UNCOVERING THE ROLE OF LAMIN A/C IN NUCLEAR MECHANICS AND REGULATION OF GENE EXPRESSION IN HEALTH AND DISEASE A Dissertation Presented to the Faculty of the Graduate School of Cornell University In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy by Melanie Elizabeth Maurer December, 2021 © 2021 Melanie Elizabeth Maurer UNCOVERING THE ROLE OF LAMIN A/C IN NUCLEAR MECHANICS AND REGULATION OF GENE EXPRESSION IN HEALTH AND DISEASE Melanie Elizabeth Maurer, Ph. D. Cornell University 2021 Lamin A/C, encoded by the LMNA gene, forms a dense protein meshwork at the nuclear envelope and functions to give the nucleus structural support, organize chromatin, and aid diverse cellular signaling pathways. Lamin A/C are particularly critical for maintaining the mechanical properties of the nucleus, and in their role as a mechanosensor of cellular forces, thereby enabling the nucleus to physically and biochemically adapt to its mechanical environment. As such, mutations in LMNA cause a host of human diseases, termed ‘laminopathies,’ that include Dilated Cardiomyopathy (LMNA-DCM) and several forms of muscular dystrophy, among others. However, an incomplete understanding of the cellular disease mechanisms driving DCM and other laminopathies has resulted in a lack of therapeutics sufficient to ameliorate disease pathogenesis. The primary goal of my thesis has been to understand changes to the mechanical conformation of nuclei and to gene expression that may be driving nuclear and cellular dysfunction in LMNA-DCM. Using an induced pluripotent stem cell-derived cardiomyocytes (iPSC-CM) derived from LMNA- DCM patients with several different gene mutations, I found that the mislocalization of Lamin A/C and Lamin B1 from the nuclear envelope in LMNA-mutant iPSC-CMs corresponds to an increase in nuclear damage and fragility, likely due to the inability of the nucleus to withstand mechanical forces. Additionally, as Lamin A/C are heavily involved in control of gene expression and downstream cellular signaling pathways, through RNA-sequencing of cardiac tissue from several Lmna-DCM mouse models and iPSC-CMs carrying the LMNA G449V mutation, I found that metabolism, extracellular matrix, cardiac, and immune response genes and associated signaling pathways are misregulated in both mouse and human models of LMNA Dilated Cardiomyopathy, and misregulated metabolism and extracellular matrix are involved in the disease onset and progression of Lmna N195K mice. A secondary goal of my thesis work was to understand how Lamin A/C mediates the transduction of forces to the nucleus. I found that the Lamin A/C Ig-fold undergoes conformational changes in response to cell seeding density, which allows for the differential binding of antibodies targeting the Ig-fold. As the Ig-fold is home to the binding sites of numerous nuclear proteins involved in mechanotransduction of forces and gene expression and is a hotspot for LMNA mutations that affect nuclear mechanotransduction, clarifying the role and context of conformational changes in the context of healthy mechanotransduction is critical for progress towards a complete understanding of disrupted nuclear organization and mechanotransduction in laminopathies. Together, these studies represent key ways in which Lamin A/C impacts nuclear mechanics and mechanotransduction in health and disease, and exhibit several novel mechanisms through which LMNA mutations drive changes in gene expression to cause cardiac dysfunction, progress which is necessary for the development of novel therapeutics for laminopathies which sufficiently target the underlying cellular disease pathology. BIOGRAPHICAL SKETCH Melanie Elizabeth Maurer obtained her B.S. in Biomedical Engineering from University of Texas at Dallas, where she gained an interest in cardiovascular disease and cellular disease modeling. Afterwards, she spent one year at the University of Stuttgart in Germany creating engineered biomaterials to advance the maturation of stem cell-derived cardiomyocytes. Upon matriculation at Cornell for her Ph.D. in Biomedical Engineering in 2017, she joined Dr. Jan Lammerding’s lab in the Meinig School for Biomedical Engineering and Weill Institute for Cell and Molecular Biology to pursue a project to employ patient-derived induced pluripotent stem cells with LMNA mutations for the study of the Dilated Cardiomyopathy cellular disease pathogenesis. Throughout her undergraduate and graduate tenures, Melanie organized and participated in numerous STEM community outreach programs and gained a passion for scientific communication. As such, upon completion of her Ph.D., she will pursue a career in scientific communication as a Product Scientist at 23andMe, authoring new genetic health reports. v Dedicated to my loving parents John and Sue, who instilled in me a love of learning and encouraged me to reach for the stars. vi ACKNOWLEDGMENTS This work would not have been possible without the steadfast support from my advisor, Dr. Jan Lammerding. His mentorship and commitment to my growth as a scientist, professional, and person has been instrumental to my success during my Ph.D. and will continue to be a positive influence in my future. Thank you for the time you invested in me, for being my advocate and problem-solver in the face of adversity, and for creating an environment where a smile and a laugh is the way of life – it was a joy to be a part of your lab! I would also like to thank my committee members Dr. Ben Cosgrove and Dr. Matthew Paszek for your thought-provoking questions that have made me grow as a scientist, and for your support and mentorship throughout my Ph.D. I would like to acknowledge the labs of the Weill Institute for Cell and Molecular Biology and the Meinig School for Biomedical Engineering for collaborations and thoughtful scientific discussions, and the staff of the Meinig School for Biomedical Engineering, particularly Belinda Floyd and Suzanne Koehl, for the support throughout my Ph.D. I thank my co-authors and scientific collaborators – Dr. Kehan Zhang, Dr. Kathy Xie, Dr. Lori Wallrath, Dr. Shuping Lai, Dr. Ivor Benjamin, Dr. Elizabeth McNally, Anthony Gacita, Dr. Todd Evans, the Gladstone Institute, and the Cure CMD foundation – for making my work possible, and the entire Lammerding Lab for the scientific support, mentorship, and camaraderie. I would specifically like to acknowledge my collaborators and co-authors in the Lammerding Lab, Dr. Hind Zahr, Dr. Greg Fedorchak, Rachel Gilbert, Jineet Patel, and Dr. Patricia Davidson, for making my work possible, and especially Dr. Tyler Kirby for his mentorship from the very beginning of my PhD – I would not be the scientist that I am today without the billion times that vii you asked me what my experimental controls were and let me barrage you with questions. To my undergraduate and M.Eng. students Huaiyao Peng, Shriya Perati, Lindsey Johnson, and Hanna Gimse: it has been a pleasure working with you over the past several years. Thank you for your time, trust, and insightful questions, and you have taught me so many lessons and for those I will always be grateful. My Ph.D. experience would not have been the same without the love and support from my friends, family, and fiancé, Bram. Bram – it has been a joy to grow together as Ph.D. students and as a couple in our time at Cornell, and I will be forever thankful for the love, unending encouragement, and daily support you gave me in getting through my Ph.D. To my parents John and Sue – thank you for making possible the many opportunities to explore science and engineering that I had growing up and for your unrelenting love and support, and to my sisters Katie, Karen, and Jamie – thank you for being my cheerleaders. To all of my friends, especially Richa Agrawal, Joseph Long, and Misha Padidar, thank you for the laughter and support that made the last few years at Cornell some of my favorite years of my life. viii TABLE OF CONTENTS BIOGRAPHICAL SKETCH .......................................................................................................... v SUMMARY PAGE ........................................................................................................................ x CHAPTER 1: Introduction: Nuclear mechanotransduction in cellular function, fate, and disease ....................................................................................................................................................... 13 CHAPTER 2: Impaired lamin localization to the nuclear envelope is responsible for nuclear damage in LMNA mutant iPSC-derived cardiomyocytes ............................................................. 48 CHAPTER 3: Investigating the disrupted signaling pathways in LMNA-Dilated Cardiomyopathy with RNA-seq ............................................................................................................................... 91 CHAPTER 4: The Lamin A/C Ig-fold undergoes cell density-dependent conformational changes that alter epitope accessibility ..................................................................................................... 123 CHAPTER 5: Conclusions and future perspectives .................................................................. 162 APPENDIX ................................................................................................................................. 165 A1. Investigating the role of cytoskeletal forces in LMNA-mutant stem cell-derived cardiomyocyte nuclear damage ..................................................................................................................................... 165 A2. Investigating altered DNA damage response in LMNA-mutant induced pluripotent stem cell derived cardiomyocytes ........................................................................................................................ 179 A3: Investigating YAP and MKL nuclear translocation and LINC complex organization in LMNA- mutant iPSC-CMs ................................................................................................................................. 190 A4: Design and validation of engineered tissue systems for the study of laminopathies ..................... 196 REFERENCES ........................................................................................................................... ccvi ix SUMMARY PAGE Chapter 1. Introduction: Nuclear mechanotransduction in cellular function, fate, and disease Nuclear mechanotransduction is the process by which the nucleus converts mechanical signals into downstream cellular response, and is mediated through numerous mechanisms, including protein conformational changes, the opening of stretch sensitive ion channels, and changes in chromatin modifications and gene expression. This chapter explores the role of nuclear mechanotransduction in examples of health and disease and outlines recent technological advances that enable the study of nuclear mechanotransduction. Chapter 2. Impaired lamin localization to the nuclear envelope is responsible for nuclear damage in LMNA mutant iPSC-derived cardiomyocytes We found that induced pluripotent stem cell-derived cardiomyocyte (iPSC-CM) lines derived from patients with three distinct LMNA mutations exhibit nuclear shape and size deformities, and in the most extreme case, reduced nuclear stiffness and increased fragility. Mislocalization of Lamin A/C and Lamin B1 from the nuclear envelope correlates to the degree of nuclear damage across healthy control and LMNA-mutant iPSC-CM lines, explaining >95% of the variability between cell lines. Chapter 3. Investigating the disrupted signaling pathways in LMNA-Dilated Cardiomyopathy with RNA-seq Using two mouse models of Lmna-Dilate Cardiomyopathy (DCM) and iPSC-CMs derived from either healthy controls or an LMNA-DCM patient with the LMNA G449V mutation, I identified x several misregulated signaling pathways common to both human and mouse models of LMNA- DCM – metabolism, the extracellular matrix, cardiac dysfunction, and immune response – that may be driving mechanisms of LMNA-DCM. Additionally, misregulation of metabolism and the extracellular matrix are involved in the onset and progression of LMNA-DCM in mice expressing the Lmna N195K mutations. This study is among the first to identify signaling pathways common to both mouse and human LMNA-DCM, and to implicate several pathways in the onset of LMNA- DCM. Chapter 4. The Lamin A/C Ig-fold undergoes cell density-dependent conformational changes that alter epitope accessibility Mechanical forces exerted on the nucleus may induce direct stretching of the nuclear lamina and/or conformational changes to the protein structure. Here, we found that the Lamin A/C Ig-fold undergoes conformational changes in response to cell-density, that allow for density-dependent binding of antibodies targeting the Ig-fold. These conformational changes are independent of nuclear compression, spread area, and force transmission through the LINC complex, and are driven by the polymerization of actin and microtubules. Appendix. Unfinished work and negative results Here, in the format of “mini chapters,” I describe several unfinished projects and preliminary results pertaining to the disease mechanisms of LMNA-DCM, including the role of forces of the cytoskeleton in nuclear damage, alterations to DNA damage response in LMNA-mutant iPSC- CMs, and an example of mechanotransduction, YAP translocation to the nucleus, that remains unchanged in LMNA-mutant iPSC-CMs. Additionally, I discuss the development of several tissue xi engineering approaches for the study of LMNA-DCM and skeletal muscle laminopathies. xii CHAPTER 1 Introduction: Nuclear mechanotransduction in cellular function, fate, and disease1 Cellular behavior is continuously affected by microenvironmental forces through the process of mechanotransduction, in which mechanical stimuli are rapidly converted to biochemical responses. Mounting evidence suggests that the nucleus itself is a mechanoresponsive element, reacting to cytoskeletal forces and mediating downstream biochemical responses. The nucleus responds through a host of mechanisms, including partial unfolding, conformational changes, and phosphorylation of nuclear envelope, modulation of nuclear import/export, and altered chromatin organization, resulting in transcriptional changes. It is unclear which of these events present direct mechanotransduction processes and which are downstream of other mechanotransduction pathways. We critically review and discuss the current evidence for nuclear mechanotransduction, particularly in the context of stem cell fate, a largely unexplored topic, and in disease, where an improved understanding of nuclear mechanotransduction is beginning to open new treatment avenues. Finally, we discuss innovative technological developments that will allow outstanding questions in the rapidly growing field of nuclear mechanotransduction to be answered. Introduction Mechanotransduction refers to the process by which cells convert mechanical stimuli from their extracellular environment or cell-generated forces into biochemical signals to induce downstream cellular responses. Mechanical forces can propagate along the cytoskeleton and travel at speeds of 1 This chapter is an adaptation from the following published manuscript: Maurer, M, and Lammerding, J (2019). The Driving Force: Nuclear Mechanotransduction in Cellular Function, Fate, and Disease. Annu Rev Biomed Eng 21, 443–468. 13 up to 30 µm/s, an impressive rate that is 25 times faster than molecular motor transport and 12.5 times faster than passive diffusion of signaling molecules (Wang et al., 2009). This rapid conversion of physical to biochemical response enables the rapid adaptation of cells to their changing physical environment (Doyle and Yamada, 2016; Cho et al., 2017). Mechanotransduction can play a critical role in cell and tissue differentiation, maintenance, and disease, for example, in the adaptation of bones and muscle to exercise or the alignment of endothelial cells to fluid shear (Jaalouk and Lammerding, 2009). Since the term mechanotransduction is often used more broadly to refer to any cellular responses to mechanical stimuli, including events downstream of the original transduction event, in this review we use the following definitions: Mechanotransmission refers to the transmission of mechanical forces through cellular components, such as along actin stress fibers or microtubules, but does not include the actual mechanotransduction process. Mechanosensing refers to the actual transduction process, which is typically limited to some specialized proteins and locations within the cell; many of these proteins, including specific focal adhesion proteins and stretch-sensitive ion channels in the plasma membrane (Wang et al., 1993; Martinac, 2004), have been recognized in the last three decades, but others may remain to be identified. Mechanotransduction signaling describes the signaling pathways that are downstream of the initial mechanosensing event. Importantly, many of these pathways, such as mitogen-activated protein kinase (MAPK) signaling or YAP/TAZ translocation, can be activated by a variety of upstream signals, not only mechanical stimuli but also biochemical signals, such as growth factor binding to cell surface receptors (González et al., 2008; Dupont et al., 2011; Driscoll et al., 2015). 14 Recently, a growing number of studies, including some on isolated nuclei, have implicated the nucleus itself as a mechanosensing element (Swift et al., 2013; Guilluy et al., 2014; Guilluy and Burridge, 2015). Several models have been proposed to explain how mechanical forces acting on the nucleus could induce changes in nuclear envelope composition, chromatin organization, and gene expression, which then drive downstream cellular responses (Szczesny and Mauck, 2017; Kirby and Lammerding, 2018), including in stem cell differentiation. At the same time, many of the reported findings linking mechanical factors and nuclear changes have been rather correlative (Cho et al., 2017; Kirby and Lammerding, 2018), and it often remains unclear whether the observed nuclear changes were upstream or downstream of other events, including established cytoplasmic mechanotransduction pathways. In this review, we provide a summary of nuclear structure and describe how these nuclear components contribute to nuclear mechanics, mechanotransmission, and mechanosensing. As many current efforts seek to understand how stem cells respond to their mechanical microenvironment to control differentiation and cell fate commitment, we discuss how nuclear mechanotransduction may be involved. Since defects in nuclear mechanics and mechanotransmission are linked to impaired mechanotransduction signaling and several human diseases, particularly affecting skeletal and cardiac muscle (Jaalouk and Lammerding, 2009), we discuss the current understanding of nuclear mechanotransduction in disease pathogenesis. Finally, we outline some of the recent technological advances in unraveling the mechanisms by which the nucleus acts in mechanotransduction and in deepening our understanding of the diseases caused by such mechanisms. 15 The Nucleus and Nuclear Mechanics Nuclear structure and organization As the compartment containing the vast majority of the genome and the site of gene transcription, the nucleus arguably plays the most important role in guiding cellular fate, behavior, and adaptation. The nucleus contains DNA that is wrapped around histones, which are organized into higher-order structures, broadly categorized as either open, transcriptionally active euchromatin or condensed, inactive heterochromatin. The nuclear envelope consists of outer and inner nuclear membranes (ONM and INM, respectively) separated by the 30–50-nm-wide perinuclear space (PNS) (Fig. 1.1). This double membrane serves as a physical barrier to protect the nuclear contents and to control exchange of large (>30-kDa) molecules between the cytoplasm and the nuclear interior through nuclear pore complexes (NPCs). NPCs regulate exchange across the nuclear envelope, directly connect to both the nucleoskeleton and the cytoskeleton, and interact with chromatin (Soheilypour et al., 2016). Below the INM exists the 10–30-nm-thick, fibrous meshwork of the nuclear lamina (Turgay et al., 2017). The nuclear lamina is composed mostly of lamins, which are type V nuclear intermediate filament proteins nearly ubiquitously expressed in differentiated cell types. Mammalian cells express two types of lamins—A-type and B-type—as products of three genes. In somatic cells, the LMNA gene gives rise to two major A-type lamin isoforms, Lamin A and Lamin C, plus some less common isoforms, as the result of alternative splicing; the LMNB1 and LMNB2 genes encode Lamins B1 and B2, respectively. Lamins A and C are developmentally regulated and appear during cellular differentiation. In contrast, all cells express at least one B-type lamin, even though recent studies show that cells lacking both Lamin B1 and B2 are viable (Yang et al., 2011; Kim et al., 16 Fig. 1.1. Constituents of the nucleus and nuclear envelope involved in mechanotransduction. The LINC (linker of the nucleoskeleton and cytoskeleton) complex— nesprins at the outer nuclear membrane (ONM) and SUN (Sad1p and UNC-84 homology)- domain proteins at the inner nuclear membrane (INM)—spans the nuclear envelope, interacting with cytoskeletal filaments and associated proteins and the nuclear lamina to enable force transmission between the cytoskeleton and nuclear interior. Nuclear lamins (A/C and B1/2) form independent yet interacting meshworks underneath the INM and are responsible for maintaining nuclear shape and stiffness. Both A- and B-type lamins interact with nuclear pore complexes (NPCs), chromatin, and various other binding partners at the nuclear envelope and the nuclear interior. NPCs enable molecular transport between the cytoplasm and nucleoplasm. 17 2013). The various lamin isoforms form independent yet interacting meshworks with a highly branched architecture at the nuclear periphery (Shimi et al., 2015; Xie et al., 2016a). Surprisingly, recent cryo–electron tomography (cryo-ET) studies indicated that A- and B-type lamins form 3.5- nm-diameter tetrameric filaments, which are substantially thinner than the 10-nm-diameter cytoplasmic intermediate filaments (Turgay et al., 2017). Notably, a fraction of lamins, particularly A-type, also exist in the nucleoplasm. Lamins have many binding partners, including chromatin, transcription factors, and LEM (LAP2, Emerin, and MAN1) family proteins, that play critical roles in gene regulation (Dorner et al., 2007). Since their discovery four decades ago (Gerace et al., 1978), nuclear lamins have attracted increased interest as more evidence of their vital roles in cellular functions and disease has emerged. Within the nucleus, lamins regulate transcription, DNA replication and repair, and chromatin organization (Gruenbaum and Foisner, 2015; de Leeuw et al., 2018). Heterochromatin exists at the nuclear periphery and interacts with the nuclear lamina in Lamin-associated domains (LADs) (Kourmouli et al., 2001; Gurudatta et al., 2010) and via Lamin-associated protein 2 (LAP2) and its binding partner, barrier to autointegration factor (BAF) (Goldman et al., 2002; Holaska et al., 2003). These interactions may directly affect chromatin organization, nuclear mechanotransduction, and gene expression and may contribute to stem cell differentiation (see Sections 3 and 4). Furthermore, Lamin A is required to retain Emerin at the INM (Vaughan et al., 2001; Guo et al., 2014), which in turn modulates expression of mechanosensitive genes (Lammerding et al., 2005) and is required for Nesprin-1-mediated nuclear envelope remodeling in response to mechanical force (Guilluy et al., 2014). Depletion of Lamin B1 results in an enlarged or loose Lamin A/C meshwork with blebs of the nuclear envelope that contain Lamin A/C and 18 euchromatin (Shimi et al., 2008; Guo et al., 2014; Shimi et al., 2015). Similarly, depletion of Lamin A/C causes loosening of the Lamin B1 meshwork and mislocalization of Emerin and other nuclear envelope proteins away from the nuclear envelope, highlighting the interconnections between various nuclear envelope components (Shimi et al., 2008, 2015). At the nuclear envelope, lamins are responsible for positioning and distribution of NPCs, with both A- and B-type lamins binding nucleoporin (Nup153). Lamins also modulate nuclear assembly and disassembly during cellular replication (de Leeuw et al., 2018), as well as nuclear shape, stiffness, and structure by regulating cytoskeletal organization (Lammerding et al., 2004b; Shimi et al., 2008; Dechat et al., 2010). Loss of Lamin A/C results in disturbed perinuclear actin, microtubule, and intermediate filament organization and changes in focal adhesions (Nikolova et al., 2004; Khatau et al., 2009; Chandar et al., 2010; Lombardi et al., 2011; Kim et al., 2017). On a larger scale, lamins play a critical role in migration through three-dimensional (3D) environments by governing the deformability of the large nucleus, which constitutes a rate-limiting factor in confined migration (reviewed in McGregor et al., 2016). Nucleo-cytoskeletal connections The LINC (linker of the nucleoskeleton and cytoskeleton) complex connects the nuclear lamina to the cytoskeleton (Crisp et al., 2006) and is critical in force transmission from the cytoskeleton to the nuclear interior, termed nucleo-cytoskeletal coupling (Lombardi et al., 2011). The LINC complex consists of SUN (Sad1p and UNC-84 homology)- and KASH (Klarsicht, ANC-1, and Syne homology)-domain proteins, named after their conserved C-terminal domains that interact across the luminal space. LINC complex proteins span the nuclear envelope (Fig. 1.1) and 19 are anchored at the nuclear envelope via lamins, NPCs, and interaction with chromatin (Crisp et al., 2006). In mammalian somatic cells, Sun1 and Sun2 constitute the SUN-domain proteins; Nesprin-1, -2, -3, and -4, each with multiple isoforms, the KASH-domain proteins. Germ cells express additional LINC complex proteins. SUN-domain protein trimers in the INM interact with the lamina at the INM and KASH-domain proteins in the PNS (Tapley and Starr, 2013). KASH- domain proteins in the ONM protrude into both the PNS and cytoplasm, where they bind the cytoskeleton. Nesprin family proteins include isoforms that can directly bind F-actin, interact with the microtubule motors kinesin and dynein and with each other, and bind the adaptor protein plectin to interact with intermediate filaments (Horn, 2014). The LINC complex plays crucial roles in mechanical processes such as nuclear movement, positioning, and shape, as well as chromatin positioning and gene expression (Rothballer and Kutay, 2013; Banerjee et al., 2014). Disruption of the interaction between nuclear lamins and LINC complex proteins, for example, through mutations in the LMNA gene responsible for various diseases (see Section 5), results in loss of nucleo-cytoskeletal coupling, perturbed cytoskeletal organization, loss of nuclear stiffness, and inability of the nucleus to properly respond to force (Isermann and Lammerding, 2013). In cells adhering on rigid two-dimensional (2D) surfaces, a perinuclear organization of dynamic apical actomyosin filaments (referred to as a perinuclear actin cap by some authors) spans the top of the nucleus and connects to the nuclear interior via the LINC complex (Khatau et al., 2009). Anchored at focal adhesions at the cellular basal surface, these perinuclear actin filaments apply compressive forces to the apical surface of the nucleus (Lele et al., 2018; Shiu et al., 2018). Together with lateral forces transmitted to the nucleus from other cytoskeletal structures, such as microtubules, these forces strongly influence nuclear shape (Tariq et al., 2017; Lele et al., 2018). 20 This concept is discussed further in the context of mechanotransduction in Section 3. Nuclear mechanics Together, the highly interconnected nuclear constituents described above mediate the transmission of mechanical forces from the cytoskeleton to the nucleus, while providing structural support to the nucleus and defining its mechanical properties (Isermann and Lammerding, 2013). The nucleus exhibits elasticity and compressibility that enable the nucleus to act as a mechanical shock absorber (Dahl et al., 2004). Both the nuclear lamina and chromatin contribute to nuclear mechanical response to strain (Stephens et al., 2017, 2018a). Nuclear lamina stretch dominates at nuclear strains above 30%, while the mechanical properties of chromatin, which exhibits viscoelastic properties, govern nuclear deformation at lower strains (Stephens et al., 2017). Physical cross- links between chromatin and INM proteins such as SUN-domain proteins can provide further mechanical stability to the nucleus (Schreiner et al., 2015). Prolonged mechanical stress can cause irreversible deformation and reorganization of chromatin, which may correspond to altered transcriptional or differentiation states (Pajerowski et al., 2007). Lamin A/C levels correlate with nuclear stiffness and ability to withstand force: Increased levels result in stiffer and more viscous nuclei (Lammerding et al., 2006; Swift et al., 2013), whereas decreased levels correspond to softer, more deformable nuclei with increased fragility (Broers et al., 2004; Lammerding et al., 2004b, 2006; De Vos et al., 2011). Lamin phosphorylation results in increased solubility of lamins, decreased polymerization of the lamina, increased lamin turnover, and reduction of cellular tension and nuclear stiffness (Swift et al., 2013; Buxboim et al., 2014). 21 The Role of the Nucleus in Mechanotransduction Many non-mutually-exclusive mechanisms of nuclear mechanotransduction have been proposed to date. In this section, we briefly discuss the major proposed mechanisms and the evidence supporting them (Fig. 1.2); we refer readers to excellent recent reviews (Szczesny and Mauck, 2017; Kirby and Lammerding, 2018) for more details and additional proposed mechanisms. Our discussion focuses on general mechanisms of nuclear mechanotransduction; it is important to recognize that different cell types may respond differently to mechanical stimuli owing to differences in both structural organization (e.g., when comparing epithelial cells, mesenchymal fibroblasts, leukocytes, and muscle cells) and cell-type-specific signaling pathways downstream of the mechanosensing event. Nuclear membrane and pore stretching During nuclear membrane stretch, NPCs take on a dilated, more open conformation (Elosegui- Artola et al., 2017; Enyedi and Niethammer, 2017). Although NPCs make up less than 10% of the nuclear surface area at rest (Mazzanti et al., 2001), the mechanically induced increase in NPC diameter accounts for one-sixth of the total increase in nuclear membrane surface area in HeLa cells during nuclear swelling (Enyedi and Niethammer, 2017). Force application to the nucleus, such as during cell spreading on a 2D substrate or by indentation with an atomic force microscopy tip, triggers partial opening of nuclear pores, promoting active nuclear import of YAP (Elosegui- Artola et al., 2017). Osmotic swelling, however, does not trigger YAP import YAP (Elosegui- Artola et al., 2017), suggesting that some elements, such as Nesprin-1 (Driscoll et al., 2015) or other LINC complex constituents, are required for the opening of nuclear pores and the mechanosensing process. 22 Fig. 1.2. Proposed mechanisms of nuclear mechanotransduction. (a) Force application to the nucleus can result in conformational changes of nuclear envelope proteins, such as partial unfolding of lamins (Swift et al, 2013; Ihalainen et al, 2015), and phosphorylation of nuclear proteins, including lamins, SUN-domain proteins, and Emerin (Guilluy et al, 2014; Swift et al, 2013; Buxboim et al, 2014). (b) Nuclear membrane stretch in response to force opens nuclear pore complexes (NPCs) (Enyedi and Niethammer, 2017; Elosegui-Artola et al, 2017) and calcium channels (Enyedi et al, 2016; Cho and Stahelin, 2005) on the cytoplasmic side, thus increasing molecular influx into the nucleoplasm. The increased import of transcription factors (TFs) into the nucleoplasm can alter gene expression (Elosegui-Artola et al, 2017). (c) Mechanical forces acting on the nucleus can induce chromatin stretching, opening, and compaction, including DNA and histone modifications, that alter accessibility to transcription factors and lead to changes in gene expression (Li et al, 2011; Jain et al, 2013; Le et al, 2016, Heo et al, 2015; Iyer et al, 2012; Tajik et al, 2016). 23 Flux of calcium in response to force application or nuclear stretching may constitute another nuclear mechanosensing mechanism. NPCs and calcium channels on the nuclear envelope, including L-type, InsP3, cyclic ADP ribose-modulate, and possibly others, regulate the influx and efflux of calcium from the nucleus (Santella and Carafoli, 1997; Malviya and Rogue, 1998; Itano et al., 2003). Cell spreading and nuclear stretching increase nuclear calcium through stretch- activated calcium channels on the nuclear membrane, which enhances transcription factor (CREB) expression and regulates gene transcription, protein import, apoptosis (Malviya and Rogue, 1998; Itano et al., 2003), and downstream mechanosignaling (Enyedi et al., 2016). In addition to the opening of NPCs and channels, nuclear envelope stretch loosens packing of the nuclear membrane phospholipid bilayers, allowing for the insertion of hydrophobic protein residues into the bilayer (Enyedi et al., 2016). Osmotic swelling in response to tissue damage triggers nuclear translocation of cytosolic phospholipase A2 (cPLA2) and 5-lipoxygenase (5-LOX) and incorporation of these proteins into the INM, where their activity triggers downstream inflammatory signaling cascades (Enyedi et al., 2016). Mechanosensitive incorporation of cPLA2 and 5-LOX is regulated by increased calcium levels in the cell, which aids residue insertion into the membrane (Cho and Stahelin, 2005; Enyedi et al., 2016), and by nuclear lamina rigidity, as a stiff nuclear lamina may not as readily allow stretching of the INM and therefore reduces protein incorporation into the INM (Enyedi et al., 2016; Stephens et al., 2018a). Protein phosphorylation and conformation change in response to mechanical force Phosphorylation states serve as common mechano-switches in mechanical response, such as cytoskeletal stretch-dependent phosphorylation of Cas (Tamada et al., 2004; Sawada et al., 2006) 24 for contraction or phosphorylation of Paxillin and Vinculin during tension-mediated focal adhesion maturation (Pasapera et al., 2010). In the nucleus, Lamin A/C and Emerin phosphorylation modulate nuclear stiffness and nucleo-cytoskeletal coupling in response to mechanical stimulation (Swift et al., 2013; Buxboim et al., 2014; Guilluy et al., 2014). Lamin A/C phosphorylation increases in cells with low cytoskeletal tension, that is, when cells are grown on soft substrates (Swift et al., 2013; Buxboim et al., 2014), increasing Lamin A/C mobility and turnover (Buxboim et al., 2014; Kochin et al., 2014). Conversely, when forces are applied to the nucleus via nesprins, Src-mediated Emerin phosphorylation recruits Lamin A/C to the nuclear periphery and promotes Sun2–Lamin A/C interactions (Guilluy et al., 2014). The precise mechanism by which mechanical forces can modulate phosphorylation of nuclear envelope proteins remains unclear, including whether this process is regulated by altering kinase activities or accessibility of the kinase substrate amino acids. Regardless, the observed mechanically induced phosphorylation implicates a structural role for phosphorylation in mechanotransduction through control of nuclear stiffening and nucleo-cytoskeletal coupling (Swift et al., 2013; Buxboim et al., 2014; Guilluy et al., 2014), which can also affect downstream transcription by downregulating some mechanoresponsive genes (VCL and SRF) and reducing YAP/TAZ translocation into the nucleus (Guilluy et al., 2014). In addition to phosphorylation, protein conformation plays a role in mechanical response at the nuclear envelope. Partial unfolding of the Lamin A C-terminal immunoglobulin (Ig)-like fold in response to mechanical forces may expose normally hidden residues, such as Cys522 (Swift et al., 2013; Ihalainen et al., 2015). This conformational change could alter interaction with binding partners, expose cryptic signaling sites, or destabilize the protein. Unfolding may expose some amino acid residues to kinases, thus allowing for altered phosphorylation and modulating 25 downstream signaling. Chromatin stretching, organization, and modification Mechanical microenvironmental cues, such as architecture and mechanical loading (e.g., tension and compression), alter chromatin modifications and condensation to control gene expression (Li et al., 2011; Iyer et al., 2012; Jain et al., 2013; Heo et al., 2015, 2016b; Le et al., 2016). Dynamic mechanical loading can cause rapid short-lived, prolonged, and even irreversible changes in chromatin condensation, depending on the intensity and duration of the mechanical load (Iyer et al., 2012; Heo et al., 2015, 2016b). Highly transcriptionally active chromosomes preferentially orient along the mechanical axis of the nucleus on anisotropic micropatterned materials (Jain et al., 2013; Maharana et al., 2016; Wang et al., 2017), demonstrating that chromatin organization is responsive to extracellular and cytoskeletal mechanical cues. Such changes in chromatin organization likely affect the transcriptional profile of the cells. Although these phenomena are widely observed, the specific mechanisms guiding mechanoresponsive gene expression are not well characterized. In particular, it remains unclear whether these changes are direct responses to forces acting on the nucleus or are downstream of other mechanotransduction events. Importantly, the observed changes in chromatin organization, condensation, and modification are dependent on the actin cytoskeleton and LINC complex (Iyer et al., 2012; Heo et al., 2015, 2016b; Kim et al., 2015). Perinuclear actin filaments bind to the LINC complex on the apical surface of the nucleus and cause accumulation of Lamin A/C and hyperacetylated, transcriptionally active euchromatin at the INM, demonstrating that the perinuclear actin filaments interact with euchromatin via nucleo-cytoskeletal coupling (Kim and Wirtz, 2015). Furthermore, cytoskeletal 26 contraction triggers mechanosensitive ATP release and calcium signaling to mediate nuclear import and activation of the histone–lysine N-methyltransferase EZH2 and histone deacetylase (HDAC) load (Iyer et al., 2012; Jain et al., 2013; Heo et al., 2015, 2016b), which stimulates gene silencing by altering methylation (Le et al., 2016) and gene transcription by increasing histone acetylation (Li et al., 2011; Jain et al., 2013). Prolonged force application drives changes in methylation states for gene regulatory control by decoupling heterochromatin from the nuclear lamina, and driving chromatin compaction and a switch from H3K9me3 to H3K27me3 to attenuate transcription and silence promotors (Le et al., 2016). Previous research suggested that force-dependent transcriptional regulation depends upon lamin– chromatin interactions (Shimi et al., 2008), but until recently, studies have struggled to show a direct effect of mechanical force on chromatin to control transcription. Wang and colleagues (Tajik et al., 2016) used 3D magnetic twisting cytometry to apply extracellular stretching with RGD- coated magnetic beads, which demonstrated the direct stretching of a reporter transgene flanked by two green fluorescent protein–labeled loci and rapid, stretch-dependent transcription of the reporter gene. This study suggests that force is transmitted through integrins, the actin cytoskeleton, the LINC complex, and then lamin–chromatin interactions, which stretch chromatin and result in upregulation of transcription (Tajik et al., 2016). Disruption of any one of these components weakens the mechanically induced response (Tajik et al., 2016). Nonetheless, studies using endogenous genes will be required to confirm these findings in a general context, and it remains unclear how chromatin stretching results in activation of specific mechanosensitive genes. Euchromatin endures greater deformation under strain than heterochromatin, which would induce larger conformational changes (Iyer et al., 2012), and may promote stretch-dependent 27 transcription. Nuclear and perinuclear actin Recently, nuclear and perinuclear actin assemblies have emerged as key players in nuclear mechanotransmission and mechanosignaling. Nuclear actin polymerization regulates nuclear structure and gene expression (Holaska et al., 2003; Lattanzi et al., 2003; Olson and Nordheim, 2010; de Lanerolle and Serebryannyy, 2011; Baarlink et al., 2013). The LINC complex mediates nuclear actin polymerization in response to cell spreading to form a nuclear scaffold (Wang et al., 2009; Plessner et al., 2015), which is accelerated by Emerin binding to the actin pointed end (Holaska et al., 2003; Lattanzi et al., 2003). Both Lamin A/C and Emerin bind nuclear actin, thereby increasing nuclear strength (Holaska et al., 2003; Lattanzi et al., 2003). Furthermore, nuclear actin acts as a transcriptional cofactor for polymerases I, II, and III (de Lanerolle and Serebryannyy, 2011). Nuclear actin polymerization can regulate transcription factor activity via increased import and export, primarily through myocardin-related transcription factor A (MRTF- A) and serum response factor (SRF), which demonstrates the downstream effects of force-driven nuclear actin dynamics (Olson and Nordheim, 2010; Baarlink et al., 2013). Highlighting the interplay between nuclear envelope proteins, actin, and MRTF-A/SRF, loss of Lamin A/C and Emerin disturbs nuclear and cytoskeletal actin dynamics and impairs MRTF-A/SRF signaling (Ho et al., 2013). Applied force can induce perinuclear actin filament assembly within minutes (Iyer et al., 2012; Shao et al., 2015), in a process that requires Lamin A/C, Emerin, and the LINC complex (Khatau et al., 2009; Lombardi et al., 2011; Kim et al., 2017; Shiu et al., 2018). The presence of perinuclear 28 actin is key in mechanotransmission of forces to the nucleus via the LINC complex, but the initial polymerization reaction likely occurs downstream of Rho GTPase (Iyer et al., 2012) and calcium mechanosignaling (Shao et al., 2015). Thus, perinuclear actin plays a crucial role in mechanotransmission to the nucleus, a requirement for nuclear mechanotransduction. At the same time, perinuclear—and nuclear—actin polymerization is downstream of other mechanoresponsive signaling pathways and can further modulate mechanotransduction signaling by interaction with MRTF-A and SRF. As a whole, this mechanosensitive mechanistic web is thought to work to guide cellular functions, such as stem cell fate (discussed in Section 4), and disruption of this intricate network can cause a host of human diseases (see Section 5). Nuclear Mechanics Guide Stem Cell Fate In addition to soluble factors, the stem cell microenvironment provides mechanical stimulation to guide lineage commitment and differentiation. Seminal research by Engler et al. (Engler et al., 2006) demonstrated that mesenchymal stem cell (MSC) fate is guided by extracellular matrix (ECM) elasticity. Motivated by these findings, researchers have focused on harnessing the mechanical environment for directing stem cell differentiation (i.e., mechanically induced differentiation), both with (Engler et al., 2006; Lee et al., 2013; Chaudhuri et al., 2016) and without (Heo et al., 2015, 2016a) the use of soluble factors. It is now recognized that matrix geometry, stiffness, adhesion, stress relaxation, micro- and nanopatterned surfaces, and applied cellular stretch can guide stem cell fate (Lee et al., 2013; Driscoll et al., 2015; Heo et al., 2015, 2016a; Chaudhuri et al., 2016). Nonetheless, the specific mechanisms by which stem cell nuclei adapt to and differentiate within their mechanical environments remain incompletely understood. Thus, this section highlights nuclear mechanotransduction mechanisms guiding stem cell fate and describes 29 how mechanotransduction can instill mechanical memory of differentiation states (Fig. 1.3). Mechanisms of stem cell nuclear mechanotransduction for guiding cell fate Compared with somatic cells, stem cells exhibit altered DNA and histone modifications (Li, 2002), including highly condensed chromatin conformations, primarily at the nuclear periphery (Labrador and Corces, 2002; West and Fraser, 2005), and altered expression of nuclear envelope proteins (Constantinescu et al., 2006; Pajerowski et al., 2007; Melcer et al., 2012). Stem cells either completely lack or have reduced levels of Lamin A/C, resulting in more deformable nuclei (Pajerowski et al., 2007). Lamin A/C interacts with chromatin to control gene expression (Peric- Hupkes et al., 2010b) and restricts heterochromatin protein dynamics (Constantinescu et al., 2006; Melcer et al., 2012). However, lamins are not required for differentiation. Embryonic stem cells (ESCs) lacking Lamin A/C, B1, and B2 [i.e., triple knockout (TKO)] differentiate into all three germ layers in vitro (Kim et al., 2013); keratinocyte-specific lamin TKO does not interfere with gestation in vivo but causes fatal skin defects upon birth (Jung et al., 2014). Together, these results suggest that lamins may be required for proper tissue architecture, rather than differentiation or organogenesis (Kim et al., 2013; Jung et al., 2014). Nonetheless, experiments with MSCs and pluripotent stem cells (PSCs) indicate an intriguing connection between Lamin A/C and mechanically induced differentiation. MSCs, which express Lamin A/C, can undergo mechanically induced differentiation (Engler et al., 2006; Lee et al., 2013; Heo et al., 2015, 2016a; Chaudhuri et al., 2016), whereas minimal progress has been made toward mechanically induced differentiation in PSCs, which express little to no Lamin A/C (Murphy et al., 2014). One potential pathway is the mechanosensitive phosphorylation of Lamin A/C, which enables nucleoplasmic Lamin A/C–LAP2α complex formation and subsequent regulation of adult stem cell proliferation 30 and differentiation pathways, such as through retinoblastoma protein, to control stemness (Gotic et al., 2010; Gesson et al., 2014). Given the intimate role of Lamin A/C in mechanotransduction, its specific contributions in regulation of mechanically induced stem cell differentiation should be further explored. Fig. 1.3. Nuclear mechanics guide stem cell fate and mechanical memory. (a) Stem cells may undergo mechanically induced differentiation in response to matrix mechanical properties, surface structure, and geometry. Nuclear mechanotransduction in response to matrix sensing alters the transcriptional program to ultimately guide downstream lineage commitment and cellular mechanical properties. (b) Substrate stiffness may enable stem cells to exhibit mechanical memory, in which a stiff phenotype is remembered upon transfer to culture on a soft substrate, via nuclear YAP retention and chromatin condensation. Mechanical memory can include increases in nuclear levels of YAP, nuclear stiffness, chromatin condensation, and expression of Runt-related transcription factor 2 (RUNX2). In addition to lamins, LINC complex proteins play a role in mechanically induced differentiation of MSCs. Nesprin-1 promotes mechanoresponsive YAP nuclear import (Driscoll et al., 2015) and is required for force transmission to the nuclear lamina and chromatin. Conversely, cyclic tensile strain downregulates Sun2 in MSCs, causing a global drop in transcription (Gilbert et al., 2018), 31 downregulation of tubulin expression (Yang et al., 2018), and disturbed perinuclear microtubule organization (Yang et al., 2018), causing nucleo-cytoskeletal decoupling. Taken together, these results suggest both positive and negative roles of the LINC complex and nuclear envelope in mechanically induced differentiation by mediating cytoskeletal organization, nucleo-cytoskeletal coupling, and regulation of gene expression through transcription factor import and signaling regulation. However, future studies should aim to further elucidate the roles of LINC complex proteins in nuclear mechanotransduction and mechanically induced differentiation of stem cells. Additionally, stem cell pluripotency genes may be subject to mechanosensitive activation and silencing via downstream transcriptional control and chromatin modifications (Chowdhury et al., 2010; Le et al., 2016). Mechanical strain independently localizes Emerin to the ONM (Guilluy et al., 2014; Le et al., 2016), which reduces H3K9me3-silenced heterochromatin, promotes the polymerization of perinuclear actin, and reduces nuclear actin levels (Le et al., 2016). The decrease in nuclear actin diminishes RNA polymerase II activity, resulting in attenuated transcription, accumulation of phosphorylation, and H3K27me3 modification of chromatin, which corresponds to a more silenced state (Le et al., 2016). Inhibiting this mechanism reduces methylation-mediated silencing, lineage commitment, and morphogenesis (Le et al., 2016). Thus, this mechanism could explain how mechanically induced differentiation without soluble factors may be achieved: through regulation of stemness or promotion of differentiation. Mechanically induced differentiation and mechanical memory Mechanically induced differentiation has introduced the intriguing concept that stem cells can exhibit so-called mechanical memory. Whereas mechanotransduction typically involves rapid responses to changes in the physical environment of cells, this mechanical memory may allow 32 stem cells to retain information and results of past mechanical conditions, which influences their future behavior and phenotype (Fig. 1.3b). For example, culture of MSCs on stiff materials results in sustained nuclear YAP localization and osteogenic RUNX2 expression, even when cells are transferred to a soft substrate, on which YAP is typically cytoplasmic and RUNX2 is not expressed (Yang et al., 2014). Initial nuclear YAP translocation is likely mediated by cytoskeletal mechanotransduction and nuclear membrane stretch to open NPCs (Elosegui-Artola et al., 2017) via nucleo-cytoskeletal coupling through Nesprin-1 (Guilluy et al., 2014; Driscoll et al., 2015). Subsequent nuclear stiffening, triggered by phosphorylation of Emerin to facilitate recruitment of Lamin A/C to the nuclear envelope (Driscoll et al., 2015), may contribute to the memory effect. Additionally, condensed chromatin stabilized via actin polymerization can persist after mechanical loading to create mechanical memory (Heo et al., 2015). Both of these mechanisms trigger chromatin condensation and nuclear stiffening, which correspond to a more differentiated state (Heo et al., 2016a). As mechanotransduction and signaling typically result in rapid adaptation to the mechanical environment, the concept of sustained mechanical memory is somewhat paradoxical: How do the classical mechanosensing mechanisms achieve permanent changes that resist further adaptation when the mechanical conditions have changed? The answer may lie in the persistent changes associated with stem cell differentiation. Mechanically induced stem cell differentiation causes altered chromatin organization, chromatin modifications, and gene expression, including that of nuclear and cytoskeletal proteins, thereby affecting nuclear mechanics, mechanotransmission, and mechanotransduction. These mechanoresponsive changes may be permanent and cannot easily be overcome by subsequent changes in the physical microenvironment, such as stiff to soft substrates 33 or cessation of loading. However, further research is needed to fully elucidate the molecular events underlying the mechanical memory of stem cells and to determine how to harness this knowledge for applications using stem cells. Nuclear Mechanotransduction Gone Wrong: Spotlight on Laminopathies Collectively, the laminopathies refer to diseases arising from mutations in the LMNA and LMNB genes. More than 450 different LMNA mutations give rise to ∼14 different human diseases (see http://www.umd.be/LMNA/). Examples of human LMNA laminopathies include autosomal dominant Emery–Dreifuss muscular dystrophy (EDMD), dilated cardiomyopathy (DCM) with conduction defects, and Hutchinson–Gilford progeria syndrome. Many laminopathies primarily affect mechanically stressed tissues such as skeletal muscle, heart, and tendons. In contrast, only two human diseases have been associated with the LMNB1 and LMNB2 genes to date: adult-onset autosomal dominant leukodystrophy, resulting from duplication of the LMNB1 gene (Padiath et al., 2006), and acquired partial lipodostrophy, associated with mutations in the LMNB1 gene (Hegele et al., 2006). Most laminopathies are currently incurable, and several result in premature death. Intriguingly, mutations in genes encoding the LINC complex proteins (Emerin, Nesprin- 1/2, Sun1/2) can cause several of the same or similar human diseases as LMNA mutations, including EDMD, DCM, and Charcot–Marie–Tooth syndrome (reviewed in Crisp et al., 2006). Thus, these diseases are also referred to as nuclear envelopathies. With similar disease phenotypes observed in these nuclear envelopathies, altered nucleo-cytoskeletal coupling, nuclear mechanics, and disturbed mechanotransduction could be clear culprits in the disease pathology. 34 Fig. 1.4. Defective mechanotransduction as a bridge between laminopathy hypotheses. Structural defects (increased nuclear fragility that leads to breakage and cell death) and gene misregulation (altered gene activation and silencing) are the two primary hypothesized mechanisms responsible for the muscle-specific defects in many laminopathies. A third hypothesis—defective nuclear mechanotransduction—synthesizes both the structural disruption and gene misregulation hypotheses, as it can explain how downstream gene misregulation might be a product of nuclear weakness due to disruption of mechanotransduction mechanisms in and on the nucleus. Classically, two cellular mechanisms by which laminopathies act in disease have been suggested: structural disruption and gene misregulation. The structural hypothesis proposes that mutant lamins cause nuclear fragility, leading to increased nuclear damage and cell death, particularly in mechanically stressed tissues. The gene regulation hypothesis suggests that lamin mutations play a tissue-specific role in gene expression by altering gene activation and silencing (Peric-Hupkes 35 et al., 2010b) or by inhibiting tissue-specific factor binding (Simon and Wilson, 2013). Impaired stem cell differentiation caused by mutant lamins has been proposed as part of the gene regulation hypothesis. A third hypothesis, disrupted nuclear mechanotransduction, can bridge the mechanistic gap between the structural and gene regulation hypotheses, as disturbed gene regulation may, at least in part, be the product of physical disruption of nuclear mechanotransmission and mechanosensing (Fig. 1.4). In the following subsections, we discuss laminopathies in the context of disrupted nuclear mechanics and mechanotransduction, particularly in light of the mechanisms discussed in Section 3. Disrupted mechanotransduction as a driver of laminopathy pathology The physical consequences of laminopathies on the structure and function of the nuclear lamina have been known for nearly two decades (Vigouroux et al., 2001; Novelli et al., 2002; Muchir et al., 2004). Mutant or mislocalized proteins can lead to disrupted interactions between lamins and their binding partners, thus disturbing the mechanical integrity of the lamina, connections to chromatin and LINC complex proteins, and transcriptional regulators. LMNA mutations associated with muscular defects frequently result in reduced nuclear stability (Lammerding et al., 2004b; Zwerger et al., 2013; Earle et al., 2019). Furthermore, LMNA mutations increase susceptibility of Lamin A to phosphorylation (Cenni et al., 2005), thereby increasing its solubility and promoting disassembly of the nuclear lamina. LMNA mutant nuclei are often subject to nuclear envelope blebbing and both spontaneous rupture and rupture due to mechanical stress (Lammerding et al., 2006; De Vos et al., 2011; Yang et al., 2018). Nuclear instability and rupture yield reduced cellular viability (Lammerding et al., 2006), loss of cellular compartmentalization that can mislocalize both proteins and whole organelles (Fidziańska et al., 2008; Gupta et al., 2010; De Vos et al., 36 2011), and DNA damage (Cho et al., 2017; Earle et al., 2019). Changes in LINC expression or anchoring at the nuclear envelope due to LMNA mutations, such as by overexpression of Sun1 (Chen et al., 2012), loss of Emerin (Hale et al., 2008a), or loss or mislocalization of Nesprin-2G (Lammerding et al., 2004b; Arsenovic et al., 2016) disrupts mechanotransmission across the nuclear envelope. This impaired nucleo-cytoskeletal coupling (Hale et al., 2008a; Chen et al., 2012; Arsenovic et al., 2016) could explain disturbed nuclear positioning in skeletal muscle (Houben et al., 2009; Folker et al., 2011; Mattioli et al., 2011) and the loss of perinuclear actin filaments in LMNA mutant cells (Khatau et al., 2009; Kim et al., 2017; Shiu et al., 2018), which is associated with increased nuclear height, abnormal nuclear shape, and impaired YAP translocation into the nucleus (Khatau et al., 2009; Shiu et al., 2018). Disruption of YAP translocation in response to cyclic stretch results in poor matrix adhesion and decreased cytoskeletal tension (Bertrand et al., 2014) that may be due to both loss of mechanically induced NPC opening (Elosegui-Artola et al., 2017) and Nesprin-1 disruption (Driscoll et al., 2015). Loss of Lamin A/C results in NPC clustering (Sullivan et al., 1999; Verga et al., 2003; Xie et al., 2016a), and mutations in the Ig fold of Lamin A/C result in defective binding to nucleoporin (Al- Haboubi et al., 2011) which may inhibit the roles of NPCs in mechanosensitive gene regulation. Altered nuclear mechanics and nucleo-cytoskeletal coupling could further disrupt mechanosensitive NPC opening and nuclear import of transcription factors and downstream gene expression (Elosegui-Artola et al., 2017). Moreover, since nucleoporins interact with transcriptionally active euchromatin (Kalverda et al., 2010; Buchwalter et al., 2014), improper distribution of NPCs and nucleoporins resulting from LMNA mutations may perturb 37 transcriptional regulation. Similarly, as lamin sequesters heterochromatin to the nuclear periphery, altered chromosome location due to LMNA mutations could dysregulate chromatin organization and gene expression. Several studies have demonstrated that relevant striated muscle genes are mislocalized to either the nuclear periphery or the center, depending on the mutation (Mewborn et al., 2010; Mattout et al., 2011). Such mislocalization could explain the altered tissue-specific gene expression observed in laminopathies (Mewborn et al., 2010; Mattout et al., 2011; Zuela et al., 2017), a concept that should be further explored using genome mapping technologies (see Section 6). Furthermore, possibly as a downstream effect of disturbed nuclear or cytoplasmic mechano- sensing, several critical signaling pathways regulating differentiation and proliferation are disrupted in LMNA mutant muscle. These include transforming growth factors β1 and 2 (Van Berlo et al., 2005; Bernasconi et al., 2018), MyoD (Frock et al., 2006), MAPK (specifically extracellular signal–regulated kinases 1 and 2, JNK, and p38α) (Muchir et al., 2007, 2012), and WNT/β-catenin (Muchir et al., 2007, 2012; Le Dour et al., 2017), which may compromise tissue homeostasis. Consequently, LMNA mutations can disrupt myogenic differentiation in skeletal muscle (Favreau et al., 2004; Frock et al., 2006; Houben et al., 2009), although other studies found that Lamin A/C– deficient myoblast differentiation into myotubes is normal (Melcon et al., 2006; Earle et al., 2019). Lamin A/C is expressed in both muscle stem cells (MuSCs) and differentiated myofibers. Mutant forms cause improper cell cycle exit, decreased MuSC fusion with myofibers, and increased apoptosis during differentiation (Favreau et al., 2004), as well as slower and less efficient differentiation (Frock et al., 2006). As a possible explanation for increased muscle wasting, DNA- dependent protein kinase (DNA-PK), which was recently linked to aging-related muscle wasting 38 (Park et al., 2017; Chung, 2018), is activated in response to DNA damage (Earle et al., 2019). This activation may drive muscle health decline in EDMD, possibly through apoptosis mediated by the activation of Caspase-3 (Earle et al., 2019). Strategies to remedy cellular pathology Targeting disrupted signaling in LMNA laminopathies may open a window for the pharmaceutical treatment of laminopathies (Azibani et al., 2014). WNT/β-catenin stimulation (Le Dour et al., 2017) and p38α MAPK inhibition (Muchir et al., 2012; National Institute of Health, 2017; Laurini et al., 2018) improve cellular pathology and disease outcomes, including improved cellular mechanical properties, cytoskeletal structure, cardiac contractility, and survival. Targeting impaired nuclear stability may present another therapeutic avenue. Pharmaceutical stabilization of microtubules, which reduces nuclear deformation, and depletion of the microtubule motor kinesin 1, involved in nuclear shuttling in skeletal muscle, prevented accrual of nuclear damage by nuclear envelope rupture and chromatin protrusions in Lamin A/C–deficient skeletal muscle cells in vitro (Earle et al., 2019). Although preliminary, these results demonstrate that reducing mechanical stress on the nucleus can positively influence laminopathic prognosis and represent a new treatment option that should be explored for laminopathies affecting skeletal muscle. Current Technologies for the Study of Mechanotransduction, Nuclear Mechanics, and Related Diseases Our knowledge of mechanotransduction and nuclear mechanics in stem cell biology and disease (laminopathies) is often the product of innovative technologies. From nuclear- to cellular- to tissue-level technologies, creative force application methods, imaging techniques, and model 39 systems have defined the study of nuclear mechanotransduction (Table 1). In this section, we discuss current technological innovations, including superresolution imaging, fluorescence molecular reporters, and engineered tissue constructs for analyzing the role of the nucleus and the corresponding mechanisms in mechanotransduction and disease (laminopathies). Molecular probes for nuclear structure Unraveling nuclear mechanotransduction has remained a challenge due to (a) the complex and interconnected nature of the nuclear constituents and (b) the microscopic scale required to mechanically probe the nuclear components. Thus, several techniques, such as superresolution microscopy, fluorescence reporters for nanoscale forces and deformation, and tools to probe nuclear structure and organization across the whole genome, stand at the forefront of technologies to overcome such obstacles. Superresolution imaging techniques, such as structured illumination microscopy (Shimi et al., 2015), dSTORM (direct stochastic optical reconstruction microscopy) (Xie et al., 2016a), and cryo-ET (cryo–electron tomography) (Turgay et al., 2017), among others, have enabled the observation of the organization of the nuclear lamina and their binding partners at the protein level and have revealed nuclear supramolecular structures and unexpected details of lamin filament organization (Turgay et al., 2017). To further probe protein–protein interactions at the nuclear envelope in living cells and animals, investigators have developed BioID, in which a protein of interest, such as Lamin A, is fused to a promiscuous version of BirA, an Escherichia coli biotin ligase. Proteins in close proximity (∼10 nm) to the protein of interest are biotinylated and can subsequently be identified by mass spectrometry (Roux et al., 2012; Xie et al., 2016a). Newer versions of BioID have been developed to reduce the interaction radius and improved control over the timing of the biotinylation (Schopp et al., 2017). BioID evades the removal of 40 proteins from their native environment or the disruption of native protein interactions, as is the case with common alternative methods of yeast two-hybrid and coimmunoprecipitation assays (Roux et al., 2012). To date, these techniques have been used primarily to interrogate native nuclear protein conformations and Lamin A binding partners (Roux et al., 2012), but they could easily be applied to examine other key protein players and interactions in mechanically stressed or lamin-mutant nuclei in order to better understand mechanotransduction. Characterization of the spatial organization of chromatin over time, termed the 4D nucleome (Dekker et al., 2017), has been a rapidly growing point of focus in cell biology. Genome interaction mapping techniques, evolved from the original 3C (chromosome conformation capture) methods to today's 4C, 5C, Hi-C, and ChIA-PET (chromatin-interaction analysis by paired-end tag sequencing), have created high-resolution interaction maps of chromatin (Dekker et al., 2017) and are beginning to be suitable for single-cell analysis. These techniques are now being applied to laminopathies (McCord et al., 2013), where they can yield novel insights into how transcription may be regulated in response to mechanical force or how chromatin may be disorganized in laminopathies. Fluorescence imaging for nuclear mechanics and mechanotransduction Chromatin reorganization, dynamics, interactions, condensation, and modifications may be better understood through fluorescence imaging techniques. Fluorescence tagging of chromatin using gene editing has been a common method of tracking reorganization of specific gene loci (Ma et al., 2015; Shao et al., 2016; Tajik et al., 2016). Such methods and the use of multiple colors may, for example, allow tracking of several mechanosensitive genes simultaneously in response to 41 mechanical force to better understand mechanosensitive chromosome reorganization. FRAP (fluorescence recovery after photobleaching) experiments, using tagging of histones or chromatin modifications via fluorescently labeled specific antigen binding fragments (Fabs), can examine chromatin dynamics in live cells (Bhattacharya et al., 2009; Hayashi-Takanaka et al., 2011; Harkness et al., 2015; Ramdas and Shivashankar, 2015). These techniques may be particularly useful to understand how mechanical stimulation affects chromatin dynamics, reorganization, and modification (Harkness et al., 2015). As an additional approach, Förster resonance energy transfer (FRET)-based reporters can be used to monitor chromatin modification and condensation (Llères et al., 2009; Sasaki et al., 2009; Ito et al., 2011) in living cells. Recent fluorescence lifetime imaging microscopy experiments have enabled high-throughput spatial tracking of condensation of fluorescently labeled chromatin in the nuclear interior simply by using the viscosity of chromatin and bypassing any gene modification, such as overexpression, required by other techniques (Llères et al., 2009). Chromatin may be labeled either through fluorescently tagged histones (Llères et al., 2009) or the use of DNA-binding dyes (Spagnol and Dahl, 2016). High chromatin condensation is associated with low viscosity and low fluorescence lifetime, while decondensation causes an increase in viscosity due to reduction in chromatin packing and therefore has a high fluorescence lifetime (Spagnol and Dahl, 2016). Thus, epigenetic modifications and changes in nuclear chromatin localization can be spatially and temporally tracked, which is useful for observing changes in response to mechanical stresses, for understanding chromatin changes during stem cell differentiation, and for studying diseases involved with disrupted interactions with chromatin. 42 Table 1.1. Prominent technologies for elucidating mechanotransduction mechanisms. Technique Description C itations Detection techniques Superresolution Imaging techniques (i.e. SIM, dSTORM, cryo-ET) with protein-level (Shimi et al., microscopy resolution. Useful for examining nuclear organization, binding partners, and 2015; Xie et al., supramolecular structure. 2016b; Turgay et al., 2017) BioID Proteins are biotinylated when in proximity to an engineered fusion protein (Xie et al., (such as Lamin A) to label and identify novel binding partners with mass 2016b; Balikov spectrometry. Can be used to examine protein interactions in mechanically- et al., 2017) stressed or lamin-mutant nuclei. 4D nucleome Genome mapping techniques, i.e. 4C, 5C, Hi-C, and ChIA-PET (chromatin- (Roux et al., interaction analysis by paired-end-tag sequencing), for observing spatial 2012) organization and condensation states of chromatin. Genomic Fluorescent tagging of chromatin using gene editing for tracking (Ma et al., labeling mechanosensitive reorganization of (multiple) gene loci. 2015; Shao et al., 2016; Tajik et al., 2016) Fluorescence A target protein is fluorescently tagged, a small area is photobleached, and (Bhattacharya recovery after time of recovery of fluorescence to the area is measured to understand the et al., 2009; photobleaching recovery dynamics, such as for chromatin histone organization or Hayashi- (FRAP) modifications. Takanaka et al., 2011; Harkness et al., 2015; Ramdas and Shivashankar, 2015) Förster Visual monitoring of the interaction between fluorescently-tagged proteins, (Llères et al., resonance energy which creates a FRET signal. Diverse applications to mechanotransduction, 2009; Sasaki et transfer (FRET) such as monitoring force-dependent protein interactions, chromatin al., 2009; modification/condensation, actin assembly or measuring tension forces. Grashoff et al., 2010; Ito et al., 2011; Iyer et al., 2012; Arsenovic et al., 2016) Fluorescence Through fluorescent tagging of chromatin and examining fluorescence (Llères et al., lifetime imaging lifetime, which corresponds to viscosity due to degree of chromatin packing, 2009; Spagnol microscopy can be used for high-throughput spatial tracking of chromatin condensation and Dahl, 2016) (FLIM) in the nucleoplasm. Mechanical manipulation Isolated nuclei Removal of the nucleus from a cell for the direct study of the nucleus and its (Arsenovic et constituents, eliminating any confounding effects from the cytoplasm and/or al., 2016; cytoskeleton. Force can be directly applied to the nucleus, such as for LINC Balikov et al., complex force measurement or examining nuclear changes. 2017) LINC complex Depletion or deletion of LINC complex proteins via gene editing. By (Crisp et al., disruption examining any subsequent defects resulting from force application, the role 2006; Lombardi of LINC complex proteins in mechanotransduction may be better et al., 2011; understood. Tajik et al., 2016) Tissue engineering 43 Engineered Cells are suspended in an extracellular matrix (ECM) solution, compact to (Legant et al., (muscle) tissues form a tissue between two flexible pillars, and tissues contract to deflect the 2009; Sakar et pillars. Useful for examining cell and tissue structures, tissue generated al., 2012; van forces, and improving maturity of tissues. Spreeuwel et al., 2014; Tiburcy et al., 2017; Abilez et al., 2018; Long et al., 2018; Maffioletti et al., 2018; Ronaldson- Bouchard et al., 2018, 2019) Micropatterning, Cells are cultured on micron- or nanometer-scale geometries/architectures. (Li et al., 2011; structured, and Examining the subsequent nuclear changes and cellular signaling, behavior, Jain et al., engineered or phenotype can give an understanding of the role of the nucleus in matrix 2013) substrates sensation, such as in stem cell differentiation. In addition to understanding chromatin dynamics, related imaging techniques can be useful for the study of other mechanotransduction mechanisms. FRET between fluorophores of a single type, known as homoFRET (Vishwasrao et al., 2012), has been used to visualize and quantitatively measure changes in F-actin/G-actin ratios upon force application, based on the homoFRET signal produced when actin molecules labeled with enhanced green fluorescent protein assemble into filaments (Iyer et al., 2012). Furthermore, tension-based FRET molecular biosensors, in which the FRET signal inversely correlates with the force transmitted across the tension-sensor-containing molecule (Arsenovic et al., 2016), enable one to probe mechanotransmission through various LINC complex proteins. This approach has already been successfully applied to measure forces across Nesprin-2G under different mechanical conditions (Arsenovic et al., 2016). FRET biosensors could be further applied to examine the interactions of and force transmission across the cytoskeleton to other nuclear envelope proteins, reorganization or binding of the nuclear lamina to its many binding partners, or mechanically induced changes within the nucleus. Engineered muscle for examining tissue mechanics in laminopathies 44 Current methods for the in vitro study of cardiac and skeletal muscle, particularly in 2D culture, insufficiently recapitulate the native structure and organization of mature muscle cells in tissues. Recently developed engineered skeletal muscle and cardiac tissues more closely mimicking native tissue structure and maturity present an intriguing opportunity for the study of laminopathies and their underlying nuclear and tissue mechanics, and offer better platforms for testing pharmacological treatments compared with 2D culture systems. Consequently, engineered muscle constructs, ranging from the micrometer to the centimeter scale, enable (a) examining cell and tissue structures, (b) examining tissue-generated forces, and (c) improving the maturity of tissues. To form tissues, muscle or heart cells or progenitors, either alone or in direct coculture with fibroblasts or other cell types (Maffioletti et al., 2018), are suspended in an ECM solution (Fig. 1.5a). Cells remodel and compact the ECM to form a tissue-like structure between two flexible pillars, which apply passive tension across the tissue that results in cytoskeletal and sarcomere alignment (Fig. 1.5b) (van Spreeuwel et al., 2014; Ronaldson-Bouchard et al., 2019). The engineered muscle tissues further compact over time and begin to contract as the muscle cells mature. The deflection of the flexible pillars (Fig. 1.5c) can used to measure the tissue-generated forces (Legant et al., 2009). Optionally, mechanical and/or electrical stimulation can be applied to engineered tissues to further improve maturation (Abilez et al., 2018; Ronaldson-Bouchard et al., 2018). To date, engineered muscle has also been employed to study cardiac muscle maturation (Tiburcy et al., 2017; Abilez et al., 2018; Ronaldson-Bouchard et al., 2018, 2019), examine cellular forces and anisotropy (Legant et al., 2009; van Spreeuwel et al., 2014), analyze generated contraction forces (Legant et al., 2009; Sakar et al., 2012; Long et al., 2018), and create disease models for 45 examining cellular phenotype, such as Duchenne's muscular dystrophy or EDMD (Maffioletti et al., 2018). Engineered muscle tissues can be useful for assessing tissue structure or nuclear morphologies for various disease-causing mutations (Maffioletti et al., 2018), assessing disruption of tissue-generated forces (Long et al., 2018), and modeling correction of disease-causing mutations (Long et al., 2018). However, tissue and sarcomere maturity still do not fully recapitulate native tissue, particularly for stem cell–derived muscle, motivating further research. Fig. 1.5. Creation of engineered muscle tissue constructs for the study of tissue morphology and generated forces. (a) Devices are loaded with a suspension of cells (brown spheres) and extracellular matrix (red fibers), and (b) cells reorganize and restructure the matrix to form a tissue around elastic pillars. (c) Tissues gradually compact and/or contract as cells elongate, thereby deflecting pillars. The force generated by the engineered tissue constructs can be calculated from the measured pillar deflection and the known material properties of the elastic pillars. Conclusions and Perspectives Over the past few decades, efforts to obtain a clearer picture of nuclear mechanotransduction have shed light on how the cellular microenvironment and mechanical force guide cellular behavior and phenotype, stem cell differentiation, and human diseases such as laminopathies. Mechanotransmission through perinuclear cytoskeletal assemblies and the LINC complex to the lamina and chromatin governs nuclear mechanical response to force and alters organization of chromatin and gene expression as well as downstream expression of LINC proteins. 46 Mechanosensitive phosphorylation and protein conformation modulate nuclear strength by altering the organization of the nuclear envelope. Nuclear membrane stretch guides downstream mechanosignaling by stretching of NPCs for increased nuclear import of transcription factors and by allowing for mechanosensitive incorporation of proteins into the INM. Finally, chromatin organization, compaction, stretching, and modification control downstream mechanosensitive gene expression, although the specific guiding mechanisms should be further explored. Together, these nuclear mechanotransduction mechanisms guide mechanically induced stem cell differentiation and can instill mechanical memory of differentiation states. Disruption of any component or mechanism, such as in LMNA laminopathies, may induce a chain reaction of disrupted nuclear nucleo-cytoskeletal coupling, altered nuclear mechanics, and defective mechanotransduction elements and downstream mechanosignaling to cause human disease. An improved understanding of defective mechanotransmission and mechanotransduction signaling that enables targeting affected pathways and components may ultimately allow these pathologies to be remedied. Future research should aim to gain a more systematic understanding of the cascade of nuclear mechanotransduction events. In particular, which nuclear mechanisms are a direct response to mechanical force (i.e., true mechanosensors), and which are a product of downstream signaling? Elucidation of the positive and negative feedback loops driving nuclear mechanotransduction would clarify how the many individual mechanisms relate and work together to guide downstream cellular phenotype and function and would shed new light on the nucleus as a mechanosensor. 47 CHAPTER 2 Impaired lamin localization to the nuclear envelope is responsible for nuclear damage in LMNA mutant iPSC-derived cardiomyocytes2 The LMNA gene encodes the nuclear envelope proteins Lamins A and C, which comprise a major part of the nuclear lamina, provide mechanical support to the nucleus, and participate in diverse intracellular signaling. LMNA mutations give rise to a collection of diseases called laminopathies, including dilated cardiomyopathy (LMNA-DCM) and muscular dystrophies. Although nuclear deformities are a hallmark of LMNA-DCM, the role of nuclear abnormalities in the pathogenesis of LMNA-DCM remains incompletely understood. Using induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) from LMNA mutant patients and healthy controls, we show that LMNA mutant iPSC-CM nuclei have altered shape or increased size compared to healthy control iPSC-CM nuclei. The LMNA mutation exhibiting the most severe nuclear deformities, R249Q, additionally caused reduced nuclear stiffness and increased nuclear fragility. Importantly, for all cell lines, the degree of nuclear abnormalities corresponded to the degree of Lamin A/C and Lamin B1 mislocalization from the nuclear envelope. The mislocalization was likely due to altered assembly of Lamin A/C. Collectively, these results point to the importance of correct lamin 2 This work was submitted to the Molecular Biology of the Cell special issue on Quantitative Cell Biology. Maurer, Melanie, Perati, Shriya, Johnson, Lindsey, Gacita, Anthony M, Lai, Shuping, Wallrath, Lori, Benjamin, Ivor, McNally, Elizabeth, Kirby, Tyler, Lammerding, Jan. MM, TK, JL, AGM, SL, LW, IB, and LW contributed to the conception and design of the work. MM, TK, SP, and LJ contributed to data acquisition and analysis. MM, TK, SP, LJ, and JL contributed to interpretation of the data. MM and JL contributed to the drafting of the manuscript. All authors contributed to editing the manuscript. 48 assembly at the nuclear envelope in providing mechanical stability to the nucleus and illustrate that defects in nuclear lamina organization can contribute to the nuclear and cellular dysfunction in LMNA-DCM. Introduction Lamin A/C are intermediate filaments that assemble to form the nuclear lamina, a dense protein meshwork that resides underneath the inner nuclear membrane. The nuclear lamina provides structural support to the nucleus and functions in diverse mechanical and biochemical signaling processes (Davidson and Lammerding, 2014; Kirby and Lammerding, 2018; Maurer and Lammerding, 2019; Donnaloja et al., 2020). More than 450 mutations have been identified in the LMNA gene that give rise to a collection of about 15 different human diseases, termed ‘laminopathies,’ which include dilated cardiomyopathy (LMNA-DCM), Emery-Dreifuss muscular dystrophy (EDMD), congenital muscular dystrophy (CMD), and Hutchinson-Gilford progeria syndrome (HGPS) (Davidson and Lammerding, 2014). LMNA-DCM has a high mortality rate, and compared to other forms of congenital DCM, LMNA-DCM has a particularly poor prognosis with early onset, a high occurrence of arrhythmias and up to 19% of all patients requiring heart transplants (Taylor et al., 2003; McNally and Mestroni, 2017; Hasselberg et al., 2018). To date, clinical treatments for LMNA-DCM have focused on slowing heart failure rather than targeting the underlying cellular pathology, largely because the molecular disease mechanisms responsible for LMNA-DCM are not yet fully understood. LMNA mutations and deletions often cause a reduction in nuclear stiffness that results in changes to nuclear shape, nuclear envelope (NE) rupture, and altered response to mechanical stress (Broers 49 et al., 2004; Lammerding et al., 2004a; Nikolova et al., 2004; Chandar et al., 2010; De Vos et al., 2011; Zwerger et al., 2013; Earle et al., 2019; Bertrand et al., 2020). However, few studies have been completed to understand nuclear mechanics and nuclear deformation in LMNA mutant human cardiomyocytes. Until the recent advent of human induced pluripotent stem cell (iPSC) models, the majority of studies on the cellular disease pathogenesis of LMNA-DCM have relied on mouse models that do not fully reflect all aspects of the human disease. For example, mutant mouse phenotypes require homozygous mutant alleles rather than dominant heterozygous allelic state in human disease (Stewart et al., 2007). As such, limited previous studies have demonstrated nuclear shape changes and NE ruptures resulting in the mislocalization of cellular components into the nucleus in cardiomyocytes of laminopathy models (Nikolova et al., 2004; Chandar et al., 2010; Cho et al., 2019; Shah et al., 2019). Despite these efforts, the cardiac-specific mechanisms through which nuclei become damaged, the consequences of nuclear damage, and the effect of specific LMNA mutations in the context of nuclear damage in cardiomyocytes remain largely unclear. Since the contractile behavior of cardiomyocytes can influence nuclear mechanical properties through lamin conformation (Swift et al., 2013; Buxboim et al., 2014; Heffler et al., 2019; Piccus and Brayson, 2020), it is critical to determine the cardiomyocyte-specific role of nuclear mechanics and damage in LMNA-DCM and how they contribute to disease progression. Here, we use three LMNA mutant patient-derived human iPSC lines, iPSCs from healthy controls, and iPSC-derived cardiomyocytes (iPSC-CMs) to provide a quantitative analysis of the causes and consequences of nuclear damage in LMNA-DCM, with a particular focus on the nuclear lamina organization and physical properties of the cell nucleus. We show that different LMNA mutant iPSC-CMs exhibit varying degrees of nuclear shape and size deformities, with only cells with the 50 most severe mutation, LMNA R249Q, exhibiting NE rupture. Furthermore, we demonstrate that the severity of nuclear deformities correlates with reduced Lamin A/C and Lamin B1 localization to the NE, likely due to defective lamin assembly. Assembly of Lamin A/C at the NE is critical for maintaining nuclear mechanical properties and NE integrity (Zwerger et al., 2015). As such, defective assembly of lamins at the NE may cause reduced nuclear stability and/or defects in organization of NE proteins that in turn may result in altered cytoskeletal organization, abnormal nuclear shape, and susceptibility of the nucleus to damage. Collectively, these findings suggest that mislocalization of Lamin A/C and Lamin B1 from the NE can explain much of the nuclear deformities and damage observed in three LMNA-DCM iPSC-CM lines. Methods and Materials Induced Pluripotent Stem Cell (iPSC) culture and cardiac differentiation Healthy control (WT) and patient-derived LMNA mutant iPSCs were either purchased from commercial sources (Cure CMD) or generously provided from other researchers (Fig. 2.1A; Table 2.1). iPSCs were cultured feeder-free in TeSR-E8 media (#05990, StemCell Technologies) on 1:30 Matrigel Basement Membrane (Matrigel; #47743-722, Corning) diluted in RPMI 1640 medium (#11875093, Gibco). Media was changed daily, and colonies were monitored for and removed with signs of spontaneous differentiation. Cardiac differentiation was performed based on a previously established protocol (Sharma et al., 2015). iPSCs were washed with RMPI 1640 medium, dissociated into single cells with TrypLE Express (#12605028, Gibco), and seeded on 1:30 Matrigel in 12-well plates in TeSR-E8 media supplemented with Y-27632 (10 μM, #10005583, Cayman Chemical). Media was changed to fresh 51 TeSR-E8 after 24 hours. After 48 hours (D0), differentiation was induced using inhibition of the WNT pathway with CHIR99021 (6 μM or 12 μM, #SML1046, Sigma-Aldrich) in RPMI 1640 media supplemented with B27 minus insulin (#A1895602, Gibco; referred to here as B27-INS). The CHIR99021 concentration, at 6 μM or 12 μM, was optimized for each cell line as each had different sensitivities to the inhibitor. Media was changed to fresh B27-INS at day 1 of differentiation (D1). IWR-1 (5 µM, #13659, Cayman Chemical) in B27-INS was added at D3 and changed to fresh B27-INS at D5. Media was changed to RPMI 1640 medium supplemented with B27 (#17504044, Gibco; referred to as B27) on D7 and was changed daily for three days until beating was induced. On D10, media was changed to RPMI 1640 without L-Glutamine (#21870076, Gibco; referred to as selection medium) supplemented with B27 and Sodium Lactate (Sigma) for metabolic selection of cardiomyocytes. On D12, media was changed to B27 supplemented with Penicillin-Streptomycin (referred to as B27+P/S), and iPSC-CMs were allowed to recover for two days before passaging (‘first CM passage’). At D14, only wells of iPSC-CMs that were beating were washed with RPMI 1640 and dissociated into single cells using Trypsin (0.25%). Trypsin was inactivated using RPMI 1640 plus 20% fetal bovine serum. Cells were centrifuged and resuspended in RPMI 1640 plus fetal bovine serum (10%) and 10 μM Y-27632 and then seeded 1:1 in 12-well plates coated with Fibronectin (4 μM). After 24 hours, media was changed to B27+P/S. After an additional 24 hours, iPSC-CMs were subjected to a second round of metabolic selection with selection medium for 48 hours. Media was then changed back to B27+P/S, and cells were used seven days after the first CM passage (D21). After 21 days of differentiation, all cell lines spontaneously contracted throughout the wells. We 52 confirmed cardiac differentiation by immunofluorescence staining for a cardiac marker, cardiac troponin T (cTnT; Fig. 1B). We quantified the percentage of cells that expressed cTnT, which demonstrated that all cell lines routinely had over 85% expression of cTnT and no significant changes in expression between cell lines (Fig. 1C). All immunofluorescence experiments were performed with a cTnT co-stain. Upon immunofluorescence labeling, only wells with predominantly cTnT-expressing cells were used for analysis (wells with few cTnT-expressing cells were excluded), and any remaining individual cells not expressing cTnT were excluded from analysis. Generation of fluorescently labeled cell lines iPSC-CMs after the first CM passage were stably modified with lentiviral constructs to express a NE rupture reporter consisting of a green fluorescent protein with a nuclear localization sequence (NLS-GFP, full vector: pCDH-CMV-NLS-copGFP-EF1-blastiS) described previously (Denais et al., 2016). Cells were cultured for at least six days after modification to allow sufficient expression of the construct. Expression of NLS-GFP was confirmed before each experiment. Lamin A/C depletion Prior to the first round of metabolic selection, WT2-iPSC-CMs were stably modified with a lentivirus to express either shRNA targeting LMNA (shLMNA; pLKO.1-shRNA-mLMNA; target sequence: CCGGGAAGCAACTTCAGGATGAGATCTCGAGATCTCATCCTGAAGTTGCTTCTTTTT G) or a non-target control (shNT). For all experiments, shRNA-modified iPSC-CMs were used as a mixed population of modified (‘knockdown’ or ‘KD’) and un-modified (‘no knockdown’ or ‘no 53 KD’) cells to serve as an additional internal control group, as described in the Image Analysis section. Long-term imaging experiments Long-term fluorescence imaging of iPSC-CMs expressing NLS-GFP imaging was performed on an IncuCyte (Sartorius) incubator imaging system. Images were acquired at 10× magnification every 15 minutes for three days, starting one week after the first CM passage. Images were exported and manually analyzed in FIJI software to quantify the frequency and duration of NE rupture, as evidenced by the transient loss of NLS-GFP from the nucleus. Immunofluorescence staining iPSC-CMs were passaged one week after the first CM passage into optically clear 96-well plates. The next day, media was changed to B27+P/S for several hours before cells were washed with 1× PBS and fixed in warm Paraformaldehyde (4%) for 10 minutes. Cells were then washed with 1× PBS and blocked in 3% BSA with 0.1% Triton-X 100 (Thermo-Fisher) and 0.1% Tween (Sigma) in PBS for one hour at room temperature. Primary antibodies (Table S2.1) were prepared in blocking solution and incubated overnight at 4°C. iPSC-CMs were then washed with a solution of 0.3% BSA with 0.1% Triton-X 100 and 0.1% Tween in PBS and stained with AlexaFluor secondary antibodies (1:250, Invitrogen) for 1 hour at room temperature. DAPI (1:1000, Sigma) was added for 15 minutes at room temperature, and cells were washed with PBS before imaging. Nuclear semi-permeabilization and soluble protein washout One week after the first CM passage, iPSC-CMs were passaged into optically clear 96-well plates 54 (Greiner). The next day, media was changed to B27+P/S for several hours before cells were washed with 1× PBS. A buffer, “CSKT,” containing NaCl (100 mM), Sucrose (300 mM), MgCl2 (3 mM), PIPES (10 mM, pH 6.8) and Triton-X 100 (0.5%) was added to cells to semi-permeabilize nuclei and incubated on ice for 1 minute. A 1-minute wash with 1× PBS was used as a control in parallel to semi-permeabilization with CSKT. CSKT buffer was removed, cells were gently washed with 1× PBS to washout soluble nuclear proteins, and then the cells were fixed with 4% PFA for 10 minutes. Following PFA fixation, cells were inspected under a microscope to ensure that nuclear semi-permeabilization and washout was successful, as evidenced by the darkened and more prominent appearance of nuclei. Image acquisition Plates were imaged on an inverted Zeiss LSM700 confocal microscope. Z-stacks were taken using a 40× water immersion (1.2 NA) objective with Airy units set to 1.0 for all images. Z-stacks for nuclear volume were acquired using the optimum step size for the 40× water immersion objective, i.e., 0.35 μm. All other image stacks were acquired using a step size of 1.5 μm and analyzed using Maximum Intensity Projections (MIPs). Image analysis All image analysis was conducted by observers blinded for genotype and treatment conditions. iPSC-CM purity was quantified from MIPs by counting the proportion of nuclei in the image that are located within cardiac troponin T (cTnT) positive cells (referred to as cTnT-positive nuclei). For image analysis of iPSC-CMs, only cTnT-positive nuclei were used. 55 For nuclear shape and area, a custom MATLAB (MathWorks) script was used to segment nuclei based on the Lamin B1 immunofluorescence signal in MIPs. Nuclei intersecting with the image edges were removed, followed by manual selection of nuclei in each image to exclude any nuclei touching each other, dead cells, or non-cTnT-positive nuclei. The area (A) and perimeter (P) were then measured for each selected nucleus, and the circularity index (CI) was computed based off 4𝜋𝐴 the formula 𝐶𝐼 = . A perfectly circular object has a circularity index of 1, and the circularity 𝑃2 index decreases as circularity decrease, e.g., in oval or irregularly shaped nuclei. Nuclear volume was computed from high resolution confocal Lamin B1 immunofluorescence 3D image stacks. Image stacks were read into the FIJI image analysis program, and a threshold value was determined from the middle slice of z-stacks. The threshold value was then input into the 3D Simple Segmentation suite (Ollion et al., 2013), and the 3D Objects Counter (Bolte and Cordelières, 2006) was used to compute nuclear volumes. Partial nuclei at the image edges, nuclei touching each other, dead cells, and non-cTnT-positive nuclei were excluded from analysis. For shRNA experiments, nuclear area, circularity index, and Lamin A/C mean fluorescence intensity were computed from a modified version of the MATLAB script for nuclear area and circularity described above. Nuclear area and circularity index were computed from the Lamin B1 immunofluorescence signal of MIPs as described above. Additionally, Lamin A/C images were read into the program, images were cross-correlated with the corresponding Lamin B1 images to analyze only the user-selected nuclei, and the mean Lamin A/C immunofluorescence intensity inside of the area of each user-selected nucleus was computed. For each individual experiment, Lamin A/C intensities for shNT- and shLMNA-treated cells were plotted, and the lowest Lamin 56 A/C intensity from the shNT group was used as a threshold, above which shLMNA nuclei were considered “no knockdown (no KD)” and below which shLMNA nuclei were considered “knock down (KD).” shLMNA nuclei were separated into “no KD” or “KD” groups for each experiment, with “no KD” nuclei serving as a secondary, internal control. Nuclear area and circularity index were analyzed based on the average of several separately thresholded experiments. Nuclear volume analysis of shRNA Lamin A/C depleted cells and controls was performed using the same FIJI image analysis pipeline described above. For these experiments, an observer blinded for treatment conditions manually classified nuclei as “no KD” or “KD” for each image based on their Lamin A/C immunofluorescence intensity. Untreated control or shNT groups, as expected, had extremely few nuclei considered as “KD” and the proportion of shLMNA no KD and KD nuclei was similar to that obtained by the objective thresholding method. Fluorescence intensity profiles were computed from high-resolution confocal z-stacks in the ZEN software (Zeiss). For each nucleus, the z-position for which the x-y plane dissects the center of the nucleus was identified, and a line was drawn across the major axis of the nucleus to obtain the Lamin A/C and Lamin B1 fluorescence intensity values along with distances. Fluorescence intensity profiles were trimmed such that the first values in the profile correspond to the outer edges of the NE. Distances across each nucleus were then normalized to a scale of 0 to 1 to account for differences in nuclear size. Fluorescence intensity profiles and normalized distances for each nucleus for each cell line were then read into a custom R script for analysis (script available upon request). Fluorescence intensity profiles across all nuclei in a cell line were averaged and normalized to the area under the curve to account for variations in fluorescence intensities between 57 cell lines due to varying lamin levels or staining conditions. The boundary of the nuclear lamina was determined by the inflection point of the average fluorescence intensity profile of all cell lines with the nuclear lamina defined as the peripheral regions with normalized nuclear distance of <0.16 and >0.84. For each nucleus, the fluorescence intensity of nucleoplasmic lamins was then determined by taking the 95th percentile of fluorescence intensity values in the nuclear lamina region (to avoid influence of spurious pixel intensity values), and the fluorescence intensity of nucleoplasmic lamins was determined by taking the average value of fluorescence intensities in the nucleoplasm, corresponding to normalized nuclear distances between 0.16 and 0.84. The ratio of nucleoplasmic to peripheral lamins was then computed for each nucleus. Lamin A/C and phospho-Lamin A/C fluorescence intensities were computed using the same MATLAB script used for shRNA analysis described above with Lamin A/C images being used for thresholding and user selection under the same exclusion criteria as previously selected. All Lamin A/C and phospho-Lamin A/C average fluorescence intensity values were normalized to the respective average of a healthy control (WT1) fluorescence intensity from images taken at the same time. Nuclear stiffness measurements iPSC-CMs were passaged one week after the first CM passage into glass-bottom 35-mm tissue culture plates (FluoroDish; World Precision Instruments) coated with 4 µM Fibronectin and allowed to adhere for 24 hours. Prior to experimentation, nuclei were fluorescently labeled with Hoechst 33342 (1:1000; Biotium). Experiments were performed at room temperature, and all indentations were performed within one hour of removing cells from the incubator, at which point 58 the majority of cells were still beating. Nanoindentation was performed with a microscope-mounted Chiaro system (Optics11). The system was calibrated prior to experimentation according to the manufacturer’s instructions, and a spherical glass tip with a radius of 3 µm and stiffness of 0.026 N/m was used to indent cells. The center of the probe tip and the center of nuclei were aligned, and an image of each fluorescent nucleus was taken prior to indentation. The probe was then lowered to approximately 5-10 µm above the top of the nucleus. If the probe accidentally touched a nucleus prior to the start of the indentation, the cell was excluded. The probe was lowered towards the nucleus at a speed of 2 µm/s, and once the system reached a contact force threshold of 5 nN, indentation continued at a rate of 10 nN/s until the probe reached a force threshold of 40 nN. The 40 nN load was held for five seconds, and then the probe was retracted at a rate of 10 nN/s. A Hertzian model with a Poisson’s Ratio of 0.5 (Guz et al., 2014) was fitted to the load vs. indentation curve from 250- 1500nm of indentation depth for each nucleus to determine the Young’s elastic modulus. Any nucleus with a fit for the Hertzian model below an R2 value of 0.95 was excluded from further analysis. Western analysis iPSC-CMs lysates were collected one week after the initial passage. Cells were lysed in High Salt RIPA buffer with protease (cOmplete EDTA-Free, Roche) and phosphatase inhibitors (PhosSTOP, Roche). After lysis, samples were vortexed for five minutes, sonicated for 30 seconds at 36% amplitude, and boiled for five minutes at 93°C. Protein content was quantified with Bio-Rad Protein Assay Dye, and 20 µg of protein was run on a 4-12% Bis-Tris polyacrylamide gel using 59 an SDS-PAGE protocol. Protein was transferred to a polyvinylidene fluoride (PVDF) membrane using a semi-dry transfer at 16 V for one hour. Membranes were blocked with Intercept (PBS) Blocking Buffer (LI-COR) for one hour at room temperature, primary antibodies were diluted in the same blocking buffer, and membranes were incubated the primary antibodies on a rocker overnight at 4°C. Loading control primary antibodies were diluted in the same blocking buffer and incubated for one hour at room temperature the next day. IRDye 680LT or IRDye 800CW (LI- COR) secondary antibodies were used to detect protein bands. Membranes were imaged on an Odyssey® CLx imaging system (LI-COR) and analyzed using Image Studio Lite (LI-COR). Gene expression analysis RNA was extracted from iPSCs before passaging when colonies were large but not touching and iPSC-CMs one week after their initial passage using the RNeasy Plus Kit (Qiagen). Quality of the RNA was assessed using a Fragment Analyzer. Libraries were prepared by the Cornell Institute of Biotechnology using a TruSeq DNA library prep kit (Illumina) and sequenced on a NextSeq 500/550 (Illumina). Raw sequencing reads were quality checked using FastQC (Andrew, 2010). STAR (Dobin et al., 2013) was used to assemble the human genome and align reads to the human genome. Aligned reads were quality control checked, and samples with alignment <70% were removed. A count matrix was generated using featureCounts (Liao et al., 2014) which was then input to build a statistical model in DESeq2 (Love et al., 2014). Normalized gene counts were then extracted for each sample for LMNA and LMNB1. Code availability Custom MATLAB scripts utilized here are available upon request. 60 Statistics All results were taken from a minimum of three independent experiments. For continuous numeric datasets in which individual nuclei were analyzed, a mixed-effects linear regression was performed in R to account for the effects of both LMNA-mutations and individual cell lines. Datasets were tested for normality, and data not following a normal distribution were linearized either by taking the natural log or the square root, whichever achieved better normalization. Pairwise t-tests were then performed using the Tukey method to correct for multiple comparisons. For continuous numeric datasets in which whole images or experiments were analyzed and followed a normal distribution, either a Student’s t-test (for two groups, two-tailed) or one-way analysis of variance (ANOVA) with multiple comparisons (for more than two groups) was performed in Prism (GraphPad). Dunnett correction was used for multiple comparisons. For NE rupture experiments in which proportional data was analyzed, a binomial regression was performed in R. Since several cell lines exhibited no NE rupture, a Firth logistic regression was then performed to reduce bias from such zero values. Pairwise t-tests were then performed using the Tukey method to correct for multiple comparisons. For correlation data, a linear regression was performed for each cell line in Graphpad Prism, and the correlation was tested for significance for a non-zero slope using Pearson correlation and compared to the other regression. Unless otherwise noted, error bars in graphs represent mean ± standard error of the mean (SEM). Results LMNA-mutant iPSCs differentiate efficiently into cardiomyocytes To quantitatively study the effects of different LMNA mutations on nuclear lamina organization and nuclear mechanics, we selected three LMNA mutant iPSC lines (LMNA L35P, R249Q, and 61 G449V) derived from patients with DCM and CMD and two healthy controls (WT1, WT2) for this study (Table 2.1). Each of the LMNA mutations cause amino acid substitutions in a different domain of the Lamin A/C protein (Fig. 2.1A), enabling the identification of common and mutation- specific effects. The iPSCs were subjected to a chemically defined cardiac differentiation protocol to obtain iPSC-derived cardiomyocytes (iPSC-CMs). All iPSCs underwent successful cardiac differentiation with over 85% of nuclei staining positive for cardiac troponin T (cTnT), an established cardiomyocyte marker (Fig. 2.1B, C). We did not detect any significant differences in iPSC-CM differentiation efficiency between cell lines (Fig. 2.1C). Table 2.1. Overview of LMNA-mutant iPSC-CM lines. The iPSC lines were isolated from patients carrying LMNA mutations that produce amino acid substitutions in different domains of the Lamin A/C protein, as indicated in the table, and cause either Congenital Muscular Dystrophy (CMD) or LMNA-DCM. The CMD patients with the L35P and G449V amino acid substitutions are expected to develop a cardiac phenotype later in life, in addition to the skeletal muscular dystrophy present at diagnosis. *The G449V patient presented with muscular dystrophy in childhood, but patient is too young to necessarily develop cardiac phenotype. Protein cDNA Exon Protein Mutation Affected Tissue Cell Source Domain Origin p.L35P c.T104C 1 Coil 1A Unknown Skeletal and Cure CMD and cardiac muscle Cellular Dynamics International Inc. p.R249Q c.G746A 1 Coil 2 De novo Skeletal and Elizabeth McNally cardiac muscle p.G449V c.G1346T 7 Ig-fold De novo Skeletal muscle* Ivor Benjamin The iPSCs were subjected to a chemically defined cardiac differentiation protocol to obtain iPSC- derived cardiomyocytes (iPSC-CMs). All iPSCs displayed successful cardiac differentiation, with over 85% of nuclei staining positive for a cardiac troponin T (cTnT), an established cardiomyocyte marker (Fig. 2.1B, C). We did not detect any significant differences in iPSC-CM differentiation efficiency between cell lines. 62 Fig. 2.1. LMNA mutant iPSCs differentiate into cardiomyocytes with high efficiency. (A) Diagram of Lamin A protein domains with the locations of the amino acid substitutions used in this study indicated, which map to separate domains of the Lamin A protein. Note that these domains are also present the Lamin C protein, which has an alternative tail domain compared to Lamin A. (B) Immunofluorescence images of iPSC-CMs show expression of the cardiac marker Cardiac Troponin T (cTnT) in the vast majority of cells. Scale bar = 50 μm. (C) Quantification of iPSC-CM purity based on cTnT labeling shows that all cell lines differentiate into cardiac myocytes with >85% efficiency, with no statistically significant differences between the cell lines. Data represented as mean ± SEM. N = 8-14 images per group. LMNA mutant iPSC-CMs express normal levels of Lamin A/C Although most LMNA mutations result in stable expression of the mutant protein, some LMNA mutations can reduce protein stability and lead to haploinsufficiency (Wolf et al., 2008; Siu et al., 2012; Cattin et al., 2013). To test whether the LMNA mutant cell lines had altered expression of Lamin A/C, we investigated protein and mRNA expression levels of Lamin A/C and Lamin B1 in the panel of iPSCs and the corresponding iPSC-CMs. We detected only very low Lamin A/C protein levels in iPSCs by both western analysis (Fig. 2.2A) and immunofluorescence labeling (Fig. S2.1A), consistent with previous reports that iPSCs express no or only little Lamin A/C (Zuo et al., 2012). As expected, Lamin A/C expression increased significantly with differentiation into iPSC-CMs (Fig. 2.2A, Fig. S2.2B). Differences in Lamin A/C expression were not significant between the various iPSC-CMs, although the R249Q mutant iPSC-CMs showed a trend towards 63 reduced Lamin A/C protein expression (p > 0.15 vs. both healthy control lines) (Fig. 2.2A-B, Fig. S2.2C-D). Lamin B1 expression did not differ significantly between iPSC-CMs cell lines (Fig. 2.2A-B), although the R249Q iPSCs had a trend towards increased Lamin B1 mRNA levels compared to (p > 0.14 vs. both healthy control lines) (Fig. S2.2B). Fig. 2.2. R249Q iPSC-CMs trend toward reduced Lamin A/C protein expression. (A) Representative western analysis, showing that iPSCs do not express detectable Lamin A/C, and that Lamin A/C expression increases in iPSC-CMs. (B) Quantification of lamin protein levels from western analysis of iPSC-CM samples, showing that G449V iPSC-CMs have normal levels of Lamin A/C, whereas R249Q iPSC-CMs have a trend towards decreased Lamin A/C protein levels that does not reach statistical significance (p > 0.15 vs both WT lines). Data presented as mean ± SEM. N = 5 protein lysates per group. LMNA-mutant iPSC-CMs exhibit altered nuclear shape and size Although altered nuclear shape and size are hallmarks of skeletal muscle and cardiac laminopathies, the severity of nuclear shape defects varies substantially across different LMNA mutations (Sullivan et al., 1999; Lammerding et al., 2004a, 2006; Muchir et al., 2004; Zwerger et al., 2013; Steele-Stallard et al., 2018; Bertrand et al., 2020). To quantify defects in nuclear morphology in the different LMNA mutant iPSC-CMs, we computed nuclear circularity index, area, and volume based on confocal 3D image stacks and cross-sections of iPSC-CMs immunofluorescently labeled for Lamin B1 (Fig. 2.3A). The circularity index has a maximum 64 value of one for perfectly circular nuclei and decreases in value for abnormally shaped nuclei (Fig. 2.3B). R249Q mutant iPSC-CMs had a significantly decreased circularity index compared to healthy controls and other LMNA mutants (Fig. 2.3C), indicating more abnormal nuclear shapes. In contrast, G449V mutant iPSC-CM nuclei exhibited no change in circularity index compared to healthy controls, and L35P mutant nuclei were slightly more circular compared to healthy controls and the other mutant iPSC-CMs (Fig. 2.3C). Notably, all three LMNA mutant iPSC-CM lines had increased nuclear cross-sectional areas compared to healthy controls (Fig. 2.3D), although only Fig. 2.3. LMNA mutant iPSC-CMs have nuclear shape and size defects. (A) Representative immunofluorescence images of iPSC-CMs stained for Lamin A/C (green) and cardiac troponin T (cTnT, gray). Scalebar = 50 µm. (B) Examples of WT2 and R249Q nuclei with corresponding circularity index. (C) R249Q nuclei have significantly decreased circularity index values compared to the WT cell lines, while L35P and G449V nuclei have an increase or no change in circularity index, respectively. (D) All LMNA-mutant iPSC-CMs exhibit an increase in nuclear area, but (E) only L35P and R249Q cells have decreased nuclear height. (F) L35P and R249Q iPSC-CMs have increased nuclear volume. Data presented as mean ± SEM. *, p < 0.05 vs. WT, ***, p < 0.001 vs. WT, ****, p < 0.0001 vs. WT. N ≥ 83 nuclei per group for all experiments. 65 R249Q and L35P mutant iPSC-CMs showed reduction in nuclear height (Fig. 2.3E) and increased nuclear volumes (Fig. 2.3F) compared to healthy controls. These data suggest more severe nuclear defects in L35P and R249Q mutant iPSC-CMs with increased nuclear area and volume, potentially pointing to changes in nuclear organization and/or nuclear stiffness, as decreased nuclear height and increased nuclear cross-sectional area can result in increased nuclear flattening under cytoskeletal tension. R249Q-iPSC-CMs exhibit increased nuclear envelope rupture In addition to changes to nuclear morphology, LMNA mutations can lead to increased incidence of nuclear envelope (NE) rupture due to weakening of the NE (De Vos et al., 2011; Cho et al., 2019; Earle et al., 2019). Therefore, we modified iPSC-CMs with a NE rupture reporter, NLS-GFP, in which a green fluorescent protein (GFP) is fused to a nuclear localization sequence (NLS) (Denais et al., 2016). NLS-GFP normally localizes to the nucleus but leaks into the cytoplasm upon NE rupture and then gradually translocates back into the nucleus following NE repair (Fig. 2.4A-B). Long-term (72 hours) time-lapse microscopy revealed that NE rupture is extremely rare in healthy control iPSC-CMs (Fig. 2.4C). In contrast, R249Q mutant iPSC-CMs, which had the most severe defects in nuclear shape, size, and volume (Fig. 2.3C-E), had a significant increase of NE ruptures compared to both healthy control cell lines (Fig. 2.4C). On the other hand, the L35P and G449V mutant iPSC-CMs, which had milder defects in nuclear shape and size, showed no increase in NE rupture or only a trend towards increased NE rupture (G449V) that did not reach statistical significance (p > 0.3 vs healthy controls). Of note, the duration of NE rupture was not statistically significant between any of the cell lines that exhibited NE rupture (Fig. S2.3), suggesting that LMNA mutations did not alter NE repair, consistent with previous studies that found that Lamin 66 A/C depletion did not alter NE rupture duration (Denais et al., 2016; Halfmann et al., 2019). The dramatic increase of NE rupture in the R249Q mutant iPSC-CMs compared to other LMNA mutant cell lines, paired with the R249Q iPSC-CMs being the only cell line to have a reduction in circularity index, suggests that R249Q mutant iPSC-CMs have more dramatic changes in nuclear stability, potentially due to changes in nuclear mechanics, that leave the nucleus more susceptible to damage. R249Q-iPSC-CMs have reduced nuclear stiffness Lamin A/C are a major determinant of nuclear stiffness (Lammerding et al., 2004a, 2006; Swift et al., 2013; Stephens et al., 2017), and LMNA-mutations can cause changes in nuclear stability that lead to nuclear damage (Sullivan et al., 1999; Nikolova et al., 2004; Gupta et al., 2010; De Vos et al., 2011; Zwerger et al., 2013; Cho et al., 2019; Earle et al., 2019). To determine whether the high degree of nuclear defects exhibited by R249Q mutant nuclei could be related to altered nuclear mechanics, we measured nuclear stiffness of R249Q mutant iPSC-CMs and healthy controls with a Chiaro Nanoindenter (Optics11; Fig. 2.4D). We confirmed that nuclei in iPSC-CMs are close to the cell surface with little to no cytoskeleton over the top of nuclei (Fig. S2.4A). Thus, when probing the nucleus in intact cells, the resulting force-indentation curves are expected to primarily reflect the mechanical properties of the cell nucleus. The R249Q iPSC-CMs had significantly reduced nuclear stiffness (Fig. 2.4E) which was also reflected by the fact that at the same indentation force, the R249Q mutant iPSC-CM nuclei had a significantly increased indentation depth compared to healthy control iPSC-CMs (Fig. S2.4B). Interestingly, although R249Q mutant iPSC-CMs had increased nuclear cross-sectional areas compared to healthy control iPSC-CMs (Fig. S2.4D), consistent with our earlier measurements (Fig. 2.2D), we did not find a significant 67 correlation between nuclear cross-sectional area and nuclear stiffness across individual cells for either cell line (Fig. S2.4B), likely due to the large variability in nuclear cross-sectional area between individual cells. Fig. 2.4. R249Q iPSC-CMs exhibit decreased nuclear stiffness and increased nuclear envelope rupture. (A) iPSC-CMs were modified with a nuclear envelope (NE) rupture reporter, NLS-GFP, consisting of GFP with a Nuclear Localization Sequence. NLS-GFP is normally packaged in the nucleus and leaks into the cytoplasm upon NE rupture. (B) Representative time lapse series of a G449V iPSC-CM undergoing NE rupture and repair. Scalebar = 30 µm. Time units are represented in hours:minutes. (C) Quantification of NE ruptures shows that R249Q iPSC-CMs have significantly increased NE rupture compared to the healthy control (WT) iPSC-CM lines. (D) iPSC-CM nuclei were indented using a spherical probe on an Optics11 Chiaro nanoindenter to quantify nuclear stiffness. The indentation probe was lowered until it reached a load of 40 nN; the load was held for 5 seconds and then the probe was slowly retracted. (E) Quantification of nuclear elastic modulus, determined by fitting a Hertzian model to the load-indentation curves, revealed that R249Q nuclei are significantly softer compared to healthy control (WT) nuclei. NE rupture data presented as mean ± SEP, nuclear stiffness data represented as mean ± SEM. *, p < 0.05 vs. WT, **, p < 0.01 vs. WT. N ≥ 68 nuclei per group for nuclear stiffness, N ≥ 1020 nuclei per group for NE rupture. 68 Reduced Lamin A/C expression can partially explain R249Q nuclear damage Since the iPSC-CM lines have a heterogeneous genetic background, we wanted to confirm whether the defects in nuclear shape and stability observed in the R249Q mutant iPSC-CMs could be attributed to the reduced levels or function of Lamin A/C levels in these cells. To do so, we assessed the effect of shRNA-mediated Lamin A/C depletion (shLMNA) in healthy control iPSC- CMs on nuclear morphology and NE rupture. Cells treated with non-target shRNA (shNT) served as isogenic controls. Depletion of Lamin A/C was confirmed by immunofluorescence labeling. Due to limited transduction efficiency in iPSC-CMs, only about 50% of the shLMNA-modified cells exhibited loss of Lamin A/C (‘shLMNA-KD’ cells) at 7 days post transfection, with the remainder showing near normal Lamin A/C levels (‘shLMNA-no KD’ cells) (Fig. 2.5A; Fig. S2.5A-B). The shLMNA-KD iPSC-CMs had significantly increased nuclear area and volume compared to shLMNA-no KD and shNT controls (Fig. 2.5C-D), consistent with the increase in nuclear area and volume observed in the R249Q cells (Fig. 2.3D-E). Surprisingly, we did not detect any changes in circularity index between shLMNA-KD nuclei and either shNT or shLMNA-no KD nuclei (Fig. 2.5B). These findings suggest that reduced Lamin A/C levels directly influence nuclear size in iPSC-CM nuclei, but other factors may contribute to abnormal nuclear shape, although it is also possible that prolonged loss of Lamin A/C is required to cause more substantial defects in nuclear shape or the timing of Lamin A/C loss during cardiac differentiation of iPSC- CMs influences nuclear shape. Time-lapse experiments with shLMNA- and shNT-modified iPSC-CMs expressing the NLS-GFP NE rupture reporter suggested that Lamin A/C depletion may increase NE rupture (Fig. 2.5E), although the difference between the shLMNA iPSC-CMs and shNT controls was not quite 69 statistically significant (p = 0.12). This lack of statistical significance was likely due to the limited transduction efficiency, in which only ~50% of the observed shLMNA cells are expected to actually lack Lamin A/C (Fig. S2.5B). Overall, these data are consistent with a previous study reporting increased NE rupture in Lamin A/C depleted cardiac myocytes (Cho et al., 2019), and suggest that reduced Lamin A/C levels or functional loss of Lamin A/C could be at least partially responsible for the increase in NE ruptures in the R249Q mutant iPSC-CMs. On the other hand, the fraction of shLMNA iPSC-CMs exhibiting NE rupture was substantially lower than in the R249Q mutant cells, even when accounting for the reduced transduction efficiency, suggesting that additional mechanism contribute to the increased incidence of NE rupture in the R249Q mutant iPSC-CMs. 70 Fig. 2.5. Depletion of Lamin A/C in healthy control iPSC-CMs causes increased nuclear area, volume, and nuclear envelope ruptures. (A) Representative immunofluorescence images for Lamin B1 and Lamin A/C in healthy control (WT) iPSC-CMs either modified with shRNA targeting LMNA (shLMNA) or a non-target control shRNA (shNT). White dotted circles mark shLMNA treated nuclei with visible Lamin A/C depletion. Scalebar = 50 µm. (B) shLMNA nuclei with depletion of Lamin A/C (knockdown; KD shLMNA) have no change in circularity index compared to both non-target (shNT) control and non-depleted shLMNA nuclei (no KD). (C) KD shLMNA nuclei have significantly increased nuclear area compared to both non-target (shNT) control and non-knockdown shLMNA nuclei. (D) KD shLMNA nuclei have significantly increased nuclear volume compared to both non-target (shNT) control and non- knockdown shLMNA nuclei. (E) Unselected populations of shLMNA treated iPSC-CMs had a trend towards increased NE rupture compared to shNT treated controls. Due to the experimental protocol using live cells, we could not distinguish cells with successful Lamin A/C depletion from those without Lamin A/C depletion in the shLMNA treated cells. Data presented as mean ± SEM. **, p < 0.01 vs. control, ***, p < 0.001 vs. control, ****, p < 0.0001 vs. control. N ≥ 83 nuclei per group for nuclear shape and area, N ≥ 47 nuclei per group for nuclear volume, N > 1,415 nuclei per group for NE rupture. 71 LMNA-mutant iPSC-CMs exhibit reduced localization of lamins to the nuclear envelope LMNA mutations may alter Lamin A/C self-assembly or lamin phosphorylation, resulting in improper assembly of the lamina and mislocalization of Lamin A/C from the NE (Cenni et al., 2005; Wiesel et al., 2008; Zwerger et al., 2015; Bertrand et al., 2020). Since defects in lamin assembly could impair the resistance of the nucleus to mechanical forces exerted by the cytoskeleton, resulting in nuclear defects and NE rupture (De Vos et al., 2011; Zwerger et al., 2013; Denais et al., 2016), we examined the intranuclear distribution of Lamin A/C and Lamin B1 in the LMNA mutant and healthy control iPSC-CMs by immunofluorescence labeling. Both the R249Q and the L35P mutant iPSC-CMs frequently had reduced Lamin A/C fluorescence at the nuclear periphery (Fig. 2.6A). For a more quantitative analysis, we measured the immunofluorescence intensity profiles of Lamin A/C (Fig. 2.6B) and Lamin B1 (Fig. S2.6) along a line across the central z-plane of nuclei. Lamin fluorescence intensity profiles were normalized to the nuclear diameter and to the area under the curve to account for variations in nuclear size and Lamin A/C levels (Fig. 2.6C). Whereas healthy control iPSC-CMs exhibited Lamin A/C fluorescence intensity profiles with large peaks at the nuclear periphery, indicating that Lamin A/C is predominantly found at the nuclear lamina in these cells, the R249Q, L35P, and, to a lesser extent, G449V mutant iPSC-CMs had significantly decreased Lamin A/C at the nuclear periphery and increased Lamin A/C in the nucleoplasm (Fig. 2.6C-D). To understand the relative contributions of assembled vs. soluble Lamin A/C to lamin mislocalization, we washed out soluble nuclear proteins in semi-permeabilized nuclei of healthy control and R249Q mutant iPSC-CMs, as a representative cell line. We imaged the remaining non- soluble Lamin A/C (Fig. S2.7A) and quantified the un-normalized Lamin A/C fluorescence 72 intensity in the nucleoplasm and at the NE to better capture the differences in protein levels at each location (Fig. S2.7B). Both R249Q mutant and healthy control iPSC-CMs had a similar pool of nucleoplasmic Lamin A/C that decreased with washout (Fig. S2.7B), indicating that much of the nucleoplasmic Lamin A/C was soluble and that R249Q mutant iPSC-CMs maintain a similar pool of nucleoplasmic Lamin A/C to healthy controls. However, the R249Q mutant iPSC-CMs had significantly decreased Lamin A/C at the NE compared to healthy controls, both in non-washout and washout conditions (Fig. S2.6B), suggesting that the R249Q mutation results in reduced Lamin A/C assembly at the NE. Lamin A/C fluorescence intensity profiles for non-washout (Fig. S2.7C) and washout (Fig. S2.7D) iPSC-CMs provide additional visual confirmation of these results. Together, these results suggest that the changes in Lamin A/C localization can be attributed to reduced levels of Lamin A/C assembled at the NE. Lamin A/C and Lamin B1 form separate but interacting protein meshworks at the nuclear periphery (Shimi et al., 2015; Xie et al., 2016a), and mutation or depletion of either protein can disrupt the structural organization of the other (Vigouroux et al., 2001; Muchir et al., 2004; Guo et al., 2014; Shimi et al., 2015; Steele-Stallard et al., 2018). Thus, we hypothesized that perturbed Lamin A/C assembly could also impact the organization of Lamin B1 in the LMNA mutant iPSC- CMs. Indeed, all three LMNA mutant iPSC-CM lines exhibited an increased nucleoplasmic to peripheral ratio of Lamin B1 compared to healthy control iPSC-CMs (Fig. 2.6E), with R249Q iPSC-CMs exhibiting the most severe defects in Lamin B1 intranuclear distribution, and L35P and G449V iPSC-CMs displaying milder abnormalities. Collectively, these data indicate that mislocalization of Lamin A/C from the NE can alter Lamin B1 distribution. Furthermore, they suggest that the particularly severe defects in nuclear morphology and stability in the R249Q 73 mutant iPSC-CMs could potentially arise in part from the mislocalization of Lamin B1 from nuclear periphery, given the critical role of B-type lamins in nuclear stability and preventing NE rupture (Lammerding et al., 2006; Hatch et al., 2013; Chen et al., 2018, 2019a). Fig. 2.6. LMNA mutant iPSC-CMs have mislocalization of Lamin A/C and Lamin B1 from the nuclear envelope. (A) Confocal immunofluorescence images of cross sections through the center of nuclei show that L35P and R249Q iPSC-CM nuclei have reduced fluorescence intensity of Lamin A/C at the nuclear periphery, and R249Q nuclei additionally have reduced fluorescence intensity of Lamin B1 at the nuclear periphery. Scalebars = 50 µm. (B) Schematic of the lamin localization profile analysis. Lamin A/C fluorescence intensity profiles were taken at the central z-plane of nuclei from confocal image slices. (C) A plot of the normalized Lamin A/C fluorescence intensity profile shows that L35P and R249Q nuclei have reduced fluorescence intensity peaks at the nuclear periphery (nuclear envelope; NE) and increased fluorescence intensity in the nucleoplasm. (D) L35P, R249Q, and G449V iPSC-CMs have a significantly increased ratio of nucleoplasmic to peripheral Lamin A/C, indicating lamin mislocalization away from the nuclear envelope. (E) L35P, R249Q, and G449V-iPSC-CMs have a significantly increased ratio of nucleoplasmic to peripheral Lamin B1, with R249Q nuclei having a particularly increased ratio, indicating lamin mislocalization away from the nuclear envelope. Data presented as mean ± SEM. *, p < 0.05 vs. WT, ***, p < 0.001 vs. WT, ****, p < 0.0001 vs. WT. N > 60 nuclei per group. 74 Lamin A/C phosphorylation is not altered in LMNA mutant iPSC-CMs An altered ratio of nucleoplasmic to peripheral Lamin A/C could be due to a reduced ability of mutant Lamin A/C to self-assemble into higher order filaments (Wiesel et al., 2008; Zwerger et al., 2013; Bertrand et al., 2020) or an increase in Lamin A/C phosphorylation, which inhibits Lamin A/C’s ability to form dimers and assemble at the nuclear periphery (Swift et al., 2013; Buxboim et al., 2014). To test whether increased phosphorylation of Lamin A/C was responsible for reduced Lamin A/C localization to the NE, we immunofluorescently labeled iPSC-CMs for Lamin A/C phosphorylated at Serine 22 (p-Lamin A/C) and for total Lamin A/C (Fig. S2.8A). We did not detect any significant increase in the ratio of phosphorylated Lamin A/C to total Lamin A/C in the LMNA mutant iPSC-CMs (Fig. S2.8B), indicating that increased Lamin A/C phosphorylation is not responsible for reduced peripheral Lamin A/C in these cells. These data suggest instead that likely altered self-assembly of Lamin A/C into higher order filaments at the NE is responsible for lamin mislocalization in LMNA-mutant iPSC-CMs. Nuclear abnormalities correlate with lamin mislocalization from the nuclear envelope Our data suggest that LMNA mutant iPSC-CMs have, to a varied degree, defects in Lamin A/C and Lamin B1 localization to the nuclear periphery and that the R249Q mutant iPSC-CMs show particularly severe defects in both Lamin A/C and Lamin B1 organization, an observation which may explain the increased susceptibility of these cells to nuclear damage. To determine the degree to which mislocalization of lamins can explain nuclear size and shape defects in different cell lines, we computed a “lamin mislocalization index” from the average of the nucleoplasmic to peripheral Lamin A/C and Lamin B1 ratios for each cell line and correlated this index with a normalized “nuclear defects score,” computed from the geometric mean of nuclear circularity index, area, and 75 volume (Fig. 2.7). Intriguingly, the nuclear defects score across the panel of iPSC-CMs strongly correlated with the lamin mislocalization index (R2 = 0.96) with a significantly positive slope (p = 0.004), indicating that the extent of impaired Lamin A/C and Lamin B1 assembly at the NE potentially explains the varying degrees of nuclear damage severities across multiple cell lines. Fig. 2.7. Lamin mislocalization correlates to nuclear damage in LMNA-mutant iPSC- CMs. Lamin mislocalization index (defined as the average of Lamin A/C and Lamin B1 nucleoplasmic to peripheral intensity ratios) shows a significant positive correlation to the nuclear defects score (defined as the normalized geometric mean of nuclear circularity index, area, and volume) for the five iPSC-CM cell lines used in this study. Discussion and conclusion The mechanisms through which different LMNA mutations cause an array of disease phenotypes and severities has long remained an open question. Here, using three LMNA mutant iPSC lines associated with LMNA-DCM, we made the novel discovery that the degree of impaired localization of Lamin A/C and Lamin B1 to the NE can explain much of the varying severities of nuclear damage observed in iPSC-CMs with different LMNA mutations. We propose a mechanism in which LMNA mutations cause defective assembly of Lamin A/C, resulting in reduced Lamin A/C levels at the nuclear periphery, which, in turn, perturbs Lamin B1 localization to the nuclear periphery. Collectively, this reduced lamin assembly at the NE leads to impaired nuclear stability, 76 resulting in defects in nuclear shape and size and increased nuclear fragility and NE rupture (Fig. 2.8). Fig. 2.8. Proposed mechanism of nuclear damage in LMNA-mutant iPSC-CMs. Reduced Lamin A/C levels and/or defective assembly of Lamin A/C causes Lamin A/C mislocalization from the nuclear envelope. This mislocalization alters nuclear mechanics, particularly reducing nuclear stiffness and stability, leaving nuclei susceptible to cytoskeletal forces exerted on the nucleus in cardiac myocytes. These forces in turn cause abnormal nuclear size and shape and increased nuclear envelope rupture. Nuclear envelope rupture, and potentially other nuclear damage, can cause a decline in muscle health, potentially through mechanisms such as DNA damage or increased apoptosis. Similar to iPSCs from healthy controls, LMNA mutant iPSCs have low expression of Lamin A/C and lack nuclear shape abnormalities. Upon differentiation into cardiomyocytes, Lamin A/C expression is upregulated, and LMNA mutant iPSC-CMs display varying degrees of nuclear abnormalities in the form of altered nuclear shape and/or increased nuclear size compared to healthy controls. R249Q mutant iPSC-CMs had the most severe nuclear size and shape defects and exhibited significantly more NE ruptures than the other iPSC-CMs, likely due to their reduced nuclear stiffness and stability. Healthy control iPSC-CMs depleted of Lamin A/C showed similar defects in nuclear morphology and increased NE rupture, as in the LMNA mutant iPSC-CMs, particularly the R249Q cells, confirming our observations in an isogenic background. Intriguingly, R249Q mutant iPSC-CMs exhibited higher levels of NE rupture than the Lamin A/C- 77 depleted iPSC-CMs. Furthermore, all LMNA mutant iPSC-CMs had essentially normal overall levels of Lamin A/C, although the R249Q iPSC-CMs showed a trend towards lower Lamin A/C expression, which may cause some degree of LMNA haploinsufficiency. These findings suggest that at least some LMNA mutations function in a dominant negative manner. In support of this idea, expression of the R249Q equivalent in Drosophila Lamin C caused dominant effects on nuclear structure, including alterations in the organization of A- and B-type lamins (Hinz et al., 2021). Furthermore, in silico analysis of the Lamin A/C R249Q dimer predicted an arched conformation in the alpha helical rod domain of the extended Lamin A/C model that could be responsible for altered lamina assembly (Hinz et al., 2021). Pointing to a potential mechanism, all three LMNA mutant iPSC-CM lines exhibited Lamin A/C and Lamin B1 mislocalization from the nuclear periphery compared to healthy controls. This is likely due to defective assembly of Lamin A/C at the nuclear lamina, as the LMNA mutant cells still showed reduced Lamin A/C levels at the NE after removing soluble lamins (Fig. S2.7). These results are in line with previous studies showing that LMNA mutations may result in Lamin A/C mislocalization from the NE (Wiesel et al., 2008; Zwerger et al., 2013; Steele-Stallard et al., 2018; Bertrand et al., 2020), and that disrupting Lamin A/C assembly at the nuclear periphery impairs nuclear stability (Zwerger et al., 2015). However, although it is now well recognized that Lamin A/C and Lamin B1 form interacting meshworks at the nuclear periphery and that depletion of one filament system disrupts and disrupts the structure of the other (Vigouroux et al., 2001; Muchir et al., 2004; Guo et al., 2014; Shimi et al., 2015; Steele-Stallard et al., 2018), little work has been done to understand the consequences of altered Lamin A/C assembly on Lamin B1 organization in laminopathies. The results presented here demonstrate that beyond disrupting the structure of 78 Lamin B1, LMNA mutations can displace Lamin B1 from the NE, which is expected to make cells more susceptible to NE rupture (Lammerding et al., 2006; Hatch et al., 2013; Denais et al., 2016). While previous studies have shown that Lamin A/C mislocalization correlates with defective nuclear shape (Wiesel et al., 2008; Zwerger et al., 2013; Steele-Stallard et al., 2018; Bertrand et al., 2020) and altered nuclear mechanics (Lammerding et al., 2004a; Zwerger et al., 2015), our results point to an intriguing discovery that the ‘lamin mislocalization index,’ describing defects in both Lamin A/C and Lamin B1 assembly at the NE, can explain >95% of the variability in nuclear defects in different LMNA mutations and healthy control cell lines. Interestingly, at the level of individual nuclei, this correlation between lamin mislocalization and degree of nuclear abnormalities is roughly maintained among healthy control iPSC-CM nuclei but not R249Q nuclei (Fig. S2.9), suggesting that the increase in nuclear defects observed in R249Q mutant iPSC-CMs cannot be solely explained by an increase in lamin mislocalization from the nuclear periphery. This could potentially be due to a threshold for amount of lamin assembled at the NE, below which nuclei are increasingly susceptible to nuclear abnormalities. R249Q mutant iPSC-CMs generally have a high degree of both Lamin A/C and Lamin B1 mislocalization from the NE and evidence of significantly decreased nuclear stiffness, which could indicate that their nuclei are below this threshold for peripheral lamin assembly and thus are more susceptible to deformation. Additionally, we hypothesize that other changes to the cellular or nuclear structure, such as differences in contractility or force exertion on the NE (Cho et al., 2019), could drive increased nuclear abnormalities in this population. However, one limitation of this study is that iPSC-CMs were neither uniformly aligned on a substrate nor electrically paced, thus resulting in disorganized cellular structure with likely variation in cellular contractility. As actin contractility is known to 79 play a role in nuclear damage (Hatch and Hetzer, 2016; Cho et al., 2019), it is plausible that the variability between individual nuclei within a cell line could be attributed at least in part to differences in contractility and force transmission to the NE: nuclei which are subject to increased amounts of force, or force that is applied to the nucleus in different ways, could increase nuclear abnormalities. The loss of nuclear mechanical strength and NE integrity due to the loss of Lamin A/C and/or Lamin B1 at the NE has drastic mechanical consequences on cellular health and survival (Denais et al., 2016; Chen et al., 2018, 2019a; Cho et al., 2019; Earle et al., 2019; Shah et al., 2021), particularly through susceptibility of the nucleus to damage from intracellular forces (Hatch and Hetzer, 2016; Cho et al., 2019; Earle et al., 2019; Shah et al., 2021). However, one interesting observation of this study is that while the degree of lamin mislocalization works well to predict the effect of the LMNA mutation on nuclear mechanics and stability, it is not necessarily a good predictor of patient phenotype. Of the three LMNA mutations used in this study, the patient with the L35P mutation had the most severe phenotype, presenting both skeletal muscle and cardiac phenotypes during childhood and living only to age 15. Conversely, the patient presenting the R249Q mutation, whose iPSC-CMs had more severe disruptions to nuclear mechanics than L35P iPSC-CMs, developed early onset muscular dystrophy in childhood, but only developed severe cardiac dysfunction later in life. This disparity between patient and cellular phenotype is not completely surprising, as additional defects caused by LMNA mutations, such as altered gene expression or biochemical signaling likely contribute to disease onset and progression. For example, altered Lamin A/C and Lamin B1 assembly at the NE can cause changes in chromatin organization and gene expression (Bertero et al., 2019; Cheedipudi et al., 2019; Kim et al., 2019), 80 including nuclear mechanotransduction and associated signaling (Kirby and Lammerding, 2018; Maurer and Lammerding, 2019; Donnaloja et al., 2020). As such, loss of assembled Lamin A/C and/or Lamin B1 at the nuclear periphery may impair mechanically-induced gene expression (Maurer and Lammerding, 2019; Donnaloja et al., 2020). Overall, our results point to the degree of lamin mislocalization from the NE as an important mechanism across LMNA-DCM mutations which determines the degree of nuclear mechanical damage in cardiomyocytes. Moreover, altered lamin assembly at the NE may present a potential link between the proposed laminopathy pathogenic mechanisms of impaired nuclear stability and altered gene expression driving LMNA mutant skeletal muscle and cardiac laminopathies (Maurer and Lammerding, 2019; Donnaloja et al., 2020). Future studies should be aimed at determining the molecular mechanism(s) by which nuclear damage causes cardiac dysfunction in LMNA-DCM, including obtaining a better understanding of changes in mechanotransduction signaling and gene expression. Additionally, our work and others showing that Lamin A/C assembly has implications in the degree of nuclear damage and altered nuclear mechanics (Wiesel et al., 2008; Zwerger et al., 2013; Bertrand et al., 2020) suggest that by understanding the mutation-specific degree of defective lamin assembly, we may ultimately be able to identify patient mutations that would most benefit from therapeutics targeting the reduction of force transmission to the nucleus (Cho et al., 2019; Earle et al., 2019; Chai et al., 2021). Although additional mechanisms may drive LMNA- DCM, such therapeutics hold promise to improve cellular health and survival (Cho et al., 2019; Earle et al., 2019; Chai et al., 2021). 81 Acknowledgements The authors thank the Cure Muscular Dystrophy Foundation for the LMNA L35P iPSCs; Hanna Gimse for analyzing the nuclear volume in shRNA experiments; the Cornell Institute of Biotechnology for performing library preparation and Illumina sequencing experiments; the Cornell Statistical Consulting Unit for developing the regressions performed in R; Dr. Kehan Zhang and Dr. Christopher Chen for hosting M.E.M. to learn iPSC culture and cardiac differentiation protocols; and Dr. Kathleen Xie for the nuclear semi-permeabilization protocol. This work was supported by awards from the National Institutes of Health (R01 HL082792 to J.L., R01 HL128075 to E.M.M., and F30 HL142187 to A.M.G.); the National Science Foundation (CBET 1715606 to J.L.; Graduate Research Fellowships 2016229710 to M.M.), the American Heart Association (20PRE35080179 to M.M.), the Muscular Dystrophy Association (MDA603238 to T.J.K), and the Volkswagen Foundation (A130142 to J.L.). 82 Supplementary Material Fig. S2.1. iPSCs express detectable levels of Lamin B1 and Lamin A/C. Representative immunofluorescence images of WT and LMNA-mutant iPSCs, revealing detectable expression of Lamin A/C and Lamin B1. Scalebar = 50 µm. 83 Fig. S2.2. LMNA mutant iPSCs and iPSC-CMs show normal levels of Lamin mRNA expression. (A) LMNA expression was low in iPSCs and increased with cardiac differentiation into iPSC-CMs. The difference in LMNA mRNA expression in the R249QiPSC-CMs was not statistically different compared to healthy controls iPSCs. (B) LMNB1 expression was similar between LMNA mutant and healthy control iPSCs and iPSC-CMs, although LMNA-mutant iPSCs showed a trend towards increased LMNB1 expression compared to healthy controls (p > 0.14 vs both WT lines). (C) Quantification of Lamin A protein expression from western analysis showed that R249Q iPSC-CMs had a trend towards decreased Lamin A (p > 0.08 vs both WT lines). (D) Quantification of Lamin C protein expression from western analysis showed that R249Q PSC-CMs had a trend towards decreased Lamin C (p > 0.18 vs both WT lines). Data presented as mean ± SEM. N = 3-4 samples for all groups for mRNA levels. N = 5 samples for all groups for western quantification. 84 Fig. S2.3. Duration of nuclear envelope rupture is not altered by LMNA mutations. Quantification of nuclear envelope rupture duration, based on the time NLS-GFP was displaced from the nucleus, in LMNA mutant and healthy control iPSC-CMs. Difference in nuclear envelope rupture duration between healthy control (WT) and LMNA-mutant iPSC-CMs exhibiting nuclear envelope rupture were not statistically significant. Data presented as mean ± SEM. N ≥ 1020 nuclei per group. 85 Fig. S2.4. R249Q iPSC-CM nuclear stiffness does not correlate to increased nuclear area. (A) An orthogonal view of representative healthy control (WT2) and R249Q iPSC-CMs shows that nuclei (labeled for Lamin B1, green) are almost directly exposed to the apical plasma membrane with very little cytoskeleton atop the nucleus, as indicated by immunofluorescence labeling for cTnT. Scalebar = 50µm. (B) Nuclear indentation depth was significantly increased in R249Q-iPSC-CM cells compared with WT2. These depths suggest that indentations were approximately 33-50% of the height of nuclei, as indicated in Figure 3D, and significantly more than the height of any cytoskeletal elements over the top of nuclei in panel A. (C) R249Q iPSC- CMs used in nuclear stiffness experiments have increased nuclear area, but (D) a scatter plot shows no correlation between the nuclear Elastic Modulus vs. nuclear cross-sectional area across individual cells. Equations indicate the results of linear regression models for each cell line. Data in B and C presented as mean ± SEM. *, p < 0.05 vs. WT. N ≥ 68 nuclei per group. 86 Fig. S2.5. Validation of shRNA mediated Lamin A/C depletion in healthy control iPSC- CMs. (A) Histogram of average nuclear Lamin A/C immunofluorescence intensities shows an increased population of nuclei with low or absent Lamin A/C expression in iPSC-CMs modified with shRNA targeting LMNA (shLMNA) or a non-target (shNT) control. Cells did not undergo selection prior to analysis. (B) Quantification of the number of nuclei with Lamin A/C Knockdown (KD) indicates >55% knockdown efficiency in the shLMNA treated cells. (C) shNT and shLMNA treated iPSC-CMs have similar durations of nuclear envelope rupture. Data in B and C presented as mean ± SEM. N > 150 nuclei per group for fluorescence intensity and Lamin A/C depletion efficiency, N > 1,415 nuclei per group for nuclear envelope rupture. Fig. S2.6. Lamin B1 fluorescence intensity profiles. Lamin B1 fluorescence intensity profiles taken at the central z-plane of nuclei from confocal image sections of healthy control and LMNA-mutant iPSC-CMs. 87 Fig. S2.7. LMNA-mutant iPSC-CMs have decreased polymerized Lamin A/C at the nuclear periphery, but normal levels of nucleoplasmic Lamin A/C. (A) Representative immunofluorescence images of Lamin A/C in healthy control (WT) and R249Q mutant iPSC- CMs following washout of soluble nuclear proteins and in non-washout controls. Scalebars = 50µm. (B) Quantification of the average nucleoplasmic and peripheral fluorescence intensities for healthy control (WT) and R249Q samples after washout and non-washout control showed that nucleoplasmic Lamin A/C levels decreased following washout but remained similar between cell lines. In contrast, R249Q iPSC-CMs had significantly reduced levels of peripheral Lamin A/C levels compared to healthy controls, and the differences remained after washout, indicating that the R249Q cells have reduced levels of assembled Lamin A/C at the nuclear lamina. These results are also visible in the Lamin A/C fluorescence intensity profiles of non- washout controls (C) and after washout (D). Data in B presented as mean ± SEM. ****, p < 0.0001 vs. WT. N > 60 nuclei per group. 88 Fig. S2.8. LMNA-mutant iPSC-CMs do not have increased Lamin A/C phosphorylation. (A) Representative immunofluorescence images of LMNA mutant and healthy control iPSC- CMs immunofluorescently labeled for phospho-Ser22 Lamin A/C (p-Lamin A/C). Scalebar = 50µm. (B) The ratio of phospho-Lamin A/C to total Lamin A/C is not increased in LMNA- mutant iPSC-CMs. Data presented as mean ± SEM. **, p < 0.01, ***, p < 0.001. N > 60 nuclei per group. Fig. S2.9. Lamin mislocalization does not correlate to nuclear damage across single cells within individual iPSC-CM lines. Correlating lamin mislocalization index and nuclear defects score on individual nuclei for WT iPSC-CMs, which exhibit a lower degree a lamin mislocalization and nuclear damage, showed only a weak positive correlation (R2 = 0.06), which was not quite statistically significant (p = 0.088). The R249Q nuclei did not show any correlation between lamin mislocalization and nuclear defects. N > 50 nuclei per cell line for single cell analysis. 89 Table S2.1. Primary and secondary antibodies used for immunofluorescence labeling and western analysis. Antibody Catalog # Vendor Dilution (IF) Dilution (Western) Lamin A/C (E-1) sc-376248 Santa Cruz 1:150 1:1000 Lamin B1 12987-1-AP Proteintech 1:200 1:5000 cTnT RV-C2 DSHB 1:20 Phospho-Lamin A/C 2026S Cell Signaling 1:500 (Ser22) Pan-Actin (D18C11) 8456S Cell Signaling 1:1000 GAPDH 6 0004-1-Ig Proteintech 1:20,000 H3 Abcam 1 :5000 Alexa Fluor 488; Goat A-21121 Invitrogen 1:250 anti-mouse IgG1 Alexa Fluor 568; Goat A-21124 Invitrogen 1:250 anti-mouse IgG1 Alexa Fluor 647; Goat A-21242 Invitrogen 1:250 anti-mouse IgG2b Alexa Fluor 488; A-21206 Invitrogen 1:250 Donkey anti-rabbit IgG Alexa Fluor 568; A-10042 Invitrogen 1:250 Donkey anti-rabbit IgG IRDye 800CW Donkey 926-32213 Li-cor 1:3000 anti-Rabbit IgG IRDye 880RD Donkey 926-68072 Li-cor 1:3000 anti-Mouse IgG 90 CHAPTER 3 Investigating the disrupted signaling pathways in LMNA-Dilated Cardiomyopathy with RNA-seq Lamin A/C form a dense protein meshwork under the inner nuclear membrane that gives the nucleus structural support, and participates in chromatin organization, gene expression, and regulation of downstream signaling pathways. Mutations in the gene encoding Lamin A/C, LMNA, are associated with numerous human diseases, termed ‘laminopathies,’ which includes dilated cardiomyopathy (LMNA-DCM) and muscular dystrophies, among others. While laminopathies are characterized by mechanical damage to nuclei and disruption of signaling pathways, our understanding of the transcriptomic changes driving disease progression remains incomplete, thus limiting the development of therapeutics targeting the underlying disease pathology. Here, through the use of two Lmna-DCM mouse models and an induced pluripotent stem cell-derived cardiomyocyte line derived from an LMNA-DCM patient carrying the G449V mutation, we implicate misregulation of metabolism and the extracellular matrix in both disease onset and phenotype. This is among the first studies directly comparing altered gene expression and cell signaling in both mouse and human models of LMNA-DCM, and employing both pre- and post- phenotype Lmna-DCM mice to identify signaling pathways not only associated with disease phenotype, but that may be driving the onset and progression of Lmna-DCM. Introduction Two classic hypotheses through which laminopathies cause disease have been proposed: the structural hypothesis, in which mutant lamins cause increased nuclear fragility, cell damage, and 91 death, particularly in mechanically active or stressed tissues, and the gene regulation hypothesis, in which lamin mutations play a tissue-specific role in alteration of gene expression through gene activation or silencing (Peric-Hupkes et al., 2010a) or inhibition of tissue-specific factor binding (Simon and Wilson, 2013). As evidence of the gene regulation hypothesis, numerous cell signaling pathways have been implicated in skeletal and cardiac laminopathy pathogenesis, including transforming growth factors β1 and 2 (Van Berlo et al., 2005; Bernasconi et al., 2018), MyoD (Frock et al., 2006), MAPK (specifically extracellular signal–regulated kinases 1 and 2, JNK, and p38α) (Muchir et al., 2007, 2012), WNT/β-catenin (Muchir et al., 2007, 2012; Le Dour et al., 2017), E2F/DNA damage response (Chen et al., 2019b; Cho et al., 2019; Pradas et al., 2020), and FOXO transcription (Auguste et al., 2018). While these studies represent progress towards understanding LMNA-DCM signaling dysfunction and potential treatment avenues, relatively few sequencing studies have been done to understand transcriptomic changes that result from alterations to gene expression in LMNA-DCM (Auguste et al., 2018; Chen et al., 2019b; Pradas et al., 2020; Shao et al., 2020). Additionally, few studies have been done to understand transcriptomic changes that occur before the onset of LMNA-DCM (Shao et al., 2020), which is critical for the development of therapeutics that can treat the underlying cause of disease rather than ameliorate disease symptoms. While RNA-sequencing (RNA-seq) studies have succeeded in identifying broader signaling pathways in progression of Lmna-DCM (Auguste et al., 2018; Chen et al., 2019b; Shao et al., 2020), several of these have been endpoint studies, at which point gene expression driving disease progression is likely already conflated with gene expression changes as a result of the disease phenotype. Only one sequencing study has aimed to understand the contributions of gene 92 expression to the development and progression of Lmna-DCM, through the use of Lmna-deficient mice (Shao et al., 2020). Moreover, only one study (Chen et al., 2019b) has examined disrupted gene expression with RNA-seq in human samples. Since mice and humans with LMNA-DCM differ in their expression of mutant alleles, with mice requiring homozygous expression of the mutant allele for phenotype development and humans typically developing the phenotype with only heterozygous expression (Stewart et al., 2007), it is difficult to assume that disrupted human gene expression may be synonymous with that observed in mouse models of LMNA-DCM. Here, I utilized RNA-seq with mouse and human LMNA-DCM models to understand the common and diverging gene expression and associated signaling pathways of LMNA-DCM. I used mouse left ventricle tissue harvested either pre-phenotype and/or post-phenotype from two Lmna-DCM mouse models, Lmna–/– (or Lmna KO) and LmnaN195K/N195K (or Lmna N195K), and demonstrated that although there are differences in gene expression across the models, several common signaling pathways are disrupted in both models, including decline of cardiac function, extracellular matrix, immune function, and metabolic dysfunction. Additionally, I identified several genes, Hef2, Eno3, and Etfb that may bridge the gap between decline of cardiac function and mitochondrial dysfunction in both Lmna KO and Lmna N195K models. Through the use of a human induced pluripotent stem cell-derived cardiomyocyte (iPSC-CM) model of LMNA-DCM with iPSCs derived from two healthy controls and patients with one LMNA mutation, G449V, I confirmed that the high-level signaling pathways observed in mouse models are mirrored in human samples, particularly ECM and metabolic dysfunction, and identified DEGs and misregulated signaling pathways within these pathways that are common to both mouse and human disease models. Together, this study implicates novel pathways in the progression of Lmna-DCM in the N195K 93 mouse model, and is among the first studies (Auguste et al., 2018) to identify disrupted signaling pathways common to both human and mouse Lmna-DCM. Materials and Methods Animals Lmna–/– (referred to here as Lmna KO) (Sullivan et al., 1999) and LmnaN195K/N195K (referred to here as Lmna N195K) (Mounkes et al., 2005) are two mouse models of Emery Dreifuss Muscular Dystrophy (EDMD) with accompanying Lmna-DCM that have been described previously. Lmna– /– and LmnaN195K/N195K mice were back-crossed a minimum of seven generations into a C57BL/6 line. For each of the Lmna KO and N195K models, heterozygous mice were crossed to obtain homozygous mutant and healthy control (WT) littermates. Following the onset of Lmna-DCM and decline of health, Lmna KO mice were fed a gel diet (Nutri-Gel Diet, BioServe) to improve food intake and hydration. All breeding, maintenance, and euthanasia of mice was performed according to guidelines and ethical regulations approved by the Cornell University Institutional Animal Care and Use Committee (IACUC), protocol number 2011-0099. Left ventricle tissue harvest Left ventricle tissue from post-phenotype Lmna KO mice was harvested at 4 weeks, pre-phenotype Lmna N195K mice at 4 weeks, and post-phenotype Lmna N195K mice at 10 weeks. Animals were euthanized, and the heart immediately removed and washed thoroughly in ice cold PBS to ensure all removal of blood by manual pumping with tweezers. All vessels and fat were trimmed off, the heart was dried, and cut in half through the septum to obtain the right and left ventricles. The left ventricle was then cut transversely to obtain two halves, the superior and the inferior, containing 94 the apex. Tissue was then snap frozen in liquid nitrogen and stored at -80oC until future processing for transcriptomic analysis. Cell culture Human induced pluripotent stem cell (iPSCs) lines which carry either the two LMNA R249Q or G449V mutation associated with LMNA-DCM and two healthy control cell lines were cultured and differentiated to cardiomyocytes according to the methods outlined in Chapter 3. RNA extraction and sequencing The inferior half of a left ventricle tissue was cut into several pieces and placed in a homogenizer tube with a metal bead (3 mm, McMaster-Carr). Trizol reagent was added to each tube, and tissues were homogenized on a TissueLyser II (Qiagen) at a rate of 30/second for three minutes. Homogenization was repeated about 8 times until the tissue was fully homogenized. Trizol was then transferred to an RNAse-free tube. RNA was isolated from iPSCs before passaging when colonies were large but not touching and iPSC-CMs one week after their initial passage as cardiomyocytes (Chapter 3) using Trizol. RNA was extracted from Trizol from both tissues and cells using the RNeasy Plus Kit (Qiagen). Quality of the RNA was assessed using a Fragment nalyzer, and samples which had an RQN < 7.0, indicating degradation of RNA, were exlucded. Libraries were prepared by the Cornell Institute of Biotechnology using a TruSeq DNA library prep kit (Illumina) and sequenced on a NextSeq 500/550 (Illumina). RNA-seq analysis Raw sequencing reads were quality checked using FastQC (Andrew, 2010). STAR (Dobin et al., 95 2013) was used to assemble the human genome and align reads to the human genome. Aligned reads were quality control checked, and samples with alignment <70% were removed. A count matrix was generated using featureCounts (Liao et al., 2014), which was then input to build statistical models in DESeq2 (Love et al., 2014). A separate DESeq2 model was made for each of Lmna KO and Lmna N195K and analyzed according to following four comparisons: post- phenotype (4 week) Lmna KO mutant (mut) vs. healthy control (WT); pre-phenotype (4 week) Lmna N195K mutant vs. healthy control; post-phenotype (10 week) Lmna N195K mutant vs. healthy control, and the interaction of genotype and timepoint for Lmna N195K, referred to here as mutant (or mut) over time. A DESeq2 model (Love et al., 2014) accounting for genotype (i.e. healthy control, or WT, vs mutant), cell line, differentiation state (i.e. iPSC or iPSC-CM), the interaction of cell line and genotype, and the interaction of cell line and differentiation state was first constructed. In addition, to focus on the G449V cell line iPSC-CM comparison, two DESeq2 models, each comparing one healthy control iPSC-CM line vs G449V mutant iPSC-CMs, were constructed. The intersection of the results of these two models, as defined by a DEG having a significant p-value in and having a fold change in the same direction for both models, was generated and used as a the set of significant DEGs for G449V mutant iPSC-CMs. Genes with low normalized read counts (<10) were excluded, and a set of differentially expressed genes (DEGs) was obtained by using a threshold of p < 0.01 and |Log2FoldChange>1|. Genes were annotated using the AnnotationDBI package (Pagés et al., 2021) using either the genome wide annotation for mouse, org.Mm.eg.db (Carlson, 2019b), or human, org.Hs.eg.db (Carlson, 2019a). 96 Volcano plots were generated using the EnhancedVolcano package (Blighe et al., 2021), and protein interaction maps were created using StringDB (Szklarczyk et al., 2019). KEGG pathway analysis was performed using KEGGprofile (Zhao et al., 2021) with a p-value threshold of 0.05, and Gene Ontology (GO) analysis for biological processes, molecular functions, and cellular components was performed using TopGO (Alexa and Rahnenfuhrer, 2021) with a p-value threshold of 0.01. Statistics A minimum of 3 independent animals per group or 3 independently collected iPSC or iPSC-CM samples was collected. For mouse studies, age-matched healthy control (wild type, WT) littermate controls were used. Only RNA samples with a RQN>7 were sequenced, and samples which had low sequence mapping to the respective genome or low normalized read counts were excluded from the study. Genes which had fewer than 10 read counts were excluded, and only genes which had a |Log2FoldChange>1| were considered to be differentially expressed. For all experiments a p-value of either 0.05, to define DEGs or KEGG pathways, or 0.01, for GO analysis, was used to define statistical significance. Results Transcriptomic analysis of Lmna-DCM mouse left ventricle tissues To identify affected signaling pathways in LMNA-DCM, I performed RNA-sequencing (RNA- seq) on left ventricle tissue from two Lmna-DCM mouse models, Lmna KO and Lmna N195K, which represent different severities of disease: Lmna KO mice die from LMNA-DCM between 4- 6 weeks of age, whereas Lmna N195K mice die from LMNA-DCM between 10-14 weeks 97 progression (Sullivan et al., 1999; Mounkes et al., 2005; Earle et al., 2019). I harvested left ventricle tissue from mice showing the Lmna-DCM phenotype (post-phenotype), at 4 weeks for Lmna KO and 10 weeks for Lmna N195K (Earle et al., 2019), and age-matched healthy littermate controls. Additionally, to understand the genes contributing to development of Lmna-DCM, I harvested tissue from Lmna N195K mice at 4 weeks, i.e. before they develop the phenotype and age-matched healthy littermate controls. As it is extremely challenging to obtain pre-phenotype Lmna KO mice due to the early disease development before normal weaning and genotyping, this genotype was excluded from the phenotype development analysis. For each timepoint and condition, tissues from 4 female and 4 male mice were harvested. A summary of the RNA-seq experiment is depicted in Fig. 3.1. Lmna KO Lmna N195K 4 week 4x 4x WT WT 4x 4x Mutant Mutant 4 week 10 week 4x 4x 4x 4x WT WT WT WT 4x 4x 4x 4x Mutant Mutant Mutant Mutant Male Female Fig. 3.1. Experimental summary of conditions for mouse models of Lmna-DCM RNA-seq. Left ventricle tissue from four male and four female mice for each healthy control (WT) and mutant (either Lmna KO or Lmna N195K) at pre-phenotype, when possible, or post-phenotype was harvested. Each circle represents a group of mice included in the analysis. 98 Post-phenotype Pre-phenotype I generated a statistical model of read counts for each Lmna-DCM model, with the Lmna KO model accounting for genotype and sex, and the Lmna N195K model accounting for genotype, sex, timepoint (i.e. 4 vs 10 weeks, for pre- or post-phenotype development), and the interaction of timepoint and genotype, since the genotype-specific gene expression is expected to be different for mutant mice as Lmna-DCM progresses. I performed Principal Component Analysis and generated PCA plots to check the clustering of samples by timepoint, genotype, and sex (Fig. 3.2). For the Lmna KO model, healthy control (WT) and Lmna KO (mutant or mut) formed separate clusters across Principal Component 1, indicating that the difference in genotype is the primary source of variability between samples (Fig. 3.2A). For the Lmna N195K model, 10-week Lmna N195K (mutant, or mut) mice formed a distinct cluster along the PC2, whereas all the other conditions that did not exhibit the Lmna-DCM phenotype (4-week WT, 4-week mut, 10-week WT) formed another one to two additional clusters along the PC1 axis, with no clear indicator of the defining difference between clusters (Fig. 3.2B). Since there was little to no difference between sex for each mouse model in the PCA plots, I compared the male vs. female mice for each genotype to determine differentially expressed genes (DEGs), and found extremely few genes that were differentially expressed and known to be sex-specific. Hence, all subsequent results are based on models in which both male and female mice were grouped together by genotype. Together, these results for the Lmna KO and N195K models were as expected, with the post-phenotype mice for each model clustering and little sex-specific differences in gene expression, and indicate that the statistical models for each of the Lmna KO and N195K mouse models of Lmna-DCM are sufficient to detect transcriptomic differences in post-phenotype mice. 99 A B Lmna KO Lmna N195K 10 week mutant Fig. 3.2. PCA plots of Lmna-DCM mouse model RNA-seq results. (A) PCA plot of Lmna KO results by genotype (healthy control or WT, and Lmna KO or mut) and sex (male or M, and female or F). (B) PCA plot of LMNA N195K results by genotype (WT or mut), sex (M or F), and timepoint (4 week pre-phenotype or 4wk, or 10 week post-phenotype or 10wk). Clusters are depicted with black circles. Few common significant DEGs are expressed between Lmna-DCM models With the aforementioned statistical models, I used the DEGs from following four statistical comparisons to understand the transcriptomic differences both the Lmna KO and N195K models: Lmna KO vs healthy control (“Lmna KO post-phenotype”), pre-phenotype (4 week) Lmna N195K vs healthy control (“N195K pre-phenotype”), post-phenotype (10 week) Lmna N195K vs healthy control (“N195K post-phenotype”), and the interaction of genotype and timepoint for Lmna N195K (“N195K phenotype development”). With thresholds of p < 0.05 and |Log2FoldChange>1| to identify significant DEGs, I obtained between 38 and 166 DEGs for each comparison (Table 3.1). Table 3.1. Differentially expressed genes for each comparison. Comparison # DEGs Lmna KO post-phenotype 166 N195K pre-phenotype 73 N195K post-phenotype 38 N195K phenotype development 71 100 With the genes identified in each comparison, I generated volcano plots to examine the distribution of up and downregulated genes (Fig. 3.3). Volcano plots revealed that both Lmna KO post- phenotype (Fig. 3.3A) and N195K phenotype development (Fig. 3.3B) had a more even distribution of both up- and downregulated DEGs, while N195K pre- (Fig. 3.3C) and post- phenotype (Fig. 3.3D) comparisons had a strong skew towards upregulated genes. P-values were most significant for the Lmna KO and N195K post-phenotype comparisons (Fig. 3.3 A, D), suggesting that with disease onset some genes become very strongly differentially expressed. Fig. 3.3. Volcano plots of Lmna-DCM mouse model RNA-seq results. Volcano plots of (A) Lmna KO post-phenotype and (B) N195K phenotype development show a similar number of up- and downregulated significant DEGs, whereas (C) N195K pre- and (D) N195K 10 week post-phenotype show a more upregulated significant DEGs. 101 To determine whether there were common DEGs among the different Lmna-DCM models and timepoints that may be indicative of overarching disease mechanisms, I compared the lists of DEGs for each comparison to identify any overlapping DEGs. However, relatively few DEGs were shared between each of the four comparisons (Fig. 3.4), and genes identified as common in between the groups had little obvious connection (Table S3.1). Only one DEG, Dio2, was shared by three or more groups, although interestingly Dio2 is implicated in disrupted fat storage in another laminopathy, type 2 Familial Partial Lipidostrophy (Pellegrini et al., 2019), and thus could be related to the reduced fat storage in Lmna mutant mice, particularly even in pre-phenotype N195K mutant mice. Of the other DEGs common between more than one comparison, no overarching pathways or mechanisms are apparent without dedicated pathway analysis. Lmna KO post-phenotype N195K phenotype development N195K pre-phenotype N195K post-phenotype Fig. 3.4. Venn diagram of the number of common and differing significant DEGs among RNA-seq comparisons. Few genes were found to be common across different comparisons, and those genes found to be common showed little obvious connection to each other (Table S3.1). 102 Protein interaction mapping and pathway analysis reveal several commonly dysregulated pathways in Lmna-DCM To identify any key groups of significant DEGs that may drive dysregulated signaling pathways in Lmna-DCM, I generated a protein interaction map for each of the four statistical comparisons that illustrates the connections between related, significant DEGs. The protein interaction map for the N195K phenotype development (Fig. 3.5) was particularly interesting, containing several clusters of protein-coding genes that have strong interactions: one with several extracellular matrix (ECM) genes (top left); two connected clusters with mitochondrial and oxidative phosphorylation genes (top and left), and one with several cardiac genes (bottom). Interestingly, three key genes connect these pathways (Fig. 3.5): Hfe2, a gene involved in iron metabolism (Gobbi et al., 2002), links the metabolism and cardiac gene clusters; Eno3, which encodes an enolase isoenzyme and plays a role in muscle development and regeneration (Peshavaria and Day, 1991), connects the cardiac genes to the metabolism/oxidative phosphorylation genes via Etfb, which is involved in mitochondrial fatty acid and amino acid catabolism (Schiff et al., 2006) and connects the metabolism and oxidative phosphorylation gene clusters. However, the other comparison protein interaction maps had either no or little clustering of DEGs: the N195K pre-phenotype map showed few interactions between DEGs (Fig S3.1), the N195K post-phenotype map (Fig S3.2) showed some similarities to the N195K phenotype development comparison, particularly with the ECM and sarcomeric genes, although the genes were primarily stemming from one cluster, and the Lmna post-phenotype map (Fig S3.3) showed one primary cluster around several enzyme-encoding and immune system-related genes, with some peripheral interactions between cytoskeletal genes. However, the identification of immune-related DEGs likely does not pertain to dysregulated signaling pathways by Lmna-DCM cardiomyocytes, as this experiment being performed on bulk 103 cardiac tissue and therefore likely included other subpopulations of cells in the heart that adapt the immune system during heart failure and tissue fibrosis (Burke et al., 2016; Sweet et al., 2018). Fig. 3.5. Protein interaction map for the N195K phenotype development comparison. Each dot represents a protein-coding gene, with size denoting expression levels and halos denoting p-value. Lines show interactions between protein-coding genes. Gene clusters representing common pathways are circled, with the pathway. Three genes connecting the metabolism and oxidative phosphorylation clusters with cardiac genes – Etfb, Hfe2, and Eno3 – are denoted with arrows. With the pathways identified by these protein interaction maps in mind, I sought to identify more specific misregulated pathways with each of the four Lmna-DCM comparisons using KEGG pathway analysis, and to identify any pathways common between the comparisons. In total, the N195K pre-phenotype comparison had no significantly dysregulated pathways, indicating that 104 there were no overarching misregulated signaling pathways, and surprisingly given the onset of Lmna-DCM, the N195K post-phenotype comparison had only 5 significantly dysregulated pathways that did not correspond to similar types signaling pathways, instead corresponding to metabolism, cardiac function, and immune response. Lmna KO post-phenotype identified 8 dysregulated signaling pathways and N195K phenotype development 23 (Fig. 3.6), both of which, identified DCM and hypertrophic cardiomyopathy (HCM), suggesting that these statistical models detected some expected altered signaling pathways, and the HIF-1 signaling pathway. Additionally, both the N195K phenotype development and N195K post-phenotype comparison identified the AGE-RAGE signaling pathway, focal adhesions, and proteoglycans in cancer. Of these pathways, both HIF-1 signaling, which is commonly upregulated in ischemic cardiac disease to confer cardioprotection (Tekin et al., 2010), and the AGE-RAGE signaling pathways are of interest as they both play a role in metabolism. In addition, the N195K phenotype development comparison pointed to several other misregulated metabolic pathways – oxidative phosphorylation, purine metabolism, carbon metabolism, Thyroid hormone signaling, and PI3K- AKT signaling – which, together, could be indicative of broader defective metabolism driving the progression of Lmna-DCM. Similar to the above protein interaction maps and GO terms, numerous pathways identified by KEGG analysis for all four statistical comparisons pointed to cardiac muscle contraction, ECM interactions, and the immune. Finally, several misregulated pathways identified by the N195K phenotype development comparison are, unsurprisingly, those involved in neurodegeneration (Alzheimer, Parkinson, and Huntington disease), likely due to either these pathways being broad KEGG categories that include overlap with other broad cardiac pathways (i.e. through the inclusion of mitochondrial dysfunction and oxidative phosphorylation), or because we performed RNA-seq on bulk cardiac tissue in mouse models which can exhibit 105 neuropathy (Stewart et al., 2007). Lmna KO and N195K mut over time N195K mut over time only Muscle-related Dilated cardiomyopathy (DCM) Oxidative phosphorylation Metabolism HIF-1 signaling pathway Purine metabolism ECM/cell-cell interaction Hypertrophic cardiomyopathy (HCM) Metabolic pathways Immune system Carbon metabolism DNA damage 10wk N195K and mut over time PI3K-Akt signaling pathway AGE-RAGE signaling pathway in diabetic complications Cardiac muscle contraction Focal adhesion Platelet activation Proteoglycans in cancer Leukocyte transendothelial migration Thermogenesis Lmna KO Only Retrograde endocannabinoid signaling Ribosome Thyroid hormone signaling pathway Phagosome Non-alcoholic fatty liver disease (NAFLD) Hematopoietic cell lineage Alzheimer disease Leishmaniasis Parkinson disease Tuberculosis Huntington disease Salmonella infection 10wk N195K only Human papillomavirus infection Malaria Amoebiasis Fig. 3.6. Dysregulated KEGG pathways. 4 week N195K had no significantly dysregulated pathways, 10 week N195K had only 5 pathways, Lmna KO had 8, and N195K mutant over time had 23. Three pathways were shared between Lmna KO and N195K mutant (mut) over time, and three pathways were shared between N195K mutant (mut) over time and 10 week N195K. The majority of pathways followed the same broader classifications as was observed with protein interaction maps, denoted by color here. To find more specific misregulated signaling pathways than those identified by KEGG analysis, I used Gene Ontology (GO) analysis with each of the four comparisons. Looking at the top 15 GO terms in each comparison (Fig. 3.7), the same common themes become evident between the GO terms and those observed in the protein interaction maps and KEGG pathway analysis (colored in Fig. 3.8): ECM, cardiac function/sarcomere, metabolism, and immune system, among a few other terms. Interestingly, each of the comparisons showed trends towards different high-level biological themes, with post-phenotype Lmna KO mice having more cardiac muscle-related and immune system genes, the N195K phenotype development comparison having metabolism and ECM- related genes, and post-phenotype N195K mice having more ECM-related genes. Pre-phenotype N195K mice GO pathways did not closely follow the same high-level signaling pathways as was identified with other comparisons, although this is unsurprising given our previous analysis of 106 protein interactions and KEGG pathways. While the identified post-phenotype Lmna KO GO terms are anticipated since these mice develop a very extreme cardiac phenotype even by tissue harvest at 4 weeks of age, the differences observed across N195K comparisons are striking: at pre- phenotype, N195K mice do not show high-level signaling pathway disruption, as disease develops (captured in the N195K phenotype development comparison), metabolism becomes strongly misregulated and some changes to ECM and muscle function begin to occur, and finally by full phenotype development at 10 weeks, N195K mice have more strongly misregulated ECM. These results are among the first (Shao et al., 2020) to implicate signaling pathways in the development of Lmna-DCM, and the identification of the pathways, particularly those in disease progression, will allow us to determine whether any of the genes contributing to misregulated signaling are already beginning to show altered expression in the pre-phenotype mice. Lmna KO N195K mut over time eosinophil chemotaxis fatty acid beta-oxidation Muscle-related regulation of slow-twitch skeletal muscle fiber contraction supramolecular fiber organization Metabolism CCR2 chemokine receptor binding protein-containing complex ECM/cell-cell interaction monocyte chemotaxis collagen-containing extracellular matrix twitch skeletal muscle contraction BMP receptor activity Immune system transition between fast and slow fiber oxidoreductase activity DNA damage regulation of skeletal muscle contraction muscle organ morphogenesis neutrophil chemotaxis lipid oxidation positive regulation of calcium ion import cardiac myofibril skeletal muscle adaptation acetyl-CoA C-acyltransferase activity muscle hypertrophy in response to stress oxidoreductase complex granulocyte migration monocarboxylic acid catabolic process regulation of muscle contraction banded collagen fibril muscle system process positive regulation of mRNA 3'-end processing extracellular space cellular component assembly involved in morphogenesis 4wk N195K 10wk N195K IgE binding extracellular space homeostasis of number of cells complex of collagen trimers serotonin transport extracellular region histone demethylation extracellular matrix structural constituent conferring tensile strength protein dealkylation cardiovascular system development import into cell anatomical structure formation involved in morphogenesis negative regulation of cellular component organization collagen fibril organization granulocyte activation integrin binding nucleobase metabolic process regulation of cellular component movement DNA repair complex epithelium migration lymphocyte activation involved in immune response collagen-containing extracellular matrix demethylase activity regulation of cell migration phagocytosis cellular component organization or biogenesis regulation of leukocyte degranulation positive regulation of cell motility enzyme binding regulation of locomotion Fig. 3.7. Top Gene Ontology (GO) terms for each comparison. Most GO terms follow the same trends in broader signaling pathways of cardiac muscle-related (red), metabolism (blue), ECM/cell-cell interactions (green), immune system (yellow), and here we observed a first mention of DNA damage (purple). 107 RNA-seq of human iPSC and iPSC-CMs with LMNA mutations show only weak clustering With these results in mind from the mouse Lmna-DCM RNA-seq experiments, we anticipated that similar disrupted signaling pathways would be observed in cardiomyocytes derived from human patients with LMNA-DCM mutations, which only have heterozygous expression of the mutant allele. To answer this question, I performed RNA-seq with iPSCs and iPSC-CMs derived from two healthy controls and two patients with LMNA mutations, R249Q and G449V, to find common and differing gene expression between the two models, and in hopes to mimic disease progression, with iPSCs modeling pre-phenotype conditions. Upon sequencing and building a DEseq2 model that takes into account the genotype (i.e. healthy control, or WT, vs mutant), cell line, differentiation state (i.e. iPSC or iPSC-CM), the interaction of cell line and genotype, and the interaction of cell line and differentiation state, I performed PCA analysis and generated corresponding PCA plots to determine whether iPSC and iPSC-CMs cluster by cell line, presence of an LMNA mutation, and differentiation stage (Fig. 3.8). A PCA plot with all samples and differentiation conditions showed a large separation between the iPSC and iPSC- CM samples along the PC1 axis, regardless of genotype (Fig. 3.8A), as expected due to the large transcriptomic differences that occur during cardiac differentiation of iPSCs (Rajala et al., 2011). Since the large separation between iPSCs and iPSC-CMs makes it challenging to look more specifically at samples within those groups, I generated separate PCA plots for iPSCs and iPSC- CMs (Fig. 3.8B and C, respectively). Interestingly, a PCA plot of only the iPSC samples (Fig. 3.8B) showed most of the WT samples clustered and some clustering of the LMNA-mutant samples, suggesting that there could already be some signaling differences before cardiac differentiation. On the other hand, a PCA plot of the iPSC-CM samples (Fig. 3.8C) showed less 108 clustering, with the G449V samples clustering more closely on the PC1 axis, and both healthy control and R249Q samples being mixed together and spread out along both the PC1 and PC2 axes. This not only suggests that even within a single cell line, iPSC-CMs have inherent variability between samples, potentially due to variations in differentiation, but also suggests that R249Q mutant iPSC-CMs have less transcriptomic differences than healthy control iPSC-CMs compared to the differences observed with G449V mutant iPSC-CMs. It is challenging to discern whether the lack of R249Q mutant transcriptomic differences are due to less disrupted gene expression compared to G449V mutant cells, or whether these differences result from heterogeneous backgrounds among the cell lines that naturally give some cell lines more similarity than others. A B C Fig. 3.8. PCA plots of iPSC and iPSC-CM samples. (A) The PCA plot of all samples showed significant variance between the iPSC and iPSC-CM samples, as expected with cardiac differentiation, but also suggests that there may be some differences between WT (BMWT1 and WTC11) and LMNA mutant iPSC samples (R249Q and G449V), with the variance between these samples being almost as large as the iPSC-CM samples on PC2. (B) The PCA plot of only the iPSC samples confirmed that most of the WT samples cluster closely, and there is a large separation from the roughly clustered LMNA mutant samples. (C) However, iPSC-CM samples clustered a bit less well, with the G449V samples closely clustering, and the other three cell lines being much more spread out and not clustering together. 109 To better understand the lack of clustering I observed with the iPSC-CM PCA plot, I created a heatmap of significant DEGs with a p<0.05 to further examine sample clustering and variation in gene expression (Fig. 3.9). Indeed, one sample from each healthy control line did not cluster with the other three samples of that cell line, instead clustering with the R249Q mutant samples. Moreover, the heatmap showed that not only was gene expression variable between samples for each of R249Q and the two healthy control cell lines, but the G449V mutant iPSC-CMs had a more distinct signature of gene expression compared to the other three cell lines. Therefore, I proceed with only comparing healthy control and G449V mutant iPSC-CMs, to determine the Fig. 3.9. A heatmap of iPSC-CM LMNA healthy control (WT) vs LMNA-mutant (mut) DEGs with a p<0.05. A heatmap shows that one sample each of WT1 (B) and WT2 (W) samples do not fully cluster to the rest of their genotype. 110 transcriptomic differences arising after cardiac differentiation from the LMNA-DCM phenotype. G449V-iPSC-CMs exhibit disrupted signaling pathways similar to the mouse RNA-seq study To identify G449V significant DEGs compared to both healthy control lines, I create a statistical model for each comparison, and identified significant DEGs that had the same directional fold change in each comparison. In total, this identified 358 significant DEGs. I created a protein interaction map (Fig. 3.10) to identify clusters of DEGs that may be driving misregulated signaling pathways. Several clusters of DEGs emerged with high-level pathway themes similar to the mouse RNA-seq results: ECM/cell-cell interactions, cardiac genes, and metabolism. I also identified one additional, novel cluster of transcription factor genes in the G449V protein interaction map, including one FOX transcription factor, a group that has been identified by another LMNA-DCM RNA-seq study (Auguste et al., 2018). Fig. 3.10. Protein interaction map for G449V significant DEGs. Each dot is a protein-coding gene, with size denoting expression levels and halos denoting p-value. Lines show interactions between protein-coding genes. Gene clusters representing common pathways are circled and labeled. 111 To understand how gene clusters relate more specifically to disrupted signaling pathways, I performed KEGG pathway analysis on the G449V significant DEGs (Fig. 3.11). KEGG analysis identified 19 significant signaling pathways, 14 of which corresponded to the four high-level pathways identified by the mouse RNA-seq study (muscle-related or cardiac genes, metabolism, ECM/cell-cell interaction, and immune system; colored in Fig. 3.11) and 8 of which were also identified by the mouse RNA-seq study (italicized in Fig. 3.11). Notably, several of the metabolism pathways identified here were also key pathways identified in the mouse RNA-seq study: AGE- RAGE signaling, PI3K-AKT signaling, and carbon metabolism. These results are particularly exciting as they not only represent misregulated signaling pathways important in both mouse and human LMNA-DCM, but as iPSC-CMs tend to be immature in structure and metabolic function, resembling fetal cardiomyocytes more closely than adult cardiomyocytes (Scuderi and Butcher, 2017), these pathways could be indicative of metabolic signaling pathways disrupted in early cardiac development, before the complete onset of LMNA-DCM. Fig. 3.11. Significant KEGG pathways altered in G449V-iPSC-CMs. 19 significant KEGG pathways were identified, 14 of which corresponded to the four major general pathways identified by the mouse RNA-seq study (muscle-related in red, metabolism in blue, ECM/cell- cell interaction in green, or immune system in yellow) and 8 of which were specific KEGG pathways also identified by the mouse RNA-seq study (italicized). 112 Twenty DEGs are shared between mouse and G449V DCM models To more directly compare the results of mouse Lmna-DCM and G449V mutant iPSC-CM RNA- seq studies, I generated a list of DEGs from all of the comparisons from the mouse left ventricle study and found the genes common with the G449V mutant iPSC-CM study. In total, 28 genes were identified (Table 3.2), including several ECM and collagen genes and sarcomere genes. Since this was a relatively small list of DEGs, rather than doing GO Term and KEGG pathway analysis separately, I performed DAVID clustering analysis (Huang et al., 2009a, 2009b) of GO Terms, KEGG pathways, and UP_Keywords. This analysis identified four primary clusters (Fig. 3.12), three of which were pertaining to collagen and the ECM, and the last pertaining to immune system response. As particularly the ECM has been a common theme through all of these analyses, even through different stages of disease phenotype (Lmna N195K mutant over time and 10 weeks) and cardiac maturity (with less mature iPSC-CMs) this represents a potentially interesting mechanism to explore in future studies. 113 Table 3.2 Significant DEGs by both mouse left ventricle and G449V-iPSC-CM RNA-seq Gene Symbol Gene Name ACTB actin beta ACTN1 actinin alpha 1 CCN2 cellular communication network factor 2 COL1A1 collagen type I alpha 1 chain COL3A1 collagen type III alpha 1 chain COL5A2 collagen type V alpha 2 chain DCN decorin EGFLAM EGF like, fibronectin type III and laminin G domains ENDOD1 endonuclease domain containing 1 FBLN5 fibulin 5 FN1 fibronectin 1 IKBIP IKBKB interacting protein KCNK6 potassium two pore domain channel subfamily K member 6 KDELR2 KDEL endoplasmic reticulum protein retention receptor 2 LUM lumican MFAP5 microfibril associated protein 5 MMP14 matrix metallopeptidase 14 PXDN peroxidasin S100A10 S100 calcium binding protein A10 S100A11 S100 calcium binding protein A11 S100A4 S100 calcium binding protein A4 SERPINE2 serpin family E member 2 SOX9 SRY-box transcription factor 9 SPARC secreted protein acidic and cysteine rich TIMP1 TIMP metallopeptidase inhibitor 1 TMEM176B transmembrane protein 176B TNNI1 trponin I1, slow skeletal type TNNT2 troponin T2, cardiac type 114 Fig. 3.12. DAVID clustering analysis identifies pathways and GO terms from DEGs common to mouse left ventricle and G499V mutant iPSC-CM studies. Of the four clusters identified by DAVID pathway analysis, the major three were related to collagen and the extracellular matrix (ECM), which the last was pertaining to immune response. Discussion and conclusion Altered gene expression and cell signaling has long been recognized to be a driver of LMNA-DCM pathogenesis (Davidson and Lammerding, 2014; Maurer and Lammerding, 2019; Donnaloja et al., 2020). Here, I report that through RNA-seq analysis of two mouse models of Lmna-DCM, both pre- and post-phenotype development, and human iPSC-CMs with an LMNA mutation, G449V, I identified several high-level misregulated signaling pathways and DEGs that not only contribute to LMNA-DCM disease phenotype but also contribute to the development of disease. 115 RNA-seq of Lmna KO and N195K mouse models of Lmna-DCM identified four notable high- level signaling pathways: cardiac function, metabolism, extracellular matrix, and the immune system. While several of these pathways, particularly cardiac function and the immune system, may be more related to decline of cardiac function and health and overall animal health due to Lmna-DCM, as they are particularly evident in the post-phenotype Lmna KO model, both metabolism and the extracellular matrix represent two exciting biological pathways that are becoming increasingly evident to play a role in LMNA-DCM (Pauschinger et al., 1999; Cai et al., 2020; Pradas et al., 2020; Shao et al., 2020; Mehrabi et al., 2021) but remain incompletely understood in the context of disease progression and decline in health. Through the study of Lmna- DCM altered gene expression in pre- (4 week) and post-phenotype (10 week) Lmna N195K mice and by generating a statistical model representing the effect phenotype development over 4- to 10- week aging and disease development period period, I have uncovered a novel perspective on metabolism and ECM misregulation through disease progression. At pre-phenotype, N195K mice showed few distinct patterns of disrupted gene expression and related signaling. However, as Lmna-DCM onsets and progresses, metabolism becomes strongly misregulated and some changes to ECM and muscle function begin to occur. Finally, by full phenotype development, N195K mice have more strongly misregulated ECM. Through the study of Lmna N195K mutant gene expression changes attributed to phenotype development, I identified several clusters of DEGs pertaining to various forms of metabolism, notably fatty acid, carbon, and purine, and oxidative stress, and several relevant genes (Eno3, Etfb, and Hfe1) that connect regulation of metabolism and oxidative phosphorylation to cardiac function. These genes represent exciting new connections between disrupted metabolism in 116 decline of cardiac function in Lmna-DCM and should be further investigated to establish their precise role in either phenotype development or progression. Additionally, gene expression changes due to phenotype development identifed several notable metabolic signaling pathways involved in the development and progression of Lmna-DCM: HIF-1, AGE-RAGE, PI3K-AKT, and thyroid hormone signaling pathways. Through comparison of one LMNA-mutant human iPSC-CM line, G449V, to several healthy control lines, I successfully identified 358 significant DEGs, 28 of which are shared with mouse models of Lmna-DCM. G449V mutant pathway analysis mirrored many of the trends observed in mouse models, with ECM and metabolism misregulation standing out, and some specific biological signaling pathways standing out, particularly the AGE-RAGE pathway, the PI3K-AKT pathway, and carbon metabolism and regulation of collagen genes. While it was not within the scope of this study to delve further into these specific mechanisms, future studies should clarify the mechanisms through which these pathways become dysregulated. However, several limitations of this study remain. Notably, the iPSC-CM cell lines used in this study are from a heterogeneous genetic background, which could potentially conflate gene expression changes attributed to differing genetic backgrounds with those from the LMNA mutation and LMNA-DCM phenotype. Although we designed our study to reduce this possibility by finding G449V-iPSC-CMs DEGs that were significant compared separately to two healthy control cell lines, we cannot rule out the possibility that the healthy control cell lines are in a more similar genetic background compared to the G449V, resulting in some significant DEGs being due to differing genetic backgrounds. Additionally, while I have identified misregulated genes and 117 pathways across several mouse models and one LMNA-DCM iPSC-CM line, we have no direct match of mutations or cell types to compare the basal gene expression differences between the relatively immature iPSC-CMs and mouse left ventricle tissue. Particularly since mouse RNA-seq studies were carried out using bulk left ventricle tissue, of which the vast majority is comprised of cardiomyocytes, numerous other cell types reside in cardiac tissue, and thus without a direct match of LMNA mutations, we have no way to compare which signaling pathways could be attributed to other cell types or immature iPSC-CMs. Although this study has made significant progress towards identifying novel biological pathways driving LMNA-DCM phenotype and progression, it remains incomplete. Particularly the human iPSC-CM study has room to expand to many unanswered questions: are there already disrupted signaling pathways in G449V at the iPSC stage of development? If so, are these genes similar to those expressed by pre-phenotype Lmna N195K mice? To what degree is gene expression altered in the LMNA R249Q mutant cell line? Are there genes common to both the G449V and R249Q mutations, and potentially even to mouse models of Lmna-DCM? Why do we observe a more drastic difference in gene expression between G449V mutant iPSC-CMs and healthy controls compared to the R249Q mutation, which showed more severe nuclear damage (Chapter 2)? To what degree can the G449V mutation gene expression differences be explained by differing genetic backgrounds, or can it be attributed to the LMNA-DCM phenotype? To answer these questions, this project will be continued by a postdoctoral scholar in the Lammerding Lab, Julien Morival. 118 Supplementary Materials Table S3.1. Common significant DEGs among Lmna KO and N195K comparisons. Gene Symbol Gene Name Group Dio2 deiodinase, iodothyronine, type II Lmna KO, 4wk and 10wk N195K Lgals3 lectin, galactose binding, soluble 3 Lmna KO and 4wk N195K Gpx3 glutathione peroxidase 3 Lmna KO and 10wk N195K Ifi27l2a interferon, alpha-inducible protein 27 like 2A Lmna KO and 10wk N195K Lyz2 lysozyme 2 Lmna KO and 10wk N195K Pitpnc1 phosphatidylinositol transfer protein, Lmna KO and 10wk N195K cytoplasmic 1 Rtn4 reticulon 4 Lmna KO and 10wk N195K Serpina3n serine (or cysteine) peptidase inhibitor, clade Lmna KO and 10wk N195K A, member 3N Cpt2 carnitine palmitoyltransferase 2 Lmna KO and N195K mut over time Hjv hemojuvelin BMP co-receptor Lmna KO and N195K mut over time Hrc histidine rich calcium binding protein Lmna KO and N195K mut over time Mgp matrix Gla protein Lmna KO and N195K mut over time Armc5 armadillo repeat containing 5 4wk and mut over time N195K Cpeb1 cytoplasmic polyadenylation element binding 4wk and mut over time N195K protein 1 Ctps2 cytidine 5'-triphosphate synthase 2 4wk and mut over time N195K Dbnl drebrin-like 4wk and mut over time N195K Dusp3 dual specificity phosphatase 3 (vaccinia virus 4wk and mut over time N195K phosphatase VH1-related) Gipc1 GIPC PDZ domain containing family, member 4wk and mut over time N195K 1 Kdm1a lysine (K)-specific demethylase 1A 4wk and mut over time N195K Kdm6a lysine (K)-specific demethylase 6A 4wk and mut over time N195K Krt10 keratin 10 4wk and mut over time N195K Pear1 platelet endothelial aggregation receptor 1 4wk and mut over time N195K Ppp1r14b protein phosphatase 1, regulatory inhibitor 4wk and mut over time N195K subunit 14B Ppp1r18 protein phosphatase 1, regulatory subunit 18 4wk and mut over time N195K Snord13 small nucleolar RNA, C/D box 13 4wk and mut over time N195K Zfpl1 zinc finger like protein 1 4wk and mut over time N195K Zscan21 zinc finger and SCAN domain containing 21 4wk and mut over time N195K Col1a1 collagen, type I, alpha 1 10wk and mut over time N195K Col3a1 collagen, type III, alpha 1 10wk and mut over time N195K Cox10 heme A:farnesyltransferase cytochrome c 10wk and mut over time N195K oxidase assembly factor 10 Dcn decorin 10wk and mut over time N195K Eno3 enolase 3, beta muscle 10wk and mut over time N195K Gas5 growth arrest specific 5 10wk and mut over time N195K Mrpl45 mitochondrial ribosomal protein L45 10wk and mut over time N195K 119 Ndufa10 NADH:ubiquinone oxidoreductase subunit A10 10wk and mut over time N195K Prkar1a protein kinase, cAMP dependent regulatory, 10wk and mut over time N195K type I, alpha Txndc17 thioredoxin domain containing 17 10wk and mut over time N195K Wscd1 WSC domain containing 1 10wk and mut over time N195K Ccl8 chemokine (C-C motif) ligand 8 4wk and 10wk N195K Lilrb4a leukocyte immunoglobulin-like receptor, 4wk and 10wk N195K subfamily B, member 4A Fig. S3.1. Protein interaction map for the N195K pre-phenotype comparison. Each dot is a protein-coding gene, with size denoting expression levels and halos denoting p-value. Lines show interactions between protein-coding genes. Very few connections between genes are observed. 120 Fig. S3.2. Protein interaction map for the N195K post-phenotype comparison. Each dot is a protein-coding gene, with size denoting expression levels and halos denoting p-value. Lines show interactions between protein-coding genes. One primary cluster is shown containing ECM genes. 121 Enzyme/immune genes Cardiac genes Fig. S3.3. Protein interaction map for the Lmna KO post-phenotype comparison. Each dot is a protein-coding gene, with size denoting expression levels and halos denoting p-value. Lines show interactions between protein-coding genes. Few genes clusters, forming one broad cluster with several enzyme-encoding or immune-related genes, and one with several cardiac genes. 122 CHAPTER 4 The Lamin A/C Ig-fold undergoes cell density-dependent conformational changes that alter epitope accessibility3 Nuclear Lamins A/C are intermediate filament proteins that impart mechanical strength to the nucleus and are involved in diverse mechanical and biochemical signaling pathways. The Lamin A/C Ig-fold is the site for a large portion of Lamin A/C’s binding partners, thus making it critical for nuclear organization, and is a hotspot for LMNA mutations causing several different diseases, termed ‘laminopathies.’ Given the importance of the Lamin A/C Ig-fold in nuclear organization and its high correlation with human disease, it is critical to understand the Ig-fold’s mechanical behavior and how it can govern cellular health and disease. Here, we report that the Lamin A/C- Ig fold undergoes mechanosensitive conformational changes in response to cell seeding density, that allows for increased binding of a Lamin A/C antibody at low cell densities. This density- dependent epitope binding is independent of protein expression and changes in nuclear mechanics. We hypothesize that the C’E and/or EF loops of the Ig-fold partially unfold in response to cell spread area, due to changes in forces from actin and microtubule on the nucleus. These results 3 This project is a continuation of preliminary work complete by a former Biomedical Engineering PhD student, Greg Fedorchak (GF), and a former Biomedical Engineering Master’s Student, Rachel Gilbert (GF), in Jan Lammerding’s lab, and will be submitted to Frontiers in Cell and Developmental Biology’s special issue 3D Architecture of Intermediate Filaments in Tissue Mechanics and Function by the end of the 2021. Maurer, Melanie, Fedorchak, Greg, Gilbert, Rachel, Patel, Jineet, Lammerding, Jan. MM, GF, RG, and JL contributed to the conception and design of the work. MM, GF, RG, and JP contributed to data acquisition and analysis. MM, GF, RG, and JL contributed to interpretation of the data. MM and JL contributed to the drafting of the manuscript. 123 suggest that it is critical that studies using the Lamin A/C-Ig1 and other Lamin A/C Ig-fold antibodies be aware of its sensitivity to cellular conformation and may be a potential link between disrupted mechanotransduction in LMNA mutations affecting the Ig-fold. Introduction One way that cells respond to their mechanical environment is through the coupling of the extracellular matrix and cytoskeleton to the nucleus and the corresponding nuclear response to mechanical stress (Wang et al., 2009). The mechanical properties of the nucleus are governed by nuclear lamins and chromatin (Stephens et al., 2017), which respond to matrix elasticity and external forces by modulating the expression and conformation of nuclear lamin proteins (Pajerowski et al., 2007; Swift et al., 2013; Buxboim et al., 2014; Ihalainen et al., 2015; Iyer et al., 2021) and changes to chromatin modifications (Furusawa et al., 2015; Stephens et al., 2018b; Nava et al., 2020). Lamin A/C expression scales with tissue stiffness and force application (Swift et al., 2013; Buxboim et al., 2014; Iyer et al., 2021), but despite evidence that the nuclear lamina can stretch in response to mechanical force (Rowat et al., 2006; Pajerowski et al., 2007; Swift et al., 2013), we still have relatively little understanding of how direct mechanical stimulus affects lamin conformation. The Lamin A/C immunoglobulin-like (Ig-) fold responds to various forms of mechanical stimuli through conformational changes that allow for the exposure of a cryptic cysteine reside, Cys522, in response to shear stress (Swift et al., 2013) and the selective binding of antibodies in response to actin-mediated cell spreading or force application to the nucleus (Ihalainen et al., 2015). The Lamin A/C Ig-fold plays an important role in the structural organization and assembly of Lamin 124 A/C (Verstraeten et al., 2009; Ahn et al., 2021) and houses the binding sites of countless Lamin A/C binding partners (Donnaloja et al., 2020). Many Ig-fold binding partners that are important for nuclear organization and function in nucleo-cytoskeletal coupling, chromatin organization, and regulation of gene expression coincide with the particular region subject to mechanical stretch (Swift et al., 2013; Ihalainen et al., 2015), including actin, DNA, emerin, LAP2α, and SUN1/2 (Donnaloja et al., 2020). Perhaps even more intriguingly, the Lamin A/C Ig-fold is a hotspot for LMNA mutations that cause several “laminopathy” diseases (Scharner et al., 2014). These mutations may cause structural changes to the Ig-fold, in combination with mechanosensitive conformational changes. Therefore, it is imperative to understand the nature of Lamin A/C mechanosensitive conformational changes to better understand how they can impact nuclear mechanics, organization, and its implication for laminopathy diseases. Here, we report that the Lamin A/C Ig-fold undergoes conformational changes in response to cell seeding density, as evidenced by the cell density-dependent binding of a commonly used Lamin A/C antibody that binds the Ig-fold (JOL-2, referred to here as Lamin A/C-Ig1). Density-dependent epitope binding is independent of protein expression. While we found no changes in nuclear mechanics with cell seeding density, we observed that Lamin A/C-Ig1 density-dependent binding is sensitive to the reduction of cytoskeletal forces exerted on the nucleus, although this effect could not fully explain the density-dependent effects of the Lamin A/C-Ig1 labeling. We hypothesize that the Lamin A/C-Ig1 antibody binds the C’E loop of the Ig-fold, which partially unfolds in response to tension on the nucleus, which is in line with previous results showing that unfolding of the C’E loop (Swift et al., 2013). These results add to previous work (Swift et al., 2013; Ihalainen et al., 2015) providing evidence for that the Lamin A/C Ig-fold is subject to 125 mechanosensitive conformational changes, and provides the first evidence that Ig-fold mechanosensitive conformational changes may be governed by forces from microtubules exerted on the nucleus. Materials and Methods Cell culture Immortalized human fibroblast, HeLa, and MDA-MB-231 cells were cultured in tissue culture- treated flasks in Dulbecco's Modified Eagle Medium (DMEM; Gibco) plus 10% Fetal Bovine Serum (FBS) and 1% Penicillin-Streptomycin (P/S), referred to here as D-10 medium. For imaging and microharpoon experiments, cells were dissociated as single cells using 0.25% Trypsin and resuspended in D-10 medium. Cells were then counted and seeded in a serial dilution, according to Table 4.1 for imaging experiments, on Fibronectin-coated (1µg/mL) sterile coverslips to achieve a range from low seeding densities in which there are rare cell-cell contacts to very confluent seeding densities. Cells were allowed to adhere to coverslips or plates for 24 hours prior to experimentation. Alternatively, cells were seeded at the same density and collected at different time points to obtain the desired density. Table 4.1. Cell seeding density for human fibroblast and HeLa immunofluorescence experiments. Cell seeding densities ranged from fully confluent with numerous cell-cell contacts to low seeding densities with extremely low confluency and rare cell-cell contacts. Dilution # of Cells Cell Density (#/cm2) % Confluency 1.5× 60,000 31,579 100% 1.0× 40,000 21,053 80% 0.5× 20,000 10,526 40% 0.25× 10,000 5,263 20% 0.125× 5,000 2,632 10% 0.0625× 2,500 1,316 5% 126 CRISPR knock-in generation CRISPR knock-in cells were generated by Rachel Gilbert, a past Biomedical Engineering Masters student in the Lammerding lab. Lamin A/C ablation was achieved using lentiCRISPR v2 directed at exon 1 site 2 (spacer sequence: CATCGACCGTGTGCGCTCGC). Gibson assembly was used to generate the repair plasmid within a DTA killer gene backbone (PGKdta bpa) and consisted of an mNeonGreen fluorophore cloned from a pCDH mNeonGreen Prelamin A plasmid. The use of site directed mutagenesis to induce silent mutations in the repair construct prevented secondary cutting by CRISPR. MDA-MB-231 cells were plated in a 6-well plate and hit with recombinant adeno-associated virus (rAAV) (Addgene, pAAV TBG FFLuc) supernatant and allowed to incubate for 48 hours. Cells were split to reduce the density and then transiently transfected with the lentiCRISPR v2 E1E2 plasmid consisting of an insertion with Blasti-P2A-mNeonGreen. Cells were placed under high selection with blasticidin and subsequently sorted to enhance the population with strong expression. Both heterogenous and clonal populations were generated and used in experiments. Cell sorting and clonal selection rAAV-modified MDA-MB-231 cells were sorted for mNeonGreen signal using the Cornell Imaging Facility FACS. Non-rAAV-modified cells were used as controls. After two rounds of sorting, mNeonGreen-positive cells were enriched from <10% to almost 100%. While some populations were left as heterogeneous, others were sorted based on intensity (high and low expression) in order to test distinct subpopulations. For clonal selection, cells were seeded at very low density in 10 cm2 dishes and grown for several weeks. Clonal rings and vacuum grease were 127 used to select individual colonies for expansion. DNA isolation and PCR gel DNAzol Reagent (Invitrogen) was added to cells seeded in a 6-well plate and genomic DNA was isolated according to the manufacturer’s instructions. The following primer pairs were used to evaluate proper insertion of the mNG tag into the correct location in the genome: mNeonGreen_seq_F = 5’ CCCAACGACAAAACCATCAT 3’, LMNA downstream_R = 5’ TGCAGTTGAGTAGGGTGGG3’, LMNA upstream_F = 5’AACTCCTTGATCCCTGGCC3’ and LHA (left homology arm) extract_R = 5’ACCGCCAAGCGATCAT 3’. Amplicons were amplified using Phusion DNA Polymerase (NEB). PCR products were run on an agarose gel using standard lab protocols and imaged. Immunoblotting Cells were either lysed in equal cell numbers in Laemmli sample buffer (Bio-Rad) containing 0.1 M of Dithiothreitol (DTT) after trypsinization or in RIPA buffer supplemented with proteinase inhibitor (Roche). Protein content was measured in the RIPA cell lysates using a standard Bradford assay. All samples were heat-denatured (5 min at 95°C) in Laemmli sample buffer and separated by SDS-PAGE (Invitrogen). Proteins were transferred onto PVDF membranes (Millipore, IPVH00010) using semi-dry transfer method (Bio-Rad). Immuno-detection was carried out with the following primary antibodies: anti-lamin A/C (Santa Cruz, sc-6215, dilution 1:2000), anti-- tubulin (Sigma, T5168, dilution 1:4000). Blots were probed with HRP-conjugated antibodies (Biorad, dilution 1:1000; Jackson Immuno Research, dilution: 1:10000). Activation and inhibition of retinoic acid pathway 128 Cultured cells were treated with all-trans retinoic acid (RA, 1 µM, Fisher Scientific), high affinity pan-RAR antagonist AGN-193109 (AGN, 1 µM, Santa Cruz Biotechnology), or vehicle control (0.15% EtOH, 0.15% DMSO in media with 10% FBS) for 48 hours. Pharmaceutical inhibitor studies Human fibroblasts were seeded, at high (10,000 cells per well) and low (1,000 cells/well) density, into 96-well plates, allowed to adhere, and treated for 24 hours (unless otherwise noted) prior to fixation. The following treatments were used: a histone deacetylase inhibitor, Trichostatin A (TSA, 200 nM), a microtubule depolymerizing agent, Nocodazole (75 nM), or an actin depolymerizing agent, Cytochalasin D (CytoD, 10 µM, 60 minutes prior to fixation). A separate human fibroblast line stably modified to express a dominant-negative KASH (DN-KASH) was treated with 500ng/mL Doxycycline for 24 hours prior to fixation to disrupt nucleo-cytoskeletal coupling. Immunofluorescence staining Cells were passaged as previously described onto Fibronectin-coated coverslips. The next day, media was either changed to fresh D-10 or D-10 treated with drugs, and then cells were washed with 1× PBS and fixed in warmed Paraformaldehyde (2% or 4%) for 10 minutes or on ice in cold 1:1 Methanol and Acetone for 15 minutes. Cells were then washed with 1× PBS and blocked in 3% BSA with 0.1% Triton-X 100 (Thermo-Fischer) and 0.1% Tween (Sigma) in PBS for one hour at room temperature. Primary antibodies were prepared in blocking solution and incubated overnight at 4°C. Cells were then washed with a solution of 0.3% BSA with 0.1% Triton-X 100 and 0.1% Tween in PBS and stained with AlexaFluor secondary antibodies (1:250; Invitrogen) for 1 hour at room temperature. DAPI (1:1000, Sigma) was added for 10 minutes at room temperature, 129 and cells were washed with PBS before imaging. Image acquisition and analysis Flasks and coverslips in cell culture plates were imaged at 20× magnification (NA=0.4) on an inverted Zeiss Observer Z1 epifluorescence microscope. Mean fluorescence intensity and area of each nucleus was obtained using a custom MATLAB (Mathworks) script to threshold nuclei. Similarly, actin fluorescence intensity was computed by dividing the total actin fluorescence intensity in an image by the number of nuclei in the image. For the majority of experiments, fluorescence intensity values were normalized for each experiment to account for variability in immunofluorescence experiments and image acquisition. Cells on coverslips were imaged as stacks on a Zeiss LSM700 confocal microscope with 63× magnification (NA=1.4) to quantify the apico-basal distribution of Lamin A/C and nuclear height. Airy units were set to 1.0 for all images. Apico-basal polarization analysis was complete with a custom MATLAB (Mathworks) script in which 63× magnification image stacks of cells labeled for Lamin A/C were thresholded, and the fluorescent intensity in each image slice was computed at the centroid of each nucleus. A measure of apico-basal polarization was computed by taking the average centroid intensity of the bottom half of image slices over the average centroid intensity of top half of image slices. Biophysical assays For micropipette aspiration experiments, a confluent T75 flask was split 1:12 (~87,500 cells) into a T25, T75 and T150 flask 48 hours prior to experimentation to establish three different densities (low, medium, high). Micropipette aspiration was performed on human fibroblasts according to a 130 previously described protocol (M. Davidson et al., 2019). A pressure differential of 1.0 psi and 0.2 psi at the inlet and outlet reservoirs drove the flow of cells through the device. Images were acquired every 5 seconds using an inverted microscope for a minimum of 60 seconds. Nuclear protrusion length was calculated using a custom MATLAB script. For microharpoon studies, human fibroblasts were seeded 24 hours prior to experimentation. The microharpoon experiment was performed as previously described (Fedorchak and Lammerding, 2016). The microharpoon was inserted ≈5 μm from the edge of the nucleus and pulled 16 μm at 4 μm/s. Images were acquired at 40× magnification (NA=0.75) every 5 seconds. Average nuclear strain and centroid displacement were calculated using a custom MATLAB script. Statistics All results were taken from two to three independent experiments. For numeric datasets in which whole images or experiments were analyzed and followed a normal distribution, either a Student’s t-test (for two groups) or one-way analysis of variance (ANOVA) with multiple comparisons (for more than two groups) was performed in GraphPad Prism. Tukey’s correction for multiple comparisons was used. Results Lamin A/C epitope immunolabeling changes in a density-dependent manner The Lamin A/C Ig-fold is known to undergo mechanosensitive conformational changes in response to mechanical stimuli (Swift et al., 2013; Ihalainen et al., 2015), which may be indicated by differential binding of a Lamin A/C Ig-fold epitope (Ihalainen et al., 2015). We hypothesized 131 that variations in cell seeding density that correspond to changes in cell spreading and cell-cell contacts could alter nuclear tension and result in mechanosensitive conformational changes. To test whether cell seeding density could induce mechanosensitive conformational changes, we seeded human fibroblasts in a serial dilution such that the lowest cell seeding density (0.0625× or 1,316 cells/cm2) corresponded to low confluency, increased cell spread area, and rare cell-cell contacts, and our highest seeding density (1.5× or 31,579 cells/cm2) corresponded to nearly 100% confluency, lower spread area, and many cell-cell contacts (Fig. 4.1A,C). We performed immunofluorescence labeling for Lamin A/C across six cell seeding densities in immortalized human fibroblasts using the JOL-2 antibody (LAC-Ig1; Fig. 4.1B) and observed a striking difference in Lamin A/C-Ig1 fluorescence intensity across cell seeding densities: Lamin A/C-Ig1 fluorescence intensity normalized to the highest seeding density, 1.5×, varied inversely with cell seeding density, and cells seeded at lower densities had significantly increased Lamin A/C-Ig1 fluorescence intensity compared to cells seeded at higher densities (Fig. 4.1D). However, the two lowest seeding densities, 0.0625× and 0.125×, exhibited similarly increased Lamin A/C-Ig1 fluorescence intensity. This could potentially be because at the lowest two seeding densities, cells are already very spread apart and have little cell-cell contacts, thus resulting in more similar cell spreading areas and/or cytoskeletal conformations. To confirm that density-dependent immunolabeling with the Lamin A/C-Ig1 antibody was not a product of fixation method, we repeated the experiment with two alternative fixation methods, 2% PFA or 1:1 Methanol:Acetone. We observed the same Lamin A/C-Ig1 significant inverse correlation with both methods, in which low cell seeding densities had significantly higher fluorescence intensity compared to high seeding densities (Fig. S4.1). Finally, to determine this trend could be observed in Lamin B1 as well as 132 Lamin A/C, we quantified Lamin B1 fluorescence intensity across the same cell seeding densities. However, Lamin B1 showed no density-dependent changes in fluorescence intensity (Fig. 4.1A,D). Fig. 4.1. Lamin A/C epitope expression changes in a density-dependent manner in human fibroblasts. (A) Representative images of human fibroblasts labeled for actin seeded in a serial dilution showing that at low cell seeding densities, cells are at a low confluency with rare cell- cell contacts, and at the highest cell seeding densities, cells are in a in a nearly confluent monolayer with numerous cell-cell contacts. (B) Representative images of human fibroblasts seeded roughly in a serial dilution and stained for Lamin A/C with the JOL-2 antibody (LAC- Ig1) and for Lamin B1. (B) Average number of human fibroblast nuclei per image shows an exponential trend, as expected. (C) Lamin A/C JOL-2 fluorescence intensity of human fibroblasts normalized to the highest cell seeding density, 1.5×, varies inversely with cell seeding density. (D) Lamin B1 fluorescence intensity normalized to the highest cell seeding density, 1.5×, does not change with cell seeding density. Scalebars = 200 μm. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. N = 30 images per group for number of nuclei. N > 100 nuclei per group for fluorescent intensity experiments. 133 To test whether density-dependent immunolabeling with Lamin A/C-Ig-1 was conserved in other human cell types, we repeated the experiment with HeLa cells, an immortalized epithelial adenocarcinoma cell line. At similar seeding densities in which number of cells follows an exponential trend (Fig. S4.2A), we observed a similar inverse relationship between Lamin A/C fluorescence intensity and cell seeding density in HeLa cells, with low densities having significantly increased fluorescence intensity compared to high seeding densities (Fig. S4.2B). Similar to previous experiments in human fibroblasts, the two lowest cell seeding densities had similar Lamin A/C-Ig1 fluorescence intensity, and we did not observe any density-dependent differences in Lamin B1 fluorescence intensity (Fig. S4.2C). To determine whether Lamin A/C-Ig1 density-dependent immunolabeling was related to cell cycle stage, as all cells in previous experiments were seeded simultaneously, we tried an additional seeding method with an additional cell line, MDA-MB-231, a triple-negative epithelial adenocarcinoma cell line, to allow for growth up to certain densities over several days, thereby varying the number of doublings cells have undergone before fixation. We seeded the same number of MDA-MB-231 cells in either T75 or T150 flasks and allowed them to grow for up to four days, fixing periodically, to obtain a similar exponential curve of average nuclei per image to our original seeding method (Fig. S4.3A). Upon immunolabeling of Lamin A/C with the Ig1 antibody (Fig. S4.3B), we observed the same inverse relationship between Lamin A/C-Ig1 fluorescence intensity and cell seeding density, with a significant increase at low seeding densities compared to high seeding densities. These results indicate that Lamin A/C-Ig1 density-dependent immunolabeling is independent of cell cycle and confirmed that it is conserved across three human cell lines. As such, we focused primarily on human fibroblasts as a representative cell line for 134 future experiments. Lamin A/C-Ig1 density-dependent variation is not due to changes in protein expression To test whether density-dependent Lamin A/C expression could be due to changes in protein expression levels, we performed western analysis using an N-terminus Lamin A/C antibody of cells seeded at three densities (Fig. 4.2A). We found no significant differences in lamin levels, particularly with the more predominant Lamin C isoform, indicating that density-dependent Lamin A/C-Ig1 variation is not caused by changes Lamin A/C protein levels. To confirm these results using a second method, we used CRISPR/Cas9 to insert an mNeonGreen (mNG) tag at the N-terminus of endogenous Lamin A/C (mNG-LMNA) into MDA-MB-231 cells (Fig. 4.2B) to more directly quantify changes in Lamin A/C protein expression. We first validated the mNG-LMNA cells to ensure that we obtained a larger DNA amplicon by PCR analysis (Fig. S4.4A, C), detect a larger Lamin A/C protein isoform with Western Blot (Fig. S4.4B), and that mNG-Lamin A/C is properly expressed and localized to the nucleus (Fig. 4.2C). Finally, we confirmed that gene expression of mNG-LMNA responds as expected to modulation of the retinoic acid pathway, which regulates LMNA expression ( Fig. S4.4D; Swift et al., 2013), and that mNG- LMNA modification did not alter nuclear stiffness (Fig. S4.4E). We then seeded mNG-LMNA cells at either high or low cell seeding densities, corresponding to high confluency and numerous cell-cell contacts or low confluency with rare cell-cell contacts, and quantified the mNG-LMNA fluorescence intensity and fluorescence intensity of immunolabeling for Lamin A/C-Ig1 as a positive control (Fig. 4.2D). For each of the parental cell 135 line, cells modified with mNG-LMNA, and a clonal mNG-LMNA cell line (Fig. 4.2E), as expected we observed large, significant increases in Lamin A/C-Ig1 fluorescence intensity in low seeding densities compared to high seeding densities. Surprisingly, and contrary to immunolabeling of Lamin A/C-Ig1, we observe some significant increase in mNG-LMNA expression at high seeding densities for mNG-LMNA expressing cells (Fig. 4.2F), although the scale of these changes is much smaller compared to those observed with the immunolabeling with the Lamin A/C-Ig1 antibody. Although this suggests that there may be some small differences in Lamin A/C expression at higher seeding densities, these results confirm that the density-dependent Lamin A/C-Ig1 immunolabeling is not due to changes in Lamin A/C protein expression. Rather, density-dependent immunolabeling results from conformational changes of the Lamin A/C Ig-fold that alter accessibility of certain regions and allow for changes in epitope binding. Lamin A/C density-dependent epitope binding does not correspond to changes in nuclear area or stiffness Since Lamin A/C is a key determinant of nuclear mechanical properties (Lammerding et al., 2006; Stephens et al., 2018a) and we observed a conformational change in the Lamin A/C Ig-fold that allows for density-dependent epitope binding, we asked whether density-dependent Lamin A/C- Ig1 variation is reflective of changes in nuclear mechanics. Since the mechanical properties of the nuclear can be influenced by cytoskeletal conformation (Liu et al., 2019) and increased nuclear area is correlated with decreased nuclear stiffness (Lammerding et al., 2006), we quantified nuclear area in human fibroblasts seeded in a serial dilution to determine whether density- dependent conformational changes could be related to nuclear mechanics. However, human fibroblasts seeded at the different densities showed no differences in nuclear area (Fig. 136 4.3A), as was the case when we repeated the quantification with two additional two cell lines (HeLa or MDA-MB-231) that exhibit Lamin A/C-Ig1 density-dependent epitope binding (Fig. S4.5A-B, respectively). These data suggest that density-dependent Lamin A/C-Ig1 immunolabeling is independent of nuclear spread area. Fig. 4.2. Density-dependent Lamin A/C immunolabeling is not due to altered protein levels. (A) A Western Blot shows that human fibroblasts Lamin A/C levels are similar at different cell seeding densities. (B) An mNeonGreen (mNG) fluorescent tag was inserted into the LMNA gene (mNG-LMNA) in between the 5’ Untranslated Region (UTR) and Exon 1. (C) mNG-Lamin A/C (LAC) is highly expressed and localized only to the nucleus in MDA-MB- 231 cells, as expected. (D) Representative images of mNG-LAC expression of immunofluorescence staining for LAC-Ig1 in MDA-MB-231 cells modified with mNG-LMNA. (E) Lamin A/C-Ig1 fluorescence intensity is significantly reduced at high cell densities for parental, mNG-LMNA, and a clonal cell line of mNG-LMNA, as expected. (F) mNG-LAC fluorescence intensity does not follow the same density-dependent effect as the Lamin A/C-Ig1 antibody, with only some small, significant increase in signal intensity at higher cell densities. **p < 0.01, ***p < 0.001. N > 60 nuclei per group. 137 To obtain a more direct measurement of nuclear mechanics, we quantified nuclear stiffness using high-throughput micropipette aspiration (M. Davidson et al., 2019) with the same number of human fibroblasts seeded into different-sized cell culture vessels (T25, T75, or T150) to achieve different cell seeding densities. In this assay, a pressure gradient is applied across a series of microfluidic channels, the nucleus is aspirated into a narrow channel (Fig. 4.3B), and the length of the nuclear protrusion into the channel is used as a proxy for nuclear stiffness (Fig. 4.3C). We did not observe any density-dependent changes in nuclear protrusion length into the channel (Fig. 4.3C), indicating that nuclear stiffness did not change with cell seeding density. However, it should be noted that we do not know how long the cells retain a “memory” of their density upon trypsinization for the experiment. To confirm these results, we also employed a microharpoon assay (Fedorchak and Lammerding, 2016), in which a microneedle is inserted at a specific distance from the nucleus and pulled a specific distance away from the nucleus(Fig. 4.3D) to infer changes in nuclear mechanics and nucleo-cytoskeletal coupling. We quantified both the nuclear strain (Fig. 4.3E) and nuclear centroid displacement (Fig. 4.3F) of human fibroblasts seeded at several cell densities in response to microharpoon pulling as metrics of nuclear mechanics. While we observed some fluctuations in nuclear strain (Fig. 4.3E) and nuclear centroid displacement (Fig. 4.3F), the overall trend did not show any correlation between changes in nuclear mechanics and cell seeding density, consistent with our results from the micropipette studies. Together, these results indicate that density- dependent Lamin A/C Ig-fold conformational changes that allow for differential epitope binding do not significantly change the mechanical properties of the nucleus. 138 Fig. 4.3. Density-dependent Ig-fold conformational changes do not correspond to changes in nuclear mechanics. (A) Human fibroblasts do not exhibit density-dependent changes in nuclear area. (B) High-throughput micropipette aspiration to compute nuclear protrusion length into the channel was used as an indicator of nuclear mechanics. (C) Nuclear protrusion length measured by high-throughput micropipette aspiration showed no trends across high, medium, and low seeding densities, indicating no changes in nuclear mechanics. (D) A microharpoon assay in which a microneedle is inserted into the cytoplasm below the nucleus and pulled away over a certain distance was used to quantify changes in nuclear mechanics. Quantification of (E) nuclear strain and (F) change in nuclear major axis length from the microharpoon assay showed no significant trends across cell seeding densities. **p < 0.01. N ≥ 98 nuclei per group for nuclear area. N > 50 nuclei for micropipette experiments. N ≥ 9 nuclei per group for microharpoon experiments. 139 Actin expression per cell correlates with Lamin A/C Ig-fold conformational changes Previous studies showing that the Lamin A/C-Ig fold undergoes mechanoresponsive conformational changes found that these were mediated by the actin cytoskeleton, particularly through apico-basal compression of the nucleus by perinuclear actin (Swift et al., 2013; Ihalainen et al., 2015). Therefore, to determine whether density-independent Lamin A/C Ig-fold conformational changes could be correlated to differing actin filament (F-actin) levels between seeding densities, we labeled F-actin with Phalloidin (Fig. 4.4A) and quantified the F-actin fluorescence intensity per cell (Fig. 4.4B). Perhaps unsurprisingly given the increased cell spread area at low cell seeding densities and confluencies, we found that lower cell densities had significantly more F-actin per cell (Fig. 4.4B) that followed a similar inverse correlation with cell seeding density to that observed with Lamin A/C-Ig1 epitope binding (Fig. 4.1). Fig. 4.4. Actin per cell varies inversely with seeding density. (A) Representative images of F-actin, labeled with Phalloidin, and DAPI at low (0.0625×) and medium (0.25×) seeding densities. (B) The average phalloidin fluorescent intensity per cell shows a significant inverse relationship between seeding cell density and actin per cell. (C) Quantification of nuclear height across seeding densities shows that the highest seeding density (1×) has decreased nuclear height compared to low (0.0625×) and medium (0.25×) densities. Scalebar = 50 µm. *p < 0.05, ***p < 0.001, ****p < 0.0001. N > 20 nuclei per group for nuclear height experiments. N ≥ 28 images per group for per-cell actin experiment. 140 Since these results suggested that increased actin polymerization at lower seeding densities may be related to density-dependent Ig-fold conformational changes, we quantified nuclear height as a metric of nuclear compression to determine whether apico-basal compression of the nucleus by perinuclear actin could explain density-dependent conformational changes, as was observed in previous studies (Swift et al., 2013; Ihalainen et al., 2015). We seeded cells at three densities to represent a range from low cell confluency and rare cell-cell contacts (0.0625×) to high cell confluency with numerous cell-cell contacts (1×), and quantified nuclear height from high- resolution confocal images of nuclei. However, contrary to what we expected with significantly more actin polymerization per cell at lower seeding densities, we observed that cells seeded at higher densities had significantly decreased nuclear height compared to cells seeded at lower densities (Fig. 4.4C). Vogel and colleagues found that apico-basal polarization of binding of a Lamin A/C Ig-fold antibody acted as a sensor for nuclear spreading, with more flattened nuclei having increased apico-basal polarization of Lamin A/C (Ihalainen et al., 2015). To determine whether the cell density-dependent differences in nuclear height (i.e., degree of nuclear compression) were sufficient to have a differential functional impact on the mechanical conformation of the nucleus, we examined apico-basal polarization of Lamin A/C and Lamin B1 as a control in human fibroblasts seeded at the same three representatives densities (Fig. 4.2A). As expected since we found the Lamin A/C-Ig1 antibody to be mechanosensitive, we found the fluorescence intensity profile of Lamin A/C-Ig1 in the z-direction to show a skew towards apical polarization and a more even apico-basal distribution of Lamin B1 (Fig. S4.6A; Ihalainen et al., 2015). We quantified the apico-basal polarization by dividing each nucleus into the vertical top and bottom halves, summing 141 the fluorescence intensity across each half, and taking the ratio of basal:apical fluorescence intensity. Human fibroblasts seeded at all three densities showed a similarly reduced ratio of reduced basal:apical fluorescence intensity, suggesting that all three densities had similar apico- basal polarization (Fig. 4.2B), and that differences in nuclear height between seeding densities were not sufficient to induce differences in the mechanical conformation of nuclei. As expected, Lamin B1 showed little apico-basal polarization that was similar across all seeding densities (Fig. 4.2C). These results indicating that the small differences in nuclear height we observed are insufficient to cause altered mechanical conformation of the nucleus, paired with our results showing nuclear area does not change with cell seeding density, demonstrate that nuclear compression and spread area is not responsible for the density-dependent Lamin A/C Ig-fold conformational changes we observe. Density-dependent conformational changes are regulated by actin and microtubule polymerization Since per cell actin polymerization had an inverse correlation with cell seeding density, similar to that of immunolabeling of Lamin A/C-Ig1 and cell seeding density, to determine whether forces exerted on the nucleus by actin could be responsible for Lamin A/C density-dependent conformational changes, we depolymerized the actin cytoskeleton with Cytochalasin D in human fibroblasts. Labeling for actin with Phalloidin in Cytochalasin D-treated human fibroblasts confirmed strong depolymerization of the actin cytoskeleton (Fig. 4.5A). Immunofluorescence labeling for Lamin A/C-Ig1 in Cytochalasin D-treated human fibroblasts and quantification of fluorescence intensity showed a significant reduction in Lamin A/C-Ig1 fluorescence intensity at each density when compared with vehicle-treated cells, although the differences in Lamin A/C- 142 Ig1 fluorescence intensity between low and high seeding densities were not ablated (Fig. 4.5B). This suggests that while actin polymerization may not fully explain density-dependent conformational changes and subsequent binding of the Lamin A/C-Ig1 epitope, it may contribute at least in part to the effect. Fig. 4.5. Density-dependent Lamin A/C-Ig1 epitope expression is modulated in part by cytoskeletal tension. (A) Representative images of labeling actin with Phalloidin and immunofluorescence labeling for Lamin A/C-Ig1 at representative low (0.0625×) and medium (0.25×) cell seeding densities show depolymerization of actin filaments after treatment with Cytochalasin-D (CytoD). (B) Cytochalasin-D treatment in human fibroblasts results in significant reduction of Lamin A/C-Ig1 fluorescent intensity across nearly all individual seeding densities compared to control cells. Scalebar = 200 µm. **p < 0.01, ****p < 0.0001. N > 100 nuclei for Cytochalasin D experiments. The relatively small reduction in density-dependent Lamin A/C Ig-1 immunolabeling was surprising given previous reports of actin-mediated Ig-fold conformational changes (Swift et al., 2013; Ihalainen et al., 2015) and the strong increase we observed in per cell F-actin at low seeding densities (Fig. 4.2). As nuclear spread area can be used as an indicator of nuclear tension, to confirm that actin depolymerization with Cytochalasin D resulted in a loss of nuclear tension, we measured nuclear area in human fibroblasts (Fig. S4.7A) treated with either a vehicle control or 143 Cytochalasin D. We found a significant reduction in nuclear area in response to Cytochalasin D treatment at nearly every seeding density (Fig. S4.7A), indicating that nuclear tension was reduced with Cytochalasin D treatment. As a second readout of loss of nuclear tension, we re-measured nuclear height of human fibroblasts treated with Cytochalasin D and found that nuclear height of Cytochalasin D-treated fibroblasts was significantly increased compared to vehicle-treated cells (Fig. S4.7B), as expected, and confirming the loss of nuclear tension with Cytochalasin D treatment. To determine whether these results showing a relationship between actin polymerization and Ig- fold conformational changes were not specific to human fibroblasts, we treated HeLa cells with Cytochalasin D, and again found a small but significant reduction in Lamin A/C-Ig1 fluorescence intensity at each seeding density compared to vehicle-treated control cells (Fig. S4.7C). We quantified nuclear area in Cytochalasin D-treated HeLa cells to confirm that nuclear tension was indeed reduced, and observed a similar significant reduction in nuclear area at nearly each seeding density (Fig. S4.7D). Together, the incomplete ablation of the Lamin A/C-Ig1 density-dependent immunolabeling with reduction of cytoskeletal tension across several human cell lines suggests that Lamin A/C Ig-fold conformational changes are not regulated purely by forces exerted by F-actin on the nucleus. In addition to actin, other cytoskeletal consitutents may extert forces on nuclei to govern mechanosensitive conformational changes of the Ig-fold. Additionally, as chromatin plays a role in governing nuclear mechanics (Stephens et al., 2017, 2018b; Stephens, 2020), chromatin compaction may influence Lamin A/C Ig-fold conformational changes. To parse whether other 144 nuclear constituents contribute to Lamin A/C Ig-fold conformational changes, we examined the effect of microtubule polymerization, chromatin compaction, or more general force transmission through the LINC complex on Lamin A/C-Ig1 density-dependent immunolabeling of human fibroblasts seeded at either low (3,125 cells/cm2) or high (31,250 cells/cm2) densities. To determine whether microtubule polymerization influences Lamin A/C Ig-fold conformational changes, we depolymerized microtubules by treating with nocodazole, and quantified Lamin A/C- Ig1 fluorescence intensity (Fig. 4.6A). We found that while there was still some difference in Lamin A/C-Ig1 fluorescence intensity, it was no longer significant, indicating that microtubule polymerization influences Lamin A/C Ig-fold conformational changes. We repeated the experiment to determine the role for chromatin compaction by inhibiting histone deacetylase with Trichostatin A (TSA) to increase chromatin compaction and found that while immunolabeling of Lamin A/C-Ig1 was increased at both seeding densities compared to vehicle controls, there was still a significant difference in Lamin A/C-Ig1 fluorescence intensity between high and low seeding densities (Fig. 4.6B). This suggests that there may be some connection between nuclear mechanical properties governed by chromatin and Lamin A/C Ig-fold conformational changes, but differences in chromatin compaction between seeding densities likely does not govern density-dependent Lamin A/C Ig-fold conformational changes. Finally, to determine whether general force transmission through the LINC complex governs Lamin A/C Ig-fold conformational changes, we severed nucleo-cytoskeletal coupling of the LINC complex by using dominant-negative KASH disruption (DN-KASH) in human fibroblasts seeded at the same low and high cell densities. Interestingly, LINC complex disruption with DN-KASH 145 did not change Lamin A/C-Ig1 fluorescence intensity in either seeding density compared to the control (Fig. 4.6C). Since we have observed that actin and microtubule polymerization do play a role in Lamin A/C Ig-fold conformational changes, this suggests that compressive forces from either cytoskeletal constituent may influence Lamin A/C Ig-fold conformation, independent of connections to the LINC complex. Fig. 4.6. Density-dependent Lamin A/C-Ig1 epitope expression is modulated in part by cytoskeletal tension. Lamin A/C (LAC)-Ig1 fluorescence intensity per nucleus normalized to the low density vehicle control of human fibroblasts seeded at either low (3,125 cells/cm2) or high (31,250 cells/cm2) cell density. (A) Treatment with Nocodazole to depolymerize microtubules ablates the significant difference in Lamin A/C-Ig1 fluorescence intensity between low and high seeding density observed in control cells. (B) Treatment with Trichostatin A (TSA) increased Lamin A/C-Ig1 fluorescence intensity in both high and low seeding densities compared to vehicle controls, but maintained a significant reduction in fluorescence intensity between in the high seeding density compared to the low seeding density. (C) Modification with inducible dominant-negative KASH (DN-KASH) to disrupt nucleo- cytoskeletal coupling did not change the reduction Lamin A/C-Ig1 fluorescent intensities at the high seeding density compared to the low seeding density. Scalebar = 200 µm. **p < 0.01, ***p < 0.001. N > 45 nuclei per group for all experiments. 146 Lamin A/C density-dependent epitope binding is most pronounced with the Ig1 antibody and constrained to the Ig-fold This is the first observation of density-dependent Lamin A/C Ig-fold conformational changes for the epitope bound by the Lamin A/C-Ig1 antibody and its behavior is distinct from other Lamin A/C-Ig fold mechanosensitive responses in that it is only partially governed by actin polymerization (Swift et al., 2013; Ihalainen et al., 2015). As such, to determine whether cell density-dependent conformational changes in the Lamin A/C Ig-fold affect the binding of antibodies targeting other epitopes in the Ig-fold, we repeated the serial dilution seeding experiment with human fibroblasts and immunofluorescence labeled Lamin A/C with three additional antibodies (Fig. 4.7A): one binding the N-terminus as a control (Lamin A/C-N) and two binding regions of the Ig-fold (Lamin A/C-Ig2 and -Ig3) that overlap with the Lamin A/C-Ig1 antibody but have more specifically identified binding sites (Dyer et al., 1997; Ihalainen et al., 2015). The Lamin A/C-Ig2 antibody had previously been shown by Vogel and colleagues to exhibit a differential apico-basal accessibility that responds to changes in nuclear tension and compression (Ihalainen et al., 2015). Previous studies have shown that only a specific region of the Ig-fold is known to have mechanosensitive conformational changes (Swift et al., 2013; Ihalainen et al., 2015), so we did not anticipate any density-dependent immunolabling with the Lamin A/C-N antibody, which binds the N-terminal head of Lamin A/C. Accordingly, the Lamin A/C-N antibody exhibited no significant changes in immunolabeling across all seeding densities (Fig. 4.7B), indicating a lack of conformational change that allows for differential antibody binding in this region of the protein. 147 Interestingly, although the Lamin A/C-Ig2 apico-basal distribution is known to be mechanosensitive (Ihalainen et al., 2015), we only observed a small and slightly significant increase in Lamin A/C-Ig2 fluorescence intensity between the lowest and highest seeding densities (Fig. 4.7C). However, we noted that Lamin A/C-Ig2 was extremely sensitive to the fixation method: the fluorescence intensity varied strongly between experiments with some being too dim to image and others showing a strong density-dependent inverse correlation with fluorescence intensity of the Lamin A/C-Ig2 antibody (Fig. S4.8). We tried several fixation times and concentrations of PFA and only achieved semi-consistent labeling with one concentration of PFA. Fig. 4.7. Density-dependent epitope binding is not fully replicated with other Lamin A/C N-terminus or Ig-fold antibodies. (A) Lamin A protein structure with locations of the Lamin A/C-Ig1 antibody and of three additional antibodies to look for density-dependent epitope binding: one N-terminus antibody (Lamin A/C-N) and two additional Ig-fold antibodies (Ig1 and Ig2) that overlap the Lamin A/C-Ig1 antibody. More specific binding locations are shown in black. (B) The Lamin A/C-N antibody shows no cell density-dependent differences in fluorescence intensity normalized to the 1.5× density. (C) The Lamin A/C-Ig2 antibody, previously shown to be mechanosensitive (Ihalainen, et al, 2015), shows a slight inverse correlation between cell seeding density and normalized fluorescence intensity that only has a significant difference between the most extreme cell densities. (E) The Lamin A/C-Ig3 antibody shows no cell density-dependent differences in fluorescence intensity normalized to the 1.5× density. *p < 0.05. N > 67 nuclei per group. 148 Together this experiment-to-experiment variability amounted to extremely large error bars at the lower cell seeding densities, potentially masking the effect of density-dependent immunolabeling observed more in some experiments compared to others. However, as the Lamin A/C-Ig2 antibody is known to be sensitive to Lamin A/C conformation, our inconsistent results may indicate that Lamin A/C conformation and thus the Lamin A/C-Ig2 epitope binding may be sensitive to fixation method. Finally, the Lamin A/C-Ig3 antibody, which is known to bind a specific region of the Lamin A/C Ig-fold at AA 477-478 (Manilal et al., 2004), overlapping with both the Ig1 and Ig2 antibodies, showed no cell density-dependent changes in Lamin A/C-Ig3 immunofluorescence labeling (Fig. 4.7D), suggesting that these specific amino acids are not involved in density-dependent conformational changes. Proposed mechanism of density-dependent epitope binding The combination of some density-dependent epitope binding of the Lamin A/C-Ig2 antibody and its sensitivity to fixation method points to this region of the Ig-fold being involved in Lamin A/C Ig-fold conformational changes that allow for conditional epitope binding, similar to what was observed by Vogel and colleagues (Ihalainen et al., 2015). The Lamin A/C-Ig2 antibody has been epitope mapped and known to in part bind AA475-497 (Ihalainen et al., 2015), which falls in the C’E loop of the Ig-fold ( Fig. 4.8A; Krimm et al., 2002). Similarly, the cryptic Cys522 residue that is known to become exposed in response to mechanical strain (Swift et al., 2013) falls in another loop domain of the Ig-fold, the EF loop (Fig. 4.8A; Krimm et al., 2002). Although the specific binding sites of the Lamin A/C-Ig1 antibody in unknown, the general region it is known to bind 149 overlaps the outer edges of the Ig-fold, including both the C’E and EF loop regions (Fig. 4.8A-B). Therefore, we hypothesize that the Lamin A/C-Ig1 antibody binds the C’E and/or the EF loop of the Ig-fold, which partially unfold in response to LINC complex-independent forces on the nucleus caused by actin and microtubule polymerization. Fig. 4.8. Lamin A/C-Ig1 antibody binding in the Ig-fold structure (A) Map of the Lamin A/C Ig-fold protein structure. Regions known to be mechanosensitive from the Lamin A/C-Ig2 epitope binding (Ihalainen, et al, 2015) and exposure of the cryptic Cys522 residue (Swift, et al, 2013) are highlighted in yellow and blue, respectively. Potential regions bound by the Lamin A/C-Ig1 antibody are shaded in red, notably overlapping two loop regions – C’E and EF – that are known to be mechanosensitive. (B) Lamin A/C Ig-fold 3D structure, in which loop regions known to be mechanosensitive are highlighted in green. 150 Discussion and Conclusions The Lamin A/C Ig-fold is known to be mechanosensitive, particularly in response to cellular shear stress and force transmission of perinuclear actin on the nucleus (Swift et al., 2013; Ihalainen et al., 2015). Here, we have demonstrated that the Lamin A/C Ig-fold undergoes conformational changes in response to cell seeding density that allows for a commonly used Lamin A/C Ig-fold antibody (JOL-2; referred to here as Lamin A/C-Ig1), previously not known to be mechanosensitive, to bind in a density-dependent manner, in which fluorescent intensity is significantly increased at low cell seeding densities. Our results, alongside two previous studies (Swift et al., 2013; Ihalainen et al., 2015), represent evidence that the Lamin A/C Ig-fold exhibits conformational changes in response to the cellular mechanical environment. Moreover, Lamin A/C is a commonly used marker for the nucleus, particularly in studies of nuclear mechanotransduction, and the quantification of Lamin A/C levels an important biomarker for countless physiological phenomenon, including development, stem cell differentiation, cancer, and aging, among others (Rober et al., 1989; Constantinescu et al., 2006; Duque and Rivas, 2006; Willis et al., 2008; Irianto et al., 2016; Dubik and Mai, 2020; Bell et al., 2021). As such, the JOL- 2 (Lamin A/C-Ig1) is a commonly used antibody. The results presented here indicate that it is critical for studies using the Lamin A/C-Ig1, Lamin A/C-Ig2, and likely other Lamin A/C Ig-fold antibodies, particularly for the quantification of Lamin A/C levels, be aware of its sensitivity to cellular mechanical conformation. While our results and others point to Lamin A/C Ig-fold mechanosensitive behavior, there are several interesting distinctions between our findings and those of similar studies (Swift et al., 2013; Ihalainen et al., 2015). First, while there is some overlap of the density-dependent epitope binding 151 of the Lamin A/C-Ig2 antibody and the Lamin A/C-Ig1, we observed only some more slight apico- basal polarization of the Lamin A/C-Ig1 antibody compared to that observed with the Lamin A/C- Ig2 antibody by Vogel and colleagues (Ihalainen et al., 2015). Second, we have observed a partial retention of density-dependent Lamin A/C-Ig1 epitope binding upon reduction of nuclear tension through the dissociation of either the actin cytoskeleton or microtubule network. Apico-basal compression of the nucleus has been shown to both induce apico-basal polarization of Lamin A/C- Ig2 epitope binding (Ihalainen et al., 2015) and regulate Lamin A/C expression through either cellular and nuclear spreading in response to stiffness (Swift et al., 2013) or apico-basal cell compression (Iyer et al., 2021). Therefore, we hypothesize that the root of the distinctions of our studies lies in the conformation of nuclei in our studies: we did not find any density-dependent differences in nuclear area and only small differences in in nuclear height that were inversely correlated to density-dependent Lamin A/C-Ig1 epitope binding. In the absence of density- dependent changes in nuclear spreading and on the same stiffness of material, it is highly plausible that density-dependent epitope binding could be independent of such other mechanosensitive mechanisms. One open question that remains with this study is the role of cell spread area and/or cell-cell contacts in Lamin A/C Ig-fold conformational changes. Both cell spread area and cell-cell contacts can induce unique patterns of force generation within the cell by varying the apical and/or lateral junctions from which the cytoskeletal network can connect to the extracellular environment (Wu et al., 2014; Yap et al., 2018). The cell seeding densities used in this study varied in confluency, thereby altering the surface area available to cells and forcing increased cell-cell contacts, but as each cell spread area and cell-cell contacts are critical for defining the mechanical conformation 152 of a cell, it is critical to understand their contributions to density-dependent conformational changes. To answer this question, in collaboration with a Ph.D. student in the laboratory of Dr. Matt Paszek, Steven Park, we will seed single human fibroblasts on micropatterned circles either 20 or 40μm in diameter to artificially control cell spread area and remove cell-cell contacts. By quantifying Lamin A/C Ig-1 fluorescence intensity on such a system, we can determine whether density-dependent conformational changes still occur and result from differential cell spreading independent of cell-cell contacts, or whether by removing cell-cell contacts, density-dependent conformational changes no longer occur. Beyond a fundamental understanding of how the Lamin A/C Ig-fold responds to mechanical stresses, the Ig-fold is a hotspot for LMNA mutations, many of which are located in the C’E and EF loop regions (Krimm et al., 2002) that we hypothesize to be responsible for the density- dependent Lamin A/C-Ig1 epitope binding and other similar mechanosensitive behaviors (Swift et al., 2013; Ihalainen et al., 2015). LMNA mutations occurring in the Ig-fold are known to destabilize the Ig-fold structure (Krimm et al., 2002; Bera et al., 2014; Dutta et al., 2018) and cause changes in nuclear organization (Vigouroux et al., 2001; Bera et al., 2014) and chromatin configuration (Vigouroux et al., 2001; Verstraeten et al., 2009). Such changes are already known to disrupt nuclear mechanotransduction processes and hence downstream biochemical signaling pathways (Davidson and Lammerding, 2014; Maurer and Lammerding, 2019; Donnaloja et al., 2020). However, our results and others demonstrating the Ig-fold exhibits mechanosensitive conformational changes (Swift et al., 2013; Ihalainen et al., 2015) point to a potential missing link through which nuclear mechanotransduction may be disturbed with LMNA Ig-fold mutations: such mutations cause changes in the Ig-fold protein structure that could disrupt the normal 153 conformational changes through with the Ig-fold goes in response to mechanical stressors, thereby causing improper adaptation of the nucleus to the cell’s mechanical environment. As more becomes known about both the Ig-fold’s adaptation to mechanical stresses and how LMNA Ig-fold mutations, particularly those falling in loop regions, disrupt normal nuclear mechanotransduction processes, this link should further be explored. In conclusion, this study demonstrates a novel means through which the Lamin A/C Ig-fold undergoes conformational changes, through the differential folding of loop regions. These conformational changes may be critical for regulating interactions with the many Ig-fold binding partners of this region that are involved in nuclear mechanotransduction and represent a point of further investigation for LMNA mutations affecting the Ig-fold that have disrupted nuclear mechanotransduction. 154 Supplementary Materials Fig. S4.1. Lamin A/C-Ig1 density-dependent immunolabeling is independent of fixation method. Fixation of human fibroblasts seeded in a serial density with (A) 2% PFA and (B) 1:1 Methanol:Acetone shows the expected significant inverse relationship between Lamin A/C-Ig1 fluorescence intensity and cell seeding density. *p < 0.05, **p < 0.01, ***p < 0.001. N > 90 nuclei per group. Fig. S4.2. HeLa cells show an inverse relationship between Lamin A/C-Ig1 fluorescence intensity and cell seeding density. (A) Average number of HeLa nuclei per image shows an exponential trend, as expected. (B) Lamin A/C (LAC)-Ig1 fluorescence intensity normalized to the highest cell seeding density, 1.5×, varies inversely with cell seeding density. (C) Lamin B1 fluorescence intensity normalized to the highest cell seeding density, 1.5×, does not change with cell seeding density. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. N = 30 images per group for number of nuclei. N > 100 nuclei per group for fluorescent intensity experiments. 155 Fig. S4.3. MDA-MB-231 cells grown to different seeding densities over four days exhibit the density-dependent Lamin A/C-Ig1 immunolabeling. (A) The average number of MDA- MB-231 nuclei per image upon growth over several days follows a similar exponential trend to serial dilution experiments. (B) MDA-MB-231 cells grown over several days show the same significant inverse relationship between Lamin A/C (LAC)-Ig1 fluorescence intensity and cell seeding density. *p < 0.05, ****p < 0.0001. N = 10 images for number of nuclei counts, N > 100 nuclei per group. 156 Fig. S4.4. Design and validation of mNeon Green knock-in into endogenous Lamin A/C (A) An mNeonGreen (mNG) fluorescent tag was inserted into the LMNA gene (mNG-LMNA) in between the 5’ Untranslated Region (UTR) and Exon 1 of MDA-MB-231 cells, and primers were designed to determine whether the mNG was successfully inserted such that an amplicon of either 971 base pairs (bp) or 2192 bp would be obtained. (B) A Western Blot shows that mNG-Lamin A/C has a larger size than endogenous Lamin A/C. (C) PCR products show the presence of a 2192 bp fraction in cells modified with mNG-LMNA. (D) mNG-LMNA cells respond similarly to the parental cell line to activation and inhibition of the retinoic acid pathway using a 48 hour treatment with retinoic acid (RA) and inhibitor (AGN). (E) The nuclear protrusion length obtained through micropipette aspiration shows that there is no change in nuclear mechanics of mNG-LMNA cells compared to the parental cell line. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. N > 67 nuclei per group for retinoic acid pathway experiments. N > 48 nuclei per group for micropipette experiments. 157 Fig. S4.5. HeLa and MDA-MB-231 cells do not exhibit major changes in nuclear area as a function of cell seeding density. (A) HeLa cells and (B) MDA-MB-231 cells that exhibited density-dependent Lamin A/C-Ig1 epitope binding do not exhibit major changes in nuclear area. *p < 0.05. N > 100 nuclei per group. 158 Fig. S4.6. All seeding densities show apico-basal polarization of binding of the Lamin A/C- Ig1 antibody. (A) Cross-sectional images immunofluorescence labeled for Lamin A/C-Ig1 or Lamin B1 in human fibroblasts and the corresponding fluorescence intensity profiles in the z- direction shows some apico-basal polarization of Lamin A/C-Ig1 in at a representative low (0.0625×) and high (1×) cell seeding density. (B) The ratio of Basal/Apical Lamin A/C-Ig1 fluorescent intensity shows some apico-basal polarization in all seeding densities. (C) The ratio of Basal/Apical Lamin B1 shows little apico-basal polarization in all seeding densities. N > 20 nuclei per group for apico-basal distribution and nuclear height experiments. 159 Fig. S4.7. Depolymerization of the actin network reduces nuclear tension and Lamin A/C- Ig1 fluorescence intensity. (A) Cytochalasin D treatment significantly reduced nuclear area of human fibroblasts at nearly all seeding densities. (B) Nuclear height is significantly increased at all seeding densities compared to controls, and is increased to similar levels at all seeding densities with Cytochalasin D treatment. (C) Cytochalasin-D treatment in HeLa cells results in significant reduction of Lamin A/C-Ig1 fluorescent intensity across all individual seeding densities compared to control cells, and but the overall density-dependent variation is still statistically significant. (D) HeLa nuclear area is significantly reduced at all seeding densities with Cytochalasin D treatment compared to controls. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 control vs Cytochalasin D. #p < 0.05 Cytochalasin D conditions. N > 98 nuclei per group for nuclear area and Lamin A/C-Ig1 fluorescent intensity experiments. N > 20 nuclei per group for nuclear height experiments. 160 Fig. S4.8. The Lamin A/C-Ig2 has high experiment-to-experiment variability. Immunofluorescence labeling for Lamin A/C-Ig2 and DAPI show that between nuclei and experiments, there is high variability in the Lamin A/C-Ig2 labeling. Scalebar = 200 µm for all images. 161 CHAPTER 5 Conclusions and future perspectives LMNA mutations have long been known to cause a collection of diseases termed ‘laminopathies,’ which are known to be characterized by nuclear damage and disrupted cell signaling. However, an incomplete understanding of how these phenotypes cause disease has hindered the generation of adequate therapeutics that both inhibit disease progression and ameliorate the underlying cellular pathology. Together, the chapters of this thesis represent progress towards understanding the three hypotheses through which LMNA mutations cause laminopathies (Chapter 1): the Structural Hypothesis, the Gene Regulation Hypothesis, and the Mechanotransduction Hypothesis. I have demonstrated that Lamin A/C organization and localization to the nuclear envelope (NE) is critical for nuclear stability (Chapter 2), and that the Lamin A/C Ig-fold undergoes an intriguing conformational change in response to cell seeding density (Chapter 4). While the precise molecular mechanisms of Ig-fold conformational changes remain to be elucidated, these results point to the possibility that Ig-fold conformational changes may act as a switch that enables cells to respond to their microenvironment (Chapter 4). Finally, I have completed one of the most comprehensive studies of transcriptomic changes in Lmna-DCM disease progression using iPSC-derived cardiomyocytes and mouse models of the disease, uncovering a role for metabolic dysfunction in the onset and progression of Lmna-DCM in Lmna-mutant or -deficient mice (Chapter 3), and one of the first studies to identify differentially expressed genes (DEGs) and disrupted signaling pathways common to both mouse and human LMNA-DCM (Chapter 3). 162 One particularly intriguing finding of this thesis is that nuclear damage (Chapter 2) and/or altered gene expression (Chapter 3) of individual LMNA mutations does not correlate well with patient disease phenotype and severity. Whereas the R249Q mutant cell line had the most severe nuclear damage, the L35P patient had the most severe presentation of LMNA-DCM and a drastically shortened lifespan (Chapter 2). Additionally, this thesis demonstrates that various LMNA-mutant patient-derived cell lines exhibit different varying combinations of nuclear damage and/or altered gene expression. The R249Q mutant cell line exhibited the most nuclear damage (Chapter 2) and little gene expression differences (Chapter 3) compared to healthy control lines, whereas the G449V mutant cell line exhibited little nuclear damage (Chapter 2) and strong gene expression changes (Chapter 3). Together these results point to a well-recognized gap in the field: that the relationship between type and severity of cellular pathology and patient disease severity and outcomes are not well understood. These findings also emphasize that likely multiple different mechanisms may be able to drive the disease progression in LMNA-DCM. In the case of the R249Q mutation, impaired nuclear stability and increased nuclear damage may be a major disease driver, whereas in the case of the G449V mutation, where nuclear stability was not substantially affected, altered gene expression may be responsible for the severe disease progression. Other mutations may present a spectrum of outcomes in between. As therapeutics emerge that either chiefly target disrupted cell signaling pathways (Muchir et al., 2012; Le Dour et al., 2017; National Institute of Health, 2017; Laurini et al., 2018) or reduction of nuclear damage (Cho et al., 2019; Earle et al., 2019; Chai et al., 2021), the idea that different LMNA mutations present varying degrees of physical damage or disrupted cell signaling suggests that therapeutics targeted towards the specific underlying cellular pathology would be more 163 appropriate than a one-size-fits-all approach. However, to target therapeutics towards the underlying cellular disease pathology, biomarkers must be developed to determine the degree to which a patient would benefit from therapeutics targeting certain cell signaling and/or reduction of nuclear damage. While the work in this thesis is still far from clinical translation, the findings presented here represent intriguing potential biomarkers for targeted therapeutics. The lamin mislocalization index (Chapter 2), which strongly correlates with the degree of nuclear damage, could be a predictor the efficacy of therapeutics targeting nuclear damage, or by measuring metabolic dysfunction (Chapter 3), for example, the suitability of using therapies targeting cellular signaling could be assessed. As the gaps in our understanding of the connection between cellular disease phenotypes and patient disease progression and severity, the development of such potential biomarkers alongside novel, effective therapeutics represent exciting progress towards the long elusive treatment of laminopathies. 164 APPENDIX A1. Investigating the role of cytoskeletal forces in LMNA-mutant stem cell-derived cardiomyocyte nuclear damage4 Lamin A/C comprises a major portion of the nuclear lamina, a dense protein meshwork inside of the inner nuclear membrane that gives the nucleus structural support and participates in diverse biochemical and mechanotransduction signaling pathways. The nuclear lamina connects to the cytoskeleton via a complex of proteins spanning the inner nuclear membrane called the LINC complex (linker of the nucleoskeleton and cytoskeleton), and thus cytoskeletal forces exerted on the nucleus can influence nuclear morphology and nuclear envelope rupture in health and disease. Mutations in the LMNA gene, encoding Lamin A/C, give rise to a collection of diseases termed ‘laminopathies,’ of which the major types affect skeletal muscle and heart. LMNA mutations can disrupt the structure of the nuclear lamina, thereby weakening the nucleus and leaving it susceptible to damage. Here, we employed human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) carrying different LMNA mutations with varying degrees of nuclear damage, R249Q and G449V, to investigate the role of two major cytoskeletal constituents – actin filaments and microtubules – in governing the abnormal nuclear shape and size in LMNA mutant iPSC-CMs. We found that actin polymerization and contractility and microtubule polymerization 4 This work was completed in collaboration with several members of the Lammerding lab. Maurer, Melanie, Zahr, Hind, Perati, Shriya, Gimse, Hanna, Lammerding, Jan. MM, HZ, and JL contributed to the conception and design of the work. MM, HZ, SP, and HG contributed to data acquisition and analysis. MM, HZ, SP, HG, and JL contributed to interpretation of the data. MM and JL contributed to the drafting of this chapter. 165 can impact nuclear area and circularity index of LMNA-mutant iPSC-CMs in a cell line-dependent manner, suggesting that the cytoskeleton plays a role in LMNA mutant iPSC-CMs nuclear damage. Introduction Lamin A/C form a dense protein meshwork inside of the inner nuclear membrane, and function to support the nucleus and participate in control of gene expression, biochemical signaling, and mechanotransduction. Nuclear lamins connect the nucleus to the cytoskeleton via the LINC complex (linker of the nucleoskeleton and cytoskeleton), which spans the nuclear envelope (NE) and plays critical roles in mechanotransduction signaling to guide nuclear shape, positioning/movement, and gene expression (Rothballer and Kutay, 2013; Banerjee et al., 2014; Maurer and Lammerding, 2019). Mutations in the gene LMNA give rise to a host of diseases, termed ‘laminopathies,’ which predominantly cause DCM and several skeletal muscular dystrophies. LMNA-DCM has a particularly poor prognosis compared to other forms of DCM, with a high occurrence of arrhythmias and up to 19% of all patients requiring heart transplants (Nicolas et al., 2019). However, current clinical strategies focus on slowing heart failure rather than targeting the cellular pathology, owing to the fact that the cellular pathology remains incompletely understood. While altered nuclear shape and the occurrence of nuclear envelope rupture have been recognized as widespread consequences of LMNA mutations or deletions (Broers et al., 2004; Lammerding et al., 2004b; Nikolova et al., 2004; Chandar et al., 2010; De Vos et al., 2011; Zwerger et al., 2013; Earle et al., 2019; Bertrand et al., 2020), only limited studies have demonstrated such nuclear damage in cardiac tissue (Nikolova et al., 2004; Chandar et al., 2010; Cho et al., 2019; Shah et al., 166 2019). Previously (Chapter 2), we found that induced pluripotent stem cell-derived cardiomyocytes with different LMNA mutations exhibit mislocalization of Lamin A/C and Lamin B1 from the NE that correlates to the degree of defective nuclear shape and increased nuclear size (Chapter 2). In the cell line with the most extreme lamin mislocalization and nuclear shape and size defects, R249Q, we also observed an increase in NE ruptures. However, the mechanisms through which weakened LMNA-mutant iPSC-CM nuclei incur nuclear damage, and the consequences of lamin mislocalization and nuclear damage to cellular health and downstream signaling remain unclear. Previous studies demonstrated that forces acting on the nucleus by the cytoskeleton govern nuclear shape and susceptibility of weakened nuclei to NE rupture (Hatch and Hetzer, 2016; Xia et al., 2018; Cho et al., 2019; Earle et al., 2019; Heffler et al., 2019). Actin compression and tension on the nucleus has been shown to cause NE rupture (Hatch and Hetzer, 2016; Xia et al., 2018), and in LMNA-DCM cardiomyocytes, actin contractility increases NE ruptures and causes DNA damage (Cho et al., 2019). Although little is known about nuclear positioning in cardiomyocytes, nuclear positioning by microtubules has been shown to cause nuclear envelope ruptures in vitro myotubes and skeletal muscle (Earle et al., 2019), and since microtubules and intermediate filaments play a key role in maintaining nuclear shape in cardiomyocytes (Heffler et al., 2019), it is critical to understand the contributions of both actin and microtubules in LMNA-DCM nuclear damage. Moreover, recent work severing nucleo-cytoskeletal force transmission through either KASH or SUN1 ablation has shown that general force transmission to the nucleus is a major driver of skeletal and cardiac cellular health and nuclear damage (Earle et al., 2019; Chai et al., 2021). 167 Here, we investigated the role of forces from either polymerization or contractility of actin or microtubule polymerization or microtubule-associated motors in governing nuclear circularity and size. Treatment with Latrunculin B and Blebbistatin, targeting actin polymerization and contractility, respectively, significantly decreased nuclear area in G449V mutant iPSC-CMs, and microtubule depolymerization with Nocodazole reduced nuclear shape defects in R249Q mutant iPSC-CMs, suggesting that microtubules may play a protective role in maintaining nuclear shape against the forces of actin or other cytoskeletal proteins. Interestingly, however, when we seeded iPSC-CMs on nanogrooved substrates to increase sarcomere alignment and force generation (Kim et al., 2010; Scuderi and Butcher, 2017), we observed a general reduction in nuclear area and circularity index across all cell lines. Although this study is only partially complete, I outline future studies using these approaches and others to further clarify these results and determine whether cytoskeletal forces acting on the nucleus cause increased NE rupture in LMNA-mutant iPSC-CMs. Results LMNA-mutant iPSC-CM nuclei respond to depolymerization of actin or microtubules in a mutation-specific manner To determine whether forces exerted on the nucleus by the cytoskeleton, chiefly actin and microtubules, cause the nuclear size and shape defects in LMNA-mutant iPSC-CMs. I treated healthy control and LMNA-mutant iPSC-CMs with drugs to stabilize or depolymerize microtubules (Paclitaxel, 10nM, and Nocodazole,0.5ug/mL, respectively) or to depolymerize or inhibit contractility of actin (Latrunculin B, 1μM, and Blebbistatin, 25μM, respectively). I performed immunofluorescence labeling for actin and microtubules to ensure that the intended effect of each drug was achieved (i.e. depolymerization of microtubules; Fig. A1.1), and 168 additionally labeled Lamin B1 to quantify nuclear area and circularity index. Vehicle Taxol Nocodazole Latrunculin B Blebbistatin Cytochalisin D Actin / α-Tubulin / Lamin B1 / DAPI Fig. A1.1. Representative images of G449V-iPSC-CMs treated with cytoskeleton- modifying drugs. Treatment with DMSO was used as a vehicle control. Paclitaxel (or Taxol) was used to stabilize microtubules, whereas Nocodazole treatment was used to depolymerize microtubules, as evidenced by the ablation of microtubule structure in the top image panel. Latrunculin B treatment depolymerized the actin cytoskeleton, as evidenced by fragmentation of actin in the top image panel, while Blebbistatin inhibited actin contractility. With the help of a former undergraduate student, Shirya Perati, to determine the role of microtubule and actin polymerization in LMNA-mutant iPSC-CM nuclear size, we quantified nuclear area in response to each drug treatment, normalizing the nuclear size of each cell line to its corresponding vehicle (DMSO) control. Nuclear area was not significantly altered for any cell line with either with Nocodazole or Paclitaxel treatment for microtubule depolymerization or stabilization, respectively (Fig. A1.2), suggesting that microtubule-associated forces exerted on the nucleus likely play little role in governing nuclear area. Unsurprisingly, actin depolymerization had a trend towards reduced nuclear area that was not statistically significant for all cell lines, likely due to the high degree of loss of tension and compression on the nucleus (Fig. A1.2B), but G449V mutant iPSC-CMs were the only cell line to have a significant decrease in nuclear area compared to its DMSO control (Fig. A1.2B). Inhibition of actin contractility with Blebbistatin significantly decreased nuclear area only in G449V mutant iPSC-CMs (Fig. A1.2B), suggesting 169 that forces from actin contraction pulling on the nucleus may increase nuclear size. Taken together, these results suggest that actin tension and contractility on G449V mutant iPSC-CM nuclei likely plays a role in the increased nuclear area we previously observed (Chapter 2). Additionally, to determine the effect of actin or microtubule polymerization or contractility on nuclear circularity in LMNA-mutant iPSC-CMs, we quantified nuclear circularity index in response to each drug treatment, normalizing the nuclear size of each cell line to its corresponding vehicle (DMSO) control. Treatment with Paclitaxel to stabilize microtubules caused no changes in circularity index in any of our cell lines (Fig. A1.2C), which was somewhat surprising given Fig. A1.2. G449V mutant iPSC-CMs nuclear area and R249Q mutant iPSC-CM nuclear circularity index respond to drugs targeting actin and microtubule polymerization. (A) Nuclear area of iPSC-CMs normalized to a vehicle (DMSO) control did not change in response to microtubule depolymerization with Nocodazole or stabilization with Taxol. (B) Nuclear area of G449V mutant iPSC-CMs significantly decreased with both actin depolymerization with Latrunculin B or inhibition of contractility with Blebbistatin. (C) Nuclear circularity index of R249Q iPSC-CMs was increased only with Nocodazole treatment to depolymerize microtubules. (D) Nuclear circularity index did not change with actin depolymerization with Latrunculin B or inhibition of contractility with Blebbistatin. *p<0.05, **p<0.01, ****p<0.0001 vs cell line vehicle control. 170 that a previous study showed microtubule impingements on nuclei may alter nuclear shape in adult cardiomyocytes (Heffler et al., 2019). However, depolymerization of microtubules caused a significant increase in circularity index in R249Q mutant iPSC-CMs compared to vehicle control (Fig. A1.2C), suggesting that microtubule polymerization may alter nuclear shape in R249Q mutant iPSC-CMs (Chapter 2). Treatments with Latrunculin B or Blebbistatin to depolymerize or inhibit contractility of actin, respectively, had no impact on nuclear circularity index in any of our cell lines (Fig. A1.2D). However, it is unsurprising that we observed no nuclear shape changes in cell lines other than R249Q, given that R249Q was the only cell line to have altered nuclear shape at baseline (Chapter 2). Together, these results point to a role for actin polymerization and contractility in governing nuclear area, particularly in G449V mutant iPSC-CMs, and a potential role for microtubule polymerization in governing nuclear shape in R249Q mutant iPSC-CMs. Nanogroove experiments Previous studies have established that actin contractility increases nuclear envelope rupture in both stromal cells and cardiomyocytes (Hatch and Hetzer, 2016; Cho et al., 2019), which is in line with our results here showing that actin polymerization and contractility may impact nuclear size in LMNA-mutant iPSC-CMs. To further test the role for sarcomere assembly and force generation in the generation of LMNA-mutant iPSC-CM nuclear damage, in collaboration with a postdoctoral scholar in the Lammerding Lab, Hind Zahr, we seeded healthy control and LMNA-mutant iPSC- CMs on nanogrooved substrates to increase sarcomere alignment, and presumably force generation, of iPSC-CMs (Kim et al., 2010; Scuderi and Butcher, 2017), compared to previous 171 results on unpatterned substrates. However, without directly measuring force generation of aligned LMNA-mutant that may have contractility defects, we cannot conclude that force generation is increased. Additionally, as the sarcomere can take up to seven days to reform after dissociation, such as after iPSC-CM passaging for experimentation (Taneja et al., 2020), we fixed iPSC-CMs seeded on nanogrooved substrates either one (Chapter 2, “D7+1”) or seven days (“D7+7”) to determine whether the increased sarcomere assembly and force generation would exacerbate LMNA-mutant iPSC-CM nuclear damage. To determine whether nanogrooved substrates promoted sarcomere alignment of iPSC-CMs, we utilized an established method, AFT – Alignment by Fourier Transform (Marcotti et al., 2021), to quantify sarcomere anisotropy in iPSC-CMs. With this method, grayscale maximum intensity projection images of iPSC-CMs labeled for cardiac troponin T (cTnT) was divided up into smaller areas, each of which were vectorized and the corresponding angle found (Fig. A1.3A; Marcotti et al., 2021). From the angles representing each vector of the image, an order parameter is calculated as 1 𝑂𝑟𝑑𝑒𝑟 𝑃𝑎𝑟𝑎𝑚𝑒𝑡𝑒𝑟 = 2(< 𝑐𝑜𝑠2𝜃𝑖𝑗 > − ), 2 in which 𝜃𝑖𝑗 represents the angle between a central reference vector and neighboring vectors (Marcotti et al., 2021). An order parameter of 1 is isotropic and closer to 0 is more anisotropic (Marcotti et al., 2021). As expected, unaligned D7+1 iPSC-CMs for all cell lines had little alignment, as evidenced by a small order parameter, whereas iPSC-CMs aligned on nanogrooves at both D7+1 and D7+7 had an increased order parameter (Fig. A1.3B). Interestingly, D7+1 iPSC- CMs were slightly more aligned compared to D7+7 iPSC-CMs, indicating that even before complete sarcomere reformation at Day 7+7 (Taneja et al., 2020), iPSC-CMs are well-aligned to 172 nanogrooves (Fig. A1.3B). Additionally, while there were small differences in alignment of iPSC- CMs between cell lines, LMNA-mutant iPSC-CMs were similarly aligned to healthy control iPSC- CMs at D7+1 and less aligned compared to healthy controls at D7+7, although statistics will need to be completed to determine this conclusively (Fig. A1.3B). As previously mentioned, typically sarcomere force generation is increased with alignment to nanogrooves. However, we have not yet A Grayscale Vectorized Angle Heatmap Order Parameter = 0.75 Order Parameter = 0.70 B Fig. A1.3. Seeding iPSC-CMs on nanogrooves increases sarcomere alignment. (A) Representative WT1 and R249Q mutant iPSC-CM cTnT grayscale, vectorized, and angle heatmap images, showing a similar degree of alignment to nanogrooves, as evidenced by a similar order parameter. Heatmap angles are indicated by the color scale on the right. (B) Quantification of order parameter shows that unaligned D7+1 iPSC-CMs have a small order parameter, as expected, which increases with alignment to nanogrooves on D7+1 and D7+7. 173 R249Q WT1 quantified contractility in LMNA-mutant iPSC-CMs on unaligned or aligned substrates, which may have contractility defects due to alterations to sarcomere organization, calcium handling, or force generation, so we cannot conclusively say that force generation is certainly increased on nanogrooved substrates. To determine if increased sarcomere alignment of LMNA-mutant iPSC-CMs to nanogrooves (Kim et al., 2010) increases nuclear damage, with the help of an undergraduate student, Hanna Gimse, we utilized a previously described MATLAB program (Chapter 2) to quantify nuclear area and circularity index. iPSC-CMs seeded on nanoogrooves at both D7+1 and D7+7 showed similar trends in nuclear area to unaligned D7+1 iPSC-CMs, with significantly increased nuclear area in LMNA-mutant iPSC-CMs (Fig. A1.5A). However, we observed some overall decrease in nuclear area across all cell lines seeded on nanogrooves at both timepoints (Fig. A1.4A), indicating that increased sarcomere alignment and force transmission to the nucleus do not cause increased nuclear area in LMNA-mutant iPSC-CMs, which, assuming that sarcomere force generation is indeed increased on nanogrooved substrates, would be the opposite of what we expected based on our results showing that reducing actin contractility with blebbistatin reduced nuclear area in G449V mutant iPSC-CMs. iPSC-CMs aligned to nanogrooves at both D7+1 and D7+7 had a general decrease in nuclear circularity index compared to unaligned D7+1 iPSC-CMs of all cell lines, suggesting an overall decrease in circularity (Fig. A1.4B). However, this is perhaps unsurprising as we would expect nuclei to become more elongated (i.e. having reduced circularity index) as sarcomere forces align to nanogrooves. Additionally, the degree of decrease in circularity index varied across cell lines and conditions, with no consistent trends, which could potentially be due to differences in alignment across experiments and images. Hanna will continue this analysis 174 to quantify nuclear eccentricity to determine whether the decrease in circularity is due to elongation of nuclei or an increased presence of irregular shaped nuclei, and to correlate the decrease in circularity index to alignment to nanogrooves. A **** **** * **** **** **** **** **** **** **** **** B **** ** **** **** **** ** * * Fig. A1.3. Alignment to nanogrooves decreases iPSC-CM nuclear area and circularity index for all cell lines. (A) Nuclear area of iPSC-CMs seeded on nanogrooves and fixed after 1 (D7+1) or 7 (D7+7) days is slightly decreased from nuclear area of unaligned iPSC-CMs at D7+1, but at each condition R249Q and G449V-mutant iPSC-CM nuclear area is significantly increased. (B) Circularity index of D7+1 and D7+7 iPSC-CMs seeded on nanogrooves is slightly decreased from unaligned D7+1 iPSC-CMs, but there is no clear correlation between circularity index and alignment to nanogrooves across cell lines. *p<0.05, **p<0.01, ****p<0.0001. 175 Discussion and Conclusion These studies, while preliminary, have hinted at cell line-specific contributions of actin and microtubules to nuclear shape and circularity index for the G449V mutant and R249Q mutant iPSC-CMs, respectively. Additionally, while healthy control and LMNA-mutant iPSC-CMs seeded on nanogrooves show similar alignment one day post-seeding, when the sarcomere is immature, LMNA-mutant iPSC-CMs have reduced alignment to the nanogrooves after one week post seeding, when the sarcomere has fully matured. This could potentially be reflective or altered matrix mechanosensing, or an inability of LMNA-mutant iPSC-CMs to properly organize sarcomeres. While we observed a general decrease on nuclear area of healthy control and LMNA-mutant iPSC- CMs seeded on nanogrooved substrates, LMNA-mutant iPSC-CMs still showed similar trends of increased nuclear area compared to healthy controls, suggesting that alignment and potential increase of sarcomere forces does not play a role in governing nuclear area. Finally, all iPSC-CMs had reduced nuclear circularity index on nanogrooved substrates, although it is unclear whether this is due to increased eccentricity of nuclei as they align to sarcomeres and nanogrooves, or whether other mechanisms could be governing irregular nuclear shape as they compensate for increased alignment of sarcomeres. As the results from these experiments are very preliminary, many open questions remain, and future analysis will be done to answer these questions. First, our results showing that G449V nuclear area is decreased by inhibiting actin contractility with blebbistatin is in stark contrast to the general reduction in nuclear area we observed across all cell lines upon alignment of iPSC- CMs to nanogrooves. If there indeed is reduced nuclear area from alignment of iPSC-CM sarcomeres to nanogrooves and increased force generation of actin exerting on the nucleus, would 176 increased maturation of iPSC-CMs and aligning them to nanogrooves further reduce nuclear area? To partially answer this question, Hind Zahr, is investigating iPSC-CMs matured to 28 days post- cardiac differentiation, instead of the 7 days in the experiments here, and seeded on unpatterened or nanogrooved substrates. Additionally, since we observed microtubule depolymerization to reduce R249Q mutant iPSC-CM nuclear defects, could microtubule polymerization around the nucleus be altered on nanogrooved substrates and thereby by driving changes in nuclear circularity? Or is a decrease in nuclear circularity index on nanogrooved substrates simply due to increased eccentricity of nuclei as they align to sarcomeres? Finally, if we increase actin contractility, such as with treatment with a β-adrenoreceptor agonist Isoproterenol, do we observe further reduced nuclear area? Additionally, one component of the cardiomyocyte cytoskeleton that is missing from our analysis to date is Desmin, an intermediate filament protein that connects the sarcomere to the nucleus. Desmin structure may be disrupted with LMNA mutations or depletion (Nikolova et al., 2004; Mounkes et al., 2005; Piercy et al., 2007; Galata et al., 2018), and its balance with microtubules is critical for maintaining proper nuclear shape in cardiomyocytes (Heffler et al., 2019). As such, particularly in the R249Q mutant iPSC-CMs that have disrupted nuclear shape, it is critical to understand desmin organization in the context of our drug-treatment and nanogroove studies to gain a full picture of how a balance of forces acting on nuclei regulates nuclear shape. Finally, perhaps even more importantly than understanding the regulation of nuclear area and shape in the context of cytoskeletal forces is understanding how they might cause increased NE rupture in R249Q mutant iPSC-CMs (Chapter 2). To answer this question, I have completed 177 experiments treating healthy control and R249Q mutant iPSC-CMs modified with a nuclear envelope rupture reporter, NLS-GFP (Chapter 2), with either Nocodazole or Taxol to depolymerize or stabilize microtubules, or with Isoproterenol to increase actin contractility. While this analysis has not yet been completed, these experiments will be critical to gaining an understanding of the role of the cytoskeleton in nuclear fragility of R249Q mutant iPSC-CMs. Together, the experiments both completed and proposed in this chapter represent progress towards understanding the mechanisms through which the cytoskeleton causes nuclear deformities and fragility in LMNA-mutant iPSC-CMs. As therapeutics targeting the reduction of force transmission to the nucleus (Cho et al., 2019; Earle et al., 2019; Chai et al., 2021) become increasingly understood, it is critical to develop an understanding of how force transmission of cytoskeletal components influence nuclear damage in LMNA-mutant iPSC-CMs in order to best employ cytoskeletal force-reducing therapies to treat the underlying cellular pathology in LMNA-DCM. Although additional mechanisms may drive LMNA-DCM, such therapeutics hold promise to improve cellular health and survival (Cho et al., 2019; Earle et al., 2019; Chai et al., 2021). 178 A2. Investigating altered DNA damage response in LMNA-mutant induced pluripotent stem cell derived cardiomyocytes Lamin A/C, encoded by the LMNA gene, are intermediate filament proteins that form a dense meshwork of intermediate filament proteins inside of the inner nuclear membrane and function to give the nucleus structural support and participate in diverse biochemical and mechanotransduction signaling pathways. LMNA mutations cause a host of tissue-specific diseases, termed ‘laminopathies,’ which most commonly affect skeletal muscle and heart in the form of dilated cardiomyopathy (LMNA-DCM) or muscular dystrophy. These laminopathies are often characterized by an increase in DNA damage, and a growing body of work has indicated that DNA damage response (DDR) pathways may be disrupted in numerous ways, potentially leading to defective DNA repair and decreased cell and muscle health. However, previous studies investigated the role of specific DDR proteins, rather than obtaining a global view of how DDR may be impacted by LMNA mutations associated with muscle defects. Here, I employed human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) with two different LMNA mutations, R249Q and G449V, to understand disrupted DDR as a potential disease mechanism of LMNA-DCM. I examined disruptions to constituents of the ATM pathway for double stranded break repair, the primary DNA repair pathway for post-mitotic cardiomyocytes, and found that while neither R249Q and G449V mutant iPSC-CMs have an increase in basal DNA damage indicated by γH2AX, a DNA damage marker, both R249Q and G449V mutant iPSC-CMs have significantly impaired phosphorylation of ATM and recruitment of 53BP1, indicating impaired DDR via the ATM pathway. While this study remained incomplete due to technical challenges, these results represent significant progress towards understanding of how LMNA mutations may 179 disrupt DDR. Introduction Increased apoptosis and senescence have long been observed in laminopathic models (Wolf et al., 2008; Gonzalez-Suarez et al., 2011; Redwood et al., 2011; Auguste et al., 2018; Kim et al., 2018; Earle et al., 2019). However, the driving mechanisms are not yet clear, and thus treatments to reduce apoptosis and senescence are lacking. Recent studies have found increased levels of DNA double-strand breaks (Redwood et al., 2011; Cho et al., 2019; Earle et al., 2019) accompanied by disrupted DNA damage response (DDR) (Manju et al., 2006; Gonzalez-Suarez et al., 2011; Moiseeva et al., 2011; Redwood et al., 2011; Chen et al., 2019b) and DNA damage repair defects (Redwood et al., 2011; Cho et al., 2019; Earle et al., 2019), suggesting that altered response to increased DNA damage may be driving apoptosis and senescence. However, to date there is no clear consensus of how DDR is dysregulated, with the literature alluding to diverse and sometimes conflicting alterations of gene expression and DDR activation for many of the DDR proteins (Fig. A2.1), including γH2AX (Manju et al., 2006; Chen et al., 2019b; Earle et al., 2019), ATM (Moiseeva et al., 2011; Chen et al., 2019b), ATR (Manju et al., 2006), 53BP1 (Liu et al., 2005; Gonzalez-Suarez et al., 2011; Redwood et al., 2011), and p53 (Varela et al., 2005; Kudlow et al., 2008; Moiseeva et al., 2011; Chen et al., 2019b; Shao et al., 2020). Moreover, the root cause(s) of DNA damage, apoptosis, and senescence remain unclear. Several non-mutually exclusive causes of DNA damage have been suggested, including mechanical damage (Lammerding et al., 2004b, 2006; Kim et al., 2018), nuclear envelope rupture (Earle et al., 2019), and oxidative stress (De Vos et al., 2011; Sieprath et al., 2012; Morales Rodriguez et al., 2018). 180 P P ATM 53BP1 P P CHK2 p53 DNA damage signal effectors Senescence Apoptosis Cell cycle arrest DNA repair Fig. A2.1. The ATM pathway for DNA double strand break repair. Phosphorylated histone 2AX (γH2AX) accumulates at sites of DNA repair. Recruitment and phosphorylation of ATM kinase activates a DNA damage response (DDR) cascade involving CHK2, p53, and 53BP1 to induce DNA repair, cell cycle arrest, senescence, and apoptosis. One reason for the scattered theories surrounding the causes and mechanisms of DNA damage, DDR, apoptosis, and senescence is that studies have reported only on either singular mechanisms or small portions of DDR signaling, and therefore give an incomplete picture of pathogenic DNA damage and DDR signaling. Similarly, these studies have been performed using different laminopathies caused by numerous mutations and in both human and mouse samples. Here, to resolve the issues of study heterogeneity and investigate altered DNA damage response as a potential pathogenic mechanism of LMNA-DCM, I used two human LMNA-mutant iPSC-CM lines, R249Q and G449V and iPSC-CMs from healthy control to characterize activation of the DDR cascade in response to a specific DNA damage-inducing treatment. I found that while LMNA- mutant iPSC-CMs have no changes in baseline DNA damage, activation or recruitment of two key DNA damage response proteins, phosphorylated ATM and 53BP1, is defective. Although several 181 technical challenges occurred to inhibit further progress on this work, I outline several key experiments that could be done to follow-up on these promising results. Results LMNA-mutant iPSC-CMs do not exhibit increased baseline DNA damage To determine whether LMNA-mutant iPSC-CMs had increased DNA damage and/or DNA damage activation, I immunofluorescently labeled a DNA damage sensor, gamma Histone 2AX (γH2AX), which binds to sites of DNA damage, and a DNA damage repair protein, 53BP1, which binds at sites of DNA damage. I quantified the number of γH2AX (Fig. A2.2A) and 53BP1 (Fig. A2.2B) foci per nucleus, and found that LMNA-mutant iPSC-CMs show no differences in the number of DNA damage foci compared with healthy control iPSC-CMs. A B Fig. A2.2. Baseline levels of γH2AX and 53BP1 in iPSC-CMs. LMNA-mutant iPSC-CMs had a similar percentage of nuclei with both (A) γH2AX foci and (B) 53BP1 foci, indicating similar levels of DNA damage and DNA repair, respectively. Differences were not statistically significant. Data represented as mean ± SEM. LMNA-mutant iPSC-CMs have defective recruitment of 53BP1 phospho-ATM Although baseline levels of DNA damage and DNA damage activation were unchanged, LMNA 182 % Nuclei with γH2AX Foci % Nuclei with 53BP1 Foci mutations can alter DNA repair, such as through recruitment and/or activation of ATM (Moiseeva et al., 2011; Chen et al., 2019b), ATR (Manju et al., 2006), 53BP1 (Liu et al., 2005; Gonzalez- Suarez et al., 2011; Redwood et al., 2011), and p53 (Varela et al., 2005; Kudlow et al., 2008; Moiseeva et al., 2011; Chen et al., 2019b; Shao et al., 2020). iPSC-CMs are post-mitotic and likely undergo primarily DNA repair through the ATM pathway and limited DNA repair through the ATR pathway, so I hypothesized that if DNA damage response is altered in LMNA-mutant iPSC- CMs, phosphorylation of ATM, recruitment of 53BP1, or activation of p53 would be impaired. As a means of studying DNA damage response activation, I used treatment of LMNA-mutant iPSC- CMs with Phelomycin to induce DNA damage (Fig. A2.3A). Accordingly, all iPSC-CM lines had a similar percentage of nuclei with γH2AX foci in response to Phleomycin treatment (Fig. A2.3B). A B WT2 R249Q Fig. A2.3. LMNA-mutant iPSC-CMs have normal recruitment of γH2AX. (A) Representative immunofluorescence images of γH2AX in iPSC-CMs treated with a DNA damaging agent, Phelomycin, for one hour. (B) Quantification of the percentage of nuclei with γH2AX show all cell lines had a similar percentage of nuclei with foci. Data represented as mean ± SEM. Results for untreated controls are shown in Figure A2.2 in Appendix 2. ATM is a kinase that is among first responders to DNA damage, and phosphorylation of ATM is critical to enable its phosphorylation of downstream substrates critical for DNA repair, cell cycle regulation, apoptosis, and senescence, among other processes (Maréchal and Zou, 2013). To examine phosphorylation of ATM in response to DNA damage, I treated healthy control and 183 γH2AX CTNT DAPI LMNA-mutant iPSC-CMs with a DNA damaging agent, Phleomycin, for one hour and performed immunofluorescence labeling for phosphorylated ATM (p-ATM) and Lamin B1 to look for increased p-ATM labeling in the nucleus (Fig. A2.4A). Although I observed significant background staining outside of the nucleus, nuclear p-ATM fluorescent intensity normalized to the vehicle-treated healthy control was fairly low across all cell lines, with some significant increase in both LMNA-mutant iPSC-CM lines, R249Q and G449V (Fig. A2.4B). With Phleomycin treatment, nuclear p-ATM fluorescent intensity increased in all cell lines, although there was significantly less nuclear p-ATM in R249Q and G449V mutant cell lines (Fig. A2.4B). This suggests that there may be some impaired activation of p-ATM, although these results were without consideration that total levels of ATM may vary across cell lines. A B Control Phleomycin **** Lamin B1 p-ATM Lamin B1 p-ATM **** **** **** ** **** **** Fig. A2.4. LMNA-mutant iPSC-CMs have defective recruitment of phospho-ATM. (A) Representative immunofluorescence images of Lamin B1 and phosphorylated ATM (p-ATM) in vehicle control iPSC-CMs and iPSC-CMs treated with a DNA damaging agent, Phelomycin. Scalebar = 50µm. (B) Quantification of nuclear p-ATM normalized to control WT2-iPSC-CMs shows that while R249Q and G449V mutant iPSC-CMs have a small but significant increase in basal nuclear p-ATM, both R249Q and G449V mutant iPSC-CMs have significantly impaired recruitment of p-ATM in response to Phleomycin treatment compared to healthy control iPSC-CMs. Data represented as mean ± SEM. **, p < 0.01, ****, p < 0.0001. 184 G449V R249Q WT2 Norm. Nuclear p-ATM (a.u.) To confirm these results, I attempted to use Western Blotting to quantify levels of both total ATM and p-ATM. However, due to technical challenges of blotting with large proteins, as ATM is 350 kDa in size, despite attempts with several Western Blot protocol modifications and antibodies, I was never able to detect either ATM or p-ATM. LMNA-mutant iPSC-CMs have defective recruitment of 53BP1 53BP1 is another critical mediator for DNA repair, and recruitment of 53BP1, a binding partner of Lamin A/C, is triggered by ATM’s phosphorylation of H2AX at sites of DNA damage within minutes (Maréchal and Zou, 2013; Georgescu et al., 2015; Bártová et al., 2019). Since 53BP1 recruitment may be altered with LMNA mutations (Liu et al., 2005; Gonzalez-Suarez et al., 2011; Redwood et al., 2011) and I observed potentially defective activation of p-ATM, to determine whether LMNA-mutant iPSC-CMs have impaired recruitment of 53BP1, which could potentially indicate disrupted DNA repair, I treated healthy control and LMNA-mutant iPSC-CMs with a DNA damaging agent, Phleomycin, immunofluorescence labeled 53BP1 (Fig. A2.5B) and quantified the number of foci per nucleus. Although 53BP1 labeling showed some level of background, foci in nuclei were still very distinct (Fig. A2.5A). While healthy control nuclei showed normal recruitment of 53BP1, with nearly all nuclei having foci, LMNA-mutant iPSC-CM nuclei had severely impaired recruitment of 53BP1 in response to Phleomycin treatment, with only 20-30% of nuclei having foci (Fig. 2.5A-B). 185 A B *** WT2 R249Q ** **** ** Fig. A2.5. LMNA-mutant iPSC-CMs have defective recruitment of 53BP1. (A) Representative immunofluorescence images of 53BP1 in iPSC-CMs treated with a DNA damaging agent, Phleomycin. Scalebar = 20μm. (B) Quantification of the percentage of nuclei with 53BP1 foci shows that R249Q and G449V mutant iPSC-CMs have severely impaired recruitment of 53BP1 in response to Phleomycin treatment. Data represented as mean ± SEM. **, p < 0.01, ***, p < 0.001. To better characterize the timing of recruitment of 53BP1 and whether impaired recruitment of 53BP1 has an impact on DNA repair in LMNA-mutant iPSC-CMs, I performed a time lapse experiment in which I treated healthy control and R249Q-iPSC-CMs with Phleomycin to induce DNA damage and fixed cells periodically in a 24-hour period following Phleomycin treatment, at which point 53BP1 should be nearly completely depleted from sites of DNA damage (Georgescu et al., 2015), and then immunofluorescence labeled 53BP1 to quantify the presence of foci. However, 53BP1 images showed a very inconsistent presence of foci across timepoints even in healthy control iPSC-CMs (Fig. A2.6), with some timepoints having few or almost no nuclei with distinct foci, even though timepoints afterwards would show many nuclei with distinct foci. Additionally, with 53BP1 immunofluorescence labeling showing some background staining, this method seemed too unreliable for tracking of DNA repair. 186 53BP1 CTNT DAPI 0:00 0:30 1:00 2:00 24:00 Fig. A2.6. Healthy control iPSC-CMs show inconsistent presence of foci across the 24 hour period following DNA damage induction with Phleomycin. Representative time lapse images of Lamin A/C and 53BP1 after induction of DNA damage with Phleomycin and washout shows inconsistent appearance of foci in nuclei across all timepoints. Time represented as hours:minutes. Scalebar = 50µm. As an additional approach to immunofluorescence staining, I stably modified iPSC-CMs with an mCherry fused to 53BP1 (mCherry-53BP1) to track live cell recruitment of 53BP1 in an overnight timelapse on a confocal microscope, rather than relying on antibody labeling. However, I ran into several issues with these experiments: first, mCherry-53BP1 expression was quite low in iPSC- CMs, and they cannot be selected with antibiotics since their health relies on a confluent culture; second, overnight experiments on the Lammerding Lab’s confocal microscope require seeding cells on a glass coverslip, and iPSC-CM adherence and health was poor on this substrate; and third, cell survival in the overnight time period was extremely poor, with cells dying after only a few hours. Due to the troubleshooting time required for these experiments and the projects in the main chapters of this thesis being more promising, I did not continue with these experiments. iPSC-CMs have little expression of p53 p53 is a transcription factor activated by phosphorylation by ATM (Turenne et al., 2001; Cheng 187 53BP1 Lamin A/C and Chen, 2010) and binding to 53BP1 (Rappold et al., 2001; Ward et al., 2003) that regulates DNA repair, cell cycle control, apoptosis, and senescence. LMNA mutations or depletion may alter p53 signaling and consequently senescence, apoptosis, and altered cell cycle control in mitotic cells (Varela et al., 2005; Kudlow et al., 2008; Moiseeva et al., 2011; Chen et al., 2019b; Shao et al., 2020). As such, altered activation or recruitment of ATM and 53BP1 may be mediators of disrupted p53 signaling in LMNA mutations. To determine whether p53 is being properly activated in LMNA-mutant iPSC-CMs, I used immunofluorescence labeling and Western Blotting for p53 with healthy control and LMNA-mutant iPSC-CMs treated with a p53-activator, Doxorubicin, or a vehicle control to look for differences in upregulation of p53. However, despite trying several p53 antibodies, doses of Doxorubicin, and a second p53-activator, Nutlin-3, I was never able to detect any p53 in iPSC-CMs by either immunofluorescence or Western Blot. Discussion and Conclusion Recent studies point to increased DNA damage (Redwood et al., 2011; Cho et al., 2019; Earle et al., 2019) and altered DNA damage response (DDR) (Manju et al., 2006; Gonzalez-Suarez et al., 2011; Moiseeva et al., 2011; Redwood et al., 2011; Chen et al., 2019b) as pathogenic mechanisms in laminopathies. However, previous work has primarily focused on LMNA mutations associated with Hutchinson-Gilford Progeria Syndrome (HGPS), not LMNA-DCM, and studies establishing a complete picture of how DDR may be disrupted to drive disease pathogenesis have been lacking. Here, I demonstrated exciting progress towards understanding disruptions to the ATM pathway for DNA repair in LMNA-mutant iPSC-CMs. Although LMNA-mutant iPSC-CMs do not exhibit increased DNA damage at baseline, I have found that phosphorylation of ATM and recruitment of 53BP1 is defective in response to induction of DNA damage with Phleomycin. 188 As I ran into many technical challenges with these experiments and did not complete this project, many limitations and open questions remain that should be answered with future studies. Both of these experiments were completed using immunofluorescence and a single method of DNA damage induction, and therefore should be confirmed using additional methods. Perhaps most importantly, from these experiments we do not yet understand whether defective phosphorylation of ATM and recruitment of 53BP1 are related, and whether they actually reduce the capacity of LMNA-mutant iPSC-CMs to repair DNA damage when it occurs. Additionally, if reduced phosphorylation of ATM and recruitment of 53BP1 do reduce DNA repair, through what mechanisms do they do so? Although we were not able to detect p53 in our experiments, several studies have implicated defective p53 signaling in DNA damage and repair in laminopathies (Varela et al., 2005; Kudlow et al., 2008; Moiseeva et al., 2011; Chen et al., 2019b; Shao et al., 2020), and as such it is critical to understand whether p53 could be a mediator between altered DDR and defective DNA repair. Finally, future work should understand the consequence of defective DDR in LMNA-mutant iPSC- CMs on apoptosis and senescence, which have long been observed in laminopathic models (Wolf et al., 2008; Gonzalez-Suarez et al., 2011; Redwood et al., 2011; Auguste et al., 2018; Kim et al., 2018; Earle et al., 2019). It remains unclear whether LMNA-mutant iPSC-CMs have increased apoptosis or senescence, or whether when DNA damage does occur, if DDR response and/or repair is altered, whether apoptosis or senescence may occur. As apoptosis and senescence may be key players in the deterioration of cellular health in LMNA-mutant cells and tissues, this relationship must be understood to develop appropriate therapeutics to ameliorate the cellular pathology. 189 A3: Investigating YAP and MKL nuclear translocation and LINC complex organization in LMNA-mutant iPSC-CMs Lamins A and C, encoded by the LMNA gene, form a major part of the nuclear lamina, a dense protein meshwork inside of the inner nuclear membrane that functions to give the nucleus structural support and govern numerous biochemical and mechanical signaling processes. Lamin A/C is critical for the organization of nuclear envelope proteins, including the constituents of the LINC complex (linker of the nucleoskeleton and cytoskeleton), which spans the nuclear envelope and connect to nuclear lamina to the cytoskeleton. As such, disruptions to the structure of the Lamin A/C meshwork, such as by mutations in LMNA, can disrupt the organization of the LINC complex and alter cellular mechanotransduction, the process by which cells converts mechanical stimuli into biochemical signals. Here, using human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) with two different LMNA mutations, R249Q and G449V, which show varying degrees of nuclear damage (Chapter 2) and disrupted gene expression (Chapter 3), I explored import of mechanoresponsive transcription factors YAP and MKL1, and investigated the organization of LINC complex components to determine whether altered mechanotransduction and nucleo-cytoskeletal force transmission may play a role in nuclear damage and/or disrupted cell signaling in LMNA-mutant iPSC-CMs. Introduction Connections between the nuclear lamina and cytoskeleton, or nucleo-cytoskeletal coupling, through the LINC complex (linker of the nucleoskeleton and cytoskeleton), which spans the nuclear envelope (NE), are critical to guide nuclear shape, positioning/movement, and gene 190 expression (Rothballer and Kutay, 2013; Banerjee et al., 2014; Maurer and Lammerding, 2019). The cytoskeleton likely plays a role in the nuclear damage and LMNA mutations may disrupt nucleo-cytoskeletal coupling by mislocalization, loss, or overexpression of several LINC complex proteins or LINC complex-associated proteins, including Emerin, Sun1, and Nesprin-2G (Hale et al., 2008b; Chen et al., 2012; Arsenovic et al., 2016). As an alternative hypothesis to LMNA- mutant iPSC-CM nuclear damage being caused by force transmission to mechanically weakened nuclei, we hypothesized that nucleo-cytoskeletal coupling may be disrupted by disorganization of the LINC complex in LMNA-mutant iPSC-CMs, in turn leading to defective nucleo-cytoskeletal coupling. LINC complex disorganization or displacement and defective transmission of forces to the nucleus could then in turn result in uneven force exertion on the nucleus, causing nuclear shape deformities in LMNA-mutant iPSC-CMs and/or disrupted downstream signaling processes and gene expression. Here, I attempted to quantify the translocation of mechanoresponsive transcription factors into the nucleus and the localization and distribution of LINC complex proteins to the nuclear envelope to test this hypothesis. Results LMNA-mutant iPSC-CMs do not exhibit altered YAP localization To determine whether LMNA-mutant iPSC-CMs have normal localization of mechanoresponsive transcription factors in the nucleus, I quantified nuclear import of YAP in healthy control and LMNA-mutant iPSC-CMs. YAP is known to have mechanoresponsive transport into the nucleus (Driscoll et al., 2015; Elosegui-Artola et al., 2017; Kassianidou et al., 2019), which may be disrupted with LMNA mutations (Bertrand et al., 2014; Owens et al., 2020). I performed immunofluorescence labeling of YAP in healthy control and R249Q mutant iPSC-CMs, as a 191 representative LMNA mutant cell line that has nuclear deformities (Fig. A3.1A) and quantified the ratio of fluorescent intensity of YAP in the nucleus compared to the cytoplasm (Fig. A3.1B). Nuclear:cytoplasmic YAP was not significantly different in R249Q mutant iPSC-CMs compared to healthy controls, indicating that translocation of YAP into the nucleus under normal culture conditions was not affected by the LMNA mutation. To test whether these results also apply to other mechanoresponsive transcription factors, I conducted experiments using immunofluorescence labeling for MKL-1 (MRTF-A), another protein known to be subject to mechanosensitive import into the nucleus (Olson and Nordheim, 2010; Baarlink et al., 2013) that can be disrupted with LMNA deficiency and mutation (Ho et al., 2013), in R249Q mutant iPSC-CMs and healthy controls. However, I was unable to detect any expression of MKL-1 in iPSC-CMs (data not shown). A B Lamin B1 YAP 0.15 0.10 0.05 0.00 Fig. A3.1. YAP nuclear localization is unchanged in R249Q mutant iPSC-CMs. (A) Representative immunofluorescence images of YAP in healthy control (WT) and R249Q-iPSC- CMs. (B) The ratio of nuclear to cytoplasmic YAP is unchanged between WT and R249Q- iPSC-CMs. Scalebar = 50µm. N>138 nuclei per group. 192 R249Q WT1 Nuclear:Cytoplasmic YAP Fluorescent Intensity (a.u.) Quantification of LINC complex localization of iPSC-CMs is inconclusive To determine whether LINC complex proteins were properly localized to and distributed across the nuclear envelope in LMNA-mutant iPSC-CMs (Hale et al., 2008b; Chen et al., 2012; Arsenovic et al., 2016), I examine three LINC complex associated proteins – Emerin, Sun1, and Sun2 – in healthy control and LMNA-mutant iPSC-CMs. Immunofluorescence labeling for Sun-1 showed substantial localization across both the nuclear envelope and the endoplasmic reticulum (ER), where Sun-1 is not typically localized, in both healthy control (Fig. A3.2A) and LMNA-mutant iPSC-CMs. This ER localization of Sun-1 masks the localization to the nucleus and was high variability in localization between cells in healthy controls (Fig. A3.2A). Although immunofluorescence labeling of Sun-1 was inconclusive, to determine whether there were differences in Sun-1 expression, which may be altered by LMNA mutations (Chen et al., 2012), I performed a Western Blot for Sun-1. However, I found no expression of the full-length Sun-1 isoform at its expected size, 90-100kDa, but rather observed highly variable labeling of some truncated isoform, approximately 65kDa in size (Fig. A3.2B), which could potentially explain the unusual localization of Sun-1 in iPSC-CMs. A B WT1 Sun-1 115 80 Lamin A Lamin B1 Lamin C Sun-1 40 GAPDH 25 H3 15 10 Fig. A3.2. iPSC-CMs express a truncated form of Sun-1 that localizes to both the nuclear envelope and endoplasmic reticulum. (A) Representative immunofluorescence image of Sun- 1 labeling in WT1-iPSC-CMs. (B) A Western Blot of two healthy control and two LMNA- mutant iPSC-CMs shows the presence of a truncated Sun-1 isoform around 65kDa rather than the full-length isoform in the 90-100kDa range, and also highly variable expression of truncated Sun-1 with no trends between healthy control and LMNA-mutant cell lines. 193 Since the Sun-1 results were inconclusive, I performed immunofluorescence labeling for Sun-2 and Emerin, as other important constituents of the LINC complex that may affect nuclear mechanotransduction. However, I observed no Sun-2 immunofluorescence labeling across several experiments with antibodies that are recognized to be compatible with human samples. Sun-2 is developmentally regulated during cardiac development, so likely due to the immature nature of our iPSC-CMs, the cells do not yet have Sun-2 expression. Similarly, I observed expression of Emerin in the nucleus but no localization to the nuclear envelope in either healthy control or LMNA-mutant iPSC-CMs, despite the same antibody detecting Emerin localized to the nuclear envelope in a human fibroblast cell line. Together, these data suggest that the LINC complex is more immature in iPSC-CMs compared to mature cardiomyocytes, and thus may not be an ideal system for the study of nuclear mechanotransduction and the role of the LINC complex in LMNA- DCM. Discussion and Conclusion Given that the cytoskeleton is involved in nuclear damage in LMNA-DCM (Hatch and Hetzer, 2016; Xia et al., 2018; Cho et al., 2019; Earle et al., 2019; Heffler et al., 2019) and the importance of nuclear mechanotransduction in maintenance of cellular function (Kirby and Lammerding, 2016, 2018; Maurer and Lammerding, 2019; Donnaloja et al., 2020), it is critical to understand the role of nucleo-cytoskeletal coupling in LMNA-DCM. Here, we utilized healthy control and LMNA- mutant iPSC-CMs to examine nuclear mechanotransduction and LINC complex organization. Through the unchanged nuclear and cytoplasmic distribution of YAP, we determined that there 194 are no changes in translocation of mechanoresponsive transcription factors in LMNA-mutant iPSC- CMs under normal culture conditions, although we could not detect the nuclear and cytoplasmic disrtibution of MKL-1 to confirm these results. However, I observed high variability in nuclear localization of YAP between nuclei, which was likely due to variation in the cytoskeletal and mechanical conformation of iPSC-CMs arranged randomly on an unpatterned substrate. As such, to confirm these results, future studies should be performed either on nanogrooved substrates to improve alignment of iPSC-CMs, or should present some mechanical challenge, such as increased substrate stiffness, as a positive control to determine that LMNA-mutant iPSC-CMs have intact mechanosensitive import of YAP into the nucleus. To determine whether the LINC complex remains organized in LMNA-mutant iPSC-CMs, we examined the localization of Emerin, Sun-1, and Sun-2, which may be affected by LMNA mutations (Hale et al., 2008b; Chen et al., 2012; Arsenovic et al., 2016). However, we could not detect Sun-2 in iPSC-CMs, Emerin was not localized to the nuclear envelope, and Sun-1 was expressed only as a truncated isoform and had unusual localization to both the nuclear envelope and the ER. Together, these results suggested that the LINC complex of iPSC-CMs may be immature compared to adult cardiomyocytes, and therefore iPSC-CMs may not be a sufficient model of LMNA-DCM to examine changes to the LINC complex and nuclear mechanotransduction. 195 A4: Design and validation of engineered tissue systems for the study of laminopathies5 Mutations in the LMNA gene, encoding Lamin A/C, give rise to a collection of human diseases, termed ‘laminopathies,’ that primarily affect skeletal muscle and the heart. Progress towards understanding the cellular pathogenesis of human skeletal muscle and cardiac laminopathies has been limited by the use of 2D in vitro systems, which do not recapitulate the structure and mechanical environment of native tissue, and in vivo mouse models, which do not fully mimic all aspects of human disease. Thus, 3D systems for the study of cardiac and skeletal muscle laminopathies are critical. Here, we present a method of high-throughput engineered skeletal muscle and cardiac tissues for the study of laminopathies. We demonstrated that engineered skeletal muscle tissues comprised of differentiated mouse primary myoblasts undergo strong myogenic differentiation, and Lmna-deficient engineered skeletal muscle tissues exhibit nuclear damage in the form of elongated nuclei and protrusions of chromatin into the cytoplasm. Future work will examine the distribution of nuclear damage and DNA damage across different regions of the tissues, to mimic the increased damage of high-force regions of the muscle and will extend these studies to cardiac tissues comprised of human LMNA-mutant iPSC-CMs. Introduction Mutations in the gene LMNA, encoding nuclear Lamin A/C, give rise about 15 human diseases, 5 This work was complete in collaboration with several members of the Lammerding Lab. Maurer, Melanie, Peng, Huaiyao, Johnson, Lindsey, Kirby, Tyler, Lammerding, Jan. MM, TK, and JL contributed to the conception and design of the work. MM, HP, LJ, and TK contributed to data acquisition and analysis. MM, HP, TK, and JL contributed to interpretation of the data. MM and JL contributed to the drafting of this chapter. 196 termed ‘laminopathies,’ which include dilated cardiomyopathy (DCM) and several forms of muscular dystrophy. While substantial progress has been made to understand the disease mechanisms of straited muscle laminopathies, studies have primarily relied on 2D cell culture, which enables the study of nuclear mechanics but fails to recapitulate the maturity and architecture of native tissue, and mouse models, which do not fully mimic human disease presentation and progression (Stewart et al., 2007) and limit nuclear mechanical studies. Additionally, as novel therapeutics for laminopathies emerge (Cho et al., 2019; Earle et al., 2019; Chai et al., 2021), there remains a large gap between testing therapies in human in vitro or mouse in vivo models and in humans. One emerging strategy to overcome these disease modeling challenges is to employ engineered tissues, in which cells compact extracellular matrix (ECM) around two flexible pillars to form tissues that exhibit improved maturity and tissue structure compared to 2D in vitro culture (Legant et al., 2009; Sakar et al., 2012; Tiburcy et al., 2017; Maffioletti et al., 2018; Ronaldson-Bouchard et al., 2018, 2019). Additionally, engineered tissues have several distinct advantages over in vivo tissue studies, such as the ability to perform live-cell imaging and measure tissue-generated forces (Legant et al., 2009) and the ability to understand tissue response to electrical (Ronaldson- Bouchard et al., 2018, 2019), mechanical (Zhao et al., 2013), or biochemical stimulation. Recent studies have demonstrated that engineered skeletal muscle tissues recapitulate the hallmark nuclear deformities of EDMD (Maffioletti et al., 2018; Steele-Stallard et al., 2018), indicating that tissue engineered approaches are promising for the study of disease mechanisms in laminopathies. Here, we describe a system for the study of tissue engineered skeletal muscle and cardiac 197 laminopathies, “microtissues,” in which arrays of 500 μm-long tissues allow for high-throughput studies (Legant et al., 2009). We describe the generation of both engineered cardiac tissues from human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) with LMNA mutations and engineered skeletal muscle tissues from Lmna-deficient (Lmna KO) primary mouse myoblasts, and present promising preliminary results showing that (Lmna KO) engineered skeletal muscle exhibit nuclear damage. Results Generation of engineered cardiac and skeletal muscle microtissues Molds to generate microtissue arrays were generously provided by the Chen lab. Microtissue arrays were manufactured by stamping a negative mold of the wells into a poly-dimethysiloxane (PDMS)-filled 35 mm dish and contain about one hundred individual wells containing two pillars, spaced about 500 μm apart (Fig. A4.1). Fig. A4.1. A size comparison of micro- and microtissue devices for generation of engineered tissues. Molds for macrotissues are shown on the top of the figure, in which cells and ECM are seeded into the yellow well areas. A microtissue array is shown in a 35 mm cell culture dish on the bottom of the figure, in which cells and ECM are seeded into the blue areas. A zoomed-in image of four microtissues in the array is depicted on the right. 198 Engineered skeletal muscle and cardiac tissue matrix and cell seeding conditions were optimized by an M.Eng. student, Huaiyao Peng, and a former undergraduate student, Lindsey Johnson, respectively. The complete cell seeding protocol is included in the Supplementary Material in this chapter. Generally, an ECM solution optimized for the type of tissue (Table A4.1) is placed on the top of the device and centrifuged down into wells, and then cells in ECM solution are placed on the top of the device and centrifuged into wells. For engineered skeletal muscle tissues, primary myoblasts (650,000 for healthy control and 800,000 for Lmna-deficient (Lmna KO) to account for increased cell death in Lmna KO engineered tissues) are seeded into wells, and for engineered cardiac tissues, 1.2 million cells of a 70:30 mixture of human iPSC-CMs and human foreskin fibroblasts are seeded into wells. ECM is allowed to polymerize for one hour, and media containing a protease inhibitor, Aprotinin) is added to devices to inhibit matrix degradation. Tissues have generally compacted around pillars within 24-hours of seeding. Myogenic differentiation is induced one day after seeding for engineered skeletal muscle tissues according to a published protocol (Pimentel et al., 2017; Earle et al., 2019). Tissues generally begin to contract after about five days after seeding, and tissues are used for endpoint studies, such as immunofluorescence staining, eight days after seeding, after which tissue health begins to decline. Table A4.1. Extracellular matrix compositions for engineered cardiac and skeletal muscle tissu es. Cardiac Tissue Concentration Skeletal Muscle Tissue Concentration Media NaHCO3 5% w/v 5% w/v NaOH 0.02 μL/1 mL collagen 0.02 μL/1 mL collagen Collagen 2.25 mg/mL 1.5 mg/mL Fibrinogen 0.5 mg/mL - Matrigel - 20% 199 Engineered muscle tissues show myotube formation and alignment Healthy control engineered muscle microtissues immunofluorescence labeled with myosin heavy chain showed strong myogenic differentiation and alignment of myofibers to the long axis of the tissue (Fig A4.2), indicating successful myogenic differentiation and tissue maturation. MHC ,Lamin B1, Hoechst Fig. A4.2. A representative engineered skeletal muscle microtissue from healthy control tissues after eight days of differentiation and maturation. A representative maximum intensity projection of a healthy control engineered skeletal muscle microtissue shows high expression of myosin heavy chain (MHC), indicating a high degree of myogenic differentiation and alignment of myofibers to the long axis of the tissue. Pillars are indicated by white dashed boxes. Lmna KO engineered skeletal muscle microtissues exhibit nuclear deformities, DNA damage, and apoptosis To confirm that engineered skeletal muscle microtissues derived from Lmna KO myoblasts exhibit nuclear damage, a hallmark of skeletal muscle laminopathies (Earle et al., 2019), we immunofluorescently labeled Lamin A/C and stained for DNA with DAPI in healthy control and Lmna KO microtissues matured to eight days post-seeding to examine nuclear morphology. As expected, we observed normal nuclear morphology in healthy control microtissues (Fig. A4.3), and elongated nuclei with accompanying protrusions of DNA into the cytoplasm in Lmna KO microtissues, as is observed in both 2D myotube culture and in vivo (Earle et al., 2019). These 200 results indicate that Lmna KO microtissues indeed recapitulate at least some aspects of skeletal muscle laminopathies. Fig. A4.3. Lmna KO (Lmna–/–) microtissues show elongated nuclear morphology and DNA protrusions from the nucleus. Representative immunofluorescence images labeled for Lamin A/C and DAPI of (Left) microtissues derived from healthy control (Lmna+/+) myoblasts show normal rounded nuclear morphology, whereas (right) microtissues derived from Lmna KO myoblasts show nuclear elongation and the protrusion of DNA out of the nucleus, indicated by white arrows. It should be noted that the Lamin A/C antibody used in this study recognizes the truncated Lamin A/C isoform expressed in Lmna KO mice. Discussion and Conclusion Here, we present optimized protocols for the generation of engineered cardiac and skeletal muscle microtissues. We have demonstrated that engineered skeletal muscle microtissues differentiate and form aligned, and mature tissues in the span of eight days, and that engineered skeletal muscle microtissues derived from Lmna KO myoblasts have elongated nuclei and protrusions of DNA from the nucleus, hallmarks of muscular dystrophy caused by Lmna mutations that are in line with other studies of Lmna-mutant engineered skeletal muscle tissues (Maffioletti et al., 2018; Steele- Stallard et al., 2018; Earle et al., 2019). 201 The proof-of-concept experiments presented here pave the way for ongoing work carried out by an M.Eng. student in the Lammerding Lab, Huaiyao Peng, to validate engineered skeletal muscle microtissues as a model for Lmna skeletal muscular dystrophies. Current studies are underway to quantify increased nuclear damage, DNA damage, cell death, and nuclear envelope rupture in Lmna KO skeletal muscle microtissues, which represent pathogenic mechanisms previously demonstrated in both in vivo and in vitro models of Lmna skeletal muscular dystrophies in the Lammerding lab (Earle et al., 2019). One exciting goal of these engineered skeletal muscle tissues is to bridge a gap in results between previous in vitro and in vivo studies: increased nuclear damage, nuclear envelope rupture, and DNA damage occur in the myotendinous junction (MTJ) of skeletal muscle compared to the muscle body, likely due to higher forces experienced by nuclei near the MTJ (Earle et al., 2019), which has not been able to replicated in vitro studies as force is more uniformly distributed in 2D culture. However, the regions of engineered tissues near the pillars are subject to increased mechanical forces (Legant et al., 2009), and as such, we hypothesize that we will be able to replicate results showing that myonuclei in higher force regions of the tissue are subject to increased damage. We will quantify the distribution of nuclear damage and cell death across engineered skeletal muscle tissues to test this hypothesis. In addition, future work should continue with the engineered cardiac tissues using LMNA-mutant iPSC-CMs. While our previous results in Chapter 2 demonstrated that LMNA-mutant iPSC-CMs have increased nuclear damage and fragility that can be explained by lamin mislocalization from the nuclear envelope are exciting, these studies are limited in that iPSC-CMs are relatively immature (Scuderi and Butcher, 2017) and 2D culture does not mimic the natural myocardium, so iPSC-CMs experience unrealistic environmental forces. As our understanding that cytoskeletal 202 forces exerted on the nucleus cause nuclear damage and decline in cellular health in striated muscle laminopathies (Cho et al., 2019; Earle et al., 2019; Chai et al., 2021), it is critical to understand the role for cell-generated and microenvironmental forces in a realistic context. Together, this chapter represents exciting progress towards one of the few tissue engineering approaches for the study of laminopathies (Maffioletti et al., 2018; Steele-Stallard et al., 2018). These results demonstrate that engineered skeletal muscle microtissues are differentiated and mature, and Lmna KO muscle tissues matured to eight days post-seeding in microtissue arrays exhibit the hallmark nuclear damage of laminopathies. Future studies will work towards demonstrating that engineered muscle microtissues are a valid in vitro model that can recapitulate disease aspects of in vivo tissue that had previously not been able to be demonstrated in vitro. 203 Supplementary Material Supplementary methods Seeding protocol for arrays of engineered microtissues. 1. Sonicate devices for 5mins in 70% Ethanol. 2. Dry devices with compressed air. 3. Fill device with 70% EtOH (about 700uL). Pipette up and down over wells using a P1000 pipette to make sure EtOH goes into wells (bubbles coming out will be visible). Check that bubbles are removed, and if they are not centrifuge 1500rpm for 2 mins. Incubate at RT for 30mins. 4. Check that large centrifuge in entrance hallway is cooled to 0oC. PDMS inserts should be placed in the well-plate baskets in the centrifuge. 5. Aspirate EtOH. 6. Dry device with compressed air in hood 7. Fill device with 0.2% pluronic, and pipette up and down as in step 3 to remove bubbles. Check that bubbles are removed, and if they are not centrifuge 1500 rpm for 2 mins. Incubate at RT for 30 mins. 8. Place collagen, medium (RPMI for cardiac tissues, IMDM for muscle tissues), NaOH, HEPES, NaHCO3, Fibrinogen and two eppi tubes on ice. 9. Prepare 1mL/device collagen solution, according to the following table: For cardiac tissues: C onc. Stock conc. Volume (μL, per 1mL) Media 226.18 NaHCO3 5% w/v 0.1 M 12.5 NaOH 0.02 uL/1 mL collagen 1 M 1.32 Collagen 2.25 mg/mL 3.17 mg/mL 710 Fibrinogen 0.5 mg/mL 10 mg/mL 50 For muscle tissues: C onc. Stock conc. Volume (μL, per 1mL) Media 586 NaHCO3 5% w/v 0.1 M 12.5 NaOH 0.02 μL/1 mL collagen 1 M 1.32 Collagen 2.25 mg/mL 10.21 mg/mL 221 Matrigel 20% 1:1 333 10. Passage cells. 11. For cardiac tissues, pipette iPSC-CMs through a cell strainer 12. Count cells a. For cardiac tissues, prepare a mixture of 1.2 million total of 70:30 iPSC- CMs:HFFs b. For muscle tissues, determine the volume needed for WT (650,000 cells) or Lmna KO (800,000 cells) 13. Centrifuge for 4 mins at 250 g 204 14. Fill device with 2% pluronic, and pipette up and down as in step 3 to remove bubbles. Check that bubbles are removed, and if they are not centrifuge 1500 rpm for 2 mins. Incubate at RT for 5 mins. 15. Wash with sterile 1× PBS, pipetting into wells as in step 1. 16. Aspirate PBS. 17. Using compressed air can with straw nozzle, dry device with compressed air until all PBS is out of wells (check under microscope). 18. Place device on ice 19. When device is cooled, pipette 700 μL collagen solution over wells of device, being careful not to create bubbles. 20. Centrifuge for 90 seconds at 1500 rpm and 0oC using the PDMS plate inserts 21. Store device in back of cell culture fridge until ready for cell seeding. 22. Place cell suspension on ice, and place device on ice. 23. Resuspend cell mixture with 300 μL of matrix solution, being careful to avoid intriducing bubbles 24. Pipette cell-matrix mixture onto device, being careful to spread cells evenly over wells. Pipette up and down a few times to spread cells more, but only if no bubbles will be introduced into solution. 25. Centrifuge for 1.5 mins at 1500 rpm and 0oC. 26. Rotate 90o, and centrifuge again for 1.5 mins 27. Check that cells were spun down into wells. 28. De-wetting of device to remove extra cells and matrix a. Tilt the plate towards you so extra matrix runs toward bottom b. Using a 2 μL pipette tip on tip of an aspiration pipette (to help improve accuracy), aspirate along the top edge of the device until you see some space between the edge and the matrix c. Continue to tilt the device and wait for get to run down to the bottom edge. d. VERY SLOWLY remove pooled gel from the bottom corners. Continue to aspriate until all the matrix has run down/basically all is removed e. If collagen is taking a long time, just be patient! 29. Centrifuge device upside down for 15 seconds at 500rpm and 0oC. Keep device inverted! 30. Pipette a large drop of sterile water onto lid of device. Polymerize device upside down for 30 mins in 37oC incubator. Place any remaining collagen matrix in the incubator to check for polymerization. 31. Aspirate water. 32. Flip device right-side up and add 2 mL of media (B27 + P/S + Y27632 for cardiac tissues, IMDM for muscle tissues) plus 800ng/mL Aprotinin to device. Add VERY slowly directly on top of pillars. (adding on the PDMS on the side will shear tissues out of wells) 33. Change media every other day. 205 REFERENCES Abilez, OJ et al. (2018). Passive Stretch Induces Structural and Functional Maturation of Engineered Heart Muscle as Predicted by Computational Modeling. Stem Cells 36, 265– 277. Ahn, J, Lee, J, Jeong, S, Kang, S mi, Park, BJ, and Ha, NC (2021). Beta-strand-mediated dimeric formation of the Ig-like domains of human lamin A/C and B1. Biochem Biophys Res Commun 550, 191–196. Al-Haboubi, T, Shumaker, DK, Köser, J, Wehnert, M, and Fahrenkrog, B (2011). Distinct association of the nuclear pore protein Nup153 with A- and B-type lamins. Nucleus 2, 500–509. Alexa, A, and Rahnenfuhrer, J (2021). topGO: Enrichment Analysis for Gene Ontology. R Packag Version 2440. Andrew, S (2010). FastQC: A Quality Control Tool for High Throughput Sequence Data [Online]. Arsenovic, PT, Ramachandran, I, Bathula, K, Zhu, R, Narang, JD, Noll, NA, Lemmon, CA, Gundersen, GG, and Conway, DE (2016). Nesprin-2G, a Component of the Nuclear LINC Complex, Is Subject to Myosin-Dependent Tension. Biophys J 110, 34–43. Auguste, G, Gurha, P, Lombardi, R, Coarfa, C, Willerson, JT, and Marian, AJ (2018). Suppression of Activated FOXO Transcription Factors in the Heart Prolongs Survival in a Mouse Model of Laminopathies. Circ Res 122, 678–692. Azibani, F, Muchir, A, Vignier, N, and Bertrand, AT (2014). Striated muscle laminopathies. Semin Cell Dev Biol 29, 107–115. Baarlink, C, Wang, H, and Grosse, R (2013). Nuclear Actin Network Assembly by Formins ccvi Regulates the SRF Coactivator MAL. Science 340, 864–867. Balikov, DA, Brady, SK, Ko, UH, Shin, JH, de Pereda, JM, Sonnenberg, A, Sung, H-J, and Lang, MJ (2017). The nesprin-cytoskeleton interface probed directly on single nuclei is a mechanically rich system. Nucleus 8, 534–547. Banerjee, I et al. (2014). Targeted Ablation of Nesprin 1 and Nesprin 2 from Murine Myocardium Results in Cardiomyopathy, Altered Nuclear Morphology and Inhibition of the Biomechanical Gene Response. PLoS Genet 10, e1004114. Bártová, E, Legartová, S, Dundr, M, and Suchánková, J (2019). A role of the 53BP1 protein in genome protection: structural and functional characteristics of 53BP1-dependent DNA repair. Aging (Albany NY) 11, 2488. Bell, ES et al. (2021). Low lamin A levels enhance confined cell migration and metastatic capacity in breast cancer. BioRxiv, 2021.07.12.451842. Bera, M, Kotamarthi, HC, Dutta, S, Ray, A, Ghosh, S, Bhattacharyya, D, Ainavarapu, SRK, and Sengupta, K (2014). Characterization of Unfolding Mechanism of Human Lamin A Ig Fold by Single-Molecule Force Spectroscopy—Implications in EDMD. Biochemistry 53, 7247–7258. Van Berlo, JH, Voncken, JW, Kubben, N, Broers, JLV, Duisters, R, van Leeuwen, REW, Crijns, HJGM, Ramaekers, FCS, Hutchison, CJ, and Pinto, YM (2005). A-type lamins are essential for TGF-β1 induced PP2A to dephosphorylate transcription factors. Hum Mol Genet 14, 2839–2849. Bernasconi, P et al. (2018). Elevated TGF β2 serum levels in Emery-Dreifuss Muscular Dystrophy: Implications for myocyte and tenocyte differentiation and fibrogenic processes. Nucleus 9, 292–304. ccvii Bertero, A et al. (2019). Chromatin compartment dynamics in a haploinsufficient model of cardiac laminopathy. J Cell Biol 218, 2919–2944. Bertrand, AT et al. (2014). Cellular microenvironments reveal defective mechanosensing responses and elevated YAP signaling in LMNA-mutated muscle precursors. J Cell Sci 127, 2873–2884. Bertrand, AT, Brull, A, Azibani, F, Benarroch, L, Chikhaoui, K, Stewart, CL, Medalia, O, Ben Yaou, R, and Bonne, G (2020). Lamin A/C Assembly Defects in LMNA-Congenital Muscular Dystrophy Is Responsible for the Increased Severity of the Disease Compared with Emery–Dreifuss Muscular Dystrophy. Cells 9, 844. Bhattacharya, D, Talwar, S, Mazumder, A, and Shivashankar, GV (2009). Spatio-Temporal Plasticity in Chromatin Organization in Mouse Cell Differentiation and during Drosophila Embryogenesis. Biophys J 96, 3832–3839. Blighe, K, Rana, S, and Lewis, M (2021). EnhancedVolcano: Publication-ready volcano plots with enhanced colouring and labeling. R Packag Version 1100. Bolte, S, and Cordelières, F (2006). A guided tour into subcellular colocalization analysis in light microscopy. J Microsc 224, 213–232. Broers, JLV, Peeters, EAG, Kuijpers, HJH, Endert, J, Bouten, CVC, Oomens, CWJ, Baaijens, FPT, and Ramaekers, FCS (2004). Decreased mechanical stiffness in LMNA-/- cells is caused by defective nucleo-cytoskeletal integrity: Implications for the development of laminopathies. Hum Mol Genet 13, 2567–2580. Buchwalter, AL, Liang, Y, and Hetzer, MW (2014). Nup50 is required for cell differentiation and exhibits transcription-dependent dynamics. Mol Biol Cell 25, 2472–2484. Burke, MA et al. (2016). Molecular profiling of dilated cardiomyopathy that progresses to heart ccviii failure. JCI Insight 1. Buxboim, A et al. (2014). Matrix Elasticity Regulates Lamin-A,C Phosphorylation and Turnover with Feedback to Actomyosin. Curr Biol 24, 1909–1917. Cai, ZJ et al. (2020). Expression of Lmna-R225X nonsense mutation results in dilated cardiomyopathy and conduction disorders (DCM-CD) in mice: Impact of exercise training. Int J Cardiol 298, 85–92. Carlson, M (2019a). org.Hs.eg.db: Genome wide annotation for Human. R Packag Version 382. Carlson, M (2019b). org.Mm.eg.db: Genome wide annotation for Mouse. R Packag Version 382. Cattin, ME et al. (2013). Heterozygous LmnadelK32 mice develop dilated cardiomyopathy through a combined pathomechanism of haploinsufficiency and peptide toxicity. Hum Mol Genet 22, 3152–3164. Cenni, V et al. (2005). Lamin A N-terminal phosphorylation is associated with myoblast activation: impairment in Emery-Dreifuss muscular dystrophy. J Med Genet 42, 214– 220. Chai, RJ et al. (2021). Disrupting the LINC complex by AAV mediated gene transduction prevents progression of Lamin induced cardiomyopathy. Nat Commun 12, 1–16. Chandar, S et al. (2010). Effects of mechanical stress and carvedilol in lamin A/C-deficient dilated cardiomyopathy. Circ Res 106, 573–582. Chaudhuri, O et al. (2016). Hydrogels with tunable stress relaxation regulate stem cell fate and activity. Nat Mater 15, 326–334. Cheedipudi, SM et al. (2019). Genomic Reorganization of Lamin-Associated Domains in Cardiac Myocytes is Associated with Differential Gene Expression and DNA Methylation in Human Dilated Cardiomyopathy. Circ Res 124, 1198. ccix Chen, C-Y, Chi, Y-H, Mutalif, RA, Starost, MF, Myers, TG, Anderson, SA, Stewart, CL, and Jeang, K-T (2012). Accumulation of the Inner Nuclear Envelope Protein Sun1 Is Pathogenic in Progeric and Dystrophic Laminopathies. Cell 149, 565–577. Chen, NY et al. (2019a). An absence of lamin B1 in migrating neurons causes nuclear membrane ruptures and cell death. Proc Natl Acad Sci, 201917225. Chen, NY, Kim, P, Weston, TA, Edillo, L, Tu, Y, Fong, LG, and Young, SG (2018). Fibroblasts lacking nuclear lamins do not have nuclear blebs or protrusions but nevertheless have frequent nuclear membrane ruptures. Proc Natl Acad Sci, 201812622. Chen, SN, Lombardi, R, Karmouch, J, Tsai, JY, Czernuszewicz, G, Taylor, MRG, Mestroni, L, Coarfa, C, Gurha, P, and Marian, AJ (2019b). DNA Damage Response/TP53 Pathway Is Activated and Contributes to the Pathogenesis of Dilated Cardiomyopathy Associated With LMNA (Lamin A/C) Mutations. Circ Res. Cheng, Q, and Chen, J (2010). Mechanism of p53 stabilization by ATM after DNA damage. Cell Cycle 9, 472. Cho, S et al. (2019). Mechanosensing by the Lamina Protects against Nuclear Rupture, DNA Damage, and Cell-Cycle Arrest. Dev Cell. Cho, S, Irianto, J, and Discher, DE (2017). Mechanosensing by the nucleus: From pathways to scaling relationships. J Cell Biol 216, 305–315. Cho, W, and Stahelin, R V. (2005). Membrane-Protein Interactions in Cell Signaling and Membrane Trafficking. Annu Rev Biophys Biomol Struct 34, 119–151. Chowdhury, F, Na, S, Li, D, Poh, Y-C, Tanaka, TS, Wang, F, and Wang, N (2010). Material properties of the cell dictate stress-induced spreading and differentiation in embryonic stem cells. Nat Mater 9, 82–88. ccx Chung, JH (2018). The role of DNA-PK in aging and energy metabolism. FEBS J 285, 1959– 1972. Constantinescu, D, Gray, HL, Sammak, PJ, Schatten, GP, and Csoka, AB (2006). Lamin A/C Expression Is a Marker of Mouse and Human Embryonic Stem Cell Differentiation. Stem Cells 24, 177–185. Crisp, M, Liu, Q, Roux, K, Rattner, JB, Shanahan, C, Burke, B, Stahl, PD, and Hodzic, D (2006). Coupling of the nucleus and cytoplasm: role of the LINC complex. J Cell Biol 172, 41–53. Dahl, KN, Kahn, SM, Wilson, KL, and Discher, DE (2004). The nuclear envelope lamina network has elasticity and a compressibility limit suggestive of a molecular shock absorber. J Cell Sci 117, 4779–4786. Davidson, PM, and Lammerding, J (2014). Broken nuclei - lamins, nuclear mechanics, and disease. Trends Cell Biol 24, 247–256. Dechat, T, Adam, SA, Taimen, P, Shimi, T, and Goldman, RD (2010). Nuclear lamins. Cold Spring Harb Perspect Biol 2, 1–23. Dekker, J et al. (2017). The 4D nucleome project. Nature 549, 219–226. Denais, CM, Gilbert, RM, Isermann, P, McGregor, AL, Te Lindert, M, Weigelin, B, Davidson, PM, Friedl, P, Wolf, K, and Lammerding, J (2016). Nuclear envelope rupture and repair during cancer cell migration. Science (80- ). Dobin, A, Davis, CA, Schlesinger, F, Drenkow, J, Zaleski, C, Jha, S, Batut, P, Chaisson, M, and Gingeras, TR (2013). STAR: Ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21. Donnaloja, F, Carnevali, F, Jacchetti, E, and Raimondi, MT (2020). Lamin A/C ccxi Mechanotransduction in Laminopathies. Cells 9, 1306. Dorner, D, Gotzmann, J, and Foisner, R (2007). Nucleoplasmic lamins and their interaction partners, LAP2α, Rb, and BAF, in transcriptional regulation. FEBS J 274, 1362–1373. Le Dour, C, Macquart, C, Sera, F, Homma, S, Bonne, G, Morrow, JP, Worman, HJ, and Muchir, A (2017). Decreased WNT/β-catenin signalling contributes to the pathogenesis of dilated cardiomyopathy caused by mutations in the lamin a/C gene. Hum Mol Genet 26, 333– 343. Doyle, AD, and Yamada, KM (2016). Mechanosensing via cell-matrix adhesions in 3D microenvironments. Exp Cell Res 343, 60–66. Driscoll, TP, Cosgrove, BD, Heo, S-J, Shurden, ZE, and Mauck, RL (2015). Cytoskeletal to Nuclear Strain Transfer Regulates YAP Signaling in Mesenchymal Stem Cells. Biophys J 108, 2783–2793. Dubik, N, and Mai, S (2020). Lamin A/C: Function in Normal and Tumor Cells. Cancers 2020, Vol 12, Page 3688 12, 3688. Dupont, S et al. (2011). Role of YAP/TAZ in mechanotransduction. Nature 474, 179–183. Duque, G, and Rivas, D (2006). Age-related changes in lamin A/C expression in the osteoarticular system: Laminopathies as a potential new aging mechanism. Mech Ageing Dev 127, 378–383. Dutta, S, Das, JK, Maganti, L, Bhattacharyya, M, Bhattacharyya, D, Mukherjee, S, and Sengupta, K (2018). Skeletal Muscle Dystrophy mutant of lamin A alters the structure and dynamics of the Ig fold domain. Sci Reports 2018 81 8, 1–16. Dyer, JA, Kill, IR, Pugh, G, Quinlan, RA, Lane, EB, and Hutchison, CJ (1997). Cell cycle changes in A-type lamin associations detected in human dermal fibroblasts using ccxii monoclonal antibodies. Chromosom Res 1997 56 5, 383–394. Earle, AJ, Kirby, TJ, Fedorchak, GR, Isermann, P, Patel, J, Iruvanti, S, Moore, SA, Bonne, G, Wallrath, LL, and Lammerding, J (2019). Mutant lamins cause nuclear envelope rupture and DNA damage in skeletal muscle cells. Nat Mater 19, 464–473. Elosegui-Artola, A et al. (2017). Force Triggers YAP Nuclear Entry by Regulating Transport across Nuclear Pores. Cell 171, 1397-1410.e14. Engler, AJ, Sen, S, Sweeney, HL, and Discher, DE (2006). Matrix Elasticity Directs Stem Cell Lineage Specification. Cell 126, 677–689. Enyedi, B, Jelcic, M, and Niethammer, P (2016). The Cell Nucleus Serves as a Mechanotransducer of Tissue Damage-Induced Inflammation. Cell 165, 1160–1170. Enyedi, B, and Niethammer, P (2017). Nuclear membrane stretch and its role in mechanotransduction. Nucleus 8, 156–161. Favreau, C, Higuet, D, Courvalin, J-C, and Buendia, B (2004). Expression of a mutant lamin A that causes Emery-Dreifuss muscular dystrophy inhibits in vitro differentiation of C2C12 myoblasts. Mol Cell Biol 24, 1481–1492. Fedorchak, G, and Lammerding, J (2016). Cell Microharpooning to Study Nucleo-Cytoskeletal Coupling. Methods Mol Biol 1411, 241–254. Fidziańska, A, Bilińska, ZT, Tesson, F, Wagner, T, Walski, M, Grzybowski, J, Rużyłło, W, and Hausmanowa-Petrusewicz, I (2008). Obliteration of cardiomyocyte nuclear architecture in a patient with LMNA gene mutation. J Neurol Sci 271, 91–96. Folker, ES, Ostlund, C, Luxton, GWG, Worman, HJ, and Gundersen, GG (2011). Lamin A variants that cause striated muscle disease are defective in anchoring transmembrane actin-associated nuclear lines for nuclear movement. Proc Natl Acad Sci U S A 108, 131– ccxiii 136. Frock, RL, Kudlow, BA, Evans, AM, Jameson, SA, Hauschka, SD, and Kennedy, BK (2006). Lamin A/C and emerin are critical for skeletal muscle satellite cell differentiation. Genes Dev 20, 486–500. Furusawa, T, Rochman, M, Taher, L, Dimitriadis, EK, Nagashima, K, Anderson, S, and Bustin, M (2015). Chromatin decompaction by the nucleosomal binding protein HMGN5 impairs nuclear sturdiness. Nat Commun 6, 6138. Galata, Z, Kloukina, I, Kostavasili, I, Varela, A, Davos, CH, Makridakis, M, Bonne, G, and Capetanaki, Y (2018). Amelioration of desmin network defects by αB-crystallin overexpression confers cardioprotection in a mouse model of dilated cardiomyopathy caused by LMNA gene mutation. J Mol Cell Cardiol 125, 73–86. Georgescu, W, Osseiran, A, Rojec, M, Liu, Y, Bombrun, M, Tang, J, and Costes, S V. (2015). Characterizing the DNA Damage Response by Cell Tracking Algorithms and Cell Features Classification Using High-Content Time-Lapse Analysis. PLoS One 10, e0129438. Gerace, L, Blum, A, and Blobel, G (1978). Immunocytochemical localization of the major polypeptides of the nuclear pore complex-lamina fraction: Interphase and mitotic distribution. J Cell Biol 79, 546–566. Gesson, K, Vidak, S, and Foisner, R (2014). Lamina-associated polypeptide (LAP)2α and nucleoplasmic lamins in adult stem cell regulation and disease. Semin Cell Dev Biol 29, 116–124. Gilbert, HTJ, Mallikarjun, V, Dobre, O, Jackson, MR, Pedley, R, Gilmore, AP, and Swift, J (2018). Nuclear decoupling, through regulation of the LINC complex, is part of a rapid, ccxiv protein-level cellular response to high-intensity mechanical loading. BioRxiv. Gobbi, M De et al. (2002). Natural history of juvenile haemochromatosis. Br J Haematol 117, 973–979. Goldman, RD, Gruenbaum, Y, Moir, RD, Shumaker, DK, and Spann, TP (2002). Nuclear lamins: building blocks of nuclear architecture. Genes Dev 16, 533–547. Gonzalez-Suarez, I, Redwood, AB, Grotsky, DA, Neumann, MA, Cheng, EH-Y, Stewart, CL, Dusso, A, and Gonzalo, S (2011). A new pathway that regulates 53BP1 stability implicates Cathepsin L and vitamin D in DNA repair. EMBO J 30, 3383–3396. González, JM, Navarro-Puche, A, Casar, B, Crespo, P, and Andrés, V (2008). Fast regulation of AP-1 activity through interaction of lamin A/C, ERK1/2, and c-Fos at the nuclear envelope. J Cell Biol 183, 653–666. Gotic, I, Schmidt, WM, Biadasiewicz, K, Leschnik, M, Spilka, R, Braun, J, Stewart, CL, and Foisner, R (2010). Loss of LAP2α delays satellite cell differentiation and affects postnatal fiber-type determination. Stem Cells 28, 480–488. Grashoff, C et al. (2010). Measuring mechanical tension across vinculin reveals regulation of focal adhesion dynamics. Nature 466, 263–266. Gruenbaum, Y, and Foisner, R (2015). Lamins: Nuclear Intermediate Filament Proteins with Fundamental Functions in Nuclear Mechanics and Genome Regulation. Annu Rev Biochem 84, 131–164. Guilluy, C, and Burridge, K (2015). Nuclear mechanotransduction: Forcing the nucleus to respond. Nucleus 6, 19–22. Guilluy, C, Osborne, LD, Van Landeghem, L, Sharek, L, Superfine, R, Garcia-Mata, R, and Burridge, K (2014). Isolated nuclei adapt to force and reveal a mechanotransduction ccxv pathway in the nucleus. Nat Cell Biol 16, 376–381. Guo, Y, Kim, Y, Shimi, T, Goldman, RD, and Zheng, Y (2014). Concentration-dependent lamin assembly and its roles in the localization of other nuclear proteins. Mol Biol Cell 25, 1287–1297. Gupta, P et al. (2010). Genetic and ultrastructural studies in dilated cardiomyopathy patients: a large deletion in the lamin A/C gene is associated with cardiomyocyte nuclear envelope disruption. Basic Res Cardiol 105, 365–377. Gurudatta, B V., Shashidhara, LS, and Parnaik, VK (2010). Lamin C and chromatin organization in Drosophila. J Genet 89, 37–49. Guz, N, Dokukin, M, Kalaparthi, V, and Sokolov, I (2014). If Cell Mechanics Can Be Described by Elastic Modulus: Study of Different Models and Probes Used in Indentation Experiments. Biophys J 107, 564–575. Hale, CM, Shrestha, AL, Khatau, SB, Stewart-Hutchinson, PJ, Hernandez, L, Stewart, CL, Hodzic, D, and Wirtz, D (2008a). Dysfunctional connections between the nucleus and the actin and microtubule networks in laminopathic models. Biophys J 95, 5462–5475. Hale, CM, Shrestha, AL, Khatau, SB, Stewart-Hutchinson, PJ, Hernandez, L, Stewart, CL, Hodzic, D, and Wirtz, D (2008b). Dysfunctional connections between the nucleus and the actin and microtubule networks in laminopathic models. Biophys J 95, 5462–5475. Halfmann, CT, Sears, RM, Katiyar, A, Busselman, BW, Aman, LK, Zhang, Q, O’Bryan, CS, Angelini, TE, Lele, TP, and Roux, KJ (2019). Repair of nuclear ruptures requires barrier- to-autointegration factor. J Cell Biol, jcb.201901116. Harkness, T, McNulty, JD, Prestil, R, Seymour, SK, Klann, T, Murrell, M, Ashton, RS, and Saha, K (2015). High-content imaging with micropatterned multiwell plates reveals ccxvi influence of cell geometry and cytoskeleton on chromatin dynamics. Biotechnol J 10, 1555–1567. Hasselberg, NE, Haland, TF, Saberniak, J, Brekke, PH, Berge, KE, Leren, TP, Edvardsen, T, and Haugaa, KH (2018). Lamin A/C cardiomyopathy: Young onset, high penetrance, and frequent need for heart transplantation. Eur Heart J 39, 853–860. Hatch, EM, Fischer, AH, Deerinck, TJ, and Hetzer, MW (2013). Catastrophic Nuclear Envelope Collapse in Cancer Cell Micronuclei. Cell 154, 47–60. Hatch, EM, and Hetzer, MW (2016). Nuclear envelope rupture is induced by actin-based nucleus confinement. J Cell Biol 215, 27–36. Hayashi-Takanaka, Y et al. (2011). Tracking epigenetic histone modifications in single cells using Fab-based live endogenous modification labeling. Nucleic Acids Res 39, 6475– 6488. Heffler, J, Shah, PP, Robison, P, Phyo, S, Veliz, K, Uchida, K, Bogush, A, Rhoades, J, Jain, R, and Prosser, BL (2019). A Balance Between Intermediate Filaments and Microtubules Maintains Nuclear Architecture in the Cardiomyocyte. Circ Res. Hegele, RA, Cao, H, Liu, DM, Costain, GA, Charlton-Menys, V, Rodger, NW, and Durrington, PN (2006). Sequencing of the Reannotated LMNB2 Gene Reveals Novel Mutations in Patients with Acquired Partial Lipodystrophy. Am J Hum Genet 79, 383–389. Heo, S-J, Thorpe, SD, Driscoll, TP, Duncan, RL, Lee, DA, and Mauck, RL (2015). Biophysical Regulation of Chromatin Architecture Instills a Mechanical Memory in Mesenchymal Stem Cells. Sci Rep 5, 16895. Heo, SJ, Driscoll, TP, Thorpe, SD, Nerurkar, NL, Baker, BM, Yang, MT, Chen, CS, Lee, DA, and Mauck, RL (2016a). Differentiation alters stem cell nuclear architecture, mechanics, ccxvii and mechano-sensitivity. Elife 5, 1–21. Heo, SJ, Han, WM, Szczesny, SE, Cosgrove, BD, Elliott, DM, Lee, DA, Duncan, RL, and Mauck, RL (2016b). Mechanically Induced Chromatin Condensation Requires Cellular Contractility in Mesenchymal Stem Cells. Biophys J 111, 864–874. Hinz, BE, Walker, SG, Xiong, A, Gogal, RA, Schnieders, MJ, and Wallrath, LL (2021). In Silico and In Vivo Analysis of Amino Acid Substitutions That Cause Laminopathies. Int J Mol Sci 2021, Vol 22, Page 11226 22, 11226. Ho, CY, Jaalouk, DE, Vartiainen, MK, and Lammerding, J (2013). Lamin A/C and emerin regulate MKL1-SRF activity by modulating actin dynamics. Nature 497, 507–513. Holaska, JM, Lee, KK, Kowalski, AK, and Wilson, KL (2003). Transcriptional repressor germ cell-less (GCL) and barrier to autointegration factor (BAF) compete for binding to emerin in vitro. J Biol Chem 278, 6969–6975. Horn, HF (2014). LINC Complex Proteins in Development and Disease. Curr Top Dev Biol 109, 287–321. Houben, F, Willems, CHMP, Declercq, ILJ, Hochstenbach, K, Kamps, MA, Snoeckx, LHEH, Ramaekers, FCS, and Broers, JLV (2009). Disturbed nuclear orientation and cellular migration in A-type lamin deficient cells. Biochim Biophys Acta - Mol Cell Res 1793, 312–324. Huang, D, Sherman, B, and Lempicki, R (2009a). Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res 37, 1–13. Huang, D, Sherman, B, and Lempicki, R (2009b). Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 4, 44–57. Ihalainen, TO, Aires, L, Herzog, FA, Schwartlander, R, Moeller, J, and Vogel, V (2015). ccxviii Differential basal-to-apical accessibility of lamin A/C epitopes in the nuclear lamina regulated by changes in cytoskeletal tension. Nat Mater 14, 1252–1261. Irianto, J, Pfeifer, CR, Ivanovska, IL, Swift, J, and Discher, DE (2016). Nuclear Lamins in Cancer. Cell Mol Bioeng 2016 92 9, 258–267. Isermann, P, and Lammerding, J (2013). Nuclear mechanics and mechanotransduction in health and disease. Curr Biol 23, R1113–R1121. Itano, N, Okamoto, S -i., Zhang, D, Lipton, SA, and Ruoslahti, E (2003). Cell spreading controls endoplasmic and nuclear calcium: A physical gene regulation pathway from the cell surface to the nucleus. Proc Natl Acad Sci 100, 5181–5186. Ito, T et al. (2011). Real-Time Imaging of Histone H4K12–Specific Acetylation Determines the Modes of Action of Histone Deacetylase and Bromodomain Inhibitors. Chem Biol 18, 495–507. Iyer, KV, Pulford, S, Mogilner, A, and Shivashankar, GV (2012). Mechanical Activation of Cells Induces Chromatin Remodeling Preceding MKL Nuclear Transport. Biophys J 103, 1416–1428. Iyer, KV, Taubenberger, A, Zeidan, SA, Dye, NA, Eaton, S, and Jülicher, F (2021). Apico-basal cell compression regulates Lamin A/C levels in epithelial tissues. Nat Commun 2021 121 12, 1–15. Jaalouk, DE, and Lammerding, J (2009). Mechanotransduction gone awry. Nat Rev Mol Cell Biol 10, 63–73. Jain, N, Iyer, KV, Kumar, A, and Shivashankar, G V (2013). Cell geometric constraints induce modular gene-expression patterns via redistribution of HDAC3 regulated by actomyosin contractility. Proc Natl Acad Sci U S A 110, 11349–11354. ccxix Jung, H-J, Tatar, A, Tu, Y, Nobumori, C, Yang, SH, Goulbourne, CN, Herrmann, H, Fong, LG, and Young, SG (2014). An absence of nuclear lamins in keratinocytes leads to ichthyosis, defective epidermal barrier function, and intrusion of nuclear membranes and endoplasmic reticulum into the nuclear chromatin. Mol Cell Biol 34, 4534–4544. Kalverda, B, Pickersgill, H, Shloma, V V., and Fornerod, M (2010). Nucleoporins Directly Stimulate Expression of Developmental and Cell-Cycle Genes Inside the Nucleoplasm. Cell 140, 360–371. Kassianidou, E, Kalita, J, and Lim, RYH (2019). The role of nucleocytoplasmic transport in mechanotransduction. Exp Cell Res 377, 86–93. Khatau, SB, Hale, CM, Stewart-Hutchinson, PJ, Patel, MS, Stewart, CL, Searson, PC, Hodzic, D, and Wirtz, D (2009). A perinuclear actin cap regulates nuclear shape. Proc Natl Acad Sci U S A 106, 19017–19022. Kim, D-H, Lipke, EA, Kim, P, Cheong, R, Thompson, S, Delannoy, M, Suh, K-Y, Tung, L, and Levchenko, A (2010). Nanoscale cues regulate the structure and function of macroscopic cardiac tissue constructs. Proc Natl Acad Sci 107, 565–570. Kim, DH, and Wirtz, D (2015). Cytoskeletal tension induces the polarized architecture of the nucleus. Biomaterials 48, 161–172. Kim, DI, Birendra, KC, and Roux, KJ (2015). Making the LINC: SUN and KASH protein interactions. Biol Chem 396, 295–310. Kim, JK, Louhghalam, A, Lee, G, Schafer, BW, Wirtz, D, and Kim, DH (2017). Nuclear lamin A/C harnesses the perinuclear apical actin cables to protect nuclear morphology. Nat Commun 8, 1–13. Kim, PH et al. (2018). Disrupting the LINC complex in smooth muscle cells reduces aortic ccxx disease in a mouse model of Hutchinson-Gilford progeria syndrome. Sci Transl Med. Kim, Y, Zheng, X, and Zheng, Y (2013). Proliferation and differentiation of mouse embryonic stem cells lacking all lamins. Cell Res 23, 1420–1423. Kim, Y, Zheng, X, and Zheng, Y (2019). Role of lamins in 3D genome organization and global gene expression. Https://DoiOrg/101080/1949103420191578601 10, 13–21. Kirby, TJ, and Lammerding, J (2016). Cell mechanotransduction: Stretch to express. Nat Mater 15, 1227–1229. Kirby, TJ, and Lammerding, J (2018). Emerging views of the nucleus as a cellular mechanosensor. Nat Cell Biol 20, 373–381. Kochin, V, Shimi, T, Torvaldson, E, Adam, SA, Goldman, A, Pack, C-G, Melo-Cardenas, J, Imanishi, SY, Goldman, RD, and Eriksson, JE (2014). Interphase phosphorylation of lamin A. J Cell Sci 127, 2683–2696. Kourmouli, N, Dialynas, G, Petraki, C, Pyrpasopoulou, A, Singh, PB, Georgatos, SD, and Theodoropoulos, PA (2001). Binding of heterochromatin protein 1 to the nuclear envelope is regulated by a soluble form of tubulin. J Biol Chem 276, 13007–13014. Krimm, I, Östlund, C, Gilquin, B, Couprie, J, Hossenlopp, P, Mornon, JP, Bonne, G, Courvalin, JC, Worman, HJ, and Zinn-Justin, S (2002). The Ig-like Structure of the C-Terminal Domain of Lamin A/C, Mutated in Muscular Dystrophies, Cardiomyopathy, and Partial Lipodystrophy. Structure 10, 811–823. Kudlow, BA, Stanfel, MN, Burtner, CR, Johnston, ED, and Kennedy, BK (2008). Suppression of Proliferative Defects Associated with Processing-defective Lamin A Mutants by hTERT or Inactivation of p53. Mol Biol Cell 19, 5238–5248. Labrador, M, and Corces, VG (2002). Setting the Boundaries of Chromatin Domains and ccxxi Nuclear Organization. Cell 111, 151–154. Lammerding, J, Fong, LG, Ji, JY, Reue, K, Stewart, CL, Young, SG, and Lee, RT (2006). Lamins A and C but not lamin B1 regulate nuclear mechanics. J Biol Chem 281, 25768– 25780. Lammerding, J, Hsiao, J, Schulze, PC, Kozlov, S, Stewart, CL, and Lee, RT (2005). Abnormal nuclear shape and impaired mechanotransduction in emerin-deficient cells. J Cell Biol 170, 781–791. Lammerding, J, Schulze, PC, Takahashi, T, Kozlov, S, Sullivan, T, Kamm, RD, Stewart, CL, and Lee, RT (2004a). Lamin A/C deficiency causes defective nuclear mechanics and mechanotransduction. J Clin Invest 113, 370–378. Lammerding, J, Schulze, PC, Takahashi, T, Kozlov, S, Sullivan, T, Kamm, RD, Stewart, CL, and Lee, RT (2004b). Lamin A / C deficiency causes defective nuclear mechanics and mechanotransduction. J Clin Invest 113, 370–378. de Lanerolle, P, and Serebryannyy, L (2011). Nuclear actin and myosins: Life without filaments. Nat Cell Biol 13, 1282–1288. Lattanzi, G, Cenni, V, Marmiroli, S, Capanni, C, Mattioli, E, Merlini, L, Squarzoni, S, and Mario Maraldi, N (2003). Association of emerin with nuclear and cytoplasmic actin is regulated in differentiating myoblasts. Biochem Biophys Res Commun 303, 764–770. Laurini, E et al. (2018). Biomechanical defects and rescue of cardiomyocytes expressing pathologic nuclear lamins. Cardiovasc Res 114. Le, HQ et al. (2016). Mechanical regulation of transcription controls Polycomb-mediated gene silencing during lineage commitment. Nat Cell Biol 18, 864–875. Lee, J, Abdeen, AA, Zhang, D, and Kilian, KA (2013). Directing stem cell fate on hydrogel ccxxii substrates by controlling cell geometry, matrix mechanics and adhesion ligand composition. Biomaterials 34, 8140–8148. de Leeuw, R, Gruenbaum, Y, and Medalia, O (2018). Nuclear Lamins: Thin Filaments with Major Functions. Trends Cell Biol 28, 34–45. Legant, WR, Pathak, A, Yang, MT, Deshpande, VS, McMeeking, RM, and Chen, CS (2009). Microfabricated tissue gauges to measure and manipulate forces from 3D microtissues. Proc Natl Acad Sci 106, 10097–10102. Lele, TP, Dickinson, RB, and Gundersen, GG (2018). Mechanical principles of nuclear shaping and positioning. J Cell Biol, jcb.201804052. Li, E (2002). Chromatin modification and epigenetic reprogramming in mammalian development. Nat Rev Genet 3, 662–673. Li, Y, Chu, JS, Kurpinski, K, Li, X, Bautista, DM, Yang, L, Paul Sung, K-L, and Li, S (2011). Biophysical Regulation of Histone Acetylation in Mesenchymal Stem Cells. Biophys J 100, 1902–1909. Liao, Y, Smyth, GK, and Shi, W (2014). FeatureCounts: An efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30, 923–930. Liu, B et al. (2005). Genomic instability in laminopathy-based premature aging. Nat Med 11, 780–785. Liu, L, Luo, Q, Sun, J, and Song, G (2019). Cytoskeletal control of nuclear morphology and stiffness are required for OPN-induced bone-marrow-derived mesenchymal stem cell migration. Https://DoiOrg/101139/Bcb-2018-0263 97, 463–470. Llères, D, James, J, Swift, S, Norman, DG, and Lamond, AI (2009). Quantitative analysis of chromatin compaction in living cells using FLIM-FRET. J Cell Biol 187, 481–496. ccxxiii Lombardi, ML, Jaalouk, DE, Shanahan, CM, Burke, B, Roux, KJ, and Lammerding, J (2011). The interaction between nesprins and sun proteins at the nuclear envelope is critical for force transmission between the nucleus and cytoskeleton. J Biol Chem 286, 26743– 26753. Long, C et al. (2018). Correction of diverse muscular dystrophy mutations in human engineered heart muscle by single-site genome editing. Sci Adv 4, 1–12. Love, MI, Huber, W, and Anders, S (2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15, 550. M. Davidson, P, R. Fedorchak, G, Solenne Mondésert-Deveraux, S. Bell, E, Philipp Isermann, Denis Aubry, Rachele Allena, and Jan Lammerding (2019). High-throughput microfluidic micropipette aspiration device to probe time-scale dependent nuclear mechanics in intact cells. Lab Chip 19, 3652–3663. Ma, H, Naseri, A, Reyes-Gutierrez, P, Wolfe, SA, Zhang, S, and Pederson, T (2015). Multicolor CRISPR labeling of chromosomal loci in human cells. Proc Natl Acad Sci 112, 3002– 3007. Maffioletti, SM et al. (2018). Three-Dimensional Human iPSC-Derived Artificial Skeletal Muscles Model Muscular Dystrophies and Enable Multilineage Tissue Engineering. Cell Rep 23, 899–908. Maharana, S, Iyer, KV, Jain, N, Nagarajan, M, Wang, Y, and Shivashankar, G V. (2016). Chromosome intermingling—the physical basis of chromosome organization in differentiated cells. Nucleic Acids Res 44, 5148–5160. Malviya, AN, and Rogue, PJ (1998). “Tell Me Where Is Calcium Bred”: Clarifying the Roles of Nuclear Calcium. Cell 92, 17–23. ccxxiv Manilal, S, Randles, K, Aunac, C, Nguyen, M, and Morris, G (2004). A lamin A/C beta-strand containing the site of lipodystrophy mutations is a major surface epitope for a new panel of monoclonal antibodies. Biochim Biophys Acta 1671, 87–92. Manju, K, Muralikrishna, B, Parnaik, VK, Pouchelet, M, Capeau, J, Courvalin, JC, and Buendia, B (2006). Expression of disease-causing lamin A mutants impairs the formation of DNA repair foci. J Cell Sci 119, 2704–2714. Marcotti, S, Freitas, DB de, Troughton, LD, Kenny, FN, Shaw, T, Stramer, BM, and Oakes, PW (2021). A workflow for rapid unbiased quantification of fibrillar feature alignment in biological images. BioRxiv, 2021.07.22.453401. Maréchal, A, and Zou, L (2013). DNA Damage Sensing by the ATM and ATR Kinases. Cold Spring Harb Perspect Biol 5. Martinac, B (2004). Mechanosensitive ion channels: molecules of mechanotransduction. J Cell Sci 117, 2449–2460. Mattioli, E, Columbaro, M, Capanni, C, Maraldi, NM, Cenni, V, Scotlandi, K, Marino, MT, Merlini, L, Squarzoni, S, and Lattanzi, G (2011). Prelamin A-mediated recruitment of SUN1 to the nuclear envelope directs nuclear positioning in human muscle. Cell Death Differ 18, 1305–1315. Mattout, A, Pike, BL, Towbin, BD, Bank, EM, Gonzalez-Sandoval, A, Stadler, MB, Meister, P, Gruenbaum, Y, and Gasser, SM (2011). An EDMD Mutation in C. elegans Lamin Blocks Muscle-Specific Gene Relocation and Compromises Muscle Integrity. Curr Biol 21, 1603–1614. Maurer, M, and Lammerding, J (2019). The Driving Force: Nuclear Mechanotransduction in Cellular Function, Fate, and Disease. Annu Rev Biomed Eng 21, 443–468. ccxxv Mazzanti, M, Bustamante, JO, and Oberleithner, H (2001). Electrical Dimension of the Nuclear Envelope. Physiol Rev 81, 1–19. McCord, RP, Nazario-Toole, A, Zhang, H, Chines, PS, Zhan, Y, Erdos, MR, Collins, FS, Dekker, J, and Cao, K (2013). Correlated alterations in genome organization, histone methylation, and DNA-lamin A/C interactions in Hutchinson-Gilford progeria syndrome. Genome Res 23, 260–269. McGregor, AL, Hsia, CR, and Lammerding, J (2016). Squish and squeeze - the nucleus as a physical barrier during migration in confined environments. Curr Opin Cell Biol 40, 32– 40. McNally, EM, and Mestroni, L (2017). Dilated cardiomyopathy: Genetic determinants and mechanisms. Circ Res 121, 731–748. Mehrabi, M et al. (2021). A Study of Gene Expression, Structure, and Contractility of iPSC- Derived Cardiac Myocytes from a Family with Heart Disease due to LMNA Mutation. Ann Biomed Eng 2021, 1–16. Melcer, S, Hezroni, H, Rand, E, Nissim-Rafinia, M, Skoultchi, A, Stewart, CL, Bustin, M, and Meshorer, E (2012). Histone modifications and lamin A regulate chromatin protein dynamics in early embryonic stem cell differentiation. Nat Commun 3, 910. Melcon, G et al. (2006). Loss of emerin at the nuclear envelope disrupts the Rb1/E2F and MyoD pathways during muscle regeneration. Hum Mol Genet 15, 637–651. Mewborn, SK et al. (2010). Altered Chromosomal Positioning, Compaction, and Gene Expression with a Lamin A/C Gene Mutation. PLoS One 5, e14342. Moiseeva, O, Bourdeau, V, Vernier, M, Dabauvalle, M-C, and Ferbeyre, G (2011). Retinoblastoma-independent regulation of cell proliferation and senescence by the p53- ccxxvi p21 axis in lamin A /C-depleted cells. Aging Cell 10, 789–797. Morales Rodriguez, B, Khouzami, L, Decostre, V, Varnous, S, Pekovic-Vaughan, V, Hutchison, CJ, Pecker, F, Bonne, G, and Muchir, A (2018). N-acetyl cysteine alleviates oxidative stress and protects mice from dilated cardiomyopathy caused by mutations in nuclear A- type lamins gene. Hum Mol Genet 00, 1–8. Mounkes, LC, Kozlov, S V., Rottman, JN, and Stewart, CL (2005). Expression of an LMNA- N195K variant of A-type lamins results in cardiac conduction defects and death in mice. Hum Mol Genet 14, 2167–2180. Muchir, A et al. (2004). Nuclear envelope alterations in fibroblasts from patients with muscular dystrophy, cardiomyopathy, and partial lipodystrophy carrying lamin A/C gene mutations. Muscle Nerve 30, 444–450. Muchir, A, Pavlidis, P, Bonne, G, Hayashi, YK, and Worman, HJ (2007). Activation of MAPK in hearts of EMD null mice: similarities between mouse models of X-linked and autosomal dominant Emery–Dreifuss muscular dystrophy. Hum Mol Genet 16, 1884– 1895. Muchir, A, Wu, W, Choi, JC, Iwata, S, Morrow, J, Homma, S, and Worman, HJ (2012). Abnormal p38 mitogen-activated protein kinase signaling in dilated cardiomyopathy caused by lamin A/C gene mutation. Hum Mol Genet 21, 4325–4333. Murphy, WL, McDevitt, TC, and Engler, AJ (2014). Materials as stem cell regulators. Nat Mater 13, 547–557. National Institute of Health, USNL of M (2017). A Study of ARRY-371797 in Patients With LMNA-Related Dilated Cardiomyopathy. Available at: https://clinicaltrials.gov/ct2/show/NCT02057341. ccxxvii Nava, MM et al. (2020). Heterochromatin-Driven Nuclear Softening Protects the Genome against Mechanical Stress-Induced Damage. Cell 181, 800-817.e22. Nicolas, HA, Akimenko, M-A, and Tesson, F (2019). Cellular and Animal Models of Striated Muscle Laminopathies. Cells 8, 291. Nikolova, V et al. (2004). Defects in nuclear structure and function promote dilated cardiomyopathy in lamin A / C – deficient mice. J Clin Invest 113, 357–369. Novelli, G et al. (2002). Mandibuloacral Dysplasia Is Caused by a Mutation in LMNA-Encoding Lamin A/C. Am J Hum Genet 71, 426–431. Ollion, J, Cochennec, J, Loll, F, Escudé, C, and Boudier, T (2013). TANGO: a generic tool for high-throughput 3D image analysis for studying nuclear organization. Bioinformatics 29, 1840–1841. Olson, EN, and Nordheim, A (2010). Linking actin dynamics and gene transcription to drive cellular motile functions. Nat Rev Mol Cell Biol 11, 353–365. Owens, DJ, Fischer, M, Jabre, S, Moog, S, Mamchaoui, K, Butler-browne, G, and Coirault, C (2020). Lamin Mutations Cause Increased YAP Nuclear Entry in Muscle Stem Cells. 1– 14. Padiath, QS, Saigoh, K, Schiffmann, R, Asahara, H, Yamada, T, Koeppen, A, Hogan, K, Ptáček, LJ, and Fu, YH (2006). Lamin B1 duplications cause autosomal dominant leukodystrophy. Nat Genet 38, 1114–1123. Pagés, H, Carlson, M, and Falcon, S (2021). AnnotationDbi: Manipulation of SQLite-based annotations in Bioconductor. R Packag Version 1541. Pajerowski, JD, Dahl, KN, Zhong, FL, Sammak, PJ, and Discher, DE (2007). Physical plasticity of the nucleus in stem cell differentiation. Proc Natl Acad Sci U S A 104, 15619–15624. ccxxviii Park, S-J et al. (2017). DNA-PK Promotes the Mitochondrial, Metabolic, and Physical Decline that Occurs During Aging. Cell Metab 25, 1135-1146.e7. Pasapera, AM, Schneider, IC, Rericha, E, Schlaepfer, DD, and Waterman, CM (2010). Myosin II activity regulates vinculin recruitment to focal adhesions through FAK-mediated paxillin phosphorylation. J Cell Biol 188, 877–890. Pauschinger, M, Knopf, D, Petschauer, S, Doerner, A, Poller, W, Schwimmbeck, PL, Kühl, U, and Schultheiss, HP (1999). Dilated cardiomyopathy is associated with significant changes in collagen type I/III ratio. Circulation 99, 2750–2756. Pellegrini, C et al. (2019). Altered adipocyte differentiation and unbalanced autophagy in type 2 Familial Partial Lipodystrophy: an in vitro and in vivo study of adipose tissue browning. Exp Mol Med 2019 518 51, 1–17. Peric-Hupkes, D et al. (2010a). Molecular maps of the reorganization of genome-nuclear lamina interactions during differentiation. Mol Cell 38, 603–613. Peric-Hupkes, D et al. (2010b). Molecular Maps of the Reorganization of Genome-Nuclear Lamina Interactions during Differentiation. Mol Cell 38, 603–613. Peshavaria, M, and Day, IN (1991). Molecular structure of the human muscle-specific enolase gene (ENO3). Biochem J 275, 427–433. Piccus, R, and Brayson, D (2020). The nuclear envelope: LINCing tissue mechanics to genome regulation in cardiac and skeletal muscle. Biol Lett 16, 20200302. Piercy, RJ, Zhou, H, Feng, L, Pombo, A, Muntoni, F, and Brown, SC (2007). Desmin immunolocalisation in autosomal dominant Emery-Dreifuss muscular dystrophy. Neuromuscul Disord 17, 297–305. Pimentel, MR, Falcone, S, Cadot, B, and Gomes, ER (2017). In Vitro Differentiation of Mature ccxxix Myofibers for Live Imaging. J Vis Exp 2017, 55141. Plessner, M, Melak, M, Chinchilla, P, Baarlink, C, and Grosse, R (2015). Nuclear F-actin formation and reorganization upon cell spreading. J Biol Chem 290, 11209–11216. Pradas, JC et al. (2020). Identification of Genes and Pathways Regulated by Lamin A in Heart. J Am Heart Assoc 9. Rajala, K, Pekkanen-Mattila, M, and Aalto-Setälä, K (2011). Cardiac differentiation of pluripotent stem cells. Stem Cells Int. Ramdas, NM, and Shivashankar, GV (2015). Cytoskeletal Control of Nuclear Morphology and Chromatin Organization. J Mol Biol 427, 695–706. Rappold, I, Iwabuchi, K, Date, T, and Chen, J (2001). Tumor suppressor p53 binding protein 1 (53BP1) is involved in DNA damage-signaling pathways. J Cell Biol 153, 613–620. Redwood, AB et al. (2011). A dual role for A-type lamins in DNA double-strand break repair. Cell Cycle 10, 2549–2560. Rober, RA, Weber, K, and Osborn, M (1989). Differential timing of nuclear lamin A/C expression in the various organs of the mouse embryo and the young animal: a developmental study. Development 105, 365–378. Ronaldson-Bouchard, K et al. (2019). Engineering of human cardiac muscle electromechanically matured to an adult-like phenotype. Nat Protoc. Ronaldson-Bouchard, K, Ma, SP, Yeager, K, Chen, T, Song, L, Sirabella, D, Morikawa, K, Teles, D, Yazawa, M, and Vunjak-Novakovic, G (2018). Advanced maturation of human cardiac tissue grown from pluripotent stem cells. Nature 556, 239–243. Rothballer, A, and Kutay, U (2013). The diverse functional LINCs of the nuclear envelope to the cytoskeleton and chromatin. Chromosoma 122, 415–429. ccxxx Roux, KJ, Kim, DI, Raida, M, and Burke, B (2012). A promiscuous biotin ligase fusion protein identifies proximal and interacting proteins in mammalian cells. J Cell Biol 196, 801– 810. Rowat, AC, Lammerding, J, and Ipsen, JH (2006). Mechanical Properties of the Cell Nucleus and the Effect of Emerin Deficiency. Biophys J 91, 4649–4664. Sakar, MS, Neal, D, Boudou, T, Borochin, MA, Li, Y, Weiss, R, Kamm, RD, Chen, CS, and Asada, HH (2012). Formation and optogenetic control of engineered 3D skeletal muscle bioactuators. Lab Chip 12, 4976. Santella, L, and Carafoli, E (1997). Calcium signaling in the cell nucleus. FASEB J 11, 1091– 1109. Sasaki, K, Ito, T, Nishino, N, Khochbin, S, and Yoshida, M (2009). Real-time imaging of histone H4 hyperacetylation in living cells. Proc Natl Acad Sci 106, 16257–16262. Sawada, Y, Tamada, M, Dubin-Thaler, BJ, Cherniavskaya, O, Sakai, R, Tanaka, S, and Sheetz, MP (2006). Force Sensing by Mechanical Extension of the Src Family Kinase Substrate p130Cas. Cell 127, 1015–1026. Scharner, J, Lu, H-C, Fraternali, F, Ellis, JA, and Zammit, PS (2014). Mapping disease-related missense mutations in the immunoglobulin-like fold domain of lamin A/C reveals novel genotype–phenotype associations for laminopathies. Proteins Struct Funct Bioinforma 82, 904–915. Schiff, M, Froissart, R, Olsen, RKJ, Acquaviva, C, and Vianey-Saban, C (2006). Electron transfer flavoprotein deficiency: Functional and molecular aspects. Mol Genet Metab 88, 153–158. Schopp, IM, Amaya Ramirez, CC, Debeljak, J, Kreibich, E, Skribbe, M, Wild, K, and Béthune, J ccxxxi (2017). Split-BioID a conditional proteomics approach to monitor the composition of spatiotemporally defined protein complexes. Nat Commun 8, 15690. Schreiner, SM, Koo, PK, Zhao, Y, Mochrie, SGJ, and King, MC (2015). The tethering of chromatin to the nuclear envelope supports nuclear mechanics. Nat Commun 6, 1–13. Scuderi, GJ, and Butcher, J (2017). Naturally Engineered Maturation of Cardiomyocytes. Front Cell Dev Biol 5, 1–28. Shah, D et al. (2019). Modeling of LMNA-Related Dilated Cardiomyopathy Using Human Induced Pluripotent Stem Cells. Cells 8, 594. Shah, P, Hobson, CM, Cheng, S, Colville, MJ, Paszek, MJ, Superfine, R, and Lammerding, J (2021). Nuclear Deformation Causes DNA Damage by Increasing Replication Stress. Curr Biol 31, 753-765.e6. Shao, S, Zhang, W, Hu, H, Xue, B, Qin, J, Sun, C, Sun, Y, Wei, W, and Sun, Y (2016). Long- term dual-color tracking of genomic loci by modified sgRNAs of the CRISPR/Cas9 system. Nucleic Acids Res 44, e86–e86. Shao, X, Li, Q, Mogilner, A, Bershadsky, AD, and Shivashankar, G V. (2015). Mechanical stimulation induces formin-dependent assembly of a perinuclear actin rim. Proc Natl Acad Sci 112, E2595–E2601. Shao, Z, Koh, W, Ni, Y, Li, W, Agatisa-Boyle, B, Merkurjev, D, and Tang, WHW (2020). RNA Sequence Analyses throughout the Course of Mouse Cardiac Laminopathy Identify Differentially Expressed Genes for Cell Cycle Control and Mitochondrial Function. Sci Rep 10, 6632. Sharma, A, Li, G, Rajarajan, K, Hamaguchi, R, Burridge, PW, and Wu, SM (2015). Derivation of Highly Purified Cardiomyocytes from Human Induced Pluripotent Stem Cells Using ccxxxii Small Molecule-modulated Differentiation and Subsequent Glucose Starvation. J Vis Exp, 3–10. Shimi, T et al. (2008). The A- and B-type nuclear lamin networks: microdomains involved in chromatin organization and transcription. Genes Dev 22, 3409–3421. Shimi, T, Kittisopikul, M, Tran, J, Goldman, AE, Adam, SA, Zheng, Y, Jaqaman, K, and Goldman, RD (2015). Structural organization of nuclear lamins A, C, B1, and B2 revealed by superresolution microscopy. Mol Biol Cell 26, 4075–4086. Shiu, J-Y, Aires, L, Lin, Z, and Vogel, V (2018). Nanopillar force measurements reveal actin- cap-mediated YAP mechanotransduction. Nat Cell Biol 20, 262–271. Sieprath, T, Darwiche, R, and De Vos, WH (2012). Lamins as mediators of oxidative stress. Biochem Biophys Res Commun 421, 635–639. Simon, DN, and Wilson, KL (2013). Partners and post-translational modifications of nuclear lamins. Chromosoma 122, 13–31. Siu, CW et al. (2012). Modeling of lamin A/C mutation premature cardiac aging using patient- specific induced pluripotent stem cells. Aging (Albany NY) 4, 803–822. Soheilypour, M, Peyro, M, Jahed, Z, and Mofrad, MRK (2016). On the Nuclear Pore Complex and Its Roles in Nucleo-Cytoskeletal Coupling and Mechanobiology. Cell Mol Bioeng 9, 217–226. Spagnol, ST, and Dahl, KN (2016). Spatially resolved quantification of chromatin condensation through differential local rheology in cell nuclei fluorescence lifetime imaging. PLoS One 11, 1–19. van Spreeuwel, ACC, Bax, NAM, Bastiaens, AJ, Foolen, J, Loerakker, S, Borochin, M, van der Schaft, DWJ, Chen, CS, Baaijens, FPT, and Bouten, CVC (2014). The influence of ccxxxiii matrix (an)isotropy on cardiomyocyte contraction in engineered cardiac microtissues. Integr Biol 6, 422–429. Steele-Stallard, HB, Pinton, L, Sarcar, S, Ozdemir, T, Maffioletti, SM, Zammit, PS, and Tedesco, FS (2018). Modeling Skeletal Muscle Laminopathies Using Human Induced Pluripotent Stem Cells Carrying Pathogenic LMNA Mutations. Front Physiol 9, 1332. Stephens, AD (2020). Chromatin rigidity provides mechanical and genome protection. Mutat Res - Fundam Mol Mech Mutagen 821. Stephens, AD, Banigan, EJ, Adam, SA, Goldman, RD, and Marko, JF (2017). Chromatin and lamin A determine two different mechanical response regimes of the cell nucleus. Mol Biol Cell 28, 1984–1996. Stephens, AD, Banigan, EJ, and Marko, JF (2018a). Separate roles for chromatin and lamins in nuclear mechanics. Nucleus 9, 119–124. Stephens, AD, Liu, PZ, Banigan, EJ, Almassalha, LM, Backman, V, Adam, SA, Goldman, RD, and Marko, JF (2018b). Chromatin histone modifications and rigidity affect nuclear morphology independent of lamins. Mol Biol Cell 29, 220–233. Stewart, CL, Kozlov, S, Fong, LG, and Young, SG (2007). Mouse models of the laminopathies. Exp Cell Res 313, 2144–2156. Sullivan, T, Escalante-Alcalde, D, Bhatt, H, Anver, M, Bhat, N, Nagashima, K, Stewart, CL, and Burke, B (1999). Loss of A-type lamin expression compromises nuclear envelope integrity leading to muscular dystrophy. J Cell Biol 147, 913–920. Sweet, ME et al. (2018). Transcriptome analysis of human heart failure reveals dysregulated cell adhesion in dilated cardiomyopathy and activated immune pathways in ischemic heart failure. BMC Genomics 2018 191 19, 1–14. ccxxxiv Swift, J et al. (2013). Nuclear Lamin-A Scales with Tissue Stiffness and Enhances Matrix- Directed Differentiation. Science (80- ) 341. Szczesny, SE, and Mauck, RL (2017). The Nuclear Option: Evidence Implicating the Cell Nucleus in Mechanotransduction. J Biomech Eng 139, 021006. Szklarczyk, D et al. (2019). STRING v11: protein–protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res 47, D607–D613. Tajik, A, Zhang, Y, Wei, F, Sun, J, Jia, Q, Zhou, W, Singh, R, Khanna, N, Belmont, AS, and Wang, N (2016). Transcription upregulation via force-induced direct stretching of chromatin. 15, 1287–1296. Tamada, M, Sheetz, MP, and Sawada, Y (2004). Activation of a Signaling Cascade by Cytoskeleton Stretch. Dev Cell 7, 709–718. Taneja, N, Neininger, AC, and Burnette, DT (2020). Coupling to substrate adhesions drives the maturation of muscle stress fibers into myofibrils within cardiomyocytes. Https://DoiOrg/101091/MbcE19-11-0652 31, 1273–1288. Tapley, EC, and Starr, DA (2013). Connecting the nucleus to the cytoskeleton by SUN–KASH bridges across the nuclear envelope. Curr Opin Cell Biol 25, 57–62. Tariq, Z et al. (2017). Lamin A and microtubules collaborate to maintain nuclear morphology. Nucleus 8, 433–446. Taylor, MRG et al. (2003). Natural history of dilated cardiomyopathy due to lamin A/C gene mutations. J Am Coll Cardiol 41, 771–780. Tekin, D, Dursun, AD, and Xi, L (2010). Hypoxia inducible factor 1 (HIF-1) and cardioprotection. Acta Pharmacol Sin 2010 319 31, 1085–1094. ccxxxv Tiburcy, M et al. (2017). Defined Engineered Human Myocardium with Advanced Maturation for Applications in Heart Failure Modelling and Repair. Circulation 135. Turenne, GA, Paul, P, Laflair, L, and Price, BD (2001). Activation of p53 transcriptional activity requires ATM’s kinase domain and multiple N-terminal serine residues of p53. Oncogene 2001 2037 20, 5100–5110. Turgay, Y, Eibauer, M, Goldman, AE, Shimi, T, Khayat, M, Ben-Harush, K, Dubrovsky-Gaupp, A, Sapra, KT, Goldman, RD, and Medalia, O (2017). The molecular architecture of lamins in somatic cells. Nature 543, 261–264. Varela, I et al. (2005). Accelerated ageing in mice deficient in Zmpste24 protease is linked to p53 signalling activation. Nature 437, 564–568. Vaughan, A, Alvarez-Reyes, M, Bridger, JM, Broers, JL, Ramaekers, FC, Wehnert, M, Morris, GE, Whitfield WGF, and Hutchison, CJ (2001). Both emerin and lamin C depend on lamin A for localization at the nuclear envelope. J Cell Sci 114, 2577–2590. Verga, L, Concardi, M, Pilotto, A, Bellini, O, Pasotti, M, Repetto, A, Tavazzi, L, and Arbustini, E (2003). Loss of lamin A/C expression revealed by immuno-electron microscopy in dilated cardiomyopathy with atrioventricular block caused by LMNA gene defects. Virchows Arch 443, 664–671. Verstraeten, VLRM et al. (2009). The R439C mutation in LMNA causes lamin oligomerization and susceptibility to oxidative stress. J Cell Mol Med 13, 959–971. Vigouroux, C, Auclair, M, Dubosclard, E, Pouchelet, M, Capeau, J, Courvalin, JC, and Buendia, B (2001). Nuclear envelope disorganization in fibroblasts from lipodystrophic patients with heterozygous R482Q/W mutations in the lamin A/C gene. J Cell Sci 114, 4459– 4468. ccxxxvi Vishwasrao, HD, Trifilieff, P, and Kandel, ER (2012). In Vivo Imaging of the Actin Polymerization State with Two-Photon Fluorescence Anisotropy. Biophys J 102, 1204– 1214. De Vos, WH et al. (2011). Repetitive disruptions of the nuclear envelope invoke temporary loss of cellular compartmentalization in laminopathies. Hum Mol Genet 20, 4175–4186. Wang, N, Butler, JP, and Ingber, DE (1993). Mechanotransduction across the cell surface and through the cytoskleton. Science (80- ) 260, 1124–1127. Wang, N, Tytell, JD, and Ingber, DE (2009). Mechanotransduction at a distance: mechanically coupling the extracellular matrix with the nucleus. Nat Rev Mol Cell Biol 10, 75–82. Wang, Y, Nagarajan, M, Uhler, C, and Shivashankar, G V. (2017). Orientation and repositioning of chromosomes correlate with cell geometry–dependent gene expression. Mol Biol Cell 28, 1997–2009. Ward, IM, Minn, K, Deursen, J van, and Chen, J (2003). p53 Binding Protein 53BP1 Is Required for DNA Damage Responses and Tumor Suppression in Mice. Mol Cell Biol 23, 2556. West, AG, and Fraser, P (2005). Remote control of gene transcription. Hum Mol Genet 14, R101–R111. Wiesel, N, Mattout, A, Melcer, S, Melamed-Book, N, Herrmann, H, Medalia, O, Aebi, U, and Gruenbaum, Y (2008). Laminopathic mutations interfere with the assembly, localization, and dynamics of nuclear lamins. Proc Natl Acad Sci U S A 105, 180–185. Willis, ND et al. (2008). Lamin A/C Is a Risk Biomarker in Colorectal Cancer. PLoS One 3, e2988. Wolf, CM et al. (2008). Lamin A/C haploinsufficiency causes dilated cardiomyopathy and apoptosis-triggered cardiac conduction system disease. J Mol Cell Cardiol 44, 293–303. ccxxxvii Wu, SK, Gomez, GA, Michael, M, Verma, S, Cox, HL, Lefevre, JG, Parton, RG, Hamilton, NA, Neufeld, Z, and Yap, AS (2014). Cortical F-actin stabilization generates apical–lateral patterns of junctional contractility that integrate cells into epithelia. Nat Cell Biol 2014 162 16, 167–178. Xia, Y et al. (2018). Nuclear rupture at sites of high curvature compromises retention of DNA repair factors. J Cell Biol 217, 3796–3808. Xie, W, Chojnowski, A, Boudier, T, Lim, JSY, Ahmed, S, Ser, Z, Stewart, C, and Burke, B (2016a). A-type Lamins Form Distinct Filamentous Networks with Differential Nuclear Pore Complex Associations. Curr Biol 26, 2651–2658. Xie, W, Chojnowski, A, Boudier, T, Lim, JSY, Ahmed, S, Ser, Z, Stewart, C, and Burke, B (2016b). A-type Lamins Form Distinct Filamentous Networks with Differential Nuclear Pore Complex Associations. Curr Biol 26, 2651–2658. Yang, C, Tibbitt, MW, Basta, L, and Anseth, KS (2014). Mechanical memory and dosing influence stem cell fate. Nat Mater 13, 645–652. Yang, SH, Chang, SY, Yin, L, Tu, Y, Hu, Y, Yoshinaga, Y, de Jong, PJ, Fong, LG, and Young, SG (2011). An absence of both lamin B1 and lamin B2 in keratinocytes has no effect on cell proliferation or the development of skin and hair. Hum Mol Genet 20, 3537–3544. Yang, Y et al. (2018). Cross-talk between microtubules and the linker of nucleoskeleton complex plays a critical role in the adipogenesis of human adipose-derived stem cells. Stem Cell Res Ther 9, 125. Yap, AS, Duszyc, K, and Viasnoff, V (2018). Mechanosensing and Mechanotransduction at Cell–Cell Junctions. Cold Spring Harb Perspect Biol 10, a028761. Zhao, R, Boudou, T, Wang, WG, Chen, CS, and Reich, DH (2013). Decoupling cell and matrix ccxxxviii mechanics in engineered microtissues using magnetically actuated microcantilevers. Adv Mater 25, 1699–1705. Zhao, S, Guo, Y, and Shyr, Y (2021). KEGGprofile: An annotation and visualization package for multi-types and multi-groups expression data in KEGG pathway. R Packag Version 1340. Zuela, N, Dorfman, J, and Gruenbaum, Y (2017). Global transcriptional changes caused by an EDMD mutation correlate to tissue specific disease phenotypes in C. elegans. Nucleus 8, 60–69. Zuo, B, Yang, J, Wang, F, Wang, L, Yin, Y, Dan, J, Liu, N, and Liu, L (2012). Influences of lamin A levels on induction of pluripotent stem cells. Biol Open 1, 1118–1127. Zwerger, M, Jaalouk, DE, Lombardi, ML, Isermann, P, Mauermann, M, Dialynas, G, Herrmann, H, Wallrath, L, and Lammerding, J (2013). Myopathic lamin mutations impair nuclear stability in cells and tissue and disrupt nucleo-cytoskeletal coupling. Hum Mol Genet 22, 2335–2349. Zwerger, M, Roschitzki-voser, H, Zbinden, R, Denais, C, and Herrmann, H (2015). Altering lamina assembly identifies lamina-dependent and -independent functions for A-type lamins Journal of Cell Science Accepted manuscript. J Cell Sci 41. ccxxxix