Heusch G (2015) Mitochondria at the heart of cardiovascularprotection: p66shc-friend or foe? Eur Heart J 36: 469–471

doi:10.1093/eurheartj/ehu409

Hill C, Würfel A, Heger J, Meyering B, Schlüter KD, Weber M

et al (2013) Inhibition of AP-1 signaling by JDP2 overexpressionprotects cardiomyocytes against hypertrophy and apoptosisinduction Cardiovasc Res 99: 121–128 doi:10.1093/cvr/cvt094.Hoppe UC, Erdmann E (2009) Kapitel 5 Chronische Herzinsufﬁzienz,aus Erdmann E Klinische Kardiologie Krankheiten des Herzens, desKreislaufs und herznaher Gefäße 7 Auﬂage, Springer VerlagHuang CY, Kuo WW, Yeh YL, Ho TJ, Lin JY, Lin DY et al (2014) ANG IIpromotes IGF-IIR expression and cardiomyocyte apoptosis byinhibiting HSF1 via JNK activation and SIRT1 degradation Cell DeathDiffer 21: 1262–1274 doi:10.1038/cdd.2014.46

Huang ZP, Chen J, Seok HY, Zhang Z, Kataoka M, Hu X et al (2013).MicroRNA-22 regulates cardiac hypertrophy and remodeling inresponse to stress Circ Res 112: 1234–1243 doi:10.1161/

Hullinger TG, Montgomery RL, Seto AG, Dickinson BA, Semus HM,Lynch JM et al (2012) Inhibition of miR-15 protects against cardiacischemic injury Circ Res 110: 71–81 doi:10.1161/

Trang 13

Huntgeburth M, Tiemann K, Shahverdyan R, Schlüter KD,

Schreckenberg R, Gross ML et al (2011) Transforming growth factor

β1oppositely regulates the hypertrophic and contractile response to

β-adrenergic stimulation in the heart PLoS One 6: e26628

doi:10.1371/journal.pone.0026628

Ikeuchi M, Tsutsui H, Shiomi T, Matsusaka H, Matsushima S, Wen J

et al (2004) Inhibition of TGF-beta signaling exacerbates early

cardiac dysfunction but prevents late remodeling after infarction

Cardiovasc Res 64: 526–535 doi:10.1016/j.cardiores.2004.07.017

Jain A, Atale N, Kohli S, Bhattacharya S, Sharma M, Rani V (2015) An

assessment of norepinephrine mediated hypertrophy to apoptosis

transition in cardiac cells: a signal for cell death Chem Biol Interact

225: 54–62 doi:10.1016/ j.cbi.2014.11.017

Khan S, Joyce J, Margulies KB, Tsuda T (2014) Enhanced bioactive

myocardial transforming growth factor-β in advanced human heart

failure Circ J 78: 2711–2718 doi:10.1253/circj.CJ-14-0511

Kinnally KW, Peixoto PM, Ryu SY, Dejean LM (2011) Is mPTP the

gatekeeper for necrosis, apoptosis, or both? Biochim Biophys Acta 4:

616–624 doi:10.1016/j.bbamcr.2010.09.013

Kolwicz SC Jr, Tian R (2011) Glucose metabolism and cardiac

hypertrophy Cardiovasc Res 90: 194–201 doi:10.1093/cvr/cvr071

Kumarswamy R, Volkmann I, Jazbutyte V, Dangwal S, Park DH,

Thum T (2012) Transforming growth factor-β-induced

endothelial-to-mesenchymal transition is partly mediated by microRNA-21

Arterioscler Thromb Vasc Biol 32: 361–369 doi:10.1161/

ATVBAHA.111.234286

Lai L, Leone TC, Keller MP, Martin OJ, Broman AT, Nigro J et al

(2014) Energy metabolic reprogramming in the hypertrophied and

early stage failing heart: a multisystems approach Circ Heart Fail 7:

1022–1031 doi:10.1161/CIRCHEARTFAILURE.114.001469

Leksa V, Godar S, Schiller HB, Fuertbauer E, Muhammad A,

Slezakova K (2005) TGF-beta-induced apoptosis in endothelial

cells mediated by M6P/IGFII-R and mini-plasminogen J Cell Sci

118 (Pt 19): 4577–4586 doi:10.1242/ jcs.02587

Li L, Chen Y, Doan J, Murray J, Molkentin JD, Liu Q (2014a)

Transforming growth factorβ-activating kinase 1 signaling pathway

critically regulated myocardial survival and remodeling Circulation

Li J, Qi M, Li C, Shi D, Zhang D, Xie D et al (2014b) Tom70 serves as a

molecular switch to determine pathological cardiac hypertrophy Cell

Res 24: 977–993 doi:10.1038/Cr.2014.9

Liao R, Jain M, Cui L, D’Agostino J, Aiello F, Luptak I et al (2002)

Cardiac-speciﬁc overexpression of GLUT1 prevents the development

of heart failure attributable to pressure overload in mice Circulation

106: 2125–2131 doi:10.1161/01.CIR.0000034049.61181

Lijnen PJ, Petrov VV, Fagard RH (2000) Induction of cardiac

ﬁbrosis by transforming growth factor-beta(1) Mol Genet Metab

71: 418–435 PMID: 11001836

Lymperopoulos A, Negussie S (2013).βArrestins in cardiac G

protein-coupled receptor signaling and function: partners in crime or "good cop,

bad cop"? Int J Mol Sci 14: 24726–24741 doi:10.3390/ijms141224726

Matkovich SJ, Wang W, Tu Y, Eschenbacher WH, Dorn LE, Condorelli

G et al (2010) MicroRNA-133a protects against myocardialﬁbrosis

and modulates electrical repolarization without affecting

hypertrophy in pressure-overloaded adult hearts Circ Res 106:

166–175 doi:10.1161/CIRCRESAHA.109.202176

Mitra A, Basak T, Datta K, Naskar S, Sengupta S, Sarkar S (2013) Role of

α-crystallin B as a regulatory switch in modulating cardiomyocyte

apoptosis by mitochondria or endoplasmic reticulum during cardiac

hypertrophy and myocardial infarction Cell Death Dis 4: e582.doi:10.1038/cddis.2013.114

Montgomery RL, Hullinger TG, Semus HM, Dickinson BA, Seto AG,Lynch JM et al (2011) Therapeutic inhibition of miR-208a improvescardiac function and survival during heart failure Circulation 124:

1537–1547 doi:10.1161/CIRCULATIONAHA.111.030932

Munger JS, Huang X, Kawakatsu H, Grifﬁths MJ, Dalton SL, Wu J

et al (1999) The integrin alpha v beta 6 binds and activateslatent TGF beta 1: a mechanism for regulating pulmonaryinﬂammation and ﬁbrosis Cell 6: 319–328 doi:10.1016/S0092-8674(00)80545-0

Narula J, Haider N, Virmani R, DiSalvo TG, Kolodgie FD, Hajjar RJ(1996) Apoptosis in myocytes in end-stage heart failure N Engl JMed 335: 1182–1189 doi:10.1056/NEJM199610173351603.Pawson AJ, Sharman JL, Benson HE, Faccenda E, Alexander SP,Buneman OP et al (2014) The IUPHAR/BPS Guide toPHARMACOLOGY: an expert-driven knowledgebase of drug targetsand their ligands Nucl Acids Res 42 (Database Issue): D1098–106.Piao L, Fang YH, Cadete VJ, Wietholt C, Urboniene D, Toth PT

et al (2010) The inhibition of pyruvate dehydrogenase kinaseimproves impaired cardiac function and electrical remodeling intwo models of right ventricular hypertrophy: resuscitating thehibernating right ventricle J Mol Med (Berl) 88: 47–60

doi:10.1007/s00109-009-0524-6

Piao L, Fang YH, Parikh KS, Ryan JJ, D’Souza KM, Theccanat T et al.(2012) GRK2-mediated inhibition of adrenergic and dopaminergicsignaling in right ventricular hypertrophy: therapeutic implications

in pulmonary hypertension Circulation 126 (24): 2859–2869.doi:10.1161/CIRCULATIONAHA.112.109868

Portbury AL, Ronnebaum SM, Zungu M, Patterson C, Willis MS(2012) Back to your heart: ubiquitin proteasome system-regulatedsignal transduction J Mol Cell Cardiol 52: 526–537 doi:10.1016/j.yjmcc.2011.10.023

Qiu H, Lizano P, Laure L, Sui X, Rashed E, Park JY et al (2011) H11kinase/heat shock protein 22 deletion impairs both nuclear andmitochondrial functions of STAT3 and accelerates the transition intoheart failure on cardiac overload Circulation 24: 406–415

doi:10.1161/CIRCULATIONAHA.110.013847

Rosenkranz S, Flesch M, Amann K, Haeuseler C, Kilter H, Seeland U

et al (2002) Alterations of beta-adrenergic signaling and cardiachypertrophy in transgenic mice overexpressing TGF-beta(1) Am JPhysiol Heart Circ Physiol 283: H1253–H1262 doi:10.1152/ajpheart.00578.2001

Rosenkranz S (2004) TGF-beta1 and angiotensin networking incardiac remodeling Cardiovasc Res 94: 2837–2842 doi:10.1016/j.cardiores.2004.04.030

Ryan JJ, Huston J, Kutty S, Hatton ND, Bowman L, Tian L et al (2015).Right ventricular adaptation and failure in pulmonary arterialhypertension Can J Cardiol 31: 391–406 doi:10.1016/j

cjca.2015.01.023

Sack MN, Rader TA, Park S, Bastin J, McCune SA, Kelly DP (1996) Fattyacid oxidation enzyme gene expression is downregulated in thefailing heart Circulation 94: 2837–2842 doi:10.1161/01

Trang 14

cardiomyocytes Cardiovasc Res 67: 87–96 doi:10.1016/j.

cardiores.2005.02.021

Schreckenberg R, Rebelo M, Deten A, Weber M, Rohrbach S, Pipicz M

et al (2015) Speciﬁc Mechanisms Underlying Right Heart Failure:

The Missing Upregulation of Superoxide Dismutase-2 and Its Decisive

Role in Antioxidative Defense Antioxid Redox Signal Epub ahead of

print, PMID:25978844

Schröder D, Heger J, Piper HM, Euler G (2006) Angiotensin II

stimulates apoptosis via TGF-beta1 signaling in ventricular

cardiomyocytes of rat J Mol Med (Berl) 84: 975–983 doi:10.1007/

s00109-006-0090-0

Taimor G, Schlüter KD, Best P, Helmig S, Piper HM (2004) Transcription

activator protein 1 mediates alpha- but not beta-adrenergic hypertrophic

growth responses in adult cardiomyocytes Am J Physiol Heart Circ

Physiol 286: H2369–H2375 doi:10.1152/ajpheart.00741.2003

Thum T, Gross C, Fiedler J, Fischer T, Kissler S, Bussen M et al (2008)

MicroRNA-21 contributes to myocardial disease by stimulating MAP

kinase signalling inﬁbroblasts Nature 456: 980–984 doi:10.1038/

nature07511

Thum T, Chau N, Bhat B, Gupta SK, Linsley PS, Bauersachs J, et al

(2011) Comparison of different miR-21 inhibitor chemistries in a

cardiac disease model J Clin Invest 121:461-462: 462–463 doi:

10.1172/JCI45938

Tijsen AJ, van der Made I, van den Hoogenhof MM, Wijnen WJ, van

Deel ED, de Groot NE et al (2014) The microRNA-15 family inhibits

the TGFβ-pathway in the heart Cardiovasc Res 104: 61–71

doi:10.1093/cvr/cvu184

Topkara VK, Mann DL (2011) Role of microRNAs in cardiac

remodeling and heart failure Cardiovasc Drugs Ther 25: 171–182

doi:10.1007/s10557-011-6289-5

Wahlquist C, Jeong D, Rojas-Muñoz A, Kho C, Lee A, Mitsuyama S

et al (2014) Inhibition of miR-25 improves cardiac contractility in

the failing heart Nature 508: 531–535 doi:10.1038/nature13073

Walther T, Tschope C, Sterner-Kock A, Westermann D,

Heringer-Walther S, Riad A et al (2007) Accelerated mitochondrial adenosine

diphosphate/adenosine triphosphate transport improveshypertension-induced heart disease Circulation 115: 333–344.doi:10.1161/CIRCULATIONAHA.106.643296

Wang J, Xu N, Feng X, Hou N, Zhang J, Cheng X et al (2005) Targeteddisruption of Smad4 in cardiomyocytes results in cardiac

hypertrophy and heart failure Circ Res 97: 821–828 doi:10.1161/01.RES.0000185833.42544.06

Watkins SJ, Borthwick GM, Oakenfull R, Robson A, Arthur HM (2012).Angiotensin II-induced cardiomyocyte hypertrophy in vitro is TAK1-dependent and Smad2/3-independent Hypertens Res 35: 393–398.doi:10.1038/hr.2011

Wharton K, Derynck R (2009) TGFbeta family signaling: novelinsights in development and disease Development 136: 3691–3697.doi:10.1242/dev.040584

Xu P, Liu J, Derynck R (2012) Post-translational regulation of TGF-βreceptor and Smad signaling FEBS Lett 586: 1871–1884 doi:10.1016/j.febslet.2012.05.010

Zaglia T, Milan G, Ruhs A, Franzoso M, Bertaggia E, Pianca N et al.(2014) Atrogin-1 deﬁciency promotes cardiomyopathy andpremature death via impaired autophagy J Clin Invest 124:

2410–2424 doi:10.1172/JCI66339

Zhang D, Gaussin V, Taffet GE, Belaguli NS, Yamada M, Schwartz RJ

et al (2000) TAK1 is activated in the myocardium after pressureoverload and is sufﬁcient to provoke heart failure in transgenic mice.Nat Med 6: 556–563 doi:10.1038/75037

Zhang YE (2009) Non-Smad pathways in TGF-beta signaling Cell Res19: 128–139 doi:10.1038/cr.2008.328

Zhang DW, Shao J, Lin J, Zhang N, Lu BJ, Lin SC et al (2009) RIP3, anenergy metabolism regulator that switches TNF-induced cell deathfrom apoptosis to necrosis Science 325: 332–336 doi:10.1126/science.1172308

Zaiman AL, Podowski M, Medicherla S, Gordy K, Xu F, Zhen L et al.(2008) Role of the TGF-beta/Alk5 signaling pathway in

monocrotaline-induced pulmonary hypertension Am J Respir CritCare Med 177: 896–905 doi:10.1164/rccm.200707-1083OC

Trang 15

Seung-Hyun Jung†, Young Sook Kim†, Yu-Ri Lee and Jin Sook Kim

Korean Medicine Convergence Research Division, Korea Institute of Oriental Medicine (KIOM),

Daejeon 305-811, Korea

Correspondence

Jin Sook Kim, Korean MedicineConvergence Research Division,Korea Institute of Oriental Medicine(KIOM), 1672 Yuseongdae-ro,Yuseong-gu, Daejeon 305-811, Korea.E-mail: jskim@kiom.re.kr

†These authors contributed equally tothis work

BACKGROUND AND PURPOSE

Although a variety of animal models have been used to test drug candidates and examine the pathogenesis of diabetic

retinopathy, time-saving and inexpensive models are still needed to evaluate the increasing number of therapeutic approaches.EXPERIMENTAL APPROACH

We developed a model for diabetic retinopathy using the early stage of transgenic zebraﬁsh (ﬂk:EGFP) by treating embryos with

130 mM glucose, from 3-6 days post fertilisation (high-glucose model) On day 6, lenses from zebrafish larvae were isolated andtreated with 3% trypsin, and changes in hyaloid-retinal vessels were analysed usingfluorescent stereomicroscopy In addition,expression of tight junction proteins (such as zonula occludens-1), effects of hyperosmolar solutions and of hypoxia, and Vegfexpression were assessed by RT–PCR NO production was assessed with a fluorescent substrate Effects of inhibitors of the VEGFreceptor, NO synthesis and a VEGF antibody (ranibizumab) were also measured

KEY RESULTS

In this high-glucose model, dilation of hyaloid-retinal vessels, on day 6, was accompanied by morphological lesions with ruption of tight junction proteins, overproduction of Vegf mRNA and increased NO production Treatment of this high-glucosemodel with an inhibitor of VEGF receptor tyrosine kinase or an inhibitor of NO synthase or ranibizumab decreased dilation ofhyaloid-retinal vessels

dis-CONCLUSIONS AND IMPLICATIONS

Theseﬁndings suggest that short-term exposure of zebraﬁsh larvae to high-glucose conditions could be used for screening anddrug discovery for diabetic retinopathy and particularly for disorders of retinal vessels related to disruption of tight junctionproteins and excessive VEGF and NO production

Trang 16

Hyperglycaemia is involved in retinal vascular dysfunction

asso-ciated with the development of diabetic retinopathy (DR), which

results in worsening vision and eventual blindness Nearly all

pa-tients with type 1 and type 2 diabetes exhibit some lesions after

20 years with disease (Kempen et al., 2004; Roy et al., 2004;

Calcutt et al., 2009) As early DR progresses, non-proliferative

DR enters an advanced proliferative stage Loss of pericytes,

reti-nal vasculature thickening and disruption of tight junction

pro-teins such as zonula occludens-1 (ZO-1) are found in the early

stages of DR (Checchin et al., 2006; Pﬁster et al., 2008)

Further-more, these events result in decreased circulation, blood vessel

dilation and obstruction, and increased leakage of blood from

the microvascular circulation, thereby increasing VEGF

signal-ling Increased VEGF signalling causes the formation of new

blood vessels from retinal tissues (Guillemin and Brew, 2004;

Armulik et al., 2005; Bergers and Song, 2005; Checchin et al.,

2006; Alvarez et al., 2010) These vascular changes, collectively,

cause a breakdown of the blood-retinal barrier and compromise

retinal function Neovascularization on the retina and posterior

surface of the vitreous, and macular oedema characterized by

retinal thickening from leaky blood vessels, develop during

vari-ous stages of retinopathy (Fong et al., 2003) Moreover,

expres-sion of endothelial NOS (eNOS), a downstream mediator of

VEGF activity, was increased in diabetic retina, and eNOS

inhibi-tion also potently reduced retinal leukocyte adhesion (Joussen

et al., 2002) Several studies indicate that NO plays a crucial role

in VEGF-induced vascular hyperpermeability and angiogenesis

(Morbidelli et al., 1996; Lakshminarayanan et al., 2000)

Addi-tionally, VEGF induces the expression of NOS and stimulates

production of NO (Abu El-Asrar et al., 2004) Therapeutic

ap-proaches such as VEGF inhibitors, aldose reductase inhibitors,

NOS inhibitors, antioxidants, anti-inﬂammatory agents and

the receptor for advanced glycation end products are proposed

interventions for treating DR (Greene et al., 1999; Joussen et al.,

2002; Starita et al., 2007; Calcutt et al., 2009)

Many species, including dogs, hamsters, rats, mice and

zebraﬁsh, have been used to provide research models for DR

(Engerman and Kern, 1995; Jorgens et al., 2012) These models

employ chemical or genetic modiﬁcations to induce early stages

of DR, including degeneration of retinal capillaries (Kempen

et al., 2004) Zebraﬁsh are an attractive animal model for

molec-ular, toxicological and drug development studies because of

their fecundity, as well as their genetic and physiological

similar-ities to mammals (Barros et al., 2008) They are generally suited

to high-throughput screening because of their small size, highproductivity and optical transparency of the embryo (Kitambi

et al., 2009) Furthermore, techniques for generating transgeniclines and gene-targeting mutations have been developed forzebrafish, enabling the establishment of disease models for drugdiscovery The immersion of adult or larvae zebrafish in glucose(0.7–25%) results in diabetic complications, which share similari-ties with streptozotocin (STZ)-induced diabetic mice and diabeticpatients (Gleeson et al., 2007; Alvarez et al., 2010; Liang et al.,2010; Jorgens et al., 2012) Additionally, in retinas of adultzebrafish exposed for 30 days to hyperglycaemia, there are mor-phological changes such as thickening of the vessels, breakdown

of interendothelial cell–cell junction integrity and vessel ment membrane thickening (Gleeson et al., 2007; Alvarez et al.,2010) In larval zebrafish, anti-diabetic compounds reduce expres-sion of phosphoenolpyruvate carboxykinase, which catalyses arate-limiting step in gluconeogenesis and is transcriptionallyregulated by glucagon and insulin Data have shown that larvaezebrafish are an appropriate model for the study of glucose metab-olism (Elo et al., 2007) Our previous study, we tested the inhibi-tory effect of single compounds isolated from herbal extracts onhyaloid-retinal vessel dilation in glucose-induced larval zebrafish(Lee et al., 2013; Yu et al., 2013) However, changes of tight junc-tions in retinal vessels in glucose-induced zebrafish larvae havenot previously been described in a short-term model for selection

base-of effective agents against DR

Here, to examine the potential of larval zebrafish as a modelfor DR, we exposed zebrafish embryos, from flk:EGFP transgenicfish, to a range of concentrations of glucose for 3 days, and iso-lated the retina Alterations in thickness of the hyaloid-retinalvessels, expression of the tight junction protein ZO-1, vascularleakage and mRNA expression for Vegf and nos were measured

in glucose-exposed zebraﬁsh larvae We also tested the effects

of an inhibitor of VEGFR tyrosine kinase, an inhibitor of NOS(L-NAME) and the clinically used VEGF antibody ranibizumab(FDA approved drug for DR)

Methods

Zebrafish maintenanceAll animal husbandry and experimental protocols complied withinstitutional guidelines and were approved by local ethical boards(Korea Institute of Oriental Medicine Animal Care and Use Com-mittee) Studies involving animals are reported in accordance with

These Tables list key protein targets and ligands in this article which are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Pawson et al., 2014) and are permanently archived in the ConciseGuide to PHARMACOLOGY 2013/14 (a bAlexander et al., 2013a,b)

BJP S-H Jung et al.

Trang 17

the ARRIVE guidelines for reporting experiments involving

ani-mals (Kilkenny et al., 2010; McGrath et al., 2010) Adult zebraﬁsh

were maintained under standard conditions at 28.5°C with a

14 h light/10 h dark cycle (Nusslein and Dahm, 2002) Embryos

were obtained from crosses between flk:EGFP transgenicﬁsh and

raised in embryonic water Transgenic zebraﬁsh (flk:EGFP) were

provided by Professor C-H Kim (Chungnam National University)

Embryonic stages were determined by days post-fertilization (dpf)

Fluorescent images were collected at 6 dpf using a ﬂuorescent

dissecting microscope SZX16 (Olympus, Tokyo, Japan)

Treatments and measurement of glucose levels

Zebraﬁsh transgenic (flk:EGFP) embryos (3 dpf) were placed in

24-well plates (ﬁve embryos per well) and maintained in 2 mL

embryonic water (sea salt, Sigma, USA, 0.06 g·L 1) containing

a range of concentrations of glucose (30, 60, 90, 120, 125, 130,

135 and 150 mM) Media were maintained for 3 days, and

embryos wereﬁxed, and survival rates were analysed at 6 dpf

In later experiments, embryos were exposed to a single

concen-tration of glucose (130mM) from 3 to 6 dpf and are referred to as

high glucose (HG)-treated larvae

To assess the effects of hyperosmolarity, embryos were

exposed to mannitol (130 mM; Sigma, USA, Cat No C8661),

in-stead of glucose, but under the same conditions To assess effects

of hypoxia, embryos were exposed to CoCl2 (120μM; Sigma,

USA, Cat No M9546) instead of glucose, but under the same

conditions The effects of the VEGF receptor tyrosine kinase

in-hibitor (0.5 or 1μM; Calbiochem, Germany, Cat No 676500),

or ranibizumab (1 or 2.5μg·mL 1

; Lucentis, Novatris, Stein,Switzerland) or L-NAME (10 or 20μM; Sigma, USA, Cat No

N5751) were tested by addition to the HG treatment for 3 days

Quantitative analysis of glucose levels was performed from

whole body lysates using a glucose assay kit (Sigma, USA) Brieﬂy,

20 zebraﬁsh larvae in each experimental group were sonicated in

500μL deionized water on ice According to the instructions,

standard curves were generated using glucose standard solution

A total of 2 mL assay reagent (Sigma) was added and incubated

for 30 min at 37°C Fluorescence (540 nm) was measured using

a BioTek plate reader equipped with GEN5 software (v.2.04,

BioTek, Winooski, VT, USA)

Lens isolation and measurement of

hyaloid-retinal vessel diameter

At 6 dpf, zebraﬁsh larvae were ﬁxed with 4%

paraformalde-hyde and stored overnight at 4°C They were then washed

with distilled water (1.5 ml per well, 3 times within 1 h) and

lenses containing hyaloid-retinal vessels were isolated after

incubation with 3% trypsin in Tris–HCl buffer (pH 7.8), for

80 min at 37°C (Figure 1) Images of hyaloid-retinal vessels,

optic disc (OD) branches and lens size were obtained using

an Olympus stereomicroscope (SZX16) or a laser scanning

confocal microscope (FV10i, Olympus, Tokyo, Japan) Their

diameters were determined byIMAGEJsoftware (Figure S1)

Total RNA extraction and RT-PCR

Total RNA from whole embryos was extracted using TRIzol reagent

according to the manufacturer’s protocol (iNtRON, Korea) and

reverse transcribed using M-MLV reverse transcriptase (Invitrogen,

USA) The reverse transcription (RT) product (1μL) was used as

a template for PCR ampliﬁcation of Vegf165 and β-actin using

speciﬁc primers (5 pM; Genotech, Korea), Taq DNA polymerase(Elpis, Korea) and the following cycles: 95°C for 30 s, 55°C for

30 s and 68°C for 40 s for 30 cycles Primer pairs were as follows:Vegf165, forward primer 5′-CTC CTC CAT CTG TCT GCT GTAAAG-3′ and reverse primer 5′-CTC TCT GAG CAA GGC TCACAG-3′; nos2a, forward primer 5′-GTG TTC CCT CAG AGAACA GAT-3′ and reverse primer 5′-GAT CAG TCC TTT GAA GCTGAC-3′; and β-actin, forward primer 5′-GAG AAG ATC TGG CATCAC AC-3′ and reverse primer 5′-ATC AGG TAG TCT GTC AGGTC-3′ PCR products were separated by electrophoresis on 1.2%agarose gels and visualized by ethidium bromide staining.Whole-mount immunostaining

Whole-mount immunostaining was performed as previouslydescribed (Jung et al., 2010) using primary antibodies againstmouse ZO-1 (1:500, Invitrogen, USA) Forﬂuorescent detection

of antibody labelling, we used Alexa Fluor 568 anti-mouse jugate (1:500, Molecular Probes, Eugene, OR, USA) and Hoechst

con-33342 (1:1000, Thermo Fisher Scientiﬁc, Waltham, MA, USA).Fluorescence images were examined using an Olympus laserscanning confocal microscope (FV10i)

NO detectionLive imaging detection of NO production was carried out asdescribed by Lepiller et al (2007) Brieﬂy, wild-type larvae treatedwith control and HG at 6 dpf were incubated in 5μM 4,5-diamino-ﬂuorescein diacetate (DAF-FM DA) (Molecular Probes)for 30 min, rinsed in egg water, anaesthetized and imaged usingOlympus laser scanning confocal microscopy (FV10i)

Data analysisResults are expressed as the mean ± SEM of multiple experi-ments Experimental data are from 40–60 embryos per treat-ment, from 2–3 independent clutches Data were analysedusing one-way ANOVAwith Tukey’s post hoc analysis or unpairedStudent’s t-tests andPRISMsoftware (GraphPad, San Diego, CA,USA) P< 0.05 was considered statistically signiﬁcant

to DR models Thus, retinal vascular changes in HG zebraﬁshmay reﬂect similar changes observed in mammalian models

To investigate changes in hyaloid-retinal vessels induced by

HG in zebraﬁsh Tg (flk:EGFP), we exposed embryos to anrange of concentrations of glucose (30–150 mM) from 3 to

6 dpf and analysed the diameter of the hyaloid-retinal vesselsobtained at 6dpf (Figure 2) The hyaloid-retinal vasculature be-gins to be generated on 2.5 dpf in normal zebraﬁsh development(Alvarez et al., 2007) Thus, we exposed embryos for 3 days, from

3 to 6 dpf to the different levels of glucose The experimentalprotocols used to measure glucose-induced changes in hyaloid-retinal vessel diameter in zebraﬁsh embryos are summarised inFigure 1 (and S1) Using retinae isolated at 6 dpf, we found that

HG-induced changes in hyaloid-retinal vessels in ZF

BJP

Trang 18

the diameter of hyaloid-retinal vessels increased

concentration-dependently, from 30 to 135 mM glucose (Figure 2A)

Further-more, we observed that the proportion of lenses containing

vessels with diameter more than 1.5 AU (artiﬁcial unit) also

increased with the concentration of glucose, from about 20%

of lenses with large vessels at 30mM glucose to just over 80%

of lenses at 130mM (Figure 2B) This effect is referred to as

inductivity The effects of these conditions on the survival of

the larvae is shown in Figure 2C Survival was unaffected,

com-pared with control (0 mM glucose) values, from 30 to 120 mM

glucose However, at 135 and 150 mM glucose there was a icant decrease in survival at 6 dpf

signif-On the basis of the results shown in Figure 2, we chose130mM glucose as a standard treatment to induce changes

in the retinal vessels and this is referred to as the HG tion Under these HG conditions, we examined the vesseldiameter, the rate of induction , overall survival rates andglucose levels in the zebraﬁsh larvae, along with other aspects

condi-of embryonic development, as described below (Figure 3) In

10 preparations, HG treatment did not induce the dilated

Figure 1

Overall strategy for rapid screening to identify therapeutic drugs for DR in zebrafish larvae (A) Fertilized embryos from zebrafish Tg (flk:EGFP) pairs wereproduced GFP-positive embryos at 3 dpf were placed in 24-well plates and maintained in embryonic water containing high concentrations of glucose(130 mM; HG) and candidate drugs (B) At 6 dpf, larvae deposited in 24-well plates werefixed with 4% paraformaldehyde and stored overnight at 4°C.After washing three times for 1 h with distilled water (DW), larvae were incubated for 80 min with 3% trypsin [Tris–HCl (pH 7.8)] at 37°C and washedthree times for 1 h with DW on an orbital shaker at room temperature Using a stereomicroscope, lenses with hyaloid-retinal vessels were detached using

a single channel pipette (P200) by strongly pipetting buffer up and down in the 24-well plates Lenses were transferred to cover glasses in dishes(35 × 10 mm) Images of hyaloid-retinal vessels were observed usingﬂuorescence stereomicroscopy

Trang 19

retinal vessels in 3 samples (Figure 3A) This phenomenon is

thought to be modulated by different individuals In the 7

prep-arations that did show these changes, the inductivity was 75%

(Figure 3B) Under the HG conditions, the overall survival rate

was approximately 80% (Figure 3C) Apart from the changes

in the hyaloid-retinal vessels, there were no obvious differences

in the structural development of the embryos, as assessed by

mi-croscopic observation, between the HG group and the control

group (Figure 3B, C and E) In addition, glucose levels in whole

embryo lysates were measured and found to be much higher

in the HG group than those in the control group (Figure 3D)

Because changes in the size of the retinal vessels could reﬂect

changes in the size of the lens itself, we measured the lens size in

the HG and control groups As shown in Figure S2A, there was

no signiﬁcant difference in lens size between these two

experi-mental groups Moreover there was no difference in the number

of OD branches in the HG and control groups (Figure S2B)

As the HG conditions affected the hyaloid-retinal vessels, we

looked for changes in other parts of the vasculature, using

con-focal microscopy With this method we were able to see through

the lens to the hyaloid-retinal vessels in zebraﬁsh larvae and

were able to conﬁrm that HG treatment increased the diameter,

compared with control conditions (Figures S3A, 3B and 3C)

However, the diameters of blood vessels in the body parts

present from the ninth segment to the 14 segment and of the

intersomitic vessels (ISVs) and the dorsal aorta (DA) showed

no signiﬁcant differences between the HG and control groups(Figures S3A and 3B)

Effects of mannitol and CoCl2on zebrafish hyaloid vessels Next, weassessed the contribution of the osmotic pressure of the HGtreatment (the glucose concentration is 130 mM) to the changes

in the hyaloid-retinal vessels in zebraﬁsh larvae, by using 130mMmannitol instead of glucose for the 3 days from 3-6dpf Asshown in Figure 4, the hyaloid-retinal vessels of mannitol-treatedlarvae were not affected compared with the control vessels,whereas the HG treatment clearly increased the vessel diameter.Another physical change associated with DR is hypoxia(Geoffrey and Sobha, 2011) and hypoxic conditions haveinduced retinopathy in adult zebraﬁsh (Cao et al., 2010) Wetherefore evaluated the effects of hypoxia induced by CoCl2

on the hyaloid-retinal vessels In these experiments, weexposed the zebraﬁsh embryos to 120 μM CoCl2, instead ofglucose, from 3-6dpf and found that this condition of hyp-oxia was as effective as HG in increasing the diameter of thehyaloid-retinal vessels (Figure 4A, B and D)

HG-induced vessel dilation leads to a defect in tight-junctionproteins In DR, pericyte loss and vessel thickening lead tocompromised integrity of the vascular walls, resulting inincreased permeability and changes in tight junctionprotein formation in retinal vessels (Bergers and Song, 2005;

BJP

Trang 20

Pﬁster et al., 2008) As shown in Figure 5A and D, at 6 dpf, after 3

days of HG treatment, the overall diameter of hyaloid-retinal

vessels in the HG-treated zebraﬁsh larvae was thicker than that

in the control vessels The tight junction protein ZO-1 is an

intracellular adaptor protein that regulates permeability in the

retinal vasculature When we measured expression of this tight

junction protein in hyaloid-retinal vessels, using a ZO-1-speciﬁc

antibody, we found that expression of this protein in the vessels

from from HG-treated larvae showed markedly irregular and

discontinuous staining, compared with the staining in control

vessels (Figure 5B and E)

Effects of VEGF signalling and the VEGFR

inhibitor in HG-induced larvae

The mitogenic cytokine VEGF exhibits potent pro-angiogenic

activity and also increases vascular permeability Moreover,

VEGF has been implicated in the pathogenesis of DR throughangiogenesis induced by binding to VEGFR2 in endothelialcells (Sheetz and King, 2002; Caldwell et al., 2003; Scheppke

et al., 2008) We therefore assessed the role of VEGF in ourmodel of DR

First, we examined the effects of inhibiting VEGF signallingusing a VEGFR tyrosine kinase inhibitor (0.5 or 1μM), added tothe HG treatment, from 3 to 6 dpf, of zebraﬁsh larvae As shown

in Figure 6A and 6B, the addition of the VEGF inhibitor blockedthe effects of HG treatment on the vessels and, at the higher con-centration (1μM), treated vessels were identical to those in con-trol larvae Next, we wanted to evaluate the expression of theVegf165 gene in hyaloid-retinal vessels However, due to the smallbody size of zebrafish larvae, it is not possible to isolate specifictissues for tissue-specific mRNA quantifications without notice-able contamination from surrounding tissues So we carried outour evaluation using RT-PCR and mRNAs extracted from whole

Figure 3

Changes in hyaloid-retinal vessel diameter induced by 130 mM glucose(HG) (A) The average diameter andﬂuorescent images of hyaloid-retinal vessels inHG-treated larvae The diameter of hyaloid-retinal vessels was measured at locations proximal (red circle) to the OD from each group The graph displaysthe mean AU for vessel diameter The vessel diameter of each lens was measured three times ***P< 0.001, signiﬁcantly different from control, n = 8–10embryos per group Scale bar = 40μm (B) Inductivity for the vessel diameter was increased in the HG-treated group, compared with the control group

***P< 0.001, significantly different from control, n = 8–10 embryos per group (C) Survival rates in HG-treated zebrafish larvae *P < 0.05, significantlydifferent from control, n = 10 embryos per group (D) After HG-treatment for 3 days, the zebrafish displayed increased levels of glucose ***P < 0.001,significantly different from control, n = 20 embryos per group (E) Control and HG-treated larvae did not show gross morphological changes at 6dpf Scalebar = 300μm

Trang 21

body lysates As shown in Figure 6C and D, Vegf165 mRNA was

up-regulated in larvae after HG-treatment from 3-6dpf, compared

with control larvae and this increase was clearly prevented by

addition of the VEGFR inhibitor

To further validate our model, we used a monoclonal

anti-body to VEGF, ranibizumab (Lucentis), which has been used

clinically for the treatment of retinal neovascularization This

antibody was added to the HG treatment for 3 days and

substan-tially prevented the increase in vessel diameter induced by HG

treatment alone (Figure 6E) The effects of the ranibizumab were

signiﬁcant only at the higher concentration used (2.5μg mL-1

;Figure 6F)

NO activation and the effect of L-NAME on the dilation of hyaloid-retinal vessels after

HG-treatmentSignalling by NO is important for regulating blood vessel dila-tion and HG conditions are known to increase NOS expression(Francesco et al., 1997) Zebraﬁsh larvae produce NO, and this

Figure 4

Effects of mannitol and CoCl2on hyaloid vessels of zebraﬁsh larvae [(A)–(D)] To assess possible effects of the hyperosmolarity of the HG treatment,mannitol (130 mM) was added instead of glucose HG-treated larvae (B) showed dilated hyaloid vessels compared to the control (A), whereas hyaloidvessels of the mannitol-treated larvae (C) were not affected (D) The effects of hypoxic conditions on hyaloid vessel development from 3-6 dpf wasstudied by adding CoCl2(120μM) instead of glucose (130 mM) for three days Under these conditions of hypoxia, we found dilated hyaloid vessels,similar to those after HG exposure (B) Scale bar = 40μm (E) Graph data are displayed as the mean AU for vessel diameter The diameter of each lenswas measured three times ###P< 0.001, signiﬁcantly different from control, n = 16–20 embryos per group

Figure 5

Expression of the tight junction protein ZO-1ion hyaloid-retinal vessels of HG-treated larvae Confocal immunoﬂuorescence microscopy images ofhyaloid-retinal vessels on the dissected lens after treatment with trypsin The overall morphology of hyaloid-retinal vessels was viewed with a confocalmicroscope in control and HG-treated larvae Scale bar = 40μm Hyaloid-retinal vessels in HG-induced larvae were signiﬁcantly thicker than control[(A) and (D)] The expression of ZO-1 showed irregular and discontinuous staining on hyaloid-retinal vessels from HG-treated larvae (E) compared tocontrol (B) To identify cells, Hoechst nuclear stain was used [(C) and (F)]

BJP

Trang 22

signal can be live imaged in larvae incubated with a DAF-FM DA

probe (Lepiller et al., 2007) In our system, , we exposed the control

and experimental groups of larvae at 6 dpf to DAF-FM DA and

observed the results using confocal microscopy We were able to

identify NO production in the region of the notochord, cleithrum,

and heart (data not shown) Interestingly, in hyaloid-retinal

vessels, NO production was induced in HG-treated larvae but

was not induced in control larvae (Figure 7A) Next, we used an

NO inhibitor,,L-NAME, (10 or 20μM) added to the HG treatmentfor 3 days and found that the hyaloid-retinal vessels from larvaetreated with HG andL-NAME showed decreased vessel diameter,similar to the control group (Figure 7B and C) We also analysedthe expression of mRNA for nos2a in the larvae As shown inFigure 7D and E, mRNA expression of nos2a was increased, relative

to control, in the larvae treated with HG and that this increase wassigniﬁcantly reduced by treatment withL-NAME

Figure 6

Vegf expression and the effect of the VEGFR inhibitor and the VEGF antibody ranibizumab in HG-treated larvae (A) The effect of the VEGFR inhibitor on lated hyaloid-retinal vessels Vessel dilation induced by HG was blocked by treatment with the VEGFR tyrosin kinase inhibitor (0.5 or 1μM) Scale bar = 40 μm.(B) The diameter of hyaloid-retinal vessels in HG-treated larvae with VEGFR inhibition was similar to control Graph data are displayed as the mean AU forvessel diameter ###P< 0.001 versus control; **P < 0.01, ***P < 0.001, signiﬁcantly different from HG, n = 8–10 embryos per group The vessel diameter

di-of each lens was measured three times [(C) and (D)] Expression di-of Vegf in HG-treated larvae increased signiﬁcantly compared with control and was reduced

by treatment with the VEGFR inhibitor No changes in the transcript levels ofβ-actin were detected between control and HG-treated larvae #P < 0.05,signiﬁcantly different from control; ***P < 0.001, signiﬁcantly different from HG (E) The effect of the VEGF antibody, ranibizumab, on dilated hyaloid-retinalvessels HG-induced vessel dilation was morphologically rescued by ranibizumab treatment (2.5μg·mL 1

) Scale bar = 40μm (F) The diameter of retinal vessels in HG-treated larvae with ranibizumab treatment was significantly reduced Graph data are displayed as the mean AU for vesseldiameter ###P< 0.001, significantly different from control; **P < 0.01, significantly different from HG, n = 8–10 embryos per group The vesseldiameter of each lens was measured three times

hyaloid-BJP S-H Jung et al.

Trang 23

Discussion and conclusions

During the last two decades, models of diabetes have been

devel-oped in many species, including pigs, dogs, cats, rats and mice,

using different inducing agents such as chemicals (STZ or alloxan)

or genetic modiﬁcations (Feit-Leichman et al., 2005; Robinson

et al., 2012) Studies of the molecular pathways involved in DR

have used STZ- or alloxan-induced diabetes models or transgenic

and knockout mouse models such as ob/ob and db/db (Midena

et al., 1989; Cheung et al., 2005; Zhang et al., 2007) The retinae

of STZ-induced diabetic animals exhibit biochemical and

physio-logical changes 1–2 months after the onset of hyperglycaemia

This model reproduces early symptoms of DR, such as thickening

of the vascular basement membrane, loss of retinal pericytes andcapillaries, vascular occlusion and increased vascular permeability(Robinson et al., 2012) A recent study suggested an animal modelfor non-proliferative DR in adult zebrafish, in which the zebrafishare subjected to 2% glucose (110 mM) immersion for 30 days(Alvarez et al., 2010) In this model, visual function is diminished,and cone photoreceptors are disrupted Hypoxia-induced retinalvascular disorder in the adult zebrafish model takes 11 days toshow retinopathy and, consequently, to assess therapeutic effects(Cao et al., 2010) In addition, these adult zebrafish models re-quire testing several drug compounds in the plastic containers,and isolation of the retina vessels is time-consuming, requiringsurgical microscissors and forceps resulting in physical stress

Figure 7

NO production by HG in hyaloid-retinal vessels (A) Comparison of NO production using a DAF-FA DA probe in hyaloid-retinal vessels in control larvae at 6 dpf

NO production is not observed in control larvae but is detected in HG-treated larvae (red asterisk) Scale bar = 50μm (B) The effect of the NO inhibitor ondilated hyaloid-retinal vessels HG-induced vessel dilation was rescued by the NOS inhibitor (L-NAME, 10 and 20μM) (C) Graph data are displayed as the mean

AU for vessel diameter The vessel diameter of each lens was measured three times Scale bar = 40μm ###P < 0.001 versus control; **P < 0.01, ***P < 0.001,significantly different from HG, n = 10 embryos per group [(D) and (E)] Expression of nos2a in HG-treated larvae increased significantly compared to controland reduced by treatment withL-NAME No changes in the transcript levels ofβ-actin were detected between control and HG-treated larvae ###P < 0.001,significantly different from control; *P < 0.05, **P < 0.01, significantly different from HG

BJP

Trang 24

Recently, Wang et al (2013) reported that combination of a

high cholesterol diet and immersion in a 3% glucose (166 mM)

solution applied to 5 to 15 dpf zebraﬁsh larvae could be used as

a model for diabetic vasculopathy Earlier studies showed that

1 dpf zebraﬁsh larvae could be exposed to 30 mM glucose and

drugs for to 6 days (Lee et al., 2013; Yu et al., 2013) In the present

study, we have attempted to optimize and characterize a

short-term in zebraﬁsh to study diabetic retinal vascular dysfunction

and provide a system for rapid drug screening We induced

hyperglycaemia by immersing 3 dpf zebraﬁsh larvae for 3 days

in 130 mM glucose (HG; Figures 2 and 3) Although there were

clear changes in the hyaloid-retinal vessels, the lens size and

number of OD branches were not affected (Figure S2) Moreover,

other bloods vessels such as ISVs and DA did not show any

differences between the control and HG groups (Figure S3)

In the present model, the hyaloid-retinal vessels were

analysed after only 3 days incubation in 24-well plates This

pro-cedure provides two advantages, a low incubation volume means

that less drug is needed per test and the 3 days is clearly much

shorter than 11 (Cao et al., 2010) or 30 days (Alvarez et al.,

2010), as used previously The structural changes in the

hyaloid-retinal vessels after HG were not osmotic events but were

inﬂu-enced by in vivo glucose metabolism in zebraﬁsh larvae Gleeson

et al (2007) and Wang et al (2013) reported that larvae treated

with HG regulate glucose metabolism through direct secretion

of insulin, phosphoenolpyruvate carboxykinase, somatostatin

and glucagon into the bloodstream These results suggest that

HG-induced changes in zebraﬁsh larvae hyaloid-retinal vessels

closely mimic the molecular mechanism for human DR

Reduction and redistribution of tight junction proteins in

retinal endothelial cells under hyperglycaemic conditions

increases vascular permeability (Antonetti et al., 1998) Kim

et al (2011) did not observe occludin and ZO-1 expression

on retinal vessels when viewed on whole mounts of frozen

sections with immunoﬂuorescence staining in zebraﬁsh

However, using our isolation method and quantitative

analy-ses (Figure 1 and S1), we were able to demonstrate expression

of the tight junction protein ZO-1 and show that its

expres-sion was irregular or discontinuous on hyaloid-retinal vessels

from HG-treated larvae (Figure 5)

At present, VEGF constitutes an exciting target for

thera-peutic intervention in DR VEGF-induced angiogenesis and

vascular hyperpermeability are mediated by NO (Van der Zee

et al., 1997) Also NO was an essential downstream

compo-nent of VEGF-induced angiogenesis in vivo and NOS

inhibi-tors blocked the VEGF-induced proliferation of endothelial

cells (Shizukuda et al., 1999) In the present experiments, we

have conﬁrmed that this model was sensitive to VEGF-related

and NO-related interventions Although early (0–1 dpf)

expo-sure to a VEGFR inhibitor induced severe developmental

defects, including heart and eye oedema, short body and

de-velopmental retardation (data not shown), later application,

at 3 to 6 dpf, blocked the increase in diameter of retinal

vessels, induced by HG treatment (Figure 6A and B)

As 75% of the protein sequence in zebraﬁsh VEGFAA is

similar to that of human VEGF-A 165 isoform (Wu et al., 2015)

and the recombinant humanized monoclonal IgG antibody

for VEGF, ranibizumab, has been used to treat human DR by

blocking angiogenesis, it was reasonable to test ranibizumab in

our model We also assessed NO production with aﬂuorescent

indicator dye (DAF-FA DA) and the expression of NOS mRNA,

in our model These results and those after application of theNOS inhibitor L-NAME together demonstrated clearly that

NO was an important mediator of the structural changes inthe hyaloid-retinal vessels of zebraﬁsh larvae following HGtreatment Overall, our results suggest that NO and VEGF inHG-treated zebraﬁsh larvae display regulatory patterns andactivities, very similar to those in mammals

Although several animal models for DR are well established,there are no short-term in vivo models to test the effects of can-didate therapeutic compounds In this study, we have described

a novel, short-term, in vivo screening method for compoundsaffecting DR, using zebraﬁsh larvae exposed to HG conditions

In this model, dilation of hyaloid-retinal vessels was nied by morphological lesions with disruption of tight junctionproteins, overproduction of Vegf mRNA and NO production Wealso showed that the VEGFR tyrosine kinase inhibitor, the VEGFantibody ranibizumab and inhibition of endogenous NO syn-thesis could all decrease or reverse the HG-induced dilation ofhyaloid-retinal vessels These results suggest that this zebraﬁshmodel system will be a powerful novel tool for the screening

accompa-of therapeutic drug candidates for DR

Con ﬂict of interest

The authors declare they have no conﬂict of interest

ReferencesAlexander SPH, Benson HE, Faccenda E, Pawson AJ, Sharman JL,Spedding M et al (2013a) The Concise Guide to PHARMACOLOGY2013/14: Catalytic Receptors Br J Pharmacol 170: 1676–1705.Alexander SPH, Benson HE, Faccenda E, Pawson AJ, Sharman JL,Spedding M et al (2013b) The Concise Guide to PHARMACOLOGY2013/14: Enzymes Br J Pharmacol 170: 1797–1867

Alvarez Y, Cederlund ML, Cottell DC, Bill BR, Ekker SC, Torres-Vazquez J

et al (2007) Genetic determinants of hyaloid and retinal vasculature inzebraﬁsh BMC Dev Biol 7: 114–130

Alvarez Y, Chen K, Reynolds AL, Waghorne N, O’Connor JJ, Kennedy

BN (2010) Predominant cone photoreceptor dysfunction in ahyperglycaemic model of non-proliferative diabetic retinopathy DisModel Mech 3: 236–245

Trang 25

Abu El-Asrar AM, Meersschaert A, Dralands L, Geboes K (2004) Inducible

nitric oxide synthase and vascular endothelial growth factor are colocalized

in the retinas of human subjects with diabetes Eye 18: 306–313

Antonetti DA, Barber AJ, Khin S, Lieth E, Tarbell JM, Gardner TW

(1998) Vascular permeability in experimental diabetes is associated

with reduced endothelial occludin content: vascular endothelial

growth factor decreases occludin in retinal endothelial cells Penn

State Retina Res Group Diab 47: 1953–1959

Armulik A, Abramsson A, Betsholtz C (2005) Endothelial/pericyte

interactions Circ Res 97: 512–523

Barros TP, Alderton WK, Reynolds HM, Roach AG, Berghmans S (2008)

Zebraﬁsh: an emerging technology for in vivo pharmacological

assessment to identify potential safety liabilities in early drug discovery

Br J Pharmacol 154: 1400–1413

Bergers G, Song S (2005) The role of pericytes in blood-vessel

formation and maintenance Neuro Oncol 7: 452–464

Calcutt NA, Cooper ME, Kern TS, Schmidt AM (2009) Therapies for

hyperglycaemia-induced diabetic complications: from animal

models to clinical trials Nat Rev Drug Discov 8: 417–429

Caldwell RB, Bartoli M, Behzadian MA, EI-Remessy AE, AI-Shabrawey

M, Platt DH et al (2003) Vascular endothelial growth factor and

diabetic retinopathy: pathophysiological mechanisms and treatment

perspectives Diabetes Metab Res Rev 19: 442–455

Cao Z, Jensen LD, Rouhi P, Hosaka K, Lanne T, Steffensen JF et al

(2010) Hypoxia-induced retinopathy model in adult zebraﬁsh Nat

Protoc 5: 1903–1910

Checchin D, Sennlaub F, Levavasseur E, Leduc M, Chemtob S (2006)

Potential role of microglia in retinal blood vessel formation Invest

Ophthalmol Vis Sci 47: 3595–3602

Cheung AK, Fung MK, Lo AC, Lam TT, So KF, Chung SS et al (2005)

Aldose reductase deﬁciency prevents diabetes-induced blood-retinal

barrier breakdown, apoptosis, and glial reactivation in the retina of

db/db mice Diabetes 54: 3119–3125

Elo B, Villano CM, Govorko D, White LA (2007) Larval zebraﬁsh as a

model for glucose metabolism: expression of phosphoenolpyruvate

carboxykinase as a marker for exposure to anti-diabetic compounds

J Mol Endocrinol 38: 433–440

Engerman RL, Kern TS (1995) Retinopathy in animal models of

diabetes Diabetes Metab Rev 11: 109–120

Feit-Leichman RA, Kinouchi R, Takeda M, Fan Z, Mohr S, Kern TS et al

(2005) Vascular damage in a mouse model of diabetic retinopathy: relation

to neuronal and glial changes Invest Ophthalmol Vis Sci 46: 4281–4287

Fong DS, Aiello L, Gardner TW, King GL, Blankenship G, Cavallerano

JD et al (2003) Diabetic retinopathy Diabetes Care 26: 226–229

Francesco C, Keiichi H, Zvonimir SK, Thomas FL (1997) High glucose

increases nitric oxide synthase expression and superoxide anion

generation in human aortic endothelial cells Circulation 96: 25–28

Geoffrey BA, Sobha S (2011) Hypoxia and oxidative stress in the

causation of diabetic retinopathy Curr Diabetes Rev 7: 291–304

Gleeson M, Connaughton V, Arneson LS (2007) Induction of

hyperglycaemia in zebraﬁsh (Danio rerio) leads to morphological

changes in the retina Acta Diabetol 44: 157–163

Greene DA, Arezzo JC, Brown MB (1999) Effect of aldose reductase

inhibition on nerve conduction and morphometry in diabetic

neuropathy Zenarestat Study Group Neurology 53: 580–591

Guillemin GJ, Brew BJ (2004) Microglia, macrophages, perivascular

macrophages, and pericytes: a review of function and identiﬁcation

J Leukoc Biol 75: 388–397

Jorgens K, Hillebrands JL, Hammes HP, Kroll J (2012) Zebraﬁsh: amodel for understanding diabetic complications Exp ClinEndocrinol Diabetes 120: 186–187

Joussen AM, Poulaki V, Qin W, Kirchhof B, Mitsiades N, Wiegand SJ

et al (2002) Retinal vascular endothelial growth factor inducesintercellular adhesion molecule-1 and endothelial nitric oxidesynthase expression and initiates early diabetic retinal leukocyteadhesion in vivo Am J Pathol 2: 501–509

Jung SH, Kim S, Chung AY, Kim HT, So JH, Ryu J et al (2010) Visualization

of myelination in GFP-transgenic zebraﬁsh Dev Dyn 2: 592–597.Kempen JH, O’Colmain BJ, Leske MC, Haffner SM, Klein R, Moss SE et al.(2004) The prevalence of diabetic retinopathy among adults in theUnited States Arch Ophthalmol 122: 552–563

Kilkenny C, Browne WJ, Cuthill IC, Emerson M, Altman DG (2010).Animal research: Reporting in vivo experiments: The ARRIVEguidlines Br J Pharmacol 160: 1577–1579

Kim JH, Yu YS, Kim KW, Kim JH (2011) Investigation of barriercharacteristics in the hyaloid-retinal vessel of zebraﬁsh J NeurosciRes 89: 921–928

Kitambi SS, McCulloch KJ, Peterson RT, Malicki JJ (2009) Smallmolecule screen for compounds that affect vascular development inthe zebraﬁsh retina Mech Dev 126: 464–477

Lakshminarayanan S, Antonetti DA, Gardner TW, Tarbell JM (2000).Effect of VEGF on retinal microvascular endothelial hydraulicconductivity: the role of NO Invest Ophthalmol Vis Sci 41: 4256–4261.Lee IS, Yu SY, Jung SH, Lee YR, Lee YM, Kim JH et al (2013)

Proanthocyanidins from Spenceria ramalana and their effects on AGEformation in vitro and hyaloid-retinal vessel dilation in larvalzebraﬁsh in vivo J Nat Prod 76: 1881–1888

Lepiller S, Laurens V, Bouchot A, Herbomel P, Solary E, Chluba J (2007).Imaging of nitric oxide in a living vertebrate using a diamino-ﬂuoresceinprobe Free Radic Biol Med 43: 619–627

Liang J, Gui Y, Wang W, Gao S, Li J, Song H (2010) Elevated glucoseinduces congenital heart defects by altering the expression of tbx5,tbx20, and has2 in developing zebraﬁsh embryos Birth Defects Res AClin Mol Teratol 88: 480–486

McGrath JC, Drummond GB, McLachlan EM, KilKenny C,Wainwright CL (2010) Guidelins for reporteing experimentsinvolving animals: the ARRIVE guidlines Br J Pharmacol 160:

1573–1576

Midena E, Segato T, Radin S, di Giorgio G, Meneghini F, Piermarocchi S

et al (1989) Studies on the retina of the diabetic db/db mouse I.Endothelial cell-pericyte ratio Ophthalmic Res 21: 106–111

Morbidelli L, Chang CH, Douglas JG, Granger HJ, Ledda F, Ziche M(1996) Nitric oxide mediates mitogenic effect of VEGF on coronaryvascular endothelium Am J Physiol 270: H411–415

Nusslein VC, Dahm R (eds) (2002) Zebraﬁsh: A Practical Approach.Oxford University Press: Oxford, UK

Pawson AJ, Sharman JL, Benson HE, Faccenda E, Alexander SP,Buneman OP et al (2014) The IUPHAR/BPS Guide toPHARMACOLOGY: an expert-driven knowledge base of drug targetsand their ligands Nucl Acids Res 42 (Database Issue): D1098–1106.Pﬁster F, Feng Y, vom Hagen F, Hoffmann S, Molema G, Hillebrands JL

et al (2008) Pericyte migration: a novel mechanism of pericyte loss

in experimental diabetic retinopathy Diabetes 57: 2495–2502.Robinson R, Barathi VA, Chaurasia SS, Wong TY, Kern TS (2012) Update

on animal models of diabetic retinopathy: from molecular approaches tomice and higher mammals Dis Model Mech 5: 444–456

BJP

Trang 26

Roy MS, Klein R, O’Colmain BJ, Klein BE, Moss SE, Kempen JH

(2004) The prevalence of diabetic retinopathy among adult type 1

diabetic persons in the United States Arch Ophthalmol 122:

546–551

Scheppke L, Aguilar E, Gariano RF, Jacobson R, Hood J, Doukas J et al

(2008) Retinal vascular permeability suppression by topical

application of a novel VEGFR2/Src kinase inhibitor in mice and

rabbits J Clin Invest 118: 2337–2346

Sheetz MJ, King GL (2002) Molecular understanding of

hyperglycemia’s adverse effects for diabetic complications JAMA

288: 2579–2588

Shizukuda Y, Tang S, Yokota R, Ware JA (1999) Vascular endothelial

growth factor-induced endothelial cell migration and proliferation

depend on a nitric oxide-mediated decrease in protein kinase Cδactivity

Circ Res 85: 247–256

Starita C, Patel M, Katz B, Adamis AP (2007) Vascular endothelial

growth factor and the potential therapeutic use of pegaptanib

(macugen) in diabetic retinopathy Dev Ophthalmol 39: 122–148

Van der Zee R, Murohara T, Luo Z, Zollmann F, Passeri J, LeKutat C et al

(1997) Vascular endothelial growth factor/vascular permeability

factor augments nitric oxide release from quiescent rabbit and

human vascular endothelium Circulation 95: 1030–1037

Wang Z, Mao Y, Cui T, Tang D, Wang XL (2013) Impact of a combined

high cholesterol diet and high glucose environment on vasculature

PLoS One 12: e81485

Wu YC, Chang CY, Kao A, Hsi B, Lee SH, Chen YH et al (2015)

Hypoxia-induced retinal neovascularization in zebraﬁsh embryos: a potential

model of retinopathy of prematurity PLoS One 15: e0126750

Yu SY, Lee IS, Jung SH, Lee YM, Lee YR, Kim JH et al (2013)

Caffeoylated phenylpropanoid glycosides from Brandisia hancei

inhibit advanced glycation end product formation and aldose

reductase in vitro and vessel dilation in larval zebraﬁsh in vivo Planta

Med 79: 1705–1709

Zhang H, Sonoda KH, Qiao H, Oshima T, Hisatomi T, Ishibashi T (2007).Development of a new mouse model of branch retinal vein occlusionand retinal neovascularization Jpn J Ophthalmol 51: 251–257

Supporting Information

Additional Supporting Information may be found in the onlineversion of this article at the publisher’s web-site:

http://dx.doi.org/10.1111/bph.13279Figure S1Diagram depicting measurement of the hyaloid-retinal vessels diameter (A) Lens in a dish with a coverglass.(B) Lens alignments under theﬂuorescence microscope Theoptic disc (OD) was displayed up (C) Fluorescence micros-copy images (D) The diameter of hyaloid vessels was mea-sured in three main branches (red arrow) from the OD intotheﬁrst branch (~30 μm) using Image J software

Figure S2Morphological effects on the patterning of thehyaloid vessels and lens size by HG (A) The size of lenses werenormal in control and HG-treated larvae Control larvae,n=38; HG-treated larvae, n=40 Scale bar = 40μm (B) Thenumber of main branches from the optic disc (OD) wascounted in control and glucose-treated larvae Control larvae,n=76; HGtreated larvae, n=72 Scale bar = 40μm

Figure S3Repersentative images of hyaloid-retinal vessels andtrunk vessels in 6 dpf zebrafish (A,A`,B,B`) There were no differ-ences in intersomite vessels (ISVs) and dorsal aorta (DA) betweenthe control (A, A`) and HG-treated groups (B, B`) 250´ magnifica-tion, n=4 in each group Scale bar = 50μm (C-E) The graph displaysthe mean artificial unit (AU) for diameter of ISVs and DA Thevessel diameter of each region was measured three times Theexperiment was repeated triplicate *P< 0.05 vs control

Trang 27

RESEARCH PAPER

human GIP receptors, but a

partial agonist and

competitive antagonist at rat

and mouse GIP receptors

A H Sparre-Ulrich1,2†, L S Hansen1,2,4†, B Svendsen2, M Christensen4,

F K Knop4, B Hartmann2,3, J J Holst2,3and M M Rosenkilde1

1Department of Neuroscience and Pharmacology, Faculty of Health and Medical Sciences, The

Panum Institute, University of Copenhagen, Copenhagen, Denmark,2NNF Center for Basic

Metabolic Research, Copenhagen, Denmark,3Department of Biomedical Sciences Faculty of Health

and Medical Sciences, University of Copenhagen, Copenhagen, Denmark, and4Center for Diabetes

Research, Gentofte Hospital, University of Copenhagen, Copenhagen, Denmark

Speciﬁc, high potency receptor antagonists are valuable tools when evaluating animal and human physiology Within theglucose-dependent, insulinotropic polypeptide (GIP) system, considerable attention has been given to the presumed GIP re-ceptor antagonist, (Pro3)GIP, and its effect in murine studies We conducted a pharmacological analysis of this ligand includinginterspecies differences between the rodent and human GIP system

EXPERIMENTAL APPROACH

Transiently transfected COS-7 cells were assessed for cAMP accumulation upon ligand stimulation and assayed in competitionbinding using125I-human GIP Using isolated perfused pancreata both from wild type and GIP receptor-deﬁcient rodents, insulin-releasing, glucagon-releasing and somatostatin-releasing properties in response to species-speciﬁc GIP and (Pro3)GIP analogueswere evaluated

CONCLUSIONS AND IMPLICATIONS

When evaluating new compounds, it is important to consider interspecies differences both at the receptor and ligand level Thus,

in rodent models, human GIP is a comparatively weak partial agonist Human (Pro3)GIP was not an antagonist at human GIPreceptors, so there is still a need for a potent antagonist in order to elucidate the physiology of human GIP

Abbreviations

GIP, glucose-dependent insulinotropic polypeptide (gastric inhibitory peptide); GLP-1, glucagon-like peptide-1; HBS,HEPES-buffered saline

published by John Wiley & Sons Ltd on behalf of British Pharmacological Society.

British Journal of Pharmacology (2016) 173 27–38 27

Trang 28

Glucose-dependent insulinotropic polypeptide (also gastric

inhibitory peptide; GIP) is a hormone secreted from the K

cells of the gut following a meal (Baggio and Drucker,

2007) Like the related hormone glucagon-like peptide-1

(GLP-1), GIP is a potent insulin secretagogue (Holst, 2004)

In contrast to the glucagonostatic effect of GLP-1 (Gutniak

et al., 1992; de Heer et al., 2008), GIP has glucagon-releasing

properties (Kreymann et al., 1987; Christensen et al., 2011)

The incentive to understand the biology of GIP was

intensi-ﬁed by the discovery of an association between rodent GIP

receptors and adiposity (Miyawaki et al., 1999; Miyawaki

et al., 2002; Nasteska et al., 2014) In humans, although less

clear, there is likewise evidence for a role of GIP in fat

metab-olism with the demonstration of GIP receptor expression in

adipose tissue (Ahlqvist et al., 2013), an association between

high body mass index and increased GIP levels (Ahlqvist

et al., 2013; Calanna et al., 2013), and increased adipose

tis-sue bloodﬂow and triacylglycerol deposition following GIP

administration in a state of high insulin and high glucose

(Asmar et al., 2010) Furthermore, obese children decrease

their basal and postprandial GIP levels, when put on a

low-calorie diet (Deschamps et al., 1980), and healthy young

men increase their fasting GIP levels, when put on a

high-fat diet (Brøns et al., 2009)

The availability of the GLP-1 receptor antagonist

exendin-(9–39) (Raufman et al., 1991; Jørgensen et al., 2013)

has been helpful for our understanding of the physiology of

GLP-1, which underlies the pharmaceutical development of

GLP-1 receptor agonists as anti-diabetic and anti-obesity

agents Inspired by this, researchers have tried to develop a

potent GIP receptor antagonist Many different strategies

have been undertaken in order to inhibit GIP function,

including prepro-GIP gene truncations (Nasteska et al.,

2014), administration of low MW receptor antagonists

(Nakamura et al., 2012), immunization against GIP (Ebert

et al., 1979; Fulurija et al., 2008; Irwin et al., 2009), various

truncations and amino acid substitutions of the GIP molecule

thought to provide antagonistic properties (Tseng et al., 1996;

Gelling et al., 1997; Hinke et al., 2001; Gault et al., 2002;

Deacon et al., 2006; Irwin et al., 2006; Kerr et al., 2011), and

recently, a potent antagonistic antibody against the GIP

receptor was reported (Ravn et al., 2013) However, most

attention has been given to the analogue (Pro3)GIP

follow-ing demonstrations that chronic treatment with this

peptide improved diabetic parameters in ob/ob mice (Irwin

et al., 2007), improved diabetic parameters and inducedweight loss in obese mice previously on a high-fat diet(McClean, 2007), and reduced weight gain and improveddiabetic parameters following the induction of a high-fatdiet (Gault et al., 2007)

Animal models are widely used as basis for predicting logical properties of given receptor-ligand systems to human(patho)physiology Often, rodent models are theﬁrst choicedue to reduced costs and space requirements compared withlarger animal models (pigs, dogs and primates) Initial drugscreening are similarly often carried out in rodent models.This creates the classical dilemma of whether the chosenrodent model is representative for human physiology orwhether the inter-species differences (stemming from struc-tural, functional, spatial or temporal differences) are too great.Among the seven-transmembrane receptor systems, such dif-ferences have been described for instance for the lipid-activated G protein-coupled receptor 119 (GPR119), where aseries of compounds, with similar binding properties to hu-man and mouse GPR119, displayed large differences in po-tency and efﬁcacy (Scott et al., 2013) A similar example can

bio-be found within the 5-HT receptor system, where a singleamino acid difference between the human and mouse recep-tor can account for large differences in binding afﬁnities andpotencies (Canal et al., 2013)

Other receptor-ligand systems, however, display very tle interspecies variations, as seen in the GLP-1 system with

lit-a llit-arge degree of sequence homology both in terms of theligand (GLP-1) and its receptor in humans, mice and ratsand where both liraglutide (a long-acting GLP-1 receptor ago-nist) and native GLP-1 have similar potencies across species-speciﬁc receptors (Knudsen et al., 2012) (Figure 1) Eventhough the GIP system is closely related to GLP-1 and theGLP-1 receptor, the GIP system is noticeably less conservedacross species (Figure 1)

In an attempt to develop GIP receptor antagonists to beused in humans, for instance, as a putative therapeuticagainst obesity (Asmar et al., 2010; Ahlqvist et al., 2013), weset out to investigate functional differences between rodentsand human within the GIP system in terms of ligand bindingand signalling properties As agonists, we chose the endoge-nous agonist GIP(1–42) from human, rat and mouse, andcharacterized this in parallel with the species-correspondingpreviously described antagonist (Pro3)GIP (Gault et al.,2002) on all three GIP receptors (from human, rat and

These Tables list key protein targets and ligands in this article which are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Pawsonet al., 2014)and are permanently archived in the Concise Guide to PHARMACOLOGY 2013/14 (Alexander et al., 2013)

BJP A H Sparre-Ulrich et al.

28 British Journal of Pharmacology (2016) 173 27–38

Trang 29

mouse) The goal was to determine whether the beneﬁcial

an-tagonistic effects of (Pro3)GIP in rodents could possibly be

transferred to humans

Methods

Animals

All animal care and experimental procedures complied with

institutional guidelines and were approved by the Danish

Animal Experiments Inspectorate (2013-15-2934-00833)

Studies involving animals are reported in accordance with

the ARRIVE guidelines for reporting experiments involving

animals (Kilkenny et al., 2010; McGrath et al., 2010) A total

of 18 animals were used in the experiments described here

Male C57Black/6 (B6) mice (25–30 g) or male Wistar rats

(220–250 g) were purchased from Taconic (Lille Skensved,

Denmark) The animals were housed in plastic-bottomed,

wire-lidded cages in air-conditioned (21C) and

humidity-controlled (55%) rooms with a 12:12 h light–dark cycle

and free access to standard rat chow and water Animals

were acclimatized for at least 1 week before use

Transfections and tissue culture

COS-7 cells (ATCC, Virginia, USA) were cultured at 10% CO2

and 37 °C in DMEM 1885 supplemented with 10% FBS,

2 mM glutamine, 180 units mL 1 penicillin and 45 g mL 1

streptomycin Transient transfection of the COS-7 cells for

cAMP accumulation and competition binding was performed

using the calcium phosphate precipitation method with the

addition of chloroquine (Kissow et al., 2012)

cAMP assay

Transiently transfected COS-7 cells were seeded out in white

96-well plates at a density of 3 * 104cells per well The day after, the

cells were washed twice with HEPES-buffered saline (HBS) buffer

and incubated with HBS and 1 mM IBMX for 30 min at 37 °C To

test agonists, ligands were added and incubated for 30 min at

37 °C In order to test for antagonistic properties, the antagonistwas pre-incubated for 10 min, and then the agonist was addedand incubated for an additional 20 min The HitHunterTMcAMP XS assay (DiscoveRx, Herlev, Denmark) was carried out ac-cording to the manufacturer’s instructions

Competition binding assayCOS-7 cells were seeded in clear 96-well plates the day aftertransient transfection The number of cells seeded per well,which is determined by the apparent expression efﬁciency

of the receptor, was aimed to result in 5–10% speciﬁc binding

of the added radioactive ligand (125I-human GIP) The ing day, cells were assayed by competition binding for 4 h at

follow-4 °C using 15–40 pM125

I-human GIP as well as unlabelledligand in 50 mM HEPES buffer (pH 7.2) supplemented with0.5% BSA (binding buffer) After incubation, the cells werewashed twice in ice-cold binding buffer and lysed using

200 mM NaOH with 1% SDS for 30 min Nonspeciﬁc bindingwas determined as the binding in the presence of 100 nMunlabelled human GIP The samples were analysed by theWallac Wizard 1470 Gamma Counter

Rat or mouse isolated perfused pancreasMale C57B6 mice (25–30 g) or male Wistar rats (220–250 g) wereanaesthetized (mice: ketamine/xylazine 100/10 mgkg 1

i.p.;rats: 0.0158 mg fentanyl citrate + 0.5 mgfluanisone + 0.25 mgmidazolam/100 g; Pharmacy Service, Denmark), and the pan-creas was dissected and perfused in situ as described previously(Deacon et al., 2006) Briefly, the pancreas was perfused in asingle-pass system through both the coeliac and the superiormesenteric artery via a catheter inserted into the aorta All otheraortic branches were ligated The venous effluent was collectedfor 1 min intervals via a catheter in the portal vein, and stored

at 20 °C until analysis The pancreas was perfused with amodiﬁed Krebs Ringer bicarbonate buffer [composition:

fu-marate, 5 mM pyruvate, 5 mM glutamate, 7 mM glucose,

-1 mouse

Figure 1

The GIP system is less conserved than the GLP-1 hormone system A) Sequence identity between rodent and human GIP receptors (GIPR) and GLP-1receptors (GLP-1R), as well as GIP and GLP-1 given in percent The receptors were compared without signalling peptide sequences The alignment wasdone in Geneious 6.0.5 B) Sequence alignment between human, rat, and mouse GIP (top) and GLP-1 (bottom) Grey indicates a mismatch and the blackspiral indicates predicted alpha helix structure

Species-speciﬁc activity of (Pro3)GIP

BJP

Trang 30

0.1% BSA, 5% dextran (Pharmacosmos, Holbaek, Denmark)].

Flow rate was kept constant at 1.5 mLmin 1

(mouse) or

4 mLmin 1

(rat), perfusion buffer was heated and

oxygen-ated (95% O2, 5% CO2), and pressure was continuously

measured throughout the experiment Rodent (rat/mouse)

forms of (Pro3)GIP and of GIP were infused as test

sub-stances through a sidearm infusion pump at a ﬂow rate of

(rat) Arginine(10 mM) was infused at the end of each experiment as a

positive control

Hormone analysis

Hormone concentrations in the perfusion efﬂuent were

measured using in-house radioimmunoassays Glucagon

was measured using a midregion-directed glucagon

rum (code no 4304) or a COOH-terminally directed

antise-rum (code no 4305), which measures fully processed

glucagon as well as N-terminally extended molecular forms, in

addition, a synthetic glucagon for standards, and 125I-labelled

glucagon (a gift from Novo Nordisk A/S) as tracer (Orskov

et al., 1991) Insulin was measured using an antibody

cross-reacting strongly with rat insulin I and II (code no 2006–3)

As the standard, we used human insulin, and the tracer was

125

I-labelled human insulin (gift from Novo Nordisk A/S)

(Brand et al., 1995) Somatostatin concentrations were

deter-mined using a rabbit antiserum (code no 1758) raised against

synthetic cyclic somatostatin, recognizing both somatostatin

14 and somatostatin 28 (Baldissera et al., 1983), somatostatin

14 as standard and 125I-labelled Tyr11-somatostatin (NEX389,

Perkin-Elmer, Skovlunde, Denmark) as tracer

Sequence alignments

The amino acid sequences of the GIP and GLP-1 systems were

acquired from GenBank of NCBI The alignment was

per-formed in Geneious 6.0.5 using MAFFT v6.814b The

BLOSUM62 matrix was applied with gap open penalty and

offset value of 1.53 and 0.123 respectively

Data analysis

Pharmacological analyses including Schild plot analyses and

statistical analyses were carried out with the GraphPad Prism

6.0 software (GraphPad, San Diego, CA, USA) and Microsoft

ExcelTM Sigmoid curves wereﬁtted logistically with a Hillslope

of 1.0 The calculations of Kivalues were based on the

Cheng-Prusoff formula (DeBlasi et al., 1989) Dose ratios (DR) for the

Schild analyses (Lazareno and Birdsall, 1993) were based on

the potency shift of GIP(1–42) in the presence of a given

antag-onist concentration, relative to the absence of antagantag-onist

Materials

Rat and mouse (Pro3)GIP (cat no 027–29 and 027–49,

re-spectively) and rat and mouse GIP (cat no 027–12 and

027–27, respectively) were purchased from Phoenix

Phar-maceuticals (Karlsruhe, Germany) Human (Pro3)GIP was

synthesized by CASLO ApS (Lyngby, Denmark) and human

GIP was purchased from Bachem (H5645: Bubendorf,

Switzerland) All peptides had a purity of more than 95%

by HPLC analyses and an MS-controlled molecular weight

purchased from Origene (Rockville, MD, USA) (SC110906,RN212314 and MC216211, respectively) and cloned into

purchased from PerkinElmer Life Sciences (Skovlunde,Denmark; NEX402)

(Pro3)GIP is an agonist with efficacies dependent on the species providing GIP receptors

Due to the reported antagonistic properties and the effects served in obese and/or diabetic animal models (Gault et al.,2007; Irwin et al., 2007; McClean, 2007), the Pro3-analogues wereinvestigated on the human, rat and mouse GIP receptors, in par-allel with the respective endogenous GIP agonists As GIPreceptor-induced cAMP accumulation is well established as an im-portant signalling pathway for GIP function, the effects ofspecies-specific GIP and (Pro3)GIP analogues were evaluated bymeasuring cAMP accumulation Compared with human GIP, hu-man (Pro3)GIP was an efficacious agonist on the human GIP re-ceptor approaching the Emaxof human GIP (Figure 2A), whileboth rat (Pro3)GIP (Figure 2B) and mouse (Pro3)GIP (Figure 2C)were partial agonists with Emaxvalues of ~50 and ~30%, respec-tively, on their respective receptors In addition to the differ-ences in efficacy, there was a dramatic shift in potencybetween GIP and the corresponding (Pro3)GIP within the threespecies On the human GIP receptor, we observed 181-fold dif-ference between human GIP and human (Pro3)GIP, while muchlarger differences (1300-fold and 2636-fold in rat and mouse, re-spectively) were observed in the rodent systems These differ-ences were due to both higher species-specific GIP potencies(from 26 pM for human GIP on human GIP receptors to 10and 11 pM for rat and mouse GIP on their GIP receptors) andlower (Pro3)GIP species-specific potencies (from 4.7 nM on hu-man GIP receptors to 13 and 29 nM on rat and mouse GIP recep-tors respectively)

ob-(Pro3)GIP is a partial agonist with competitive antagonistic properties in rodent GIP receptors

To determine the potential of the two partial agonists (rat andmouse (Pro3)GIP) as antagonists of GIP-induced activation,

Trang 31

cAMP production was measured as a function of increasing

concentration of rat and mouse GIP in the absence or

pres-ence of various concentrations of rat and mouse (Pro3)GIP

on the corresponding GIP receptors (Figure 3A and B

respectively) Reﬂected by the agonistic properties, (Pro3)

GIP increased the cAMP production in the absence or at low

GIP concentrations on both receptors However, a

concomi-tant rightward shift in potency of GIP was observed, which

is an indication of a competitive antagonistic nature At 10,

100, and 1000 nM of rat (Pro3)GIP, the potency of rat GIP

was shifted 2-fold, 4-fold and 16-fold compared with the

absence of (Pro3)GIP (Figure 3A) Using the calculated EC50

values from these curves, a Schild plot analysis was made

(Figure 3C) The Hill coefﬁcient was 0.55 ± 0.20, and the

X-axis intercept, which represents the dissociation

con-stants (Ki) for rat (Pro3)GIP, was 13 nM A similar pattern

was observed in the mouse system Here, the highest

concen-tration of mouse (Pro3)GIP (1 uM) resulted in an 11-fold

right-shift in the potency of mouse GIP The Hill coefﬁcient

in the Schild Plot in the mouse system was 0.79 ± 0.16 and

Kiof rat (Pro3)GIP was calculated to be 61 nM (Figure 3D)

As the Hill coefﬁcient was <1 in both cases, this indicates

confounding factors, presumably determined by the

con-comitant (partial) agonist and antagonist effects of (Pro3)GIP

Lower activities of all three (Pro3)GIP in the

rodent GIP systems compared with the human

system

To determine whether the species-speciﬁc differences of

(Pro3)GIPs were due to ligand or receptor differences, we

investigated the effects of the three (Pro3)GIP ligands

across all three GIP receptors The approaching full

agonism as observed for human (Pro3)GIP at the human

GIP receptor was not seen with the two rodent (Pro3)GIP

li-gands, which both demonstrated an Emaxof ~70% as

com-pared with human GIP (Figure 4A) In contrast, the (Pro3)

GIP analogues were equally potent at the human GIP

re-ceptor (Table 1) At the rat GIP rere-ceptor, there was a lower

efﬁcacy of all three (Pro3)GIP analogues, with human

(Pro3)GIP again showing the highest efﬁcacy (67 vs

~40%), but with similar potencies for the three (Pro3)GIPanalogues (Figure 4B, Table 1) The same tendency was seen

at the mouse GIP receptor with an even lower efﬁcacy ofhuman (Pro3)GIP (41%) and rodent (Pro3)GIP (~30%) com-pared with mouse GIP (Figure 4C)

Higher activities of GIP in the rodent GIP system as compared with the human systemInspired by thefindings that the glutamic acid to proline sub-stitution at the third position in GIP resulted in a lower activ-ity in the rodent systems, we decided to examine whethersuch species-dependent differences were present for nativeGIP Interestingly, the opposite picture arose here, as thehuman GIP displayed a lower potency and efficacy on allthree receptors compared with the rodent GIP ligands(Figure 5A–C respectively) The rodent GIPs were equallypotent and efficacious on all tested receptors, without anyconsiderable change in potency between the receptors(EC50~ 10 pM) (Table 1) In contrast, human GIP had a lower

efﬁcacy on all three receptors with the largest difference inthe rodent receptors with an Emax of 75% on the rat GIPreceptor and 60% on mouse GIP receptors compared withthe corresponding rodent GIP Likewise, human GIP had alower potency than the rodent GIPs on all three receptorswith the largest difference being 23-fold on the mouse GIPreceptor followed by 10-fold and 5-fold on the rat and humanGIP receptors respectively

No inter-species differences in the binding affinities of GIP analogues or (Pro3)GIP analogues

In order to determine whether the differences at both the tor and the ligand level in terms of cAMP accumulation weredue to differences in binding afﬁnity, competition binding ex-periments were conducted using 125I-labelled human GIP asthe radioligand Human GIP was found to bind to all three

recep-Figure 2

Human (Pro3)GIP is a full agonist, rat and mouse (Pro3)GIP are partial agonists COS-7 cells were transiently transfected with either human GIPreceptors (GIPR; A), rat GIPR (B), or mouse GIPR (C) and assayed for cAMP accumulation following increasing concentrations of WT GIP and(Pro3)GIP from the corresponding species The data was normalised to Emaxof the endogenous GIP on every receptor and shown as mean

±SEM, N≥3 Nonlinear regression was used to calculate the EC50value and Emax

BJP

Trang 32

Figure 3

The rodent partial agonist peptides, (Pro3)GIP, are functional antagonists at their corresponding GIP receptors (GIPR) (A) rat GIP and (B) mouseGIP cAMP accumulation concentration–response curves with or without increasing concentrations (10, 100 and 1000nM) of rat (Pro3)GIP ormouse (Pro3)GIP The results were normalized to Emaxof the species-speciﬁc GIP in the absence of (Pro3)GIP and shown as mean + SEM, N = 3.Nonlinear regression was used to calculate EC50values Schild plot analysis of the dose–response curves of (C) rat GIP and (D) mouse GIP revealed

Ki-values of 13 and 61 nM respectively

Figure 4

Human (Pro3)GIP is a partial agonist on rodent receptors COS-7 cells were transiently transfected with either human GIPR (A), rat GIPR (B), ormouse GIPR (C) and assayed for cAMP accumulation following increasing concentrations of human, rat, and mouse (Pro3)GIP The data was nor-malised to Emaxof the species speciﬁc GIP on every receptor and shown as mean±SEM, N≥3 Nonlinear regression was used to calculate the EC50

value and Emax

Trang 33

receptors with equally high afﬁnities (Kdof 0.90 to 1.1 nM)

(Table 2) Likewise, both rodent GIP analogues were similar to

human GIP on the tested receptors with a Kiin the nM range

from 0.27 nM of rat GIP on the mouse GIP receptor to 2.0 nM

for mouse GIP on the human GIP receptor Thus, the functional

discrepancies between human and rodent GIP are not due to

ac-tual differences in afﬁnity The (Pro3)GIP analogues displayed in

average 10-fold lower afﬁnities than the native GIP analogues

on all three receptors with the lowest afﬁnity observed for the

human ligand on mouse GIP receptor (Kiof 39 nM) whereas

the rest displayed afﬁnities (Ki) between 4.8 and 14 nM Thus,

similar to the native GIP ligands, the functional differences

be-tween the human (Pro3)GIP and the rodent counterparts are

not caused by differences in binding afﬁnities

(Pro3)GIP stimulates insulin, glucagon and

somatostatin release

As most of the studies describing (Pro3)GIP were carried out

in rodent models (Gault et al., 2007; Irwin et al., 2007;

McClean, 2007), we determined how the receptor bindingand activation would translate into hormone secretion, usingperfused pancreata from rodents In brief, the tested ana-logues were perfused through the pancreata of the rodents,and the venous efﬂuent was collected at 1 min interval andanalysed for insulin, glucagon and somatostatin content Inboth perfusion models, we used 100-fold higher concentra-tions of (Pro3)GIP (100 nM), compared with that of GIP(1 nM), because of the 1000-fold difference in potencies(Table 1) and the 10-fold difference in afﬁnity (Table 2) Inthe mouse pancreas, (Pro3)GIP as well as GIP-induced secre-tion of insulin, glucagon and somatostatin (Figure 6A–C) asexpected from the agonistic nature of (Pro3)GIP (Figure 2C).However, compared with GIP, there were differences in themagnitude of secretion Although given at 100-fold higherconcentration, (Pro3)GIP did not reach the same levels ofinsulin secretion as GIP (Figure 6A), while there was no differ-ence in somatostatin and glucagon release (Figure 6B and C)

Table 1

cAMP accumulation induced by human, rat, and mouse GIP and (Pro3)GIP

Human GIPRlogEC50±SEM

EC50

(nM)

Emax±SEM (N)

Rat GIPRlogEC50±SEM

EC50

(nM)

Emax±SEM (N)

Mouse GIPRlogEC50±SEM

EC50

(nM)

Emax±SEM (N)

Figure 5

Human GIP is less efﬁcacious and has a lower potency than rodent GIP COS-7 cells were transiently transfected with either human GIP receptors(GIPR; A), rat GIPR (B), or mouse GIPR (C) and assayed for cAMP accumulation following increasing concentrations of human, rat and mouse GIP.The data was normalized to Emaxof the species speciﬁc GIP on every receptor and shown as mean±SEM, N≥3 Nonlinear regression was used tocalculate the EC50value and Emax

BJP

Trang 34

Similar results were obtained in the rat perfused pancreas

model with a signiﬁcantly lower insulin and somatostatin

secretion by (Pro3)GIP compared with GIP (Figure 6D and F,

respectively) but with an equal glucagon release by both

ligands (Figure 6E)

In order to conﬁrm that the observed hormone secretion

from perfused pancreata was mediated through the GIP

receptors, the same experimental set-up was repeated using

GIP receptor KO mice As seen in Figure 7, GIP

receptor-deﬁcient mice did not respond to mouse GIP or mouse

(Pro3)GIP with either insulin, glucagon or somatostatin

secretion (Figure 7A–C respectively) As with the pancreata

from wild-type mice and rats, the positive control, arginine,

induced secretion of the respective hormones, validatingthe experimental set-up

Discussion

Our study demonstrated that human (Pro3)GIP was not anantagonist at human GIP receptors, but rather approachesfull agonism At rodent GIP receptors, (Pro3)GIP was a par-tial agonist with competitive antagonistic properties whenevaluated in terms of cAMP accumulation In terms of its ef-fect on pancreatic hormone secretion, (Pro3)GIP followedthe partial agonistic pattern seen in cAMP with regard to in-

Figure 6

(Pro3)GIP is a agonist on rodent pancreata (A/D) insulin, (B/E) glucagon, and (C/F) somatostatin secretion from perfused mouse (A, B, C) and rat(D, E, F) pancreata following stimulation with either species speciﬁc 100 nM (Pro3)GIP, 1 nM GIP, or 10 mM arginine (N=6-9) The glucose con-centration was 7 mM and data are mean±SEM

Table 2

Similar displacement of125I-labeled hGIP by GIP analogues or (Pro3)GIP analogues

Human GIPRlogIC50± SEM Ki/Kd(nM) (N)

Rat GIPRlogIC50± SEM Ki/Kd(nM) (N)

Mouse GIPRlogIC50± SEM Ki/Kd(nM) (N)

Trang 35

(Pro3)GIP induced a glucagon release equal to that induced

by native GIP in rats

(Pro3)GIP is not an effective GIP receptor

antagonist for future studies of the GIP system

Theﬁrst GIP receptor antagonists that showed promise were

the truncated and amidated forms, GIP(6–30)-NH2and GIP

200 nM obtained in heterologous binding experiments

(Tseng et al., 1996; Gelling et al., 1997; Hinke et al., 2001)

However, elaborate in vivo characterization has not been

car-ried out The naturally occurring GIP metabolite GIP(3–42)

binds with similar afﬁnity (IC50of 22 nM); however, no tagonistic effect could be demonstrated in vivo in pigs(Deacon et al., 2006) Low MW compounds have also beenpresented, e.g the antagonist SKL-14959 with an afﬁnity (Kivalue) of 55 nM and the ability to increase blood glucoselevels and inhibit insulin secretion during an oral glucose tol-erance test in mice (Nakamura et al., 2012) Further develop-ment of this compound has not been published Recently, apromising GIP receptor antibody, Gipg013, with a Ki of6.8 nM was reported (Ravn et al., 2013) Based on the numer-ous rodent studies that displayed promising improvements inthe diabetic and/or obese state, (Pro3)GIP showed potential

an-toﬁll the role as a promising antagonist in the context of

Figure 7

The effects seen in perfused pancreata are mediated through GIP receptors (GIPR) A) insulin, B) glucagon, and C) somatostatin secretion ing stimulation of perfused pancreata from GIPR KO mice by either 100 nM mouse (Pro3)GIP, 1 nM mouse GIP, or 10 mM arginine (N=3) Theglucose concentration was 7 mM and data are mean±SEM

follow-Figure 8

The changes in potency and efﬁcacy of the tested GIP receptors (GIPR) and GIP analogues The calculated EC50(A/C) and Emax(B/D) values termined from the cAMP accumulation assays and normalised to human GIP on each species receptor (A, B, and D) or human (Pro3)GIP on eachspecies receptor

de-Species-speciﬁc activity of (Pro3)GIP

BJP

Trang 36

GIP physiology (Gault et al., 2007; Irwin et al., 2007;

McClean, 2007; De Toro-Martin et al., 2014) In ob/ob mice,

chronic treatment with (Pro3)GIP improved glucose

toler-ance and insulin sensitivity (Irwin et al., 2007), while

treat-ment in mice previously on a high-fat diet resulted in

weight loss, improved insulin sensitivity and glucose

toler-ance (McClean, 2007)

Consistent with previous rodent studies (Gault et al.,

2002; Gault et al., 2003; Gault Victor et al., 2007), weﬁnd

that rodent (Pro3)GIP ligands act as competitive

antago-nists on the rodent receptors (Figure 3) But when it comes

to the human GIP system, human (Pro3)GIP acted as a

efﬁ-cacious agonist (Figure 2A), which is in line with very

re-cent studies that reported substantial agonist activity of

human (Pro3)GIP with efﬁcacies up to 83% of human GIP

in cAMP release from transiently transfected HEK293 cells

(Ravn et al., 2013) or CHL cells (Pathak et al., 2015), and

in the reporter gene expression for cAMP-response element

(Al-Sabah et al., 2014) However, this contrasts to a previous

study demonstrating (Pro3)GIP to have 9% of GIP’s efﬁcacy

on transiently transfected Chinese hamster lung cells (CHL)

expressing the human GIP receptor (Gault Victor et al.,

2007) These efﬁcacy discrepancies may rely on differences

in cell types (CHL, HEK283 and COS-7 cells); however,

con-sistent for all studies is theﬁnding that (Pro3)GIP was not a

neutral antagonist, but has agonist properties in

cAMP-dependent pathways

Structural GIP divergence between species has

markedly affects the pharmacology

Our study emphasizes important interspecies variations

within the GIP system The rodent GIP receptorligands were

more potent and efﬁcacious than human GIP on all the tested

receptors (Figure 5), with up to 22-fold and 25-fold increase in

potency of rat GIP and mouse GIP on the mouse GIP

recep-tors (Figure 8A and B) These changes were not matched by

a similar increase in binding afﬁnities (Table 2) Only few

amino acids are altered among the three GIP receptor ligands

(Figure 1A) The most dramatic change is located at position

18 where human GIP has a histidine instead of an arginine,

which is found in rat and mouse GIP Among the few

non-identities observed (in positions 18, 30, 34 and 40), this is

the only case where the rodent sequences are identical and,

because of the non-conservative nature (Arg/His), a

signiﬁ-cant structural divergence is possible A previous study

sup-ports the functional importance of this position, as an

alanine substitution in human GIP increased insulin

secre-tion from BRIN-BD11 cells compared with native GIP

(Venneti et al., 2011) Thus, position 18 seems to have a role

in the activation mechanism of the GIP receptor, as also

sup-ported by a recent study predicting a direct interaction

be-tween His18 of human GIP and Ala32 of the N-terminus of

the GIP receptor (Tikhele et al., 2010) Thus, theseﬁndings

add some detail to the general view that hydrophobic

interac-tions between the C-terminal part of GIP and the N-terminus

receptor parts account for binding, while the N-terminus of

GIP interacts with other extracellular receptor domains

in-ducing activation (Malde et al., 2007; Parthier et al., 2007)

This so-called two-step receptor activation not only describes

GIP receptor activation but presents a global mechanism for

ligand-binding and receptor activation among class B1 tors (e.g glucagon, GLP-1, secretin, vasoactive intestinalpolypeptide and parathyroid hormone receptor) (Gourlet

recep-et al., 1996; Hjorth and Schwartz, 1996; Gardella and Jüppner,2001; Castro et al., 2005; Vilardaga et al., 2011) It also extendsbeyond class B1, as also some class A seven-transmembranereceptors (e.g chemokine, C3a and C5a receptors) are acti-vated by their endogenous peptide ligands in a two (ormulti)-step model involving receptor recognition and activa-tion by different regions in the ligands (Allen et al., 2007; Klos

et al., 2013; Thiele and Rosenkilde, 2014)

Taken together, the rodent GIP systems seem more activecompared with the human system, in terms of GIP (1–42)-mediated activation and more sensitive to GIP ligandmodiﬁcations, as judged by the lower activity of (Pro3)GIP.This is relevant and should be considered in studies using ro-dent models As our study indicates, when using human GIP

in rodent models, researchers are in fact injecting a less tent partial agonist (compared with rodent GIP) that maylead to underestimation of the true GIP activity in such exper-imental systems

po-The therapeutic potential of targeting the GIP receptor

In the wake of the GIP receptor KO studies demonstrating sistance to diet-induced obesity (Miyawaki et al., 2002), manyhave tried to antagonize the GIP receptors as a potential tar-get for treating obesity Accumulating evidence links GIP bi-ology to adiposity and due to the lack of an effective GIPreceptor antagonist, the therapeutic potential of antagoniz-ing this receptor in obese patients remains to be assessed Inthis context, our study brings attention to a signiﬁcant inter-species difference within the GIP system, and thus, pharma-cological evaluation of novel compounds in rodent modelsshould interpreted cautiously with respect to their relevancefor humans

re-Acknowledgements

This work was supported by the Novo Nordisk FoundationCenter for Basic Metabolic Research and the University ofCopenhagen

Con ﬂicts of interest

A H S U., L S H., B S., M C., B H and M M R declare thatthey have no conflict of interest J J H has served as a consul-tant or advisor to Novartis Pharmaceuticals, Novo Nordisk,Merck Sharp & Dohme and Roche and has received fees forlectures from Novo Nordisk, Merck Sharp & Dohme andGlaxoSmithKline F K K has received fees for consultancy,lectures or being part of an advisory board from AstraZeneca,Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly, Gilead,Merck Sharp & Dohme, Novo Nordisk, Sanofi and ZealandPharma, and has received research support from Sanofi andNovo Nordisk

Trang 37

Ahlqvist E, Osmark P, Kuulasmaa T, Pilgaard K, Omar B, Brøns C et al

(2013) Link between GIP and osteopontin in adipose tissue and

insulin resistance Diabetes 62: 2088–2094

Alexander SPH, Benson HE, Faccenda E, Pawson AJ, Sharman JL,

Spedding M et al (2013) The Concise Guide to PHARMACOLOGY

2013/14: G Protein-Coupled Receptors Br J Pharmacol 170:

1459–1581

Allen SJ, Crown SE, Handel TM (2007) Chemokine:receptor structure,

interactions, and antagonism Annu Rev Immunol 25: 787–820

Al-Sabah S, Al-Fulaij M, Ahmed HA (2014) Selectivity of peptide

ligands for the human incretin receptors expressed in HEK-293 cells

Eur J Pharmacol 741: 311–315

Asmar M, Simonsen L, Madsbad S, Stallknecht B, Holst JJ, Bülow J

(2010) Glucose-dependent insulinotropic polypeptide may enhance

fatty acid re-esteriﬁcation in subcutaneous abdominal adipose tissue

in lean humans Diabetes 59: 2160–2163

Baggio LL, Drucker DJ (2007) Biology of incretins: GLP-1 and GIP

Gastroenterology 132: 2131–2157

Baldissera FGA, Munoz-Perez MA, Holst JJ (1983) Somatostatin 1–28

circulates in human plasma Regul Pept 6: 63–69

Brand CL, Jorgensen PN, Knigge U, Warberg J, Svendsen I, Kristensen

JS et al (1995) Role of glucagon in maintenance of euglycemia in fed

and fasted rats Am J Physiol 269: E469–E477

Brøns C, Jensen CB, Storgaard H, Hiscock NJ, White A, Appel JS et al

(2009) Impact of short-term high-fat feeding on glucose and insulin

metabolism in young healthy men J Physiol 587: 2387–2397

Calanna S, Christensen M, Holst JJ, Laferrère B, Gluud LL, Vilsbøll T

et al (2013) Secretion of glucose-dependent insulinotropic

polypeptide in patients with type 2 diabetes: systematic review and

meta-analysis of clinical studies Diabetes Care 36: 3346–3352

Canal CE, Cordova-Sintjago T, Liu Y, Kim MS, Morgan D, Booth RG

(2013) Molecular pharmacology and ligand docking studies reveal a

single amino acid difference between mouse and human serotonin

5-HT2A receptors that impacts behavioral translation of novel

4-phenyl-2-dimethylaminotetralin ligands J Pharmacol Exp

Therapeut 347: 705–716

Castro M, Nikolaev VO, Palm D, Lohse MJ, Vilardaga J-P (2005) Turn-on

switch in parathyroid hormone receptor by a two-step parathyroid

hormone binding mechanism Proc Natl Acad Sci U S A 102: 16084–16089

Christensen M, Vedtofte L, Holst JJ, Vilsboell T, Knop FK (2011)

Glucose-Dependent insulinotropic polypeptide: a bifunctional

glucose-dependent regulator of glucagon and insulin secretion in

humans Diabetes 60: 3103–3109

De Toro-Martin J, Fernandez-Millan E, Lizarraga-Mollinedo E,

Lopez-Oliva E, Serradas P, Escriva F et al (2014) Predominant role of

GIP in the development of a metabolic syndrome-like phenotype in

female Wistar rats submitted to forced catch-up growth

Endocrinology 155: 3769–3780

Deacon CF, Plamboeck A, Rosenkilde MM, deHeer J, Holst JJ (2006)

GIP-(3–42) does not antagonize insulinotropic effects of GIP at physiological

concentrations Am J Physiol Endocrinol Metabol 291: E468–E475

Deblasi A, O’reilly K, Motulsky HJ (1989) Calculating receptor

number from binding experiments using same compound as

radioligand and competitor Trends Pharmacol Sci 10: 227–229

Deschamps I, Heptner W, Desjeux JF, Baltakse V, Machinot S, Lestradet

H (1980) Effects of diet on insulin and gastric inhibitory polypeptide

levels in obese children Pediatr Res 14: 300–303

Ebert R, Illmer K, Creutzfeldt W (1979) Release of gastric inhibitorypolypeptide (GIP) by intraduodenal acidiﬁcation in rats and humansand abolishment of the incretin effect of acid by GIP-antiserum inrats Gastroenterology 76: 515–523

Fulurija A, Lutz TA, Sladko K, Osto M, Wielinga PY, Bachmann MF et al.(2008) Vaccination against GIP for the treatment of obesity PLoSOne 3: e3163

Gardella TJ, Jüppner H (2001) Molecular properties of thePTH/PTHrP receptor Trends Endocrinol Metabol 12: 210–217.Gault Victor A, Hunter K, Irwin N, Greer B, Green Brian D, Harriott P

et al (2007) Characterisation and glucoregulatory actions of a novelacylated form of the (Pro3)GIP receptor antagonist in type 2 diabetes.Biol Chem 388: 173

Gault VA, O’harte FPM, Harriott P, Flatt PR (2002) Characterization ofthe cellular and metabolic effects of a novel enzyme-resistantantagonist of glucose-dependent insulinotropic polypeptide.Biochem Biophys Res Commun 290: 1420–1426

Gault VA, O’harte FPM, Harriott P, Mooney MH, Green BD, Flatt PR(2003) Effects of the novel (Pro3)GIP antagonist and exendin(9–39)amide on GIP- and GLP-1-induced cyclic AMP generation, insulinsecretion and postprandial insulin release in obese diabetic (ob/ob)mice: evidence that GIP is the major physiological incretin.Diabetologia 46: 222–230

Gault VA, Mcclean PL, Cassidy RS, Irwin N, Flatt PR (2007) Chemicalgastric inhibitory polypeptide receptor antagonism protects againstobesity, insulin resistance, glucose intolerance and associated disturbances

in mice fed high-fat and cafeteria diets Diabetologia 50: 1752–1762.Gelling RW, Coy DH, Pederson RA, Wheeler MB, Hinke S, Kwan T et al.(1997) GIP(6-30amide) contains the high afﬁnity binding region ofGIP and is a potent inhibitor of GIP1-42 action in vitro Regul Pept 69:151–154

Gourlet P, Vilardaga J-P, De Neef P, Waelbroeck M, Vandermeers A,Robberecht P (1996) The C-terminus ends of secretin and VIP interactwith the N-terminal domains of their receptors Peptides 17: 825–829.Gutniak M, Orskov C, Holst JJ, Ahrén B, Efendic S (1992)

Antidiabetogenic effect of glucagon-like peptide-1 (7–36)amide innormal subjects and patients with diabetes mellitus New Engl JMedMassachusetts Medical Society 326: 1316–1322

de Heer J, Rasmussen C, Coy DH, Holst JJ (2008) Glucagon-likepeptide-1, but not glucose-dependent insulinotropic peptide,inhibits glucagon secretion via somatostatin (receptor subtype 2) inthe perfused rat pancreas Diabetologia 51: 2263–2270

Hinke SA, Manhart S, Pamir N, Demuth HU, Gelling W, Pederson RA

et al (2001) Identiﬁcation of a bioactive domain in the terminus of glucose-dependent insulinotropic polypeptide (GIP).Biochim Biophys Acta Protein Struct Mol Enzymol 1547: 143–155.Hjorth SA, Schwartz TW (1996) Glucagon and GLP-1 receptors: lessonsfrom chimeric ligands and receptors Acta Physiol Scand 157: 343–345.Holst JJ (2004) On the Physiology of GIP and GLP-1 Horm Metab Res36: 747–754

amino-Irwin N, Green BD, Parker JC, Gault VA, O’harte FPM, Flatt PR (2006).Biological activity and antidiabetic potential of synthetic fragmentpeptides of glucose-dependent insulinotropic polypeptide, GIP(1–16)and (Pro3)GIP(1–16) Regul Pept 135: 45–53

Irwin N, Mcclean PL, O’harte FPM, Gault VA, Harriott P, Flatt PR (2007).Early administration of the glucose-dependent insulinotropicpolypeptide receptor antagonist (Pro3)GIP prevents the development ofdiabetes and related metabolic abnormalities associated with geneticallyinherited obesity in ob/ob mice Diabetologia 50: 1532–1540

BJP

Trang 38

Irwin N, Mcclean PL, Patterson S, Hunter K, Flatt PR (2009) Active

immunisation against gastric inhibitory polypeptide (GIP) improves

blood glucose control in an animal model of obesity-diabetes Biol

Chem 390: 75

Jørgensen NB, Dirksen C, Bojsen-Møller KN, Jacobsen SH, Worm D,

Hansen DL et al (2013) Exaggerated glucagon-like peptide 1

response is important for improvedβ-cell function and glucose

tolerance after roux-en-Y gastric bypass in patients with type 2

diabetes Diabetes 62: 3044–3052

Kerr BD, Flatt AJS, Flatt PR, Gault VA (2011) Characterization and

biological actions of N-terminal truncated forms of

glucose-dependent insulinotropic polypeptide Biochem Biophys Res

Commun 404: 870–876

Kilkenny C, Browne W, Cuthill IC, Emerson M, Altman DG (2010)

NC3Rs Reporting Guidelines Working Group Br J Pharmacol 160:

1577–1579

Kissow H, Hartmann B, Holst JJ, Viby NE, Hansen LRS, Rosenkilde

MM et al (2012) Glucagon-like peptide-1 (GLP-1) receptor agonism

or DPP-4 inhibition does not accelerate neoplasia in carcinogen

treated mice Regul Pept 179: 91–100

Klos A, Wende E, Wareham KJ, Monk PN (2013) International Union

of Basic and Clinical Pharmacology LXXXVII Complement Peptide

C5a, C4a, and C3a Receptors Pharmacol Rev 65: 500–543

Knudsen LB, Hastrup S, Underwood CR, Wulff BS, Fleckner J (2012)

Functional importance of GLP-1 receptor species and expression

levels in cell lines Regul Pept 175: 21–29

Kreymann B, Ghatei MA, Williams G, Bloom SR (1987) Glucagon-like

peptide-1 7–36: A physiological incretin in man Lancet 330:

1300–1304

Lazareno S, Birdsall NJ (1993) Estimation of competitive antagonist

afﬁnity from functional inhibition curves using the Gaddum, Schild

and Cheng-Prusoff equations Br J Pharmacol 109: 1110–1119

Malde AK, Srivastava SS, Coutinho EC (2007) Understanding

interactions of gastric inhibitory polypeptide (GIP) with its G-protein

coupled receptor through NMR and molecular modeling J Pept Sci

13: 287–300

Mcclean PLI (2007) GIP receptor antagonism reverses obesity, insulin

resistance, and associated metabolic disturbances induced in mice by

prolonged consumption of high-fat diet Am J Physiol Endocrinol

Metabol 293: E1746–E1755

McGrath J, Drummond G, Kilkenny C, Wainwright C (2010)

Guidelines for reporting experiments involving animals: the ARRIVE

guidelines Br J Pharmacol 160: 1573–1576

Miyawaki K, Yamada Y, Yano H, Niwa H, Ban N, Ihara Y et al (1999)

Glucose intolerance caused by a defect in the entero-insular axis: a

study in gastric inhibitory polypeptide receptor knockout mice Proc

Natl Acad Sci 96: 14843–14847

Miyawaki K, Yamada Y, Ban N, Ihara Y, Tsukiyama K, Zhou H et al

(2002) Inhibition of gastric inhibitory polypeptide signaling

prevents obesity Nat Med 8: 738–742

Nakamura T, Tanimoto H, Mizuno Y, Tsubamoto Y, Noda H (2012)

Biological and functional characteristics of a novel low-molecular

weight antagonist of glucose-dependent insulinotropic polypeptide

receptor, SKL-14959, in vitro and in vivo Diabetes Obes Metabol 14:

511–517

Nasteska D, Harada N, Suzuki K, Yamane S, Hamasaki A, Joo E et al (2014).Chronic reduction of GIP secretion alleviates obesity and insulin resistanceunder high-fat diet conditions Diabetes 63: 2332–2343

Orskov C, Jeppesen J, Madsbad S, Holst JJ (1991) Proglucagonproducts in plasma of noninsulin-dependent diabetics andnondiabetic controls in the fasting state and after oral glucose andintravenous arginine J Clin Invest 87: 415–423

Parthier C, Kleinschmidt M, Neumann P, Rudolph R, Manhart S,Schlenzig D et al (2007) Crystal structure of the incretin-boundextracellular domain of a G protein-coupled receptor Proc Natl AcadSci 104: 13942–13947

Pathak V, Gault VA, Flatt PR, Irwin N (2015) Antagonism of gastricinhibitory polypeptide (GIP) by palmitoylation of GIP analogues with N-and C-terminal modiﬁcations improves obesity and metabolic control inhigh fat fed mice Mol Cell Endocrinol 401: 120–129

Pawson AJ, Sharman JL, Benson HE, Faccenda E, Alexander SP,Buneman OP et al (2014) The IUPHAR/BPS Guide toPHARMACOLOGY: an expert-driven knowledge base of drug targetsand their ligands Nucl Acids Res 42 (Database Issue): D1098–D1106.Raufman JP, Singh L, Eng J (1991) Exendin-3, a novel peptide fromHeloderma horridum venom, interacts with vasoactive intestinalpeptide receptors and a newly described receptor on dispersed acinifrom guinea pig pancreas Description of exendin-3(9–39) amide, aspeciﬁc exendin receptor antagonist J Biol Chem 266: 2897–2902.Ravn P, Madhurantakam C, Kunze S, Matthews E, Priest C, O’brien S

et al (2013) Structural and pharmacological characterization ofnovel potent and selective monoclonal antibody antagonists ofglucose-dependent insulinotropic polypeptide receptor J Biol Chem288: 19760–19772

Scott JS, Brocklehurst KJ, Brown HS, Clarke DS, Coe H, Groombridge

SD et al (2013) Conformational restriction in a series of GPR119agonists: differences in pharmacology between mouse and human.Bioorg Med Chem Lett 23: 3175–3179

Thiele S, Rosenkilde M (2014) Interaction of chemokines with theirreceptors– from initial chemokine binding to receptor activatingsteps Curr Med Chem 21: 3594–3614

Tikhele SH, Pissurlenkar RRS, Srivastava S, Saran A, Coutinho EC(2010) Mapping interactions of gastric inhibitory polypeptide withGIPR N-terminus using NMR and molecular dynamics simulations JPept Sci 16: 383–391

Tseng CC, Kieffer TJ, Jarboe LA, Usdin TB, Wolfe MM (1996).Postprandial stimulation of insulin release by glucose-dependentinsulinotropic polypeptide (GIP) Effect of a speciﬁc glucose-dependent insulinotropic polypeptide receptor antagonist in the rat

J Clin Invest 98: 2440–2445

Venneti KC, Malthouse JP, O’harte FPM, Hewage CM (2011).Conformational, receptor interaction and alanine scan studies ofglucose-dependent insulinotropic polypeptide Biochim BiophysActa Protein Proteonomics 1814: 882–888

Vilardaga J-P, Romero G, Friedman P, Gardella T (2011) Molecularbasis of parathyroid hormone receptor signaling and trafﬁcking: afamily B GPCR paradigm Cell Mol Life Sci 68: 1–13

Trang 39

RESEARCH PAPER

Contractile function

assessment by intraventricular

balloon alters the ability of

regional ischaemia to evoke

Catherine D E Wilder, Radwa Masoud, Duygu Yazar, Brett A O ’Brien,

Thomas R Eykyn and Michael J Curtis

Cardiovascular Division, King‘s College London, London, UK

Correspondence

Michael J Curtis, CardiovascularDivision, Faculty of Life Sciences andMedicine, The Rayne Institute, StThomas’ Hospital, London SE17EH,UK

E-mail: michael.curtis@kcl.ac.uk -

In drug research using the rat Langendorff heart preparation, it is possible to study left ventricular (LV) contractility using anintraventricular balloon (IVB), and arrhythmogenesis during coronary ligation-induced regional ischaemia Assessing both con-currently would halve animal requirements We aimed to test the validity of this approach

CONCLUSIONS AND IMPLICATIONS

IVB inﬂation does not inhibit VF suppression by a standard drug, but it has profound antiarrhythmic effects of its own, likely to bedue to inﬂation-induced localized ischaemia This means rhythm and contractility cannot be assessed concurrently by this ap-proach, with implications for drug discovery and safety assessment

Abbreviations

BG, bigeminy; IVB, intraventricular balloon; IZ, ischaemic zone; LV, left ventricular; NMR, nuclear magnetic resonance; NO,nitric oxide; PCr, phosphocreatine; Pi, inorganic phosphate; TVW, total ventricular weight; UZ, uninvolved zone; VF, ven-tricularﬁbrillation; VPB, ventricular premature beat; VT, ventricular tachycardia

Trang 40

The present study reports a reﬁnement of a common model

for studying drug actions on the heart, identiﬁes an

unex-pected alteration in the model’s bioassay characteristics,

explains this mechanistically and provides guidance on

future usage

An intraventricular balloon (IVB) is commonly used in

the rat isolated Langendorff perfused heart in order to

measure left ventricular (LV) contractile function (Ridley

and Curtis, 1992; Farkas et al., 1999; Crook and Curtis,

2012) In addition, the rat isolated heart is also used to

evalu-ate ischaemia-induced and reperfusion-induced cardiac

arrhythmias such as ventricular ﬁbrillation (VF) (Ridley

et al., 1992; Clements-Jewery et al., 2006; Crook and Curtis,

2012) Assessment of rhythm and LV contractility

simulta-neously in a single experiment is a strategy that would give

a more comprehensive assessment of drug effects with

reduced animal usage Indeed, the growing interest in the

rat Langendorff preparation with an IVB for cardiac safety

assessment (Henderson et al., 2013) coupled with its

longstanding value for examining drug effects on regional

ischaemia-induced arrhythmias (Curtis et al., 2013) indicates

that combining assessment into a single experiment is a

log-ical next step

It is not known whether it is feasible to do this, and there

are reasons to suspect that IVB assessment of function may

affect arrhythmogenesis owing to effects of LV stretch Studies

on electrically induced arrhythmias with and without

ischae-mia suggest an increase in VF susceptibility and complexity

with LV stretch (Coronel et al., 2002; Parker et al., 2004;

Barrabés et al., 2013), and importantly, from a

pharmacologi-cal perspective, possible alteration of the effects of

antiar-rhythmic drugs (Chorro et al., 2000) However, the effects of

LV stretch on VF induced by ischaemia (rather than by

electri-cal stimulation) have not been systematielectri-cally established,

especially in the versatile rat Langendorff preparation

In rat isolated hearts subjected to regional ischaemia, the

incidence of VF is dependent on the size of the ischaemic

territory (Ridley et al., 1992) as it is in larger species (Austin

et al., 1982) Moreover, in the rat isolated heart, the size of

the territory, that is, the ischaemic zone (IZ), can be varied

easily by selecting the position of the coronary ligature (Ridley

et al., 1992) Proximal ligation of the left coronary artery in rat

isolated hearts results in approximately 80% of hearts

devel-oping VF, whereas placing the ligature more distally to the

atrial appendage reduces VF incidence proportionately (Ridley

et al., 1992) This allows evaluation of an alteration in the tionship between VF and IZ by drugs, or by other manipula-tion such as IVB inﬂation

rela-In order to determine the feasibility of combining IVBassessment of contractility with assessment of ischaemia-induced arrhythmias, and assess whether an IVB facilitatesischaemia-induced arrhythmias in the rat isolated heart, wedeliberately made the IZ smaller (~35% mean of the ventricu-lar mass, with individual values<40%) than is typically usedfor antiarrhythmic drug studies, thereby generating a moremoderate baseline incidence of VF (~33%) to permit detec-tion of any increase in VF incidence In separate studies, weexamined whether IVB inﬂation affected the antiarrhythmicaction of verapamil, a drug well characterized in the unloadedpreparation (Farkas et al., 1999) and which has been reported

to alter the pattern of VF in rabbit hearts subjected to IVB ﬂation (Chorro et al., 2000) Coronary efﬂuent lactate releaseand 31P nuclear magnetic resonance (NMR) spectroscopy(providing data on intracellular pH, ATP and other variablessusceptible to ischaemia) were analysed in two separate stud-ies in order to explore whether IVB-induced myocardial in-jury was the mechanism that determined outcome

in-Methods

Concordance with ARRIVE, ethical and legal requirements and new guidance on design and analysis

Animal housing and husbandry were exactly as previouslydescribed (Andrag and Curtis, 2013), in full compliance withARRIVE (Kilkenny et al., 2010) and United Kingdom HomeOfﬁce Guide on the Operation of the Animals (Scientific Proce-dures) Act 1986, and the new guidance on design and analysisfor British Journal of Pharmacology (Curtis et al., 2015) withblinding and randomization throughout, and are thereforenot described in detail here A total of 164 animals were used

in the experiments described here

Animals and general experimental methodsRats (male Wistar; Harlan UK; 280–400 g) were anaesthetizedwith a lethal dose of sodium pentobarbitone (170 mg kg 1i.p.) and given sodium heparin (160 IU kg 1 i.p.) in order

to preclude blood clot formation within the coronary ture The high dose of pentobarbitone obtained a rapid onset

BJP C D E Wilder et al.

Định dạng
Số trang	251
Dung lượng	45,69 MB