RENÉE G.C. MAAS POTENTIALS OF STEM CELL-DERIVED CARDIOMYOCYTES ‘FROM DISEASE MODELING TO THERAPEUTIC STRATEGIES’
POTENTIALS OF STEM CELL-DERIVED CARDIOMYOCYTES ‘FROM DISEASE MODELING TO THERAPEUTIC STRATEGIES’ De Potentie van Pluripotente Stamcel Gedifferentieerde Hartspiercellen ‘Van Ziekte Nabootsing tot Therapeutische Toepassingen’ Renée G. C. Maas
Potentials of Stem Cell derived cardiomyocytes. From disease modelling to therapeutic strategies. PhD Thesis with summary in Dutch. Utrecht University. Cover design: Ridderprint | www.ridderprint.nl Layout: Ridderprint | www.ridderprint.nl Print: Ridderprint | www.ridderprint.nl ISBN: 978-94-6483-793-3 Copyright © Renée G.C. Maas 2024. All rights reserved. No parts of this thesis may be reproduced, stored in a retrieval system of any nature, or transmitted in any form or by any means, without prior written consent of the author. The copyright of the articles that have been published has been transferred to the respective journals.
POTENTIALS OF STEM CELL-DERIVED CARDIOMYOCYTES ‘FROM DISEASE MODELING TO THERAPEUTIC STRATEGIES’ De Potentie van Pluripotente Stamcel Gedifferentieerde Hartspiercellen ‘Van Ziekte Nabootsing tot Therapeutische Toepassingen’ (Met een samenvatting in de Nederlandse taal) Proefschrift ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de rector magnificus, Prof. dr. H.R.B.M. Kummeling, ingevolge het besluit van het college voor promoties in het openbaar te verdedigen op Woensdag 13 Maart 2024 des middags te 2.15 uur door Renée Goverdina Catharina Maas geboren op 9 februari 1994 te Amersfoort
Promotoren: Prof. dr. J.P.G. Sluijter Prof. dr. P.A.F.M. Doevendans Copromotoren: Dr. J.W. Buikema Dr. M. Harakalová Het onderzoek beschreven in dit proefschrift is uitgevoerd binnen het Universitair medisch centrum Utrecht, het Regenerative Medicine Center te Utrecht, Nederland en aan de Universiteit van Stanford, Californië, Verenigde Staten van Amerika. Het onderzoek beschreven in dit proefschrift werd (mede) mogelijk gemaakt met financiële steun van de stichting PLN (2018 ‘Cure PLN’, 2019 ‘Crazy-idea’ en 2020 ‘Crazy idea’), de Leducq grant (CURE-PLaN no.18CVD01), de Nederlandse Hartstichting (Dekker 03-003-2021-T025) en van de Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO 2021 ‘HARVEY18747-OTP’). De druk van dit proefschrift werd mede mogelijk gemaakt door financiële steun van de Hartstichting en de PLN stichting.
Beoordelingscommissie: Prof. dr. J.M. Beekman Prof. dr. A.N. Bovenschen Prof. dr. P.C.J.J. Passier Prof. dr. E. van Rooij Prof. dr. J.P. van Tintelen (voorzitter) Paranimfen: Floor van den Dolder Elana van Rooden-Meijer
Opgedragen aan mijn ouders en grootouders die mij gezegend hebben met mijn kostbaarste talent; de gave van nieuwsgierigheid en het verlangen naar een leven vol ontdekkingen.
TABLE OF CONTENTS Chapter 1 General Introduction and Thesis Outline 11 Part I - From Heart Development Towards Optimal Cardiac in Vitro Models Chapter 2 Harnessing Developmental Cues for Cardiomyocyte Production Development 2023. 27 Chapter 3 Wnt Activation and Reduced Cell-Cell Contact Synergistically Induce Massive Expansion of Functional Human iPSC-Derived Cardiomyocytes Cell Stem Cell 2020. 55 Chapter 4 Massive Expansion and Cryopreservation of Functional Human Induced Pluripotent Stem cell-Derived Cardiomyocytes Cell STAR protocols 2021. 87 Chapter 5 Sarcomere Disassembly and Transfection Efficiency in Proliferating Human iPSC-Derived Cardiomyocytes Journal of Cardiovascular Development and Disease 2022. 107 Chapter 6 Metabolic Maturation Increases Susceptibility to Hypoxia-Induced Damage in Human iPSC-Derived Cardiomyocytes Stem Cells Translational Medicine 2022. 125 Part II - Modelling the Phospholamban R14del Mutation using Patient-Specific hiPSC-CMs Chapter 7 Phospholamban p.Arg14del Cardiomyopathy: a Systematic Summary of the Pathophysiological Mechanisms. Manuscript submitted to Circulation Research. 149 Chapter 8 Transcriptional Regulation Profiling Reveals PPARA-mediated Fatty Acid Oxidation as a Novel Therapeutic Target in Phospholamban R14del Cardiomyopathy Preprint online (Research Square) 2022. 199 Chapter 9 Unfolded Protein Response as a Compensatory Mechanism and Potential Therapeutic Target in PLN R14del Cardiomyopathy Circulation 2021. 239
Chapter 10 Generation of Human Induced Pluripotent Stem Cell (hiPSC) Lines Derived from Six Patients Carrying the Pathogenic PhospholambanR14del (PLN-R14del) Variant and Two non-carrier Family Members Manuscript in preparation. 263 Chapter 11 Generation, High Throughput Screening, and Biob anking of Human iPSC-Derived Cardiac Spheroids Journal of Visualized Experiments 2023. 277 Chapter 12 Modeling and Rescue of PLN-R14del Cardiomyopathy Phenotype in Human Induced Pluripotent Stem Cell-Derived Cardiac Spheroids Manuscript in preparation. 297 Part III - General Discussion and Summary Chapter 13 General Discussion - hiPSC-CMs Disease Modelling and Future Perspectives 347 Chapter 14 General Summary 383 Appendices Summary in Dutch 390 List of publications 395 Acknowledgements 401 About the author 409 UU PhD Portfolio 411
Bright field image of hiPSC-derived cardiomyocytes after 90 days of culturing.
Chapter 1 General Introduction and Thesis Outline Renée G.C. Maas
12 Chapter 1 GENERAL INTRODUCTION This chapter describes the scientific content that incited the hypotheses and ideas investigated in this thesis. Firstly, it introduced the clinical significance of a genetic variation in cardiomyopathies, especially the dilated cardiomyopathy and arrhythmogenic cardiomyopathy, caused by a deleterious mutation of the arginine 14 codon in the phospholamban (PLN) gene (p.Arg14del). Secondly, it describes the knowledge derived from cardiogenesis and reproducible methods for the efficient generation of cardiomyocytes (CMs) derived from human induced pluripotent stem cells (hiPSCs). Lastly, it discusses the use of hiPSC-CMs for disease modeling and contemporary challenges encountered in drug research and postulates methodological alternatives. 1.1 Clinical relevance In the past 30 years, the importance of cardiomyopathies as causes of morbidity and mortality, particularly in sudden cardiac death and heart failure, has been highlighted by the recognition of disease-causing genetic variants.1 Genetic cardiomyopathies affect families following a Mendelian inheritance pattern with variable phenotype expression. They typically affect young patients and are important causes of sudden cardiac death in individuals who might otherwise be asymptomatic. The individual genetic makeup and environmental circumstances are responsible for a highly variable disease onset and progression. Despite major efforts to improve their condition with lifestyle alterations and medication, the natural course of cardiomyopathies cannot be halted, and gradual progress towards severely impaired cardiac function and death is generally inevitable. Progress has also been made in the management of several types of cardiomyopathies. However, advances in understanding these diseases show that cardiomyopathies represent complex genotypic and phenotypic entities.2 Therefore, in the past decade, major progress has been made in detecting and, understanding the molecular and genetic basis of disease, pathophysiology, and clinical and radiological assessment of genetic cardiomyopathies.3,4 These insights can potentially fuel enormous improvements for the early detection and novel therapeutic strategies of the future to prevent the detrimental effects of genetic cardiomyopathies. 1.2 PLN-R14del cardiomyopathy The major inherited cardiomyopathies, dilated cardiomyopathy (DCM), arrhythmogenic cardiomyopathy (ACM), and hypertrophic cardiomyopathy (HCM), are characterized by arrhythmias and/or cardiac dysfunction often leading to progressive heart failure and sudden cardiac death.5 In 40-60% of the patients with DCM and HCM, underlying pathogenic variants can be found, mainly located in genes encoding sarcomeric proteins6,7. In contrast, ACM is mainly caused in 60% of the patients by pathogenic variants in desmosomal genes.8 Interestingly, one of the pathogenic mutations in both 10% of the DCM and 15% of
13 General Introduction and Thesis Outline 1 the ACM patients in the Netherlands was caused by a deleterious mutation of the arginine 14 codon in the phospholamban (PLN) gene (p.Arg14del).9 Since the discovery of this mutation in a Greek family in 2006, thousands of patients have been identified in not only the Netherlands but also the USA, Canada, China, Germany, and Spain.10–15 PLN is a 52-amino acid protein located in the sarcoplasmic reticulum (SR) membrane that acts as a crucial reversible regulator of Ca2+ uptake in the CM.16 In contrast to wildtype PLN, a deletion of arginine 14 codon (R14del) in the PLN gene (PLN-R14del) disrupts the conformational changes upon phosphorylation, resulting in inhibition of PLN pentamer formation and thus constitutive inhibition of sarcoplasmic/endoplasmic reticulum Ca2+ ATPase2a (SERCA2A/ATP2A2).12,17 Therefore, PLN-R14del has been associated with irreversible super-inhibition of SERCA2A activity, preventing the influx of calcium into the sarcoplasmic reticulum. This process, in theory, delays the Ca2+ reuptake and induces prolonged muscle contraction. The pathophysiological mechanism causes several clinical features, including a dilated and/or arrhythmogenic heart muscle and the presence of cardiac fibrofatty replacement18 and protein aggregates.19 Compared to other mutation carriers, patients with a PLN mutation have a higher frequency of left ventricle structural and functional abnormalities, and they show the most pronounced diminished left ventricle function detected by echocardiography and cardiac magnetic resonance imaging (MRI).20 Moreover, PLN-R14del hearts were compared with hearts with desmosomal, lamin A/C, sarcomeric, and desmin mutations and presented the highest amount of myocardial fibrosis, which is found in a distinct pattern in the posterolateral left ventricular wall.21 Ultimately, it has been determined that PLN-R14del mutation carriers have a higher incidence of malignant ventricular arrhythmias with left ventricular ejection fraction <45%, premature sudden cardiac death, and end-stage heart failure when compared to DCM patients that do not carry this pathogenic variant.22 This mutation presents a highly variable phenotype, ranging from asymptomatic to cardiomyopathic. Its awareness is rather low, therefore, thousands of people can be carriers unknowingly.20 This caveat is highly relevant because a substantial proportion of individuals who carry disease-causing genetic variants and are at risk of disease complications have incomplete and/or late-onset disease expression. To raise awareness of the mutation phenotype and more efficiently identify carriers and possibly develop new therapies, it is key to better understand the pathological mechanisms underlying this disease. It remains unclear how exactly the PLN-R14del mutation leads to such severe cardiomyopathy and malignant arrhythmias. Unfortunately, despite almost two decades worth of research about PLN-R14del cardiomyopathy, targeted treatment for these patients is lacking. Currently, efforts to identify a tailored therapy are ongoing, using state-of-the-art technology and through synergistic collaborations. With the development in finding therapeutics for this specific cardiovascular disease, bridges will be built toward the utility of these strategies for a plethora of other diseases.
14 Chapter 1 1.3 Lessons learned from cardiogenesis The important inventions of the microscope, sterile culturing, and defined mediums led to the creation of the first human immortal cell line in 1951.23 These so-famously called HeLa cells were created by a tissue sample taken from a young woman with cervical cancer and quickly became invaluable to medical research. However, until today, due to the limited regenerative capacity of the heart, no successful passaging of adult human primary CMs nor cardiac stem cell isolation has been possible. This made the culturing of adult human CM models a challenge. For many years, signaling pathways that specify cardiac mesoderm and regulate cardiac proliferation have been extensively studied. Hereafter, wingless-related integration site (Wnt) signaling has proven to be essential in heart development 24 and the balance of Wnt regulation appears to play a critical role in cardiogenesis and, later on, shaping the cardiac fields.25,26 In adult organisms, Wnt proteins regulate diverse cellular processes such as gene transcription and cell proliferation, migration, polarity, or division.24 Since the discovery of embryonic stem cells by James Thomson and later, in 2006, the hiPSCs by Shinya Yamanaka have been heralded as major breakthroughs in stem cell research.27,28 The knowledge derived from the developmental studies has been translated into reproducible methods for the efficient generation of cardiomyocytes (CMs) derived from human induced pluripotent stem cells (hiPSCs). The first attempts to differentiate hiPSCs into CMs about a decade ago resulted in very low numbers of hiPSC-CMs (5-10%), and the re-plating of these cells was a challenge.29 Hereafter, the combination of monolayer cell culture with defined serum-free media, supplemented with growth factors involved in normal human embryological heart development, like Wnt agonists and antagonists, resulted in much higher efficiency (80– 99%).30,31 Altogether, the improvements over the last 70 years have tremendously improved the generation of stem cell-derived cardiomyocytes as a human cell source for in vitro disease modeling. 1.4 Cardiomyocyte generation and expansion The discovery of hiPSCs offers unprecedented opportunities to study early human physiology and pathology at a cellular level.32 Multiple embryonic pathways have been implicated in cardiac differentiation of pluripotent stem cells.33 Understanding the particular role of the Wnt signaling pathway in heart formation has helped to develop pluripotent stem cell differentiation protocols that produce relatively pure cardiomyocyte populations34 Currently, various hiPSC- CM differentiation protocols incorporate Wnt signaling activation via glycogen synthase kinase 3 beta (GSK-3β) inhibition (usually with a small molecule) from days 0 to 3. This is followed by Wnt inhibition via porcupine inhibition at days 3-5 to induce highly pure and dense hiPSC-CM cultures (Figure 1).35
15 General Introduction and Thesis Outline 1 Figure 1. Generation of hIPSC-CMs by modulating the Wnt signaling pathway. Created with BioRender.com A major limitation, however, remains the batch-to-batch variability of hiPSC-CM efficiency and the inability to robustly expand generally dense-cultured functional hiPSC-CMs for more than 5 fold.36–39 Recently, we described that concomitant Wnt pathway regulation and removal of cell-cell contact inhibition via low cell density serial passaging resulted in a massive proliferative response of hiPSC-CMs (Figure 2).40,41 The developmental clues and the role of the Wnt signaling pathway resulted not only in the efficient generation of hiPSC-CMs but also in a highly efficient detailed method for the expansion and passaging of functional hiPSC-CMs. However, the lack of maturity of the hiPSC-CMs generated by the described di erentiation protocols is an important limitation to overcome for optimal human in vitro modeling. Figure 2. Expansion of hiPSC-CMs by modulating the Wnt signaling pathway. Created with BioRender.com 1.5 Cardiomyocyte maturation HiPSC-CMs have emerged as a promising experimental tool for translational heart research. In theory, with the cardiac expansion method, hiPSC-CMs can provide an unlimited source of human cardiomyocytes that entail the use of human cardiac tissue or cells with minimal ethical and practical concerns. However, their usability as a human adult CM model is limited by their immature phenotype, represented in general by structural underdevelopment, metabolism based on glucose or lactate instead of fatty acids, slow Ca2+ signals, and negative force-frequency relationship that impact the features of electrophysiological parameters.42–45 Applying hiPSC-CMs for adult cardiomyopathy disease modeling or drug research purposes, the immature status may influence to some extent the observed effects by impacting the excitation-contraction coupling. Therefore, to overcome this limitation,
16 Chapter 1 many strategies have been invented to induce a mature phenotype of hiPSC-CMs. These strategies include 1) biochemical strategies such as; prolonged culture time; alterations in energy sources (Figure 3); hormones; cell-cell interactions/co-culturing; genetic manipulation and 2) biophysical strategies including extracellular matrices/substrate stiffness; biophysical stimulation; in vivo maturation; mechanical stretch and 3D cell culture have been described.46,47 These different maturation methods can improve the modeling of complex adult cardiac physiology and disease. Figure 3. Maturation of hiPSC-CMs by modulating the physiologically appropriate levels of glucose and albuminbound fatty acids. Created with BioRender.com 1.6 hiPSC-CM models Next to cardiomyocyte generation, expansion, and maturation developments, the human- induced pluripotent stem cell (hiPSC) technology has yielded patient-derived cardiomyocytes that exhibit some of the hallmarks of cardiovascular disease and are therefore being used to model disease states. Some of the technical challenges were solved, such as the scaled production of pure cardiomyocytes in a quality-controlled way and the long-term cryopreserved hiPSC-CM biobanks. The generated ‘more mature’ cardiomyocytes have been used to study physiological and disease states, screen for novel therapeutic targets, and generate heart tissue for pharmaceutical testing. However, with the unlimited production of hiPSC-CMs, challenges arise for the scalability, sensitivity, and costs of hiPSC-CM disease modeling and robotic functional screening platforms. Each technique has its strengths, such as scalability or sensitivity, and needs to be considered carefully or combined to develop a full phenotypic screening (Table 1). By combining the different hiPSC-CM models, a full phenotypic screening would make it possible to study physiological disease states or to screen pharmaceutical or novel therapeutic compounds.
17 General Introduction and Thesis Outline 1 Table 1: Advantages and disadvantages in human iPSC-derived cardiomyocytes in vitro disease modeling platforms Model Advantage Disadvantages Format Costs Sensitivity Scalability EHTs Sensitivity to cardiotoxins, physiological force, and slow action potential Hypoxic conditions in center, risk of breaking. 24 wells High High Low Heart-on-a- chip Allows ECM manipulation, mimics 3D cardiac environment Limited imaging possibilities, Nonlinear cell alignment 48 wells Middle Middle Low Cardiac Spheroids Recapitulating cell-cell interactions, reproducible, small 3D model Lack of ECM, Nonlinear cell alignment. 384 wells Low Middle High Individual CMs Heterogeneity analyses, Strong proliferation capacity Far from physiological conditions, loss of cellcell interactions 384 wells Low Low Middle 2D Monolayers Action potential and calcium wave propagation measurements, IF Low on maturity, unable to recapitulate some disease phenotypes 384 wells Low Low High 1.7 High-throughput integrative disease modeling and drug screenings New therapies for genetic heart diseases have a high attrition rate, with only 1% reaching the stadium of a clinical trial.48 This low number can partly be explained in part by a reliance on animal models, transformed cell lines, and heterologous recombinant systems for drug discovery. Because of their ease of culture, cell-based assays used in drug screening have historically depended on animal cells. However, these cells are generally short-term cultured and are limited in how well they reflect human biology. Moreover, the use of more physiologically relevant primary cells is restricted by availability and inherent variability. As previously described, the advent of hiPSC technology has opened up the possibility of technology platforms to perform compound screens of hiPSC-derived cardiomyocytes with relatively high throughput. It is essential to realize their potential for drug discovery. To date, hiPSCs have been used to model a growing list of heart diseases, providing proof of concept that their differentiated derivatives can recapitulate disease-associated pathologies. Moreover, in some cases, it has been shown that pathologies expressed by these cell-based disease models can be ameliorated by drugs known to be therapeutic for patients.30 Based on an evaluation of multiple parameters including gene/mRNA/protein omics, mRNA/protein expression and electro-pathophysiology such as contractility and calcium handling, researchers will now be able to identify compounds that target molecules or pathways known to modulate cardiomyocytes, as well as those not previously associated with cardiomyocyte function
18 Chapter 1 (Figure 4). This data will define a subset of chemical probes for interrogating cardiomyocyte phenotyping and provide validation of a platform for high-throughput screening of hiPSC- derived cardiomyocytes. Figure 4. Graphical abstract of patient-derived human cardiomyocyte models for therapeutic screenings and disease modeling in vitro. THESIS OUTLINE The aim of this thesis is to shed light on the utilization of functional patient-derived humaninduced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) for disease modeling and high throughput screening of novel therapeutic strategies for genetic cardiomyopathies. This goal is achieved in two ways: firstly, by investigating the developmental clues of cardiogenesis for the generation of hiPSC-CMs, and secondly, by showing that the patient hiPSC-CMs can be conducted with scientifically sound methods using pragmatic in vitro disease models and innovative disease pathway analysis.
19 General Introduction and Thesis Outline 1 PART I: CLUES FROM HEART DEVELOPMENT TOWARDS OPTIMIZED HIPSCCMS MODELS In the first part of this thesis, we describe the steps of cardiogenesis and the pathways involved (Chapter 2). This leads to the identification of particular mechanisms involved in cardiomyocyte proliferation and the strategies for cardiomyocyte production, which are also presented here. By this rationale, differentiation strategies that have proven to generate hiPSC-CMs effectively may be repurposed for the massive expansion of hiPSC-CMs. We describe that concomitant GSK-3β inhibition and removal of cell-cell contact inhibition via low cell density serial passaging resulted in a massive proliferative response of hiPSC-CMs (Chapter 3). With this knowledge, we invented a highly efficient method for the expansion and passaging of functional hiPSC-CMs that can routinely be cryopreserved and subsequently used as a stable cell source for downstream applications (Chapter 4). This method is put in a new perspective in Chapter 5, in which live imaging in a hiPSC-CM culture system is used to follow the sequence of sarcomere breakdown during the mitotic phases of CM cell division. Again, going one step deeper in understanding and utilizing the mechanism of action, we will describe the magnitude of Wnt activation in hiPSC-CMs, which results in increased efficiency of non-viral vector incorporation. These findings give an insight into the regulation of sarcomere homeostasis during mitotic cell phases and provide a tool for further molecular and engineering studies (Chapter 5). However, the expanded hiPSC-CMs physiological immaturity severely limits their utility as a model system and their adoption for drug discovery. Therefore, we avail ourselves of a maturation media designed to provide oxidative substrates adapted to the metabolic needs of hiPSC-CMs. Part I concludes with a model of cardiac ischemic damage in metabolically matured hiPSC-CMs and exemplarily evaluates the cardioprotective effect of the RIP1 kinase inhibitor necrostatin-1 (Chapter 6). PART II: MODELING THE PHOSHOLAMBAN R14DEL MUTATION USING PATIENT-SPECIFIC HIPSC-CMS In the second part of this thesis, we aim to provide the most complete approach to investigating the molecular mechanism behind the genetic cardiomyopathy caused by the deletion of arginine 14 in the phospholamban gene (PLN-R14del). First, we will use a systematic review describing studies conducted to investigate the PLN-R14del disease (Chapter 7). We describe the currently available observational evidence that suggests a possible molecular mechanism and the therapeutic strategies used to improve the disease phenotype. Hereafter, we combined the approach of transcriptional regulation analysis in human primary tissue and validation in a unique long-term (160 days) matured hiPSC-CM model. We demonstrate a dysregulated PPARA-mediated mitochondrial fatty acid oxidation (FAO) signaling in PLN-R14del hearts and hiPSC-CMs. By activating PPARA in PLN-R14del hiPSC-CMs using bezafibrate, we observed an improved mitochondrial structure and calcium handling function, further indicating the importance of FAO in the molecular mechanism behind the PLN-R14del disease (Chapter 8). In
20 Chapter 1 Chapter 9, we started by using transcriptomics of hiPSC-CMs and compared the differentially expressed pathways in PLN-R14del vs isogenic control. Here, we found the unfolded protein response as a compensatory mechanism of the PLN-R14del disease. We will show how this mechanism, directly and indirectly, suggests a mechanistic link between protein toxicity and PLN aggregate formation in the PLN-R14del-induced pathophysiology. Furthermore, we explored the therapeutic potential of activating the UPR with a small molecule activator, BiP (Chapter 9). In Chapters 8 and 9, we made use of only one severely affected PLN-R14del patient. To study the molecular mechanism of the PLN-R14del of various patients, we generated hiPSC lines derived from six patients carrying the pathogenic PLN-R14del variant and two non-carrier family members. (Chapter 10) In order to investigate these hiPSC-CMs of many PLN-R14del patients, we present a scalable, high-throughput screening-compatible workflow for the generation, maintenance, and optical analysis of cardiac spheroids in a 96- well-format (Chapter 11). Additionally, these small cardiac spheroids can be cryopreserved, allowing researchers to create next-generation living biobanks. Lastly, we use the spheroid model to study the PLN-R14del disease, screening hiPSC-CMs generated from 6 hiPSC lines. Here, we found the calcium handling parameters such as decay time, rise time, calcium transient duration, and peak value (amplitude) to be reduced in PLN-R14del spheroids derived from three individual patients (Chapter 12). Lastly, translating these findings back to clinical care, we will investigate the potential improvement of an AAV-mediated I-1c gene augmentation therapy on the PLN-R14del disease. Findings from the trial should be able to serve as pivotal evidence for therapy potentials and guidelines. GENERAL DISCUSSION AND SUMMARY As concluding considerations, we will put all the aforementioned findings into contemporary and future perspective, provide recommendations for future research, and provide an outlook for the future of hiPSC-CMs in disease modeling and therapeutic screening (Chapter 13). The thesis ends with a summary of the previous chapters (Chapter 14). Table 2. Overview of the general introduction and related thesis chapters on the specific introduction sections. Introduction section Related thesis chapter number 1.1 Clinical relevance Chapter 7 1.2 PLN-R14del cardiomyopathy Chapters 7-10 & 12 1.3 Lessons learned from cardiogenesis Chapter 2 1.4 Cardiomyocyte generation and expansion Chapters 3-5 1.5 Cardiomyocyte maturation Chapters 6, 8, 9, 12 1.6 hiPSC-CM models Chapters 3-6 & 8-12 1.7 High-throughput integrative disease modeling and drug screenings Chapters 11,12
21 General Introduction and Thesis Outline 1 REFERENCES 1. Towbin, J. A. INHERITED CARDIOMYOPATHIES. Circ. J. 78, 2347 (2014). 2. Robson, A. New insights into the genetics of cardiomyopathies. Nat. Rev. Cardiol. 18, 229–229 (2021). 3. Kaviarasan, V., Mohammed, V. & Veerabathiran, R. Genetic predisposition study of heart failure and its association with cardiomyopathy. The Egyptian Heart Journal 74, 1–17 (2022). 4. Lahoti, N., Jabbour, R. J., Ariff, B. & Wang, B. X. Cardiac MRI in cardiomyopathies. Future Cardiology vol. 18 51–65 Preprint at https://doi.org/10.2217/fca-2020-0233 (2022). 5. Bhuiyan, Z. Review of: ‘Penetrance and disease expression of (likely) pathogenic variants associated with inherited cardiomyopathies in the general population’. Preprint at https://doi.org/10.32388/c17hcm (2022). 6. Pugh, T. J. et al. The landscape of genetic variation in dilated cardiomyopathy as surveyed by clinical DNA sequencing. Genetics in Medicine vol. 16 601–608 Preprint at https://doi.org/10.1038/gim.2013.204 (2014). 7. Akhtar, M. & Elliott, P. The genetics of hypertrophic cardiomyopathy. Glob Cardiol Sci Pract 2018, 36 (2018). 8. Groeneweg, J. A. et al. Clinical Presentation, Long-Term Follow-Up, and Outcomes of 1001 Arrhythmogenic Right Ventricular Dysplasia/Cardiomyopathy Patients and Family Members. Circ. Cardiovasc. Genet. 8, 437–446 (2015). 9. van der Zwaag, P. A. et al. Recurrent and founder mutations in the Netherlands-Phospholamban p.Arg14del mutation causes arrhythmogenic cardiomyopathy. Neth. Heart J. 21, (2013). 10. Cheung, C. C. et al. Phospholamban cardiomyopathy: a Canadian perspective on a unique population. Neth. Heart J. 27, 208–213 (2019). 11. DeWitt, M. M., MacLeod, H. M., Soliven, B. & McNally, E. M. Phospholamban R14 deletion results in late-onset, mild, hereditary dilated cardiomyopathy. J. Am. Coll. Cardiol. 48, 1396–1398 (2006). 12. Haghighi, K. et al. A mutation in the human phospholamban gene, deleting arginine 14, results in lethal, hereditary cardiomyopathy. Proc. Natl. Acad. Sci. U. S. A. 103, 1388–1393 (2006). 13. Jiang, X. et al. The phenotypic characteristic observed by cardiac magnetic resonance in a PLN-R14del family. Scientific Reports vol. 10 Preprint at https://doi.org/10.1038/s41598-020-73359-8 (2020). 14. López-Ayala, J. M. et al. Phospholamban p.arg14del Mutation in a Spanish Family With Arrhythmogenic Cardiomyopathy: Evidence for a European Founder Mutation. Revista Española de Cardiología (English Edition) vol. 68 346–349 Preprint at https://doi.org/10.1016/j.rec.2014.11.012 (2015). 15. Posch, M. G. et al. Genetic deletion of arginine 14 in phospholamban causes dilated cardiomyopathy with attenuated electrocardiographic R amplitudes. Heart Rhythm 6, 480–486 (2009). 16. Haghighi, K., Bidwell, P. & Kranias, E. G. Phospholamban Interactome in Cardiac Contractility and Survival: A New Vision of an OLD Friend. J. Mol. Cell. Cardiol. 0, 160 (2014). 17. Vafiadaki, E., Haghighi, K., Arvanitis, D. A., Kranias, E. G. & Sanoudou, D. Aberrant PLN-R14del Protein Interactions Intensify SERCA2a Inhibition, Driving Impaired Ca2+ Handling and Arrhythmogenesis. Int. J. Mol. Sci. 23, (2022). 18. Te Rijdt, W. P. et al. Myocardial fibrosis as an early feature in phospholamban p.Arg14del mutation carriers: phenotypic insights from cardiovascular magnetic resonance imaging. Eur. Heart J. Cardiovasc. Imaging 20, 92–100 (2019). 19. Te Rijdt, W. P. et al. Phospholamban p.Arg14del cardiomyopathy is characterized by phospholamban aggregates, aggresomes, and autophagic degradation. Histopathology 69, 542–550 (2016). 20. Hof, I. E. et al. Prevalence and cardiac phenotype of patients with a phospholamban mutation. Neth. Heart J. 27, 64–69 (2019). 21. Sepehrkhouy, S. et al. Distinct fibrosis pattern in desmosomal and phospholamban mutation carriers in hereditary cardiomyopathies. Heart Rhythm 14, 1024–1032 (2017). 22. Zwaag, P. A. van der et al. Phospholamban R14del mutation in patients diagnosed with dilated cardiomyopathy or arrhythmogenic right ventricular cardiomyopathy: evidence supporting the concept of arrhythmogenic cardiomyopathy. European Journal of Heart Failure vol. 14 1199–1207 Preprint at https://doi.org/10.1093/eurjhf/ hfs119 (2012). 23. Vp, P. Cultivation of large cultures of HeLa cells in horse serum. Science 121, (1955).
22 Chapter 1 24. Olson, E. N. & Srivastava, D. Molecular Pathways Controlling Heart Development. Science vol. 272 671–676 Preprint at https://doi.org/10.1126/science.272.5262.671 (1996). 25. Solloway, M. J. & Harvey, R. P. Molecular pathways in myocardial development: a stem cell perspective. Cardiovasc. Res. 58, 264–277 (2003). 26. Ueno, S. et al. Biphasic role for Wnt/beta-catenin signaling in cardiac specification in zebrafish and embryonic stem cells. Proc. Natl. Acad. Sci. U. S. A. 104, 9685–9690 (2007). 27. Thomson, J. A. et al. Isolation of a primate embryonic stem cell line. Proc. Natl. Acad. Sci. U. S. A. 92, (1995). 28. Takahashi, K. & Yamanaka, S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, (2006). 29. Carvajal-Vergara, X. et al. Patient-specific induced pluripotent stem cell derived models of LEOPARD syndrome. Nature 465, 808 (2010). 30. Lian, X. et al. Directed cardiomyocyte differentiation from human pluripotent stem cells by modulating Wnt/βcatenin signaling under fully defined conditions. Nat. Protoc. 8, (2013). 31. Burridge, P. W., Holmström, A. & Wu, J. C. Chemically Defined Culture and Cardiomyocyte Differentiation of Human Pluripotent Stem Cells. Curr. Protoc. Hum. Genet. 87, 21.3.1 (2015). 32. Musunuru, K. et al. Induced Pluripotent Stem Cells for Cardiovascular Disease Modeling and Precision Medicine: A Scientific Statement From the American Heart Association. Circ Genom Precis Med 11, e000043 (2018). 33. Lee, J. H., Protze, S. I., Laksman, Z., Backx, P. H. & Keller, G. M. Human Pluripotent Stem Cell-Derived Atrial and Ventricular Cardiomyocytes Develop from Distinct Mesoderm Populations. Cell Stem Cell 21, 179–194.e4 (2017). 34. Batalov, I. & Feinberg, A. W. Differentiation of Cardiomyocytes from Human Pluripotent Stem Cells Using Monolayer Culture. Biomarker Insights vol. 10s1 BMI.S20050 Preprint at https://doi.org/10.4137/bmi.s20050 (2015). 35. Pahnke, A. et al. The role of Wnt regulation in heart development, cardiac repair and disease: a tissue engineering perspective. Biochem. Biophys. Res. Commun. 473, 698 (2016). 36. Mills, R. J. et al. Drug Screening in Human PSC-Cardiac Organoids Identifies Pro-proliferative Compounds Acting via the Mevalonate Pathway. Cell Stem Cell 24, 895–907.e6 (2019). 37. Sharma, A. et al. Stage-specific Effects of Bioactive Lipids on Human iPSC Cardiac Differentiation and Cardiomyocyte Proliferation. Sci. Rep. 8, 6618 (2018). 38. Titmarsh, D. M. et al. Induction of Human iPSC-Derived Cardiomyocyte Proliferation Revealed by Combinatorial Screening in High Density Microbioreactor Arrays. Sci. Rep. 6, (2016). 39. Uosaki, H. et al. Identification of chemicals inducing cardiomyocyte proliferation in developmental stagespecific manner with pluripotent stem cells. Circ. Cardiovasc. Genet. 6, (2013). 40. Buikema, J. W. et al. Wnt Activation and Reduced Cell-Cell Contact Synergistically Induce Massive Expansion of Functional Human iPSC-Derived Cardiomyocytes. Cell Stem Cell 27, 50–63.e5 (2020). 41. Maas, R. G. C. et al. Massive expansion and cryopreservation of functional human induced pluripotent stem cell-derived cardiomyocytes. STAR Protoc 2, 100334 (2021). 42. Mummery, C. L. et al. Differentiation of human embryonic stem cells and induced pluripotent stem cells to cardiomyocytes: a methods overview. Circ. Res. 111, (2012). 43. Denning, C. et al. Cardiomyocytes from human pluripotent stem cells: From laboratory curiosity to industrial biomedical platform. Biochim. Biophys. Acta 1863, 1728–1748 (2016). 44. Knight, W. E. et al. Maturation of Pluripotent Stem Cell-Derived Cardiomyocytes Enables Modeling of Human Hypertrophic Cardiomyopathy. Stem Cell Reports 16, 519–533 (2021). 45. Karakikes, I., Ameen, M., Termglinchan, V. & Wu, J. C. Human Induced Pluripotent Stem Cell-Derived Cardiomyocytes: Insights into Molecular, Cellular, and Functional Phenotypes. Circ. Res. 117, 80 (2015). 46. Ahmed, R. E., Anzai, T., Chanthra, N. & Uosaki, H. A Brief Review of Current Maturation Methods for Human Induced Pluripotent Stem Cells-Derived Cardiomyocytes. Front Cell Dev Biol 8, 178 (2020). 47. Tu, C., Chao, B. S. & Wu, J. C. Strategies for Improving the Maturity of Human Induced Pluripotent Stem CellDerived Cardiomyocytes. Circ. Res. 123, 512–514 (2018). 48. Wong, C. H., Siah, K. W. & Lo, A. W. Estimation of clinical trial success rates and related parameters. Biostatistics 20, 273 (2019).
23 General Introduction and Thesis Outline 1
PART I From Heart Development Towards Optimal Cardiac in Vitro Models
Cardiac growth of the mouse heart: from embryonic day 12.5 to postnatal day 60.
Chapter 2 Harnessing Developmental Cues for Cardiomyocyte Production Published in: Development, 2023 Renée G.C. Maas, Floor van den Dolder*, Qianliang Yuan*, Jolanda van der Velden, Sean M. Wu, Joost P.G. Sluijter, Jan W. Buikema *These authors contributed equally.
28 Chapter 2 ABSTRACT Developmental research has attempted to untangle the exact signals that control heart growth and size, with knockout studies in mice identifying pivotal roles for Wnt and Hippo signaling during embryonic and fetal heart growth. Despite this improved understanding, no clinically relevant therapies are yet available to compensate for the loss of functional adult myocardium and the absence of mature cardiomyocyte renewal that underlies cardiomyopathies of multiple origins. It remains of great interest to understand which mechanisms are responsible for the decline in proliferation in adult hearts and to elucidate new strategies for the stimulation of cardiac regeneration. Multiple signaling pathways have been identified that regulate the proliferation of cardiomyocytes in the embryonic heart and appear to be upregulated in postnatal injured hearts. In this Review, we highlight the interaction of signaling pathways in heart development and discuss how this knowledge has been translated into current technologies for cardiomyocyte production.
29 Harnessing developmental cues for cardiomyocyte production 2 INTRODUCTION Heart failure often results from the irreversible loss of functional myocardium or malfunctioning of the individual cardiomyocytes that comprise this tissue (De Boer et al., 2003; Fox et al., 2001; Gerber et al., 2000). This loss of functional cardiomyocytes can be acute or gradual and results in adverse remodeling of the remaining healthy myocardium (Olivetti et al., 1997; Saraste et al., 1997). In turn, myocardial dysfunction results in mechanical stress and upregulation of factors including angiotensin and norepinephrine, which all act to further promote detrimental myocardial remodeling (Colucci, 1997; Adhyapak, 2022). These alterations to extracellular matrix composition, cytoskeletal architecture and cell-cell connections occur in parallel with changes to the cardiac gene profile, such as reinduction of a fetal gene program (Parker et al., 1990; Bray et al., 2008). Although clinical therapies for heart failure have significantly improved with multiple lines of heart failure drugs and mechanical circulatory support devices (Cook et al., 2015; Mancini and Burkhoff, 2005; Shen et al., 2022; Ponikowski et al., 2016), these treatments do not repair or replace malfunctioning myocardium. A central hurdle is that the adult heart is a largely postmitotic organ, where annual turnover is between 1-2% in young adults, and less than 0.5% in older adults (Bergmann et al., 2009; 2015). Thus, it is unsurprising that the adult heart lacks regenerative capacity post injury. In contrast, before birth, expansion of fetal cardiomyocytes is crucial for proper cardiogenesis and is tightly regulated by Wnt and Hippo signaling, with varying cardiomyocyte proliferation rates depending on location and developmental stage (Drenckhahn et al., 2008; Sturzu et al., 2015; Buikema et al., 2013; Rochais et al., 2009; von Gise et al., 2012; Qyang et al., 2007). Remarkably, the early postnatal mammalian heart possesses regenerative potential (Haubner et al., 2016), but this regenerative response is lost 1 week after birth and scar tissue is formed in response to injury instead (Porrello et al., 2011; Ye et al., 2018; Zhu et al., 2018b). Translational cardiology aims to repair or replace a broken heart with autologous material (Laflamme and Murry, 2011; Ptaszek et al., 2012). The advent of patient-specific pluripotent stem cell sources represented a major advance for the field (Burridge et al., 2012), but the generation of large numbers of functional cardiomyocytes remains challenging due to their low proliferative rates (Tani et al., 2022). Here, we review how lessons from in vivo cardiomyocyte proliferation during mammalian cardiac development have been translated into technology to generate cardiomyocytes from human pluripotent stem cell sources. LESSONS FROM CARDIAC DEVELOPMENT Mesoderm formation During development, most heart structures arise from the mesodermal germ layer. Mesoderm formation takes place at Carnegie stages (CS) 6-7 in humans and embryonic day (E) 6.5 in
30 Chapter 2 mice, at the onset of gastrulation (Zhai et al., 2022). During gastrulation, cells from the blastocyst migrate to give rise to the endoderm and mesoderm. At the end of gastrulation, the three germ layers (endoderm, mesoderm and ectoderm) are specified. At the onset of mesoderm formation, the intraembryonic mesoderm subdivides into four distinct groups: the chordamesoderm, paraxial mesoderm, intermediate mesoderm and lateral plate mesoderm (Ivanovitch et al., 2021; Chan et al., 2013). Cardiopotent cells, also called cardiac progenitor cells, are derived from the lateral plate mesoderm during early gastrulation (Yamada and Takakuwa, 2012; Moretti et al., 2006; Garry and Olson, 2006). Cardiac progenitors are prepatterned within the primitive streak, with the atrial and ventricular cells arising at different anterior-posterior positions (Chan et al., 2013). Foxa2+ cardiac progenitors give rise primarily to the cardiovascular cells of the ventricles and are the first cardiogenic cells that migrate to the anterior side of the embryo (Bardot et al., 2017). These cells comprise approximately half of the cardiomyocyte population (Bardot et al., 2017). The right ventricle and outflow tract progenitors are found in anterior/distal primitive streak, where cells are exposed to a higher ratio of activin A to bone morphogenetic protein 4 (BMP4) signaling, whereas atrial progenitors are specified in the proximal primitive streak, where the activin A to BMP4 ratio is low (Ivanovitch et al., 2021). Brachyury (T) gene expression is also required for the formation of posterior mesoderm in mice and zebrafish (Schulte-Merker and Smith, 1995; Herrmann et al., 1990). Basic fibroblast growth factor (FGF) and Activin A can promote the expression of the T homolog, Xbra, in the Xenopus presumptive ectoderm (Smith et al., 1991). T-box transcription factors and T are intrinsic factors that are crucial for the initiation of mesoderm differentiation and patterning of the primitive streak and are regulated by the Wnt signaling pathway from the adjacent embryonic midline and posterior regions of the embryo, indicating the importance of Wnt signaling at this stage of development (Yamaguchi et al., 1999). Cardiac specification Several transcription factors for cardiac development have been identified (Olson, 2006). These include MESP1, which specifies the cardiac mesodermal population and is expressed in the primitive streak around CS6-7 or E6.5. In MESP1-null embryos, severe cardiac abnormalities are observed, leading to cardiac lethality by E10.5 (Saga et al., 1999). Furthermore, in a double knockout of MESP1 and its homolog MESP2, mesoderm progenitors do not contribute to heart development, indicating that MESP1/2 expression is essential for cardiac mesoderm formation (Kitajima et al., 2000). By CS8 or E7.5, specific regions of the mesoderm differentiate to form cardiovascular progenitor cells, which can be divided into cells that form the first heart field (FHF) and cells that form the second heart field (SHF) (Paige et al., 2015). When Wnt inhibitors such as dickkopf 1 (DKK1) or crescent are administered to posterior lateral plate mesoderm, heart
31 Harnessing developmental cues for cardiomyocyte production 2 muscle development is induced and erythropoiesis is suppressed (Marvin et al., 2001, Naito et al., 2006). Meanwhile, ectopic expression of WNT8 or WNT3a in precardiac mesoderm inhibits heart muscle formation and promotes erythropoiesis (Marvin et al., 2001). In mouse embryonic stem cells, Wnt/β-catenin signaling acts biphasically; early treatment with Wnt3a stimulates mesoderm induction and cardiac differentiation, whereas late activation of β-catenin signaling impairs cardiac differentiation (Ueno et al., 2007). In combination with BMP and FGF, Wnt inhibition results in the activation of key upstream cardiac transcriptional regulatory genes NKX2-5, GATA4 and TBX5, which are required for the initiation of cardiac-like gene expression (Kelly et al., 2014). The formation of the linear heart tube is mostly initiated by the contribution of the FHF, which eventually gives rise to the inflow tract and the majority of the left ventricle (Brade et al., 2013). FHF progenitors at this stage specifically express TBX5 and HCN4 (Später et al., 2013; Bruneau et al., 1999). The SHF develops slightly later and is less differentiated, providing cardiac progenitors that proliferate to promote the expansion of the heart tube (Kelly, 2012). The right ventricle and outflow tract are exclusively generated by the SHF (Buckingham et al., 2005). These SHF cardiomyocytes are marked with TBX1, ISL1 and HAND2 (Moretti et al., 2006; Stanley et al., 2002). Together, these studies demonstrate that Wnt signals in different parts of the mesoderm are repressed as required for cardiac specification of these regions. Most in vitro protocols for directed cardiac differentiation of pluripotent stem cells incorporate this inhibition of Wnt signaling through the application of porcupine small-molecule inhibitors including IWP-2, IWR1 and Wnt-C59 (Lian et al., 2012; Burridge et al., 2014). Developmental heart growth Organ size regulation is an important aspect of cardiac development. The heart must grow large enough to generate sufficient cardiac output, and regional under- or overgrowth may result in septal wall defects, hypoplastic ventricle(s) or obstruction of the outflow tract(s) (Heallen et al., 2011). Heart growth during development is regulated by a combination of cardiomyocyte differentiation, proliferation and hypertrophy. The human heart undergoes a dramatic proliferation period from CS9-CS16, resulting in a 600fold increase in heart volume in just 3 weeks (de Bakker et al., 2016). First, the cardiac crescent fuses at the midline and gives rise to the FHF-derived linear heart tube, which subsequently commences beating and undergoes looping. Hereafter, the linear heart tube expands by drastic proliferation and recruitment of SHF cardiac progenitor cells that migrate from the pericardial cavity to the dorsal and caudal heart tube regions while undergoing differentiation (Kelly et al., 2014). Their rapid proliferation is regulated by canonical Wnt signaling (Günthel et al., 2018; Kwon et al., 2007). Upon the presence of the receptor-bound ligands WNT5A and WNT11, active β-catenin enters the nucleus, where it acts as a transcriptional co-activator of
32 Chapter 2 the T-cell factor (TCF) and leukemia enhancer factor (LEF) transcription factors to activate Wnt target genes involved in cell proliferation such as axin 2 (AXIN2), cyclin D1 (CCND1) and lymphoid enhancer binding factor 1 (LEF1) (Cadigan and Waterman, 2012; Stamos and Weis, 2013) (Figure 1). Conversely, in the absence of Wnt ligands, a complex containing adenomatous polyposis coli (APC), casein kinase 1 (Ck1) and glycogen synthase kinase-3β (GSK-3β) mediates phosphorylation, ubiquitylation and, ultimately, degradation of β-catenin (Cadigan and Waterman, 2012; Stamos and Weis, 2013) (Figure 1). Conditional knockout studies for β-catenin in the SHF produce outflow tract abnormalities and impaired right ventricular development (Qyang et al., 2007; Lin et al., 2007). After specification and terminal differentiation of cardiac progenitors into cardiomyocytes, the developing heart predominantly increases its size and mass via the proliferation of differentiated cardiomyocytes (Günthel et al., 2018) (Table 1). In mice, between E8.0 and E11.0, cardiomyocyte numbers increase 100-fold from ∼700 to ∼70,000 (De Boer et al., 2012) (Figure 2A,C). During this massive growth phase, the size of the individual cardiomyocytes remains relatively constant. After E11.0, proliferation continues but at a slower rate, with cardiomyocyte numbers approaching 1,000,000 by E18.5 (de Boer et al., 2012). The ballooning ventricles exhibit the highest proliferation rates during this period (Moorman and Christoffels, 2003; Moorman et al., 2010). By contrast, the atrioventricular canal, outflow tract and inner curvature regions have lower proliferation rates and thereby preserve the slow contraction characteristics of the heart tube (De Jong et al., 1992). Proliferation rates may also vary within the same region depending on the developmental stage. Within the ventricles, proliferation rates are low during the formation of the trabecular network (Sedmera and Thompson, 2011). The trabeculae contribute to cardiac contractility, channel expression and energy metabolism, and start to develop at the luminal side of the myocardium by E9.5 or CS12 (Meyer et al., 2020; Günthel et al., 2018). Later in development, from CS12 until CS16, proliferation rates increase and the compacted ventricular chamber myocardium shows high expression of proliferative markers such as Ki67 and pHH3 (Buikema et al., 2013; Ye et al., 2015; Lin et al., 2015). These changes to ventricular proliferation rates are due to Wnt signaling. Canonical Wnt signaling is downregulated in the trabecular myocardium, which is consistent with the lower proliferation rates observed here (Buikema et al., 2013; Ye et al., 2015). By contrast, conditional knockout studies in mice have shown that epidermal growth factor receptor (EGFR) and Notch signaling are essential for trabecular development (Gassmann et al., 1995; Grego-Bessa et al., 2007). When proliferation rates in the ventricle increase, Wnt/β-catenin is essential for the exponential growth of the compacted ventricular myocardium; the increase in cardiomyocyte proliferation rates as you move from the inner trabeculae to outer compact myocardium corresponds with the graded activity of canonical Wnt signaling (Buikema et al., 2013; Ye et al., 2015). Consistent with this, β-catenin is mainly active in the compact myocardium, where it is expressed by the majority of proliferating cardiomyocytes (Buikema et al., 2013).
www.ridderprint.nlRkJQdWJsaXNoZXIy MTk4NDMw