病理生理学病理生理学 (6).pdf

Modelling ischemia-reperfusion injury(IRI)in vitrousing metabolically matured induced pluripotentstem cell-derived cardiomyocytesAlejandro Hidalgo,1,2Nick Glass,1Dmitry Ovchinnikov,2Seung-Kwon Yang,3Xinli Zhang,4Stuart Mazzone,3Chen Chen,4Ernst Wolvetang,2,a)andJustin Cooper-White1,5,6,a)1Tissue Engineering and Microfluidics Group,Australian Institute for Bioengineeringand Nanotechnology,The University of Queensland,Brisbane 4072,Australia2Stem Cell Engineering Group,Australian Institute for Bioengineering and Nanotechnology,The University of Queensland,Brisbane 4072,Australia3Laboratory for Respiratory Neuroscience and Mucosal Immunity,School of BiomedicalSciences,The University of Queensland,St.Lucia 4072,Australia4Laboratory for Endocrinology and Metabolism,School of Biomedical Sciences,The University of Queensland,St.Lucia 4072,Australia5School of Chemical Engineering,The University of Queensland,Brisbane 4072,Australia6CSIRO,Manufacturing Flagship,Biomedical Manufacturing Program,Clayton,Victoria 3810,Australia(Received 17 August 2017;accepted 27 February 2018;published online 20 March 2018)CoronaryinterventionfollowingST-segmentelevationmyocardialinfarction(STEMI)is the treatment of choice for reducing cardiomyocyte death but paradoxi-cally leads to reperfusion injury.Pharmacological post-conditioning is an attractiveapproach to minimize Ischemia-Reperfusion Injury(IRI),but candidate drugs identi-fied in IRI animal models have performed poorly in human clinical trials,highlightingthe need for a human cell-based model of IRI.In this work,we show that when weimposed sequential hypoxia and reoxygenation episodes mimicking the ischemia(I)and reperfusion(R)events to immature human pluripotent stem cell-derived cardio-myocytes(hPSC-CMs),they display significant hypoxia resistance and minimal celldeath(?5%).Metabolic maturation of hPSC-CMs for 8days substantially increasedtheir sensitivity to changes in oxygen concentration and led to up to?30%cell deathpost-hypoxia and reoxygenation.To mimic the known transient changes in the intersti-tial tissue microenvironment during an IRI event in vivo,we tested a new in vitro IRImodel protocol that required glucose availability and lowering of media pH during theischemic episode,resulting in a significant increase in cell death in vitro(?60%).Finally,we confirm that in this new physiologically matched IRI in vitro model,phar-macological post-conditioning reduces reperfusion-induced hPSC-CM cell death by50%.Our results indicate that in recapitulating key aspects of an in vivo IRI event,ourin vitro model can serve as a useful method for the study of IRI and the validation andscreening of human specific pharmacological post-conditioning drug candidates.VC2018 Author(s).All article content,except where otherwise noted,is licensedunderaCreativeCommonsAttribution(CCBY)license(http:/creativecommons.org/licenses/by/4.0/).https:/doi.org/10.1063/1.5000746INTRODUCTIONCurrent clinical practice aimed at limiting injury following STEMI involves prompt man-agement with thrombolytic therapy or primary percutaneous coronary intervention(PPCI).1,2a)Authors to whom correspondence should be addressed:e.wolvetanguq.edu.au and j.cooperwhiteuq.edu.au.Telephone:33463838.2473-2877/2018/2(2)/026102/14VCAuthor(s)2018.2,026102-1APL BIOENGINEERING 2,026102(2018)These therapeutic interventions however cause ischemia reperfusion injury(IRI)that exacer-bates myocardial cell death,which is one of the main contributors to coronary heart disease(CHD).1,3IRI has been widely studied in in vivo and in vitro animal models,4,5revealingimportant roles for local acidification,autophagy,reactive oxygen species(ROS)-production,mitochondrial-induced cell death,and associated adenosine triphosphate(ATP)-decline.46Pharmacological post-conditioning(PPC)in previous studies has identified reoxygenation proto-cols and compounds that,when administered after the ischemic event,minimise IRI-inducedmyocardial injury.79However,a large number of potential PPC-compounds identified usingthese models have largely failed to translate into successful human clinical trials.7Intrinsic dif-ferences between human and animal heart physiology and/or differences in experimental designare the major source of error.To investigate the molecular mechanisms underlying IRI and dis-cover better PPC drugs,a human cell-based model that recapitulates the transient microenviron-mental changes experienced by cardiomyocytes during an IRI is highly desirable.10Human pluripotent stem cell-derived cardiomyocytes(hPSC-CMs)have enabled human dis-ease modelling and screening of pharmacologically relevant drugs in vitro.1114However,hPSC-derived cardiomyocytes resemble a more fetal-like state,which relies on glycolysisand therefore allows better survival in low oxygen environments.Due to this limitation,hPSC-derived cardiomyocytes have thus not been used to study IRI to date.11,1519Immediatelyafter birth,cardiomyocytes display a rapid increase in mitochondrial mass and an accompanyingmetabolic shift from anaerobic glycolysis to mitochondrial b-oxidation of fatty acids,2022andthis increased reliance on mitochondrial respiration is thought to underlie the susceptibility ofadult cardiomyocytes to a hypoxic insult and reperfusion.20,2325Herein,we detail a novel in vitro human cell-based IRI model that mimics the knownphysiological changes in terms of temporal transients in oxygen concentration,local pH,andglucose(glycogen)availability experienced by cardiomyocytes during an IRI event.We firstshow that,post-induction of hPSCs to a cardiomyocyte fate,a short period of metabolic shiftfrom glycolysis to oxidative phosphorylation of fatty acids is critical in promoting cardiomyo-cyte maturation and rendering the hPSC-CMs sensitive to sequential exposure to hypoxiaand reoxygenationour first pass at mimicking an IRI insult.We next show that maturedhPSC-CM sensitivity to such an insult is further enhanced by recapitulating other microenviron-mental pathological conditions present during the ischemic episode in the myocardium,result-ing in the creation of a new in vitro IRI model.We finally demonstrate that this new in vitroIRI model can be used as a simple and scalable in vitro screening platform for the validation ofknown and discovery of novel PPC drugs.RESULTSImmature human pluripotent stem cell-derived cardiomyocytes are resistant to a simpli-fied model of an IRI eventTo model an ischemic-reperfusion insult in vitro,H9-NCX1cardiomyocytes(hereafterNCX1-CMs)were exposed to sequential periods of hypoxia and reoxygenation.First,mainte-nance media were exchanged with oxygen depleted,glucose-free medium,and the cells wereincubated in a 0%oxygen incubator for 2hthis constitutes the ischemic event.28To modelthe reperfusion event,the NCX1-CMs were next exposed to oxygenated glucose-rich mediumand transferred to a normoxic incubator for 1 or 4h.This temporal sequence of events mimicsa best-case intervention scenario for acute myocardial infarction(AMI)patients.Under thisimposed simplified mimic of an IRI event,quantification of cell death by cleaved-Caspase 3immunostaining revealed the absence of cell death after hypoxia-reoxygenation and persistenceof beating activity regardless of the timeframe of hypoxia(14h)(supplementary material,Fig.1).These findings were expected to be the result of newly derived hPSC-CMs resembling fetalcardiomyocytes,a cell type able to cope much better with low oxygen environments than CMsof the adult myocardium.29026102-2Hidalgo et al.APL Bioeng.2,026102(2018)Metabolic maturation of human pluripotent stem cell derived cardiomyocytesWe hypothesized that hPSC-CMs with metabolic phenotypes comparable to postnatal cardi-omyocytes would be more susceptible to IRI-induced cell death.To this end,we subjectedimmature hPSC-CMs to a metabolic maturation protocol previously reported to increase CMrespiration rates.30Depriving hPSC-CMs of glucose while providing galactose promotes mito-chondrial beta-oxidation of long-chain fatty acids(palmitate and oleic acid),which mimics themetabolic switch from glycolysis to oxidative phosphorylation observed after birth,31,32and fos-ters cardiomyocyte metabolic maturation.30Immature green fluorescent protein(GFP)-sortedNCX1-CM cells were cultured in either standard glucose-containing maintenance media here-after Standard Media(SM)or glucose free,galactose supplemented fatty acid(palmitate andoleic)-rich media hereafter fatty acid(FA)for 8days.Because expression of Sodium-Calciumexchanger 1(NCX1)decreases during CM maturation,33we used GFP expression driven by acardiomyocyte-specific NCX1 promoter as an initial readout of maturation.ImmunofluorescenceFig.1(b)and flow cytometry Fig.1(c)analyses revealed a significant reduction in mean fluo-rescence intensity(MFI)of GFP expression in CMs exposed to the FA culture condition ascompared to the standard medium.Whilst the immunofluorescence images confirm this clearreduction in GFP expression as a result of exposure to the FA media,from our flow cytometrydata,it appears that the reduction in overall MFI is supported by the appearance of subpopula-tions of cells that have varying levels of strongly decreased fluorescence,whereas some cellsremain similar to those cells in SM,suggesting that across the population of cells,not all cellsare transitioning at the same time,likely due to differences in the initial cell state.Shortly afterbirth,postnatal cardiomyocytes undergo an increase in mitochondrial mass and intracellularlocalization of mitochondria also shifts from the peri-nuclear space to a one that spreads acrossthe cytoplasm.In vivo CM maturation and PGC1A-enforced maturation21of hPSC-derived CMsin vitro are similarly accompanied by an increase in the mitochondrial content and cytoplasmicrelocalization of mitochondria.To evaluate if our maturation protocol is capable of inducing anincrease in mitochondrial mass and the re-localization phenotype in hPSC-CMs,we nextlabelled SM-and FA-treated cultures with mitotracker-FM.FA-treated cardiomyocytes werereplated onto aligned microgrooved substrates to promote cell elongation for visualisation ofmitochondrial intracellular localization.Mitochondria occupied a large volume of the intracellularFIG.1.Immature NCX1-CMs are unaffected by hypoxia-reoxygenation requiring maturation by modulation of substrateavailability.(a)FACS dot plots for NCX1-CM sorting based on GFP expression.(b)and(c)Decreased GFP expression inNCX1-CMs during 8days in Standard media(SM)and Fatty Acid media(FA)treatment(quantified by FACS)supportscardiomyocyte maturation with FA treatment.The Blue Mean Fluorescence Intensity(MFI)peak represents the SM cardio-myocytes,and the Red MFI peak represents the FA treated cells.(d)and(e)Mitotracker labelling of NCX1-CM culturesfollowing 8days of culture in SM and FA media(quantified by FACS)revealed increased cytoplasmic mitotracker stainingand a 40%increase in mitochondrial mass in FA treated cultures(n3).026102-3Hidalgo et al.APL Bioeng.2,026102(2018)space(supplementary material,Fig.2 and Video 2).Flow cytometry analysis showed that therewas a 40%increase in mitochondrial mass in FA-treated CMs Fig.1(d).In addition,immuno-fluorescence staining showed that mitochondria are localized to the perinuclear space in SM-treated CMs,while in FA-treated CMs,larger mitochondria are arranged throughout the cyto-plasm Fig.1(e).We next examined mRNA expression of a set of sarcomeric cardiac proteins in SM andFA-treated NCX1-CMs by quantitative polymerase chain reaction(qPCR).In the FA-treatedcultures,we observed significantly increased expression of MHY7 and MLY2(MLC2v)anddownregulation of MLY7(MLC2a)but no significant differences in NKX-2.5,MYH6,or TNNT2(cTnT)expression between SM and FA-treated CMs Fig.2(a).The elevated MLC2v expres-sion34and increased ratio between b-MHC(MYH7)and a-MHC(MYH6)35are consistent witha maturing CM phenotype.17FA-treated cells also exhibited robust upregulation of PGC-1amRNA Fig.2(a),a gene directly involved in mitochondrial biogenesis and maturation ofhPSC-CMs.21The cardiac chamber specification marker natriuretic peptide type A(NPPA)isalso upregulated in FA-treated CMs compared to SM-treated cells.Flow cytometric analysisrevealed an increase in the cardiomyocyte size(Forward Scatter)in FA-treated cultures Fig.2(b),as expected for physiological hypertrophy in maturing cardiomyocytes,and elevated sarco-meric protein expression.We further found increased expression of MLC2v protein in FA-treatedcultures that persisted up to 24days as indicated by immunofluorescence staining Fig.2(c).Since adult cardiomyocytes have a slower beating frequency than fetal cardiomyocytes,wequantified the beating frequency of SM and FA-treated cardiomyocytes by intracellular calciumtransients.To this end,we employed a footprint-free human induced pluripotent stem cell(iPSC)line C32 that was transduced with a lentivirus carrying the GCaMP6f calcium biosen-sor36(see Methods section and real-time output in supplementary material,Video 3).Calciumflux analysis revealed that FA-treated CM cultures exhibited clear differences in intracellularcalcium handling example traces shown in Fig.3(a),line scan images of GCaMP6f mature sin-gle cardiomyocytes in supplementary material,Fig.3),with an increase in the rising slope,ashorter time-to-peak,a shorter time-to-base,and a decrease in beating frequency from 0.9FIG.2.Characterisation of changes in gene and protein expression of NCX1-CM phenotypic maturation.(a)Gene expres-sion analysis of SM vs FA-treated NCX1-CMs by qPCR(n3)after 8days of metabolic maturation,Gene expression rel-ative to GAPDH(DCt).(b)FACS-based quantification of cTnT expression and size of SM versus FA-treated NCX1-CMs(n3)after 8days in culture.(c)Immunofluorescence detection of MLC2v(red)expression in FA-treated NCX1-CMs(far RHSenlargement of the merged image).026102-4Hidalgo et al.APL Bioeng.2,026102(2018)beats/s vs 0.45 beats/s,but no change in the amplitude Fig.3(b).Collectively,the featuresobserved in the calcium transients in FA-treated CMs are consistent with a more mature CMphenotype than those observed in SM-treated CMs.To confirm these data,we collected electrophysiological recordings from SM and FA-treated NCX1-CM cardiomyocytes.Figure 3(c)shows traces of patch-clamp recordings com-paring SM and FA-treated cardiomyocytes with ventricular electrophysiological characteristics.FA-treated cardiomyocytes exhibit higher action potential depolarization(APD50 and APD90)and lower frequency compared to the SM-treated cardiomyocytes Fig.3(d).We conclude that,compared to SM-treated,FA-treated CMs exhibit mitochondrial,cellsize,protein expression,mRNA expression,calcium handling,and membrane action potentialcharacteristics consistent with a mature cardiomyocyte phenotype.Mimicking physiological changes in pH and glucose availability during IRI rendermatured hPSC-derived cardiomyocytes susceptible to ischemia-reperfusion in vitroWe nex