British journal of pharmacology 2016 volume 173 part 1

ISSN 0007-1188 ISSN 1476-5381 January 2016 www.brjpharmacol.org NUMBER British Journal of Pharmacology VOLUME 173 BJP BJP British Journal of Pharmacology Editor-in-Chief J.C (Ian) McGrath Glasgow, UK & Sydney, Australia Senior Editors 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 Senior Online Editor Reviews Editor Stephen Alexander Angelo A Izzo Naples, Italy David MacEwan Liverpool, UK Clare Stanford London, UK Susan Wonnacott Bath, UK Nottingham, UK Annette Gilchrist Downers Grove, USA Press Editors Y.S Bakhle Caroline Wedmore 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 Editorial Board 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 The British Journal of Pharmacology is a broad-based journal giving leading international coverage of all aspects of experimental pharmacology The Editorial 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 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 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 Copyright and Copying Copyright © 2016 The British Pharmacological Society All rights reserved No part of this publication may be reproduced, stored or transmitted in any form or 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 local Reproduction Rights Organisation (RRO), e.g Copyright Clearance Center (CCC), 222 Rosewood Drive, Danvers, MA 01923, USA (www.copyright.com), provided the 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 BJP DOI:10.1111/bph.13344 www.brjpharmacol.org British Journal of Pharmacology REVIEW Correspondence Gerhild Euler, Institute of Physiology, Justus Liebig University, Giessen, Germany E-mail: Gerhild.Euler@physiologie.med uni-giessen.de Molecular switches under TGFβ signalling during progression from cardiac hypertrophy to heart failure - Commissioning Editor: Peter Ferdinandy - Received 28 April 2015 Revised 23 July 2015 J Heger, R Schulz and G Euler Accepted 29 September 2015 Institute of Physiology, Justus Liebig University, Giessen, Germany Cardiac hypertrophy is a mechanism to compensate for increased cardiac work load, that is, after myocardial infarction or upon pressure overload However, in the long run cardiac hypertrophy is a prevailing risk factor for the development of heart failure During pathological remodelling processes leading to heart failure, decompensated hypertrophy, death of cardiomyocytes by apoptosis or necroptosis and fibrosis as well as a progressive dysfunction of cardiomyocytes are apparent Interestingly, the induction of hypertrophy, cell death or fibrosis is mediated by similar signalling pathways Therefore, tiny changes in the signalling cascade are able to switch physiological cardiac remodelling to the development of heart failure In the present review, we will describe examples of these molecular switches that change compensated hypertrophy to the development of heart failure and will focus on the importance of the signalling cascades of the TGFβ superfamily in this process In this context, potential therapeutic 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 nucleic acid; LV, left ventricle; miRNA, microRNA; MPTP, mitochondrial permeability transition pore; NLRP3, nucleotide-binding domain 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 aortic constriction; TAK1, TGFβ activated kinase 1; TGFBR1, TGFβ receptor I; TGFBR2, TGFβ receptor II; TOM, translocase of the mitochondrial outer membrane; UPS, ubiquitin proteasome system; VDAC1, voltage-dependent anion channel-1 Tables of Links TARGETS GPCRs LIGANDS a Enzymes d Angiotensin II Losartan Myostatin α-adrenoceptors AMPK PDK (PDHK) Bcl-2 β-adrenoceptors Caspase RIP1 (RIPK1) Captopril Nitric oxide (NO) ERK RIP3 (RIPK3) Isoprenaline (ISO) Noradrenaline (NA) GRK2 Sirtuin L-NAME TGFβ1 AT1 receptor Catalytic receptors NLRP3 Transporters c b JNK TAK1 p38 TGFBR1 (ALK5) ANT1 TGFBR2 GLUT1-4 SERCA2 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 Concise Guide to PHARMACOLOGY 2013/14 (a,b,c,dAlexander et al., 2013a,b,c,d) © 2015 The British Pharmacological Society British Journal of Pharmacology (2016) 173 3–14 BJP J Heger et al Introduction 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 myocardial perfusion resulting in acute myocardial infarction or chronic hypoperfusion (Hoppe and Erdmann, 2009) 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 remodelling 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 myocardial 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 function 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 dysfunction 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, inflammation and differentiation of cardiac progenitor cells In spite of this, broad inhibition of TGFβ signalling does not only have positive effects on heart failure progression Administration of the TGFβ receptor I (TGFBR1) inhibitor (SM16) after aortic banding prevented cardiac fibrosis and attenuated 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 infarction 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 needs to be used to inhibit the detrimental TGFβ pathways British Journal of Pharmacology (2016) 173 3–14 TGFβ signals through binding at a heterotetrameric receptor complex of type II and type I receptor serine/threonine kinases Upon TGFβ binding, TGFBR2 phosphorylates and thereby activates type I receptor serine/threonine kinases that in turn phosphorylates and activates SMAD transcription factors Depending on the subtype of type I receptor serine/threonine kinases, also known as activin receptors or activin receptor-like kinases (ALKs), different receptor-activated SMADs (R-SMADs) become activated In cardiomyocytes and also many other cell types, TGFβ1 signalling is attributed to TGFBR1, also called ALK5, which then results in SMAD2/3 activation In endothelial cells TGFβ has also been shown to signal via ALK1 and SMAD1/5/8 (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 SMAD1/5 and SMAD2/3 as well (Wharton and Derynck, 2009) We also identified both responses in cardiomyocytes After stimulation of ventricular cardiomyocytes, from adult rats, with ngÁmlÀ1 TGFβ1 for 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 indicating that TGFβ signalling is even more complicated than originally thought Activated R-SMADs form a complex with SMAD4 that translocates into the nucleus and acts as a transcription factor The binding specificity of SMADs to promoters can be influenced by their association with other transcription factors like activator protein (AP-1) In addition to this canonical SMAD pathway, another prominent signalling molecule of TGFβ is TGFβ-activated kinase (TAK1) TAK1 activation is also mediated by TGFBR2 Downstream targets of TAK1 are c-Jun, Nterminal kinase (JNK) and p38 Furthermore, via binding to its receptor, TGFβ can activate other kinases like ERK, phosphoinositide 3-kinase (PI3K) or small GTPases like Rho (reviewed by Zhang, 2009) This huge variety of TGFβ signalling pathways already implies that the effects of TGFβ in tissues will be complex In this review, we highlight the signalling molecules that are induced by TGFβ and modulate adverse cardiac remodelling by interfering with adrenoceptor-mediated signalling, 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-apoptotic and pro-fibrotic factor in the heart TAK1 and not SMADs seems to be the main mediator of TGFβ-induced hypertrophic growth effects TAK1 is found to be up-regulated in vivo after aortic banding, and TAK1 overexpression promotes cardiac hypertrophy in transgenic mice (Zhang et al., 2000) (Figure 3) Furthermore, in neonatal cardiomyocytes, angiotensin IIinduced hypertrophic growth could be prevented by knockdown of TAK1 with silencing RNA (siRNA), but not with siRNA against SMAD2/3 (Watkins et al., 2012) This indicates that SMAD signalling is not involved in angiotensin II – TGFβ1-induced hypertrophic growth In addition to its prohypertrophic effects, TAK1 antagonizes the apoptosis and TGFβ-guided switches to heart failure Figure TGFβ signals via the SMAD2/3 and SMAD1/5 pathway in cardiomyocytes Ventricular cardiomyocytes of adult rat were stimuÀ1 lated with ngÁml TGFβ1 for h Protein extracts of these cells were analysed by Western blots with antibodies specific against phosphoSMAD2, phosphoSMAD1/3 or phosphoSMAD1/5 Phosphorylation, which is indicative of SMAD activation, was detected for all these SMADs *P < 0.05 versus unstimulated controls, n = independent culture preparations Figure Overview about TGFβ influence on components of cardiac remodelling 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 and fibrosis All these processes are involved in heart failure progression necroptosis induced by TNFα stimulation and prevents adverse 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) complexes (Zhang et al., 2009) During TNFα stimulation, TAK1 BJP associates with RIP1 thereby preventing RIP1 interaction with other death signalling proteins, that is, with caspase or RIP3 This results in a reduction in apoptosis and necroptosis (Li et al., 2014a) As TAK1 is not only induced by TGFβ 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 counteract hypertrophy, because hypertrophic growth of cardiomyocytes induced by stimulation of α-adrenoceptors was hampered by simultaneous overexpression of SMADs (Heger et al., 2009) Hypertrophic growth of cardiomyocytes induced by stimulation of α-adrenoceptors is mediated via the transcription factor AP-1 (Taimor et al., 2004) Under simultaneous SMAD4 overexpression, AP-1/SMAD complexes are formed, which may detract AP-1 from its hypertrophypromoting target genes Indeed, a shift from hypertrophy to the induction of apoptosis is found in α-adrenoceptorstimulated and SMAD4 overexpressing cardiomyocytes (Heger et al., 2009) Furthermore, cardiac-specific SMAD4 knock-out mice displayed cardiac hypertrophy (Wang et al., 2005) This indicates that SMAD4 acts as a molecular switch for transition from hypertrophy to apoptosis In addition, TGFβ induces apoptosis in adult cardiomyocytes via enhancement of SMAD and AP-1 activity (Schneiders et al., 2005) Similar to these findings, inhibition of SMAD signalling in vivo may preserve the compensating character of hypertrophic growth in cardiac remodelling while preventing the transition to apoptosis (Figure 3) That AP-1 is a mediator of hypertrophy and apoptosis in βadrenoceptor stimulated cardiomyocytes has been shown by use of transgenic mice overexpressing the AP-1 inhibitor jun dimerization protein (JDP2) JDP2 overexpression 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 contractile function of cardiomyocytes because AP-1 inhibition by JDP2 overexpression attenuated contractile responses induced by β-adrenoceptor stimulation (Hill et al., 2013) Therefore, to prevent adverse remodelling, inhibition of SMAD signalling seems to be the better choice than inhibition of AP-1, because this would negatively influence the contractile function of the heart Modulation of β-adrenoceptor responses in the presence of TGFβ During heart failure, progressive desensitization of βadrenoceptors occurs β-adrenoceptors are members of the GPCR superfamily whose stimulation results in activation of PKA via AC and cAMP, which regulate different intracellular, sarcolemmal and myofibrillar substrates Thus, cAMP exerts the cellular effects on cardiac contractile function induced by activation of β-adrenoceptors However, stimulation of β-adrenoceptors also results in agonist-dependent desensitization of these receptors, a phenomenon found during the development of heart failure This process is mediated by the receptor adapter protein β-arrestin that binds to β-adrenoceptors This binding either results in direct British Journal of Pharmacology (2016) 173 3–14 BJP J Heger et al Figure Influence of TGFβ-SMAD and TGFβ-TAK1 signalling on adrenoceptor-mediated pathways in LV heart failure progression Adrenoceptors (AR) stimulate the expression of genes promoting hypertrophic growth via the transcription factor AP-1 Under simultaneous presence of SMAD4, the prohypertrophic response to adrenoceptor stimulation is shifted to a pro-apoptotic gene transcription via AP-1/SMAD complexes Also, under TGFβ stimulation of cardiomyocytes, AP-1 and SMADs mediate apoptosis In addition to these effects on cardiomyocytes, activation of the TGFβ/SMAD pathway or induction of SMADs via β-arrestins induces the transcription of fibrotic genes In contrast to the SMAD pathway, TAK1 activation stimulates hypertrophic growth while inhibiting cardiac necroptosis and apoptosis by interacting with RIP1 Strong β-adrenoceptor (ADRB) activation results in β-adrenoceptor desensitization via β-adrenoceptor /β-arrestin complexes This process can be inhibited by TGFβ Depicted in red are switch molecules that can modulate the response of the cell to receptor stimulation and thereby influence the outcome of this stimulation on the remodelling process inhibition of β-adrenoceptors, known as functional desensitization, or in internalization of β-adrenoceptors that reduces their density (reviewed by Lymperopoulos and Negussie, 2013) Studies in β-arrestin1 knock-out mice demonstrated a major role for β-arrestin1 in cardiac dysfunction, because contractile responses to β-adrenoceptor stimulation were enhanced in these knock-out animals (Conner et al., 1997) Furthermore, knock-down of βarrestin1 prevented adverse cardiac remodelling after myocardial infarction by inhibiting apoptosis and preserving cardiac function (Bathgate-Siryk et al., 2014) Interestingly, an increase in myocardial β-adrenoceptor density and a reduction in negative regulators such as Giα and βadrenoceptor-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 prevent β-adrenoceptor desensitization in cardiomyocytes and thereby promote pro-hypertrophic signalling Whether this response is mediated by the down-regulation of βarrestin1 by TGFβ has not yet been clarified But TGFβ may be a plausible target in order to prevent β-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 British Journal of Pharmacology (2016) 173 3–14 fibroblasts during heart failure Overexpression of β-arrestin in cardiac fibroblasts results in the uncoupling of βadrenoceptors and activation of SMAD2/3, thereby promoting a pro-fibrotic phenotype This may cause enhanced stiffness of the ventricular wall and contribute to the development of heart failure Although TGFβ stimulation prevents the uncoupling of β-adrenoceptors and enhances the pro-hypertrophic signalling, the inotropic β-adrenoceptor-mediated response was diminished in TGFβ-overexpressing mice This is due to an up-regulation of mitochondrial uncoupling proteins during β-adrenoceptor stimulation, which results in a decreased mitochondrial energy production Thus, TGFβ-overexpressing mice resemble a phenotype occurring at the transition to heart failure, namely, displaying cardiomyocytes hypertrophy and promoting apoptosis as well as mitochondrial and contractile dysfunction (Schneiders et al., 2005; Huntgeburth et al., 2011) That these interacting pathways of ADRB-TGFβ signalling are even more complex was indicated by the findings that GPCRs not only activate TK receptors but also also transactivate the serine/threonine kinase TGFBR1 in different cell types (Burch et al., 2012) The proposed mechanism for this transactivation is activation of integrin by GPCRs Subsequently, integrin binding to the large latent TGFβ complex causes a conformational change and allows TGFβ to bind and activate TGFBR2/TGFBR1, thereby resulting in SMAD TGFβ-guided switches to heart failure activation (Munger et al., 1999) Whether this βadrenoceptor-induced SMAD signalling holds true in 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 analysis of the proteasome, as degradation of proteins is changed in cardiac hypertrophy The primary cellular response to βadrenoceptor stimulation in the heart is an increased pool of 20S subunits with catalytic activity, while chronic βadrenoceptor stimulation enhanced the 26S proteasome but decreased 20S proteasomal activity, accompanied by a decrease 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 different β-subunits of the proteasome, is decisive for the development of hypertrophy and depends again on the strength of β-adrenoceptor activation Proteins involved in cardiac hypertrophy are targeted by musclespecific 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 upregulation of an autophagy controlling protein, the endosomal sorting complex required for transport III (ESCRT-III) family protein charged multivesicular body protein 2B (CHMP2B) Thus, Zaglia et al (2014) demonstrated the interplay between the ubiquitin proteasome system (UPS) and autophagy and the importance of controlled degradation 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 mechanism for angiotensin II–induced apoptosis in hearts that is accompanied by activation of insulin-like growth factor receptor II (IGF2R) signalling Heat shock transcription factor (HSF1) acts as a repressor of IGF2R gene expression only if deacetylated by sirtuin However, angiotensin II and subsequently JNK activation mediates sirtuin 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 remarkable 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 evaluated, 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 intracellular activity of TGFβ controlled by UPS but also the BJP stability and levels of TGFβ receptor complexes are determined by ubiquitination (Xu et al., 2012) Influence of TGFβ on mitochondria, energy metabolism and heart failure Mitochondria are the power houses of the cell, generating ATP via oxidative phosphorylation On average, 30% of the cardiomyocytes volume is filled with mitochondria (Barth et al., 1992) One side product of the major respiratory enzyme complexes is the generation of reactive oxygen species (ROS) that modifies the redox potential of the cell and is essential for numerous signalling pathways (Chen and Zweier, 2014) Mitochondrial dysfunction occurs under pathophysiological conditions and involves malfunction of complexes of oxidative phosphorylation, and an increase in ROS production that leads to cell death contributing to the development of heart failure The enzymes of the respiratory chain seem to be the main site of ROS formation, but many other enzymes contribute to ROS production in failing hearts, including monoamine oxidases and the cytosolic adaptor protein p66Shc (Di Lisa et al., 2009) Cellular stress signals lead to translocation of p66Shc into the mitochondrial intermembrane space, where it oxidizes cytochrome c and generates ROS (Heusch, 2015) Factors that influence ROS production, therefore, critically determine the cell’s fate A newly identified signalling molecule in the control of mitochondrial ROS production that is under the control of TGFβ signalling is nucleotide-binding domain and leucinerich repeat containing PYD-3 (NLRP3), a pattern recognition receptor that is involved in the pathogenesis of chronic diseases and inflammation NLRP3 is expressed in the heart, localized in mitochondria, and interacts with components of the redox system (Figure 4) Upon TGFβ stimulation of cardiac fibroblasts, NLRP3 increases mitochondrial ROS production, which supports SMAD2 phosphorylation and results in the differentiation of cardiac fibroblasts into myofibroblasts, an important process in adverse cardiac remodelling (Bracey et al., 2014) The involvement of NLRP3 in cardiac fibrosis has been confirmed in an in vivo model of hypertension: angiotensin II infusion for 28 days resulted in TGFβ-mediated fibrosis in wild-type mice, but NLRP3-deficient mice were protected against this angiotensin II-induced fibrosis NLRP3, therefore, is a newly identified mitochondrial signalling factor in TGFβ-induced cardiac remodelling that may promote the transition to heart failure as it facilitates ROS-mediated fibrosis Increased ROS production induces the opening of the mitochondrial permeability transition pore (MPTP) (Figure 4) that changes the permeability of the inner mitochondrial membrane, leading to mitophagy, fusion/fission events and biogenesis (Brenner and Moulin, 2012) Opening of the MPTP facilitates the release of pro-apoptotic factors from the mitochondria that stimulates the activation of caspases and finally leads to cell death (Kinnally et al., 2011) The addition of noradrenaline induced a concentration-dependent decrease in mitochondrial membrane potential that was associated with a switch from compensated hypertrophy to apoptosis, thereby indicating that MPTP opening is involved in British Journal of Pharmacology (2016) 173 3–14 BJP J Heger et al Figure The central role of mitochondria in LV heart failure can be modulated by TGFβ Hypertrophy, fibrosis and apoptosis can be controlled by mitochondria via generation of ROS NLRP3 is a newly identified molecule that enhances mitochondrial ROS production and that is controlled by TGFβ or angiotensin II (AngII) miR181c enhances ROS production via modulation of complex IV of the respiratory chain TOM70, acting as a repressor of mitochondrial ROS production, is found to be reduced in heart failure This reduction then provokes enhancement of ROS Enhancement of mitochondrial uncoupling protein during stimulation of β-adrenoceptors by noradrenaline (NA) and TGFβ-receptor activation results in reduced energy production and impaired contractile function Opening of the MPTP plays a central role in the induction of apoptosis Opening of this pore can 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 adverse remodelling (Jain et al., 2015) Inhibiting MPTP opening by overexpression of adenine nucleotide translocase (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 contribution MPTP opening has to the adverse cardiac remodelling induced 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 (Bcl-2) family is a well-known gate keeper in mitochondriamediated 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 mitochondrial 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 British Journal of Pharmacology (2016) 173 3–14 action on the mitochondrial level is Hsp22 (Qiu et al., 2011) Overexpression of Hsp22 results in physiological hypertrophy via up-regulation of NFκB, and binding of Hsp22 to signal transducer and activator of transcription (STAT3), which is a marker of cardiac stress responses Down-regulation of Hsp22 leads to an increased remodelling of the heart and death in knock-out mice after transverse aortic constriction (TAC) by modulating the nuclear and mitochondrial function of STAT3 and STAT3-dependent genes A further mitochondria-associated candidate, mediating a switch to pathophysiological hypertrophy is TOM70, a translocase of the mitochondrial outer membrane (TOM) complex that mediates the import of mitochondrial preproteins (Figure 4) Li et al (2014b) nicely showed a down-regulation of TOM70 in pathophysiological hypertrophy in humans as well as animal models This results in the reduced import of optical atrophy-1 (OPA1) – a protein important for mitochondrial fusion –, a reduction in complex I activity and finally in ROS production As a consequence, changes in the outer mitochondrial membrane and/or inner mitochondrial membrane occurred, followed by apoptotic events as discussed above In addition, increased TOM70 levels made cardiomyocytes completely resistant to the effects of various pro-hypertrophic stimuli These findings explain the significance of the modulation of mitochondrial pores by, for example, VDAC1, crystalline B or TOM70 for cardiac hypertrophy to progress to heart failure At the onset of the development of heart failure, a metabolic shift from fatty acid to glucose metabolism has been TGFβ-guided switches to heart failure described This shift is due to the down-regulation of enzymes for fatty acid oxidation, whereas glycolytic enzymes 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 related to adverse cardiac remodelling A key regulator of energy 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 protects 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) uptake into cardiomyocytes as well as a reduction of transporters 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 progression Furthermore, prevention of the metabolic switch, probably via AMPK, is a promising target for therapeutic approaches against heart failure development BJP MicroRNAs in heart failure An increased ability to regulate the processes involved in cardiac remodelling is attributed to miRNAs miRNAs are small noncoding RNAs that target the 3´-untranslated region or 5’untranslated region of mRNA transcripts This results in the destabilization or translational repression of mRNAs (Bartel, 2004) Furthermore, miRNAs can regulate gene transcription by inducing histone modifications or DNA methylations (Hawkins and Morris, 2008) In fact, one single miRNA can affect many target genes generating a broad network of miRNA-controlled gene expression that has a huge effect on different biological processes including cardiac remodelling Analysing the role of miRNAs in heart failure development has already identified some promising new therapeutic targets The RNase III endonuclease Dicer is essential for the processing of pre-miRNA into its mature form In the adult myocardium, a loss of Dicer-induced biventricular enlargement is accompanied by hypertrophic growth of cardiomyocytes, ventricular fibrosis and functional defects (da Costa Martins et al., 2008) A similar study by Chen et al (2008) revealed signs of dilated cardiomyopathy and heart failure after cardiac-specific deletion of Dicer Furthermore, they found that the level of Dicer protein was significantly reduced in in human patients with dilated cardiomyopathy and failing hearts These findings indicate that miRNAs have a major function in the control of heart failure development and progression Either an up-regulation or down-regulation of miRNAs under pressure overload can mediate cardiac remodelling, for example, when miR25 is increased the activity of the sarcoplasmic reticulum (SR) Ca2+ ATPase (SERCA2A) is reduced Figure Influence of miRNAs on LV adverse remodelling can be modulated by β-adrenoceptors (ADRBs) or TGFβ miRNAs that have been demonstrated to reverse or promote adverse cardiac remodelling are depicted Up-regulation of miR15 or miR22 prevents the induction of fibrosis or apoptosis under pressure overload (TAC) or β-adrenoceptor stimulation (ISO) while preserving effects on moderate, compensatory hypertrophy, as when these miRs are inhibited adverse remodelling develops In contrast, up-regulation of miR25 or miR21 under TAC enhances adverse cardiac remodelling, 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 are depicted in red Green arrows and symbols indicate interference of miR expression by anti-miRs or transgenic overexpression British Journal of Pharmacology (2016) 173 3–14 BJP J Heger et al 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 overload 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 epigenetic regulation of gene expression during cardiac hypertrophy These findings indicate that during β-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 mitochondria (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 modulate exercise capacity directly at the mitochondrial level Interestingly, this miR181c-induced dysfunction was not accompanied by cardiac hypertrophy Unfortunately, these observations were only carried out in vitro with a simulated overexpression of miR181c and its role in cardiovascular disease 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 family members that are up-regulated in human heart failure and can inhibit numerous components of the TGFβ signalling pathway (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 SMADsignalling 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 order to restrict adverse remodelling However, due to the ubiquitous 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 regulator 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 myocardium (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 transition during TGFβ stimulation (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 10 British Journal of Pharmacology (2016) 173 3–14 (including miR21) Fibroblast-derived miR21 acts as a potent paracrine RNA molecule that induces cardiomyocyte hypertrophy (Bang et al., 2014) miR155, secreted by macrophages, also has paracrine effects on the heart, because miR155 knockout in macrophages prevented angiotensin II-induced or TACinduced cardiac hypertrophy and dysfunction although fibrosis was still present These findings showing that miRNAs can exert paracrine effects implies that systemic pharmacological interference of miRNA signalling by the use of anti-miRNAs might be useful for preventing heart failure progression Such a systemic therapeutic intervention would be easy to apply clinically However, in cardiac pathophysiology, the systemic application of anti-miRNAs is still restricted to basic science studies A promising approach in this direction has been shown by Montgomery et al (2011) using locked nucleic acid (LNA)modified miR208a-antisense oligonucleotides Systemic delivery of these oligos silenced miR208a-expression in the heart, thereby preventing hypertension-induced heart failure in Dahl hypertensive rats by reducing cardiomyocyte hypertrophy, cardiac fibrosis and improving cardiac function Because cardiac miR208 overexpression in transgenic mice induced cardiac hypertrophy, the reduction of miR208a in the heart by LNA-antisense oligos is at least in part responsible for cardioprotection in hypertensive Dahl rats However, reductions in the levels of circulating miR499 and miR208b were also found during the treatment with LNA-modified 208a antisense oligos Therefore, a combination of local and systemic effects may contribute to the protective effects All these studies demonstrate the enormity of the miRNA network and its influence on heart failure progression Fine tuning of specific miRNAs is essential for physiological hypertrophy or decompensation, and thus has great therapeutic potential for the treatment of heart failure patients Right heart failure For many years, analysis of left ventricular systolic dysfunction was at the centre of heart failure research, which is also the focus of our review However, in recent years, some promising advances in the analysis of right heart failure have been made that should be discussed here One major cause for right ventricular dysfunction is pulmonary hypertension (PAH) Due to an increase in pulmonary vascular resistance, afterload on the right ventricle (RV) increases, leading to right heart failure which determines the prognosis of patients with PAH Therefore, it is of utmost importance to define new therapeutic strategies to reduce RV remodelling in order to improve patient prognosis (Ryan et al., 2015) The RV, similar to the LV, compensates an increased work load due to hypertension by hypertrophic growth processes However, compensatory remodelling is limited and over time, the RV decompensates, finally leading to heart failure There seem to be chamber-specific responses, and thus, a simple extension of LV findings to the RV is not possible An interesting new finding in this respect comes from Schreckenberg et al (2015) who analysed the effect of chronic NO deprivation on the remodelling processes in the LV and RV Treatment of rats with the NOS inhibitor L-NAME resulted in moderate ventricular hypertrophy without signs of dysfunction However, the Nicergoline inhibits platelet Ca2+ signalling BJP Figure 2+ A summary of the localization of fluorescent Ca indicators used in this study The diagram shows a simplified structural diagram of a platelet including key cellular structures discussed in this paper These include the dense tubular system (DTS; the platelet equivalent of the smooth endoplasmic reticulum), the open canalicular system (OCS; a complex invagination of the platelet plasma membrane), the membrane complex (MC; a close apposition of the OCS and DTS), the cortical microtubule bundle (CMB; made up of a number of microtubule coils; labelled with 2+ TubulinTracker) and the acidic Ca stores (which probably encompass the lysosomes as well as the α- and dense granules) Note the presence of KDEL-containing proteins solely within the DTS (van Nispen tot Pannerden et al., 2009), which allows this structure to be labelled distinctly 2+ from the acidic Ca stores A cortical actin ring is also found in platelets and occupies a space similar to the cortical microtubule bundle (labelled with CytoPainter Phalloidin-iFluor555) and treated samples were randomly assigned to samples within each of these groups before the start of the experiment Blinding Data files were labelled with a date and sample identifier (e.g letter, number or time of experiment) Data were analysed in this format and then subsequently reassigned to their experimental condition using lab records Normalization Data were subjected to statistical analyses before normalization Data sets are presented as mean % of control to allow for comparison of results obtained between different preparations, as there were significant variations in the magnitude of agonist-evoked Ca2+ signals observed in the control responses of samples taken from different donors In the Mn2+ quench experiments, normalization to baseline fluorescence levels (F/F0) was used to allow for differences in resting Fura-2 fluorescence of samples Statistical comparison Values are presented as the mean ± SEM of the number of independent observations (n) indicated Analysis of statistical significance was performed on independently acquired values using Student’s paired t-test or using a one-way ANOVA test followed by a post hoc Bonferroni multiple comparisons test P < 0.05 was considered significant Results Nicergoline inhibits thrombin-evoked Ca2+ signalling in human platelets Experiments were performed to examine whether pretreating platelets with nicergoline at a concentration able to trigger reorganization of the OCS and DTS (Le Menn et al., 1979) affected Ca2+ signalling These experiments were initially performed in the absence of extracellular Ca2+, as these conditions allow the clearest examination of pericellular Ca2+ recycling by preventing direct Ca2+ entry through channels in the plasma membrane As shown in Figure 2A, preincubation with 10 μM nicergoline for 10 at 37°C had no effect on thrombin-evoked rises in [Ca2+]cyt (n = 6; P > 0.05), whereas pretreatment with 50 or 100 μM nicergoline elicited a significant inhibition of thrombin-evoked rises in [Ca2+]cyt (Figure 2B; both n = 6; P < 0.05) In addition, pretreatment with higher concentrations of nicergoline was found to elicit a small, but significant fall in the resting [Ca2+]cyt observed after EGTA treatment (Figure 2C; both n = 6; P < 0.05) compared with the control samples (n = 6) No significant effect on resting [Ca2+]cyt was observed in cells pretreated with 10 μM nicergoline (n = 6; P > 0.05) Further experiments found that nicergoline itself induced no change in [Ca2+]cyt either in the presence or absence of external Ca2+, but it did trigger a slow, small reduction in the baseline measured Ca2+ (Supporting Information Figure S1) Pretreatment of platelets with 100 μM nicergoline also significantly inhibited thrombin-evoked release of Ca2+ from British Journal of Pharmacology (2016) 173 234–247 237 BJP T Walford et al Figure 2+ Nicergoline inhibits resting and thrombin (Thr)-evoked Ca signalling in human platelets Fura-2-loaded (A–C) or Fluo-5N-loaded (D) human platelets were suspended in supplemented HBS Platelets were pretreated with 10, 50 or 100 μM nicergoline for 10 at 37°C in the presence À1 of continuous magnetic stirring One millimolar EGTA was added before the cells were stimulated with 0.5 UÁmL thrombin (B) and (C) show 2+ 2+ graphs summarizing the effect of nicergoline on thrombin-evoked changes in [Ca ]cyt (B) and resting [Ca ]cyt (C) respectively Results presented are representative of five and six experiments respectively intracellular stores (73.9 ± 4.9% of control; n = 5; P < 0.05; Figure 2D) However, unlike [Ca2+]cyt, baseline Fluo-5N fluorescence was found to be unaffected by nicergoline treatment (101.8 ± 3.1% of control; n = 5; P > 0.05) These results demonstrate that nicergoline reduces thrombin-stimulated increases in [Ca2+]cyt, through reducing thrombin-evoked release of Ca2+ from intracellular stores and not by changing their basal Ca2+ content In addition, the lack of change in resting [Ca2+]st suggests that the nicergoline-induced reduction in [Ca2+]cyt is not mediated through increasing Ca2+ uptake into the intracellular Ca2+ stores Given the nicergoline concentrations utilized here are significantly higher than those required to fully inhibit α-adrenoreceptors; it seems most likely that the inhibitory effect of nicergoline on resting and thrombin-evoked Ca2+ signalling is related to the ultrastructural reorganization of the platelets previously reported (Le Menn et al., 1979) Nicergoline causes disruption to the cortical microtubule network Le Menn et al (1979) demonstrated that nicergoline treatment results in reorganization of the platelet cortical microtubule bundle, with the bundle remaining present but with an apparent reduction in the number of microtubules, with the remainder displaced centrally The authors also reported that reorganization of the cortical microtubule bundle also caused the cells to become more spherical, in line with the known role of this cytoskeletal structure in maintaining platelets in their resting discoid form (Italiano et al., 2003) To confirm that nicergoline causes disorganization of the cortical microtubule network, the microtubule structure of 238 British Journal of Pharmacology (2016) 173 234–247 nicergoline-treated and DMSO-treated platelets was examined in resting platelets co-loaded with both TubulinTracker and the cytosolic label, Fura red Platelets treated with DMSO (the vehicle for nicergoline) frequently had prominent cortical bundles of microtubules with only weak fluorescence observed from the central regions of the cells (Figure 3) In contrast, platelets treated with 100 μM nicergoline were found to have a significant disruption to the cortical microtubule bundle with a greater percentage of platelets in each field, demonstrating a complete lack of a discernible cortical bundle (24.5 ± 7.4% of nicergoline-treated platelets compared with 11.2 ± 5.2% of DMSO-treated platelets; n = 9; P < 0.05) or showing an incomplete cortical ring (21.0 ± 2.9% of nicergoline-treated platelets compared with 9.2 ± 3.6% of DMSO-treated platelets; n = 9; P < 0.05) A reduced proportion of the TubulinTracker fluorescence was observed in the cortical regions in nicergoline-treated platelets (63.9 ± 5.1% in DMSO-treated platelets compared with 36.3 ± 6.4% in nicergoline-treated platelets; n = 9; P < 0.05) Nicergoline pretreatment also reduced the thickness of the microtubule bundles in cells displaying an observable cortex (197 ± 50 nm in nicergoline-treated platelets compared with 401 ± 64 nm in DMSO-treated platelets; n = 9; P < 0.05) These results confirm the findings of Le Menn et al (1979) that nicergoline pretreatment disrupts the structure of the cortical microtubule bundle In addition, experiments were performed to examine the effect of nicergoline on the actin cytoskeleton in fixed cells Although no obvious difference was seen in the morphology of live nicergoline-treated cells plated onto poly-L-lysinecoated coverslips, when experiments were performed using fixed cells, the nicergoline-treated cells could be observed Nicergoline inhibits platelet Ca2+ signalling BJP Figure Nicergoline causes disruption of the cortical microtubule network Platelets co-loaded with both TubulinTracker and Fura Red were suspended in supplemented HBS Cells were pretreated with either DMSO or 100 μM nicergoline for at 37°C in the presence of continuous magnetic stirring The platelets were then added to a poly-L-lysine-coated chambered slide and allowed to settle for Excess platelet suspension 2+ was removed, and the slides were washed twice with Ca -free HBS Fluorescence was then monitored using an Olympus Fluoview FV1200 confocal microscope The upper panels show the overlay of both TubulinTracker and Fura Red fluorescence; the lower panels show the fluorescence from TubulinTracker alone The results presented are representative of nine experiments to take a consistent rounded form (Supporting Information Figure S2) – consistent with them becoming more spherical in shape The observation that fixed nicergoline-treated platelets plated onto coverslips appear spherical, whilst live cells appear similar to the control cells suggests that platelets are able to spread similarly to control cells through remodelling of their cortical F-actin layer Further analysis of these cells found that nicergoline caused a slight thickening of the cortical actin ring without altering the amount of F-actin within the cells These data are presented and discussed further in the Supporting Information (Supporting Information Figure S2) Nicergoline triggers a reorganization of the dense tubular system and inhibits thrombin-evoked Ca2+ removal and pericellular Ca2+ accumulation in human platelets Le Menn et al (1979) demonstrated that high concentrations of nicergoline cause a reorganization of the membrane of the DTS, with these intracellular Ca2+ stores remaining associated with the disorganized microtubule system Experiments were therefore performed to assess the effect of nicergoline on the subcellular localization of the DTS To confirm the nicergolineinduced change in the DTS, fixed nicergoline or DMSO-treated platelets were permeabilized and stained using an antibody to proteins containing the endoplasmic reticulum retention signal, KDEL This approach has previously been used to demonstrate the specific localization of protein disulfide isomerase (PDI) in the DTS (van Nispen tot Pannerden et al., 2009) Fluorescent imaging of the labelled cells found that control cells possessed an inhomogeneous, punctate distribution (Figure 4A) – which is consistent with previous observations of the distribution of PDI in human platelets (van Nispen tot Pannerden et al., 2009) In contrast, nicergoline-treated cells did not show such obvious puncta, and fluorescence appeared to be more homogenously distributed This was confirmed in line scan analysis of the cells, with DMSO-treated cells showing one or two obvious spikes in the fluorescence levels in the localized areas of the cell (Figure 4B); in contrast, nicergoline-treated cells did not show such spikes and were instead seen to have a homogenous fluorescence in their interior, with fluorescence only falling at the cell margins Quantitative analysis of these results found that mean cellular fluorescence was unaffected by nicergoline pretreatment (97.3 ± 4.6% of control; n = 5; P > 0.05; Figure 4), yet the variance in platelet pixel fluorescence was found to be significantly lower in nicergoline-treated platelets (SD = 451.5 ± 24.4 arbitrary units in nicergoline-treated cells compared with 565.6 ± 312 arbitrary units in DMSO-treated cells; n = 5; P < 0.05) The maximum cell pixel fluorescence in nicergoline-treated platelets was also found to be consistently reduced across all donors (82.4 ± 4.1% of control; n = 5; P < 0.05) These data suggest that there is no difference in DTS content of nicergoline-treated platelets, but the DTS is redistributed away from its normal concentration at the membrane complex Similar effects of nicergoline upon the distribution of intracellular Ca2+ stores were also observed in Fluo-5N-loaded human platelets (Supporting Information Figure S3) These results are consistent with the previous work that indicates that the cortical microtubule bundle is likely to be important in British Journal of Pharmacology (2016) 173 234–247 239 BJP T Walford et al Figure Nicergoline triggers redistribution of the DTS (A) Platelets were treated with either DMSO or 100 μM nicergoline for at 37°C in the presence -1 of continuous magnetic stirring Cells were then fixed, permeabilized and incubated with 1% (v v ) Fluor®488-labelled anti-KDEL antibody for 30 at room temperature The cells were washed and then resuspended in supplemented HBS The labelled cells were then allowed to settle for on poly-L-lysine-coated chambered slide Fluorescence was then monitored using an Olympus Fluoview FV1200 confocal microscope Images for the fluorescence alone (left) or overlaid over the transmitted light image (right) are shown (B) A linescan analysis of the three numbered cells indicated in (A) is shown The results presented are representative of five experiments maintaining the normal inhomogenous distribution of the DTS in platelets (White, 1968; Le Menn et al., 1979) Our previous work has indicated that the close interaction of membranes at the MC is required for the near-maximal rate of thrombin-evoked Ca2+ efflux and calculated that this high rate of transport would be needed to generate the observed thrombin-evoked pericellular Ca2+ signals (Sage et al., 2013) Experiments were therefore performed to examine whether the nicergoline-induced redistribution of the intracellular Ca2+ stores affected Ca2+ removal from the cell and accumulation in the pericellular space As shown in Figure 5A and B, nicergoline treatment significantly inhibited thrombinevoked Ca2+ efflux from the cells (37.1 ± 3.9% of control; n = 5; P < 0.05; Figure 5) and also inhibited the resulting increase in [Ca2+]peri (23.3 ± 8.4% of control; n = 6; P < 0.05; Figure 5), suggesting that re-organization of the intracellular membranes may affect thrombin-evoked rises in [Ca2+]cyt by interfering with pericellular Ca2+ recycling by inhibiting Ca2+ removal into the pericellular space by disrupting the 240 British Journal of Pharmacology (2016) 173 234–247 close apposition of the DTS with the OCS This conclusion was supported by additional experiments showing that nicergoline induces no additional inhibitory effect when pericellular Ca2+ rises are also blocked (Supporting Information Figure S4) Interestingly, there was no significant alteration observed in either the resting [Ca2+]ext or [Ca2+]peri (99.8 ± 4.7% and 98.6 ± 12.1% of control for [Ca2+]ext and [Ca2+]peri respectively; n = and respectively; P > 0.05 for both) As the data indicate that nicergoline induces no active redistribution of Ca2+ from either the extracellular fluid or the intracellular stores – these results suggest that the most likely reason for the nicergoline-induced reduction in resting [Ca2+]cyt is a small increase in the cytosolic volume caused by the loss of the cortical microtubule bundle leading to platelet becomes more spherical in shape Given that platelets rapidly change shape from discoid to spherical upon activation, such a change in cell volume is unlikely to significantly alter thrombin-evoked Ca2+ signals Nicergoline inhibits platelet Ca2+ signalling BJP triggered a reduction in [Ca2+]cyt (35.0 ± 8.8 nM for nicergolinetreated versus 50.8 ± 5.6 nM for control cells; n = 7; P < 0.05), when cells were treated with taxol before nicergoline, there was no significant difference in the observed resting [Ca2+]cyt (98.3 ± 17.2 nM for nicergoline- and taxol-treated cells versus 80.6 ± 20.3 nM for taxol-treated cells; n = 7; P > 0.05) Similar experiments were performed to examine whether taxol pretreatment also prevented the effect of nicergoline on thrombin-evoked changes in [Ca2+]st, [Ca2+]ext and [Ca2+]peri Consistent with the previous experiments, nicergoline significantly inhibited thrombin-evoked changes in [Ca2+]st (52.4 ± 8.3% of control; n = 8; P < 0.05; Figure 6B), [Ca2+]ext (41.0 ± 16.6% of control; n = 10; P < 0.05; Figure 6C) and [Ca2+]peri (11.0 ± 11.0% of control; n = 7; P < 0.05; Figure 6E) However, upon pretreatment with taxol, nicergoline had no statistically significant effect on any of these parameters (81.3 ± 12.9% of control, 86.0 ± 20.7% of control and 263.3 ± 99.8% of control for [Ca2+]st, [Ca2+]ext and [Ca2+]peri respectively; n = 8, 10 and respectively; P > 0.05 for all conditions; Figure 6B, D and F) These data indicate that nicergolineinduced changes in Ca2+ signalling are due to the structural alteration induced by disruption of the cortical microtubule bundle, probably resulting in disruption of the MC Nicergoline treatment inhibits the initial accumulation and spread of Ca2+ from the pericellular Ca2+ hotspot Figure 2+ Nicergoline inhibits thrombin (Thr)-evoked Ca removal and 2+ pericellular Ca accumulation in human platelets (A and B) Washed platelets suspended in supplemented HBS containing 2.5 μM + Fluo-4 K salt (A) or FFP-18-loaded platelets (B) were pretreated with either DMSO or 100 μM nicergoline for at 37°C in the presence of continuous magnetic stirring One millimolar EGTA was then À1 added before the cells were stimulated with 0.5 UÁmL thrombin The results presented are representative of five and six experiments respectively Stabilization of platelet microtubules by prior treatment with taxol prevents nicergolineinduced disruption of thrombin-evoked Ca2+ signalling in platelets As the nicergoline-induced ultrastructural changes in the DTS and OCS appear to be related to a change in the cortical microtubule structure (Le Menn et al., 1979), further experiments were performed to investigate whether preventing microtubule reorganization by pre-treatment of platelets with the microtubule-stabilizing agent, taxol, prior to incubation with nicergoline negated the inhibitory effect of nicergoline on thrombin-evoked Ca2+ signalling As previously observed, pretreatment with nicergoline significantly inhibited thrombin-evoked rises in [Ca2+]cyt to 64.6 ± 5.1% of control (n = 7; P < 0.05; Figure 6A) In contrast, incubation of taxolpretreated platelets with nicergoline resulted in no significant defect in the thrombin-evoked rise in [Ca2+]cyt compared with the taxol-pretreated cells (107.2 ± 7.0% of control; n = 7; P > 0.05) In line with a role for the nicergoline-induced structural alterations in eliciting the reduced resting [Ca2+]cyt previously observed, we find that whilst nicergoline alone again Single cell imaging was employed to examine whether nicergoline treatment alters the characteristics of the pericellular Ca2+ accumulations observed in single platelets As previously shown, virtually, all DMSO-treated platelets were observed to generate a pericellular Ca2+ hotspot within the platelet boundary upon thrombin stimulation (e.g as indicated by yellow arrow in Figure 7B; 93.5 ± 3.1% of cells/field, n = 5; Sage et al., 2013) Whilst most nicergolinetreated platelets were still able to generate a similar microdomain of raised [Ca2+]peri, the proportion of cells showing a Ca2+ hotspot was significantly reduced (62.9 ± 7.6% of cells per field in treated cells; n = 5; P 0.05) These results indicate that the nicergoline-induced redistribution of the DTS reduces Ca2+ accumulation within the British Journal of Pharmacology (2016) 173 234–247 241 BJP T Walford et al Figure 2+ Stabilization of platelet microtubules by pretreatment with taxol prevents nicergoline-induced disruption of thrombin (Thr)-evoked Ca signalling Fura-2-loaded (A), Fluo-5N-loaded (B) or FFP-18-loaded (E and F) human platelets, or platelets suspended in supplemented HBS with 2.5 μM + Fluo-3 K salt (C and D) were pretreated with either 100 μM taxol (A, B, D and F) or its vehicle, DMSO (A–C and E), for 30 min, followed by a further 5-min incubation with either 100 μM nicergoline or its vehicle, DMSO Platelets were subjected to continuous magnetic stirring and held at 37°C À1 throughout EGTA (1 mM) was added before the cells were stimulated with 0.5 UÁmL thrombin pericellular hotspot and is consistent with the hypothesis that the MC is responsible for the efficient accumulation of Ca2+ at the pericellular Ca2+ hotspot Pretreatment with taxol partially reverses the inhibitory effect of nicergoline on thrombin-evoked dense granule secretion Previously, we have demonstrated a role for pericellular Ca2+ recycling in potentiating dense granule secretion from thrombin-stimulated cells (Sage et al., 2013) Experiments were performed to investigate whether nicergoline inhibits dense granule secretion As shown in Figure 8A, nicergoline pretreatment almost completely ablated thrombin-evoked dense granule secretion (15.1 ± 3.6% of control; n = 6; P < 0.05) In contrast to our finding with thrombin-evoked Ca2+ signalling, taxol pretreatment only partially reversed the inhibitory effect of nicergoline on granule secretion, with a partial inhibition still observed (67.0 ± 8.6% of control; n = 6; P < 0.05) These data therefore suggest that reversing the effects of nicergoline on pericellular Ca2+ recycling restores a significant dense granule secretion, in line with our previous data, suggesting that pericellular 242 British Journal of Pharmacology (2016) 173 234–247 Ca2+ recycling provides a secondary Ca2+ source which potentiates dense granule secretion (Sage et al., 2013) However, there is also a taxol-insensitive effect of nicergoline on dense granule secretion suggesting that there is a secondary action of this drug which prevents secretion even in the presence of normal pericellular Ca2+ recycling Further experiments were performed to examine whether nicergoline’s inhibition of dense granule secretion could be the cause, and not the consequence, of nicergoline’s inhibition of thrombin-evoked Ca2+ signalling These data suggested that nicergoline’s effect on thrombin-evoked Ca2+ signalling occurred upstream of to its effect on dense granule secretion (Supporting Information Figure S5) Pretreatment with taxol partially reverses the inhibitory effect of nicergoline on thrombin-evoked Ca2+ signalling elicited when platelets are stimulated in the presence of extracellular Ca2+ Experiments were also performed to examine whether nicergoline inhibits thrombin-evoked Ca2+ signalling in the presence of extracellular Ca2+ As shown in Figure 9A, Nicergoline inhibits platelet Ca2+ signalling BJP Figure 2+ 2+ Nicergoline treatment causes a dispersion of the pericellular Ca hotspot and inhibits the spread of the pericellular Ca signal across the platelet Cells were pretreated with either DMSO (A–E) or 100 μM nicergoline (F–I) for at 37°C in the presence of continuous magnetic stirring The platelets were then added to a poly-L-lysine-coated chambered slide and allowed to settle for Excess platelet suspension was removed, and the slides were 2+ + À1 washed twice with Ca -free HBS containing mM EGTA and 2.5 μM Fluo-4 K salt, and platelets were then stimulated by addition of 0.5 UÁmL thrombin Fluorescence was then monitored using an Olympus Fluoview FV1200 confocal microscope at 0.5 Hz for Scale bar indicates μm Images shown at selected points during the recording period after thrombin addition (A and F – s; B – 52.8 s; C – 162.8 s; D and E – 224.4 s; G – 70.4 s; H and I – 222.2 s) Images show either Fluo-4 fluorescence alone (A–D; F–H) or overlaid with the transmitted light image (E and I) (A–E) DMSO-treated cells demonstrate a pericellular hotspot (yellow arrow; panel B), which then spreads across the cell (panel C for top cell; panels C and D for bottom cell) Note the re-appearance of the pericellular hotspot in the same location it was initially seen in the top cell in panel D, in line with our previous findings (F–I) Nicergoline-treated cells show an increased fluorescence over time from a more dispersed cortical hotspot As 2+ can be seen in panel G, nicergoline-treated cells show no (top cell) or weak spreading of Ca (bottom cell; yellow arrow) in the pericellular region, which rarely crosses to the other side of the cell Figure Pretreatment with taxol partially reverses the inhibitory effect of nicergoline on thrombin (Thr)-evoked dense granule secretion Washed -1 platelet suspensions containing 1% (v v ) luciferin–luciferase were pretreated with either 100 μM taxol or its vehicle, DMSO, for 30 min, followed by a further 5-min incubation with either 100 μM nicergoline or its vehicle, DMSO Platelets were subjected to continuous magnetic stirring and held at 37°C throughout EGTA (1 mM) was added immediately before the start of the experiment, and platelets were subsequently À1 stimulated with 0.5 UÁmL thrombin The results presented are representative of six experiments pretreatment of platelets with nicergoline significantly inhibited thrombin-evoked rises in [Ca2+]cyt in the presence of mM extracellular Ca2+ (31.5 ± 5.1% of control; n = 13; P < 0.05; Figure 9) However, unlike most of the other findings previously discussed, taxol pretreatment was not able to fully prevent the inhibitory effects of nicergoline on thrombin-evoked Ca2+ signalling (58.9 ± 8.9% of control; n = 13; P < 0.05) Interestingly, this inability of taxol to fully reverse the inhibitory effect of nicergoline appeared to be related to an inability of the cells to maintain the prolonged, secondary plateau phase of the Ca2+ signal, as the initial Δ[Ca2+]cyt in the initial spike phase of the Ca2+ response was not affected by nicergoline in taxol pre-treated cells (101.2 ± 14.5% of control; n = 13; P > 0.05) However, the inhibitory effect on the plateau elicited by nicergoline treatment caused the integral of the full Ca2+ signal to be inhibited (51.1 ± 4.3% of control; n = 13; P < 0.05) Examination of the individual Ca2+ fluxes which contribute to the rise in [Ca2+]cyt showed that in the presence of extracellular Ca2+, nicergoline significantly inhibited thrombin-evoked removal of Ca2+ from the cell (18.9 ± 10.6% of control; n = 9; P < 0.05), Ca2+ release from intracellular stores (59.8 ± 6.1% of control; n = 8; P < 0.05) and initial opening of Ca2+-permeable ion channels as assessed by Mn2+ quench of Fura-2 fluorescence (40.0 ± 6.5% of control; n = 8; P < 0.05) However, pretreatment with taxol prevented nicergoline from eliciting a statistically British Journal of Pharmacology (2016) 173 234–247 243 BJP T Walford et al Figure 2+ Pretreatment with taxol partially reverses the inhibitory effect of nicergoline on thrombin (Thr)-evoked Ca signalling elicited when platelets are 2+ stimulated in the presence of extracellular Ca (A, E and F) Fura-2-loaded (A) or Fluo-5N-loaded (E and F) human platelets were pretreated with either 100 μM taxol (A and F) or its vehicle, DMSO (A and F), for 30 min, followed by a further 5-min incubation with either 100 μM nicergoline or 2+ its vehicle, DMSO Platelets were subjected to continuous magnetic stirring and held at 37°C throughout The extracellular Ca concentration À1 was raised to mM by addition of 800 μM CaCl2 before platelets were stimulated with 0.5 UÁml thrombin (B) Fura-2-loaded platelets 2+ À1 suspended in Ca -free HBS supplemented with 0.1 UÁmL apyrase were pre-incubated with either 100 μM taxol or its vehicle, DMSO, for 30 min, followed by a further 5-min incubation with either 100 μM nicergoline or its vehicle, DMSO Platelets were subjected to continuous mag2+ netic stirring and held at 37°C throughout Extracellular Ca was chelated by addition of 500 μM EGTA, followed 30 s later by mM MnCl2 PlateÀ1 lets were stimulated 30 s later with 0.5 UÁmL thrombin (C and D) Washed platelets suspended in a supplemented HBS containing 300 μM CaCl2 + and μM Rhod-5N K salt were pretreated with either 100 μM taxol (D) or its vehicle, DMSO (D), for 30 min, followed by a further 5-min incubation with either 100 μM nicergoline or its vehicle, DMSO Platelets were subjected to continuous magnetic stirring and held at 37°C throughout À1 Platelets were subsequently stimulated by addition of 0.5 UÁmL thrombin The results presented are representative of 6–13 experiments significant effect on each of these component Ca2+ fluxes (83.9 ± 24.7%, 115.3 ± 27.0% and 81.3 ± 12.8% of control respectively; n = 9, and respectively; all P > 0.05) These results suggest that pericellular Ca2+ recycling at the MC is required to potentiate the maximum amplitude of the initial thrombin spike by facilitating Ca2+ release and Ca2+ entry (probably by facilitating store depletion and thus triggering store-operated Ca2+ entry; Figure 8; Braun et al., 2008; Sage et al., 2013); however, nicergoline has taxol-insensitive effects on the plateau phase which prevents the full reinstatement of the Ca2+ signal seen under these conditions Previous work on platelets from patients with storage pool disorder has demonstrated that the plateau phase of thrombin-evoked Ca2+ responses requires the maintained opening of receptor-operated channels triggered by the secretion of ATP, ADP and 5-HT from the dense granules (Lages and Weiss, 1999) Given the observed taxol-insensitive 244 British Journal of Pharmacology (2016) 173 234–247 effects on dense granule secretion (Figure 8) and the effect only being observed in the presence of extracellular Ca2+, we suggest the possibility that nicergoline prevents autocrine-dependent signalling either via an inhibitory effect on secretion or via an inhibitory effect on a downstream receptor or channel Discussion In our previous work, we hypothesized that the nanojunction created at the membrane complex played a key role in platelet function by regulating Ca2+ release from intracellular stores (Sage et al., 2013) In this paper, we have demonstrated that nicergoline-induced disruption of platelet ultrastructure results in analogous inhibitory effects on thrombin-evoked Nicergoline inhibits platelet Ca2+ signalling Ca2+ signalling as previously reported for a range of experimental manipulations that prevented pericellular Ca2+ recycling (Sage et al., 2013) Previously, we used a quantitative analysis of our Ca2+ signalling data to demonstrate that the creation of a pericellular Ca2+ hotspot was most likely due to the action of the MC Here, we demonstrate that disruption of the normal organization of the OCS and DTS by nicergoline also disrupts the creation of this highly concentrated pericellular Ca2+ hotspot These results therefore support our hypothesis that the MC may be responsible for the creation of the pericellular Ca2+ hotspot Le Menn et al (1979) previously suggested that the change in distribution of the two component membrane systems of the MC appeared to be due to disruption of the cortical microtubule system In line with this, we have demonstrated that pre-treatment of platelets with taxol reverses the majority of the effects of nicergoline on thrombin-evoked Ca2+ signalling In addition, previous studies have also shown that the microtubule-disrupting agent colchicine was able to elicit a reduction in thrombin-evoked Ca2+ release to about 80% of control (Redondo et al., 2007) This effect is similar BJP in magnitude to that observed using nicergoline and suggests that normal microtubule structure is required to elicit normal Ca2+ signalling in human platelets These data suggest that the cortical microtubule bundle is likely to play a role in scaffolding the nanojunctions created by the close apposition of the OCS and DTS This is supported by previous electron microscope studies performed by Behnke (1967) and White (1968; 1972), who both showed an association of DTS, OCS and cortical microtubules at the MC Further investigations will need to examine how nicergoline is able to disrupt microtubule structure Recent work by Sadoul et al has suggested a number of possible mechanisms, including an alteration in the activity of microtubule-associated molecular motors such as myosin, kinesin and dynein, as well as the possibility of destabilizing the microtubule bundle through altering the post-translational modifications of tubulin (e.g acetylation or detyrosination; Sadoul et al., 2012; Diagouraga et al., 2013) Given the ability of high concentrations of nicergoline to inhibit Ca2+ signalling as well as agonist-evoked platelet adhesion, secretion and aggregation (Lanza et al., 1986), we hypothesize that the MC plays a key role in amplifying Figure 10 2+ Proposed model for the role of the membrane complex in thrombin (Thr)-evoked Ca signalling and its disruption by nicergoline (A) We propose that the membrane complex is held together by the cortical microtubule bundle, which helps hold the DTS in close apposition to the OCS One possible mechanism that may facilitate this interaction of the microtubules with the DTS is the known ability of stromal interaction molecule (STIM1) to interact with microtubules via an EB1-containing complex (Grigoriev et al., 2008) The membrane complex helps potentiate the initial 2+ 2+ phase of Ca entry by helping to transporting Ca out of the cell via the NCX in large quantities (Sage et al., 2013), allowing it to accumulate at 2+ 2+ high concentrations in the OCS (Figure 7) This Ca can recycle back into the cell contributing to the rise in [Ca ]cyt directly, as well as indirectly 2+ 2+ by facilitating store unloading via Ca -induced Ca release (Sage et al., 2011; Sage et al., 2013; Figure 9E and F), and the activation of the storeoperated channels via STIM1-dependent activation of an Orai1-containing ion channel (Braun et al., 2008; Figure 7B) Upon nicergoline-treatment, the DTS stays attached to the disorganized microtubule bundle (Le Menn et al., 1979) leading to its redistribution around the cell (Figure 3) 2+ 2+ and dissociation of the membrane complex disperses the Ca removal over a larger area of the OCS This prevents Ca accumulation within a 2+ small subregion of the OCS and thus reduces the pericellular Ca concentration, preventing its ability to recycle back into the cell where it can 2+ 2+ both directly and indirectly affect the initial phase of the Ca signal (Sage et al., 2013; Figure 9B, E and F) The reinstatement of Ca recycling 2+ by taxol treatment is able to reinstate some of the Ca signal and thus trigger a partial reversal of the effect of nicergoline on dense granule se2+ 2+ cretion Ca levels remain high due to the role of maintained Ca entry through receptor–operated channels brought about by dense granule secretion, as previously demonstrated by studying platelets from patients with storage pool disorders (Lages and Weiss, 1999) Nicergoline has a 2+ taxol-insensitive effect on this phase of the Ca signal by an additional effect on dense granule secretion and/or the signalling pathways triggered by ATP, ADP or 5-HT British Journal of Pharmacology (2016) 173 234–247 245 BJP T Walford et al platelet Ca2+ signals and modulating platelet activation This hypothesis is supported by a number of clinical case reports which have demonstrated that patients with abnormal membrane complexes suffer from bleeding disorders (Green et al., 1981; Meiamed et al., 1984; Canizares et al., 1990; Parker et al., 1993) Of particular note here is one report in which an inherited MC defect was found to be related to a deficit in thrombin-evoked Ca2+ signalling (Parker et al., 1993), in line with our findings presented here One limitation of our current findings is that nicergoline was also found to have a secondary taxol-insensitive effect on dense granule secretion (Figure 8) These data suggest that nicergoline can also inhibit dense granule secretion by a mechanism beyond its effect on pericellular Ca2+ recycling (Figure 10) From the data provided here, it seems most likely that either nicergoline’s effect on the cortical actin cytoskeleton or a possible effect on dense granule motility may account for the taxol-insensitive secretory defect observed As previous studies have shown that using jasplakinolide to induce actin polymerization and thickening of the cortical actin ring can also reduce dense granule secretion (Cerecedo et al., 2010), it is possible that the nicergoline-induced thickening of the cortical actin cytoskeleton might underlie the taxol-insensitive effects on granule secretion Alternatively, it is possible that nicergoline could directly or indirectly inhibit the activity of a molecular motor which could elicit a reorganization of the cortical microtubule bundle as well as affecting agonist-induced platelet granule motility For example, previous work has suggested that kinesin may play an important role in mediating platelet granule motility (Cerecedo et al., 2010), as well as in maintaining normal cortical microtubule bundle shape (Diagouraga et al., 2013) If nicergoline works through the inhibition of a molecular motor such as kinesin, this might potentially account for both the taxol-dependent and taxolindependent effects of nicergoline Further work will be needed to examine whether it is possible to separate out the taxol-sensitive and insensitive aspects of nicergoline on platelet function Conclusion From the data presented here and elsewhere (Le Menn et al., 1979; Lanza et al., 1986), we suggest that nicergoline provides an initial proof-of-concept that a MC-disrupting drug could be developed and would be effective as an anti-thrombotic We propose that by studying the molecular mechanisms by which nicergoline interferes with the platelet ultrastructure, we may be able to better understand the molecular structures underlying the formation of the MC As such, nicergoline may be useful as a prototype compound which can be used to help identify new pharmacological targets for the development of more selective MC-disrupting drugs here T W was supported by a Vacation Studentship from The Physiological Society F M was supported by a PhD studentship from the British Heart Foundation (FS/12/48/ 29719) A G S H was supported by a research grant from the Physiological Society Author contributions Experiments were designed by A G S H T W., F I M and A G S H collected and analysed the data The manuscript was written by T W., F I M and A G S H All authors participated in manuscript revision and have given final approval for publication Conflict of interest Authors declare that they have not any conflict of interest References Alexander SPH, 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 J Pharmacol 170: 1459–1581 Alexander SPH, Benson HE, Faccenda E, Pawson AJ, Sharman JL, Spedding M et al (2013b) The Concise Guide to PHARMACOLOGY 2013/14:ligand-gated ion channels Br J Pharmacol 170: 1582–1606 Alexander SPH, Benson HE, Faccenda E, Pawson AJ, Sharman JL, Spedding M et al (2013c) The Concise Guide to PHARMACOLOGY 2013/14: transporters Br J Pharmacol 170: 1706–1796 Behnke O (1967) Electron microscopic observations on the membrane systems of the rat blood platelet Anat Rec 158: 121–138 Blaustein MP, Lederer WJ (1999) Sodium/calcium exchange: its physiological implications Physiol Rev 79: 763–854 Braun A, Varga-Szabo D, Kleinschnitz C, Pleines I, Bender M, Austinat M et al (2008) Orai1 (CRACM1) is the platelet SOC channel and essential for pathological thrombus formation Blood 113: 2056–2063 Canizares C, Vivar N, Grijalva J (1990) Thrombocytopathy due to a defect of the platelet membrane complex Acta Haematol 83: 99–104 Cerecedo D, Cisneros B, Mondragon R, Gonzalez S, Galvan IJ (2010) Actin filaments and microtubule dual-granule transport in human adhered platelets: the role of alpha-dystrobrevins Br J Haematol 149: 124–136 Diagouraga B, Grichine A, Fertin A, Wang J, Khochbin S, Sadoul K (2013) Motor-driven marginal band coiling promotes cell shape change during platelet activation J Cell Biol 204: 177–185 Acknowledgements Green D, Ts’ao CH, Cohen I, Rossi EC (1981) Haemorrhagic thrombocytopathy associated with dilatation of the platelet— membrane complex Br J Haematol 48: 595–600 We would like to thank Dr Stewart Sage and Dr Mike Mason for their help in performing the dense granule secretion assays, as well as for valuable discussions of the data presented Grigoriev I, Gouveia SM, van der Vaart B, Demmers J, Smyth JT, Honnappa S et al (2008) STIM1 is a MT-plus-end-tracking protein involved in remodelling of the ER Curr Biol 18: 177–182 246 British Journal of Pharmacology (2016) 173 234–247 Nicergoline inhibits platelet Ca2+ signalling Harper AGS, Mason MJ, Sage SO (2009) A key role for dense granule 2+ secretion in potentiation of the Ca signal arising from store-operated calcium entry in human platelets Cell Calcium 45: 413–420 + 2+ Harper AGS, Sage SO (2007) A key role for reverse Na /Ca exchange 2+ influenced by the actin cytoskeleton in store-operated Ca entry in human platelets: evidence against the de novo conformational coupling hypothesis Cell Calcium 42: 606–617 Heemskerk JWM, Mattheij NJA, Cosemans JMEM (2013) Plateletbased coagulation: different populations, different functions J Thromb Haemost 11: 2–11 BJP Sage SO, Pugh N, Farndale RW, Harper AGS (2013) Pericellular Ca 2+ recycling potentiates thrombin-evoked Ca signals in human platelets Physiol Rep 1: e00085 2+ Sage SO, Pugh N, Mason MJ, Harper AGS (2011) Monitoring the 2+ intracellular store Ca concentration in agonist-stimulated, intact human platelets by using Fluo-5N J Thromb Haemost 9: 540–551 Van Breemen C, Fameli N, Evans AM (2013) Pan-junctional sarcoplasmic 2+ reticulum in vascular smooth muscle: nanospace Ca transport for 2+ site- and function-specific Ca signalling J Physiol 591: 2043–2054 Italiano JE Jr, Bergmeier W, Timari S, Falet H, Hartwig JH, Hoffmeister KM et al (2003) Mechanisms and implications of platelets discoid shape Blood 101: 4789–4796 van Nispen tot Pannerden HE, van Dijk SM, Du V, Heijnen HFG (2009) Platelet protein disulfide isomerase is localized in the dense tubular system and does not become surface expressed after activation Blood 114: 4738–4740 Lages B, Weiss HJ (1999) Secreted dense granule adenine nucleotides promote calcium influx and the maintenance of elevated cytosolic calcium levels in stimulated human platelets Thromb Haemost 81: 286–292 van Nispen tot Pannerden H, de Haas F, Geerts W, Posthuma G, van Dijk S, Heijnen HFG (2010) The platelet interior revisited: electron tomography reveals tubular α-granule subtypes Blood 116: 1147–1156 Lanza F, Cazenave JP, Beretz A, Sutter-Bay A, Kretz JG, Kieny R (1986) Potentiation by adrenaline of human platelet activation and the inhibition by the alpha-adrenergic antagonist nicergoline of platelet adhesion, secretion and aggregation Agents Actions 18: 586–595 White JG (1968) Effects of colchicine and vinca alkaloids on human platelets II Changes in the dense tubular system and formation of an unusual inclusion in incubated cells Am J Pathol 53: 447–461 Le Menn R, Migne J, Probst-Djovakovich RJ (1979) Ultrastructural study on the effect of an inhibition of platelet aggregation Arzneimittelforschung 29: 1278–1282 Meiamed I, Djaldetti M, Joshua H, Seligsohn U (1984) Association of the hemophilia A carrier state and Hemorrhagic thrombocytopathy with dilatation of the platelet membrane complex Acta Haematol 71: 381–387 Parker RI, Bray GL, McKeown LP, White JG (1993) Failure to mobilize intracellular calcium in response to thrombin in a patient with familial thrombocytopathy characterized by macrothrombocytopenia and abnormal platelet membrane complexes J Lab Clin Med 122: 441–449 Pawson AJ, Sharman JL, Benson HE, Faccenda E, Alexander SP, Buneman OP et al (2014) The IUPHAR/BPS Guide to PHARMACOLOGY: an expert-driven knowledgebase of drug targets and their ligands Nucl Acids Res 42 (Database Issue): D1098–106 Redondo PC, Harper AGS, Sage SO, Rosado JA (2007) Dual role of tubulin-cytoskeleton in store-operated calcium entry in human platelets Cell Signal 19: 2147–2154 Rink TJ, Sage SO (1990) Calcium signalling in human platelets Ann Rev Physiol 52: 429–447 Sadoul K, Wang J, Diagouraga B, Vitte AL, Buchou T, Rossini T et al (2012) HDAC6 controls the kinetics of platelet activation Blood 120: 4215–4218 White JG (1972) Interaction of membrane systems in blood platelets Am J Pathol 66: 295–312 Supporting Information Additional Supporting Information may be found in the online version of this article at the publisher’s web-site: http://dx.doi.org/10.1111/bph.13361 Figure S1 Nicergoline does not itself elicit a Ca2+ signal, but does trigger a small reduction in the baseline [Ca2+]cyt in both the presence and absence of extracellular Ca2+ Figure S2 Nicergoline causes a slight thickening of the cortical F-actin layer without altering the polymerisation state of F-actin within the platelets Figure S3 Nicergoline triggers reorganisation of the subcellular location of the intracellular Ca2+ stores Figure S4 Nicergoline elicits no additional inhibitory effect when pericellular Ca2+ accumulation is prevented by pretreatment with an NCX inhibitor Figure S5 Nicergoline-induced inhibition of thrombinevoked Ca2+ signalling is a cause, and not a consequence, of the nicergoline-related inhibition of dense granule secretion British Journal of Pharmacology (2016) 173 234–247 247 BJP DOI:10.1111/bph.13335 www.brjpharmacol.org British Journal of Pharmacology LETTER TO THE EDITOR Correspondence From mouse to man: predicting biased effects of beta-blockers in asthma Brian J Lipworth, Scottish Centre for Respiratory Research, University of Dundee, Ninewells Hospital and Medical School, Dundee, DD1 9SY, UK E-mail: b.j.lipworth@dundee.ac.uk - Received 25 August 2015 Accepted B J Lipworth, W J Anderson and P M Short September 2015 Scottish Centre for Respiratory Research, Ninewells Hospital and Medical School, Dundee, UK LINKED ARTICLES This article is a Commentary on Thanawala VJ, Valdez DJ, Joshi R, Forkuo GS, Parra S, Knoll BJ, Bouvier M, Leff P and Bond RA (2015) Beta-blockers have differential effects on the murine asthma phenotype Br J Pharmacol 172: 4833–4846 doi: 10.1111/ bph.13253 The authors reply in Bond RA, Thanawala VJ, Parra S and Leff P (2016) Differences in asthma study models and the effectiveness of β2-adrenoceptor ligands: response to Lipworth et al Br J Pharmacol 173: 250–251 doi: 10.1111/bph.13334 Tables of Links TARGETS LIGANDS Enzymes Adrenaline Nadolol ERK1/2 Histamine Propranolol Methacholine 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 Concise Guide to PHARMACOLOGY 2013/14 (Alexander et al., 2013) We read with interest the elegant data from Thanawala et al (2015) in ovalbumin-sensitized adrenaline-deficient or wild-type mice, which suggested differential effects of biased signalling with propranolol and nadolol We were particularly intrigued by the observation in adrenaline-deficient mice that there was a disconnect between propranolol and nadolol in restoring the asthma phenotype compared with controls It is tempting to simplistically extrapolate these data in mice to what might happen in human subjects with asthma in order to explain the negative effects of propranolol on airway hyper-responsiveness (AHR), reported in two separate placebo-controlled double-blind trials in patients receiving inhaled corticosteroids (ICS), which were powered to detect a one doubling dilution difference in the provocative concentration of methacholine (n = 18) or histamine (n = 16) to produce a 20% fall in forced expiratory volume in 1s (i.e the PC20 FEV1 threshold) (Short et al., 2013; Anderson et al., 2014) In this regard, in patients with persistent asthma, the PC threshold for FEV1 is closely related to PC threshold for airway resistance (Short et al., 2015) 248 British Journal of Pharmacology (2016) 173 248–249 Upon close inspection of the data for methacholine AHR (Thanawala et al., 2015), the provocative concentration to induce a 100% increase (PC100 threshold) in airway resistance was unaltered in wild-type mice (n = 6) treated with propranolol in contrast to an increase with nadolol (n = 7) The blunting of methacholine AHR with nadolol which was statistically significant (P < 0.05) amounted to approximately a 0.6 doubling dilution shift compared with vehicle-treated mice (n = 10) Such an effect in mice with nadolol on AHR would be considered clinically irrelevant in human patients as it less than the minimal important difference of one doubling dilution shift in PC threshold It is therefore difficult to extrapolate the magnitude of this effect with nadolol on methacholine AHR in mice to what has previously been reported in two unblinded studies with nadolol in human asthmatic subjects which amounted to an approximate two doubling dilution shift in methacholine PC20, albeit in mild intermittent asthmatics who were not taking ICS (Hanania et al., 2008; Hanania et al., 2010) © 2015 The British Pharmacological Society BJP It is however unclear how the relative mg per body weight dose of propranolol in mice (80–140 mg·LÀ1 in water) equates to that in humans (80 mg slow release tablet per day) Moreover, if propranolol at usual therapeutic doses of 80 mg per day does indeed confer arrestin-independent, partial agonist activity at the ERK1/2 activation pathway in humans, then one might expect to see an increase in Th2-mediated inflammatory biomarkers For example, in persistent asthmatics, there was no worsening in eosinophils, eosinophilic cationic protein or exhaled breath NO when oral propranolol 80 mg per day was added to a low dose of ICS, while a higher dose of ICS in conjunction with oral placebo produced further suppression of the same Th2 biomarkers (Anderson et al., 2014) Moreover, asthma control and disease-specific quality of life were also unaltered by propranolol (Short et al., 2013; Anderson et al., 2014) In order to properly confirm the putative beneficial effects of biased inhibitory signalling in mice, this will require a placebo-controlled trial to demonstrate clinically relevant improvements in methacholine PC20, inflammatory markers and asthma control with nadolol on top of existing ICS therapy in persistent asthma The placebo-controlled clinical trial (clinicaltrials.gov NCT01804218) evaluating effects of nadolol in ICS naïve mild intermittent asthmatics will unfortunately not answer this clinically important question Conflict of interest BJL has received previous grant support from the Chief Scientist Office, Scotland, to evaluate effects of propranolol in patients with persistent treated asthma BJL has also received unrestricted grant support from Chiesi, Meda, Almirall and Teva to evaluate small airways in persistent asthma and COPD; as well as multicentre pharmaceutical support from Astra Zeneca, Teva, Janssen and Roche In addition, BJL has received personal payment for consultancy and advisory boards with the following pharmaceutical companies: Astra Zeneca, Chiesi, Teva, Boehringer Ingelheim and Meda BJL has also received personal payment for giving speaker talks with Chiesi, Teva, Meda and Mitsubishi Tanabe as well as support to attend educational meetings from Chiesi, Boehringer Ingelheim and Teva WJA and PMS have no conflict of interest References Alexander SPH, Benson HE, Faccenda E, Pawson AJ, Sharman JL, Spedding M et al (2013) The Concise Guide to PHARMACOLOGY 2013/14: Enzymes Br J Pharmacol 170: 1797–1867 Anderson WJ, Short PM, Williamson PA, Manoharan A, Lipworth BJ (2014) The inverse agonist propranolol confers no corticosteroidsparing activity in mild-to-moderate persistent asthma Clin Sci (Lond) 127: 635–643 Hanania NA, Mannava B, Franklin AE, Lipworth BJ, Williamson PA, Garner WJ et al (2010) Response to salbutamol in patients with mild asthma treated with nadolol Eur Respir J 36: 963–965 Hanania NA, Singh S, El-Wali R, Flashner M, Franklin AE, Garner WJ et al (2008) The safety and effects of the beta-blocker, nadolol, in mild asthma: an open-label pilot study Pulm Pharmacol Ther 21: 134–141 Pawson AJ, Sharman JL, Benson HE, Faccenda E, Alexander SP, Buneman OP et al (2014) The IUPHAR/BPS Guide to PHARMACOLOGY: an expert-driven knowledge base of drug targets and their ligands Nucl Acids Res 42 (Database Issue): D1098–1106 Short PM, Anderson WJ, Manoharan A, Lipworth BJ (2015) Usefulness of impulse oscillometry for the assessment of airway hyperresponsiveness in mild-to-moderate adult asthma Ann Allergy Asthma Immunol 115: 17–20 Short PM, Williamson PA, Anderson WJ, Lipworth BJ (2013) Randomized placebo-controlled trial to evaluate chronic dosing effects of propranolol in asthma Am J Respir Crit Care Med 187: 1308–1314 Thanawala VJ, Valdez DJ, Joshi R, Forkuo GS, Parra S, Knoll BJ et al (2015) Beta-blockers have differential effects on the murine asthma phenotype Br J Pharmacol 172: 4833–4846 British Journal of Pharmacology (2016) 173 248–249 249 DOI:10.1111/bph.13334 www.brjpharmacol.org British Journal of Pharmacology BJP LETTER TO THE EDITOR Correspondence Differences in asthma study models and the effectiveness of β2-adrenoceptor ligands: response to Lipworth et al Richard A Bond, Department of Pharmacological and Pharmaceutical Sciences, University of Houston, Houston, TX, USA E-mail: rabond@uh.edu - Received 14 September 2015 Accepted 14 September 2015 Richard A Bond1, Vaidehi J Thanawala1,2, Sergio Parra3 and Paul Leff4 Department of Pharmacological and Pharmaceutical Sciences, University of Houston, Houston, TX, USA, 2Department of Integrative Biology and Pharmacology, University of Texas Health Science Center at Houston, Houston, TX USA, 3Vapogenix Inc., Houston, TX, USA, and 4Consultant in Pharmacology, Cheshire, UK LINKED ARTICLES This article is a reply to Lipworth BJ, Anderson WJ and Short PM (2016) From mouse to man: predicting biased effects of betablockers in asthma Br J Pharmacol 173: 248–249 doi: 10.1111/bph.13335, commenting on Thanawala VJ, Valdez DJ, Joshi R, Forkuo GS, Parra S, Knoll BJ, Bouvier M, Leff P and Bond RA (2015) Beta-blockers have differential effects on the murine asthma phenotype Br J Pharmacol 172: 4833–4846 doi: 10.1111/bph.13253 Tables of Links TARGETS GPCRs a Nadolol 2-adrenoceptors Enzymes LIGANDS Propranolol b PKA 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 a b permanently archived in the Concise Guide to PHARMACOLOGY 2013/14 ( Alexander et al., 2013a,b) We thank Lipworth et al (2016) for their positive comments about our work We agree with the letter that animal models of a disease have many limits compared with the human disease and can have additional substantial shortcomings such as dosing comparisons and not being representative of the spectrum of disease severity (Lipworth et al., 2016) The reason we discussed the different outcomes of clinical trials using propranolol or nadolol in our study is that, given the different signalling profiles of these compounds (Wisler et al., 2007; Walker et al., 2011), there is no reason to expect the same results or assume the generalizations that were discussed in the original propranolol report (Short et al., 2013) It seems to us the published differences in signalling would be a highly likely explanation for the differences 250 British Journal of Pharmacology (2016) 173 250–251 between clinical trials using propranolol and nadolol (Hanania et al., 2008; Short et al., 2013), but this alternative was not considered in the original propranolol study (Short et al., 2013) Our study also shows that the signalling differences first documented in cell-based assays also have in vivo relevance in an animal model of airway disease and are consistent with the differences observed in the clinical trials using propranolol or nadolol (Thanawala et al., 2015) Indeed, we are now fortunate enough to have a situation where the spectrum of data includes mathematical modelling, cell-based studies, in vivo studies in an animal model of disease and clinical trials, and all are supportive of current receptor theory (Wisler et al., 2007; Hanania et al., 2008; Short et al., 2013; Thanawala et al., 2015) However, we agree © 2015 The British Pharmacological Society ... Sci U S A 10 5: 211 1– 211 6 doi :10 .10 73/pnas.0 710 22 810 5 Chen Y-R, Zweier JL (2 014 ) Cardiac mitochondria and reactive oxygen species generation Circ Res 11 4: 524–537 doi :10 .11 61/ CIRCRESAHA 11 4.300559... Circulation 10 6: 212 5– 213 1 doi :10 .11 61/ 01. CIR.0000034049. 611 81 Lijnen PJ, Petrov VV, Fagard RH (2000) Induction of cardiac fibrosis by transforming growth factor-beta (1) Mol Genet Metab 71: 418 –435... 282: 3 918 –3928 doi :10 .10 74/jbc.M608867200 British Journal of Pharmacology (2 016 ) 17 3 3 14 11 BJP J Heger et al Biesemann N, Mendler L, Wietelmann A, Hermann S, Schäfers M, Krüger M et al (2 014 )