University of Southern California Dissertations and Theses

Nanoscale dynamics and nuclear envelope organization of the muscular dystrophy related protein emerin

(USC Thesis Other)

Nanoscale dynamics and nuclear envelope organization of the muscular dystrophy related protein emerin

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content NANOSCALE DYNAMICS AND NUCLEAR ENVELOPE ORGANIZATION OF THE MUSCULAR DYSTROPHY RELATED PROTEIN EMERIN by Anthony Michael Fernandez A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY MOLECULAR BIOLOGY May 2022 ii DEDICATION To my family, especially Lui, Teresa, Ezra, my parents, and everyone who helped me in so many ways throughout the years it took to earn my doctorate. I also want to remember my grandfathers and uncle whom I love and always remember. None of this would have been possible without the consistent and patient support of my family. May the Lord bless all of you and reward you for everything you have done for me. iii ACKNOWLEDGEMENTS This work was supported by the National Science Foundation, Division of Material Research, under Grant No. 1406812, the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health under award number R21AR076514, and the USC Provost Fellowship. Additional assistance was provided by Matthew Michael for use of his confocal microscope, Jongseung Yoon for assistance in PDMS stamp manufacture, Juliet Ellis for the cDNA containing emerin, Howard Worman for providing emerin null and normal human dermal fibroblasts, and the UCLA IMT core/Vector Core for assistance with viral preparations. Chapter Two is a combined version of the following papers: Fernandez, A., Bautista, M., Stanciauskas, R., Chung, T. & Pinaud, F. Cell-Shaping Micropatterns for Quantitative Super-Resolution Microscopy Imaging of Membrane Mechanosensing Proteins. ACS Appl Mater Interfaces 9, 27575-27586, doi:10.1021/acsami.7b09743 (2017). Bautista, M., Fernandez, A. & Pinaud, F. A Micropatterning Strategy to Study Nuclear Mechanotransduction in Cells. Micromachines (Basel) 10, doi:10.3390/mi10120810 (2019). Chapter Three is a version of the following paper: Fernandez, A., Bautista, M. & Pinaud, F. Emerin oligomerization and nucleoskeletal coupling at the nuclear envelope regulate nuclear mechanics against stress bioRxiv 2021.02.12.429834; doi: https://doi.org/10.1101/2021.02.12.429834 iv TABLE OF CONTENTS DEDICATION ii ACKNOWLEDGEMENTS iii LIST OF FIGURES viii ABSTRACT xii CHAPTER 1: INTRODUCTION 1 1.1 THE NUCLEAR ENVELOPE 1 1.2 THE MOLECULAR ROLE OF EMERIN 3 1.3 EMERIN IN DISEASE 5 1.4 THE MECHANOADAPTABILITY OF NUCLEI 8 1.5 THE CELLULAR ROLE OF EMERIN IN MECHANOTRANSLATION AND NUCLEUS POSITIONING 10 1.6 LIMITATION OF CURRENT APPROACHES FOR STUDYING EMERIN 12 1.7 BENEFITS OF SUPER RESOLUTION MICROSCOPY 14 1.7.1 BENEFITS OF LOCALIZATION MICROSCOPY 16 1.8 DETERMINING THE SPATIAL ORGANIZATION OF EMERIN 23 1.9 MICROPATTERNING TO MIMIC EXTRACELLULAR STRESS TO THE NUCLEUS 28 CHAPTER 2: ENGINEERING MICROPATTERNED SUBSTRATES TO EVALUATE THE MECHANOTRANSDUCING FUNCTION OF EMERIN 31 v 2.1: INTRODUCTION 31 2.2: ENGINEERING OF PATTERNED SUBSTRATES FOR CELL ATTACHMENT 34 2.3 PERFORMING LIGHT MICROSCOPY ON PATTERNED COVERSLIPS 36 2.3: NUCLEAR ADAPTATION TO EXOGENOUS MECHANICAL CUES AND IMPACT OF EMERIN EXPRESSION 40 2.4: EVALUATION OF SINGLE MOLECULE MICROSCOPY ON PATTERNED COVERSLIPS 43 2.5 EFFECTS OF NUCLEAR DEFORMATION ON THE DIFFUSION OF EMERIN 45 2.6 CHANGES IN THE NANOSCALE ORGANIZATION OF EMERIN AT THE ENVELOPE OF DEFORMED NUCLEI 48 2.7 DISCUSSION AND CONCLUSION 52 CHAPTER 3: EMERIN OLIGOMERIZATION AND NUCLEOSKELETAL COUPLING AT THE NUCLEAR ENVELOPE REGULATE NUCLEAR MECHANICS AGAINST STRESS 67 3.1 INTRODUCTION 67 3.2 EMERIN DISPLAYS DISTINCT DIFFUSIVE BEHAVIORS AT THE NUCLEAR ENVELOPE 69 3.3 EMERIN ORGANIZES AS SLOWLY DIFFUSING MONOMERS OR OLIGOMERS AT THE INNER NUCLEAR MEMBRANE 73 vi 3.4 BAF BINDING MODULATES THE MOBILITY OF BOTH EMERIN MONOMERS AND OLIGOMERS DISTRIBUTED ACROSS THE INNER NUCLEAR MEMBRANE 77 3.5 EMERIN FORMS DISCRETE OLIGOMERIC NANODOMAINS SURROUNDED BY MONOMERS 80 3.6 EMERIN OLIGOMERS ARE STABILIZED BY LAMIN A/C AND SUN1 AND MODULATED BY NUCLEAR ACTIN AND BAF 82 3.7 INCREASED DIFFUSION OF EMERIN MONOMERS AND OLIGOMERIZATION UPON NUCLEAR ADAPTATION TO MECHANICAL STRESS 86 3.8 EMERIN MUTANTS INDUCE ABNORMAL NUCLEAR DEFORMATION AGAINST MECHANICAL STRESS 89 3.9 EMERIN MUTANTS DISPLAY DEFECTIVE OLIGOMERIZATION AT THE INNER NUCLEAR MEMBRANE 91 3.10 Δ95-99 MUTATION 94 3.11 P183H MUTATION 95 3.12 ABNORMAL REORGANIZATION OF Δ95-99 EMERIN IN RESPONSE TO MECHANICAL STRESS 98 3.13 DISCUSSION 100 3.14 MATERIALS AND METHODS 105 CHAPTER 4: CONCLUSIONS 118 vii 4.1: CONCLUSIONS 118 4.2: FUTURE DIRECTIONS 122 BIBLIOGRAPHY 125 viii LIST OF FIGURES Figure 1.1 Emerin at the Inner Nuclear Membrane. 1 Figure 1.2 Binding domains of emerin. 3 Figure 1.4. Process of Localization Microscopy. 15 Figure 1.5. TIRF and HILO Illumination Diagram. 17 Figure 1.6. Experimental approach to PALM/STORM. 19 Figure 1.7. Calculation of Mean Square Displacement From Trajectories. 23 Figure 1.8. Comparison of Ripley’s K and Neighborhood Density Function. 25 Figure 1.9. Fitting exponential decay curve to autocorrelation data. 26 Figure 2.1. Cell micropatterning on HMDS-treated and fibronectin-functionalized coverslips. 34 Figure 2.2. PDMS stamps for fibronectin microstamping. 36 Figure 2.3. Fluorescence confocal and TIRF microscopy of cells micropatterned on HMDS coverslips. 37 Figure 2.4. Thickness of PDMS after spin-coating on glass coverslips. 39 Figure 2.5. Emerin deficient nuclei improperly adapt nuclear shape to exogenous mechanical cues. 40 Figure 2.6. Fluorescence confocal imaging of emerin distribution in wild type Emd +/y human skin fibroblasts as a function of nuclear strains. 41 Figure 2.7. Quantification of emerin redistribution in response to nuclear mechanical strains. 42 ix Figure 2.8. Single molecule tracking of emerin at the nuclear envelope of micropatterned cells. 44 Figure 2.9. FRAP of emerin at the bottom nuclear membrane. 46 Figure 2.10. Localization accuracy in 3D-dSTORM images of emerin at the nuclear membrane. 48 Figure 2.11. Super-resolution imaging and cluster analyses of nuclear membrane emerin in cells. 50 Figure 3.1. Emerin displays multiple diffusive behaviors at the nuclear envelope. 70 Table 3.1: Diffusion coefficients of wild-type and mutated emerin determined by PDSD analyses after sptPALM or CALM imaging. 71 Figure 3.2. FRAP of emerin in U2OS cells. 73 Figure 3.3. Immunoblotting and quantifications of siRNA knock down against lamin A/C, IPO9, BAF and SUN1. 74 Figure 3.4. Immunostaining to assess the direct and indirect effects of RNAi on the nuclear localization of lamin A/C, SUN1 and BAF 75 Figure 3.5. Emerin diffuses as inner nuclear membrane monomers and oligomers that interact with lamin A/C, nuclear actin and BAF. 76 Figure 3.6. Emerin forms oligomeric nanodomains stabilized by lamin A/C and surrounded by larger emerin monomer areas across the nuclear membrane. 80 Table 3.2. Molecular densities and domain sizes of wild-type and mutated emerin determined by spatial distribution analyses of emerin in super-resolved images. 82 x Figure 3.7. Mechanical stress on the nucleus increases emerin mobility and the formation of emerin oligomers. 86 Figure 3.8. Comparison of wild-type emerin diffusion and nanoscale organization after nuclear actin depletion or cell micropatterning. 87 Figure 3.9. Emerin mutations induce defective nuclear shape adaptation against mechanical stress. 90 Figure 3.10. Diffusive behaviors of Q133H and wild-type emerin. 91 Figure 3.11. Emerin mutants exhibit modified lateral mobilities and defective oligomerization at the inner nuclear membrane 92 Figure 3.12. Insufficient oligomerization of EDMD-inducing Δ95-99 emerin mutant against mechanical stress. 98 Figure 3.13. Model of emerin re-organization and oligomerization at the nuclear envelope in response to mechanical challenges. 100 xi ABBREVIATIONS BAF – Barrier-to-autointegration factor dSTORM – Direct Stochastic Optical Reconstruction Microscopy EDMD – Emery Dreifuss Muscular Dystrophy ER – Endoplasmic Reticulum HDAC – Histone Deacetylase IDP – Intrinsically disordered protein INE – Inner Nuclear envelope LEM – LAP2, emerin, and Man1 LINC – Linker of Nucleoskeleton and Cytoskeleton NE – Nuclear Envelope NET – Nuclear Envelope Transmembrane ONE – Outer Nuclear envelope PALM – Photoactivation Light Microscopy PSF – Point Spread Function SIM – Structured Illumination Microscopy sptPALM – Single Particle Tracking Photoactivation Light Microscopy STORM – Stochastic Optical Reconstruction Microscopy TM – Transmembrane xii ABSTRACT Emery-Dreifuss muscular dystrophy (EDMD) is a laminopathy that results in progressive muscle degeneration and early death. EDMD is caused by mutations in lamin and the inner nuclear envelope protein emerin, both of which are responsible for nucleo-cytoskeleton mechanotransductions, maintenance of nuclear shape, and genome organization. The molecular pathogenesis of the disease, however, remains unclear. To further understand the role of emerin at the molecular level, we have employed single particle tracking and three-dimensional super resolution microscopy and we quantified the diffusion and spatial organization of emerin and its clinically relevant mutations. Using diffusion and pair-correlation analyses on both live and fixed cells, we identified subpopulations of emerin associated with the endoplasmic reticulum, outer nuclear envelope, and for the first time showed that emerin organizes into two distinct populations of monomers and oligomeric nanoclusters at the inner nuclear envelope. The diffusion and clustering state of these subpopulations are directly impacted by EDMD-associated mutations and are linked to a complex interplay between lamin A/C binding, nuclear actin binding, BAF interaction, LINC Complex association, and emerin oligomerization. We also studied how mechanical strains applied to the nucleus influence the normal and pathogenic organization of emerin and how emerin participates in the adaptation of nuclear shape to forces. To do so, we developed a cell micropatterning strategy that is based on microcontact printing of fibronectin after organosilane monolayer functionalization of glass coverslips. This approach uniquely combines exogenous control of nuclear shape with nanometer accuracy single molecule microscopy imaging of cells. Super-resolution imaging of emerin in micropatterned cells reveals that the mechanotransducing functions of emerin are intimately coupled to its nanoscale oligomerization and its dynamics within the inner nuclear envelope. We additionally show that emerin mutations xiii result in its defective oligomerization and aberrant dynamics at the nuclear envelope, which leads to abnormal nuclear mechanics and improper adaptation to mechanical strains. Combining single molecule imaging and cell micropatterning techniques, we reveal key molecular mechanisms of EDMD pathogenesis that would have otherwise remained undetected by traditional diffraction- limited microscopy. CHAPTER 1: INTRODUCTION 1.1 THE NUCLEAR ENVELOPE Mammalian cells are defined by having membrane bound organelles, most notable among them is the one that contains the genetic contents of the cell: the nucleus. A unique feature of the nucleus, as opposed to most other cellular organelles, is its periphery, which is defined by two separate phospholipid bilayers. Most other organelles have a single lipid bilayer that separates their contents from the cytoplasm and allows them to carry out compartmentalized biochemical reactions. Contrarily, the nucleus contains two unique lipid bilayers, which are distinguished by their respective lipid and protein contents 5 . The outer nuclear envelope (ONE) is contiguous with the endoplasmic reticulum (ER), is exposed to the cytoplasm, and interacts with the cytoskeleton. Proteins found here bind to and interact with the cell cytoskeleton (Fig. 1.1). Approximately 10 Figure 1.1 Emerin at the Inner Nuclear Membrane. Emerin localizes to the inner nuclear membrane along with the other LEM domain proteins Lap2 and Man1 1 . Emerin interacts with the lamin network, BAF, and chromatin 2 . Also shown are the SUN proteins which are part of the LINC complex, nesprin, and cystoplasmic actin 3 , which are all linked together (adapted from reference 4 ). 2 nm below the ONE is the inner nuclear envelope (INE), which faces the contents of the nucleus. Across both ONE and INE, nuclear pore complexes (NPCs) operate as sieves to regulate the traffic of proteins and other macromolecules into and out of the nucleus 6 (Fig. 1.1). Below the INE is a meshwork of type V intermediate filaments: lamins 7 . Lamins provide stiffness to the nucleus and act as a scaffold onto which nuclear actin and DNA organize 8 . Lamins intimately interact with many proteins of the INE and are thought to be responsible for retaining and enriching these proteins on the nuclear side of the nuclear envelope 9 . Access to the INE is constrained by NPCs, which restrict diffusion of membrane proteins from the ONE to the INE by size exclusion 10 . Most of the non-membrane bound proteins do not access the nucleoplasm unless they contain a nuclear trafficking signal. One group of proteins that are enriched at the nuclear envelope are known as nuclear envelope transmembrane proteins (NETs) 11 . These NET proteins, including SUN proteins, nesprins, and our protein of interest emerin 12 , have been implicated in a wide variety of cellular functions, ranging from mechanotransduction, silencing of heterochromatin, activation of growth factor signaling, and reformation of nuclei after mitosis 13-16 . The primary localization of NET proteins at the INE make these favorable proteins for organizing chromatin in a specific manner. Indeed, one function of many of these proteins is to regulate DNA compaction, which makes regions of the genome accessible or inaccessible to transcription factors 17,18 . In addition to being critical for mitosis, these proteins are also crucial for cell differentiation 19 . A subset of NETs proteins bind to the lamin meshwork and are characterized by the presence of a LAP2-Emerin-MAN1 (LEM) domain in their peptide sequence, which is named after the proteins that contain this domain 20 . Mutations in some NETs proteins that bind lamin (e.g. FHL1B, LUMA 21 ), in lamin A/C itself, or in some LEM- 3 domain protein, such as emerin, cause diseases known as laminopathies 22 . More specifically, mutations in emerin lead to a disease known as X-linked Emery Dreifuss Muscular Dystrophy 23 . 1.2 THE MOLECULAR ROLE OF EMERIN Emerin is a small protein found in the INE of mammalian cells 24 . It is characterized by an N- terminal LEM domain, a large intrinsically disordered domain, a one pass transmembrane domain, and a small C-terminal domain 23,25,26 (Fig. 1.2). As an intrinsically disordered protein (IDP), emerin can adopt a variety of conformations, is known to oligomerize in vitro and has many binding partners at the nuclear envelope. This has made study of the protein difficult, as individual mutations can result in the loss of, or gain of binding ability to different proteins. Indeed, IDPs are characterized by a lack of a fixed three-dimensional structure, which allow them to fill a variety of roles that other, more structured proteins are unable to perform. The only structured domain of emerin, beside its transmembrane domain, is the LEM domain, which is found at the N-terminus of the protein. This LEM domain is a 40 amino acid sequence composed of a helix-loop-helix motif and is primarily responsible, among other functions, for binding to BAF and indirectly to DNA 27 (Fig. 1.2). Between the LEM domain and the Figure 1.2 Binding domains of emerin. A few of the binding partners of emerin and the regions of the protein to which they bind are shown, in addition to the transmembrane domain and the small C- terminus. The LEM domain is responsible for binding to BAF and DNA 27 . (adapted from reference 28 ). 4 transmembrane domain is an intrinsically disordered middle domain of emerin, which is a region responsible for lamin and nuclear actin interactions, as well as for emerin oligomerization (Fig. 1.2). Interactions with the nucleoskeleton, and in particular with the lamin A/C meshwork, is thought to be critical for retaining emerin at the INE and for modulating nuclear responses to mechanical stress via nuclear lamina reorganizations 29 . The intrinsically disordered region of emerin has also been implicated in binding to Lmo7 and HDAC3 30 , two proteins involved in epigenetic modifications and which, together with BAF, could be vehicles through which emerin might influence heterochromatin organization and gene expression. While emerin has long been known to interact with the lamin meshwork, more underappreciated has been its role in binding nuclear actin. In addition to its cytoplasmic roles, actin is found in the nucleus 31 and its polymerization is modulated by both emerin and lamin 32,33 . This modulation of actin polymerization might be responsible for altering the organization of the nucleus and its response to stress. Indeed, under nuclear stress, perinuclear actin is polymerized 34,35 , which could also influence the polymerization of the nuclear actin via the LINC complex 36 which itself is organized by emerin 37-39 . In vitro, emerin has been shown to polymerize actin 32 , but how exactly this is manifested in the nucleus and how the presence of nuclear actin impacts the behavior and organization of emerin is yet unknown. Initial biophysical studies of the dynamics of emerin by Fluorescence Recovery After Photobleaching FRAP 40 have revealed significantly slower diffusion in the nuclear envelope as compared to the ER (Fig 1.3). While this indicates that emerin binding partners are found in the nuclear envelope, the slow diffusion rates of emerin and its incomplete total recovery limits the usefulness of this ensemble analysis technique to understand emerin diffusion in details. The fact that the protein does not fully recover by FRAP implies that a large portion of the emerin 5 population cannot be thoroughly analyzed using photobleaching. While large differences in diffusion coefficient can reveal differences between monomer and multimer populations, a large, unmeasured population can leave multimer populations unstudied 41 . 1.3 EMERIN IN DISEASE Emerin was first discovered as a protein implicated in causing Emery-Dreifuss Muscular Dystrophy (EDMD) 42 , a progressive muscle wasting disease that leads to abnormal skeletal muscle coordination, degenerative heart failure, and ultimately early death 43 . Dystrophies are a wide variety of diseases lacking clear symptomatic differences, as mutations in an array of genes can cause similar dystrophic symptoms 44 . While early on most dystrophy related proteins were found to be associated with the sarcolemma 45 , emerin, interestingly, was found to localize primarily to the INE 24 and interact with the nuclear lamina 46 . EDMD was later found to be related to a more specific group of muscular dystrophies called laminopathies, all of which are caused by mutations Figure 1.3. FRAP analysis of emerin diffusion in INE and ER. Early studies of emerin diffusion found significantly faster diffusion in the ER versus at the nuclear membrane. Diffusion coefficients of 0.32 ± 0.01 and 0.10 ± 0.01 µm 2 /s in the ER and nuclear envelope respectively (adapted from reference 40 ). 6 in proteins found at the nuclear envelope 47 . Most surprisingly, emerin itself is expressed in all somatic tissues yet causes a muscle specific phenotype 48,49 . This has produced questions as to how a protein without tissue specific expression could lead to a muscle specific phenotype and why, in a subset of mutations (notably P183H, Q133H, and Δ95-99), emerin seems to organize similarly to the wildtype form yet still leads to the dystrophic phenotype 50,51 . While most mutations involving emerin are either nonsense or indels that produce frameshifts and cause a complete ablation of protein expression 52 , a small subset of point mutations instead produce mostly normal expression and localization of the protein, yet still cause muscular dystrophy 50,51,53 . Clearly, these point mutations ablate critical roles for emerin, which would make sense given that these are mostly found in the nucleoplasmic domain of emerin. In this thesis, we have studied these specific mutations as well as wild-type emerin to discover how emerin maintains normal nuclear responses to mechanical stimuli and to define key aspects of the pathogenesis of EDMD. After the initial discovery of emerin, other proteins were also implicated in causing EDMD. Mutations in lamins were found to cause the same disease 22 , which is not surprising given that emerin and lamin were ultimately found to be binding partners (Fig. 1.2) 46 . Emerin and its binding partners have been implicated in multiple roles, including nuclear reorganization after mitosis 54 , heterochromatin recruitment to the nuclear periphery 55 , lamin recruitment to the nuclear periphery 54 , capping of actin filaments 32 , and transmission of mechanical force to the nucleus 16 . It is this last role that has sparked the most interest into emerin and its relation to EDMD. Proper regulation of nuclear stiffness is lacking in cells without emerin 56 . This would implicate emerin as having a role in mechanosensing and mechanotransduction at the nuclear envelope. As cells can conform to a variety of shapes and are known to rearrange their cytoskeletons to adapt to 7 extracellular mechanical constraints 57 , mechanosensors fill the role of regulators during these processes by controlling when and how a cell responds to the mechanical cues 58 . For many dystrophy-related proteins, the connection to muscular dystrophy is relatively straightforward because they physically connect to the sarcolemma 59 and are critical for proper mechanical orientation. For nuclear envelope proteins involved in laminopathies, such as emerin, relationships to muscular defects has been more complicated to define. It is only recently that the nucleus became more appreciated as a mechanosensitive cellular structure, as it must itself be reshaped and repositioned during cell motility, differentiation, and application of exogenous mechanical force 60 . Many of the proteins involved in aberrant mechanotransduction and induction of laminopathies are themselves found at or near the nuclear envelope and are known binding partners of emerin, including SUN 1/2, nesprins, and lamins (Fig. 1.2). Defects in mechanosensing proteins, which are involved with the cell adaptation to mechanical signals 61 , can lead to a variety of diseases though they mostly manifest in muscle tissues, since mechanosensors are critical in force generation 62 and muscle development 63 . With emerin, this process is not as obvious because at the inner nuclear envelope emerin is not directly bound to any of the cytosplasmic cytoskeletal proteins. So how is it that an inner nuclear envelope protein, such as emerin, is critical for maintaining mechanical stability? Two possible theories have been proposed to explain the muscle specific phenotype of emerin mutations. The first theory involves the mechanical role of emerin. It argues that cells lacking emerin improperly transmit mechanical force to the nucleus, thus leading to intracellular damage and eventually to cell death such as apoptosis 56 . Since this occurs mostly in muscle cells, muscle cells eventually waste away and are replaced with adipose tissue. The second theory hinges on improper chromosomal organization in emerin-null cells 64 . In addition its role in the stabilization 8 of the nucleus and the organization of chromosomes and membranes during mitosis, emerin was also associated regulation of gene transcription 65 . Because cells lacking emerin are known to misorganize portions of chromosomes 66 , it is possible that silencing of muscle specific genes could lead to the muscle specific phenotypes. Along the same lines, studies in C. elegans have shown that its emerin analogue EMR-1, is found more frequently co-localized at genes related to muscle and neuron function 65 . These mechanical and chromosomal hypotheses are not necessarily mutually exclusive and altered genomic regulations may be a consequence of the impaired nuclear adaptability to mechanical stresses. With the significant advances made in understanding the mechano-adaptability of nuclei in the past 10 years, recent studies have been more focused on the mechanical hypotheses of EDMD and the role of emerin in mediating mechanical response of the nucleus. 1.4 THE MECHANOADAPTABILITY OF NUCLEI The nucleus itself is known to have its own stiffness which is adaptable to the stresses that are placed on it 67 . This is important in many cellular processes, including cell differentiation and cell migration. For instance, differentiated muscle cells exhibit unique nuclear positioning as opposed to their undifferentiated progenitor cells 68 . Myoblasts have their nuclei at the cell periphery, while myotubes, which are the product of multiple myoblasts fused together, line up their nuclei in the same axis as the cell 69,70 . This process requires moving nuclei from their central position towards a defined alignment axis in myotubes. During this process, the nucleus must be squeezed as it moves through the cell’s cytoskeleton. Cells lacking inner nuclear envelope proteins that are 9 important for maintaining proper nuclear stiffness have difficulty undergoing this process and as a result exhibit impaired ability to properly differentiate into muscle cells 64 . This modulation of mechanical forces is not only important to maintain cellular integrity but also to prevent breaks in the chromatin, especially in the heterochromatin which is found at the nuclear periphery. Real time imaging of the chromatin in isolated nuclei show that the chromatin is highly dynamic and able to reorganize and respond to changes in both epigenetics and mechanical stress 71 . These dynamic changes, however, are only one part of the overall response of the nucleus. The nuclear periphery additionally reacts to changes in the cellular environment. Previous experiments have shown that the nucleus stiffens in response to consecutive tugging forces, and that nuclear stiffening is impacted by key nuclear envelope proteins 72 . The stiffening response itself is likely achieved by a rearrangement of lamin A/C intermediate filaments 73-75 , found immediately beneath the INE 76 . Lamin itself is directly implicated in muscular dystrophy, as mutations of the protein lead to autosomal EDMD 77 . Emerin, with its LEM domain, directly binds to the nuclear lamina and is critical for proper mechanotransduction 56 . How emerin, itself, adapts and imparts this response is yet unknown. In addition to lamin and chromatin, nuclear actin is also present in the nucleus and can contribute to nuclear stability 78 . In the cytoplasm actin is important for structural stability of the cell, trafficking of proteins, and is adaptable in response to force applied on the cell. Although far less is known about the role of actin in the nucleus, it enters the nucleus through IPO9 nuclear channels 79 where it can form transient filaments 80 . Additionally, it can interact with lamin through its tail, which bundles actin filaments 81 . Interestingly, emerin is found at the interface of all of these interactions and, with both the ability to bind lamin and actin (Fig. 1.3), it likely plays a role 10 in organizing all of these proteins, maintaining their correct associations and orchestrating the adaptation of the nucleus to changing mechanical environments. 1.5 THE CELLULAR ROLE OF EMERIN IN MECHANOTRANSLATION AND NUCLEUS POSITIONING Previous studies have shown that emerin is critical to both the stiffening response of the nucleus 72 and the direct organization of the cytoskeleton 82 . Emerin was also shown to be one of the first proteins to surround the reforming nucleus at the end of mitosis 83,84 , and it is required for proper nuclear reformation as gross nuclear abnormalities are more common in cells lacking emerin 85 . Emerin surrounds the condensed nuclei and helps to reform the nuclear envelope, interacting with lamin A/C and BAF to get all of the chromatin back into a single nucleus after division 84 . In addition to these roles, emerin is also crucial for maintaining the proper mechanical stiffness of the nucleus. Cells without emerin have been shown to lack the nuclear stiffening response 72 , resulting in an inability to resist exogenous stresses 86 . Additionally, cells lacking emerin cannot properly organize their cytoskeletons and, as a consequence, cannot position their nuclei correctly both in fibroblasts 82,87 and muscle cells 88 . These roles are likely interconnected, as the nucleus must respond locally to the perturbations transmitted by the cytoskeleton. In cells undergoing mechanical changes, instead of forming parallel actin filaments and a proper actin cap, the actin filaments organize randomly around nuclei lacking emerin 89 . This is likely responsible for the improper localization and organization of nuclei which are characteristic of dystrophic muscle cells 90 . Muscle cells organize into long, multi- nucleated structures called myotubes that are the product of fusion of neighboring cells 91 . When 11 these cells lack emerin, they are characterized by disordered nuclear organization and higher rates of apoptosis 66,92 . How emerin prevents apoptosis and encourages proper nuclear and cytoskeletal organization is unclear. Emerin also interacts with the LINC complex, an ensemble of proteins that are responsible for coupling the nuclear lamina to the cytoskeleton across the nuclear envelope 93 When cells are placed under stress, these LINC complex proteins form transmembrane actin-associated nuclear (TAN) lines. Along these TAN lines, an actin cap forms around the nucleus 94 . This actin cap puts pressure on the nucleus, helps to compress it, but is also important for reducing the overall stress on its structure. Cells lacking lamin are unable to form an actin cap, which results in a disordered nucleus when stress is placed on the cell 95 . However, cells lacking emerin are also unable to correctly form an actin cap, and display disorganized actin bundles above the nucleus, that are linked to the altered nuclear morphology and improper nucleus positioning of emerin-null cells 82 . In addition to its role in adapting the nucleus to mechanical stress and positioning nuclei correctly, emerin also plays a role in cell division, and intriguingly, emerin is found to localize in intercalated discs in heart muscle cells 96 . Emerin mutations, when expression is maintained, induced an improper localization of emerin with an enrichment at the endoplasmic reticulum and outer nuclear envelope 54 . Additionally, cells undergoing excessive stress on the nucleus also show relocation of emerin from the INE to the ER 97 . The purpose of this process is unknown, but the fact that emerin partially leaves the nuclear envelope when cells are subjected to physical strains is reminiscent of similar observations made for some emerin mutations 98 . This process could indicate that cells cannot adequately respond to mechanical stress when emerin is mutated or is not properly retained at the nuclear envelope. Another explanation could be that the mechanical response is always 12 improperly engaged, which could lead to nuclear fragility when cells are placed under stress and could explain the higher rates of apoptotic cells expressing mutated forms of emerin. 1.6 LIMITATION OF CURRENT APPROACHES FOR STUDYING EMERIN While emerin mutations that cause an EDMD phenotype likely induce changes in emerin organization and dynamics, traditional microscopic techniques have until now been unable to determine how such changes correlate with defective transmission of mechanical stress to the nucleus and subsequent pathogenic cell behaviors. Traditionally, investigations into the dynamics of emerin have been performed at the ensemble level, using fluorescence recovery after photobleaching (FRAP) 40,99 . In FRAP, large areas of fluorescently labeled emerin are photobleached and fluorescence recovery is recorded as fluorescent emerins diffuse back into the photobleached area and photobleached emerins diffuse out of this same area. However, determining diffusion coefficients by FRAP has limitations that become especially pronounced for proteins having multiple diffusion modalities, notably because faster diffusion behaviors can obscure slower ones 41 . Because FRAP does not probe individual protein diffusion, but rather extrapolates diffusive behaviors from the dynamics of an entire population at the ensemble-level, studying slow-diffusing membrane proteins such as emerin becomes problematic. For instance, emerin is known to diffuse significantly faster in the ER than at the nuclear envelope 40 . Because the ER is contiguous with the nuclear envelope, both populations of ER-associated emerin and nuclear envelope-associated emerin are averaged during FRAP experiment analyzes. Faster diffusion of emerin in the ER can therefore obscure the diffusive behaviors of slower emerin populations at the nuclear envelope, making detailed analyses of emerin dynamics particularly difficult. 13 Beyond the current limitation in determining the membrane dynamics of emerin, little is known about its spatial organization at the nuclear envelope and it is unclear if emerin mutations induce significant changes in this organization. While in vitro biochemical assays clearly showed that emerin can oligomerize 100 and that different emerin mutations can affect the binding of emerin to specific nucleoskeletal partners 30 , emerin oligomerization and the role of emerin mutations have not been established in vivo. In most cases, the cellular localization of emerin and its mutants appears homogenous and somewhat normal at the ensemble level, when assessed by traditional diffraction-limited optical microscopy such as confocal fluorescence imaging 101 . However, it is well known that diffraction-limited optical imaging cannot report on protein oligomerization or on changes in protein localization patterns at the nanometer scale, beyond ~200 nm. Current studies on emerin have only been able to establish that most mutants lead to an absence of emerin expression or, in the case of a small subset of mutations, to a normal expression associated with a slight decrease in retention at the nuclear envelope. It remains unclear if and how emerin oligomerizes in live cells and how emerin mutations that impact binding to nucleoskeletal partners lead to manifestation of muscular dystrophy. While it is known that emerin can form dimers and higher order structures, such organizations are unlikely to be revealed by traditional and diffraction-limited microscopy techniques. Similarly, any differences in these organizations due to emerin mutations would remain inaccessible unless they are probed with nanometer scale precision. As such, in the context of nuclear mechanotransduction, the role of emerin oligomerization, the importance of its interactions with nuclear binding partners, and the impact of EDMD-inducing emerin mutations on these processes remains uninvestigated. In this thesis manuscript, I describe how we used single molecule super-resolution optical microscopy techniques to: (i) directly quantify the cellular dynamics of emerin and emerin mutants 14 at the nanometer scale in living cells and (ii) reveal their nanoscale spatial distribution and the importance of emerin oligomerization at the nuclear envelope in the context of nuclear mechanical strains and EDMD. Our results uncovered key molecular mechanisms of emerin mechanotranslation and of EDMD pathogenesis that would have otherwise remained undetected by traditional diffraction-limited microscopy. 1.7 BENEFITS OF SUPER RESOLUTION MICROSCOPY As an alternative to traditional microscopic techniques, which have thus far shown limited differences between wildtype and mutated versions of emerin, we used super resolution optical techniques which are perfectly suited to investigating small differences in the organization and the diffusion of emerin at the nanoscale. The reason why diffraction-limited and ensemble microscopy techniques have not been able to establish differences between wildtype and mutated emerin is that their optical resolution is limited to ~200 nm, a size far greater than that of a typical individual protein (e.g. 2-3 nm). Due to the wave nature of light, molecular structures smaller than a few hundred nanometers cannot be resolved and information regarding the nanoscale distribution and dynamics of proteins are lost using traditional optical imaging techniques 103,104 . In order to resolve separate objects, they must be separated by a distance of ∆𝑥 ≥ 𝜆 /2𝑁𝐴 , which represents the Abbe criterion, or ∆𝑥 ≥ 0.61𝜆 /𝑁𝐴 , where λ is the wavelength of the light and NA is the numerical aperture of the objective 105,106 . 15 Figure 1.4. Process of Localization Microscopy. Individual PSF’s are found on a detector and gaussian curves are fitted to the signal distribution to accurately determine the centroid position and estimate the real location of the emitter with a resolution far lower than the original acquisition. (adapted from reference 102 ). 16 1.7.1 BENEFITS OF LOCALIZATION MICROSCOPY Single molecule localization microscopy techniques, such as Photoactivated Localization Microscopy (PALM) 108 , single particle tracking photoactivated localization microscopy (sptPALM) 109 , Stochastic Optical Reconstruction Microscopy (STORM) 110 and the related technique direct Stochastic Optical Reconstruction Microscopy (dSTORM) 111 , can overcome this technical limitation and allow for estimation of the real position of individual proteins at the nanoscale by preventing the overlapping of multiple emission signals inherent to ensemble measurements 112 . These approaches rely on the stochastic activation of a small subset of labeled proteins over multiple frames, such that the entire point spread function (PSF) of an individual emitter can be visualized. The PSF information can be used to estimate the centroid position of the emitter using Gaussian curve fitting, and the position of the labeled emitter can be determined with a precision on the order of a few tens of nanometers for membrane diffusing molecules in live cells or below ten nanometer for immobilized molecules in fixed cells (Fig. 1.4). The only limitation on the localization precision is the number of photons emitted during each detection according to the following equation: 𝜎 𝜇𝑖 = √ 𝑠 𝑖 2 𝑁 + 𝑎 2 /12 𝑁 + 8𝜋 𝑠 𝑖 4 𝑏 2 𝑎 2 𝑁 2 , where σ is the localization precision, N is the number of collected photons, a is the pixel size, b is the standard deviation of the background, and s is the width of the PSF distribution 113 . Previous studies have reported localization precisions as high as 1.5 nm 114 , which is on the scale of the size of a typical protein. 17 To maximize localization precision, illumination techniques have been developed in order to minimize excitation of background emitters while maximizing excitation of in-focus molecules in cells. For proteins embedded in the cell plasma membrane, total internal reflection fluorescence (TIRF) illumination only excites membrane bound emitters because the TIRF evanescent wave penetrates only a few hundred nanometers beyond the plasma membrane, and thus drastically reduces background signal 115,116 . Alternative approaches to image structures deeper within cells while still maximizing signal-to-background ratios include forming thin optical light sheets 117,118 that can be swept across the cell. One simple way to obtain an optical light sheet is to implement highly inclined and laminated optical sheet (HILO) 107 microscopy, a derivative of TIRF. In both TIRF and HILO, laser illumination of cells is performed at a specific angle. In TIRF, a critical angle is chosen such that laser excitation light is completely reflected at the coverslip/buffer interface to create an evanescent field at the cell plasma membrane. In HILO, this angle is adjusted just off the TIRF critical angle to obtain an oblique and thin illumination light sheet that penetrate the cell to excite cellular structures such as the nuclear envelope (Fig. 1.5). In HILO, the illumination and detection paths are coupled in the same microscope objective, which results in an increased thickness of the light sheet when trying to image deep inside cells. Other light sheet microscopy techniques, where illumination and detection paths are uncoupled, are more suited to Figure 1.5. TIRF and HILO Illumination Diagram. TIRF and HILO change the incident angle of the emission laser to minimize background excitation and only illuminate in-focus molecules to achieve the highest localization precision. A) Light path diagramed at the objective showing the path for standard epifluorescence and TRIF and HILO. B) Light path through the sample for HILO illumination diagramming the thin light path passing through the cell. (adapted from reference 107 ). 18 penetrating deeper into tissues 119 , though, at the bottom nuclear envelope, HILO produces a light sheet only a few microns thick, which is sufficient for performing single molecule localization microscopy 107 . In super-resolution microscopy, the PSF of individual molecules are imaged separately in time and space by stochastic activation to limit PSF overlap of all the emitters in a cell. Given long enough acquisitions, the entire population of emitters can be imaged, localized, and mapped to produce an image with a resolution tenfold that of traditional light microscopy (Fig. 1.6). Additionally, characteristics such as local protein density and nanoscale cluster sizes can be extracted by combining this method with spatial pattern analyses 104 . Super-resolution imaging approaches is critical for proteins such as emerin, which are known to dimerize and to form higher order structures. Indeed, dimers and multimers are far smaller than the diffraction limit of light, and thus are indistinguishable from monomers with the lower spatial resolution of traditional light microscopy. In this work, we used direct stochastic optical reconstruction microscopy (dSTORM) 110,111 to examine the nanoscale distribution of emerin at the nuclear envelope of cells and determine if there are indeed difference in membrane organization between wildtype and mutated versions of emerin. Because emerin is known to self-associate in vitro and to form filament-like structures that are modified by EDMD-inducing mutations 28 , we anticipate that such changes in emerin oligomerization and organization might be revealed and quantified in vivo using dSTORM and advanced spatial pattern analytical tools. Previous experiments using Structured Illumination Microscopy (SIM) 121 have found that some emerin binding partners organize into a distinct fiber meshwork at the nuclear membrane 122 , which suggests that emerin itself could be non-randomly organized at the nuclear envelope. As discussed in the next section, quantifying the nanoscale 19 organization of emerin demands that point reconstruction super resolution imaging techniques, such as dSTORM, be combined with robust statistical analytics of its spatial distribution such as point-by-point spatial autocorrelation function analyses 104 , including O-ring point pattern statistics 123 that we have utilized in this work. In order to achieve stochastic activation and visualize individual PSF, special fluorescent chromophores and buffer conditions must be used to prevent all probes from switching on at once 124 . Many Alexa dyes have been found to be good candidates for this purpose, as their blinking characteristics and relatively good photostability are well suited for the requirements of Figure 1.6. Experimental approach to PALM/STORM. By activating a small subset of molecules at one time, the precise location of each individual labeled protein can be determined with high precision. Over time, all labeled proteins can be localized and plotted to produce high resolution images of protein distributions (adapted from reference 120 ). 20 dSTORM 125 . However, this alone is insufficient since most fluorophores are quickly destroyed upon illumination, and only a small proportion of the sample can be localized upon reaching single molecule detection densities. Buffer systems that provide chromophore photoswitching capabilities, while preventing excessive photobleaching have been developed in order to limit this issue 124 . For instance, multiple oxygen scavenging systems based on glucose oxidase and catalase pairs 126 are available to prevent photobleaching. In addition, fluorophore photoswitching can be accomplished with thiols, the most common being Cysteamine Hydrochloride (CHCL) and β- mercaptoethanol (ßME) 127 . Upon interaction with the chromophore, these thiol disrupts electron delocalization in the chemical structure and interferes with its spectral properties 128 . In dSTORM, removal of the thiols and switching the chromophore back to its original state is achieved by photoexcitation at lower wavelengths 129 . This allows for the controlled photoswitching of the fluorophores to attain fluorescence detection densities that are sufficiently low for single molecule localization microscopy. It also prevents photodestruction, which would impede image reconstruction. Fluorophore blinking, however, is not eliminated, and can interfere with studies of spatial organization based on single point analyses, as it induces over-counting effects in superresolved images. Some of these effects can be mitigated during image post-processing 130 and by insuring 1:1 stoichiometric labeling of chromophore per biomolecule of interest, in our case emerin. Traditional labeling of emerin by primary and secondary antibody pairing can introduce multiple dyes on a single protein, and lead to an overestimation of emerin density 131 . 1:1 ratio conjugations of chromophore to primary antibodies can be achieved but are complicated and tend to be inefficient. Additionally, large antibodies can produce steric hindrance, which may lead to an undersampling of protein density. As an alternative, a SNAP tag fusion allows 1:1 labeling of 21 protein with choromophore-benzylguanine (BG) conjugates. A fusion of emerin to the relatively small SNAP tag (19.4 kDa) ensures a 1:1 fluorophore labeling of emerin, which is important for good estimations of putative emerin oligomer characteristics. 1.7.2 SINGLE PARTICLE TRACKING As an alternative to the drawbacks inherent to FRAP that were discussed previously, we investigated the dynamics of emerin in living cells using single particle tracking 132,133 . Single particle tracking was first developed for use in cells with gold nanoparticles targeted to plasma membrane proteins 134 . Later, following the development of highly sensitive electron-multiplying charged-coupled device (EMCCD) cameras, single particle tracking was extended with the use of fluorescent dyes in artificial membranes 135 and at the plasma membrane of cells 136 . Most single particle tracking experiments are done at the plasma membrane, though, with emerin, we track diffusion at the nuclear membrane. Single particle tracking is a more suitable technique than FRAP for determining the complex dynamic behaviors of emerin because it reports on the diffusion trajectories of individual emerin molecules in cells 137 . As such, it allows us to separately examine the fast ER diffusion of emerin from its slower diffusion rates at the nuclear envelope and to analyze nanoscale diffusion dynamics of emerin and its EDMD-inducing mutants that would otherwise be obscured in FRAP experiments 135 . There are two prominent single molecule techniques for studying this type of diffusion: fluorescence correlation spectroscopy 138,139 (FCS) and single particle tracking. While FCS is a viable method for studying the diffusion of emerin at the nuclear membrane 140 , single particle tracking has been found to be a more accurate method with lower uncertainty 141 . The major 22 limitation of single particle tracking is the ability to accurately identify unique emitters, which can lead to errors in diffusion estimation at high emitter densities 141 . While emerin is present at high density at the nuclear membrane, this limitation can be overcome by using the technique in conjunction with PALM, in a technique coined sptPALM 109 . This combination of techniques allows for high density single particle tracking to be performed in live cells and is a good method for estimating the dynamics of slow membrane proteins Single particle tracking involves localizing an individual emitter from frame to frame and measuring its square displacement over multiple time intervals. This can easily be done with long- lived emitters such as quantum dots 142,144 , though this unavoidably limits the number of proteins that can be labeled in an individual cell since the emitter density must be sufficiently low to visualize the entire PSF per frame. Alternatively, fluorescent protein fusions and sptPALM can be used, although it generates shorter length trajectories because of photobleaching 109 . sptPALM has been used with EGFP and with different photoactivatable and photoconvertable fluorescent proteins such as mEOS2, Dendra2, pA-GFP or pA-mCherry, which is particularly well suited to the technique 145 . Once the emitters are localized frame by frame and linked together to build trajectories, trajectories are subsequently analyzed using a stepwise square displacement analysis. Briefly, distances are squared and averaged at multiple time lags as diagramed in Fig. 1.7. MSD curves can be generated trajectory by trajectory, though with the use of short trajectories this introduces significant error and obscures multiple diffusive behaviors that a single protein can exhibit 146 . An alternative method, Probability Density of Square Displacement (PDSD) 135 , similarly measures square distances, but instead of averaging displacement for a single trajectory it reports on all the square displacements for each time lag of every trajectory. Multiple square displacement populations are fitted to the data at each time lag to generate multiple square 23 displacement curves, which can be further fitted with diffusion models to characterize the multiple modes of diffusion of an entire population. PDSD analyses allow the discovery of differential modes of diffusion, whether confined, free, or directed (Fig. 1.7), and provides diffusion coefficients for these different modes. As such PDSD analyses can help distinguish between monomeric and oligomeric populations, but can also identify differential modes of diffusion in different membrane environments 109 . In our specific case, emerin diffusion data will be compared to super resolution data to cross-validate putative monomer/oligomer identifications and to further describe the heterogenous organization/behavior of emerin at the nuclear envelope. 1.8 DETERMINING THE SPATIAL ORGANIZATION OF EMERIN As discussed previously, one of the biggest obstacles to understanding the molecular differences between wildtype and mutated emerin is the limited resolution afforded by standard light microscopy techniques. In the case of mutations where emerin expression is maintained, the Figure 1.7. Calculation of Mean Square Displacement From Trajectories. A) Distances are measured, squared, and averaged for multiple time intervals. B) MSD values for multiple time lags are generated according to the displayed formula. C) MSD curves are generated to reveal both diffusion coefficient and the mode of diffusion (whether random, directional, or confined). Brownian motion will be a straight line, while directed motion is parabolic and confined motion has a horizontal asymptote. (Adapted from references 142,143 ). 24 protein localizes properly at the nuclear envelope, but its nanoscale distribution and oligomeric organization is unknown. As previously demonstrated for super-resolved single molecule localization images of plasma membrane proteins 147-149 , the structural organization of membrane- embedded proteins, such as emerin, can be quantified and compared across different conditions using spatial pattern analytical tools. Spatial data sets in super resolved images of emerin will consist of millions of mapped points corresponding to the respective locations of the protein at the nuclear envelope. There are a variety of robust statistical techniques that allow the detection and the analysis of spatial point patterns for such large data sets in order to detect the clustering state of emerin (e.g. monomers or oligomers), but also the structure and the molecular density of these clusters at different length scales along the nucleus membrane. These techniques have been used to describe, for example, the clustering of the IgE receptor in mast cells via pair autocorrelation analyses 150 and the reorganization of the IL-15 receptor upon binding of its ligand via the Getis and Franlkin local clustering analysis 151 . Spatial point pattern analytical functions were originally developed in the field of ecology research to describe the organization of trees over wide land areas, but they can similarly be used to describe clustering present in any spatial distribution of individual proteins 153 . For instance, Ripley’s K algorithms were one of the early functions developed to characterize spatial organization 154 . This function counts the number of events found within a circle of radius r for every point in the distribution and effectively describes spatial homogeneity in a distribution of points (Fig. 1.8). A criticism of Ripley’s K analyses is that clustering at small distances can lead to a perceived clustering at larger distances, even if at larger distances no contribution to clustering is made. To solve this issue and determine only the distance at which new clustering is added, a variant of Ripley’s K called the Neighborhood Density Function (NDF) was developed 155 . NDF is similar to 25 Ripley’s K, in that it determines the number of neighboring points around each data point for all radii, except that, at each radius, it also subtracts the points from the previous radius. Effectively, it only searches for points within an annulus with large radius r i and small radius ri-1 (Fig. 1.8). Additionally, the number of detected points at each radius can be divided by the area of the annulus, and this density can be compared to the overall density on the entire region of interest. This allows for averaging over multiple ROI’s to increase statistical power and allow for running tests of statistical significance. Alternative spatial point pattern analytical tools include spatial autocorrelation and cross-correlation function 150 , raster scans 156 , and nearest neighbor analyses 157 . In addition the aforementioned global clustering analyses, local clustering analyses can also be used to directly map cluster values back onto the analyzed ROI in microscopy images 155 . This allows for generation of cluster maps and provides a direct visualization of the cluster themselves. From such maps, one can visually compare cluster distributions and positions. Indeed, differences in organization could arise not just from protein cluster density and size, which we can extract from global analyses, but additionally from inter-cluster organization within cells, which cannot be probed from a global analysis Figure 1.8. Comparison of Ripley’s K and Neighborhood Density Function. A) Ripley’s K estimates the density of points for all circles of radius r i. B) Neighborhood density function eliminates the contribution of clustering at smaller distances by measuring only the density of points for an annulus of width r i – r i-1 (Adapted from 152 ). 26 Another problem that one must keep in mind during the spatial analysis of individual protein points are edge effects. Points near the edge of a cellular structure might appear to have increased clustering values 158 , especially as analytical radii become larger and sample the empty space beyond the cell structure of interest. To overcome this limitation, a number of edge effect correction algorithms have been developed, which we have employed 158 . Global NDF analyses performed across multiple ROI’s and multiple cells are characterized by an exponential decay of the NDF as a function of increasing radial distances (Fig. 1.9). From the number of exponential decays, the NDF amplitude and its decay rate, one can (i) identify different clusters in the spatial distribution of molecules and (ii) evaluate their relative molecular densities and their cluster sizes. This involves fitting the NDF curve with algorithms adapted for super resolution analyses and that includes mathematical corrections for spatial localization errors due to the photo-blinking behavior typically observed when individual fluorophores are imaged 104 . It is worth mentioning that significant pre-analytical treatments of super-resolved images need to be implemented before running spatial pattern analyses. Those include spatial drift corrections with fiducial markers, to limit the effect of drift in image reconstruction which can mislocalize events Figure 1.9. Fitting exponential decay curve to autocorrelation data. A) Data points were randomly organized into clusters of ten randomly shaped and distributed. B) Zoom in from data in panel A showing the distribution of points into discrete clusters. C) Autocorrelation analyses were performed to extract densities at all tested r distances. Exponential decay curves fit the extracted cluster data well, and the parameters of the peak and decay can be used to determine average cluster density and size. (Adapted from 104 ). 27 by hundreds of nanometers relative to other localizations. Fiducial markers that are constantly present around a cell sample during data acquisition can be used to eliminate drift simply by adjusting the positions of all points relative to those of the fiducials. Additional image treatments also include correction for single molecule over-counting, an inherent issue of STORM imaging, that may lead to an overestimation of cluster density 150 . One way around the over-counting issue is to account for successive appearances of an individual fluorophore within the estimated error in positioning. While this is efficient for single molecules reappearing over consecutive frames during imaging, it does not completely eliminate overcounting issues linked to fluorophore blinking, where an individual fluorophore temporally disappears (OFF) before re-appearing (ON) a few seconds or even minutes later during imaging, leading to multiple localizations for only one protein 128 . In order to solve this issue, we followed a previous analytical protocol 159 , where the localization error is included in the fitting parameters to estimate cluster size and density by being convoluted at every points along the NDF exponential decay curves. Once this convolution is applied, fitting multiple exponential decay curves can extract general parameters for cluster densities and sizes found over multiple ROIs. As shown in the result section of this dissertation, these analytical approaches allow (i) for a quantitative measurement of emerin distribution at the nuclear envelope and (ii) for a comparison of its nanoscale clustering modalities across multiple biochemical states of the cell nucleus and across various emerin mutants known to induce EDMD. It is only through this method that we can see if the molecular organization of emerin at the nuclear envelope is truly altered by the known perturbed binding capabilities of emerin mutants. Using comparison to RNAi experiments that knockdown known binding partners of emerin, we also defined the contribution of different emerin interactors in establishing the molecular distribution of emerin at the nuclear membrane and 28 studied the role of these unique organizational states in the maintenance of proper nuclear architecture and the prevention of cellular fragility in the context of EDMD. 1.9 MICROPATTERNING TO MIMIC EXTRACELLULAR STRESS TO THE NUCLEUS Emerin is implicated in transmitting mechanical stress from the cytoskeleton into the nucleus. In addition to defining the nanoscale organization and the dynamics of wildtype emerin and emerin mutants at the nuclear envelope of cells, we also aimed at providing new mechanistic insights into the mechanotransducing functions of emerin and into the pathogenesis of EDMD. Doing so demands means to controllably apply exogenous mechanical stress to cell nuclei in a way that mimics the mechanical strains placed on muscle cells where EDMD manifests itself. Cell micropatterning techniques have increasingly been employed as in vitro tools to investigate intracellular mechanotransduction processes and there are many approaches available to modulate cellular shapes and study cellular responses to specific mechanical clues, including microcontact printing, microfluidic patterning, UV-based deep etching and micro-stencils 160-164 . Previous studies of micropatterned cells have shown that there are reorganization of many nuclear binding partners of emerin, such as lamin A/C, when mechanical stress is placed on the cells 165,166 . For instance, TAN lines composed of LINC complex proteins form at the nuclear envelope and overlap with perinuclear actin cables that compress the nucleus 167 . Because emerin interacts with LINC complex proteins, one can expect that its dynamics and spatial organization are also altered by nuclear mechanical stress, and that absence of emerin or emerin mutations might result in a perturbed ability of nuclei to properly adapt to mechanical force applied using micropatterns. In non-muscle cells that do not experience significant mechanical stress, changes in the organization 29 of wildtype emerin or EDMD-inducing emerin mutants might not be obvious and might be altogether missed because their nuclei undergo minimal strains. As I show in this work, imposing increasing nuclear deformations to non-muscle cells by micropatterning them on narrow adhesion substrates, reveals the impact of nuclear stress on the organization of wildtype emerin. This approach also exposes dystrophic responses of the cell nucleus and changes in the nanoscale spatial distribution of emerin mutants. Specifically, we designed a novel cell micropatterning strategy based on fibronectin microcontact printing onto hydrophobic organosilane monolayers that is optimized for nanometer accuracy single molecule tracking and super-resolution microscopy. Plating cells on increasingly narrow rectangular fibronectin islands having widths smaller than the dimension of cells provides a simple means to impose steady-state nuclear stress and study how emerin responds to gradually increasing mechanical forces on the nucleus. In addition, the parallelization provided by micropatterning allows these forces to be applied homogeneously and simultaneously over hundreds of cells. A key benefit of this microcontact printing strategy on organosilane monolayers is that, compared to previous microprinting approaches, it is fully compatible with the stringent optical demands of super resolution imaging. Specifically, it allows high localization precision of emitter positions by insuring high photon counts, minimal background interference and good optical coupling of microscope objectives with cells through the nanometer-thin organosilane monolayer. In comparison, more traditional and thicker microstamping substrates, such as polydimethylsiloxane (PDMS), interfere with efficient laser excitation of cell samples and introduce additional background that can deteriorate localization precisions. Altogether, by combining super-resolution microscopy techniques and cell micropatterning on thin microstamped substrates, the research presented in this dissertation sheds light on the 30 mechanotranslation abilities of emerin at the nanoscale and on the perturbations in diffusion dynamics and spatial organizations that are induced by introduction of emerin mutations or mechanical stress. More generally, this work generates new means to investigate the role, the activities, and the molecular functions of mechanosensing proteins, with unprecedented details. 31 CHAPTER 2: ENGINEERING MICROPATTERNED SUBSTRATES TO EVALUATE THE MECHANOTRANSDUCING FUNCTION OF EMERIN 2.1: INTRODUCTION Muscular dystrophy is a disease that has been known for decades, yet its treatment and understanding has remained limited. In many manifestations of the disease, the causative gene is expressed exclusively in muscle cells. In other varieties of the disease, the gene is expressed ubiquitously, yet the phenotype is limited to muscle cells. Emerin has been implicated in the transduction of mechanical force and the adaptation of cells to mechanical stress 86,168 , but few studies have been conducted to examine how emerin itself adapts in response to these mechanical signals. As a mechanotransducing protein, cells without emerin exhibit altered nuclear stiffness 72 and grossly distorted nuclear shapes 56 . Interestingly, it has been shown that nuclei have the ability to stiffen and relax as needed to adapt to mechanical force 72 . Physiologically, this flexibility of the nucleus is important for cellular functions such as nuclear positioning in burgeoning muscle cells during development 169 and migration of fibroblasts during wound healing 82 . Intriguingly dystrophic muscle cells show improper nuclear organization in sarcomeres 90 , and this could be linked with the weakness of muscle contraction consistent with EMD mutations. Emerin, as a crucial regulator of the LINC complex 170 , is responsible for correctly adapting the nucleus to perturbations in mechanical equilibrium 16 . In our study of emerin mechanotransducing functions at the nuclear envelope, we wanted to examine how emerin and its mutants behaves in cells experiencing mechanical stress similar to what muscle cells would experience in vivo. Because emerin mutations affect the behavior of muscle cells 45 more than any other cells, their effects is presumably due to the unique mechanical stresses that are normally placed on muscle 32 cells 171 , where nuclei can be significantly deformed during stretching and contractions 172 . We therefore studied how emerin behaves in non-muscle cells that are forced to deform their nucleus in response to specific mechanical cues. Previous works have shown that, in cells exposed to exogenous stress, emerin traffic from the INE to the ONE 173 and that cells lacking emerin lose the ability to stretch chromatin and display increase transcription of genes normally upregulated during stretching 174 . Yet, it is unknown how the nanoscale organization of emerin is impacted by mechanical challenges. Finding the proper conditions to study the mechanotransducing function of emerin has proved difficult. While muscle cells can be cultured and studied in vitro, the mechanical stress that they experience in vivo are not easy to mimic in the laboratory. Differences in emerin distribution have been found in cultured cells with causative mutations compared to wildtype cells, yet they are generally small. We inferred that significant alterations in the behavior of emerin might better observed if we apply mechanical stress. To accomplish this, micropatterned cell substrates needed to be engineered to satisfy the stringent optical demands of super resolution microscopy. A promising technique to introduce stress on cultured cells is lithography, which can be used to control the adhesion of cultured cells and force them to adapt to boundary cues 175 . Most frequently this approach requires the use of the polymer polydimethylsiloxane (PDMS) spin-coated onto coverslips and on which cell adhesion patterns are microprinted to control cell spreading 176 . While this method is suitable for confocal and other ensemble microscopy techniques, the thickness of PDMS can interfere with imaging modalities such as total internal reflection fluorescence (TIRF) and highly inclined and laminated optical sheet (HILO) microscopies, which are often employed for high contrast single molecule imaging. In addition to possible variation in PDMS thickness 177 , PDMS spin coating is time consuming because each coverslip is made one at a time. Other 33 approaches to create activated cell adhesion surfaces involve PEG coating and UV etching through masks, which demand specialized and expensive equipment 178 , and microstenciling, which has a drawback of cells rapidly escaping patterns once the stencil is removed 179 . As an alternative, we developed an ultrathin silanization chemistry of glass coverslips by hexamethyldisilazane (HMDS) vapor deposition that preserves optical coupling for microscopy and can be produced in large quantity without special equipment. This approach allowed us to maintain many cells under a variety of mechanical constraints, to retain optimal conditions for high precision single molecule fluorescence imaging and to study changes in nuclear morphology, intracellular localization of emerin, and nanoscale emerin dynamics and organization. 34 2.2: ENGINEERING OF PATTERNED SUBSTRATES FOR CELL ATTACHMENT To avoid the drawbacks of traditional lithographical techniques, we devised an approach for printing cell-adhesion micropatterns that is based on vapor deposition of a hydrophobic silane monolayer, microcontact printing of an extracellular matrix protein, and the addition of an anti- Figure 2.1. Cell micropatterning on HMDS-treated and fibronectin-functionalized coverslips. a) Representation of rectangular and circular micropatterned cells and substrate thickness comparison of HMDS vapor coated to PDMS spin coated coverslips. b) Schematic for microcontact printing fibronectin on HMDS treated coverslips with Pluronic F-127.c) Water contact angles (θ) of untreated, Piranha-cleaned, and HMDS-treated coverslips with representative images. d) Fluorescently labeled fibronectin on HMDS coverslips after microcontact printing. Scale bars: 50 µm (left) and 10 µm (right). e) Differential interference contrast images of cells on microcontact printed HMDS coverslips with circular and rectangular patterns. Scale bar: 50 µm, insets 10 µm. 35 fouling agent to prevent cell spreading outside of printed areas (Fig. 2.1a, b). We used the reagent hexadimethysilazane (HMDS) to create a hydrophobic self-assembled monolayer (SAM) on the surface of glass coverslips suitable for super resolution imaging (Fig. 2.1a, b) 180 . Surface hydrophobicity is important for the latter application of Pluronic F-127 (PF-127), an anti-fouling agent that prevents cell attachment outside of the stamped areas. HMDS which forms a single siloxane bond with the silanol groups on glass coverslips, was deposited as a monolayer by vapor coating to make the coverslips hydrophobic 181 and to provide a balanced glass surface wettability for microstamping of fibronectin and for blocking non-stamped areas with antifouling agents. The efficiency of the HDMS surface coating process was assessed by water contact angle measurements. Coverslips were first treated with a piranha solution to activate hydrophilic -OH groups on the glass surface, which renders the glass very hydrophilic and results in a drop of the water contact angles from ~65° to below 10°. Vapor silanization with HMDS on piranha-activate coverslips renders the surface hydrophobic as the trimethysilyl groups form single siloxane bonds on the silanol groups of the glass surface with subsequent release of NH3 gas. A kinetic study of HMDS deposition indicates that a complete glass surface coating is rapidly achieved within 30 min of reaction when contact angles saturate at 87° ± 1° (Fig. 2.1c). These contact angle values are similar to those previously observed for monolayer depositions of HMDS on substrates 182 , confirming complete and homogenous monolayer deposition. Cell attachment areas were then defined by first designing PDMS stamps with 52 µm diameter circular or 210 x 10 µm rectangular micropatterns (Fig. 2.2) and inking fibronectin on those stamps. Stamps were dried and rinsed quickly with ethanol before being pressed onto the HMDS coverslips. After a few minutes, the stamps were removed, leaving an imprint of fibronectin on the surface suitable for cell attachment. Non-patterned areas were then blocked with PF-127, a nontoxic and amphiphilic triblock 36 copolymer composed of a hydrophobic poly(propylene glycol) domain that interacts with HMDS and two poly(ethyleneglycol) domains that confer antifouling properties and prevent cell adhesion. Proper deposition of fibronectin was confirmed using Cy3B-succinimidyl ester fluorescent labeling directly on glass, since the only amines present on the coverslip are those of the stamped fibronectin. This staining revealed a successful deposition of fibronectin at the surface where it formed micropatterns specific of the designed PDMS stamps used for transfer (Fig. 2.1d), indicating that the stamping process is robust and that fibronectin remained firmly and specifically attached on the coverslips after washing off excess PF-127. To verify the proper functionalization of our surfaces and the antifouling properties of PF-127, U2OS cells were allowed to attach onto the coverslip surface and we assessed whether their adhesion was restricted to the micropatterned fibronectin areas. After a few hours, cells attached and spread out only in the areas where the PDMS stamp was applied. Within 6 hours we saw no signs of escaping as cells spread out and conformed to the adhesion area provided (Fig. 2.1e), confirming that our micropatterning method is robust and effective at restricting cell adhesion. 2.3 PERFORMING LIGHT MICROSCOPY ON PATTERNED COVERSLIPS Figure 2.2. PDMS stamps for fibronectin microstamping. a) Bright field microscopy image of a 52 µm circular PDMS stamp. Scale bar: 100 μm. b) Bright field microscopy image of a 210x10 µm rectangular PDMS stamp. Scale bar: 100 μm. 37 Having shown that cultured cells indeed adapt and mechanically rearrange to fit into defined fibronectin micropatterns on our HMDS-treated coverslips, we next evaluated the effect of micropatterning on cytoskeletal arrangements and verified that cells could be imaged with high contrast imaging techniques such TIRF and HILO microscopy. Cells were chemically fixed and further stained with DAPI and phalloidin-iFluor 488 before imaging by fluorescence confocal Figure 2.3. Fluorescence confocal and TIRF microscopy of cells micropatterned on HMDS coverslips. (a) Confocal microscopy images of fixed cells grown on circular and rectangular micropatterns and stained for actin and the nucleus. Scale bars: 50 μm (top left) and 10 μm. (b) TIRF microscopy images at the plasma membrane of fixed cells stained for actin and caveolin-1. Scale bars: 8 μm. (c) Confocal microscopy images and selective photoactivation of PA-TagRFP–emerin at the nuclear envelope of live cells. The delineated region of interest (ROI) (white) was photoactivated by confocal scanning with a 405 nm laser. Scale bars: 50 and 10 μm (zoom). (d) Confocal images and photoconversion of actin in live micropatterned cells transiently transfected with LifeAct−tandem dimer EOS (LifeAct–tdEos). The + and – signs correspond to a transfected and a nontransfected cell, respectively. Green to red photoconversion of LifeAct–tdEos was done over the entire field of view using a 405 nm laser excitation. Scale bar: 20 μm. 38 microscopy to determine changes in nuclear shape and actin organization in both circular and rectangular micropatterns. Cells occupying circular patterns displayed “ring-like” distributions of F-actin bundles and a rounded nucleus, whereas rectangular-shaped cells had thick apical actin stress fibers that arched over and projected on either side of the nucleus deformed along the major cell axis, as previously reported 183,184 (Fig. 2.3a). To confirm the suitability of our treated coverslips for TIRF illumination, we imaged caveolin-1, which localizes to the plasma membrane. Using the same cell fixation protocol and additional staining for caveolin-1 to localize PM structures 185,186 , cells were then imaged by TIRF illumination. On our HMDS treated coverslips, high contrast TIRF imaging of cortical actin and caveolae at the PM could easily be achieved (Fig. 2.3b). As expected, DAPI nuclear staining was not visualized by TIRF because the nucleus is positioned too far above the basal PM, at the coverslip/cell interface. In contrast, imaging cells micropatterned on PDMS-coated coverslips using TIRF microscopy was not possible because the polymer layer was too thick to allow evanescent wave excitation of the cell basal PM. Indeed, measurements performed to assess the thickness of a PDMS layer spin coated on coverslips indicated that the PDMS layer extends 5-7 µm above a coverslip surface, far beyond the evanescent field afforded by TIRF illumination, and in good agreement with the expected thickness of spin coated PDMS 187 (Fig. 2.4). This confirms that PDMS is unsuitable for the optical demands of single molecule microscopy and that our vapor deposition of HMDS as a monolayer will not interfere with high contrast fluorescence imaging by introducing a thick substrate. 39 We next ensured that photoswitching and photoactivation of fluorescent probes often used for single molecule microscopy was efficiently achieved on our engineered surfaces. Using live U2OS cells transfected with a pA-TagRFP-emerin fusion and plated on rectangular patterns, we showed that emerin can be specifically photoactivated upon cells exposure to activating 405 nm laser excitation in specific region of interests (Fig. 2.3c). As expected, photoactivated pA-TagRFP- emerin localizes at the nuclear periphery and only the activated cells show fluorescence (Fig.2.3c). A similar photoconversion efficiency and specificity was obtained for fluorescently labeled actin in live cells transiently transfected with cDNA coding for the actin binding peptide LifeAct fused to tdEOS (LifeAct-tdEOS, Fig. 2.3d). Together these results demonstrate that our HMDS-coated coverslips are suitable for high contrast imaging of cellular structures with probes typically used with single molecule microscopy. Importantly, the micropatterns provide robust and stable cell confinement and a simple means to induce mechanical responses of cells, as observed by their modified actin organization and there change in shape. Figure 2.4. Thickness of PDMS after spin-coating on glass coverslips. The thickness of PDMS was determined over five different coverslips by confocal imaging and z-scanning of fluorescent Tetraspeck beads spread on the spin-coated coverslip after local etching of the PDMS layer with sulfuric acid to exposed the coverslip glass surface. The measured thickness corresponds to the z-traveling distances between focused beads on the coverslip glass surface and focused beads at the surface of the non-etched PDMS layer. The average thickness was estimated at about 5.5 μm, in good agreement with the expected film thickness for a 1 min spin-coating at 6000 rpm of a 100% PDMS solution. 40 2.3: NUCLEAR ADAPTATION TO EXOGENOUS MECHANICAL CUES AND IMPACT OF EMERIN EXPRESSION In addition to change in cytoskeletal organization, rectangular micropatterns also induce a deformation of the nucleus from a circular shape to an oblong shape (Fig. 2.3a, c). This underlines a response to the mechanical stress imposed by the reorganization of actin filaments as cells adapt and fit into narrow adhesion areas. We therefore tested if this nuclear deformation could be further modulated on rectangular micropatterns having different widths. To do so, we used immortalized human skin fibroblasts derived from normal patients (EMD +/y ) or from patients affected by EDMD, where emerin expression is lost (EMD -/y ). Using confocal microscopy imaging and calculation of the nuclear shape index (NSI), we compared the circularity of the nucleus in these cells after plating on increasingly narrow rectangular micropatterns having widths of 15, 10 or 5 µm (Fig. 2.5). For normal EMD +/y cells grown randomly on fibronectin-coated coverslips, nuclei are almost circular with a NSI value close to 1, but they become increasingly elliptical and display lower NSI values as they absorb increased mechanical stress in micropatterns (Fig. 2.5). Interestingly, Figure 2.5. Emerin deficient nuclei improperly adapt nuclear shape to exogenous mechanical cues. (a) Confocal imaging of Emd+/y skin fibroblasts rescued with WT emerin and grown in different width micropatterns (15, 10 or 5 µm) to controllably deform the nucleus (blue: nucleus, red: actin). Scale bar: 50 µm. (b) Changes in NSI as a function of micropattern width for normal Emd+/y, emerin null Emd-/y, and Emd-/y cells rescued with WT emerin-fusion (WT SNAP-emerin). Cell nuclei become more elliptical as the width of the cell decreases. In the absence of emerin, nuclei are consistently less circular, revealing an improper adaptation to nuclear stress. Rescue of emerin expression with a SNAP-emerin fusion restores normal nuclear shapes. n=50-60 cell nuclei per condition. T-test comparison to Emd+/y cells: ** P<0.01, NS: non-significant. 41 emerin-null EMD -/y cells are systematically more deformed than their wildtype counterparts under the same conditions (Fig. 2.5a). This apparent modified nuclear envelope stiffness in the absence of emerin expression is consistent with previous results in mouse fibroblasts 56 . Importantly, normal nucleus responses to mechanical stress are rescued when wildtype emerin is expressed in EMD -/y Figure 2.6. Fluorescence confocal imaging of emerin distribution in wild type Emd +/y human skin fibroblasts as a function of nuclear strains. The distribution of the entire cellular pool of emerin (ER, ONE, and INE) is imaged in cells treated with Triton X-100 as a detergent to permeabilize both the plasma and the nuclear membrane (left panels). The pool of emerin only associated with the ER and the ONE is imaged in cells treated with saponin to permeabilize only the plasma membrane (right panels). Cells grown randomly on fibronectin (non-patterned) or on increasingly narrow rectangular micropatterns (15–5 µm) are immunostained for emerin (green), lamin A/C (red) and the nucleus (blue). Scale bar for all images: 20 µm. 42 cells, with NSI values and nuclear circularity statistically indistinguishable from wildtype cells (Fig. 2.5b). Curiously, we also noticed a large-scale redistribution of emerin in micropatterned cells. For instance, while emerin is primarily localized at the nuclear envelope of non-patterned cells, it becomes significantly enriched in the ER for micropatterned cells (Fig. 2.6). In order to quantify this phenomena, we used variable membrane permeabilization with Triton X-100 or saponin in order to selectively stain for entire cellular pool of emerin (ER, ONE, and INE) using Triton X- 100 or the pool of emerin only associated with the ER and the ONE, since saponin permeabilizes only the plasma membrane and not nuclear membranes 188 . As shown in Fig. 2.7, emerin is trafficked from the nuclear envelope into the ER membrane as cells deform their nucleus and adapt to narrow adhesion patterns. To ensure that this apparent redistribution of emerin was not an Figure 2.7. Quantification of emerin redistribution in response to nuclear mechanical strains. (A) ratio of ER emerin to nuclear envelope (ONE and INE) emerin for non-patterned Emd +/y skin fibroblasts (Triton X-100, n = 16 cells) or fibroblasts grown on 15 µm-wide (Triton X-100, n = 15 cells), 10 µm-wide (Triton X-100, n = 15 cells), and 5 µm-wide (Triton X-100, n = 20 cells) fibronectin rectangular micropatterns. The fluorescence intensity quantification is done on sum slices images of full cell confocal z-scans obtained after Triton X-100 permeabilization and emerin immunostaining. For each cell, the intensity ratio is normalized to the mean ER/nuclear envelope ratio of non-patterned cells; (B) relative amount of ER and outer nuclear envelope emerin for non-patterned Emd +/y skin fibroblasts (saponin, n = 17 cells) or fibroblasts grown on 15 µm-wide (saponin, n = 14 cells), 10 µm-wide (saponin, n = 18 cells), and 5 µm-wide (saponin, n = 24 cells) fibronectin rectangular micropatterns. The fluorescence intensity quantification is done on sum slices images of full cell confocal z-scans obtained after saponin permeabilization and emerin immunostaining. For each cell, the quantified emerin intensity is normalized to the mean emerin intensity for non-patterned cells. For both (A) and (B), the thick bars and the error bars represent the mean and the standard deviation of each distribution, respectively. T-tests: ** p < 0.01, ns: non-significant. 43 artifact of the altered cellular shape and organelle repositioning, we quantified both the fluorescent intensity in the ER versus that in the nuclear membrane for non-patterned and patterned cells and we also looked at the relative amount of ER emerin in patterned and non-patterned cells. With both metrics we found that significantly more emerin localizes in the ER relative to non- patterned cells. For instance, the ratio of ER to total nuclear envelope emerin increases by about 25% as the nucleus adapts to strains (p < 0.01, Fig. 2.7a), indicating that adjustment of nuclear shapes against mechanical strains involves an initial relocation of emerin to the ER membrane, independent of the overall emerin expression level. Similarly, the amount of emerin associated with the ER and the ONE grows by 50% in micropatterned fibroblasts (p < 0.01, Fig. 2.7b). While there was, however, no statistically significant differences between the different pattern widths we tested, these results confirm that a significant amount of emerin is translocated from the nuclear envelope to the ER membrane as cells adapt to nuclear stress. As such, a key mechanotransducing function of emerin is to partially relocate from the INE to the ONE and the ER membrane to guarantee adequate nuclear deformation when a cell nucleus is exposed to mechanical challenges. Taken together, these results show that nuclear shape is dependent on the environment of cells, and that nuclear envelope components, including emerin, are critical for adapting the nucleus to changing mechanical environments. Next, we investigated how emerin re-organizes at the nanoscale to ensure proper nuclear adaptation to forces. 2.4: EVALUATION OF SINGLE MOLECULE MICROSCOPY ON PATTERNED COVERSLIPS 44 We next determined if our HMDS-coated micropatterning substrates were suitable for single molecule microscopy and we used HILO illumination 107 to achieve high contrast imaging of a Figure 2.8. Single molecule tracking of emerin at the nuclear envelope of micropatterned cells. (a) Localization (top) and diffusion trajectories (bottom) of individual PA-TagRFP–emerins at the bottom nuclear membrane of live cells randomly grown on HMDS-treated (left) or PDMS-coated coverslips (right). Scale bars: 5 μm. (b) Localization precision of emerin in cells grown on HMDS-treated or PDMS-coated coverslips. (c) Detection efficiency of individual PA-TagRFP–emerins in cells imaged on HMDS- or PDMS-coated coverslips. The central squares and bars represent the mean of the distribution and its median, respectively. The box length represents the interquartile range, and the error bars are the standard deviation of the mean (**: T-test, P < 0.01). (d) Diffusion trajectories of emerin at the bottom nuclear membrane of a deformed nucleus for a cell grown on a 210 × 10 μm2 micropattern. Scale bars: 10 and 2 μm (zoom). (e) Diffusion coefficient analysis by probability distribution of the squared displacement (PDSD) for the slow (left) and fast (right) diffusive behaviors of emerin at the bottom nuclear membrane of nonpatterned (black) and rectangular-patterned cells (red). Error bars represent the standard error of each mean at each time lag. (f) Diffusion trajectories of emerin at the top nuclear membrane of a deformed nucleus for a cell grown on a 210 × 10 μm2 micropattern. Some trajectories have been highlighted in red to show their distribution along thin linear structures reminiscent of apical actin fibers. Scale bar: 5 μm. 45 photoactivatable protein fusion to emerin. Live cells expressing a pA-TagRFP-emerin were imaged by spt-PALM on HMDS-functionalized coverslips or on PDMS spin-coated coverslips, for comparison. Individual pA-TagRFP-emerin were localized and their diffusion at the nuclear envelope was tracked using identical laser photactivitation at 405 nm and laser excitation at 561 nm for both types of coverslips (Fig. 2.8a). As expected, the thicker PDMS-coated coverslips required a steeper HILO angle to image emerin the nuclear envelope, high above the coverslip surface. Although there was no significant difference in the localization precision of individual photoactivated emerins between both types of coverslips (Fig. 2.8b), the number of single molecule detections per unit time was significantly lower on PDMS surfaces than on HMDS surfaces (Fig. 2.8c). This lower detection efficiency on PDMS-coated coverslips can be attributed to the increased thickness of the laminated optical sheet and to the wavelength-dependent spatial mismatch between the 405 nm activating and the 561 nm imaging HILO beams at steep excitation angles, which resulted in a reduced photoactivation efficacy of PA-TagRFP-emerin at the imaging focal plane. Comparatively, on HMDS-treated surfaces, better optical coupling between the HILO beams and cells permitted a 3-fold higher density mapping of single emerin diffusion trajectories (Fig. 2.8a). This reflects the fact that proper laser alignment over multiple colors is impeded by the thickness of the deposited PDMS layer. 2.5 EFFECTS OF NUCLEAR DEFORMATION ON THE DIFFUSION OF EMERIN 46 Using our sptPALM single molecule tracking data of emerin, we then studied how the nanoscale diffusion of emerin changes as the cell nucleus deforms and adapts to steady state mechanical forces when cells are micropatterned in 210 x 10 µm rectangular adhesion areas. Before analyzing sptPALM data, we first studied the diffusion of emerin by ensemble FRAP measurements of PA-TagRFP-emerin in non-patterned cells grown freely on fibronectin (Fig. 2.9). At the bottom of the nuclear membrane, two population of fast (D fast = 0.85 x 10 -1 ± 0.1 x 10 -1 µm 2 /s, 19%) and slow (Dslow = 3.3 x 10 -3 ± 0.1 x 10 -3 µm 2 /s, 81%) diffusing emerin was observed. These diffusion coefficients are in good agreement with previously reported diffusion values of emerin and the presence of a nearly immobile fraction of this protein at the nuclear envelope 40,99 . Figure 2.9. FRAP of emerin at the bottom nuclear membrane. a) Sequence of confocal images of PA- TagRFP-emerin before and after photobleaching a 7 μm diameter circular region of interest at the bottom nuclear membrane of one cell (top right). Scale bar: 10 µm. b) Averaged fluorescence recovery curve for 16 cells. The curve was well described by a two-component lateral diffusion model (red fit) with apparent diffusion coefficients (D) D fast = 0.85x10 -1 ± 0.1x10 -1 µm 2 /s (19%) and D slow = 3.3x10 -3 ± 0.1x10 -3 µm 2 /s (81%). Grey bars represent the standard deviation of the mean at each time points. 47 For the analysis of sptPALM data, individual pA-TagRFP-emerin were localized by two- dimensional Gaussian fitting of their PSF and by linking the localized position of each molecule from frame to frame to build trajectories (Fig. 2.8d). Diffusion coefficients were determined using a two-parameter fit of the probability distribution of the squared displacements (PDSD) 135 to account for the fast and slow diffusions observed by FRAP. This analysis revealed that, for non- micropatterned cells, most emerin (78%) diffused slowly at the bottom nuclear envelope (D slow = 3.65 x 10 -3 ± 0.14 x 10 -3 µm 2 /s), and a smaller population (22%) diffused more rapidly (D fast = 1.19 x 10 -1 ± 0.02 x 10 -1 µm 2 /s), which is in good agreement with the ensemble FRAP data. This confirms the validity of using our HMDS coverslips to measure single molecule emerin diffusion. We next wanted to see how force applied onto the nucleus via the rectangular micropatterns would affect emerin diffusion. We found that both the fast and slow populations had about 2-fold slower diffusion (D fast = 0.73 × 10–1 ± 0.02 × 10–1 μm2/s, 15%; and Dslow = 1.32 × 10–3 ± 0.1 × 10–3 μm2/s, 85%) compared to non-patterned cells (Fig. 2.9e). Interestingly, at the top of the nucleus of these micropatterned cells, many emerin trajectories also distributed along thin linear structures reminiscent of the apical actin fibers projecting above elongated nuclei (Fig. 2.9, f). These structures resemble those of TAN lines, which are composed of LINC complex proteins and are thought to help organize nuclear movement 189 . Both the slower diffusion of emerin and its apparent partial reorganization along potential TAN lines are consistent with the expected mechanosensing role of emerin at the nuclear envelope 56,72,86 , and its proposed ability to reinforce interactions between actin, the LINC complex, and the nuclear lamin meshwork to maintain the stiffness of the nucleus when it is mechanically stressed or compressed 16 . Indeed, direct binding of emerin to nuclear lamin reduces its diffusion at the nuclear membrane 190 and compressive forces from apical actin fibers on deformed nuclei induce the alignment of lamin 48 and LINC complex proteins along actin cables 191 . The observed slower membrane diffusion of emerin and its distribution along actin-like fibers imply an increased interaction of emerin with the nuclear lamin meshwork and the LINC complex 38 in deformed nuclei compared to normal nuclei. Importantly, these data also demonstrate that micropatterning of cells on HMDS-treated coverslips effectively induces nuclear mechanical stress and dynamic responses from mechanosensing proteins that can easily be probed by ultrahigh precision imaging with single molecule sensitivity. 2.6 CHANGES IN THE NANOSCALE ORGANIZATION OF EMERIN AT THE ENVELOPE OF DEFORMED NUCLEI We then assessed the suitability of our micropatterning chemistry for super-resolution microscopy imaging. We used HILO 3D-dSTORM super-resolution imaging to study the effects of nuclear deformation on the nanoscale spatial distribution of emerin at the bottom nuclear membrane of cells grown randomly on fibronectin-coated HMDS surface or cell grown on our micropatterns. Figure 2.10. Localization accuracy in 3D-dSTORM images of emerin at the nuclear membrane. Localization accuracy was determined from individual emerin molecules for cells grown on HMDS-coated coverslips and functionalized with fibronectin. Distribution histograms of localizations for individual emerin that could be imaged over at least 3 continuous frames were built as previously described4 using a bin size of 5 nm. Gaussian fit of the histograms give standard deviations of 10.3 nm, 15.1 nm, and 18.9 nm in the x, y and z directions for cells imaged by HILO on HMDS treated glass. 49 U2OS cells were transfected with a SNAP-emerin fusion, chemically fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and stained with BG-Alexa 647 to visualize emerin distribution at the nuclear envelope. Using a standard super resolution buffer composed of glucose oxidase, catalase and thiols, individual Alexa-647-labeled SNAP-emerin were detected with a 488 nm activation laser and a 647 nm excitation laser. Individual emerin were localized by Gaussian fitting of their PSF. Using an astigmatic lens, we additionally estimated the z-position of each emerin after generating z- calibration curves with a piezo stage and fiducial markers. Experimental localization precisions in in each dimension (x, y and z) were determined from the standard deviation of localization coordinates for individual emerins repeatedly re-appearing in consecutive frames, using the algorithm of Huang et al 192 . Under HILO excitation, dSTORM localization precisions of emerin were 10.3 nm in x, 15.1 nm in y, and 18.9 nm in z (Fig. 2.10). These values are well within the range expected for dSTORM super-resolution imaging with organic fluorophores. We then quantitatively compared the organization of emerin at the nuclear envelope of non- patterned cells or of mechanically stressed cells constrained to 210 x 10 µm rectangular adhesion micropatterns on HMDS-coated coverslips (Fig. 2.11a, b). To do so we performed spatial pattern NDF analyses to characterize the distribution of emerin. This approach allows us to quantitatively measure the extent of clustering at different distances while minimizing the confounding effects of clustering at smaller distances inherent to other approaches such as Ripley’s K analyses 193 . Fitting the NDF curves with an exponential decay function and convolving for our experimental localization error, we determine the relative clustering densities and the average cluster lengths of emerin, similar to the approach of Sengupta et al. 104 (Fig. 2.11c). 50 Although the nuclear membrane distribution of emerin appeared homogenous in diffraction- limited confocal and HILO images (Fig. 2.11a, b), super-resolution 3D-dSTORM revealed that emerin is actually extensively clustered at the nuclear envelope of both patterned and non- patterned cells (Fig. 2.11a, b). In non-patterned cells with a normal nuclear shape, emerin was 4- fold more clustered than expected for a completely random distribution of the protein at the nuclear membrane, and the typical size of emerin clusters was 104 ± 6 nm (Fig. 2.11c). At the membrane of deformed nuclei in the rectangular micropatterned cells, the size of emerin clusters was Figure 2.11. Super-resolution imaging and cluster analyses of nuclear membrane emerin in cells. (a) HILO imaging (left half) and 3D-dSTORM reconstruction (right half) of SNAP–emerin at the bottom nuclear membrane of a nonpatterned cell. Emerin clusters from a region of interest (red) are shown in more detail on the right. Scale bars: 2 μm (left) and 1 μm (right). (b) HILO imaging (left half) and 3D-dSTORM reconstruction (right half) of emerin at the bottom nuclear membrane of a cell grown on a 210 × 10 μm2 fibronectin micropattern. Emerin clusters from a region of interest (red) are shown in more detail on the right. Scale bars: 3 μm (left) and 1 μm. (c) Cluster analysis of emerin in normal nuclei for nonpatterned cells. The neighborhood density is best fit by a single exponential (red), revealing the clustering of emerin into nanodomains 104 ± 6 nm in size and at a density 4-fold above that expected for a random distribution (inset). (d) Cluster analysis of emerin in deformed nuclei for cells grown on 210 × 10 μm2 rectangular micropatterns. The size of emerin clusters is similar to that of nonpatterned cells (110 ± 4 nm), but the clustering density increases 11-fold above that expected for a random nuclear membrane distribution (inset). 51 unchanged (110 ± 4 nm), but emerin was 3-fold more clustered in these nanodomains than in normal nuclei, with an apparent clustering 11-fold higher than expected for a random distribution (Fig. 2.11d). This indicates that mechanical stress imposed at the nuclear envelope by the rectangular micropatterns result in the spatial redistribution of emerin into denser nanoclusters and that its mechanosensing functions are coupled to its clustering state at the nuclear membrane. Such an increased clustering of emerin is consistent with its slower diffusion observed on deformed nuclear envelops and with its expected enhanced binding to LINC complex proteins and to the nuclear lamin network in mechanically stressed nuclei 16 . Interestingly, super-resolution images also indicate that significantly more emerin is detected at the ER membrane surrounding the nucleus in micropatterned cells compared to non-patterned cells (Fig. 2.11a, b). This result is similar to that observed for endogenous emerin in micropatterned EMD +/y human skin fibroblasts (Fig. 2.6 and 2.7). As we discussed previously, this partial relocation of emerin to the ER membrane and its increased clustering appear to be important steps for the mechanotransducing functions of emerin and the appropriate deformation of the cell nucleus when it is exposed to mechanical challenges. Considering the strong interactions of emerin with the nuclear lamina and its importance in modulating nuclear adaptations to forces, it is tempting to assign these changes in emerin distribution to an altered organization of the nucleoskeleton at the nuclear envelope. Indeed, emerin self-association can be modulated by binding lamin A/C 25 and it is possible that the increased clustering of emerin we observed in deformed nuclei stems from changes in emerin/lamin A/C interactions. Effectively, binding of emerin to lamin A/C and nuclear actin in the nucleoskeleton, but also to BAF, directly impact the diffusion of emerin and its ability to oligomerize at the INE, as we discuss in more detail in Chapter 3. 52 2.7 DISCUSSION AND CONCLUSION Previous studies into the adaptations of cells to mechanical forces have relied on thick PDMS substrates. While those are useful to study large scale changes in cellular architectures, they are not well suited for dSTORM and PALM imaging and, concurrently, for studying the nanoscale adaptations that proteins undergo in response to mechanical cues. Our cell patterning approach, however, uses a simple self-assembled monolayer of HMDS and microprinted fibronectin patterns on activated glass coverslips to study how cells adapted to steady state mechanical cues. Importantly, this approach is fully compatible with the stringent requirements of high precision optical microscopy as it provide good optical coupling with micropatterned cells and optimal detection efficiencies for single molecule tracking and super-resolution microscopy. On our micropatterns surface, cells show similar cytoskeletal and nuclear rearrangements as those found on more classical patterned substrates and can still be transfected to express fusion tags and fluorescent probes suitable for single molecule microscopy imaging. The micropatterned substrates are also amenable to TIRF and HiLo illuminations, which are crucial for achieving high contrast imaging and good localization precisions for sptPALM and 3D-dSTORM. The robust and stable adhesion of extracellular matrix proteins and antifouling agents on HMDS coverslips allows the culture of individual cells with, potentially, any micropatterning geometries in order to induce variety of different cellular stress. This property is critical for studying mechanosensitive proteins like emerin, which play a key role in adapting the cell nucleus to mechanical constraints. While the organization and dynamics of emerin had previously been analyzed at the ensemble level using diffraction-limited imaging 40 , no previous studies had looked at its spatial organization 53 and its diffusion at the nanoscale and how those are altered by the introduction of mechanical stress. Lost in these previous studies was how emerin adapts its organization and its behavior to transmit changes in mechanical environment to cytoskeletal and nucleoskeletal components. Here we showed that emerin is clustered in nanodomains with sizes of ~100 nm and displays distinct diffusive behaviors at the nuclear envelope. Importantly our use of rectangular micropatterns having different width allowed us to examine the relation of these emerin clusters and diffusional states to mechanical stress applied on the nucleus. Indeed, using 10 µm wide patterned substrates, we found that emerin diffusion and clustering state are significantly altered by the introduction of mechanical stress at the nuclear envelope. Specifically, the diffusion of emerin slows significantly in mechanically constrained nuclei, while emerin clustering density in nanodomains increases, in a process reminiscent of the mechano-induced clustering of integrins for transmembrane force transmission at the cell surface 194,195 . The enhanced clustering of emerin might reflect a need to maintain the stiffness of compressed nuclei via reinforced emerin/lamin/LINC complex/actin linkages. As we show in Chapter 3, further studies of emerin nanoscale clustering as a function of nuclear deformation and of emerin mutants that induce laminopathies but localize correctly at the nuclear membrane could yield new insights into the normal mechanosensing functions of emerin and into the pathogenesis of EDMD. In conclusion, our micropatterning and imaging techniques open up new prospects to quantitatively study the molecular mechanobiology of emerin with single molecule sensitivity and nanometer precision. We have shown, for the first time, that the behavior and the organization of emerin itself is altered upon introduction of mechanical stress into cells. Next we will investigate the molecular basis of EDMD and how mutated emerins that cause the disease display modified 54 nanoscale distribution and diffusion compared to wild-type emerin, in particular when mechanical stress is imposed at the nuclear envelope. 2.8 MATERIALS AND METHODS Plasmids, Cell Lines, and Cell Labeling For the expression of PA-TagRFP–emerin, a pEGFP-N1 plasmid backbone encoding emerin fused to the C-terminus of PA-TagRFP was produced by XbaI and KpnI insertion and polymerase chain reaction fusion of human emerin cDNA. A stable monoclonal U2OS cell line constitutively expressing PA-TagRFP–emerin was obtained after transfection with XtremeGene HP (Roche), selection with 100 μM of Geneticin (G418), and clonal isolation by serial dilution. For confocal imaging, live U2OS cells expressing PA-TagRFP–emerin and grown in micropatterns were further stained with 1:1000 Hoechst 33342 (Thermo) in Hank’s balanced salt solution (HBSS) buffer (Corning), rinsed three times, and imaged in 37 °C HBSS. For single molecule tracking of emerin by sptPALM, no additional nuclear staining was performed. For the expression of SNAP–emerin, emerin was fused to the C-terminus of a SNAP tag by inserting human emerin cDNA in a pSNAP-tag(m) plasmid (NEB) via AscI and XhoI restriction sites. U2OS cells grown in six-well plates were transiently transfected with this plasmid using XtremeGene HP, trypsinized 24 h after transfection, and seeded on micropatterned coverslips. Forty-eight hours post transfection, cells were fixed with 4% paraformaldehyde in PBS for 15 min, permeabilized with 0.1% Triton X-100 for 15 min, and blocked with 4% bovine serum albumin (BSA) + 0.1% Tween-20 for 30 min, all at RT. Cells were then stained with 1:1000 SNAP 55 AlexaFluor 647 (BG-A647, NEB) in 4% BSA + 0.1% Tween-20 for 1 h at 37 °C, then thoroughly washed before super-resolution imaging. For the expression of Cav1, a multiclonal cell line stably expressing Cav1–SNAP was generated after transfection of a pEGFP-N1 plasmid backbone encoding the sequence for canine Cav1 fused to the N-terminus of a SNAP tag in 3T3 MEF KO cells originating from Cav1 knockout mouse (ATCC CRL-2753). Cell imaging was performed after cell fixation and Cav1–SNAP staining with BG-A647 as previously described. For actin imaging with the photoconvertible tandem dimer Eos fluorescent protein (tdEos), U2OS cells grown in micropatterns were transiently transfected with a pDendra2-N plasmid backbone encoding the LifeAct peptide fused to the N-terminus of tdEos (LifeAct–tdEos). For actin and Cav1 immunostaining, U2OS cells were fixed as previously described. Actin was stained with 1:1000 CytoPainter phalloidin-iFluor 488 (Abcam), and Cav1 was detected with 1:1000 anti-Cav1 rabbit antibody (N-20, Santa Cruz Biotechnology) followed by a 1:1000 staining with Alexa-647 goat antirabbit antibody (Life Technologies). Coverslips were mounted using Fluormount G with DAPI (Hatfield, PA) to visualize the nuclei. Imaging Confocal images were taken using an Olympus Fluoview microscope, 60X objective, and at RT. FRAP experiments were conducted by activating all pA-TagRFP-Emerin fusions with 100% 405 nm laser power and by bleaching a 7 µm diameter area at the bottom of the nucleus with 100% 488 and 561 nm lasers. Four frames were collected prior to bleaching, and bleaching was 56 performed for 10 s. Acquisition was performed using a 561 nm laser and 100% 405 nm laser throughout imaging. Additional confocal microscopy images were acquired on a Zeiss LSM 700 confocal laser scanning microscope equipped with a C-Apochromat 40×/1.2 W Korr objective, excitation lasers at 405 nm, 488 nm, 555 nm, and 639 nm, a multiband 405/488/555/639 beam splitter and appropriate emission filters for the detection of DAPI, Alexa Fluor 488, Cy3B, Rhodamine and Alexa Fluor 647. Images were acquired in 12-bit mode and the same settings were used across all samples. Confocal z-stacks were collected over the entire thickness of each cell in 0.34-μm slice intervals. Wide field microscopy images were acquired on a Nikon Eclipse Ti-E microscope equipped with a 40× objective (Nikon), an X-Cite 120XL fluorescence illumination system, an iXon Ultra EMCCD camera (Andor) and appropriate filter sets for DAPI (Ex: 430DF24, Dich.:458DiO2,Em: 483DF32, Semrock) and A647 (Ex: 628DF40, Dich.:660DiO2,Em: 692DF40, Semrock) detections. HMDS Coverslip Preparation High precision microscope glass coverslips (Marienfeld, #1.5, Ø25 mm) were cleaned using a Piranha solution made of a 3:1 (v/v) mixture of 18 M sulfuric acid and 30% hydrogen peroxide for 15 min and rinsed thoroughly with deionized (DI) water. Following drying with nitrogen gas, the coverslips were heated to 95 °C in a sealed glass jar containing 100 μL of silane reagents for incremental reaction times of 3 to 90 min (silanization kinetics) or for a fixed reaction time of 90 min (cell micropatterning). The silane reagents used were: hexamethyldisilazane (HMDS, Sigma- 57 Aldrich, St. Louis, MO, USA), (3-Aminopropyl)triethoxysilane (APTES, Sigma-Aldrich,) and (3- Glycidyloxypropyl)trimethoxysilane (GPTMS, Sigma-Aldrich). After vapor coating, the silane- coated coverslips were cleaned by sonication in water for 5 min and dried with N2 gas. The silane- treated coverslips were then stored in a separate and sealed glass container flushed with nitrogen. Static contact angle measurements were done on a ramé-hart 290-F1 Contact Angle Goniometer (ramé-hart instrument co., Succasunna, NJ, USA) with 5 μL water droplet volumes tested over 5– 8 different positions per coverslip on at least 6 coverslips per reaction points. Contact angle measurements Sessile drop goniometry measurements of the HMDS, PDMS, and Piranha cleaned coverslips were determined using a Tantec Contact Angle Meter. Measurements were made by placing a small drop of water on the coverslip and measuring the contact angle between the drop and the surface. 10 measurements were made per coverslip, with 5 coverslips used per condition. PDMS stamp preparation Polydimethylsiloxane (PDMS) stamps with rectangular micropatterns were produced from silicon masters fabricated using a chrome photomask (Minnesota Nano Center, Minneapolis, MN, USA). The micropatterns are 210 μm in length and have widths of 5 μm, 10 μm or 15 μm, respectively, with a constant periodic intervals of 30 μm. The etching depth on these silicon masters is 10 μm. The silicon masters were treated with tridecafluoro-1,1,2,2-tetrahydrooctyl-1-trichlorosilane (Gelest Inc., Morrisville, PA, USA) for 90 min under vacuum to induce the formation of 58 fluorosilane vapors and surface fluorosilanization of the masters. A 10:1 mixture of PDMS and curing agent (Sylgard 184 elastomer kit, Dow Corning, Midland, MI, USA) was combined in a plastic beaker and thoroughly mixed with a glass stirring rod for 10 min. As this step generates bubbles, the PDMS mixture is degassed under vacuum for 30 min (or by centrifugation at 3000 rpm for 10 min). The degassed PDMS mixture was then slowly poured onto the silicon masters, and another degassing step was done to avoid unwanted bubbles. After curing for 3 h at 65 °C and overnight at room temperature (RT), the PDMS stamps were removed slowly from the silicon master using a razor blade. Before being used in microcontact printing, the PDMS stamps were cleaned by 5 min sonication in water followed by 5 min sonication in ethanol. We routinely use the PDMS stamps for a period of 3–4 months without loss of stamping efficiency. Fibronectin Patterning For microcontact printing, 100 μL of 100 mg/mL fibronectin in phosphate-buffered saline (PBS, 154 mM NaCl, 5.6 mM Na2HPO4, 1.1 mM KH2PO4, pH 7.5) were placed on the PDMS stamps and incubated at 25 °C for 30 min. The fibronectin solution was then aspirated off and the PDMS surface was rinsed twice with PBS. Upon drying, the inked PDMS surface was brought into contact with the silane-functionalized glass coverslips for 1–2 min, applying light pressure to guarantee good contact between the stamp and the glass. To block non-patterned areas, the coverslips were immersed in 1% (m/v) Pluronic F-127 (PF-127, Sigma-Aldrich) in Milli-Q water (EMD Millipore, Billerica, MA, USA) for 20 min, then rinsed three times with PBS prior to cell seeding. To evaluate the quality of the stamping before and after PF-127 treatment, the fibronectin-stamped HMDS-, GPTMS- and APTES-coated coverslips were treated with 31 nM Cy3B-maleimide (GE 59 Healthcare Life Sciences, Marlborough, MA, USA) in PBS for 1 h and rinsed three times with PBS, before imaging on a LSM 700 Confocal Laser Scanning Microscope (Zeiss, White Plains, NY, USA). To assess the antifouling efficacy of PF-127, 100 µL of 10 mg/mL (150 µM) of bovine serum albumin (BSA, Alfa Aesar-Thermo Fisher Scientific Chemicals, Ward Hill, MA, USA) was fluorescently labeled with 10 µL of 10 mg/mL (8 mM) of Alexa-Fluor 647 Succinimidyl Ester (Life Technologies, Carlsbad, CA, USA) for 1 h in PBS. The reaction was quenched with 150 µL of 20 mM Tris buffer (pH 7.8) for 1 h and the mixture was diluted to a final concentration of 2.5 mg/mL (37 µM) of A647-BSA with PBS. A647-BSA at 2.5 mg/mL was applied for 1 h on fibronectin-stamped coverslip previously labeled with Cy3B-maleimide and treated with PF-127. After three rinses with PBS, the coverslips were imaged by fluorescence confocal microscopy. After stamping, the PDMS stamps were immersed in DI water and cleaned in an ultrasonic bath at 60 °C for 10 min, then immersed in 100% ethanol and sonicated for another 10 min at 60 °C before drying with nitrogen gas and storage. PDMS Spin Coating and Thickness Measurement Circular glass coverslips (Marienfeld, #1.5, Ø25 mm) were first cleaned using a PDC-32G Plasma Cleaner (Harrick Plasma, Ithaca, NY) set to medium for 5 minutes. PDMS was spin-coated onto the cleaned coverslips by mixing a 10:1 ratio of Sylgard 184 with curing agent (Dow Corning, Midland, MI) and placing a single 40 µl drop onto the coverslips. The coverslips were then spun at 6000 rpm for 1 min, after which the PDMS was cured for 2 hours at 65°C and overnight at room temperature. To determine the thickness of the deposited PDMS layer, PDMS was specifically etched in a small area in the middle of the coverslip using concentrated sulfuric acid applied with 60 a Pasteur pipet. The Pasteur pipet was placed on the PDMS layer and sulfuric acid was added to define a small and circularly etched area. After a few minutes, PDMS in the etched area detached from the coverslip surface and the acid was neutralized in the pipet by slowly adding a solution of sodium bicarbonate. After neutralization, the pipet was removed, exposing the glass surface in the etched area. Coverslips were thoroughly rinsed with water before addition of fluorescent Tetraspeck beads at the PDMS surface. Fluorescence confocal imaging was then performed by taking z-stacks (0.5 µm step sizes) of the fluorescent beads present at the exposed glass surface of the etched area and at the surface of the PDMS layer. The thickness of the PDMS layer is reported as the z-traveling distances between focused beads on the coverslip glass surface in the etched area and focused beads at the surface of the non-etched PDMS layer. Measurement of emerin enrichment at the endoplasmic reticulum For nuclear shape index (NSI) measurements, cells were fixed with 4% paraformaldehyde in PBS for 15 min, permeabilized with 0.1% Triton X-100 for 10 min, and blocked with 4% bovine serum albumin (BSA) + 0.1% Tween-20 for 1 h, all at RT. Wild type (EMD +/y ) and emerin-null fibroblasts (EMD -/y ) were then stained with Rhodamine Phalloidin (1:1000, Abcam Inc., Cambridge, MA, USA) for 1 h, and washed three times with PBS for 5 min each. Coverslips were mounted on a glass slide using DAPI-Fluoromount G (Electron Microscopy Sciences, Hatfield, PA, USA) and sealed with clear nailpolish. U2OS cells and emerin-null human dermal fibroblasts (EMD -/y ) expressing SNAP-emerin were fixed, permeabilized, and blocked as mentioned. Before mounting in DAPI-Fluoromount G, cells were then additionally stained with 1:1000 of SNAP- Surface® Alexa Fluor 647 benzylguanine substrate (BG-A647; New England Biolabs, Ipswich, MA, USA) in 4% BSA + 0.1% Tween-20 for 1 h at 37 °C and washed three times with PBS for 5 61 min each. Microscopy images were acquired by wide-field on an inverted Eclipse Ti-E microscope (Nikon Instruments Inc., Melville, NY, USA) or by confocal imaging on a Zeiss LSM 700 Confocal Laser Scanning Microscope. For experiments involving emerin redistribution as a function of nuclear mechanical strains on micropatterns, wild type human dermal fibroblasts (EMD +/y ) were fixed with 4% paraformaldehyde in PBS for 15 min, permeabilized with either 0.1% Triton X-100 (cell membrane and nuclear permeabilization, Sigma-Aldrich) or 0.1 % saponin (cell membrane permeabilization only, EMD Millipore) for 10 min, and blocked with 2% BSA + 1% normal goat serum (NGS) for 1 h, all at RT. Cells were stained with rabbit anti-emerin (1 μg/mL, Abcam Inc., Cambridge, MA, USA) and mouse anti-Lamin A/C (1:1000, Santa Cruz Biotechnology, Dallas, TX, USA) primary antibodies for 1 h at RT, then rinsed 3x with blocking buffer for 5 min. Staining with a goat anti rabbit-Alexa Fluor 488 (1:500, Life Technologies) and a goat anti mouse-Alexa Fluor 647 (1:1000, ThermoFisher, Life Technologies) secondary antibodies was then done for 1 h. Following three washing steps of 5 min in blocking buffer and three additional washing steps of 5 min in PBS, the coverslips were mounted and sealed as stated before. Images and z-scans through each cell were acquired on a Zeiss LSM 700 Confocal Laser Scanning Microscope. Fluorescence recovery after photobleaching (FRAP) Fluorescence recovery after photobleaching (FRAP) was performed in 37ºC HBSS buffer (Corning) on an Olympus Fluoview FV1000 confocal microscope equipped with a 60x/1.40 NA oil immersion objective and with U2OS cells stably expressing PA-TagRFP-emerin and grown to 70% confluency on fibronectin coated coverslips. PA-TagRFP-emerin at the bottom nuclear 62 membrane was briefly photoactivated by laser scanning fields of view with a 405 nm laser line and PA-TagRFP fluorescence was monitored every 2.8 seconds for 200 frames with 10% of a 543 nm HeNe laser line. After imaging four frames, circular regions of interest, 7 µm in diameter, were photobleached for 10 s by laser scanning with 100% of the 543 nm laser line. Typically, the ROI were bleached to ~20 % of the original intensity. After background subtraction, fluorescence recovery curves were doubly normalized as described by Phair et al. 196 to correct for loss of fluorescence due to bleaching during acquisition. Fluorescence recovery curves from measurements on multiple cells were averaged after normalization to full scale. The apparent diffusion coefficients for emerin were obtained by curve fitting the averaged recovery curve with a two-component lateral diffusion model using equations for a uniform circular bleach region described by Soumpasis et al. 197 Single molecule tracking sptPALM Live cells expressing mutational variants of pA-TagRFP-Emerin were mounted on 1.5H coverslips with HBSS prewarmed to 37°C. Single particle tracking was performed on a Nikon Eclipse Ti microscope with a 100X 1.49 NA TIRF objective (Nikon) and a 600 bp 50 emission filter (Chroma). Cells were illuminated with 40% 561 nm laser and activated with variable 0.1%-5% 405 nm laser filtered through a 405/488/561/647x excitation filter and 405/488/561/647rpc dichroic mirror (Chroma) and using HiLo illumination 107 . Movies were taken for no longer than 3 minutes per cell to limit UV damage of live cells. 63 Localization and tracking analyses were performed using the software package SLIMfast, which uses multiple-target tracing algorithms 198 and was kindly provided by Christian Ritcher and Jacob Piehler. Localizations were performed by 2D-gaussian fitting of the PSF of each activated pA- TagRFP in each frame. Localization precision was determined by a formula considering the number of detected photons. 113 Trajectories were built by linking localizations frame to frame and accounting for blinking statistics and local particle densities. Trajectories with fewer than three steps were discarded. Diffusion coefficients were estimated using a probability density of square displacement (PDSD) analysis 135 . For each time lag t, the Pr 2 curve was fitted with the following model: 𝑃 ( 𝑟 ⃗ 2 , 𝑡 ) = 1 − ∑ 𝑎 𝑖 ( 𝑡 ) 𝑒 −𝑟 2 /𝑟 𝑖 2 ( 𝑡 ) 𝑛 𝑖 =1 ∑ 𝑎 𝑖 ( 𝑡 ) = 1 𝑛 𝑖 =1 Where ri 2 (t) is the square displacement and ai(t) is the population density of i numbers of diffusive behaviors at each time lag t. The PDSD distributions of pA-TagRFP-Emerin were fit with 4 populations. Error bars for square displacement were determined using 𝑟 𝑖 2 √ 𝑁 , where N for each ri, or alternatively 𝑁 𝑎 𝑖 . Diffusion coefficients were determined by fitting ri 2 (t) curves with Origin software (OriginLab) and a free Brownian diffusion model with position error 𝑟 2 = 4𝐷𝑡 + 4𝜎 2 64 All reported diffusion coefficients D are reported in µm 2 /s ± standard deviation of fit value. Super-reolution imaging by dSTORM Prior to dSTORM, cells were fixed in 4% PFA for 15 min at RT, followed by permeabilization in 0.1% Triton X-100 for 15 min at RT, then blocking in 4% BSA + 0.1% Tween-20 for 30 min at RT. Cells were stained using a 1:1000 dilution of SNAP-Surface Alexa Fluor 647 (NEB) at 37° C for 1 hr. Cells were thoroughly washed with the blocking buffer and mounted in standard STORM buffer. An astigmatic lens was used for 3D positioning, and a 700 bp 75 emission filter (Chroma) was used. Cells were patterned as described previously 199 . Briefly, HMDS activated glass coverslips were stamped with fibronectin coated PDMS of 5, 10, and 15 µm widths respectively. Attachment outside of patterned areas was blocked with a 1% solution of Pluronic F-127. Cells were then seeded onto coverslips, where cell attachment and spreading was restricted to stamped areas. After 6 hours, cells were removed for either sptPALM or fixed for dSTORM. Movies were analyzed using RapidSTORM 200 to determine positions of emitters. Localizations were filtered using PALMsiever 201 to correct for overcounting of an emitter on for multiple frames by discarding additional emissions within 8 nm on in consecutive frames. Drift correction was also performed using the fiducial drift correction feature on the program. 65 Spatial Pattern Analysis Neighborhood Density Function analysis was performed as described previously 158,202,203 . Briefly, the function determines clustering values similar to O-ring statistics by measuring the density of localizations within a ring of outer radius r and width Δr for all r inside of a region of interest. The average clustering density around a localization at a distance r was determined with: 𝐷 𝑟 = ∑ 𝑁 𝑟 ∑ 𝐴 𝑟 Where Nr is the number of localizations and Ar is the area summed over all proteins. This was determined over a 1000 nm distance with a step size of 10 nm. In order to average NDF statistics over multiple ROI’s and multiple cells, Dr was standardized by dividing by the mean density of detected proteins across the entire ROI. Therefore, the NDF value represents the average cluster density above random and should return a value of 1 if the density is the same as the average density. To determine confidence intervals, Monte Carlo simulations of random distributions of localizations were performed with the same number of events and area of every ROI. Since the probability density of proteins in 2D clusters decays as an exponential function 104 , the NDF curves across all ROI’s were fit using a two-population model that corrects for position error 150 𝑟𝑒𝑙𝑎𝑡𝑖𝑣𝑒 𝑁 𝐷 𝐹 = {𝐴 1 exp( −𝑟 𝜀 1 ) + 𝐴 2 exp( −𝑟 𝜀 2 ) + 1} ∗ 𝑔 ( 𝑟 ) 𝑃𝑆𝐹 Where ε is the average cluster length at which cluster density is half maximal. A is the cluster density and * denotes a two-dimensional convolution. g(r) PSF is the correlation function for a PSF 66 of uncertainty in position for a dSTORM experiment. This term corrects for the effect that position error has on NDF. g(r) PSF was defined as 𝑔 ( 𝑟 ) 𝑃𝑆𝐹 = 1 4𝜋 𝜎 2 exp( −𝑟 2 4𝜎 2 ) Where σ is the uncertainty in position. Curve fitting was performed in Matlab. One and two population models were fitted to the data, and the two-population model chosen when a significant gain in r 2 was achieved and when it was visually confirmed that the data were better fitted. Cluster density values were reported as A + 1, since the data decay to 1 and to report how many times over random the density in the clusters was found to be. The extracted ε value was expressed as the cluster radius, but more accurately reflects the length at which most clusters have decayed to half of their peak value. 67 CHAPTER 3: EMERIN OLIGOMERIZATION AND NUCLEOSKELETAL COUPLING AT THE NUCLEAR ENVELOPE REGULATE NUCLEAR MECHANICS AGAINST STRESS 3.1 INTRODUCTION Emerin is an inner nuclear membrane (INM) protein that organizes and maintains nuclear architecture by interacting with the nucleoskeletal lamina and elements of the linker of the nucleoskeleton and cytoskeleton (LINC) complex 30,204 . Emerin also has a role in organizing genetic material via its role in tethering chromatin at the nuclear envelope by binding the DNA- bridging barrier-to-autointegration factor (BAF) 205 , and it regulates the activity of chromatin compaction modulators 206 . Both the nucleoskeletal organization and the compaction state of chromatin contribute to the nuclear response to mechanical force 207 and emerin is a critical player in orchestrating mechanotransducing processes at the nuclear envelope 208 . The importance of this protein is highlighted by the fact that when this gene is mutated or absent, X-linked Emery- Dreifuss muscular dystrophy (EDMD) manifests 43,209 , a disease part of a larger of group of laminopathies 210 associated with structural perturbations of the nuclear envelope and its underlying lamina. While emerin is expressed in all somatic cells of the body, mutations in the EMD gene seem to primarily affect cells exposed to extensive mechanical stress, such as skeletal and cardiac muscle cells. Cultured muscle tissues lacking emerin display deformed and disorganized nuclei, and in the body is observed impaired myogenesis and improper muscle fiber formation, which contribute to the muscle wasting and cardiac disease phenotypes of EDMD 211-213 . Loss of emerin has been shown to lead to altered nuclear envelope elasticity and increased nuclear fragility 214 , impaired expression of mechanosensitive genes 56 and enhanced apoptosis after continuous 68 mechanical strain 56 . EDMD symptoms might be caused by both an altered structural integrity of the nuclear envelope and modified gene expression profiles. The structure of emerin and its binding to various partners have been extensively characterized in vitro 30 . Emerin is 254 amino acids (aa) long, with a N-terminal globular LAP2-emerin-MAN1 (LEM, aa: 1-45) domain, followed by a long intrinsically disordered region (IDR, aa: 46-222) 215 and a transmembrane domain close to the C-terminus (aa: 223-234) that anchors the protein to the nuclear envelope. Among the binding partners of emerin are lamin A/C and lamin B 25,46,216 , actin 53 , BAF 217 , nesprins, SUN proteins from the LINC complex 218,219 , and chromatin modification enzymes 30 . The LEM domain is responsible for binding of emerin to BAF, while interactions with other binding partners and other emerin proteins 25,28 are mediated by the IDR. Emerin can oligomerize via regions of the IDR that serve as LEM binding and self-assembly sites between emerin monomers 25 , and, in vitro, emerin self-association into filamentous structures 100 impact binding to lamin A/C and BAF 25 . Post-translational modifications in the IDR and the LEM domain also influence the oligomerization of emerin and its interactions with binding partners 30,220,221 . Phosphorylation of emerin residues Y74 and Y95 are required for lamin A/C recruitment to the LINC complex and nuclear envelope stiffening in response to mechanical stress 222 . Additional post-translational modifications regulate the affinity of the LEM domain to BAF 221 and binding to actin 223 . How emerin organizes at the nuclear envelope and participates in protecting the nucleus against mechanical strains remains unclear. EDMD is often caused by frame-shift deletions and nonsense mutations of EMD that eliminate any expression in cells 224 . Yet, a few sets of point mutations and small deletions, including Δ95- 69 99, Q133H and P183H/T, also cause EDMD despite an apparently correct localization of these mutated emerins at the nuclear envelope and normal cellular expression levels 53,217,225,226 . These emerin mutants, however, display altered self-association and binding to nucleoskeletal proteins in vitro 28,30 and possible changes in molecular organizations at the nuclear envelope might explain why they still result in full manifestation of EDMD. Here, we used super-resolution imaging by stochastic optical reconstruction microscopy (STORM) and single particle tracking by photoactivated localization microscopy (sptPALM) to investigate the nanoscale organization of wild-type emerin and emerin mutants at the nuclear envelope in cells. By altering the mechanical environment of nuclei using cell micropatterning techniques 227 , we also studied how emerin participates in nuclear envelope mechanotransduction processes during stress. We show that emerin oligomerization and its differential interactions with key structural partners at the nuclear envelope is central to nuclear shape adaptation against mechanical challenges. 3.2 EMERIN DISPLAYS DISTINCT DIFFUSIVE BEHAVIORS AT THE NUCLEAR ENVELOPE To study the nuclear envelope dynamics of wild-type and dystrophy causing mutations of emerin, we fused PA-TagRFP to emerin and expressed this protein in emerin-null human dermal fibroblasts (HDF) from an EDMD patient (EMD -/y ) 228 . As expected, PA-TagRFP-emerin localized to the nuclear envelope, as observed for endogenous emerin in HDF from a healthy individual (EMD +/y ) (Fig. 3.1a). For sptPALM experiments, PA-TagRFP-emerin was activated at 405 nm 70 and excited at 561 nm using highly inclined and laminated optical sheet (HILO) 107 illumination to Figure 3.1. Emerin displays multiple diffusive behaviors at the nuclear envelope. a) Immunostaining and confocal imaging of emerin in emerin-null HDF (EMD -/y ), HDF with endogenous emerin expression (EMD +/y ) and EMD -/y HDF after expression of wild-type PA-TagRFP-emerin. Scale: 10 µm. b) Emerin trajectories in the nucleus built from localization and tracking of individual PA-TagRFP-emerin at the nuclear envelope. Scale: 5 µm. c) Probability distribution of square displacement (PDSD) for wild-type emerin (71004 trajectories in 14 cells) at time lag Δt = 160 ms. The distribution is composed of four distinct diffusive behaviors. d) Square displacement curves (r 2 i) for each emerin diffusive behavior generated from PDSD analyses on the first ten time lags Δt 40-400 ms. Curves were fitted over the first four values with a Brownian diffusion model (red line) to estimate the respective diffusion coefficients of each emerin behaviors. e) Emerin trajectory maps segregated by estimated diffusion coefficient ranges (D i). Scale: 5 µm. 71 photoactivate single emerin molecules and track their diffusion at the nuclear envelope. Diffusion tracks were generated from emerin appearances in successive frames (Fig. 3.1b) and diffusion analyses were performed using the probability distribution of square displacement (PDSD) 229 to characterize distinct diffusive behaviors (Fig. 3.1c). From the PDSD analysis of tens of thousands PA-TagRFP-emerin trajectories, we observed that emerin displays four distinct diffusive behaviors with diffusion values D 1: 2.21x10 -1 ± 4.9x10 -2 µm 2 s -1 , similar to previously reported for diffusion at the ER membrane 40 , D2: 1.48x10 -2 ± 1.5x10 - 3 µm 2 s -1 , D3: 1.73x10 -3 ± 1.1x10 -4 µm 2 s -1 , and D4: 2.6x10 -4 ± 1x10 -5 µm 2 s -1 (Fig. 3.1d and Table 3.1). The putative ER population D1 makes up 1% of the detected behaviors at the nucleus, while the next slowest population D2 represents 9%, a small fraction of emerin similar to that expected at the outer nuclear membrane (ONM) 204,208 . The two slowest populations D3 and D4 collectively Table 3.1: Diffusion coefficients of wild-type and mutated emerin determined by PDSD analyses after sptPALM or CALM imaging. 72 represent 90% of emerin diffusive behaviors. We next examined the individual trajectories of emerin using MSD analysis to determine if these diffusive behaviors were associated with distinct localizations of emerin in the cell. Based on their estimated diffusion coefficients, we grouped trajectories into four diffusion ranges spanning the D values measured from PDSD analysis and plotted them as maps (Fig. 3.1e). In these generated maps, the fastest emerin population D1 is primarily distributed in the immediate proximity of nuclear envelope (Fig. 3.1e), a location consistent with diffusion in the ER membrane. The three slower populations D 2, D3 and D4 display significant enrichment at the nuclear membrane (Fig. 3.1e). While diffusion analysis by fluorescence recovery after photobleaching could only detect only two diffusive behaviors of emerin (Fig. 3.2), our single molecule measurements revealed more complex dynamics of emerin at the nuclear membrane. 73 3.3 EMERIN ORGANIZES AS SLOWLY DIFFUSING MONOMERS OR OLIGOMERS AT THE INNER NUCLEAR MEMBRANE While emerin primarily localizes at the INM, a small fraction is detected at the ONM 204,208 , where it cannot interact with the nuclear lamina because the protein there extends to the cytoplasm. We therefore tested if interactions with lamin A/C would affect only some of diffusing populations of wild-type emerin. We found that downregulation of lamin A/C expression by siRNA (Fig. 3.3 and 3.4) increases the mobility of only the two slowest populations D3 and D4 (p<0.01, Fig. 2a and Table 3.1), but does not affect the faster populations D1 and D2 (p not significant, Fig. 3.5a). This shows that D3 (40%) and D4 (50%) are two distinct emerin populations diffusing at the INM and interacting with lamin A/C, while population D2 (9%) corresponds to emerin diffusing at the ONM. Figure 3.2. FRAP of emerin in U2OS cells. a) U2OS cells transfected with pA-TagRFP-emerin fusions after photoactivation and photobleaching of a circular region of interest at the nuclear envelope. Scale bar: 5 µm. b) FRAP recovery curves of wild-type (WT), P183H, Q133H, and Δ95-99 mutated emerin at the nuclear envelope. Grey bars represent the standard deviation of the mean at each time points. n= 9-12 cells per condition. c) Diffusion coefficients of wild-type and mutated emerin determined by fitting FRAP recovery curves with a two-components fit. 74 A similar increase in the mobility of emerin populations D 3 and D4 is also observed when nuclear actin is depleted after downregulating the expression of the actin nuclear import factor, importin- 9 79 (Fig. 3.3), although decreased nuclear actin levels additionally induce slightly faster emerin diffusion at the ONM (Fig. 3.5a and Table 3.1). These results indicate that a large majority of emerin is found in two distinct INM pools, both of which interact with the nucleoskeleton. Figure 3.3. Immunoblotting and quantifications of siRNA knock down against lamin A/C, IPO9, BAF and SUN1. a) Western blots of lamin A/C, nuclear actin, BAF, SUN1 and histone H2A in the nuclear fraction (left) and of actin and GAPDH in the cytoplasmic fraction (right) of EMD+/y human dermal fibroblasts after RNA interference. b) Relative change in lamin A/C, nuclear actin, BAF, SUN1 and cytoplasmic actin expression levels compared to untreated cells after siRNA (t-test, *: p<0.05). Error bars represent the standard deviation of the mean. 75 Emerin has also been found to self-assemble in vitro 28,100 , and its oligomerization at the INM could produce the slow mobility we observed, particularly if emerin oligomers are tethered by the nucleoskeleton. To test if diffusing populations D3 or D4 correspond to emerin monomers or oligomers, we tracked individual emerins exclusively in their dimeric and oligomeric forms. To do this, we co-expressed emerin fusions to complementary split-GFP fragments 230 (sGFP-emerin and M3-emerin, Fig. 3.5b) to irreversibly induce the formation of complemented emerin-GFP- Figure 3.4. Immunostaining to assess the direct and indirect effects of RNAi on the nuclear localization of lamin A/C, SUN1 and BAF. a) Confocal fluorescence imaging of lamin A/C, SUN1 and the nucleus (DAPI) after siRNA-induced depletion of lamin A/C, SUN1 and IPO9. All scale bars: 20 µm. b) Confocal fluorescence imaging of endogenous BAF, the nucleus (DAPI) and SUN1 after siRNA-induced depletion of lamin A/C, SUN1, IPO9 or BAF. All scale bars: 20 µm. c) Confocal fluorescence imaging of GFP-BAF L58R, lamin A/C and the nucleus (DAPI) after siRNA-induced depletion of endogenous BAF. Scale bar: 20 µm. 76 emerin species and studied the diffusion of this complemented sGFP-fusion by complementation activated light microscopy 231 (CALM). We found that emerin-GFP-emerin species localize almost exclusively at the nuclear envelope (Fig. 3.5c), where they diffuse as two separate populations: a dominant 90% population with a mobility similar to D 4 (p not significant, Fig. 2d), and a minor 10% population with a mobility similar to the ONM population D2 (p not significant, Fig. 3.5d). Figure 3.5. Emerin diffuses as inner nuclear membrane monomers and oligomers that interact with lamin A/C, nuclear actin and BAF. a) Diffusion coefficients (± s.e.m.) and population percentages of emerin (untreated, n=71004 trajectories in 14 nuclei) after depletion of lamin A/C (lamin A/C KD, n=60569 trajectories in 11 nuclei), depletion of nuclear actin (IPO9 KD, n=74501 trajectories in 17 nuclei) or replacement of endogenous BAF with BAF L58R mutant (BAF KD + BAF L68R, n=62714 trajectories in 8 nuclei) (F-test, ns: non-significant, **: p<0.01). b) Diagram of emerin fused to complementary split-GFP fragments used to track the mobility of emerin oligomers by single molecule imaging of complemented GFP. c) Map of individual trajectories for complemented emerin-GFP-emerin species at the nuclear envelope (left) and examples of oligomer domains (squares) where trajectories often overlap (left). Scales: 2 µm (left) and 50 nm (right). d) Comparison of diffusion coefficients (± s.e.m.) and population percentages for individual wild-type emerin assessed by sptPALM and complemented emerin-GFP-emerin species (n=4833 trajectories in 13 nuclei) assessed by CALM (F-test, ns: non- significant). e) Diffusion map of wild-type PA-TagRFP-emerin at the nuclear envelope where emerin oligomers form slow mobility domains (blue, O) surrounded by larger areas where emerin monomers diffuse faster (red, M). Scale: 500 nm. 77 No trajectories with diffusion coefficients matching those of the ER population D 1 or the INM population D3 were detected. Multiple emerin-GFP-emerin trajectories were often found to spatially overlap, indicating that emerin forms multimers rather than strictly dimers at the INM (Fig. 2c). These results show that population D3 detected by sptPALM but not by CALM represents emerin monomers, while the slowest population D 4 represents some multimeric state of emerin which could be strictly dimers or, more likely, oligomers at the INM, as we subsequently confirmed by super-resolution imaging. The nearly 10-fold difference in translational diffusion coefficient between populations D3 and D4 is close to the 10-40 fold reduction in mobility reported when monomeric membrane proteins transition to oligomers 232 , further evidence that the D4 population cannot represent a strictly dimer organizational state. The detection of a few emerin- GFP-emerin species at the ONM is likely due to split-GFP-induced emerin dimers that cannot translocate through peripheral nuclear pores channels due to their size 210 and thus cannot access the INM. 3.4 BAF BINDING MODULATES THE MOBILITY OF BOTH EMERIN MONOMERS AND OLIGOMERS DISTRIBUTED ACROSS THE INNER NUCLEAR MEMBRANE BAF is known to bind the LEM domain of emerin with high affinity 25 and since it also binds to lamin A/C 2 and chromatin 233 , it likely participates in the slow mobility of emerin at the INM 234 . To study how binding of emerin monomers and oligomers to BAF might influence their respective diffusion, endogenous BAF was knockdown by siRNA (Fig. 3.3 and 3.4) and replaced by a mutant form of the protein (L58R) which cannot bind LEM domains 235 . When unable to bind BAF due to this mutation, emerin monomers diffuse significantly faster with a lateral mobility higher than for lamin A/C or IPO9 knock down (p<0.01, Fig. 3.5a and Table 3.1). Emerin oligomers also diffuse 78 faster (p<0.01, Fig. 3.5a), indicating that BAF modulates the INM dynamics of emerin when it oligomerizes. While ER associated dynamics of emerin remain unaffected by these experimental conditions, the mobility at ONM emerin increases (p<0.01, Fig. 3.5a), which implies that cytoplasmic BAF additionally interacts with emerin at the ONM. BAF is known to be highly mobile in the nucleus 234 and the comparatively slow mobility of BAF-bound emerin at the INM likely stems from the formation of ternary complexes with lamin A/C or chromatin. The stronger influence of BAF L58R on emerin dynamics at the INM compared to lamin A/C depletion indicates that BAF binding modulates the mobility of emerin not only via lamin A/C, but also via interactions with other nuclear components, potentially chromatin. We additionally built contour maps of diffusion coefficients of emerin at the nuclear envelope by spatial rendering of diffusion coefficients at the position of each tracked molecule (Fig. 3.5e). These maps are dominated by the two slowest INM emerin populations, which represent 90% of the total pool detected. They reveal that domains with very slow diffusion and attributed to emerin oligomers are interspersed throughout the nuclear envelope and are surrounded by areas where the mobility of emerin is faster and matches that of monomers (Fig. 3.5e). This indicates a membrane- wide, yet locally structured distribution of emerin at the INM, with monomers surrounding emerin oligomer domains in an organization compatible with emerin monomer/oligomer exchanges in distinct INM areas. Together, these diffusion analyses show that newly synthesized emerin diffuses rapidly at the ER membrane on their way to enrichment at the ONM, where emerin diffusion is slowed, in part by interaction with cytoplasmic BAF. Once past the nuclear pores and in the INM, the mobility of 79 emerin is further reduced, with emerin monomers and oligomers interacting with nuclear BAF and components of the nucleoskeleton. This distribution is consistent with a combined BAF and nuclear lamina-induced model for retention and accumulation of emerin at the INM. 80 3.5 EMERIN FORMS DISCRETE OLIGOMERIC NANODOMAINS SURROUNDED BY MONOMERS Figure 3.6. Emerin forms oligomeric nanodomains stabilized by lamin A/C and surrounded by larger emerin monomer areas across the nuclear membrane. a) Super-resolved 3D-dSTORM localizations (top half) and maximum intensity projection (bottom half) of wild-type SNAP-emerin at the bottom nuclear envelope of an EMD-/y HDF. Scale: 5 µm. b) Neighborhood densities (± s.d.) of wild-type emerin at the nuclear membrane of untreated HDF (n=189,331 localizations in 10 nuclei) and HDF after lamin A/C knock down (n=178,206 localizations in 6 nuclei). Neighborhood densities at various length scales are compared to MonteCarlo simulations of complete spatial randomness and fitted to measure molecular densities above random and length scale of significant clustering (green). c) Relative nuclear envelope molecular densities above random (± s.e.m.) for wild-type emerin oligomers (O) and monomers (M) in untreated EMD-/y HDF treated with control siRNA (n= 180,546 localizations in 5 nuclei), depleted for lamin A/C, depleted for nuclear actin after IPO9 knock down (n=225,394 localizations in 9 nuclei), with endogenous BAF replaced by BAF L58R (n=90,241 localizations in 6 nuclei) or depleted for SUN1 (n=258,300 localizations in 6 nuclei). Values in parenthesis represent the typical length scale (± s.e.m.) of each domain in nanometers. d) Local cluster map of wild-type emerin at the nuclear envelope of an untreated HDF, for a search radius of 25 nm (L(r25)). e) Local cluster map of wild-type emerin after lamin A/C knock down. f) Local cluster map of wild-type emerin after destabilization of the LINC complex by SUN1 knock down. Cluster values in maps are assigned to all localized emerin, and values L(r25)=25 represent areas where emerin is randomly distributed while L(r25)=70 values represent areas with emerin local densities (70/25)2 = ∼8-fold higher than expected for a random distribution. M: monomer areas, O: oligomer nanodomains. Scales for d, e and f: 250 nm. 81 We then analyzed the structural organization of emerin at the nuclear membrane in more detail using super-resolution microscopy. Wild-type emerin fused to a SNAP-tag (SNAP-emerin) was expressed in EMD -/y HDF and, after chemical fixation, cells were stained with benzylguanine- FluorAlexa647 to image individual emerins by dSTORM (Fig. 3.6a). An emerin neighborhood density function (NDF) built from localized emerin positions across multiple ROIs in multiple nuclei was compared to MonteCarlo simulations of completely random distributions and fitted with exponential decay models 236 to determine relative density and length scale of significant clustering. We found that emerin exhibits organization that is significantly non-random at the nuclear envelope, as it displays density distributions significantly higher than expected for a completely random organization (Fig. 3.6b). The generated NDF curve is best fitted with a two exponential decay model, indicating that emerin organizes into two distinct clustering states across the nuclear envelope, which evokes the two unique diffusive behaviors of emerin found via sptPALM. The first type of cluster is small with a typical cluster length of 22±11 nm and a molecular density 8.2±0.2 fold higher than that expected for a random distribution (Fig. 3.6c and Table 3.2). The second type of cluster is significantly larger with a typical cluster length of 236±30 nm and a density slightly above the expected value of 1 for complete randomness (1.3±0.1 fold, Fig. 3.6c). In contour maps of density built from a local cluster analysis algorithm 155 , emerin forms small, high density clusters that are interspersed throughout the nuclear envelope and surrounded by areas with densities above yet closer to random (Fig. 3.6d). These density contour maps resemble our sptPALM diffusion maps in live cells (Fig. 3.5e) and indicate that small, high emerin density clusters are oligomers while the surrounding larger areas are populated by emerin monomers. Unexpected the emerin monomers appear as slightly clustered instead of random, however, a fully random distribution is not possible because emerin is excluded from areas of the 82 membrane occupied by other proteins which are themselves organized, most significantly the nuclear pores. These observations confirm that, across the INM, emerin organizes into small and discrete oligomeric units surrounded by dispersed monomers. 3.6 EMERIN OLIGOMERS ARE STABILIZED BY LAMIN A/C AND SUN1 AND MODULATED BY NUCLEAR ACTIN AND BAF We then determined the importance of lamin A/C, nuclear actin, SUN1, and BAF in maintaining the nanoscale spatial organization of emerin at the INM. Emerin was imaged by super-resolution microscopy after siRNA knock down of lamin A/C, IPO9, SUN1, or replacement of endogenous BAF with BAF L58R. After lamin A/C depletion, the molecular density of emerin is dramatically Table 3.2. Molecular densities and domain sizes of wild-type and mutated emerin determined by spatial distribution analyses of emerin in super-resolved images. 83 reduced at all length scales (Fig. 3b). The NDF distribution is comprised of small, 35±17 nm emerin clusters with molecular density close to random (1.2±0.1 fold, Fig. 3.6c and Table 3.2) and of much larger 950±173 nm emerin monomer domains with density just above random (1.1±0.1 fold, Fig. 3.6c). Compared to emerin distributions in cells having normal lamin A/C levels, emerin monomers are dispersed over larger areas and, importantly, emerin oligomers nearly vanish. Such changes in organization are seen in emerin cluster maps of lamin A/C depleted nuclei, where most oligomeric nanodomains have dissipated, and where monomers cover a significantly larger part of the nuclear envelope (Fig. 3.6d). This shows that lamin A/C plays an important structural role for the stabilization of emerin oligomers and the general spatial distribution of emerin monomers at the INM. In the case when nuclear actin is depleted after knocking down of IPO9, wild-type emerin forms oligomers having a reduced molecular density of 3.6±0.1 fold above random and a larger size of 65±13 nm compared to untreated cells (Fig. 3.6c and Table 3.2). The monomers are dispersed over domains having a typical length scale of 382±81 nm, which is larger than in untreated cells (236±30 nm, Fig. 3.6c), but not as large as for lamin A/C knock down (950±173 nm, Fig. 3.6c). Despite reduced nuclear actin levels, emerin still retains the ability to oligomerize, although it disperses over large INM areas. These results show that, together with lamin A/C, nuclear actin contributes to maintaining the overall spatial distribution of emerin at the INM and additionally participates in the structural maintenance of emerin oligomers, albeit to a lesser degree than the lamina itself. When endogenous BAF is replaced by BAF L58R, emerin oligomers are still formed but they have a reduced molecular density of 2.0±0.1 fold above random and a larger size of 85±19 nm compared to cells expressing wild-type BAF (Fig. 3.6c and Table 3.2). The inability to bind BAF also 84 induces a re-distribution of emerin monomers over large INM domains with typical sizes of 718±250 nm (Fig. 3.6c). Consistent with our diffusion results, emerin binding to BAF modulates the organization of both oligomeric and monomeric emerin and strongly influences the spatial distribution of emerin monomers at the INM. When SUN1 is depleted to destabilize LINC complexes (Fig. 3.3 and 3.4), the formation of emerin oligomers is reduced to levels equivalent to those of lamin A/C depletion. Emerin distributes randomly over 75±41 nm nanodomains with molecular densities of 1.2±0.1 fold above random, an extensive decrease compared to oligomeric clusters in cells with endogenous SUN1 expression levels (Fig. 3.6c and Table 3.2). Emerin monomers also organize over nuclear envelope domains with typical sizes of 268±110 nm, similar to the spatial distribution of emerin monomers in untreated cells (Fig. 3.6c and Table 3.2). This indicates that reduced SUN1 expression levels primarily influence the formation of emerin oligomers at the INM. In density contour maps, destabilization of SUN1 LINC complexes effectively results in a largely random distribution of emerin at the INM and, like for lamin A/C depletion, few emerin clusters are visible (Fig. 3.6f). While SUN1 interacts with lamin A 38,237,238 , the expression and nuclear envelope localization of lamin A/C was not disrupted by SUN1 knockdown (Fig. 3.3 and 3.4) and, reciprocally, lamin A/C depletion did not modify the expression of SUN1 nor its distribution at the nuclear membrane (Fig. 3.3 and 3.4), as often reported 38,237,239-241 . Thus, together with lamin A/C, SUN1 is required for the proper organization of emerin oligomers. SUN1 itself was shown to self-assemble into nearly immobile oligomeric platforms 241,242 that could potentially serve as sites for macromolecular assemblies at the INM. The observation that a core LINC component, such as SUN1, is required for the self-assembly of emerin at the INM indicates that emerin locally oligomerizes at LINC complexes. 85 Together, these results show that emerin monomers are distributed throughout the INM where their dispersion is modulated by direct binding to BAF and additional spatial constrains imposed by lamin A/C and nuclear actin. Emerin also assembles into discrete and small oligomeric nanodomains that are structurally co-stabilized by lamin A/C and the LINC complex protein SUN1 and are maintained, to a lesser extent, by BAF and nuclear actin. 86 3.7 INCREASED DIFFUSION OF EMERIN MONOMERS AND OLIGOMERIZATION UPON NUCLEAR ADAPTATION TO MECHANICAL STRESS We then sought to analyze the mechanotransducing function of emerin by studying how its diffusion and its nanoscale organization change when nuclei are subjected to increasing mechanical stress. SptPALM and super-resolution imaging of wild-type emerin was performed in Figure 3.7. Mechanical stress on the nucleus increases emerin mobility and the formation of emerin oligomers. a) Schematic for increasing mechanical stress on the nucleus by cell micropatterning on rectangular fibronectin strips with decreasing widths of 15, 10 and 5 µm. Arrows represent forces. b) Fluorescence confocal imaging of actin (green) and the nucleus (blue) in micropatterned EMD +/y HDF. Scale: 50 µm. c) Nuclear shape index as a function of micropattern width for untreated EMD +/y HDF (n=54, 70, 61 and 62 nuclei for non- patterned, 15, 10 and 5 µm wide patterns, respectively), HDF depleted for lamin A/C (n=75, 60, 70 and 66 nuclei) and HDF depleted for nuclear actin (n=62, 71, 66 and 46 nuclei). Box length: index interquartile range; central square: mean; central bar: median; error bars: ± s.d. (T-test, ** p < 0.01). d) Diffusion coefficients (± s.e.m.) and population percentages of wild-type emerin in non-patterned cells or after nuclear deformation on 15 µm wide (n=27266 trajectories in 10 nuclei) and 10 µm wide micropatterns (n=12915 trajectories 8 nuclei) (F-test, ns: non-significant, **: p<0.01. e) Relative nuclear envelope molecular densities above random (± s.e.m.) for wild- type emerin oligomers (O) and monomers (M) in non-patterned EMD -/y HDF or after nuclear deformation on 15 µm wide (n=151,647 localizations in 10 nuclei) or 10 µm wide micropatterns (n=56,563 localizations in 6 nuclei). Values in parenthesis represent the typical length scale (± s.e.m.) of each domain in nanometers. f) Increase in emerin oligomers as a function of nuclear stress on 15 and 10 µm wide micropatterns compared to non-deformed nuclei in non-patterned cells. 87 EMD -/y HDF grown on increasingly narrow rectangular micropatterns with widths of 15 µm, 10 µm or 5 µm, to impose steady-state mechanical stress to the nucleus 98,208,227 (Fig. 3.7a, b). First, we verified that changes in nuclear shape index (NSI) on these micropatterns reflect the mechanical adaptation of nuclei against forces by depleting lamin A/C or nuclear actin, two key nucleoskeletal proteins involved in maintaining nuclear shape during stress. In non-patterned cells, nuclei are slightly more deformed after lamin A/C knock down or nuclear actin depletion compared to controls (p<0.01, Fig. 3.7c). These effects are further exhibited on micropatterns where nuclei become increasingly deformed as cell growth area narrows (p<0.01, Fig. 3.7c). This demonstrates that cell micropatterning incites increased nuclear stress and effectively induces mechanical responses from the nucleus that require adaptations of the nucleoskeleton. Figure 3.8. Comparison of wild-type emerin diffusion and nanoscale organization after nuclear actin depletion or cell micropatterning. a) Diffusion coefficients (± s.e.m.) and population percentage of wild-type emerin after nuclear actin depletion by IPO9 knock down (IPO9 KD, n=74501trajectories in 17 nuclei), after nuclear deformation on 15 µm wide (n n=27266 trajectories in 10 nuclei) or after nuclear deformation 10 µm wide micropatterns (n=12915 trajectories in 8 nuclei, F-test, ns: non-significant). b) Relative nuclear envelope molecular densities above random (± s.e.m.) for wild-type emerin oligomers (O) and monomers (M) in non-patterned Emd-/y fibroblasts after nuclear actin depletion (IPO9 KD, n=225,394 localizations in 9 nuclei) and in micropatterned fibroblasts with deformed nucleus on 15 µm wide (n=151,647 localizations in 10 nuclei) or 10 µm wide micropatterns (n=56,563 localizations in 6 nuclei). Values in parenthesis represent the typical length scale (± s.e.m.) of each domain in nanometers. 88 When we tracked PA-tag-RFP-emerin on nuclei in 15 µm and 10 µm patterns, we found that its ER mobility is unchanged compared to non-patterned cells (p not significant, Fig. 3.7d and Table 3.1), while its ONM diffusion is faster (p<0.01, Fig. 3.7d). At the INM, emerin monomers also diffuse significantly faster (p<0.01, Fig. 3.7d) and the mobility of oligomers increases compared to non-patterned cells, though short of statistical significance for 15 µm patterns (p not significant and p<0.01, Fig. 3.7d). Interestingly, there is no significant difference in the mobility of all four ER, ONM and INM emerin populations in both types of micropatterns compared to nuclear actin depletion in non-pattened cells (p not significant, Fig. 3.8). This suggests that nuclear shape adaptation to mechanical cues entails modified interactions of emerin with nuclear actin, notably for monomers. Super-resolution imaging and spatial distribution analyses of wild-type SNAP-emerin further indicate that as nuclei adapt to narrower areas, monomers disperse over increasingly large INM domains with sizes of 382±62 nm and 460±136 nm, compared to 236±30 nm for non-deformed nuclei (Fig. 3.7e). Concurrently, oligomer densities drop from 8.2±0.2 fold to 3.6±0.1 fold above random in 15 µm patterns, before increasing slightly to 4.6±0.1 fold above random in 10 µm patterns (Fig. 3.7e). Remarkably, oligomeric nanodomains become larger during nuclear stress and their size expands from 22±11 nm to 60±13 nm in both micropatterns (Fig. 3.7e). When considering this wider spatial distribution of oligomers, the relative oligomerization of emerin compared to non-deformed nuclei increases by 3.4 fold and by 4.1 fold as the nucleus adapts to incremental mechanical stress (Fig. 3.7f). This shows that nuclear shape adaptation to forces is associated with a gradual change in the oligomerization potential of emerin at the INM. This 89 enhanced oligomerization of emerin over larger nanodomains is accompanied by a faster lateral mobility of emerin and is likely triggered by nucleoskeletal re-arrangements. Indeed, the stress- induced spatial reorganizations of emerin in micropatterns, including oligomer densities and emerin domain sizes, is strikingly similar to those observed when nuclear actin is depleted in non- patterned cells (Fig. 3.8 and Table 3.2). This indicates that nuclear shape deformation involves a disengagement of nuclear actin from the nucleoskeleton that leads to a faster diffusion of emerin monomers and to their increased oligomerization. An increased lateral mobility of emerin at the INM could indeed facilitate molecular collisions between monomers and the formation of oligomers stabilized by lamin A/C and SUN1 at LINC complexes. These results indicate that the mechanotransducing functions of emerin are coupled to changes in its oligomeric state along the INM and are modulated by emerin interactions with nucleoskeletal partners, including nuclear actin, to ensure appropriate nuclear deformation and response to mechanical challenges. Together, these results show that the mechanotransducing functions of emerin are coupled to its oligomeric state along the INM and are modulated by emerin interactions with nucleoskeletal partners, including nuclear actin, in order to ensure appropriate nuclear deformation and response to mechanical challenges. 3.8 EMERIN MUTANTS INDUCE ABNORMAL NUCLEAR DEFORMATION AGAINST MECHANICAL STRESS To establish the biological significance of emerin oligomerization at the INM and its importance for nuclear adaptation to mechanical stress, we then compared the organization of wild-type emerin with that of mutated forms of emerin known to induce EDMD. We studied how emerin 90 mutation Q133H, deletion Δ95-99, and mutation P183H (Fig. 3.11a) impact the lateral diffusion and the nanoscale distribution of emerin at the nuclear envelope. First, we verified that mutated emerins effectively induce defective mechanical responses of nuclei when expressed in EMD -/y HDF and we compared changes in NSI after random cell plating or when grown on increasingly narrow rectangular micropatterns (Fig. 3.9b, c). In randomly plated cells expressing mutated emerins, nuclei are slightly less circular compared to cells expressing wild-type emerin (Fig. 3.9b). With increasing mechanical stress in micropatterns, cells expressing emerin mutants display significantly higher NSI values than wild-type (Fig. 3.9b), indicative of a failure to correctly modulate the shape of the nucleus in response to exogenous mechanical force. These irregular Figure 3.9. Emerin mutations induce defective nuclear shape adaptation against mechanical stress. a) Diagram of wild-type emerin (WT) showing its transmembrane (TM) and LEM domains, its disordered central region with relevant binding domains and self-association domain (self-ass.) and the position of Δ95-99, Q133H and P183H EDMD-inducing mutations. b) Nuclear shape index as a function of micropattern width for EMD -/y HDF expressing wild-type emerin (n=38, 33, 26 and 57 nuclei for non-patterned, 15, 10 and 5 µm wide patterns, respectively), Q133H emerin (n= 74, 58 and 37 nuclei), Δ95-99 emerin (n= 64, 89, 45 and 46 nuclei) and P183H emerin (n=82, 78 and 28 nuclei). Box length: index interquartile range; central square: mean; central bar: median; error bars: ± s.d. (T-test, ns: non-significant, *: p<0.05, **: p<0.01. c) Fluorescence wide-field imaging of actin (green) and the nucleus (blue) in micropatterned EMD -/y HDF expressing Q133H, Δ95-99 or P183H emerin mutants. Scales: 50 µm. 91 changes in nuclear shape are accompanied with an improper positioning of the nucleus relative to the cell major axis, nuclear crumpling, abnormal organization of the actin cytoskeleton, and failure of cells to properly fit within micropatterns, specifically in cell areas adjacent to the misshaped nucleus (Fig. 3.9c). It indicates that expression of mutated emerin in EMD -/y HDF impedes nuclear adaptation to mechanical stress. 3.9 EMERIN MUTANTS DISPLAY DEFECTIVE OLIGOMERIZATION AT THE INNER NUCLEAR MEMBRANE Figure 3.10. Diffusive behaviors of Q133H and wild-type emerin. Comparison of Q133H emerin diffusion (± s.e.m.) and population percentage in non-patterned cells with that of wild-type emerin under mechanical stress on micropatterns (F-test, ns: non-significant, **: p<0.01). 92 Using sptPALM and super-resolution imaging, we then characterized the nuclear envelope Figure 3.11. Emerin mutants exhibit modified lateral mobilities and defective oligomerization at the inner nuclear membrane. a) Diffusion coefficients (± s.e.m.) and population percentages of Q133H emerin (n=105050 trajectories in 13 nuclei) compared to wild-type emerin (n=71004 trajectories in 14 nuclei) (F-test, ns: non- significant, *: p<0.05, **: p<0.01). b) Relative nuclear envelope molecular densities above random (± s.e.m.) for Q133H emerin oligomers (O) and monomers (M) (n=149,340 localizations in 6 nuclei) compared to wild-type emerin (n=189,331 localizations in 10 nuclei). c) Local cluster map of Q133H emerin at the nuclear envelope for a search radius of 25 nm (L(r 25)). M: monomer areas, O: oligomer nanodomains. Scale: 250 nm. d) Diffusion coefficients (± s.e.m.) and population percentages of Δ95-99 emerin (n=76944 trajectories in 14 nuclei) compared to wild-type emerin (F-test, ns: non-significant, *: p<0.05, **: p<0.01. e) Relative nuclear envelope molecular densities above random (± s.e.m.) for Δ95-99 emerin oligomers (O) and monomers (M) (n=208,092 localizations in 8 nuclei) compared to wild-type emerin. f) Local cluster map of Δ95-99 emerin. M: monomer areas, O: oligomer nanodomains. Scale: 250 nm. g) Diffusion coefficients (± s.e.m.) and population percentages of P183H emerin (n=86529 trajectories in 21 nuclei) compared to wild-type emerin and compared to complemented P183H emerin- GFP-emerin species (n=10519 trajectories in 21 nuclei) assessed by CALM (F-test, ns: non-significant, *: p<0.05, **: p<0.01). h) Relative nuclear envelope molecular densities above random (± s.e.m.) for P183H emerin oligomers (O) and monomers/dimers (M/D) (n=138,075 localizations in 6 nuclei) compared to wild-type emerin. i) Local cluster map of P183H emerin. M/D: monomer/dimer areas, O: oligomer nanodomains. Scale: 250 nm. Values in parenthesis in b, e and h represent the typical length scale (± s.e.m.) of each domain in nanometers. 93 dynamics of each mutated emerin and their respective nanoscale organization, starting with Q133H emerin. We found that the lateral mobility of Q133H PA-TagRFP-emerin is similar to that of wild- type emerin at the ER and the ONM (p not significant, Fig. 3.11a and Table 3.1). However, both Q133H monomers and oligomers diffuse significantly faster at the INM (p<0.01 and p<0.05, Fig. 3.11a). The Q133H mutation disrupts emerin binding to actin 32 but does not impede interactions with lamin A/C 226 , SUN1 38 , BAF, or other partners 243 . The increased lateral diffusion of Q133H, only at the INM, therefore confirms that it does not bind nuclear actin. This faster INM diffusion of Q133H also resembles the increased mobility of wild-type emerin when nuclear actin is depleted (Fig. 3.5a), which further underlines that nucleoskeletal actin modulates the diffusion of emerin monomers and oligomers. This is consistent with previous observations that emerin expression influences the mobility of nuclear actin 244 and indicative of a reciprocal effect of emerin/nuclear actin interactions on their respective mobility. The diffusion of Q133H at the INM is also similar to that of wild-type emerin under mechanical stress, in both 15 µm and 10 µm wide micropatterns (p not significant, Fig. 3.10 and Table 3.1). This signifies that nuclear deformations against stress involve a dissociation of emerin from nuclear actin that leads to faster emerin diffusion at the INM. Although, it was reported that Q133H has a reduced capacity to self-assemble in vitro 28 , our analyses of its spatial distribution and cluster maps show that, like wild-type emerin, it organizes into monomers and oligomers across the INM (Fig. 3.11b, c). Q133H monomers are distributed over length scales of 213±62 nm, similar to those of wild-type emerin (Fig. 3.11b), indicating that the inability to bind the nuclear actin induces a faster diffusion of Q133H monomers, but does not affect their overall spatial distribution at the INM. Q133H also retains the ability to self-assemble at the INM where it forms oligomeric clusters (Fig. 3.11c). These oligomers have sizes of 19±12 nm, similar to the 22±11 nm size of wild-type emerin oligomers, but with a molecular density 94 12.2±0.2 fold above random, a 50% increase in oligomerization compared to wild-type emerin (Fig. 3.11b and Table 3.2). This indicates that the deficient binding of Q133H to nucleoskeletal actin leads to a disproportionate self-association of emerin into oligomeric nanodomains. It also implies that direct binding to nuclear actin normally reduces the oligomerization potential of wild- type emerin. 3.10 Δ95-99 MUTATION When we performed similar studies with Δ95-99 emerin, we found that its diffusion at the ER membrane and the ONM is similar to that of wild-type emerin (p not significant, Fig. 3.11d and Table 3.1), but that its lateral mobility at the INM is significantly reduced (p<0.05 and p<0.01, Fig. 3.11d). This suggests that Δ95-99 interacts more strongly or more frequently than wild-type emerin with some of its binding partners on the nucleoplasmic side of the nuclear envelope. Previous biochemical studies have shown that the small Δ95-99 deletion eliminates emerin interactions with most of its binding partners, including lamin A/C and actin, but not BAF 30 . Considering that binding BAF strongly influences the mobility of wild-type emerin (Fig. 3.5a), the reduced diffusion of Δ95-99 at the INM could stem from its increased interaction with BAF, as recently proposed 100 . To destabilize these interactions, we attempted to track Δ95-99 in EMD -/y HDF knocked down for endogenous BAF and expressing BAF L58R. Co-expression of both Δ95- 99 and BAF mutants was toxic to cells, indicating that interactions between Δ95-99 emerin and BAF are important to maintain cell viability. Beside a slower emerin mobility at the INM, the Δ95-99 deletion also induces defective oligomerization. Δ95-99 is distributed randomly over large, 420±51 nm nuclear envelope domains and in smaller, 48±14 nm nanodomains where the molecular density of 1.3±0.1 fold above random 95 is dramatically reduced compared to wild-type emerin oligomers (Fig. 3.11e and Table 3.2). In cluster maps, Δ95-99 also displays less dense and fewer oligomerization nanodomains (Fig. 3.11f). Thus, Δ95-99 does not efficiently oligomerize at the INM, consistent with its impaired self- assembly in vitro and observations that the deletion lowers emerin/emerin proximity 28 . Interestingly, despite an intact lamina, the molecular densities and cluster maps of Δ95-99 are similar to those of wild-type emerin when the expression of lamin A/C is reduced or when LINC complexes are disrupted (Fig. 3.6b, c). The decreased oligomerization of Δ95-99 to levels seen after lamin A/C depletion implies that binding to lamin A/C is required to stabilize emerin oligomers at LINC complexes at the INM. Such a stabilizing role of lamin A/C is consistent with prior proximity ligation assays where Δ95-99 was found less close to lamin A/C than wild-type emerin 28 and in vitro studies showing that the deletion abolishes lamin A/C binding to emerin 217 . Together, these results show that Δ95-99 fails at oligomerizing and primarily distributes at random throughout the INM due to its reduced self-association, its inability to directly bind lamin A/C and its slow mobility. 3.11 P183H MUTATION SptPALM tracking of P183H emerin reveals that its lateral diffusion at the ER membrane and the ONM is unchanged compared to wild-type emerin (p not significant, Fig. 3.11g Table 3.1). However, its mobility at the INM is reduced for the populations attributed to monomers and oligomers (p<0.05 and p<0.01, Fig. 3.11g). This slow diffusion is similar to that observed for Δ95- 99 and again implies that P183H interacts more frequently than wild-type emerin with some of its nucleoplasmic binding partners. In vitro, P183H maintains its binding to many emerin partners, including SUN1 38 , BAF and actin 30 , and displays enhanced binding to lamin A/C compared to 96 wild-type emerin 30,217 . The observed reduced INM diffusion of P183H might therefore be linked to an increased binding frequency to BAF, like for Δ95-99, or to its enhanced binding to lamin A/C. This slow INM mobility could also be due to the formation of dimers as recently suggested based on the strong propensity of P183H to self-assemble in vitro 28 and observations that residue P183 is positioned in the 168-186 emerin region required to limit emerin-emerin association 25 . To determine if P183H tends to form dimers at the nuclear envelope, we performed single particle tracking by CALM after co-expression of sGFP-P183H and M3-P183H emerin. Three different populations of fluorescently activated P183H-GFP-P183H emerin species were detected at the nuclear envelope: a 2% population with a mobility slightly slower than P183H at the ER membrane (p<0.05, Fig. 3.11g), a larger 39% population with a mobility comparable to the ONM behavior of P183H (p not significant, Fig. 3.11g), and a dominant 59% population with a lateral diffusion similar to the INM population initially attributed to P183H monomers by sptPALM (p not significant, Fig. 3.11g and Table 3.1). Surprisingly, no fluorescent species with diffusion coefficient matching that of the slowest P183H oligomers were detected (Fig. 3.11g). The high frequency detection of complemented split-GFP for P183H compared to wild-type emerin at both the ER membrane and the ONM (Fig. 3.5d) indicates that P183H is more prone to self-assemble and to form dimers than wild-type emerin before reaching the INM. This apparent biased monomer:dimer equilibrium towards dimers is maintained at the INM where it precludes an efficient association of P183H into oligomer domains, in particular when dimers are further stabilized by irreversible assembly of the split-GFP fragments. This shows that P183H has a propensity to form dimers that could impact oligomerization at the INM. Consistent with these observations, spatial analyses reveal that the dimerization of P183H leads to a significantly reduced oligomerization at the INM. P183H monomers/dimers are distributed in 97 domains with typical sizes of 321±29 nm and molecular densities of 1.3±0.1 fold above random (Fig. 3.11h and Table 3.2). Smaller nanodomains with sizes of 35±12 nm are also observed, but their molecular density is reduced to 2.2±0.1 compared to wild-type emerin oligomers (Fig. 3.11h). Indeed, while P183H can still bind lamin A/C and SUN1, it forms oligomeric domains having lower molecular density than wild-type emerin in contour density maps (Fig. 3.11j). This indicates that the dimerization of P183H hinders further self-association into dense oligomers at the INM. 98 3.12 ABNORMAL REORGANIZATION OF Δ95-99 EMERIN IN RESPONSE TO MECHANICAL STRESS To further characterize the importance of emerin oligomerization for its mehcanotranducing function and ultimately the nuclear responses to mechanical stress, we studied the nanoscale organization of Δ95-99 in micropatterned cells. For deformed nuclei in 15 µm patterns, Δ95-99 monomers are dispersed over large, 810±215 nm INM domains, almost double the size of monomer domains in non-stressed nuclei (420±51 nm, Fig. 3.12a). However, as abnormal nuclear Figure 3.12. Insufficient oligomerization of EDMD-inducing Δ95-99 emerin mutant against mechanical stress. a) Relative nuclear envelope molecular densities above random (± s.e.m.) for Δ95-99 emerin oligomers (O) and monomers (M) in non-patterned EMD -/y HDF or after nuclear deformation on 15 µm wide (n=138,119 localizations in 5 nuclei) or 10 µm wide micropatterns (n=135,143 localizations in 6 nuclei). Values in parenthesis represent the typical length scale (± s.e.m.) of each domain in nanometers. b) Increase in Δ95-99 emerin oligomers as a function of nuclear stress on 15 and 10 µm wide micropatterns compared to non-deformed nuclei in non- patterned cells. c) Local cluster maps of wild-type and Δ95-99 emerin after nuclear deformation on 15 µm wide micropatterns. M: monomer areas, O: oligomer nanodomains. Scale: 250 nm. 99 deformations become more pronounced in 10 µm patterns, the distribution of Δ95-99 monomers return to non-stress levels, with dispersions in domains 499±250 nm in size (Fig. 3.12a). Contrary to wild-type emerin, a progressive dispersion of Δ95-99 monomers over increasingly large INM areas is not observed with growing stress. At the same time, the formation of oligomers remains very limited, with Δ95-99 molecular densities increasing slightly from 1.3±0.1 to 1.7±0.1 and 2.0±0.1 fold above random for 15 µm and 10 µm micropatterns (Fig. 3.12a and Table 3.2). With the concurrent size enlargement of nanodomains from 48±14 nm to 81±16 nm and 75±20 nm, the relative oligomerization of Δ95-99 compared to non-mechanically stressed nuclei increases by 3.9 fold initially, but does not rise further as nuclear stress intensifies (Fig. 3.12b). Δ95-99 oligomeric nanodomains remains sparser and significantly less dense than for wild-type emerin in cluster maps of nuclei under stress (Fig. 3.12c). Thus, compared to wild-type emerin, the failure of Δ95- 99 to gradually oligomerize at sufficiently high molecular density leads to a defective nuclear responses to force. It underlines the importance of the modular oligomerization of emerin as a response to mechanical stress for adaptive nuclear deformations. 100 3.13 DISCUSSION Figure 3.13. Model of emerin re-organization and oligomerization at the nuclear envelope in response to mechanical challenges. Emerin monomer unbinding from nuclear actin and BAF induces increased lateral mobility at the inner nuclear membrane and favors LEM domain interactions with binding sites along the intrinsically disordered region of other emerin, for the controlled formation of emerin oligomers at SUN1 LINC complexes and their stabilization by lamin A/C. 101 Using a combination of single molecule imaging and quantitative analyses, we revealed that emerin organizes as distinct monomers and oligomers at the INM, and that clinically relevant EDMD mutations impact emerin mobility and nanoscale organization. The oligomerization of emerin at specific sites across the nuclear envelope is modulated by its ability to engage or disengage interactions with different structural elements juxtaposed to the INM, including lamin A/C, nuclear actin, SUN1, and BAF. Lamin A/C and LINC complexes are essential for the stabilization of emerin oligomers and balanced interactions of emerin with BAF and nuclear actin further modulate its diffusion and its oligomerization potential. We also showed that the mechanotransducing functions of emerin are intimately coupled to its oligomerization, with the formation, stabilization and maintenance of emerin oligomers being crucial to nuclear shape adaptation to forces. Indeed, EDMD-inducing mutations that affect the oligomerization of emerin and binding to different structural elements at the INM lead to abnormal nuclear shape adaptation and defective nuclear positioning in response to mechanical stress. While the altered in vitro binding properties of Q133H, Δ95-99 and P183H emerin mutants result in expected differences in organization in vivo, they do not necessarily induce fully predicable changes in diffusion and distribution at the nuclear envelope. In effect, via its flexible IDR, emerin appears to mediate complex interactions with itself, lamin A/C, SUN1, nuclear actin, and BAF, where binding to one partner impacts oligomerization and, interdependently, affects interactions with other partners. As shown from the organization of wild-type emerin after nuclear actin depletion and our study of the Q133H emerin mutant, binding to nuclear actin significantly influences the mobility and the oligomerization potential of emerin. Compared to emerin, actin is relatively large and its binding to the IDR of emerin could mask interaction domains that normally serve as LEM binding and association sites between separate emerin monomers 25 . IDR masking, 102 combined with the reduced lateral mobility of actin-bound emerin, might therefore modulate the formation of oligomers by limiting molecular collisions between emerin monomers. Consistent with the need to precisely regulate emerin oligomerization for appropriate nuclear shape adaptation, increased emerin diffusion and controlled formation of emerin oligomers appear to be coupled molecular events during nuclear adaptation against stress, as our measurements in micropatterns indicate. The deficient binding of Q133H emerin to nuclear actin and its over- oligomerization at INM LINC complexes might thus be linked to the abnormal nuclear deformation observed in cells expressing this mutant. The influence of emerin differential interactions with nucleoplasmic components on its mobility and its oligomerization is also underlined by the surprisingly slow lateral diffusion of the Δ95-99 emerin mutant, which does not bind lamin A/C nor actin in vitro but retains binding to BAF. The 48-118 region of emerin, where Δ95-99 is located, was suggested to act as a potential binding site for the LEM domain 25 and through altered region flexibility, the deletion might reduce the efficacy of such interactions, potentially causing the LEM domain to bind BAF more frequently. Repeated interactions of Δ95-99 with BAF and with the lamina 2 or chromatin 233 via ternary BAF complexes, could slow its diffusion, limit molecular collisions, and impede its oligomerization by sequestering the LEM domain and reducing bridging interactions with LEM binding sites on other emerins 25 . The Δ95-99 deletion, within the 55-132 lamin tail-binding region of emerin 25 , could also prevent a stabilization of already sparse Δ95-99 oligomers by lamin A/C. As such, the aberrant nuclear deformation against stress observed for cells expressing Δ95-99 might stem from out of balance interactions of the deletion mutant with BAF and its inability to form lamin A/C-stabilized oligomers at LINC complexes. 103 A similar impediment of in trans interactions between the LEM domain and self-association sites along the joined IDR of P183H dimers likely result in the slow mobility and the reduced ability of this emerin mutant to oligomerize, despite its retained binding to lamin A/C and SUN1. The reported enhanced binding of lamin A/C to P183H emerin 30,217 , near these same self-association sites could also interfere with inter-emerin bridging interactions. In both cases, the inability of the LEM domain to access binding sites along the IDR would elicit its repeated interactions with BAF, leading to the observed slow diffusion of P183H at the INM. The defective nuclear shape adaptation to mechanical challenges of cells expressing P183H might therefore stem from emerin dimerization and excessive interactions with BAF or lamin A/C that prevent an efficient formation of emerin oligomers and a remodeling of the nucleoskeleton. As we have shown, Δ95-99 emerin does not assemble into dense oligomers at the INM, and it induces aberrant nuclear shape remodeling on micropatterns. The phosphorylation of emerin residues Tyr 74 and Tyr 95 by Scr kinase 245 was recently shown to mediate the recruitment of lamin A/C to the LINC complex during nuclear stiffening in response to force 222 . Our observation that Δ95-99 fails at promoting normal nuclear deformation against mechanical stress by its reduced ability to form lamin A/C and SUN1 stabilized emerin oligomers is consistent with the major role played by Tyr 95 phosphorylation for emerin-mediated mechanotransduction at the nuclear envelope. The abnormal organization of cytoskeletal actin in mechanically challenged cells expressing Δ95-99, akin to disorganizations observed with a non-phosphorylable tyrosine 74-95FF emerin mutant 222 , also suggests that emerin oligomers are required to strengthen the connection between lamin A/C, the LINC complex and cytoplasmic actin filaments in order to promote correct nucleus positioning and deformation. This points towards a link between Tyr 95 phosphorylation, emerin oligomerization and recruitment of lamin A/C for stiffening the nuclear envelope at LINC 104 complexes, that are likely driven by phosphorylation-induced changes in emerin conformation and binding to nuclear partners 221,222,245 . It is thus possible that oligomeric nanodomains at the INM are enriched in Tyr 95 phosphorylated emerin, lamin A/C and LINC complex components, while emerin monomers populate the rest of the INM, consistent with emerin enrichments in distinct nucleoskeletal “niches” at the nuclear envelope 221 . We propose that, during stress responses, transient unbinding of emerin monomers from BAF and nuclear actin favors oligomerization by increasing lateral diffusion at the INM and exposing both the LEM domain and self-association sites along the IDR for intermolecular binding between emerins. Within emerin oligomers, stabilization by direct interaction with the lamin A/C and additional modulation of emerin self-association by nuclear actin and BAF likely allow for a precise regulation of the oligomerization state and the size of oligomeric domains, as required for adaptive nuclear deformation in response to forces. The localized INM distribution of emerin oligomers and their reliance on a part of the emerin IDR that requires phosphorylation for the recruitment of lamin A/C to the LINC complex 222 , also suggest that they are sites where interactions between emerin, lamin A/C and LINC components are strengthened. Emerin oligomerization might therefore contribute to the increased connectivity between the nucleoskeleton, the nuclear envelope and the cytoskeleton for anchoring the nucleus and for providing force absorption contact points during nuclear deformation. In response to surging mechanical stress, the incremental oligomerization of emerin in larger nanodomains we observed could be part of a mechanism that redistributes increasing forces over wider areas to maintain a basal membrane pressure at local anchoring points between the nucleoskeleton, the nuclear envelope and the cytoskeleton. Outside these anchoring oligomeric domains, disengagement of emerin monomers from nuclear actin likely provides additional modulations of nucleoskeletal 105 contacts with the nuclear envelope, for instance by reducing the connections of the INM with the nucleoskeleton, while maintaining its connectivity with nuclear chromatin, via BAF. Indeed, both tethering of chromatin to the nuclear envelope and the nucleoskeleton participate in nuclear mechanics 207,246 . Together, strengthening the connections between the nuclear envelope and the lamina at specific cytoskeletal anchor points, but relaxing them in the rest of the membrane could provide a means to couple controlled nuclear deformation with nucleus positioning in cells. Such coupling is defective with emerin mutants, which display abnormal oligomerization and modified interactions with nucleoplasmic partners. The structural interdependency between emerin monomers, emerin oligomers, BAF, and key elements of the nucleoskeleton likely provides an integrated mechanism where INM/lamina contacts and INM/chromatin contacts are spatially regulated and where reinforced connectivity between the nuclear envelope and the cytoskeleton is provided where needed. At the center of such processes, emerin diffusion and monomer/oligomer exchanges ensure that mechanical cues are transduced throughout the nuclear envelope, for coordinated changes in local nuclear stiffness and appropriate remodeling of the nuclear shape. 3.14 MATERIALS AND METHODS Cell culture, emerin expression and cell staining Emerin-null human dermal fibroblast (EMD -/y HDF) and normal dermal fibroblasts (EMD +/y HDF) were kindly provided by Dr. Howard Worman, Columbia University, New York, USA. EMD -/y Fibroblasts are derived from a male EDMD patient (G-9054) and carry a 59 nucleotide deletion within the EMD gene (EMD g.329del59) that excludes emerin expression 219 . Fibroblasts were grown in DMEM (Lonza) with 10% fetal bovine serum (Gibco-Life Technologies), 50 units ml -1 106 penicillin and 50 μg ml -1 streptomycin and maintained at 37°C in a humidified atmosphere with 5% CO2. Human wild-type emerin cDNA was kindly provided by Dr. Juliet Ellis, University College London, UK. For the expression of PA-TagRFP-emerin, a pEGFP-N1 plasmid backbone encoding emerin fused to the C-terminus of PA-TagRFP was produced by XbaI and KpnI insertion and PCR fusion of the human emerin cDNA. Cells plated on fibronectin-coated glass coverslips were transfected with PA-TagRFP-emerin using X-tremeGENE HP (Roche). 48-72 hours post- transfection, live cells were imaged by sptPALM in HBSS buffer at 37°C. For micropatterning experiments, cells grown on 6-well plates were trypsinized after 48-72 hours of transfection and plated on fibronectin-micropatterned coverslips. To express SNAP-emerin, human emerin was first fused to the C-terminus of a SNAP tag by AscI and XhoI insertion in a pSNAP-tag(m) plasmid (NEB). SNAP-emerin was then subcloned into a modified pFUW lentiviral vector by NheI and AgeI insertion. Lentiviral particles for the expression SNAP-emerin were produced by the UCLA Vector Core. Transduction of HDF grown at 70% confluence on 6-well plates was done for 48 hours, using 25 ng ml -1 of lentiviral particles in complete growth medium containing 8 μg ml -1 of polybrene, after which the medium was replaced. Following another 24 hours incubation, cells were trypsinized, and plated on fibronectin- coated or fibronectin-micropatterned coverslips. For imaging, cells were fixed with 4% paraformaldehyde in PBS for 15 min, permeabilized with 0.1% Triton X-100 (Sigma-Aldrich) for 15 min and blocked with 4% bovine serum albumin (BSA, Sigma-Aldrich) + 0.1% Tween-20 (Sigma-Aldrich) for 30 min, at room temperature. Cells were then stained with 1 µM of SNAP- Surface-AlexaFluor 647 (BG-A647, NEB) in 4% BSA + 0.1% Tween-20 for 1 hour at 37°C, then thoroughly washed before super-resolution imaging. 107 For the expression of sGFP-emerin, humanized cDNA for split-GFP 1-10 231 was inserted by NheI and XbaI digestion in the pEGFP-N1 plasmid backbone encoding emerin, and expressed as an N- terminal fusion to emerin. The shorter 11 th β-sheet M3 fragment 231 was also fused to the N- terminus of emerin by PCR cloning using primers encoding the M3 fragment sequence and subcloning into the pEGFP-N1 backbone via NheI and XhoI digestions. Both plasmids were co- transfected in HDF using X-tremeGENE HP (Roche) as described for PA-TagRFP-emerin, and cells were imaged by CALM 48-72 hours post-transfection. All constructs were verified by sequencing. Upon re-expression of wild-type emerin fusions in EMD -/y HDF, nuclear deformations against mechanical stress recover to levels similar to those of normal EMD +/y HDF 98 . For immunostaining of emerin, cells were grown on coverslips, fixed and permeabilized as described for SNAP-emerin staining. Cells were labeled with a rabbit anti-emerin antibody (1:500, Santa Cruz Biotechnology, sc-15378) for 45 min, washed in PBS and further labeled with a goat anti-rabbit Alexa fluor 488 antibody (1:400, Invitrogen) for 45 min. After washing, cells were mounted in a DAPI-Fluoromount G (Electron Microscopy Sciences) and imaged by confocal microscopy. To label cytoplasmic actin and measure nuclear shape indices, cells were fixed with 4% paraformaldehyde in PBS for 15 min, permeabilized with 0.1% Triton X-100 for 10 min, and blocked with 4% bovine serum albumin + 0.1% Tween-20 for 1 hour. Cells were stained with phalloidin-iFluor 488 (1:1000, Abcam) for 1 hour, washed with PBS, mounted in DAPI- Fluoromount G and imaged by confocal or wide field microscopy. Mutations, siRNA and BAF L58R expression 108 Emerin mutations Q133H and P183H were introduced in PA-TagRFP-emerin, SNAP-emerin, sGFP-emerin and M3-emerin by site-directed mutagenesis using QuickChange Lightning Site Directed Mutagenesis (Agilent) and mutagenic primer pairs for: Q133H: 5’-CGCTTTCCATCACCATGTGCATGATGA-3’ and 5’-GATCGTCATCATGCACATGGTGATGGA-3’ P183H: 5’-CCTGTCCTATTATCATACTTCCTCCTC-3’ and 5’-GTGGAGGAGGAAGTATGATAATAGGA-3’. The Δ95-99 emerin deletion was produced using partially phosphorothioated PCR primers and T7 exonuclease digestion, as previously described 247 . Primers pairs for the Δ95-99 deletion were: 5’-GACTACTTCACCA*C*C*A*GGACTTAT-3’ and 5’-GGTGAAGTAGTCG*T*C*A*TTGTAGCC-3’, where * denotes phosphorothioate modifications. All primers were obtained from Integrated DNA Technologies (IDT) and all mutations were verified by sequencing. siRNA duplex for BAF and Dicer siRNA for lamin A/C were obtained from IDT. The sequences of sense nucleotides were as follows: BAF siRNA: 5’-AGAUUGCUAUUGUCGUACUUU-3’, lamin A/C DsiRNA: 5’-AGCUGAAAGCGCGCAAUACCAAGaa-3’. siRNA duplex for IPO9 was obtained from Ambion (id # S31299). All siRNA were transfected or co-transfected with emerin plasmids at 25 nM using X-tremeGENE HP (Roche). When associated with lentiviral expression of emerin, siRNA transfection was done 2 hours before viral titer application. BAF L58R was expressed from an EGFP-BAF L58R lentiviral plasmid 235 (Addgene #101776) and lentiviral particles were produced by the UCLA Vector Core. HDF cells were transduced with 25 ng ml -1 of lentiviral particles as described for SNAP-emerin. 109 Cell micropatterning and nuclear shape index measurements HDF were micropatterned as described previously 98,227 . Briefly, hexamethyldisilazane-activated glass coverslips (Marienfeld, #1.5, Ø25 mm) were stamped with rectangular and fibronectin- coated polydimethylsiloxane stamps having lengths of 210 µm and widths of 15 µm, 10 µm, or 5 µm respectively. Cell attachment outside the fibronectin strips was blocked with a 1% solution of Pluronic F-127. After attachment for 1 hour and removal of unattached cells, HDF were allowed to spread out on the micropatterns for 6 hours at 37°C before being prepared for microscopy imaging. Using ImageJ 248 , the nuclear shape index (NSI) 183 was determined by measuring the nuclear cross- sectional area and the nuclear perimeter of DAPI-stained nuclei imaged by wide-field microscopy, and by calculating the ratio: NSI = 4×π×area perimeter 2 (1) The NSI measures the roundness of the nucleus with an NSI of 1 corresponding to a circular nuclear shape. Mean NSI values ± standard deviation of the mean are reported for multiple cell nuclei per condition. Microscopy imaging Confocal imaging of immunostained emerin in normal HDF, emerin-null HDF and emerin null HDF after expression of wild-type PA-TagRFP-emerin were performed on an Olympus Fluoview FV1000 confocal microscope equipped with a 60x/1.40 NA objective, 405 nm and 488 nm lasers 110 and appropriate emission filters for imaging DAPI (450/20 nm) and Alexa-488 labeled antibodies against emerin (510/10 nm). Confocal imaging of nuclear deformation and actin organization in micropatterned EMD +/y HDF was performed on a Zeiss LSM 700 microscope equipped with a C-Apochromat 63×/1.15 W Korr objective, excitation lasers at 405 nm and 488 nm and appropriate beamsplitter and emission channel settings for dual detection of DAPI and phalloidin-iFluor 488. Wide-field imaging of labeled actin and labeled nuclei for NSI measurements were performed on an inverted Nikon Eclipse Ti-E microscope equipped with a 40x objective (Nikon), an iXon Ultra EMCCD camera (Andor), 405 nm and 488 nm lasers, a T495lpxr dichroic mirror and a 525/50 emission filter (Chroma) for phalloidin-iFluor 488 or a 458Di02 dichroic mirror and a 483/32 emission filter (Semrock) for DAPI. SptPALM, dSTORM and CALM imaging were performed on an inverted Nikon Eclipse Ti-E microscope equipped with a 100x 1.49 NA objective (Nikon), an iXon EMCCD camera (Andor), perfect focus drift compensation optics, an astigmatic lens for 3D super-resolution imaging, a piezo z-scanner for calibration of 3D super-resolution images (Mad City Labs), laser lines at 405, 488, 561 and 647 nm (Agilent), a multiband pass ZET405/488/561/647x excitation filter (Chroma), a quad-band ZT405/488/561/647 dichroic mirror (Chroma) and appropriate emission filters for sptPALM imaging of PA-tagRFP (600/50 nm, Chroma), 3D-dSTORM imaging of Alexa-647 (700/75 nm, Chroma) and CALM imaging of complemented split-GFP (525/50 nm, Semrock). sptPALM of PA-TagRFP-emerin was performed in 37ºC HBSS buffer (Corning) by HILO excitation of the bottom nuclear membrane of cells with a continuous and low power photoactivation at 405 nm and an excitation at 561 nm. The HILO illumination angle was θ= 51.6º. Images were acquired continuously at a frame rate of 40 ms per frame for no longer than 3 minutes 111 per cell to limit UV damage. CALM imaging of complemented emerin-GFP-emerin species was done as described for sptPALM but with a single HILO excitation at 488 nm. dSTORM of SNAP-emerin labeled with BG-A647 was performed at room temperature in a photoswitching buffer composed of 10% glucose, 0.5 mg ml -1 glucose oxidase (Sigma), 40 μg ml - 1 catalase (Sigma), and 1% β-mercaptoethanol (Sigma). Continuous photoswitching was achieved with a low power 488 nm laser and imaging was done with a 647 nm laser excitation at a frame rate of 80 ms per frame. Z-calibration and sample drift corrections were done using a few 40 nm TransFluoSphere beads (488/685 nm, Life Technologies) as fiducial markers spread on the cell samples. Analyses of diffusion coefficients Localization and tracking analyses were performed using the software package SLIMfast, which uses multiple-target tracing algorithms 198 and was kindly provided by Christian Ritcher and Jacob Piehler. Localizations were performed by 2D-gaussian fitting of the point-spread-function of each activated PA-TagRFP-emerin or activated emerin-GFP-emerin species in each frame. Localization precision was determined as previously described 113 , and individual PA-TagRFP-emerin were localized with a precision of 13±5 nm. Diffusion trajectories were built by linking localizations frame to frame and accounting for blinking statistics and local particle densities. Trajectories with fewer than three steps were discarded. Diffusion coefficients were estimated using a probability density of square displacement (PDSD) analysis 229 . For each time lag t, the PDSD curve was fitted with the following model: 𝑃 ( 𝑟 ⃗ 2 , 𝑡 ) = 1 − ∑ 𝑎 𝑖 ( 𝑡 ) 𝑒 −𝑟 2 /𝑟 𝑖 2 ( 𝑡 ) 𝑛 𝑖 =1 (2) 112 ∑ 𝑎 𝑖 ( 𝑡 ) = 1 𝑛 𝑖 =1 where ri 2 (t) is the square displacement and ai(t) is the population density of i numbers of diffusive behaviors at each time lag t. To limit the risks of overfitting or underfitting the PDSD curves and select an appropriate model for i numbers of diffusive behaviors in each data set, we used both an Akaike information criterion (AIC) and a Bayesian information criterion (BIC) after fitting PDSD with models where 1≤ i ≥5. Square displacement curves (ri 2 (t)) were extracted from PDSD analyses and reported with error bars determined using 𝑟 𝑖 2 √ 𝑁 , where N is the number of analyzed trajectories per time lag, as previously described 229 . Diffusion coefficients (D) representative of each behaviors were determined by fitting each ri 2 (t) curves over the first four time lags (t1-t4) using OriginPro 2020 software (OriginLab) and a 2D Brownian diffusion model with position error: 𝑟 2 = 4𝐷𝑡 + 4𝜎 2 (3) All diffusion coefficients D are reported in µm 2 s -1 ± standard error of fit value (± s.e.m.). Statistical comparisons between D values were done using F-tests. Population percentages are derived from the averaged ai(t) values over the considered time lags. Individual diffusion coefficients (Di) were obtained by fitting the individual mean square displacement (MSD) for each detected emerin over the first three time lags (t1-t3), using again a 2D Brownian diffusion model. Based on their individual Di value, emerin trajectories were grouped into four diffusion ranges (Di1> 0.1 µm 2 s -1 , 0.1<Di2> 0.01 µm 2 s -1 , 0.1<Di3> 0.001 µm 2 s -1 , and D i4< 0.001 µm 2 s -1 ) and plotted as maps. 113 Spatial distribution and cluster analyses from super-resolution images After 3D-dSTORM super-resolution imaging, the localization of individual emerin molecules and z-position assignments were performed by Gaussian fitting using rapidSTORM (version 3.3.1) 200 . Sample drift and overcounting corrections for molecules appearing in consecutive frames were done using PALMsiever 201 and renderings of super-resolved images were done using ImageJ 248 . Localization precisions (σ) in the x, y, and z dimensions were evaluated as previously described 227 and were σx:8.3 nm, σy:13.0 nm, and σz:28.4 nm. 2D spatial pattern analyses of emerin distributions were performed on 2 µm x 2 µm regions of interest (ROI) typically chosen in nuclear envelope areas having homogenous z ranges and away from the edges of nuclei, in order to limit 3D effects. Emerin clustering was determined using an edge-corrected neighborhood density function (NDF) as previously described 227 . Briefly, the NDF is a pairwise-correlation function similar to O-ring statistics that tallies the density of detected emerin within a ring of outer radius r and width Δr located at a distance r from an emerin position in the ROI and for all r + Δr in the ROI. The density of emerin as a function of distance from an average emerin was obtained with: 𝐷 𝑟 = ∑ 𝑁 𝑟 ∑ 𝐴 𝑟 (4) where Nr is the number of neighbors and Ar is the area summed over all detected emerin. NDF analyses were done over a 1µm distance on selected ROIs and with a fixed ring width of 10 nm and a ring radius increasing by 10 nm steps. To average NDF statistics from multiple ROIs across different nuclei and make them sample-size independent, Dr was further standardized by dividing it by the mean density of detected emerin across the entire ROI. As such, an NDF value at a given radius indicates the relative clustering of emerin as compared to the average density across the entire sample. This relative NDF gives a value of 1 for a completely random spatial distribution as 114 determined by Monte Carlo simulations of random emerin distributions with area and number of randomly seeded emerin equal to that of each experimental ROIs. Relative NDF curves averaged across multiple ROIs and multiple nuclei were fitted with a previously described model 227 , which accounts for a distribution of cluster lengths that includes two populations of emerin (monomer and oligomers) and for a probability density of emerin in 2D clusters that decays approximately as an exponential function 236 : Relative NDF = {𝐴 1 × exp( −𝑟 𝜀 1 ) + 𝐴 2 × exp( −𝑟 𝜀 2 ) + 1} ∗ 𝑔 ( 𝑟 ) 𝑃𝑆𝐹 (5) where, 𝐴 is the clustering density, 𝜀 is the typical half-maximum cluster length, ∗ denotes a 2D convolution, and 𝑔 ( 𝑟 ) 𝑃𝑆𝐹 is the correlation function of the effective point spread function of uncertainty in position determination for the dSTORM experiments. As described previously 227 , 𝑔 ( 𝑟 ) 𝑃𝑆𝐹 corrects the NDF for contribution of multiple single molecule appearances (blinking) to the overall spatial distribution. After fitting relative NDF curves, the molecular density above random for emerin clusters are reported as 𝐴 ± standard error of the fit (±s.e.m) and their typical length scales as 2 × 𝜀 ± standard error of the fit and localization precision (±s.e.m). Relative increases in emerin oligomer formation during nuclear stress were determined by considering a circular shape of oligomer nanodomains and multiplying the area of oligomerization by the measured molecular density. Cluster maps Cluster maps were generated from drift- and overcounting-corrected super-resolved emerin positions by determining local cluster values around each emerin using the Getis and Franklin L function 155 in spPack 249 and for a distance of 25 nm. Spatial positions in x and y, and cluster values were plotted as maps in MATLAB (MathWorks) using the meshgrid and griddata functions, a 1 115 nm x 1 nm pixel size and the 'v4' option when calculating pixel density values. The contour maps were generated using the countourf function with 200 levels. In contour maps, values L(r25)=25 represent areas where emerin is randomly distributed and L(r25)=70 values represent areas with emerin local density (70/25) 2 = ~8-fold higher than expected for a random distribution. Cell extracts and immunoblotting After siRNA treatment, cells were harvested and fractionated as described previously7. Cells were scraped, centrifuged for 5 min at 4000 g and washed three times with PBS. Cell pellets were then flash frozen and stored at -80°C. Cells were thawed for 10 minutes on ice, and then for 10 min in 300 µl ice cold hypotonic lysis buffer (20 mM HEPES, pH 7.4, 50 mM N-acetylglucosamine, 1 mM DTT, 100 µM PMSF, 1 µg/ml pepstatin A, 1x protease inhibitor (Roche) and 1x PhosSTOP phosphatase inhibitor (Roche) before being resuspended and set on ice for another 10 min. Cells were then spun down at 17,000 g for 1 min to collect the supernatant cytoplasmic fraction. The pellet was then washed three times in PBS with spinning at 17,000 g for 1 min in between washes to remove any residual cytoplasmic components. The pellet was then resuspended in nuclear lysis buffer (50 mM Tris-HCL, pH 7.4, 300 mM NaCl, 0.3% Triton X-100, 50 mM N- acetylglucosamine, 1 mM DTT, 100 µM PMSF, 1 µg/ml pepstain A, 1x protease inhibitor (Roche) and 1x PhosSTOP phosphatase inhibitor (Roche) before vortexing to release nuclear contents. Nuclear fractions were additionally sonicated 20 times in 0.5 s bursts to liberate dense nuclear aggregates. Cytoplasmic and nuclear fractions were analyzed by SDSPAGE. Protein concentrations were determined using a Bradord assay, and equal amounts of protein were loaded on gels before running in a Laemmli Buffer. Proteins were transferred to a nitrocellulose membrane (Bio-Rad) by wet transfer at 4°C. The membrane was then rinsed with Tris-buffered 116 saline (TBS, pH 7.5) and blocked with 5% milk in TBS for 1 hour. Membranes were probed with the following primary antibodies: mouse anti-lamin A/C (1:1000, Santa Cruz Biotechnology, sc- 7292), mouse anti-actin (1:1000, Santa Cruz Biotechnology, sc-8432), mouse anti-BAF (1:1000, Santa Cruz Biotechnology, sc-166324), rabbit anti-SUN1 (1:2000; HPA008346, Sigma-Aldrich), mouse anti-histone H2A (1:1000, Santa Cruz Biotechnology, sc2 515808) and mouse anti- GAPDH (1:1000, GeneTex, GTX627408). A goat anti-mouse IgG (H+L) HRP conjugate (1:3000, Invitrogen, 62-6520) and a goat anti-rabbit IgG (H+L) HRP conjugate (1:5000, Invitrogen, 65- 6120) were used as secondary antibodies. Blots were developed with SuperSignal West Femto Maximum Sensitivity Substrate (Thermo) and imaged on a Chemidoc XRS+ (BioRad). All assays were performed in triplicate and immunoblot quantification was done with ImageJ, using paired T-tests for statical comparisons with wild-type. Immunostaining and fluorescence imaging of RNAi effects EMD +/y human dermal fibroblasts were transfected with control siRNAs, or siRNAs against lamin A/C, IPO9, BAF, and SUN1, respectively as described in Materials and Methods. Cell were fixed with 4% PFA in 1x PBS for 15 min, permeabilized with 0.1% Triton X-100 for 10 min, and blocked with 4% BSA + 0.1% Tween-20 for 1h, all at RT. For lamin A/C and SUN1 immuno- staining, cells were incubated with mouse anti-lamin A/C (1:400; sc-7292, Santa Cruz Biotechnology) and rabbit anti-SUN1 (1:500; HPA008346, Sigma-Aldrich) for 1hr at RT, rinsed 3x with 1x PBS for 5 minutes each, then stained with goat anti-mouse-Alexa Fluor 647 (1:500, Life Technologies) and goat anti-rabbit-Alexa Fluor 488 (1:500, Life Technologies) for 1hr at RT. Following 3x PBS rinse, the coverslips were mounted in DAPI-Fluoromount G (Electron Microscopy Sciences) and imaged. For BAF staining, cells were incubated with mouse anti-BAF (1:100; sc-166324, Santa Cruz Biotechnology) for 2hr at RT, then with goat anti-mouse-Alexa 117 Fluor 647 (1:500) for further 2hr. To assess the effect of BAF L58R re-expression on lamin A/C organization, 24 hours after siRNA treatment against BAF, cells were transduced with the lentivirus for EGFP-BAF L58R expression and fixed two days later, before staining for lamin A/C as described above. Confocal fluorescence imaging was performed on a Zeiss LSM 780 microscope equipped with a C-Apochromat 63×/1.15 W Korr objective, excitation lasers at 405 nm, 488 nm and 633 nm, and appropriate beam splitter and emission channel settings for detection of DAPI, GFP, Alexa Fluor-488 and Alexa Fluor-647 labeled secondary antibodies. Other than effective and specific knock down of targeted protein expressions, there was no obvious indirect or off-target effects of the different siRNA treatments on the organization other studied proteins. 118 CHAPTER 4: CONCLUSIONS 4.1: CONCLUSIONS The contribution of the EMD gene to muscular dystrophy has long been known 23 , but the mechanical reasons for why an inner nuclear envelope protein expressed in all somatic tissues leads to improper muscle cell coordination and differentiation has remained elusive 49,250 . Only with more recent technological innovations have we been able to appreciate the unique role that emerin plays in adapting the nucleus to mechanical stress 72 . Furthermore, emerin has been identified as a contributor to a larger group of diseases called laminopathies which all involve proteins at the nuclear envelope and cause similar pathologies 251 . Study of the disease has proved difficult because it requires monitoring the adaptation of the nucleus to exogenous mechanical force 252 . Without techniques that allow for manipulation of cells and observation of their responses or higher resolution imaging techniques, we could not observe significant differences between wildtype and mutated forms of emerin. Only with super resolution techniques are we able to observe differences in the organization and dynamics of emerin when mutated. While fascinating, this alone is unable to address the role of these unique features of emerin that heretofore have been uncharacterized and does not explain how these perturbations contribute to the manifestation of the disease. These experiments cannot apply these findings to the role of emerin in mechanotransduction because the exogenous force that culture cells experience is not controlled. Only with techniques that apply force to individual cells can we understand the role of the unique organization and dynamics of emerin in contributing to the mechno-adaptibility of nuclei and prevention of nuclear fragility. 119 While techniques to apply mechanical stress have been used previously, the methods are unsuitable for use with single molecule microscopy, and the features that we found would be obscured with these approaches. Only by adapting these experimental conditions have we been able to demonstrate how the organization and dynamics of emerin are altered upon introduction of mechanical stress. By combining analysis of nanoscale emerin features and the application of force to nuclei via micropatterning, emerin has been shown to be vital to cellular response to mechanical perturbations and necessary for proper nuclear restructuring to adapt to the extracellular environment. Since the proper regulation of mechanical stress at the nucleus requires not just the presence of emerin, but also proper modification of its organization and dynamics, it is no surprise that the mutations of the protein which lead to aberrant interaction with nuclear proteins, LINC complex components, and improper oligomerization, can in the presence of stress result in muscular dystrophy. We have demonstrated that, as in the case when cells completely lack expression of emerin, these mutations prevent proper adaptation of the nucleus to mechanical stress, and as such they are unable to properly change their shape when presented with a defined cell spreading area. This process is crucial for processes like muscle contraction, when significant mechanical stress is placed on nuclei which must adapt to prevent breakage of the nucleus, but also for differentiation since this requires proper relocation of the nucleus to specific areas in the cell. In patients lacking emerin, muscle cells often fail to differentiate, or, when they do, they often apoptize and are replaced by adipose tissue, leading to the symptoms of low muscle mass and myotonia. Specifically the LINC complex has been found to be crucial to this process 253 , and it is not entirely surprising that altered emerin organization would impair nuclear responses, since these perturbed 120 emerin microdomains likely are critical for proper establishment of LINC complex as emerin is itself a major member of the LINC complex 170 . While emerin has long been known to be crucial to avoiding the occurrence of muscular dystrophy, little was known about how the protein prevents the disease and how, even in cases where the localization of the protein seemed normal, that the disease still fully manifests. Since we knew that the standard microscopic techniques that had been used to study the protein are insufficient for determining the nanometer scale organization and dynamics of proteins, we expected to find significant differences in the behavior of the protein using high resolution optical approaches. Using super resolution optical microscopy techniques, we determined that emerin exists in two distinct diffusive and organizational states at the inner nuclear envelope. We also found that upon mutation of the protein, these diffusive and organizational states are altered, differences that are invisible since they occur well below the diffraction limit of visible light. Though we did for the first time show significant differences in the behavior and organization of emerin in mutated forms, we still needed to address how these differences could account for the altered mechanical stability of cells deficient in emerin. To this end, we developed a surface silanization and micro-contact printing method to uniformly induce mechanical stress onto cells while still being able to perform our single molecule measurements. Unlike previous approaches to impose mechanical stress on cultured cells, our technique is technically simple, can be scaled up easily, and does not sacrifice optical quality. Additionally, since the patterns are printed onto our glass coverslips from manufactured PDMS stamps, an unlimited number of designs are possible to study mechanical stress on cells. Using this micropatterning technique, we determined that the presence of emerin is critical to the proper regulation of nuclear shape under mechanical stress. Cells lacking emerin showed an 121 inability to properly adapt to mechanical stress as expected 72 manifested by improper nuclear shapes. This is critical to understanding the increased apoptosis in cultured cells of patients with muscular dystrophy 66 . It also explains the lack of proper differentiation of muscle cells 254 since they require a distinct nuclear localization to properly differentiate into myocytes 255 . We were also able to examine the unique diffusion and organization of emerin upon induction of this mechanical stress. We showed that both inner nuclear envelope populations of emerin slow in response to the constrained spreading area provided by the microcontact printed stamps, and we also found that the distribution of emerin itself is impacted by the introduction of mechanical stress. Whereas emerin is normally found almost exclusively at the nuclear periphery, introduction of cell stress induced a trafficking of emerin outside of the nucleus and into the ER, consistent with altered lamin interactions which are critical for proper emerin retention in the nucleus. We also showed that the distribution of emerin itself is significantly altered in the presence of mechanical stress. Emerin cluster sizes increased in order to maintain constant pressure at the nucleus, while apparent density decreased as a result of this larger area. Similarly, the diffusion of emerin itself was impacted by this mechanical rearrangement, as the two inner nuclear envelope populations slowed considerably, indicative of a distinct interaction with lamin. This adaptation of dynamics and organization are ablated when emerin is mutated or when nuclear architecture components are altered. Under these conditions we find no significant changes from non-patterned to patterned conditions. That is, emerin adapts a behavior that is either like a cell that is consistently stressed or one that is constantly relaxed. These adaptations thus seem to be critical for properly adapting the nucleus to exogenous stress, as the nucleus must either correctly relax or stiffen in response to exogenous mechanical cues. 122 Several responses must occur when a cell is rearranging to fit into a constricted environment. Among these, emerin must be shuttled out of the nuclear envelope, while the emerin that remains must be highly organized (and concurrently, it must also diffuse more slowly). Emerin mutations do not prevent its trafficking outside of the nuclear envelope, but they do prevent its proper clustering at the nucleus. These clustered domains of slow diffusion must be critical for properly orienting the nucleus to respond to mechanical force, locally stiffen, and relax, and organize the cytoskeleton such that intracellular stress can be placed on the nucleus which will allow it to adapt to the constriction rather than shear. Mutations which cannot form these clusters instead have altered nuclear mechanics, and this likely explains the dystrophy phenotype. 4.2: FUTURE DIRECTIONS Our single molecule imaging techniques have been critical for discovering the nanoscale organization of emerin and characterizing its slow diffusion. While we have found that emerin is critical for the reorganization of the nucleus, we still want to understand how the nucleus reorganizes. We would like to see how the lamin network reorganizes in the presence of mechanical stress and how the organization of the lamin networks relates to the emerin clusters. Another interesting experiment would be to see how the actin cytoskeleton is reorganized in the presence of emerin mutations. Since we have found that nuclear actin is critical for maintaining proper emerin organization, it will be interesting to see how the organization of actin itself is impacted in the presence of the mutations that are known to alter actin polymerization and binding. Using Integrated exchangeable single-molecule localization (IRIS) 256 , we can use a Lifeact peptide conjugated to Atto488 to directly stain and localize actin in the nucleus. 123 In addition to imaging the structural features of the nucleus, we can also assess the impact of mechanical stress and emerin mutations on multiple emerin binding partners and even genome organization. Using the engineered APEX enzyme 257 , we can create fusions to emerin to label and detect those proteins and DNA fragments that are in close proximity to the protein. This will allow us to investigate the impact of emerin mutations on the spatial organization of genomic DNA in response to muscular dystrophy and introduction of mechanical stress. The latter can be probed by plating cells onto micropatterned substrates. These plates can allow for more than a million cells to be forced into a narrow pattern certain widths, much like our micropatterned coverslips, though the plates would contain many more cells which would allow for mass spec and mass DNA sequencing which can require hundreds of thousands of cells. Another unaddressed question is the role of emerin in response to a cell not in a steady state equilibrium after application of mechanical stress, but what emerin does as force is first applied to a cell. Other studies have shown that force can be directly applied using a micropipette onto plated cells. With a cell in focus on a microscope, we can use a micropipette to apply a suction force to a cell at known force 86 . We can then perform sptPALM on live nuclei with suction applied onto the nucleus. Through this method we can see how emerin helps to adapt the nucleus to direct application of force rather than seeing how emerin behaves once the nucleus has already adapted to force. Overall, our work has addressed a question until now has been underappreciated, which is the role of the nanoscale organization and dynamics of the mechanosensitive inner nuclear membrane protein emerin. This blend of the latest optical microscopic methods and quantitative analysis has addressed questions that previously were infeasible to research due to the limitations of classical imaging techniques. Now with the combined approaches, it is now possible to investigate the 124 formerly unexplored universe of nanoscale adaptations of nuclear architecture from a wide family of nuclear membrane proteins which prevent disease and contribute to normal cellular operations and ultimately organismal survival. 125 BIBLIOGRAPHY 1 Laguri, C. et al. Structural characterization of the LEM motif common to three human inner nuclear membrane proteins. Structure 9, 503-511, doi:10.1016/s0969-2126(01)00611-6 (2001). 2 Samson, C. et al. Structural analysis of the ternary complex between lamin A/C, BAF and emerin identifies an interface disrupted in autosomal recessive progeroid diseases. Nucleic Acids Res, doi:10.1093/nar/gky736 (2018). 3 Bouzid, T. et al. The LINC complex, mechanotransduction, and mesenchymal stem cell function and fate. J Biol Eng 13, 68, doi:10.1186/s13036-019-0197-9 (2019). 4 Dutta, S., Bhattacharyya, M. & Sengupta, K. Changes in the Nuclear Envelope in Laminopathies. Adv Exp Med Biol 1112, 31-38, doi:10.1007/978-981-13-3065-0_3 (2018). 5 Ledeen, R. W. & Wu, G. Nuclear lipids: key signaling effectors in the nervous system and other tissues. J Lipid Res 45, 1-8, doi:10.1194/jlr.R300015-JLR200 (2004). 6 Beck, M. & Hurt, E. The nuclear pore complex: understanding its function through structural insight. Nat Rev Mol Cell Biol 18, 73-89, doi:10.1038/nrm.2016.147 (2017). 7 Dittmer, T. A. & Misteli, T. The lamin protein family. Genome Biol 12, 222, doi:10.1186/gb-2011-12-5-222 (2011). 8 Dutta, S., Bhattacharyya, M. & Sengupta, K. Implications and Assessment of the Elastic Behavior of Lamins in Laminopathies. Cells 5, doi:10.3390/cells5040037 (2016). 9 Holmer, L. & Worman, H. J. Inner nuclear membrane proteins: functions and targeting. Cell Mol Life Sci 58, 1741-1747, doi:10.1007/PL00000813 (2001). 10 Strasser, C., Grote, P., Schauble, K., Ganz, M. & Ferrando-May, E. Regulation of nuclear envelope permeability in cell death and survival. Nucleus 3, 540-551, doi:10.4161/nucl.21982 (2012). 11 Tingey, M., Mudumbi, K. C., Schirmer, E. C. & Yang, W. Casting a Wider Net: Differentiating between Inner Nuclear Envelope and Outer Nuclear Envelope Transmembrane Proteins. Int J Mol Sci 20, doi:10.3390/ijms20215248 (2019). 12 Janin, A., Bauer, D., Ratti, F., Millat, G. & Mejat, A. Nuclear envelopathies: a complex LINC between nuclear envelope and pathology. Orphanet J Rare Dis 12, 147, doi:10.1186/s13023-017-0698-x (2017). 13 Wong, X., Luperchio, T. R. & Reddy, K. L. NET gains and losses: the role of changing nuclear envelope proteomes in genome regulation. Curr Opin Cell Biol 28, 105-120, doi:10.1016/j.ceb.2014.04.005 (2014). 14 Choi, J. C. & Worman, H. J. Nuclear envelope regulation of signaling cascades. Adv Exp Med Biol 773, 187-206, doi:10.1007/978-1-4899-8032-8_9 (2014). 15 Kim, D. I., Birendra, K. C. & Roux, K. J. Making the LINC: SUN and KASH protein interactions. Biol Chem 396, 295-310, doi:10.1515/hsz-2014-0267 (2015). 16 Osmanagic-Myers, S., Dechat, T. & Foisner, R. Lamins at the crossroads of mechanosignaling. Genes Dev 29, 225-237, doi:10.1101/gad.255968.114 (2015). 17 Kornberg, R. D. & Lorch, Y. Chromatin structure and transcription. Annu Rev Cell Biol 8, 563-587, doi:10.1146/annurev.cb.08.110192.003023 (1992). 18 Mellor, J. The dynamics of chromatin remodeling at promoters. Mol Cell 19, 147-157, doi:10.1016/j.molcel.2005.06.023 (2005). 19 Talamas, J. A. & Capelson, M. Nuclear envelope and genome interactions in cell fate. Front Genet 6, 95, doi:10.3389/fgene.2015.00095 (2015). 126 20 Herrada, I. et al. Purification and Structural Analysis of LEM-Domain Proteins. Methods Enzymol 569, 43-61, doi:10.1016/bs.mie.2015.07.008 (2016). 21 Brull, A., Morales Rodriguez, B., Bonne, G., Muchir, A. & Bertrand, A. T. The Pathogenesis and Therapies of Striated Muscle Laminopathies. Front Physiol 9, 1533, doi:10.3389/fphys.2018.01533 (2018). 22 Astejada, M. N. et al. Emerinopathy and laminopathy clinical, pathological and molecular features of muscular dystrophy with nuclear envelopathy in Japan. Acta Myol 26, 159-164 (2007). 23 Bione, S. et al. Identification of a novel X-linked gene responsible for Emery-Dreifuss muscular dystrophy. Nat Genet 8, 323-327, doi:10.1038/ng1294-323 (1994). 24 Manilal, S., Nguyen, T. M., Sewry, C. A. & Morris, G. E. The Emery-Dreifuss muscular dystrophy protein, emerin, is a nuclear membrane protein. Hum Mol Genet 5, 801-808 (1996). 25 Berk, J. M. et al. The molecular basis of emerin-emerin and emerin-BAF interactions. J Cell Sci 127, 3956-3969, doi:10.1242/jcs.148247 (2014). 26 Wolff, N. et al. Structural analysis of emerin, an inner nuclear membrane protein mutated in X-linked Emery-Dreifuss muscular dystrophy. FEBS Lett 501, 171-176, doi:10.1016/s0014-5793(01)02649-7 (2001). 27 Cai, M. et al. Solution structure of the constant region of nuclear envelope protein LAP2 reveals two LEM-domain structures: one binds BAF and the other binds DNA. EMBO J 20, 4399-4407, doi:10.1093/emboj/20.16.4399 (2001). 28 Herrada, I. et al. Muscular Dystrophy Mutations Impair the Nuclear Envelope Emerin Self- assembly Properties. ACS Chem Biol 10, 2733-2742, doi:10.1021/acschembio.5b00648 (2015). 29 Donnaloja, F., Carnevali, F., Jacchetti, E. & Raimondi, M. T. Lamin A/C Mechanotransduction in Laminopathies. Cells 9, doi:10.3390/cells9051306 (2020). 30 Berk, J. M., Tifft, K. E. & Wilson, K. L. The nuclear envelope LEM-domain protein emerin. Nucleus 4, 298-314, doi:10.4161/nucl.25751 (2013). 31 Clark, T. G. & Merriam, R. W. Diffusible and bound actin nuclei of Xenopus laevis oocytes. Cell 12, 883-891, doi:10.1016/0092-8674(77)90152-0 (1977). 32 Holaska, J. M., Kowalski, A. K. & Wilson, K. L. Emerin caps the pointed end of actin filaments: evidence for an actin cortical network at the nuclear inner membrane. PLoS Biol 2, E231, doi:10.1371/journal.pbio.0020231 (2004). 33 Mehta, I. S., Elcock, L. S., Amira, M., Kill, I. R. & Bridger, J. M. Nuclear motors and nuclear structures containing A-type lamins and emerin: is there a functional link? Biochem Soc Trans 36, 1384-1388, doi:10.1042/BST0361384 (2008). 34 Tojkander, S., Gateva, G. & Lappalainen, P. Actin stress fibers--assembly, dynamics and biological roles. J Cell Sci 125, 1855-1864, doi:10.1242/jcs.098087 (2012). 35 Shao, X., Li, Q., Mogilner, A., Bershadsky, A. D. & Shivashankar, G. V. Mechanical stimulation induces formin-dependent assembly of a perinuclear actin rim. Proc Natl Acad Sci U S A 112, E2595-2601, doi:10.1073/pnas.1504837112 (2015). 36 Luxton, G. W., Gomes, E. R., Folker, E. S., Worman, H. J. & Gundersen, G. G. TAN lines: a novel nuclear envelope structure involved in nuclear positioning. Nucleus 2, 173-181, doi:10.4161/nucl.2.3.16243 (2011). 127 37 Gudise, S., Figueroa, R. A., Lindberg, R., Larsson, V. & Hallberg, E. Samp1 is functionally associated with the LINC complex and A-type lamina networks. J Cell Sci 124, 2077-2085, doi:10.1242/jcs.078923 (2011). 38 Haque, F. et al. Mammalian SUN protein interaction networks at the inner nuclear membrane and their role in laminopathy disease processes. J Biol Chem 285, 3487-3498, doi:10.1074/jbc.M109.071910 (2010). 39 Chang, W., Worman, H. J. & Gundersen, G. G. Accessorizing and anchoring the LINC complex for multifunctionality. J Cell Biol 208, 11-22, doi:10.1083/jcb.201409047 (2015). 40 Ostlund, C., Ellenberg, J., Hallberg, E., Lippincott-Schwartz, J. & Worman, H. J. Intracellular trafficking of emerin, the Emery-Dreifuss muscular dystrophy protein. J Cell Sci 112 ( Pt 11), 1709-1719 (1999). 41 Weiss, M. Challenges and artifacts in quantitative photobleaching experiments. Traffic 5, 662-671, doi:10.1111/j.1600-0854.2004.00215.x (2004). 42 Hopkins, L. C., Jackson, J. A. & Elsas, L. J. Emery-dreifuss humeroperoneal muscular dystrophy: an x-linked myopathy with unusual contractures and bradycardia. Ann Neurol 10, 230-237, doi:10.1002/ana.410100306 (1981). 43 Emery, A. E. Emery-Dreifuss muscular dystrophy - a 40 year retrospective. Neuromuscul Disord 10, 228-232 (2000). 44 Shieh, P. B. Muscular dystrophies and other genetic myopathies. Neurol Clin 31, 1009- 1029, doi:10.1016/j.ncl.2013.04.004 (2013). 45 Ellis, J. A. Emery-Dreifuss muscular dystrophy at the nuclear envelope: 10 years on. Cell Mol Life Sci 63, 2702-2709, doi:10.1007/s00018-006-6247-8 (2006). 46 Clements, L., Manilal, S., Love, D. R. & Morris, G. E. Direct interaction between emerin and lamin A. Biochem Biophys Res Commun 267, 709-714, doi:10.1006/bbrc.1999.2023 (2000). 47 Burke, B. & Stewart, C. L. Life at the edge: the nuclear envelope and human disease. Nat Rev Mol Cell Biol 3, 575-585, doi:10.1038/nrm879 (2002). 48 Mora, M. et al. X-linked Emery-Dreifuss muscular dystrophy can be diagnosed from skin biopsy or blood sample. Ann Neurol 42, 249-253, doi:10.1002/ana.410420218 (1997). 49 Hutchison, C. J., Alvarez-Reyes, M. & Vaughan, O. A. Lamins in disease: why do ubiquitously expressed nuclear envelope proteins give rise to tissue-specific disease phenotypes? J Cell Sci 114, 9-19 (2001). 50 Holt, I., Clements, L., Manilal, S. & Morris, G. E. How does a g993t mutation in the emerin gene cause Emery-Dreifuss muscular dystrophy? Biochem Biophys Res Commun 287, 1129-1133, doi:10.1006/bbrc.2001.5708 (2001). 51 Yates, J. R. et al. Genotype-phenotype analysis in X-linked Emery-Dreifuss muscular dystrophy and identification of a missense mutation associated with a milder phenotype. Neuromuscul Disord 9, 159-165 (1999). 52 Nagano, A. et al. Emerin deficiency at the nuclear membrane in patients with Emery- Dreifuss muscular dystrophy. Nat Genet 12, 254-259, doi:10.1038/ng0396-254 (1996). 53 Fairley, E. A., Kendrick-Jones, J. & Ellis, J. A. The Emery-Dreifuss muscular dystrophy phenotype arises from aberrant targeting and binding of emerin at the inner nuclear membrane. J Cell Sci 112 ( Pt 15), 2571-2582 (1999). 54 Fairley, E. A., Riddell, A., Ellis, J. A. & Kendrick-Jones, J. The cell cycle dependent mislocalisation of emerin may contribute to the Emery-Dreifuss muscular dystrophy phenotype. J Cell Sci 115, 341-354 (2002). 128 55 Pradhan, R., Ranade, D. & Sengupta, K. Emerin modulates spatial organization of chromosome territories in cells on softer matrices. Nucleic Acids Res 46, 5561-5586, doi:10.1093/nar/gky288 (2018). 56 Lammerding, J. et al. Abnormal nuclear shape and impaired mechanotransduction in emerin-deficient cells. J Cell Biol 170, 781-791, doi:10.1083/jcb.200502148 (2005). 57 Fletcher, D. A. & Mullins, R. D. Cell mechanics and the cytoskeleton. Nature 463, 485- 492, doi:10.1038/nature08908 (2010). 58 Martino, F., Perestrelo, A. R., Vinarsky, V., Pagliari, S. & Forte, G. Cellular Mechanotransduction: From Tension to Function. Front Physiol 9, 824, doi:10.3389/fphys.2018.00824 (2018). 59 Ohlendieck, K. et al. Duchenne muscular dystrophy: deficiency of dystrophin-associated proteins in the sarcolemma. Neurology 43, 795-800 (1993). 60 Maurer, M. & Lammerding, J. The Driving Force: Nuclear Mechanotransduction in Cellular Function, Fate, and Disease. Annu Rev Biomed Eng, doi:10.1146/annurev-bioeng- 060418-052139 (2019). 61 Fedorchak, G. R., Kaminski, A. & Lammerding, J. Cellular mechanosensing: getting to the nucleus of it all. Prog Biophys Mol Biol 115, 76-92, doi:10.1016/j.pbiomolbio.2014.06.009 (2014). 62 Linari, M. et al. Force generation by skeletal muscle is controlled by mechanosensing in myosin filaments. Nature 528, 276-279, doi:10.1038/nature15727 (2015). 63 Jani, K. & Schock, F. Molecular mechanisms of mechanosensing in muscle development. Dev Dyn 238, 1526-1534, doi:10.1002/dvdy.21972 (2009). 64 Dubinska-Magiera, M., Zaremba-Czogalla, M. & Rzepecki, R. Muscle development, regeneration and laminopathies: how lamins or lamina-associated proteins can contribute to muscle development, regeneration and disease. Cell Mol Life Sci 70, 2713-2741, doi:10.1007/s00018-012-1190-3 (2013). 65 Gonzalez-Aguilera, C. et al. Genome-wide analysis links emerin to neuromuscular junction activity in Caenorhabditis elegans. Genome Biol 15, R21, doi:10.1186/gb-2014- 15-2-r21 (2014). 66 Meaburn, K. J. et al. Primary laminopathy fibroblasts display altered genome organization and apoptosis. Aging Cell 6, 139-153, doi:10.1111/j.1474-9726.2007.00270.x (2007). 67 Broers, J. L. et al. 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, doi:10.1093/hmg/ddh295 (2004). 68 Bruusgaard, J. C., Liestol, K., Ekmark, M., Kollstad, K. & Gundersen, K. Number and spatial distribution of nuclei in the muscle fibres of normal mice studied in vivo. J Physiol 551, 467-478, doi:10.1113/jphysiol.2003.045328 (2003). 69 Cadot, B. et al. Nuclear movement during myotube formation is microtubule and dynein dependent and is regulated by Cdc42, Par6 and Par3. EMBO Rep 13, 741-749, doi:10.1038/embor.2012.89 (2012). 70 Metzger, T. et al. MAP and kinesin-dependent nuclear positioning is required for skeletal muscle function. Nature 484, 120-124, doi:10.1038/nature10914 (2012). 71 Sardo, L. et al. Real-time visualization of chromatin modification in isolated nuclei. J Cell Sci 130, 2926-2940, doi:10.1242/jcs.205823 (2017). 72 Guilluy, C. et al. Isolated nuclei adapt to force and reveal a mechanotransduction pathway in the nucleus. Nat Cell Biol 16, 376-381, doi:10.1038/ncb2927 (2014). 129 73 Swift, J. et al. Nuclear lamin-A scales with tissue stiffness and enhances matrix-directed differentiation. Science 341, 1240104, doi:10.1126/science.1240104 (2013). 74 Ihalainen, T. O. et al. 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, doi:10.1038/nmat4389 (2015). 75 Fisher, D. Z., Chaudhary, N. & Blobel, G. cDNA sequencing of nuclear lamins A and C reveals primary and secondary structural homology to intermediate filament proteins. Proc Natl Acad Sci U S A 83, 6450-6454 (1986). 76 Krohne, G. & Benavente, R. The nuclear lamins. A multigene family of proteins in evolution and differentiation. Exp Cell Res 162, 1-10 (1986). 77 Worman, H. J. & Courvalin, J. C. The nuclear lamina and inherited disease. Trends Cell Biol 12, 591-598 (2002). 78 Olave, I. A., Reck-Peterson, S. L. & Crabtree, G. R. Nuclear actin and actin-related proteins in chromatin remodeling. Annu Rev Biochem 71, 755-781, doi:10.1146/annurev.biochem.71.110601.135507 (2002). 79 Dopie, J., Skarp, K. P., Rajakyla, E. K., Tanhuanpaa, K. & Vartiainen, M. K. Active maintenance of nuclear actin by importin 9 supports transcription. Proc Natl Acad Sci U S A 109, E544-552, doi:10.1073/pnas.1118880109 (2012). 80 Goley, E. D. et al. Dynamic nuclear actin assembly by Arp2/3 complex and a baculovirus WASP-like protein. Science 314, 464-467, doi:10.1126/science.1133348 (2006). 81 Simon, D. N., Zastrow, M. S. & Wilson, K. L. Direct actin binding to A- and B-type lamin tails and actin filament bundling by the lamin A tail. Nucleus 1, 264-272, doi:10.4161/nucl.1.3.11799 (2010). 82 Chang, W., Folker, E. S., Worman, H. J. & Gundersen, G. G. Emerin organizes actin flow for nuclear movement and centrosome orientation in migrating fibroblasts. Mol Biol Cell 24, 3869-3880, doi:10.1091/mbc.E13-06-0307 (2013). 83 Haraguchi, T. et al. Live fluorescence imaging reveals early recruitment of emerin, LBR, RanBP2, and Nup153 to reforming functional nuclear envelopes. J Cell Sci 113 ( Pt 5), 779-794 (2000). 84 Dabauvalle, M. C. et al. Distribution of emerin during the cell cycle. Eur J Cell Biol 78, 749-756, doi:10.1016/S0171-9335(99)80043-0 (1999). 85 Dubinska-Magiera, M. et al. Emerin Is Required for Proper Nucleus Reassembly after Mitosis: Implications for New Pathogenetic Mechanisms for Laminopathies Detected in EDMD1 Patients. Cells 8, doi:10.3390/cells8030240 (2019). 86 Rowat, A. C., Lammerding, J. & Ipsen, J. H. Mechanical properties of the cell nucleus and the effect of emerin deficiency. Biophys J 91, 4649-4664, doi:10.1529/biophysj.106.086454 (2006). 87 Ognibene, A. et al. Nuclear changes in a case of X-linked Emery-Dreifuss muscular dystrophy. Muscle Nerve 22, 864-869, doi:10.1002/(sici)1097- 4598(199907)22:7<864::aid-mus8>3.0.co;2-g (1999). 88 Fidzianska, A., Toniolo, D. & Hausmanowa-Petrusewicz, I. Ultrastructural abnormality of sarcolemmal nuclei in Emery-Dreifuss muscular dystrophy (EDMD). J Neurol Sci 159, 88- 93, doi:10.1016/s0022-510x(98)00130-0 (1998). 89 Khatau, S. B. et al. A perinuclear actin cap regulates nuclear shape. Proc Natl Acad Sci U S A 106, 19017-19022, doi:10.1073/pnas.0908686106 (2009). 130 90 Sewry, C. A. et al. Skeletal muscle pathology in autosomal dominant Emery-Dreifuss muscular dystrophy with lamin A/C mutations. Neuropathol Appl Neurobiol 27, 281-290 (2001). 91 Capers, C. R. Multinucleation of skeletal muscle in vitro. J Biophys Biochem Cytol 7, 559- 566 (1960). 92 Haraguchi, T. et al. Emerin binding to Btf, a death-promoting transcriptional repressor, is disrupted by a missense mutation that causes Emery-Dreifuss muscular dystrophy. Eur J Biochem 271, 1035-1045 (2004). 93 Burke, B. Chain reaction: LINC complexes and nuclear positioning. F1000Res 8, doi:10.12688/f1000research.16877.1 (2019). 94 Chambliss, A. B. et al. The LINC-anchored actin cap connects the extracellular milieu to the nucleus for ultrafast mechanotransduction. Sci Rep 3, 1087, doi:10.1038/srep01087 (2013). 95 Dahl, K. N. et al. Distinct structural and mechanical properties of the nuclear lamina in Hutchinson-Gilford progeria syndrome. Proc Natl Acad Sci U S A 103, 10271-10276, doi:10.1073/pnas.0601058103 (2006). 96 Cartegni, L. et al. Heart-specific localization of emerin: new insights into Emery-Dreifuss muscular dystrophy. Hum Mol Genet 6, 2257-2264 (1997). 97 Essawy, N. et al. An Emerin LEM-Domain Mutation Impairs Cell Response to Mechanical Stress. Cells 8, doi:10.3390/cells8060570 (2019). 98 Bautista, M., Fernandez, A. & Pinaud, F. A Micropatterning Strategy to Study Nuclear Mechanotransduction in Cells. Micromachines (Basel) 10, doi:10.3390/mi10120810 (2019). 99 Shimi, T. et al. Dynamic interaction between BAF and emerin revealed by FRAP, FLIP, and FRET analyses in living HeLa cells. J Struct Biol 147, 31-41, doi:10.1016/j.jsb.2003.11.013 (2004). 100 Samson, C. et al. Emerin self-assembly mechanism: role of the LEM domain. FEBS J 284, 338-352, doi:10.1111/febs.13983 (2017). 101 Bengtsson, L. & Wilson, K. L. Multiple and surprising new functions for emerin, a nuclear membrane protein. Curr Opin Cell Biol 16, 73-79, doi:10.1016/j.ceb.2003.11.012 (2004). 102 Sahl, S. J. & Moerner, W. E. Super-resolution fluorescence imaging with single molecules. Curr Opin Struct Biol 23, 778-787, doi:10.1016/j.sbi.2013.07.010 (2013). 103 Shen, H. et al. Single Particle Tracking: From Theory to Biophysical Applications. Chem Rev 117, 7331-7376, doi:10.1021/acs.chemrev.6b00815 (2017). 104 Sengupta, P. et al. Probing protein heterogeneity in the plasma membrane using PALM and pair correlation analysis. Nat Methods 8, 969-975, doi:10.1038/nmeth.1704 (2011). 105 Born, M. & Wolf, E. Principles of Optics: Electromagnetic Theory of Propagation, Interference and Diffraction of Light. 7 edn, (Cambridge University Press, 1999). 106 Sheppard, C. J. R. Resolution and super-resolution. Microsc Res Tech 80, 590-598, doi:10.1002/jemt.22834 (2017). 107 Tokunaga, M., Imamoto, N. & Sakata-Sogawa, K. Highly inclined thin illumination enables clear single-molecule imaging in cells. Nat Methods 5, 159-161, doi:10.1038/nmeth1171 (2008). 108 Hess, S. T., Girirajan, T. P. & Mason, M. D. Ultra-high resolution imaging by fluorescence photoactivation localization microscopy. Biophys J 91, 4258-4272, doi:10.1529/biophysj.106.091116 (2006). 131 109 Manley, S. et al. High-density mapping of single-molecule trajectories with photoactivated localization microscopy. Nat Methods 5, 155-157, doi:10.1038/nmeth.1176 (2008). 110 Rust, M. J., Bates, M. & Zhuang, X. Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM). Nat Methods 3, 793-795, doi:10.1038/nmeth929 (2006). 111 Endesfelder, U. & Heilemann, M. Direct stochastic optical reconstruction microscopy (dSTORM). Methods Mol Biol 1251, 263-276, doi:10.1007/978-1-4939-2080-8_14 (2015). 112 Henriques, R. & Mhlanga, M. M. PALM and STORM: what hides beyond the Rayleigh limit? Biotechnol J 4, 846-857, doi:10.1002/biot.200900024 (2009). 113 Thompson, R. E., Larson, D. R. & Webb, W. W. Precise nanometer localization analysis for individual fluorescent probes. Biophys J 82, 2775-2783, doi:10.1016/S0006- 3495(02)75618-X (2002). 114 Yildiz, A. et al. Myosin V walks hand-over-hand: single fluorophore imaging with 1.5-nm localization. Science 300, 2061-2065, doi:10.1126/science.1084398 (2003). 115 Funatsu, T., Harada, Y., Tokunaga, M., Saito, K. & Yanagida, T. Imaging of single fluorescent molecules and individual ATP turnovers by single myosin molecules in aqueous solution. Nature 374, 555-559, doi:10.1038/374555a0 (1995). 116 Sase, I., Miyata, H., Corrie, J. E., Craik, J. S. & Kinosita, K., Jr. Real time imaging of single fluorophores on moving actin with an epifluorescence microscope. Biophys J 69, 323-328, doi:10.1016/S0006-3495(95)79937-4 (1995). 117 Voie, A. H., Burns, D. H. & Spelman, F. A. Orthogonal-plane fluorescence optical sectioning: three-dimensional imaging of macroscopic biological specimens. J Microsc 170, 229-236, doi:10.1111/j.1365-2818.1993.tb03346.x (1993). 118 Power, R. M. & Huisken, J. A guide to light-sheet fluorescence microscopy for multiscale imaging. Nat Methods 14, 360-373, doi:10.1038/nmeth.4224 (2017). 119 Huisken, J., Swoger, J., Del Bene, F., Wittbrodt, J. & Stelzer, E. H. Optical sectioning deep inside live embryos by selective plane illumination microscopy. Science 305, 1007-1009, doi:10.1126/science.1100035 (2004). 120 Feng, H., Wang, X., Xu, Z., Zhang, X. & Gao, Y. Super-Resolution Fluorescence Microscopy for Single Cell Imaging. Adv Exp Med Biol 1068, 59-71, doi:10.1007/978- 981-13-0502-3_6 (2018). 121 Gustafsson, M. G. Surpassing the lateral resolution limit by a factor of two using structured illumination microscopy. J Microsc 198, 82-87, doi:10.1046/j.1365-2818.2000.00710.x (2000). 122 Shimi, T. et al. Structural organization of nuclear lamins A, C, B1, and B2 revealed by superresolution microscopy. Mol Biol Cell 26, 4075-4086, doi:10.1091/mbc.E15-07-0461 (2015). 123 Nicovich, P. R., Owen, D. M. & Gaus, K. Turning single-molecule localization microscopy into a quantitative bioanalytical tool. Nat Protoc 12, 453-460, doi:10.1038/nprot.2016.166 (2017). 124 Vogelsang, J. et al. A reducing and oxidizing system minimizes photobleaching and blinking of fluorescent dyes. Angew Chem Int Ed Engl 47, 5465-5469, doi:10.1002/anie.200801518 (2008). 132 125 Dempsey, G. T., Vaughan, J. C., Chen, K. H., Bates, M. & Zhuang, X. Evaluation of fluorophores for optimal performance in localization-based super-resolution imaging. Nat Methods 8, 1027-1036, doi:10.1038/nmeth.1768 (2011). 126 Nahidiazar, L., Agronskaia, A. V., Broertjes, J., van den Broek, B. & Jalink, K. Optimizing Imaging Conditions for Demanding Multi-Color Super Resolution Localization Microscopy. PLoS One 11, e0158884, doi:10.1371/journal.pone.0158884 (2016). 127 Xu, J., Ma, H. & Liu, Y. Stochastic Optical Reconstruction Microscopy (STORM). Curr Protoc Cytom 81, 12 46 11-12 46 27, doi:10.1002/cpcy.23 (2017). 128 Dempsey, G. T. et al. Photoswitching mechanism of cyanine dyes. J Am Chem Soc 131, 18192-18193, doi:10.1021/ja904588g (2009). 129 Jensen, E. & Crossman, D. J. Technical review: types of imaging-direct STORM. Anat Rec (Hoboken) 297, 2227-2231, doi:10.1002/ar.22960 (2014). 130 Wang, S., Moffitt, J. R., Dempsey, G. T., Xie, X. S. & Zhuang, X. Characterization and development of photoactivatable fluorescent proteins for single-molecule-based superresolution imaging. Proc Natl Acad Sci U S A 111, 8452-8457, doi:10.1073/pnas.1406593111 (2014). 131 Nieuwenhuizen, R. P. et al. Quantitative localization microscopy: effects of photophysics and labeling stoichiometry. PLoS One 10, e0127989, doi:10.1371/journal.pone.0127989 (2015). 132 Anderson, C. M., Georgiou, G. N., Morrison, I. E., Stevenson, G. V. & Cherry, R. J. Tracking of cell surface receptors by fluorescence digital imaging microscopy using a charge-coupled device camera. Low-density lipoprotein and influenza virus receptor mobility at 4 degrees C. J Cell Sci 101 ( Pt 2), 415-425 (1992). 133 Ghosh, R. N. & Webb, W. W. Automated detection and tracking of individual and clustered cell surface low density lipoprotein receptor molecules. Biophys J 66, 1301-1318, doi:10.1016/S0006-3495(94)80939-7 (1994). 134 Kusumi, A., Sako, Y. & Yamamoto, M. Confined lateral diffusion of membrane receptors as studied by single particle tracking (nanovid microscopy). Effects of calcium-induced differentiation in cultured epithelial cells. Biophys J 65, 2021-2040, doi:10.1016/S0006- 3495(93)81253-0 (1993). 135 Schutz, G. J., Schindler, H. & Schmidt, T. Single-molecule microscopy on model membranes reveals anomalous diffusion. Biophys J 73, 1073-1080, doi:10.1016/S0006- 3495(97)78139-6 (1997). 136 Hess, S. T. et al. Dynamic clustered distribution of hemagglutinin resolved at 40 nm in living cell membranes discriminates between raft theories. Proc Natl Acad Sci U S A 104, 17370-17375, doi:10.1073/pnas.0708066104 (2007). 137 Saxton, M. J. & Jacobson, K. Single-particle tracking: applications to membrane dynamics. Annu Rev Biophys Biomol Struct 26, 373-399, doi:10.1146/annurev.biophys.26.1.373 (1997). 138 Hink, M. A. Fluorescence correlation spectroscopy. Methods Mol Biol 1251, 135-150, doi:10.1007/978-1-4939-2080-8_8 (2015). 139 Digman, M. A. & Gratton, E. Analysis of diffusion and binding in cells using the RICS approach. Microsc Res Tech 72, 323-332, doi:10.1002/jemt.20655 (2009). 140 Slaughter, B. D., Unruh, J. R. & Li, R. Fluorescence fluctuation spectroscopy and imaging methods for examination of dynamic protein interactions in yeast. Methods Mol Biol 759, 283-306, doi:10.1007/978-1-61779-173-4_17 (2011). 133 141 Struntz, P. & Weiss, M. The hitchhiker's guide to quantitative diffusion measurements. Phys Chem Chem Phys 20, 28910-28919, doi:10.1039/c8cp06158k (2018). 142 Pinaud, F., Clarke, S., Sittner, A. & Dahan, M. Probing cellular events, one quantum dot at a time. Nat Methods 7, 275-285, doi:10.1038/nmeth.1444 (2010). 143 Wang, Y., Jeong, Y., Jhiang, S. M., Yu, L. & Menq, C. H. Quantitative characterization of cell behaviors through cell cycle progression via automated cell tracking. PLoS One 9, e98762, doi:10.1371/journal.pone.0098762 (2014). 144 Lidke, D. S. et al. Quantum dot ligands provide new insights into erbB/HER receptor- mediated signal transduction. Nat Biotechnol 22, 198-203, doi:10.1038/nbt929 (2004). 145 Lippincott-Schwartz, J. & Patterson, G. H. Photoactivatable fluorescent proteins for diffraction-limited and super-resolution imaging. Trends Cell Biol 19, 555-565, doi:10.1016/j.tcb.2009.09.003 (2009). 146 Michalet, X. Mean square displacement analysis of single-particle trajectories with localization error: Brownian motion in an isotropic medium. Phys Rev E Stat Nonlin Soft Matter Phys 82, 041914, doi:10.1103/PhysRevE.82.041914 (2010). 147 Kiskowski, M. A., Hancock, J. F. & Kenworthy, A. K. On the use of Ripley's K-function and its derivatives to analyze domain size. Biophys J 97, 1095-1103, doi:10.1016/j.bpj.2009.05.039 (2009). 148 Owen, D. M. et al. PALM imaging and cluster analysis of protein heterogeneity at the cell surface. J Biophotonics 3, 446-454, doi:10.1002/jbio.200900089 (2010). 149 Shivanandan, A., Unnikrishnan, J. & Radenovic, A. On characterizing protein spatial clusters with correlation approaches. Sci Rep-Uk 6, doi:10.1038/srep31164 (2016). 150 Veatch, S. L. et al. Correlation functions quantify super-resolution images and estimate apparent clustering due to over-counting. PLoS One 7, e31457, doi:10.1371/journal.pone.0031457 (2012). 151 Balint, S., Lopes, F. B. & Davis, D. M. A nanoscale reorganization of the IL-15 receptor is triggered by NKG2D in a ligand-dependent manner. Sci Signal 11, doi:10.1126/scisignal.aal3606 (2018). 152 Wiegand, T. & A. Moloney, K. Rings, circles, and null-models for point pattern analysis in ecology. Oikos 104, 209-229, doi:10.1111/j.0030-1299.2004.12497.x (2004). 153 Prodanov, D., Nagelkerke, N. & Marani, E. Spatial clustering analysis in neuroanatomy: applications of different approaches to motor nerve fiber distribution. J Neurosci Methods 160, 93-108, doi:10.1016/j.jneumeth.2006.08.017 (2007). 154 Ripley, B. D. The second-order analysis of stationary point processes. Journal of Applied Probability 13, 255-266, doi:10.2307/3212829 (1976). 155 Getis, A. & Franklin, J. Second-Order Neighborhood Analysis of Mapped Point Patterns. Ecology 68, 473-477, doi:doi:10.2307/1938452 (1987). 156 Hoshen, J. & Kopelman, R. Percolation and cluster distribution. I. Cluster multiple labeling technique and critical concentration algorithm. Physical Review B 14, 3438-3445, doi:10.1103/PhysRevB.14.3438 (1976). 157 Clark, P. J. & Evans, F. C. Distance to Nearest Neighbor as a Measure of Spatial Relationships in Populations. Ecology 35, 445-453, doi:Doi 10.2307/1931034 (1954). 158 Haase, P. Spatial pattern analysis in ecology based on Ripley's K-function: Introduction and methods of edge correction. Journal of Vegetation Science 6, 575-582, doi:doi:10.2307/3236356 (1995). 134 159 Sengupta, P., Jovanovic-Talisman, T. & Lippincott-Schwartz, J. Quantifying spatial organization in point-localization superresolution images using pair correlation analysis. Nat Protoc 8, 345-354, doi:10.1038/nprot.2013.005 (2013). 160 Martinez-Rivas, A., Gonzalez-Quijano, G. K., Proa-Coronado, S., Severac, C. & Dague, E. Methods of Micropatterning and Manipulation of Cells for Biomedical Applications. Micromachines (Basel) 8, doi:10.3390/mi8120347 (2017). 161 Thery, M. & Piel, M. Adhesive micropatterns for cells: a microcontact printing protocol. Cold Spring Harb Protoc 2009, pdb prot5255, doi:10.1101/pdb.prot5255 (2009). 162 Azioune, A., Carpi, N., Tseng, Q., Thery, M. & Piel, M. Protein Micropatterns: A Direct Printing Protocol Using Deep UVs. Method Cell Biol 97, 133-146, doi:10.1016/S0091- 679x(10)97008-8 (2010). 163 Tu, C. L. et al. A Microfluidic Chip for Cell Patterning Utilizing Paired Microwells and Protein Patterns. Micromachines-Basel 8, doi:UNSP 1 10.3390/mi8010001 (2017). 164 Sahni, G., Yuan, J. & Toh, Y. C. Stencil Micropatterning of Human Pluripotent Stem Cells for Probing Spatial Organization of Differentiation Fates. J Vis Exp, doi:10.3791/54097 (2016). 165 Belaadi, N., Aureille, J. & Guilluy, C. Under Pressure: Mechanical Stress Management in the Nucleus. Cells 5, doi:10.3390/cells5020027 (2016). 166 Cho, S., Irianto, J. & Discher, D. E. Mechanosensing by the nucleus: From pathways to scaling relationships. J Cell Biol 216, 305-315, doi:10.1083/jcb.201610042 (2017). 167 Maninova, M., Caslavsky, J. & Vomastek, T. The assembly and function of perinuclear actin cap in migrating cells. Protoplasma 254, 1207-1218, doi:10.1007/s00709-017-1077- 0 (2017). 168 Holaska, J. M. & Wilson, K. L. An emerin "proteome": purification of distinct emerin- containing complexes from HeLa cells suggests molecular basis for diverse roles including gene regulation, mRNA splicing, signaling, mechanosensing, and nuclear architecture. Biochemistry 46, 8897-8908, doi:10.1021/bi602636m (2007). 169 Folker, E. S. & Baylies, M. K. Nuclear positioning in muscle development and disease. Front Physiol 4, 363, doi:10.3389/fphys.2013.00363 (2013). 170 Meinke, P., Nguyen, T. D. & Wehnert, M. S. The LINC complex and human disease. Biochem Soc Trans 39, 1693-1697, doi:10.1042/BST20110658 (2011). 171 Ostlund, C. & Worman, H. J. Nuclear envelope proteins and neuromuscular diseases. Muscle Nerve 27, 393-406, doi:10.1002/mus.10302 (2003). 172 Chapman, M. A. et al. Disruption of both nesprin 1 and desmin results in nuclear anchorage defects and fibrosis in skeletal muscle. Hum Mol Genet 23, 5879-5892, doi:10.1093/hmg/ddu310 (2014). 173 Le, H. Q. et al. Mechanical regulation of transcription controls Polycomb-mediated gene silencing during lineage commitment. Nat Cell Biol 18, 864-875, doi:10.1038/ncb3387 (2016). 174 Tajik, A. et al. Transcription upregulation via force-induced direct stretching of chromatin. Nat Mater 15, 1287-1296, doi:10.1038/nmat4729 (2016). 175 Alves, N. M., Pashkuleva, I., Reis, R. L. & Mano, J. F. Controlling cell behavior through the design of polymer surfaces. Small 6, 2208-2220, doi:10.1002/smll.201000233 (2010). 135 176 Faid, K. et al. Rapid fabrication and chemical patterning of polymer microstructures and their applications as a platform for cell cultures. Biomed Microdevices 7, 179-184, doi:10.1007/s10544-005-3023-8 (2005). 177 Mata, A., Fleischman, A. J. & Roy, S. Characterization of polydimethylsiloxane (PDMS) properties for biomedical micro/nanosystems. Biomed Microdevices 7, 281-293, doi:10.1007/s10544-005-6070-2 (2005). 178 Nakanishi, J. Dynamic photochemical silane micropatterning. Methods Cell Biol 120, 117- 129, doi:10.1016/B978-0-12-417136-7.00007-0 (2014). 179 Wright, D., Rajalingam, B., Selvarasah, S., Dokmeci, M. R. & Khademhosseini, A. Generation of static and dynamic patterned co-cultures using microfabricated parylene-C stencils. Lab Chip 7, 1272-1279, doi:10.1039/b706081e (2007). 180 Li, N. & Ho, C. M. Photolithographic patterning of organosilane monolayer for generating large area two-dimensional B lymphocyte arrays. Lab Chip 8, 2105-2112, doi:10.1039/b810329a (2008). 181 Mujawar, L. H., Norde, W. & van Amerongen, A. Spot morphology of non-contact printed protein molecules on non-porous substrates with a range of hydrophobicities. Analyst 138, 518-524, doi:10.1039/c2an36104c (2013). 182 Kawai, A. & Kawakami, J. Wetting Analysis of Hydrophobic Substrate treated by HMDS Primer. Journal of Photopolymer Science and Technology 20, 815-816, doi:10.2494/photopolymer.20.815 (2007). 183 Versaevel, M., Grevesse, T. & Gabriele, S. Spatial coordination between cell and nuclear shape within micropatterned endothelial cells. Nat Commun 3, 671, doi:10.1038/ncomms1668 (2012). 184 Li, Q., Kumar, A., Makhija, E. & Shivashankar, G. V. The regulation of dynamic mechanical coupling between actin cytoskeleton and nucleus by matrix geometry. Biomaterials 35, 961-969, doi:10.1016/j.biomaterials.2013.10.037 (2014). 185 Pol, A. et al. Cholesterol and fatty acids regulate dynamic caveolin trafficking through the Golgi complex and between the cell surface and lipid bodies. Mol Biol Cell 16, 2091-2105, doi:10.1091/mbc.e04-08-0737 (2005). 186 Scherer, P. E. et al. Cell-type and tissue-specific expression of caveolin-2. Caveolins 1 and 2 co-localize and form a stable hetero-oligomeric complex in vivo. J Biol Chem 272, 29337-29346, doi:10.1074/jbc.272.46.29337 (1997). 187 Koschwanez, J. H., Carlson, R. H. & Meldrum, D. R. Thin PDMS films using long spin times or tert-butyl alcohol as a solvent. PLoS One 4, e4572, doi:10.1371/journal.pone.0004572 (2009). 188 Jacob, M. C., Favre, M. & Bensa, J. C. Membrane cell permeabilization with saponin and multiparametric analysis by flow cytometry. Cytometry 12, 550-558, doi:10.1002/cyto.990120612 (1991). 189 Luxton, G. W., Gomes, E. R., Folker, E. S., Vintinner, E. & Gundersen, G. G. Linear arrays of nuclear envelope proteins harness retrograde actin flow for nuclear movement. Science 329, 956-959, doi:10.1126/science.1189072 (2010). 190 Ostlund, C., Sullivan, T., Stewart, C. L. & Worman, H. J. Dependence of diffusional mobility of integral inner nuclear membrane proteins on A-type lamins. Biochemistry 45, 1374-1382, doi:10.1021/bi052156n (2006). 191 Versaevel, M. et al. Super-resolution microscopy reveals LINC complex recruitment at nuclear indentation sites. Sci Rep 4, 7362, doi:10.1038/srep07362 (2014). 136 192 Huang, B., Wang, W., Bates, M. & Zhuang, X. Three-dimensional super-resolution imaging by stochastic optical reconstruction microscopy. Science 319, 810-813, doi:10.1126/science.1153529 (2008). 193 L. W. Perry, G., Miller, B. & Enright, N. A comparison of methods for the statistical analysis of spatial point patterns in plant ecology. Vol. 187 (2006). 194 Knies, Y., Bernd, A., Kaufmann, R., Bereiter-Hahn, J. & Kippenberger, S. Mechanical stretch induces clustering of beta1-integrins and facilitates adhesion. Exp Dermatol 15, 347-355, doi:10.1111/j.0906-6705.2006.00422.x (2006). 195 Iwamoto, D. V. & Calderwood, D. A. Regulation of integrin-mediated adhesions. Curr Opin Cell Biol 36, 41-47, doi:10.1016/j.ceb.2015.06.009 (2015). 196 Phair, R. D., Gorski, S. A. & Misteli, T. Measurement of dynamic protein binding to chromatin in vivo, using photobleaching microscopy. Methods Enzymol 375, 393-414 (2004). 197 Soumpasis, D. M. Theoretical analysis of fluorescence photobleaching recovery experiments. Biophys J 41, 95-97, doi:10.1016/S0006-3495(83)84410-5 (1983). 198 Serge, A., Bertaux, N., Rigneault, H. & Marguet, D. Dynamic multiple-target tracing to probe spatiotemporal cartography of cell membranes. Nat Methods 5, 687-694, doi:10.1038/nmeth.1233 (2008). 199 Fernandez, A., Bautista, M., Stanciauskas, R., Chung, T. & Pinaud, F. Cell-Shaping Micropatterns for Quantitative Super-Resolution Microscopy Imaging of Membrane Mechanosensing Proteins. ACS Appl Mater Interfaces 9, 27575-27586, doi:10.1021/acsami.7b09743 (2017). 200 Wolter, S. et al. rapidSTORM: accurate, fast open-source software for localization microscopy. Nat Methods 9, 1040-1041, doi:10.1038/nmeth.2224 (2012). 201 Pengo, T., Holden, S. J. & Manley, S. PALMsiever: a tool to turn raw data into results for single-molecule localization microscopy. Bioinformatics 31, 797-798, doi:10.1093/bioinformatics/btu720 (2015). 202 Condit, R. et al. Spatial patterns in the distribution of tropical tree species. Science 288, 1414-1418 (2000). 203 L.W. Perry, G. SpPack: Spatial point pattern analysis in Excel using Visual Basic for Applications (VBA). Vol. 19 (2004). 204 Kirby, T. J. & Lammerding, J. Emerging views of the nucleus as a cellular mechanosensor. Nat. Cell Biol. 20, 373-381, doi:10.1038/s41556-018-0038-y (2018). 205 Cui, Y. et al. Cyclic stretching of soft substrates induces spreading and growth. Nat. Commun. 6, 6333, doi:10.1038/ncomms7333 (2015). 206 Demmerle, J., Koch, A. J. & Holaska, J. M. The nuclear envelope protein emerin binds directly to histone deacetylase 3 (HDAC3) and activates HDAC3 activity. The Journal of Biological Chemistry 287, 22080-22088, doi:10.1074/jbc.M111.325308 (2012). 207 Stephens, A. D., Banigan, E. J., Adam, S. A., Goldman, R. D. & Marko, J. F. Chromatin and lamin A determine two different mechanical response regimes of the cell nucleus. Mol. Biol. Cell 28, 1984-1996, doi:10.1091/mbc.E16-09-0653 (2017). 208 Maurer, M. & Lammerding, J. The Driving Force: Nuclear Mechanotransduction in Cellular Function, Fate, and Disease. Annu. Rev. Biomed. Eng. 21, 443-468, doi:10.1146/annurev-bioeng-060418-052139 (2019). 209 Emery, A. E. & Dreifuss, F. E. Unusual type of benign x-linked muscular dystrophy. J Neurol Neurosurg Psychiatry 29, 338-342 (1966). 137 210 Maggi, L., Carboni, N. & Bernasconi, P. Skeletal Muscle Laminopathies: A Review of Clinical and Molecular Features. Cells 5, doi:10.3390/cells5030033 (2016). 211 Frock, R. L. et al. Lamin A/C and emerin are critical for skeletal muscle satellite cell differentiation. Genes Dev. 20, 486-500, doi:10.1101/gad.1364906 (2006). 212 Fidzianska, A. & Hausmanowa-Petrusewicz, I. Architectural abnormalities in muscle nuclei. Ultrastructural differences between X-linked and autosomal dominant forms of EDMD. J. Neurol. Sci. 210, 47-51 (2003). 213 Park, Y.-E. et al. Nuclear changes in skeletal muscle extend to satellite cells in autosomal dominant Emery-Dreifuss muscular dystrophy/limb-girdle muscular dystrophy 1B. Neuromuscul. Disord. 19, 29-36, doi:10.1016/j.nmd.2008.09.018 (2009). 214 Rowat, A. C., Lammerding, J. & Ipsen, J. H. Mechanical properties of the cell nucleus and the effect of emerin deficiency. Biophys. J. 91, 4649-4664, doi:10.1529/biophysj.106.086454 (2006). 215 Yuan, J. & Xue, B. Role of structural flexibility in the evolution of emerin. J Theor Biol 385, 102-111, doi:10.1016/j.jtbi.2015.08.009 (2015). 216 Sakaki, M. et al. Interaction between emerin and nuclear lamins. J Biochem 129, 321-327 (2001). 217 Lee, K. K. et al. Distinct functional domains in emerin bind lamin A and DNA-bridging protein BAF. J Cell Sci 114, 4567-4573 (2001). 218 Mislow, J. M. K. et al. Nesprin-1α self-associates and binds directly to emerin and lamin A in vitro. FEBS Lett. 525, 135-140, doi:10.1016/S0014-5793(02)03105-8 (2002). 219 Haque, F. et al. Mammalian SUN protein interaction networks at the inner nuclear membrane and their role in laminopathy disease processes. J. Biol. Chem. 285, 3487-3498, doi:10.1074/jbc.M109.071910 (2010). 220 Roberts, R. C. et al. The Emery-Dreifuss muscular dystrophy associated-protein emerin is phosphorylated on serine 49 by protein kinase A. FEBS J. 273, 4562-4575, doi:10.1111/j.1742-4658.2006.05464.x (2006). 221 Berk, J. M. et al. O-Linked beta-N-acetylglucosamine (O-GlcNAc) regulates emerin binding to barrier to autointegration factor (BAF) in a chromatin- and lamin B-enriched "niche". J Biol Chem 288, 30192-30209, doi:10.1074/jbc.M113.503060 (2013). 222 Guilluy, C. et al. Isolated nuclei adapt to force and reveal a mechanotransduction pathway in the nucleus. Nat. Cell Biol. 16, 376-381, doi:10.1038/ncb2927 (2014). 223 Hirano, Y. et al. Dissociation of Emerin from Barrier-to-autointegration Factor Is Regulated through Mitotic Phosphorylation of Emerin in a Xenopus Egg Cell-free System*. J. Biol. Chem. 280, 39925-39933, doi:10.1074/jbc.M503214200 (2005). 224 Yates, J. R. & Wehnert, M. The Emery-Dreifuss Muscular Dystrophy Mutation Database. Neuromuscul Disord 9, 199 (1999). 225 Ellis, J. A., Craxton, M., Yates, J. R. & Kendrick-Jones, J. Aberrant intracellular targeting and cell cycle-dependent phosphorylation of emerin contribute to the Emery-Dreifuss muscular dystrophy phenotype. J. Cell Sci. 111 ( Pt 6), 781-792 (1998). 226 Holt, I., Clements, L., Manilal, S. & Morris, G. E. How does a g993t mutation in the emerin gene cause Emery-Dreifuss muscular dystrophy? Biochemical and Biophysical Research Communications 287, 1129-1133, doi:10.1006/bbrc.2001.5708 (2001). 227 Fernandez, A., Bautista, M., Stanciauskas, R., Chung, T. & Pinaud, F. Cell-Shaping Micropatterns for Quantitative Super-Resolution Microscopy Imaging of Membrane 138 Mechanosensing Proteins. ACS Appl Mater Interfaces 9, 27575-27586, doi:10.1021/acsami.7b09743 (2017). 228 Talkop, U. A. et al. Early onset of cardiomyopathy in two brothers with X-linked Emery- Dreifuss muscular dystrophy. Neuromuscul. Disord. 12, 878-881, doi:10.1016/s0960- 8966(02)00134-7 (2002). 229 Schutz, G. J., Schindler, H. & Schmidt, T. Single-molecule microscopy on model membranes reveals anomalous diffusion. Biophys. J. 73, 1073-1080, doi:10.1016/S0006- 3495(97)78139-6 (1997). 230 Cabantous, S., Terwilliger, T. C. & Waldo, G. S. Protein tagging and detection with engineered self-assembling fragments of green fluorescent protein. Nat. Biotechnol. 23, 102-107 (2005). 231 Pinaud, F. & Dahan, M. Targeting and imaging single biomolecules in living cells by complementation-activated light microscopy with split-fluorescent proteins. Proceedings of the National Academy of Sciences 108, E201-E210, doi:10.1073/pnas.1101929108 (2011). 232 Iino, R., Koyama, I. & Kusumi, A. Single Molecule Imaging of Green Fluorescent Proteins in Living Cells: E-Cadherin Forms Oligomers on the Free Cell Surface. Biophys. J. 80, 2667-2677, doi:10.1016/S0006-3495(01)76236-4 (2001). 233 Burger, M., Schmitt-Koopmann, C. & Leroux, J.-C. DNA unchained: two assays to discover and study inhibitors of the DNA clustering function of barrier-to-autointegration factor. Sci. Rep. 10, 12301, doi:10.1038/s41598-020-69246-x (2020). 234 Shimi, T. et al. Dynamic interaction between BAF and emerin revealed by FRAP, FLIP, and FRET analyses in living HeLa cells. J. Struct. Biol. 147, 31-41, doi:10.1016/j.jsb.2003.11.013 (2004). 235 Samwer, M. et al. DNA Cross-Bridging Shapes a Single Nucleus from a Set of Mitotic Chromosomes. Cell 170, 956-972 e923, doi:10.1016/j.cell.2017.07.038 (2017). 236 Sengupta, P. et al. Probing protein heterogeneity in the plasma membrane using PALM and pair correlation analysis. Nat. Methods 8, 969-975, doi:10.1038/nmeth.1704 (2011). 237 Crisp, M. et al. Coupling of the nucleus and cytoplasm: role of the LINC complex. J Cell Biol 172, 41-53, doi:10.1083/jcb.200509124 (2006). 238 Bone, C. R., Tapley, E. C., Gorjanacz, M. & Starr, D. A. The Caenorhabditis elegans SUN protein UNC-84 interacts with lamin to transfer forces from the cytoplasm to the nucleoskeleton during nuclear migration. Mol Biol Cell 25, 2853-2865, doi:10.1091/mbc.E14-05-0971 (2014). 239 Padmakumar, V. C. et al. The inner nuclear membrane protein Sun1 mediates the anchorage of Nesprin-2 to the nuclear envelope. J Cell Sci 118, 3419-3430, doi:10.1242/jcs.02471 (2005). 240 Lei, K. et al. SUN1 and SUN2 play critical but partially redundant roles in anchoring nuclei in skeletal muscle cells in mice. Proc Natl Acad Sci U S A 106, 10207-10212, doi:10.1073/pnas.0812037106 (2009). 241 Lu, W. et al. Sun1 forms immobile macromolecular assemblies at the nuclear envelope. Biochim Biophys Acta 1783, 2415-2426, doi:10.1016/j.bbamcr.2008.09.001 (2008). 242 Hennen, J., Saunders, C. A., Mueller, J. D. & Luxton, G. W. G. Fluorescence fluctuation spectroscopy reveals differential SUN protein oligomerization in living cells. Mol Biol Cell 29, 1003-1011, doi:10.1091/mbc.E17-04-0233 (2018). 139 243 Bengtsson, L. & Wilson, K. L. Multiple and surprising new functions for emerin, a nuclear membrane protein. Curr. Opin. Cell Biol. 16, 73-79, doi:10.1016/j.ceb.2003.11.012 (2004). 244 Ho, C. Y., Jaalouk, D. E., Vartiainen, M. K. & Lammerding, J. Lamin A/C and emerin regulate MKL1–SRF activity by modulating actin dynamics. Nature 497, 507-511, doi:10.1038/nature12105 (2013). 245 Tifft, K. E., Bradbury, K. A. & Wilson, K. L. Tyrosine phosphorylation of nuclear- membrane protein emerin by Src, Abl and other kinases. J Cell Sci 122, 3780-3790, doi:10.1242/jcs.048397 (2009). 246 Schreiner, S. M., Koo, P. K., Zhao, Y., Mochrie, S. G. J. & King, M. C. The tethering of chromatin to the nuclear envelope supports nuclear mechanics. Nat. Commun. 6, 7159, doi:10.1038/ncomms8159 (2015). 247 Stoynova, L., Solórzano, R. & Collins, E. D. Generation of large deletion mutants from plasmid DNA. BioTechniques 36, 402-404, 406, doi:10.2144/04363bm05 (2004). 248 Schneider, C. A., Rasband, W. S. & Eliceiri, K. W. NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 9, 671-675 (2012). 249 Perry, G. L. W. SpPack: spatial point pattern analysis in Excel using Visual Basic for Applications (VBA). Environ. Model. Software 19, 559-569, doi:10.1016/j.envsoft.2003.07.004 (2004). 250 Ho, C. Y., Jaalouk, D. E. & Lammerding, J. Novel insights into the disease etiology of laminopathies. Rare Dis 1, e27002, doi:10.4161/rdis.27002 (2013). 251 Hah, J. & Kim, D. H. Deciphering Nuclear Mechanobiology in Laminopathy. Cells 8, doi:10.3390/cells8030231 (2019). 252 Roux, K. J. & Burke, B. Nuclear envelope defects in muscular dystrophy. Biochim Biophys Acta 1772, 118-127, doi:10.1016/j.bbadis.2006.06.001 (2007). 253 Lee, Y. L. & Burke, B. LINC complexes and nuclear positioning. Semin Cell Dev Biol 82, 67-76, doi:10.1016/j.semcdb.2017.11.008 (2018). 254 Frock, R. L. et al. Lamin A/C and emerin are critical for skeletal muscle satellite cell differentiation. Genes Dev 20, 486-500, doi:10.1101/gad.1364906 (2006). 255 Cadot, B., Gache, V. & Gomes, E. R. Moving and positioning the nucleus in skeletal muscle - one step at a time. Nucleus 6, 373-381, doi:10.1080/19491034.2015.1090073 (2015). 256 Kiuchi, T., Higuchi, M., Takamura, A., Maruoka, M. & Watanabe, N. Multitarget super- resolution microscopy with high-density labeling by exchangeable probes. Nat Methods 12, 743-746, doi:10.1038/nmeth.3466 (2015). 257 Martell, J. D. et al. Engineered ascorbate peroxidase as a genetically encoded reporter for electron microscopy. Nat Biotechnol 30, 1143-1148, doi:10.1038/nbt.2375 (2012).

Abstract (if available)

Abstract Emery-Dreifuss muscular dystrophy (EDMD) is a laminopathy that results in progressive muscle degeneration and early death. EDMD is caused by mutations in lamin and the inner nuclear envelope protein emerin, both of which are responsible for nucleo-cytoskeleton mechanotransductions, maintenance of nuclear shape, and genome organization. The molecular pathogenesis of the disease, however, remains unclear. To further understand the role of emerin at the molecular level, we have employed single particle tracking and three-dimensional super resolution microscopy and we quantified the diffusion and spatial organization of emerin and its clinically relevant mutations. Using diffusion and pair-correlation analyses on both live and fixed cells, we identified subpopulations of emerin associated with the endoplasmic reticulum, outer nuclear envelope, and for the first time showed that emerin organizes into two distinct populations of monomers and oligomeric nanoclusters at the inner nuclear envelope. The diffusion and clustering state of these subpopulations are directly impacted by EDMD-associated mutations and are linked to a complex interplay between lamin A/C binding, nuclear actin binding, BAF interaction, LINC Complex association, and emerin oligomerization. We also studied how mechanical strains applied to the nucleus influence the normal and pathogenic organization of emerin and how emerin participates in the adaptation of nuclear shape to forces. To do so, we developed a cell micropatterning strategy that is based on microcontact printing of fibronectin after organosilane monolayer functionalization of glass coverslips. This approach uniquely combines exogenous control of nuclear shape with nanometer accuracy single molecule microscopy imaging of cells. Super-resolution imaging of emerin in micropatterned cells reveals that the mechanotransducing functions of emerin are intimately coupled to its nanoscale oligomerization and its dynamics within the inner nuclear envelope. We additionally show that emerin mutations result in its defective oligomerization and aberrant dynamics at the nuclear envelope, which leads to abnormal nuclear mechanics and improper adaptation to mechanical strains. Combining single molecule imaging and cell micropatterning techniques, we reveal key molecular mechanisms of EDMD pathogenesis that would have otherwise remained undetected by traditional diffraction-limited microscopy.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Investigations on the muscular dystrophy protein emerin: from nuclear mechanotransduction to molecular interactions

PDF

Physical principles of membrane mechanics, membrane domain formation, and cellular signal transduction

PDF

Unraveling the molecular mechanisms of heterochromatic double strand break repair in Drosophila cells

PDF

The organization of small RNA pathways within C. elegans germ granules: mutator foci formation, regulation, and interaction

PDF

The physics of emergent membrane phenomena

PDF

Spatial and quantitative mapping of caveolin-1 membrane nanodomains and their functional dynamic response to mechanosignaling cues

PDF

The role of endoplasmic reticulum chaperone protein GRP78 in breast cancer

PDF

From single molecules to bacterial nanowires: functional and dynamic imaging of the extracellular electron transfer network in Shewanella oneidensis MR-1

PDF

Distinct mechanisms of DDK recruitment to Fkh-activated and CEN-proximal origins control replication timing program in S. cerevisiae

PDF

Decrypting Escherichia coli DNA polymerase V mutasome ATP regulation

PDF

Subnuclear localization of replication origins is controlled by Fkh1-dependent recruitment of DDK to origins in S. cerevisiae

PDF

Nuclear respiratory factor-1 regulation by PTEN signaling

PDF

The role of WAVE proteins in the macropinocytosis-like entry of human papillomavirus

PDF

Targeted cell imaging by in-situ assembly and activation of hot spot SERS nanoprobes using split-fluorescent protein scaffolds

PDF

Molecular mechanism of the recruitment of SWI/SNF chromatin remodeling complex and histone acetyltransferase to estrogen-responsive promoters

PDF

Mammalian secretome profiling to identify adaptive and maladaptive signaling in homeostasis and diet induced obesity

PDF

The role of envelope protein in SARS-CoV-2 evasion of CD1d antigen presentation pathway

PDF

Dps contributes to typical growth, survival, and genome organization in E. coli

PDF

A novel construct to study the pulsatility of insulin secretion in single cells, islets and whole pancreas

PDF

The folding, misfolding and refolding of membrane proteins and the design of a phosphatidylserine-specific membrane sensor

Asset Metadata

Creator Fernandez, Anthony Michael (author)

Core Title Nanoscale dynamics and nuclear envelope organization of the muscular dystrophy related protein emerin

Contributor Electronically uploaded by the author (provenance)

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Molecular Biology

Degree Conferral Date 2022-05

Publication Date 12/21/2021

Defense Date 12/21/2021

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag emerin,microscopy,muscular dystrophy,nuclear membrane,OAI-PMH Harvest,single molecule

Format application/pdf (imt)

Language English

Advisor Pinaud, Fabien Florent (committee chair), Chiolo, Irene (committee member), Haselwandter, Christoph (committee member), Michael, Matthew (committee member)

Creator Email fernanam@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC18807268

Unique identifier UC18807268

Legacy Identifier etd-FernandezA-10319

Document Type Dissertation

Format application/pdf (imt)

Rights Fernandez, Anthony Michael

Type texts

Source 20211223-wayne-usctheses-batch-906-nissen (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu