CRWN FAMILY PROTEINS REGULATE NUCLEAR ORGANIZATION AND NUCLEAR FUNCTION
IN ARABIDOPSIS THALIANA
A Dissertation Presented to the Faculty of the Graduate School
of Cornell University In Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy
Haiyi Wang May 2014

© 2014 Haiyi Wang

CRWN FAMILY PROTEINS REGULATE NUCLEAR ORGANIZATION AND NUCLEAR FUNCTION
IN ARABIDOPSIS THALIANA
Haiyi Wang Ph. D. Cornell University [2014]
Abstract:
The molecular components and processes that shape and organize nuclei in plant cells are poorly understood. This thesis describes genetic, cytological, and biochemical studies of CRWN (CROWDED NUCLEI) proteins, which are required for proper nuclear structure in the flowering plant, Arabidopsis thaliana. These plant-specific proteins feature a long coiled-coil motif, with a conserved C-terminal domain. CRWN proteins are expressed primarily in proliferating tissues, and are located at the nuclear periphery. CRWN1 and CRWN4 belong to a nuclear fraction resistant to high salt and mild detergent extraction. I hypothesize that CRWN proteins are structural components of plant nuclei.
Genetic analysis of the whole family of crwn mutants in Arabidopsis thaliana revealed a variety of phenotypic changes, including altered nuclear shape, reduced nuclear size,

and heterochromatin aggregation or dispersion, mildly decreased endopolyploidy levels, and increased nuclear DNA density. In addition, some crwn mutants were dwarfed with early flowering times, shorter internodes, heavier branches, and delayed senescence. A subsequent mRNA-seq profiling in representative crwn mutants illustrated genome-wide transcriptional mis-regulation, and identified candidate genes responsible for phenotypic changes in crwn mutants. I propose that the loss of CRWN proteins primarily alters nuclear organization, leading to alterations in gene expression.
Phylogenetic analysis partitioned this protein family into CRWN1-like and CRWN4like sub-categories, and non-redundant morphological alterations observed in crwn1 versus crwn4 mutants further support this idea. Nonetheless, the transcriptomic data uncovered shared profiles of mis-expressed loci regulated by CRWN1-like and CRWN4-like genes. Moreover, regulatory relationships among CRWN paralogs exist on both the mRNA and protein level. A physical interaction between CRWN1 and CRWN4 proteins was demonstrated by immunoprecipitation experiments. These findings suggest that different CRWN proteins function together to maintain nuclear organization in plants.

BIOGRAPHICAL SKETCH
Haiyi Wang earned the bachelor degree of biological sciences from Fudan University in 2006. In college, Haiyi studied the ecology and conservation of grassland in Inner Mongolia, assisted curating the herbarium in Fudan University, and used molecular ecology tools to explore the genetic diversity in wild soybean populations from Yunnan Provinces.
Later, she joined the Graduate Program of Plant Biology in Cornell University, and started her thesis project in Richards’ lab since Jan 2009, focusing on nuclear organization and function in Arabidopsis thaliana. In graduate school, Haiyi explored various aspects of epigenetics, nuclear organization and evolution in Arabidopsis thaliana, and published first author article “Arabidopsis CROWDED NUCLEI (CRWN) proteins are required for nuclear size control and heterochromatin organization” on BMC Plant Biology. She attended many symposia and conferences, and presented posters on Chromosome Dynamics Gordon Conference (2010) and Plant Epigenetics, Stress and Evolution CSHA (Cold Spring Harbor Asia) Conference (2012). She had assisted teaching both graduate and undergraduate level courses, and supervised undergraduate students in the lab on CRWN projects.
Haiyi also actively participates in academic services, including volunteering in organization of conferences, serving on Graduate School panels, Seminar Committees, in Boyce Thompson Institute, and Plant Biology Graduate Student Association.
iii

ACKNOWLEDGMENTS
It has been unusual seven years in my life. I first want to express my gratefulness to Dr. Eric Richards, a very special scientist, for guiding me into the fun side of biology. I can never forget the moments when a mystic smile fulfilled with satisfaction flashes through his eyes, when the curiosity is lighted up by the beauty of nature.
I cherish the efforts my parents had made recent years as a remote support.
I want to say thank you to Dr. Wojtech Pawlowski, Jun Liu, June Nasrallah for their guidance and equipment supports on my research; to Dr. Jeff Doyle, Tom Owens and Jian Hua for their advices along my graduate life; to people in Dr. Tom Brutnell and Dr. Fei Zhangjun lab for many helpful discussions; to my current lab members as well as Travis Dittmer, a previous graduate student who initiated the CRWN project, for the blended intellectual inputs to my studies. My experiments could not be completed without the help from BTI and Cornell colleagues, including Lauren Dedow, Silin Zhong, Choonlin Tang, Patrick Boyle, Mamta Srivastava, and I got advices for the computational analysis from Dr. Qi Sun (CBSU) and bioinformaticians in BTI.
Life in these years has been enriched with many adventures. I would love to express my thankfulness to Molly Shook, Blake Meyers, Caroline Sartain and Aziana Ismail, for breaking the ice of diverse cultures with their genuineness, respects, and helps. I highly appreciated Chongshan Huang, Xiao Chen, Youyu Zhang, Xiangyi Xu,
iv

Lingyan Jin, Nan Zheng and Ting Shi for their company, regardless where they are in the world; and Lin Zhuang, Shu Huang and Sheldon Zhang, for sharing their immigration experiences with me. I will also remember the company and support from my Cornell fellows Xiaohua Yang, Xinsying Jiang, Jing Jin, Jun Qian, Lin Xu and Yuan Si. My stipend, tuition and research were funded by Olin Fellowship, TAship, NSF grants and BTI start-up fund for the Richards’ Lab.
v

TABLE OF CONTENTS Chapter1 Introduction: The nuclear organization and the regulation of nuclear function ( 1 ) Chapter 2 Arabidopsis CROWDED NUCLEI (CRWN) proteins are required for nuclear size control and heterochromatin organization ( 48 ) Chapter 3 The loss of CRWN proteins lead to broad transcriptional mis-regulation ( 98 ) Chapter 4 The nuclear coiled-coil proteins CRWN1 and CRWN4 physically interact to regulate nuclear organization in Arabidopsis thaliana ( 162 ) Chapter 5 Future directions ( 204 )
vi

LIST OF FIGURES
Chapter 1 • Figure 1.1 Organization of the nuclear envelope in animal cells. (8) • Figure 1.2 Tripartite structures of intermediate filament protein and lamin protein. (12)
Chapter 2 • Figure 2.1 Phylogenetic relationships among CRWN proteins. (53) • Figure 2.2 Whole plant phenotypes of crwn mutants. (59) • Figure 2.3 Nuclear phenotypes of crwn mutants. (63) • Figure 2.4 The effects of crwn mutations on nuclear size and nuclear DNA density in leaf cells. (65) • Figure 2.5 Average leaf guard cell nuclear sizes in crwn mutants. (72) • Figure 2.6 Chromocenter morphology changes in crwn mutants. (73) • Figure 2.7 Chromocenter organization is altered in crwn1 crwn2 and crwn4 mutants. (80) • Supplementary Figure 2.1 Amino acid sequences comprising the exreme C termini of 28 CRWN-like proteins, including ten CRWN4-like proteins (55) • Supplementary Figure 2.2 Transcript analysis of the crwn3-1 and crwn4-1 alleles used in this study (56) • Supplementary Figure 2.3 Nuclear shape changes in crwn1 and crwn4
vii

mutants (67) • Supplementary Figure 2.4 Leaf nuclear preparation and confocal imaging
reveals a consistent nuclear thickness across a range of nuclear sizes (69) • Supplementary Figure 2.5 Chromocenter changes in crwn double mutants
(75) • Supplementary Movie 2.1 - 2.3 Three-dimensional reconstruction of nuclei
imaged in the fluorescence in situ hybridization experiment (please refer to online data: http://www.biomedcentral.com/1471-2229/13/200/additional) Chapter 3 • Figure 3.1 The global pattern of transcriptional mis-regulation in crwn mutants (105) • Figure 3.2 Transposon activation in heterochromatic regions (107) • Figure 3.3 The relationship among significantly mis-regulated loci in crwn mutants (110) • Figure 3.4 Relative expression levels of all statistically tested loci in crwn mutants (113) • Figure 3.5 Functional categorization of mis-expressed loci in crwn mutants (119) • Figure 3.6 The activation of stress response pathways in crwn mutants (123) • Figure 3.7 Partial silencing of CRWN3 and CRWN4 in one of the crwn1 crwn2 replicates (142) • Supplementary Figure 3.1 Partial silencing of CRWN3 and CRWN4 proteins in one of the three crwn1 crwn2 replicates (144)
viii

Chapter 4 • Figure 4.1 CRWN protein domains and sequence polymorphism (168) • Figure 4.2 Specificity test for antisera against CRWN1 and CRWN4 proteins (173) • Figure 4.3 CRWN1 and CRWN4 proteins are insoluble under mild detergent and high salt extraction conditions (176) • Figure 4.4 CRWN4 abundance is reduced in crwn1 backgrounds (182) • Figure 4.5 The expression of CRWN mRNAs in various crwn mutants (183) • Figure 4.6 CRWN1 and CRWN4 proteins interact with each other in vivo (188) • Figure 4.7 A balancing model of CRWN1-like and CRWN4-like functions (195) • Supplementary Figure 4.1 The optimization of conditions for CRWN1CRWN4 co-immunoprecipitation (178)
ix

LIST OF TABLE
Chapter 1 • Table 1.1 A summary of nuclear organization relevant proteins in Arabidopsis thaliana (17)
Chapter 2 • Supplementary Table 2.1 CRWN-like proteins used in this study (54) • Supplementary Table 2.2 The nuclear phenotype data for crwn mutants used to construct Figure 2.4 (62)
Chapter 3 • Table 3.1 Summary of phenotypic changes in crwn mutants (102) • Table 3.2 Total number of tested genes and TEs (103) • Table 3.3 Epigenetically controlled loci are affected in crwn mutants (115) • Table 3.4 Summary of mis-regulated nuclear proteins in crwn mutants (134) • Supplementary Table 3.1 The activation of stress response pathways (124)
Chapter 4 • Table 4.1 Summary of CRWN protein sequence polymorphisms (170)
x

CHAPTER ONE
INTRODUCTION: THE NUCLEAR ORGANIZATION AND THE REGULATION OF NUCLEAR FUNCTION
History
The first description of the cell nucleus as an independent organelle was recorded in 1682 when Antonie van Leeuwenhoek drew a “lumen” structure that he observed in cod and salmon erythrocytes [1]. In 1802, Franz Bauer also described the putative cell nucleus in orchid, Bletia tankervilliae [2], and this observation was later popularized by the botanist Robert Brown when he presented to the Linnaean Society of London a structure that he called the “areola,” an opaque area in the outer layer cells of flowers [3].
These discoveries were followed by many studies and continuing debate on the functional role of the nucleus. In the year of 1838, Matthias Schleiden suggested that the nucleus was a structure he termed the “cytoblast,” or cell builder, that gave rise to cells to form the next cell generation. As a strong opponent of the cytoblast view, Franz Meyen proposed that cells multiply via cell division, and hypothesized that many cells would not have nuclei. In 1850s Robert Remak and Rudolf Virchow argued against the de novo cytoblast hypothesis and put forward a new paradigm that new cells can only originate from the division of pre-existing cells [4].
1

An important transition in cell theory occurred in the 1870s, when Oscar Hertwig showed that the sperm nucleus entered the oocyte and fused with its nucleus during sea urchin egg fertilization. Similar observations were later made in plants by Eduard Strasburger, and this work laid the foundation for the discovery of the essential function that the nucleus plays in heredity. In the beginning of the 20th century, Mendelian rules and the chromosome theory of heredity were developed, and the nucleus was recognized as the primary carrier of genetic information [4]. From that point forward attention and efforts were focused on understanding the structure and function of genes and how they control the phenotype of the cell and the organism. However, in the past few decades, research on nuclear organization has become popular again, uncovering new levels of regulation of transcription and other fundamental nuclear processes. In this introductory chapter, I will first selectively review work on the structural composition and functional impact of major nuclear components in non-plant systems, with a emphasis on three different topics: 1) the principles of chromosome and chromatin organization; 2) the molecular components of the nuclear envelope and nuclear pores; and 3) the different constituents of the nucleoskeleton. Then, I will turn my attention to work in plants, primarily Arabidopsis, which is directly relevant to my research.
2

Review of nuclear organization in non-plant systems
Chromosomal organization is important for transcriptional regulation Walther Flemming, who first described the process of mitosis in the mid-19th century, is credited with the discovery of chromosomes as visible or “stainable” units in the cell [5]. Carl Rabl later described a distinct bouquet-like organization of chromosomes that is often seen after cell division, with centromeres clustered on one side of the nucleus and chromosome ends located on the opposite side [6]. This observation indicated that the hereditary material had some higher-order threedimensional organization. Subsequent studies uncovered substructures within chromosomes. Cytological analysis of interphase nuclei in the liverwort Pellia epiphylla led Emil Heitz to identify two types of chromatin - the invisible or poorly stained regions of chromosomes as “euchromatin,” and the deeply stained visible knots as “heterochromatin” [7]. In most eukaryotic cells, decondensed euchromatin is associated with active transcription, and resides in the interior of the nucleus; while the tightly packed heterochromatin is associated with transcriptional repression, and is usually localized to the nuclear periphery [8]. However, exceptions exist. For example, in retinal rod cells of mice, the typical pattern is reversed, and heterochromatin is located in the interior of the nucleus. This unusual arrangement is thought to maximize the transparency of the rod cells to enhance the light transmission to the photoreceptors in these nocturnal mammals that need to see in dim light [8].
3

The mechanisms that mediate the spatial compartmentalization of eu- and heterochromatin are not well understood, but some clues have been generated from work in Drosophila and human demonstrating that an interaction between Heterochromatin Protein 1 (HP1) and the lamin B receptor help anchor heterochromatin to the nuclear envelope [9-11]. This non-random positioning of chromatin also impacts transcription [12, 13]. One classic example is position-effect variegation (PEV) of the expression of genes after relocation to chromosomal positions close to heterochromatin. Hermann Müller discovered PEV from the study of X-ray-induced chromosomal rearrangements in Drosophila [14]. He observed that the function of the euchromatic gene white, which was now placed adjacent to heterochromatin from the centromeric region due to the rearrangement, became variably inactivated. This inactivation led to clonal sectors of differential pigment in the insect’s compound eye. Later work demonstrated that the reversible inactivation of the white gene was due to heterochromatin spreading across the euchromatin/heterochromatin breakpoint, silencing white and giving rise to a variegated phenotype [15]. Insertion of transposable elements into euchromatic regions can also promote the formation of local heterochromatin and result in PEV [16]. Some researchers propose that selection has acted on these types of chromosomal position effects and resulted in the linear grouping of co-regulated genes along chromosome arms. One particularly striking example is the HOX genes that are ordered along the chromosome corresponding to their spatio-temporal expression for organ development [17]. Genomic studies in higher eukaryotes have illustrated the occurrence of clusters of highly-transcribed genes or tissue-specific genes along the chromosome, suggesting a position effect on transcriptional regulation [18].
4

The non-random compartmentalization of different regions along the chromosomes is likely achieved via alternate layers of compaction, which can also control or at least affect gene expression. A view emerged in the 1980s that chromatin fibers were organized into loop domains, anchored via S/MAR (scaffold or matrix attachment regions) sequences to a scaffolding system in interphase nuclei [19]. These sequences were first identified by purifying DNA fragments that remain associated with the “nuclear matrix,” generally defined as a salt- or detergent-resistant remnant left after extraction of nuclei [19]. S/MARs are typically a few hundred base pairs in length, enriched with AT nucleotides, and localized in the noncoding regions flanking genes [20]. Nonetheless, some studies have reported that S/MAR regions are transcriptionally active [21, 22]. The hypothesis that higher-order chromatin structure consists of confined and discrete chromosome territories is supported by the development of Fluorescence in situ Hybridization (FISH) techniques coupled with “chromosome painting” [23]. The more recent development of chromosome conformation capture techniques reveals that specific long-range interactions can occur, but that local interactions predominate. These Hi-C and related techniques have generally supported the looping model [24]. Simulations of the Hi-C data suggest that chromatin is organized like a polymer arranged in a linear array of loops compressed longitudinally [25]. These types of studies have the power to identify genes or regulatory sequences widely separated in cis, if they co-localize to a common niche in the nuclei for transcriptional regulation. One such example exists in erythroid cells in mouse, where the transcription factor Klf1 co-associates with genes that the protein
5

regulates at specialized transcription factories [26]. In parallel with the development of new genomics approaches to identify sequences involved in higher-order chromatin organization, computational tools are being applied to the problem. For example, algorithms were used to identify matrix attachment regions (MARS) with an effect on gene activation in the mammalian genome, and these MARS are being engineered into constructs for generating transgenic mice in an effort to alleviate the epigenetic silencing of transgenes and boost the expression of recombinant proteins [27]. The human ENCODE project surveyed all suspected chromatin boundary regions, and discovered 453 nuclear scaffold attachment sites predominately near expressed genes [28]. This highly compartmentalized and cell type-specific chromosome organization appears to be limited to interphase. In metaphase, a locus-independent homogenous folding state exists among all chromosomes in various cell types [24].
General physical mechanisms may also affect the positioning of sub-nuclear compartments, and consequentially affect biological functions [29]. It has been shown that macromolecular crowding contributes to the organization of nucleoli and promyelocytic leukemia (PML) nuclear bodies [30]. Also, osmotic perturbations can alter the local compaction state of chromatin and affect the formation and maintenance of heterochromatin packaging [30]. However, conformational change based on biophysical properties is insufficient to initiate or maintain the densely packed heterochromatin structures. In this case, specific regulatory proteins are still required for regulating chromatin states independent of organizing principles imposed by molecular crowding [30].
6

The nuclear envelope regulates chromosome positioning and transcription
The nuclear envelope defines the nucleus by its double bilayers of lipid molecules, with embedded nuclear pores for macromolecular trafficking (Figure 1.1). The outer nuclear membrane (ONM) resembles and connects to the endoplasmic reticulum, and the inner nuclear membrane (INM) is involved in dynamic nuclear functions. As noted earlier in this chapter, transcriptionally silent heterochromatin typically localizes close to the nuclear envelope or the nucleolus, but is excluded from the area adjacent to nuclear pores. Examples of heterochromatin preferentially localized to the nuclear periphery in mammals and yeast include repetitive centromeric sequences and transcriptionally silent telomeric chromatin [31], while highly transcribed genes are commonly found at nuclear pores [32].
Various anchoring mechanisms and pathways participate in position-dependent transcriptional regulation in the nucleus. In yeast, the SUN-domain protein, Mps3, anchors silent telomere chromatin to the nuclear rim. In mps3 mutants, telomeres become partially detached from the periphery and epigenetic repression of the telomeric regions is abrogated [33]. In a similar manner, the yeast LEM (LAP2, emerin, MAN1) protein Src1 (homolog of Man1 in mammals) is co-localized with sub-telomeric chromatin and src1 mutants mis-regulate the expression of subtelomeric genes [34]. In Drosophila and human cells, lamin (see below) proteins are important for positioning genomic regions at the nuclear periphery, and genome-wide profiling methods using tagging techniques, such as DamID, have revealed an enrichment of
7

Figure 1.1 Organization of the nuclear envelope in animal cells Figure 1.1 This figure describes the important components of the nuclear envelope in animals. Double lipid layers separate the nucleoplasm from the cytoplasm. Nuclear Pore Complexes (NPC) are imbedded in the nuclear envelope, with the nuclear baskets of the pore complex being oriented toward the inner nuclear membrane (INM) and the cytoplasmic filaments positioned at the cytoplasmic face with the the outer nuclear membrane (ONM). The SUN and KASH domain proteins form the LINC complexes (Linker of Nulceoskeleton and Cytoskeleton), bridging the nuclear lamina and cytoskeleton across the nuclear envelope.
8

transcriptionally silent loci at lamin-associated domains (LADs) [35, 36].
The functional significance of spatial compartmentalization was further validated by genetic perturbation of nuclear architecture. In mammalian cell culture systems, transgenic tandem arrays of E. coli lac operator (lacO) sites can be tethered to the nuclear envelope by lacI proteins fused with Lamin B1 or lamin-associated Emerin and Lap2β proteins [37, 38]. This perinuclear tethering down-regulates the expression of reporter genes close to the engineered lacO clusters [37, 38]. In yeast, the partial disruption of a cis-acting silencer element near the mating-type HMR locus alleviates its transcriptional repression, but silencing is restored when HMR is artificially recruited to the nuclear envelope [39]. However, this restoration was not successful in a genetic background with disrupted foci of repressive SIR (silent information regulator) factors, indicating that being anchored to the nuclear envelope facilitates silencing, but that this epigenetic regulation requires additional factors or pathways [40].
Nuclear pore: in and beyond the nucleoplasm-cytoplasm transportation
The nuclear pore complexes (NPC) are molecular channels that pass through the nuclear envelope. They are permeable to small molecules but highly selective for macromolecules. Transport of large molecules depends on nuclear transport factors (NTFs) that bind to the transport signals, such as the short amino-acid sequences of nuclear localization signals (NLSs) and nuclear export signals (NESs) in cargo
9

proteins, to facilitate transport through the NPCs [41]. Various subtypes of NPCs exist and these are differentially distributed among different cell types and at specific developmental stages to achieve a diverse selectivity in nucleus-cytoplasmic transport [41].
NPC-mediated transport also can exert an effect on nuclear morphology. The differences in nuclear size between the pseudotetraploid frog Xenopus laevis and its relative with a smaller diploid genome (Xenopus tropicalis) was mimicked in Xenopus egg extracts by altering the relative amounts of importin α and Ntf2, which are critical for lamin B3 import into the nucleus [42]. Also, Xenopus oocytes do not export a specific type of actin, resulting in its nuclear accumulation and the formation of actin filaments, which might be important for stabilization of the giant nucleus in this cell type [43].
Beyond the conventional function of nuclear-cytoplasmic transport, the NPCs seem to be the hubs of a multifunctional network on both sides of the nuclear envelope. On the cytoplasmic side, NPCs are connected to the cytoskeleton and translation machinery to coordinate messenger RNP (mRNP) export and the initiation of protein synthesis [44]. On the nucleoplasmic side, a network, which consists of macromolecular complexes and ribonucleoproteins, spreads from the baskets of NPCs into the nuclear periphery to connect to the neighboring NPCs. In animals, this network is referred to as the nuclear lamina, which plays important roles in maintaining the structural integrity of the nucleus as well as regulating gene expression and genomic integrity [45].
10

The nuclear lamina and the nucleoskeleton system in animals and non-plants
The nuclear lamina network in animal cells is largely composed of lamin proteins. These proteins were originally identified as the major components of the insoluble nuclear residue resistant to salt and detergent extraction [46]. Lamins are type V intermediate filament (IF) proteins with a conserved tri-partite structure, consisting of an N-terminal globular domain, a central coiled-coil rod domain, and a C-terminal globular domain (Figure 1.2, lamin A protein and IF protein) [47]. The rod domain is comprised of four coiled-coils (i.e., 1A, 1B, 2A, 2B), and a conserved 42 amino-acid motif within 1B distinguishes type V from other IF proteins [48].
Long coiled-coil domains are frequently involved in oligomerization. The coils consist of multiple copies of a core heptad repeat with hydrophobic amino acids on the first (a) and fourth (d) position, often isoleucine, leucine or valine. These hydrophobic residues coil around the helix to form an amphipathic secondary structure [49]. In a hydrophilic cell environment, two such helices wrap around each other to bury the hydrophobic surfaces, which drives the oligomerization. The tight packing in a coiledcoil interface is stabilized by van der Waals contact between the side chains of a and d residues [49, 50]. This type of oligomerization has been shown for many cases of long coiled-coil domain proteins, such as SMC (Structural Maintenance of Chromosomes) proteins, tropomyosins, and lamin proteins [51-53]. In vitro experiments have shown that lamin proteins form dimers and then polymerize into higher-order filaments through lateral associations of dimers [51]. The lattice of lamin-based filaments
11

Figure 2

A cytoplasmic intermediate filament

globular head domain

central alpha coiled-coil domain 1A 1B 2A 2B

globular tail domain

flexible linker region

B lamin, type V intermediate filament

globular head domain

central alpha coiled-coil domain

1A 1B

2A 2B

additional heptad repeats

globular tail domain
NLS Caax

Adapted from CJ Hutchison et al. Nature Cell Biology 2004
1A, 1B, 2A, 2B: four major long alpha coiled-coils NLS: Nuclear Localization Signal Caax: a site for carboxyl methylation, farnesylation and proteolytic cleavage

Figure 1.2 Tripartite structures of intermediate filament protein and lamin protein

12

Figure 1.2 This cartoon illustrates the tripartite structure of intermediate filament protein and lamin protein. Figure 1.2A shows the general composition of an intermediate filament protein, consisting of a variable globular head domain, a long central alpha coiled-coil rod domain, and a globular tail domain. The central coiledcoil domain contains four major long coils, named 1A, 1B, 2A and 2B. Figure 1.2B shows the structure of lamin, a type V intermediate filament, where the N terminal globular head domain is short, 1B coil domain contains six heptad repeats, and globular tail domain includes a nuclear localization signal as well as CAAX site for farnesylation and proteolytic cleavage and carboxyl methylation during lamin protein maturation.
13

provides a cage for the mechanical support of the nuclei, as well as a platform for the interaction with a number of regulatory proteins [45].
In non-human metazoa, lamins are important determinants of nuclear organization. A variety of different experiments in experimental animal systems have been used to manipulate lamins to monitor the effect on chromatin and nuclear structure. For example, RNAi knockdown of the only gene encoding lamin in C. elegans, Ce-lamin, leads to pleiomorphic nuclei [54]. In mice, the ectopic expression of the germ-linespecific lamin, B3, in somatic cells causes the development of hook-shaped nuclei, similar to the nuclear structure seen in spermatocytes [55]. Expression of a dominant negative version of human lamin B1 in somatic cells causes dramatic deformation in the nuclear envelope [56]. A lamin-A mutation in cultured fibroblasts alters nuclear morphology and re-distributes histone modification marks for silenced heterochromatin, which appears to lose its association with the nuclear periphery [45]. Work in cell-free systems supports the conclusions of the genetic experiments discussed above. For instance, incubation of condensed chromatin together with lamin-depleted extracts from Xenopus oocytes results in small and fragile nuclei [57].
In human cells, lamin proteins also form a meshwork structure underneath the inner nuclear envelope. This nuclear lamina extends from nuclear pore to pore, as well as throughout the nucleoplasm, and binds a large number of specific genomic domains and lamin-associated proteins, such as emerin (encoded by the EMD gene, null mutations lead to Emery-Dreifuss muscular Dystrophy), LBR (Lamin B Receptor)
14

and LAP2 (Lamin Associated Polypeptide 2) [45]. Mutations and polymorphisms in lamin and lamin-associated proteins lead to malfunctioning nuclei and impair normal development. More than two hundred polymorphisms in the LMNA gene, which encodes lamin A and C, were discovered to be associated with a variety of human diseases collectively called laminopathies [58]. For instance, specific polymorphisms in LMNA cause the premature aging disease Hutchinson-Gilford Progeria Syndrome (HGPS). Cells isolated from laminopathy patients display fragile and irregularlyshaped nuclei with redistributed histone modification marks [58]. These observations suggest that a properly assembled nuclear lamina is important to maintain nuclear morphology and function, such as epigenetic modification and transcriptional regulation.
Plant, fungi and unicellular organisms do not have obvious lamin protein homologs. In yeast, the nucleus lack lamins. The functional compartmentalization in yeast nuclei is achieved by specific sequence elements, protein-protein interactions, nuclear envelope and nuclear pore anchorage sites, as well as chromosomal long-range interactions [59]. In the unicellular Trypanosomes, a large coiled-coil protein NUP-1 (nuclear peripheral protein 1) locates to the inner face of the nuclear envelope, as part of a stable network [60]. NUP-1 knockdown changes nuclear shape, disrupts the organization of nuclear pore complexes, and alters chromatin states, especially in the telomere-proximal region where silenced variant surface glycoprotein (VSG) genes are located. A loss of silencing in this region can increase VSG switching, which
15

allows the parasite to escape detection by the host immune systems. Therefore, NUP-1 is hypothesized to function like a lamin analog in this non-metazoan system [61]. Nuclear organization in Arabidopsis Here, I review the current understanding of three aspects of nuclear organization in Arabidopsis thaliana, using representative studies: 1) nuclear proteins important for organelle structure; 2) the chromosome and chromatin scaffolding system; and 3) chromatin remodeling associated with dynamic nuclear organization. The proteins essential for plant nuclear organization are summarized in Table 1.1.
16

Table 1.1 A summary of nuclear organization relevant proteins in A. thaliana Table 1.1 This table summarizes nuclear proteins contributing to the nuclear organization in Arabidopsis thaliana mentioned in this chapter. The gene names, locus ID, morphological alterations in the absence of corresponding gene, are listed.
17

proteins SUN WIP
CRWN1/LINC1
CRWN2/LINC2 CRWN3/LINC3
CRWN4/LINC4 GIP WIT

location Nuclear*Envelope* Nuclear*Envelope* Nuclear*Envelope*
Nuclear*Envelope* Nuclear*Envelope
Nuclear*Envelope Nuclear*Envelope Nuclear*Envelope

XIIi Nup136 Nup1 LNO1 MOS7 MOS3 Nup160 MFP1 MAF1
AHL22 TEK ARP4 ARP5 ARP6 ARP7
SMC1/3 SMC2/4 SMC5/6
DMS3 DDM1
NRPD HDA6

Nuclear*Envelope Nuclear*Pore*Complex Nuclear*Pore*Complex Nuclear*Pore*Complex Nuclear*Pore*Complex Nuclear*Pore*Complex Nuclear*Pore*Complex S/MAR*interacting*protein S/MAR*interacting*protein
AT*Hook*S/MAR*interacting*protein AT*Hook*S/MAR*interacting*protein Actin*Related*Protein*in*the*nucleus Actin*Related*Protein*in*the*nucleus Actin*Related*Protein*in*the*nucleus Actin*Related*Protein*in*the*nucleus Structural*Maintainence*of*Chromosomes Structural*Maintainence*of*Chromosomes Structural*Maintainence*of*Chromosomes
SMC*like,*chromatin*remodeling chromatin*remodeling
chromatin*remodeling chromatin*remodeling

function SUN*domain KASH*domain nuclear*and*chromosome*organization
nuclear*and*chromosome*organization nuclear*and*chromosome*organization
nuclear*and*chromosome*organization nuclearIcytoplasmic*connection
nuclear*positioning*and*movement
nuclear*positioning*and*movement nucleoporin*/*nuclear*shape nucleoporin*/*nuclear*shape nucleoporin*/*mRNA*export
nucleoporin*/*immune*response nucleoporin*/*immune*response
nucleoporin*/*mRNA*export chromatin*anchorage* chromatin*anchorage* chromatin*remodeling chromatin*remodeling chromatin*remodeling chromatin*remodeling chromatin*remodeling chromatin*remodeling
cohesin,*chromosome*organization codensin,*chromosome*organization cohesion*and*HR*mediated*DNA*repair RNA*directed*DNA*methylation*(RdDM)
maitainence*of*DNA*methylation Pol*IV*sub*unit,*RdDM histone*acetylation

mutants

nuclear/morphology/

sun1sun2

round

wip1)wip2)wip3

round

crwn1

small*and*round

not*changed*or*slightly*

crwn2

rounder*/*smaller

crwn3

not*chnaged

round*with*irregular*

crwn4

margin

gip1)gip2

round

wit1)wit2

invaginated*NE*and*

impaired*nuclear*

kaku1

movement

nup136)nup1

round

nup136)nup1

round

lno1

mos7

mos3

nup160

mfp1

maf1

ahl22

decondensed*chromatin

tek decondensed*chromatin

arp4

arp5

arp6

arp6

smc

smc

smc

dms3

decondensed*chromatin

ddm1

decondensed*chromatin

nrpd1/2 decondensed*chromatin

hda6

decondensed*chromatin

Arabidopsis nuclear proteins important for organelle structure
Plants appear to have evolved an independent nucleoskeletal system divergent from other eukaryotes [62]. Among the nucleoskeletal components in animals, very few of them have homologous proteins in plants (Table 1.1). One example is Sad1/UNC-84 (SUN)-domain proteins. There are five homologs in Arabidopsis thaliana: AtSUN1 and AtSUN2 contain the highly conserved C-terminal SUN domain, resembling SUN protein structure in yeast, worms, flies and mammals; while AtSUN3 – AtSUN5 harbor a SUN domain in the middle of the proteins [63]. Fluorescently tagged protein fusions for AtSUN1 and AtSUN2 form homo- and hetero-dimers, which localize to the inner nuclear envelope and show low mobility in vivo [64, 65]. The inner nuclear membrane AtSUN proteins interact with the outer nuclear membrane AtWIP(WPP domain-interacting proteins) proteins to form SUN-KASH (Klarsicht/ANC-1/Syne homology) complexes in Arabidopsis, as the linkers of the nucleoskeleton to the cytoskeleton (LINC). This LINC complex is conserved from yeast to human, and is involved in the positioning of nuclei and chromatin. The sun1 sun2 double mutant has round nuclei, while the wip1 wip2 wip3 triple mutants partially disrupt the spindle shape of the nuclei [65]. These findings indicate that the interaction between AtWIPs and AtSUNs are required to maintain the elongated nuclear shape typically seen in enlarged Arabidopsis cells [65].
Besides SUN-KASH proteins, a newly characterized GIP protein is also hypothesized to be involved in the nucleo-cytoplasmic connection [66]. GIP (γ-TuC protein 3
18

Interacting Protein 1) proteins are small components of γ-Tubulin Complexes (γ-TuCs) important for perinuclear localization and nucleation of microtubules (MTs). gip1 gip2 mutants exhibited a severe alteration of nuclear shape, associated with an abnormal distribution of the NPCs and AtSUN1. Further, AtGIP1 also interacts with the nuclear envelope protein AtTSA1 [66].
Another type of nucleo-cytoplasmic linker consists of a plant-specific myosin motor XI-i and the nuclear membrane proteins WIT1 and WIT2, which might enable rapid nuclear positioning in response to environmental stimuli [67, 68]. XI-i is one of the thirteen myosin XI motor proteins responsible for transporting plant organelles on the actin cytoskeleton [69]. Myosin XI-i, which is encoded by the KAKU1 gene in Arabidopsis, is anchored to the nuclear membrane by the outer nuclear membrane proteins WIT1 and WIT2. The kaku1 mutant exhibits abnormal invagination of the nuclear envelope, and a deficiency of either myosin XI-i or WIT proteins diminishes dark-induced nuclear positioning in plant mesophyll cells [67].
Nuclear pore proteins implicated in nuclear organization
As in animal cells, the nuclear pore complexes (NPCs) primarily facilitate nucleocytoplasmic transport in plants. Proteomic studies showed that Arabidopsis contains at least 30 nucleoporins similar to nucleoporins in human and yeast. NUA (Nuclear Pore Anchor), an Arabidopsis ortholog of the long coiled-coil filament proteins of the nuclear pore basket, locates at the inner surface of the nuclear envelope and is
19

important in controlling mRNA export, as well as SUMO protease activity at the nuclear pore [70]. LNO1, homologous to Nup214 in human and Nup159 in yeast, encodes a nucleoporin protein required for mature mRNA export, and is essential for embryogenesis and seed viability [71]. MOS7, an Arabidopsis homolog to human and fly nucleoporin Nup88, localizes to the nuclear envelope and is required for the nuclear accumulation of SNC1, EDS1 and NPR1. mos7-1 single mutant plants exhibit defects in basal and R protein-mediated immunity and in systemic acquired resistance [72]. In the same pathway, MOS3, an Arabidopsis homolog of Nup96 in vertebrates, is required for constitutive plant immunity mediated by the R gene SNC1 [73]; while Arabidopsis Nup160 is needed for nuclear mRNA export and full expression of resistance dependent on EDS1 [74]. Additionally, a plant specific nucleoporin Nup136/Nup1 shows greater mobility on the nuclear envelope than other nucleoporins, and nup136/nup1 mutants exhibit a rounder nuclear shape than wild type, as well as various defects in plant development [73].
CRWN family proteins as candidate molecular components of the Arabidopsis nucleoskeleton
In the 1990s, Masuda and colleagues reported the discovery of Nuclear Matrix Constituent Protein 1 (DcNMCP1) a protein residing on the periphery of carrot nuclei and a component of the salt-resistant nuclear matrix [75]. The localization of the protein and its structure, which featured a long coiled-coil domain, suggested that this protein might serve an architectural role in plant nuclei. This result was furthered
20

supported by the discovery of AcNMCP1 in onion (Allium cepa) and AgNMCP1 in celery (Apium graveolens), which are homologous to DcNMCP1 and belong to the nuclear matrix [76, 77]. Immuno-staining showed that AcNMCP1 proteins were concentrated around the nuclear rim, with some punctate signals in the nucleoplasm [75], while AgNMCP1 co-localized with the mitotic spindle and segregating chromosomes during mitosis [77]. NMCP1 and related proteins are plant-specific and share no significant amino-acid similarity to lamins, but their tripartite structure with a central coiled-coil domain is reminiscent of lamins. Moreover, the localization of these proteins at the nuclear periphery suggests that NMCP1-related plant proteins might function like lamin analogs [78, 79]. However, it is important to note that other structural analyses has suggested that NMCP proteins share more similarity with animal myosins or paramyosins than with lamins [80].
The Arabidopsis genome encodes four NMCP homologs, which were first identified in bioinformatic surveys of predicted proteins that contain coiled coil domains. Based on reverse genetics studies undertaken previously in the Richards lab and extended in this thesis, we refer to the Arabidopsis NMCP proteins as CRWN (CROWDED NUCLEI) based on the reduced nuclear size observed in crwn mutants [78] (also refer to Chapter 2]. These proteins are also known in the literature by their previous name, LINC (LITTLE NUCLEI) [78]. Our previous studies showed that the Arabidopsis CRWN1 and CRWN2 expressed as fluorescently-tagged fusion proteins in transgenic plants primarily localize to the nuclear periphery. The over-expression of CRWN2 leads to its distribution throughout the nucleoplasm, however [78]. A recent study
21

from another group shows that CRWN4 (LINC4) proteins are located at the nuclear periphery. CRWN1 (LINC1) proteins co-localize with mitotic chromosomes during the anaphase of cell cycle while the other CRWN proteins were distributed throughout the cell during mitosis [81].
Previous work in the Richards lab demonstrated that CRWN1 and CRWN2 play important roles in specifying nuclear shape and size, and crwn1 crwn2 double mutants have reduced numbers of chromocenters, conspicuous heterochromatin aggregates visible at interphase [78]. Recently, another group also reported that disruption of CRWN4 reduced nuclear size and caused loss of the elongated nuclear shape in differentiated cells [81]. Chapter 2 of this thesis presents a comprehensive study of the phenotypic effects caused by mutations in the CRWN gene family.
Scaffold proteins interacting with chromosomes
A genome scale analysis in Arabidopsis demonstrates that genes containing intragenic S/MARs predicted by SMARTest [82] are enriched for genes encoding transcription factors and are associated with more pronounced regulation (in contrast to constitutively expressed ‘housekeeping’ genes) [83]. These observations suggest a role for chromatin structural characteristics in mediating transcriptional regulation in a tissue-specific manner [83]. Transgenic constructs with strong MAR sequences flanking reporter genes stimulated median gene expression by five- to ten-fold in Arabidopsis [84]. Proteins associated with chromatin at MAR regions are considered
22

to be important for dynamic chromatin organization in various fundamental nuclear processes in all eukaryotic cells. An investigation of isolated plant nuclear matrix identified a candidate structural protein MFP1 (MAR binding filament-like protein 1) that is located at the nuclear periphery and is associated with speckle-like structures via its predicted N-terminal trans-membrane domain [85]. MFP1, a nuclear and plastid protein [86], binds to MAR DNA and may function in attaching the chromosomes to the nuclear envelope [85, 87]. Another small and soluble serine/threonine-rich protein MAF1 (MFP1-associated factor 1) also localizes to the nuclear periphery and interacts with MFP1 [88]. These two proteins are hypothesized to be candidate components of plant nuclear matrices [85].
Nuclear localized AT-hook proteins
Another emerging group of candidate nuclear structural proteins are AHL proteins (AT-Hook motif containing nuclear Localized), which contain two AT-hook motifs and one PPC (plant and prokaryote conserved) domain. These proteins localize to the nucleus and bind to AT-rich DNA sequences, such as S/MAR regions. There are 29 AHL genes in the Arabidopsis genome [89]. The AT-hook protein AHL22 (AT-hook motif nuclear localized 22) regulates flowering time by modifying chromatin at the FT gene (Flowering Locus T) [90]. Over expression of the AHL22 gene reduces histone H3 acetylation and elevates histone H3K9 di-methylation, and activates the FT gene, resulting in delayed flowering [89, 90]. Another AT-hook DNA binding protein, TEK (Transposable Element silencing via AT-hooK), participates in silencing transposable
23

elements (TE) and TE-like sequence containing genes, such as Ler FLC (Flowering Locus C), MAF4 (MADS Affecting Flowering 4), MAF5, AtMu1 and FWA. Knockdown of TEK using miRNA leads to expression of these genes and reactivation of transposable elements, associated with increased histone acetylation, reduced histone H3K9 di-methylation, and DNA hypomethylation in these loci [91-93].
Nuclear localized Actin Related proteins (ARP)
The actin-related proteins (ARPs) that locate to the nucleus are also involved in dynamic chromatin organization and transcriptional regulation. In Arabidopsis, there are four ARP proteins, and all of them are present in chromatin modifying complexes. AtARP4, a conserved homolog of human BAF53 and yeast Arp4, belongs to chromatin modifying complexes and is concentrated in the nucleoplasm of plant cells. The arp4-1 mutant, which exhibits reduced ARP protein activity, has defects in anther development and is partially sterile. A complete knockdown of ARP4 by RNAi led to strong pleiotropic phenotypes in organ organization, timing of flowering, and high levels of sterility [94]. Another ARP member in Arabidopsis, AtARP5, corresponds to a conserved subunit of the INO80 chromatin-remodeling complex in yeast and mammals. The AtARP5 protein is ubiquitously expressed and localizes to the nucleoplasm of interphase cells. A null arp5 mutant produces moderately dwarfed plants, with reduced cell size and delayed stomatal development. These plants are also hypersensitive to DNA-damaging reagents [95]. Moreover, AtARP6, a homolog of ARP6 in yeast and other organisms, is a component of the SWR1 chromatin-
24

remodeling complex. AtARP6is also universally expressed and localizes to the nucleus in interphase. Null mutations in AtARP6 caused numerous developmental defects, female fertility and early flowering. Some of these phenotypic changes could be explained by a down-regulation of FLC (Flowering Locus C), MAF4 and MAF5 (MADS Affecting Flowering 4 and 5) in arp6 mutants [96, 97]. AtARP7 is a constitutively expressed nuclear protein. The arp7-1 T-DNA mutant is homozygous lethal and produces abnormal homozygous embryos that are blocked in their development at or before the torpedo stage. A drastic knockdown of AtARP7 by RNAi results in severe dwarfism, pleiotropic phenotypes, and defective cell expansion and trichome morphology [98]. It has been proposed that these transcriptional and developmental changes in various arp mutants are a result of transcriptional misregulation mediated by altered chromatin structures due to the loss of ARP proteins in the chromatin modifying complexes [99].
Structural Maintenance of Chromosome (SMC) complexes
Similar to other eukaryotes, plants use SMC (Structural Maintenance of Chromosome) protein complexes and their interacting partners for sister chromatid cohesion and chromosome condensation [100]. Arabidopsis contains single copies of the cohesion subunits SMC1 and SMC3. Both of these subunits are required for sister chromatid cohesion along chromosome arms in somatic interphase nuclei [100]. Condensin subunits include SMC2A/B and SMC4A/B, which are important for chromosome
25

condensation in mitosis and meiosis [100]. Besides scaffolding the chromosomes, the SMC complexes also participate in other basic nuclear functions, such as DNA repair and epigenetic regulation. In Arabidopsis, the incompletely aligned sister-chromatids in interphase nuclei increase the chromatin accessibility and facilitate frequent interchromosomal recombination [101]. In somatic cells, this feature is important for efficient DNA repair via homologous recombination guided by the SMC5/6 complex [102]. Moreover, DMS3 (At3g49250) encodes a protein similar to the hinge-domain region of SMC proteins, which functions together with RDR1 and Pol IVb subunits in the RdDM (RNA-directed DNA methylation) machinery. These complexes establish cytosine methylation on DNA regions homologous to the Dicer-generated small RNAs to achieve gene silencing [103]. These examples suggest that chromosome architectural proteins play various functional roles in maintaining the chromosome structure and function.
Chromatin remodeling and dynamic nuclear organization in Arabidopsis
Epigenetic reprogramming during plant development is also accompanied by significant alteration in chromatin organization. At the beginning of seed maturation, embryonic cotyledon nuclei display significantly reduced nuclear size and highly condensed chromatin [104]. These morphologies are released during germination, when an increase in nuclear size occurs, along with a partial decondensation of the chromatin and restoration of transcription [104]. In the floral transition, chromosomes go through another round of decondensation and transcriptional reprogramming to
26

accomplish developmental changes [105]. Moreover, in experimentally cultured plant cells, dedifferentiation of mesophyll cells into protoplasts disrupts chromocenters and decondenses many major heterochromatic repeat regions. Under specific conditions, this process could be reversed by re-condensation of heterochromatin into chromocenters in a stepwise manner [106].
Several layers of epigenetic remodeling are associated with reorganization of chromatin compaction in the nuclei. Examples include DNA methylation, small RNA silencing, histone modification, and crosstalk among these pathways [106]. DNA methylation plays important roles in transcriptional silencing and heterochromatin maintenance. The ddm1 mutant in Arabidopsis, with its defect in the maintenance of DNA methylation, shows de-condensed chromocenters in the nuclei of mature cells [106]. Small interfering RNAs also participate in this silencing process via siRNAdirected DNA methylation. The Arabidopsis null mutants for plant-specific RNA polymerases IV impair the biogenesis of siRNAs, release the silencing of transposable elements and pericentromeric repeats, and cause the dispersal of heterochromatin [107]. Histone acetylation and deacetylation are also involved in mediating transcriptional activation and silencing. Arabidopsis plants carrying a mutation in HDA6, a class I histone deacetylase gene, is associated with increased histone H3K4 methylation, decondensation of the chromatin, and a loss of transcriptional silencing at several transgenic loci as well as in endogenous repetitive regions [108].
The re-organization of heterochromatin in Arabidopsis can also be induced by
27

environmental cues. One example comes from the exploration of Arabidopsis natural variation, which revealed a positive correlation among chromatin compaction, latitude of geographic origin and local light intensity. The natural strain Cape Verde Islands-0 (Cvi-0) contains a polymorphism in the PHYB gene and within the HDA6 promoter. This genotype leads to a plant that resembles either hda6 or phyB5 mutant in having reduced chromatin compaction [110]. Cvi-0 also displayed decreased methylation levels of DNA and histone H3 K9 methylation at the ribosomal RNA gene clusters as the loss of HDA6 protein does [110]. Thus both PHYB (Phytochrome-B) and HDA6 promote global chromatin compaction level in response to light, suggesting that acclimation of Arabidopsis to its environment could be associated with chromatin plasticity [109]. In another example, prolonged heat stress showed transcriptional activation of several repetitive elements in the heterochromatic region that are under epigenetic regulation. These gene expression changes were accompanied by loss of nucleosomes and by heterochromatin decondensation. However, this change was transient, and did not involve clear repressive epigenetic changes such as DNA hypomethylation and changes in histone modifications. The nucleosome loading and transcriptional silencing was reestablished after the plant recovered from heat stress, but a delayed recovery was observed in mutants with chromatin assembly deficiencies [110]. Therefore, chromatin organization can react to environmental cues independent of epigenetic modifications, and this feature might facilitate more permanent epigenetic changes [110].

Summary

28

Architecture is something that makes space meaningful. The same thing holds true for nuclei. Certain structures were needed to define the margin of the first nucleus. Once formed, the nucleus and its organization started to evolve and pick up new functions along the evolutionary path. Progress in the study of nuclear organization - the dynamic architectural components and various molecular machinery functioning in that environment - promises to uncover a new layer of spatial regulation of nuclear functions in both non-plant and plant systems.
Overview of Dissertation
In this dissertation, I took advantage of Arabidopsis thaliana as a model system in plants, to study nuclear organization and its impact on nuclear function. The work focuses on a plant-specific long coiled-coil domain protein family, CRWN, which stands for “CRoWded Nuclei”.
Chapter 2 contains a description of my genetic analysis of all the viable crwn mutant combinations. Phenotypes I observed in these mutants included whole plant dwarfism, nuclear size reduction, increased nuclear DNA density, and disruption of heterochromatin organization. The phylogenetic analysis of CRWN proteins, as well
29

as the morphological alterations I recorded on both the whole plant and nuclear levels point to a functional divergence between CRWN1-like and CRWN4 proteins in the family. Chapter 2 has been published in BMC Plant Biology 2013, 13:200.
In Chapter 3, I performed mRNA-seq profiling on six select crwn mutant genotypes. This analysis confirmed the synergistic relationship among CRWN1, CRWN2 and CRWN3 proteins, and uncovered a suppression between crwn1 and crwn4 mutations in crwn1 crwn4 double mutants. Moreover, the most severe mutants, crwn4 and crwn1 crwn2, shared common genomic targets in transcriptional mis-regulation. Many misexpressed genes were identified as candidates for corresponding morphological changes in crwn mutants. Thes mutants also display a mild release of epigenetic silencing and activation of various stress response pathways. Many structural and functional nuclear proteins were mis-expressed, potentially leading to disruption of both nuclear organization and basic nuclear processes, such as DNA replication, epigenetic modification, and DNA repair.
In Chapter 4, I obtained antisera specifically recognizing CRWN1 or CRWN4 proteins. Cell fractionation experiments confirmed that, as like other NMCP proteins, both CRWN1 and CRWN4 proteins are resistant to high salt and mild detergent treatment, thus could be candidate components of nucleoskeleton. Interestingly, I demonstrated that CRWN4 protein is down-regulated in crwn1 mutants. Also, a compensatory regulation among different CRWN genes exists on the mRNA level,
30

suggesting a complex set of interactions among these paralogs. Last, I used coimmunoprecipitation to demonstrate that CRWN1 and CRWN4 proteins physically interact in vivo, supporting a model that CRWN1-like and CRWN4 proteins are working together to organize plant nuclei. I hypothesize that a balance between the divergent CRWN1-like and CRWN4 function is essential for proper nuclear organization. Chapter 5 contains a brief summary of this thesis and a discussion of potential future directions.
31

REFERENCES
1. Dobell C, Leeuwenhoek Av: Antony van Leeuwenhoek and his "Little animals"; being some account of the father of protozoology and bacteriology and his multifarious discoveries in these disciplines. New York,: Harcourt, Brace and company; 1932.
2. Harris H: The birth of the cell. New Haven, Conn.: Yale University Press; 1999.
3. Brown R: On the organs and mode of fecundation and Orchidex and Asclepiadea. In: Miscellaneous Botanical Works I: 511-514. 1866.
4. Cremer T: Von der Zellenlehre zur Chromosomentheorie : naturwissenschaftliche Erkenntnis und Theorienwechsel in der frühen Zell- und Vererbungsforschung. Berlin ; New York: Springer-Verlag; 1985.
5. Lukacs D: [Walter Flemming, discoverer of chromatin and mitotic cell division]. Orvosi hetilap 1981, 122(6):349-350.
6. Rabl C: "Uber Zelltheilung" Morphologisches Jahrbuch. 1885, 10. 7. E H: Das Heterochromatinder Moose. I. Jahrh wiss Bot 1928, 69:762-818. 8. Solovei I, Kreysing M, Lanctot C, Kosem S, Peichl L, Cremer T, Guck J, Joffe
B: Nuclear architecture of rod photoreceptor cells adapts to vision in mammalian evolution. Cell 2009, 137(2):356-368. 9. Pyrpasopoulou A, Meier J, Maison C, Simos G, Georgatos SD: The lamin B receptor (LBR) provides essential chromatin docking sites at the nuclear envelope. The EMBO journal 1996, 15(24):7108-7119.
32

10. Ye Q, Callebaut I, Pezhman A, Courvalin JC, Worman HJ: Domain-specific interactions of human HP1-type chromodomain proteins and inner nuclear membrane protein LBR. The Journal of biological chemistry 1997, 272(23):14983-14989.
11. Ye Q, Worman HJ: Interaction between an integral protein of the nuclear envelope inner membrane and human chromodomain proteins homologous to Drosophila HP1. The Journal of biological chemistry 1996, 271(25):14653-14656.
12. Makatsori D, Kourmouli N, Polioudaki H, Shultz LD, McLean K, Theodoropoulos PA, Singh PB, Georgatos SD: The inner nuclear membrane protein lamin B receptor forms distinct microdomains and links epigenetically marked chromatin to the nuclear envelope. The Journal of biological chemistry 2004, 279(24):25567-25573.
13. Hirano Y, Hizume K, Kimura H, Takeyasu K, Haraguchi T, Hiraoka Y: Lamin B receptor recognizes specific modifications of histone H4 in heterochromatin formation. The Journal of biological chemistry 2012, 287(51):42654-42663.
14. Muller HJ: Types of visible variations induced by X-rays in Drosophila. Journal of Genetics 1930, 22(3):299-334.
15. Wallrath LL, Elgin SC: Position effect variegation in Drosophila is associated with an altered chromatin structure. Genes & development 1995, 9(10):1263-1277.
33

16. Dorer DR, Henikoff S: Expansions of transgene repeats cause heterochromatin formation and gene silencing in Drosophila. Cell 1994, 77(7):993-1002.
17. Duboule D: The rise and fall of Hox gene clusters. Development 2007, 134(14):2549-2560.
18. Caron H, van Schaik B, van der Mee M, Baas F, Riggins G, van Sluis P, Hermus MC, van Asperen R, Boon K, Voute PA et al: The human transcriptome map: clustering of highly expressed genes in chromosomal domains. Science 2001, 291(5507):1289-1292.
19. Mirkovitch J, Mirault ME, Laemmli UK: Organization of the higher-order chromatin loop: specific DNA attachment sites on nuclear scaffold. Cell 1984, 39(1):223-232.
20. Heng HH, Goetze S, Ye CJ, Liu G, Stevens JB, Bremer SW, Wykes SM, Bode J, Krawetz SA: Chromatin loops are selectively anchored using scaffold/matrix-attachment regions. Journal of cell science 2004, 117(Pt 7):999-1008.
21. Cai S, Lee CC, Kohwi-Shigematsu T: SATB1 packages densely looped, transcriptionally active chromatin for coordinated expression of cytokine genes. Nature genetics 2006, 38(11):1278-1288.
22. Wang L, Di LJ, Lv X, Zheng W, Xue Z, Guo ZC, Liu DP, Liang CC: InterMAR association contributes to transcriptionally active looping events in human beta-globin gene cluster. PLoS One 2009, 4(2):e4629.
34

23. Dietzel S, Jauch A, Kienle D, Qu G, Holtgreve-Grez H, Eils R, Munkel C, Bittner M, Meltzer PS, Trent JM et al: Separate and variably shaped chromosome arm domains are disclosed by chromosome arm painting in human cell nuclei. Chromosome research : an international journal on the molecular, supramolecular and evolutionary aspects of chromosome biology 1998, 6(1):25-33.
24. Dekker J, Marti-Renom MA, Mirny LA: Exploring the three-dimensional organization of genomes: interpreting chromatin interaction data. Nature reviews Genetics 2013, 14(6):390-403.
25. Lieberman-Aiden E, van Berkum NL, Williams L, Imakaev M, Ragoczy T, Telling A, Amit I, Lajoie BR, Sabo PJ, Dorschner MO et al: Comprehensive mapping of long-range interactions reveals folding principles of the human genome. Science 2009, 326(5950):289-293.
26. Schoenfelder S, Sexton T, Chakalova L, Cope NF, Horton A, Andrews S, Kurukuti S, Mitchell JA, Umlauf D, Dimitrova DS et al: Preferential associations between co-regulated genes reveal a transcriptional interactome in erythroid cells. Nature genetics 2010, 42(1):53-61.
27. Lee TH, Kim SJ, Han YM, Yu DY, Lee CS, Choi YJ, Moon HB, Baik MG, Lee KK: Matrix attachment region sequences enhanced the expression frequency of a whey acidic protein/human lactoferrin fusion gene in the mammary gland of transgenic mice. Molecules and cells 1998, 8(5):530536.
35

28. Keaton MA, Taylor CM, Layer RM, Dutta A: Nuclear scaffold attachment sites within ENCODE regions associate with actively transcribed genes. PloS one 2011, 6(3):e17912.
29. Hancock R: A role for macromolecular crowding effects in the assembly and function of compartments in the nucleus. Journal of structural biology 2004, 146(3):281-290.
30. Walter A, Chapuis C, Huet S, Ellenberg J: Crowded chromatin is not sufficient for heterochromatin formation and not required for its maintenance. Journal of structural biology 2013, 184(3):445-453.
31. Palladino F, Laroche T, Gilson E, Axelrod A, Pillus L, Gasser SM: SIR3 and SIR4 proteins are required for the positioning and integrity of yeast telomeres. Cell 1993, 75(3):543-555.
32. Brickner JH, Walter P: Gene recruitment of the activated INO1 locus to the nuclear membrane. PLoS biology 2004, 2(11):e342.
33. Bupp JM, Martin AE, Stensrud ES, Jaspersen SL: Telomere anchoring at the nuclear periphery requires the budding yeast Sad1-UNC-84 domain protein Mps3. The Journal of cell biology 2007, 179(5):845-854.
34. Grund SE, Fischer T, Cabal GG, Antunez O, Perez-Ortin JE, Hurt E: The inner nuclear membrane protein Src1 associates with subtelomeric genes and alters their regulated gene expression. The Journal of cell biology 2008, 182(5):897-910.
35. van Steensel B, Delrow J, Henikoff S: Chromatin profiling using targeted DNA adenine methyltransferase. Nature genetics 2001, 27(3):304-308.
36

36. Orian A: Chromatin profiling, DamID and the emerging landscape of gene expression. Curr Opin Genet Dev 2006, 16(2):157-164.
37. Reddy KL, Zullo JM, Bertolino E, Singh H: Transcriptional repression mediated by repositioning of genes to the nuclear lamina. Nature 2008, 452(7184):243-247.
38. Kumaran RI, Spector DL: A genetic locus targeted to the nuclear periphery in living cells maintains its transcriptional competence. The Journal of cell biology 2008, 180(1):51-65.
39. Taddei A, Gasser SM: Multiple pathways for telomere tethering: functional implications of subnuclear position for heterochromatin formation. Biochimica et biophysica acta 2004, 1677(1-3):120-128.
40. Taddei A, Van Houwe G, Nagai S, Erb I, van Nimwegen E, Gasser SM: The functional importance of telomere clustering: global changes in gene expression result from SIR factor dispersion. Genome research 2009, 19(4):611-625.
41. Strambio-De-Castillia C, Niepel M, Rout MP: The nuclear pore complex: bridging nuclear transport and gene regulation. Nature reviews Molecular cell biology 2010, 11(7):490-501.
42. Levy DL, Heald R: Nuclear size is regulated by importin alpha and Ntf2 in Xenopus. Cell 2010, 143(2):288-298.
43. Feric M, Brangwynne CP: A nuclear F-actin scaffold stabilizes ribonucleoprotein droplets against gravity in large cells. Nature cell biology 2013, 15(10):1253-1259.
37

44. Akhtar A, Gasser SM: The nuclear envelope and transcriptional control. Nature reviews Genetics 2007, 8(7):507-517.
45. Dechat T, Pfleghaar K, Sengupta K, Shimi T, Shumaker DK, Solimando L, Goldman RD: Nuclear lamins: major factors in the structural organization and function of the nucleus and chromatin. Genes & development 2008, 22(7):832-853.
46. Aaronson RP, Blobel G: Isolation of nuclear pore complexes in association with a lamina. Proceedings of the National Academy of Sciences of the United States of America 1975, 72(3):1007-1011.
47. Hutchison CJ, Worman HJ: A-type lamins: guardians of the soma? Nature cell biology 2004, 6(11):1062-1067.
48. Dechat T, Adam SA, Taimen P, Shimi T, Goldman RD: Nuclear lamins. Cold Spring Harbor perspectives in biology 2010, 2(11):a000547.
49. Mason JM, Arndt KM: Coiled coil domains: stability, specificity, and biological implications. Chembiochem : a European journal of chemical biology 2004, 5(2):170-176.
50. Hanukoglu I, Ezra L: Proteopedia: Coiled-coil structure of keratins. Biochemistry and molecular biology education : a bimonthly publication of the International Union of Biochemistry and Molecular Biology 2013.
51. Strelkov SV, Schumacher J, Burkhard P, Aebi U, Herrmann H: Crystal structure of the human lamin A coil 2B dimer: implications for the headto-tail association of nuclear lamins. Journal of molecular biology 2004, 343(4):1067-1080.
38

52. Hirano T: At the heart of the chromosome: SMC proteins in action. Nature reviews Molecular cell biology 2006, 7(5):311-322.
53. McLachlan AD, Stewart M: Tropomyosin coiled-coil interactions: evidence for an unstaggered structure. Journal of molecular biology 1975, 98(2):293304.
54. Gruenbaum Y, Lee KK, Liu J, Cohen M, Wilson KL: The expression, lamindependent localization and RNAi depletion phenotype for emerin in C. elegans. Journal of cell science 2002, 115(Pt 5):923-929.
55. Furukawa K, Hotta Y: cDNA cloning of a germ cell specific lamin B3 from mouse spermatocytes and analysis of its function by ectopic expression in somatic cells. The EMBO journal 1993, 12(1):97-106.
56. Butin-Israeli V, Ben-nun-Shaul O, Kopatz I, Adam SA, Shimi T, Goldman RD, Oppenheim A: Simian virus 40 induces lamin A/C fluctuations and nuclear envelope deformation during cell entry. Nucleus 2011, 2(4):320330.
57. Jenkins H, Holman T, Lyon C, Lane B, Stick R, Hutchison C: Nuclei that lack a lamina accumulate karyophilic proteins and assemble a nuclear matrix. Journal of cell science 1993, 106 ( Pt 1):275-285.
58. Butin-Israeli V, Adam SA, Goldman AE, Goldman RD: Nuclear lamin functions and disease. Trends in genetics : TIG 2012, 28(9):464-471.
59. Taddei A, Gasser SM: Structure and function in the budding yeast nucleus. Genetics 2012, 192(1):107-129.
39

60. Rout MP, Field MC: Isolation and characterization of subnuclear compartments from Trypanosoma brucei. Identification of a major repetitive nuclear lamina component. The Journal of biological chemistry 2001, 276(41):38261-38271.
61. DuBois KN, Alsford S, Holden JM, Buisson J, Swiderski M, Bart JM, Ratushny AV, Wan Y, Bastin P, Barry JD et al: NUP-1 Is a large coiled-coil nucleoskeletal protein in trypanosomes with lamin-like functions. PLoS biology 2012, 10(3):e1001287.
62. Rose A, Patel S, Meier I: Plant nuclear envelope proteins. Symposia of the Society for Experimental Biology 2004(56):69-88.
63. Graumann K, Runions J, Evans DE: Characterization of SUN-domain proteins at the higher plant nuclear envelope. The Plant journal : for cell and molecular biology 2010, 61(1):134-144.
64. Oda Y, Fukuda H: Dynamics of Arabidopsis SUN proteins during mitosis and their involvement in nuclear shaping. The Plant journal : for cell and molecular biology 2011, 66(4):629-641.
65. Zhou X, Graumann K, Evans DE, Meier I: Novel plant SUN-KASH bridges are involved in RanGAP anchoring and nuclear shape determination. J Cell Biol 2012, 196(2):203-211.
66. Janski N, Masoud K, Batzenschlager M, Herzog E, Evrard JL, Houlne G, Bourge M, Chaboute ME, Schmit AC: The GCP3-interacting proteins GIP1 and GIP2 are required for gamma-tubulin complex protein localization,
40

spindle integrity, and chromosomal stability. The Plant cell 2012, 24(3):1171-1187. 67. Tamura K, Iwabuchi K, Fukao Y, Kondo M, Okamoto K, Ueda H, Nishimura M, Hara-Nishimura I: Myosin XI-i links the nuclear membrane to the cytoskeleton to control nuclear movement and shape in Arabidopsis. Current biology : CB 2013, 23(18):1776-1781. 68. Brkljacic J, Zhao Q, Meier I: WPP-domain proteins mimic the activity of the HSC70-1 chaperone in preventing mistargeting of RanGAP1anchoring protein WIT1. Plant Physiol 2009, 151(1):142-154. 69. Sattarzadeh A, Schmelzer E, Hanson MR: Arabidopsis myosin XI subdomains homologous to the yeast myo2p organelle inheritance sub-domain target subcellular structures in plant cells. Frontiers in plant science 2013, 4:407. 70. Xu XM, Rose A, Muthuswamy S, Jeong SY, Venkatakrishnan S, Zhao Q, Meier I: NUCLEAR PORE ANCHOR, the Arabidopsis homolog of Tpr/Mlp1/Mlp2/megator, is involved in mRNA export and SUMO homeostasis and affects diverse aspects of plant development. The Plant cell 2007, 19(5):1537-1548. 71. Braud C, Zheng W, Xiao W: LONO1 encoding a nucleoporin is required for embryogenesis and seed viability in Arabidopsis. Plant physiology 2012, 160(2):823-836. 72. Cheng YT, Germain H, Wiermer M, Bi D, Xu F, Garcia AV, Wirthmueller L, Despres C, Parker JE, Zhang Y et al: Nuclear pore complex component
41

MOS7/Nup88 is required for innate immunity and nuclear accumulation of defense regulators in Arabidopsis. The Plant cell 2009, 21(8):2503-2516. 73. Zhang Y, Li X: A putative nucleoporin 96 Is required for both basal defense and constitutive resistance responses mediated by suppressor of npr1-1,constitutive 1. The Plant cell 2005, 17(4):1306-1316. 74. Wiermer M, Cheng YT, Imkampe J, Li M, Wang D, Lipka V, Li X: Putative members of the Arabidopsis Nup107-160 nuclear pore sub-complex contribute to pathogen defense. The Plant journal : for cell and molecular biology 2012, 70(5):796-808. 75. Masuda K, Xu ZJ, Takahashi S, Ito A, Ono M, Nomura K, Inoue M: Peripheral framework of carrot cell nucleus contains a novel protein predicted to exhibit a long alpha-helical domain. Experimental cell research 1997, 232(1):173-181. 76. Kimura Y, Kuroda C, Masuda K: Differential nuclear envelope assembly at the end of mitosis in suspension-cultured Apium graveolens cells. Chromosoma 2010, 119(2):195-204. 77. Ciska M, Masuda K, Moreno Diaz de la Espina S: Lamin-like analogues in plants: the characterization of NMCP1 in Allium cepa. Journal of experimental botany 2013, 64(6):1553-1564. 78. Dittmer TA, Stacey NJ, Sugimoto-Shirasu K, Richards EJ: LITTLE NUCLEI genes affecting nuclear morphology in Arabidopsis thaliana. The Plant cell 2007, 19(9):2793-2803.
42

79. Ciska M, Moreno S: NMCP/LINC proteins: Putative lamin analogs in plants? Plant signaling & behavior 2013, 8(12).
80. Gardiner J, Overall R, Marc J: Putative Arabidopsis homologues of metazoan coiled-coil cytoskeletal proteins. Cell biology international 2011, 35(8):767-774.
81. Sakamoto Y, Takagi S: LITTLE NUCLEI 1 and 4 regulate nuclear morphology in Arabidopsis thaliana. Plant & cell physiology 2013, 54(4):622-633.
82. Rudd S, Frisch M, Grote K, Meyers BC, Mayer K, Werner T: Genome-wide in silico mapping of scaffold/matrix attachment regions in Arabidopsis suggests correlation of intragenic scaffold/matrix attachment regions with gene expression. Plant physiology 2004, 135(2):715-722.
83. Tetko IV, Haberer G, Rudd S, Meyers B, Mewes HW, Mayer KF: Spatiotemporal expression control correlates with intragenic scaffold matrix attachment regions (S/MARs) in Arabidopsis thaliana. PLoS computational biology 2006, 2(3):e21.
84. Van der Geest AH, Welter ME, Woosley AT, Pareddy DR, Pavelko SE, Skokut M, Ainley WM: A short synthetic MAR positively affects transgene expression in rice and Arabidopsis. Plant biotechnology journal 2004, 2(1):13-26.
85. Meier I, Phelan T, Gruissem W, Spiker S, Schneider D: MFP1, a novel plant filament-like protein with affinity for matrix attachment region DNA. The Plant cell 1996, 8(11):2105-2115.
43

86. Jeong SY, Rose A, Meier I: MFP1 is a thylakoid-associated, nucleoidbinding protein with a coiled-coil structure. Nucleic acids research 2003, 31(17):5175-5185.
87. Gindullis F, Meier I: Matrix attachment region binding protein MFP1 is localized in discrete domains at the nuclear envelope. The Plant cell 1999, 11(6):1117-1128.
88. Gindullis F, Peffer NJ, Meier I: MAF1, a novel plant protein interacting with matrix attachment region binding protein MFP1, is located at the nuclear envelope. The Plant cell 1999, 11(9):1755-1768.
89. Xiao C, Chen F, Yu X, Lin C, Fu YF: Over-expression of an AT-hook gene, AHL22, delays flowering and inhibits the elongation of the hypocotyl in Arabidopsis thaliana. Plant molecular biology 2009, 71(1-2):39-50.
90. Yun J, Kim YS, Jung JH, Seo PJ, Park CM: The AT-hook motif-containing protein AHL22 regulates flowering initiation by modifying FLOWERING LOCUS T chromatin in Arabidopsis. The Journal of biological chemistry 2012, 287(19):15307-15316.
91. Xu Y, Gan ES, He Y, Ito T: Flowering and genome integrity control by a nuclear matrix protein in Arabidopsis. Nucleus 2013, 4(4):274-276.
92. Xu Y, Gan ES, Ito T: The AT-hook/PPC domain protein TEK negatively regulates floral repressors including MAF4 and MAF5. Plant signaling & behavior 2013, 8(8).
93. Xu Y, Wang Y, Stroud H, Gu X, Sun B, Gan ES, Ng KH, Jacobsen SE, He Y, Ito T: A matrix protein silences transposons and repeats through
44

interaction with retinoblastoma-associated proteins. Current biology : CB 2013, 23(4):345-350. 94. Kandasamy MK, Deal RB, McKinney EC, Meagher RB: Silencing the nuclear actin-related protein AtARP4 in Arabidopsis has multiple effects on plant development, including early flowering and delayed floral senescence. The Plant journal : for cell and molecular biology 2005, 41(6):845-858. 95. Kandasamy MK, McKinney EC, Deal RB, Smith AP, Meagher RB: Arabidopsis actin-related protein ARP5 in multicellular development and DNA repair. Developmental biology 2009, 335(1):22-32. 96. Choi K, Kim S, Kim SY, Kim M, Hyun Y, Lee H, Choe S, Kim SG, Michaels S, Lee I: SUPPRESSOR OF FRIGIDA3 encodes a nuclear ACTINRELATED PROTEIN6 required for floral repression in Arabidopsis. The Plant cell 2005, 17(10):2647-2660. 97. Martin-Trillo M, Lazaro A, Poethig RS, Gomez-Mena C, Pineiro MA, Martinez-Zapater JM, Jarillo JA: EARLY IN SHORT DAYS 1 (ESD1) encodes ACTIN-RELATED PROTEIN 6 (AtARP6), a putative component of chromatin remodelling complexes that positively regulates FLC accumulation in Arabidopsis. Development 2006, 133(7):1241-1252. 98. Kandasamy MK, McKinney EC, Deal RB, Meagher RB: Arabidopsis ARP7 is an essential actin-related protein required for normal embryogenesis, plant architecture, and floral organ abscission. Plant physiology 2005, 138(4):2019-2032.
45

99. Meagher RB, Kandasamy MK, Smith AP, McKinney EC: Nuclear actinrelated proteins at the core of epigenetic control. Plant signaling & behavior 2010, 5(5):518-522.
100. Schubert V: SMC proteins and their multiple functions in higher plants. Cytogenetic and genome research 2009, 124(3-4):202-214.
101. Schubert V, Klatte M, Pecinka A, Meister A, Jasencakova Z, Schubert I: Sister chromatids are often incompletely aligned in meristematic and endopolyploid interphase nuclei of Arabidopsis thaliana. Genetics 2006, 172(1):467-475.
102. Watanabe K, Pacher M, Dukowic S, Schubert V, Puchta H, Schubert I: The STRUCTURAL MAINTENANCE OF CHROMOSOMES 5/6 complex promotes sister chromatid alignment and homologous recombination after DNA damage in Arabidopsis thaliana. The Plant cell 2009, 21(9):26882699.
103. Kanno T, Bucher E, Daxinger L, Huettel B, Bohmdorfer G, Gregor W, Kreil DP, Matzke M, Matzke AJ: A structural-maintenance-of-chromosomes hinge domain-containing protein is required for RNA-directed DNA methylation. Nature genetics 2008, 40(5):670-675.
104. van Zanten M, Koini MA, Geyer R, Liu Y, Brambilla V, Bartels D, Koornneef M, Fransz P, Soppe WJ: Seed maturation in Arabidopsis thaliana is characterized by nuclear size reduction and increased chromatin condensation. Proceedings of the National Academy of Sciences of the United States of America 2011, 108(50):20219-20224.
46

105. Tessadori F, Schulkes RK, van Driel R, Fransz P: Light-regulated large-scale reorganization of chromatin during the floral transition in Arabidopsis. The Plant journal : for cell and molecular biology 2007, 50(5):848-857.
106. Fransz P, De Jong JH, Lysak M, Castiglione MR, Schubert I: Interphase chromosomes in Arabidopsis are organized as well defined chromocenters from which euchromatin loops emanate. Proceedings of the National Academy of Sciences of the United States of America 2002, 99(22):1458414589.
107. Pikaard CS, Haag JR, Ream T, Wierzbicki AT: Roles of RNA polymerase IV in gene silencing. Trends in plant science 2008, 13(7):390-397.
108. Probst AV, Fagard M, Proux F, Mourrain P, Boutet S, Earley K, Lawrence RJ, Pikaard CS, Murfett J, Furner I et al: Arabidopsis histone deacetylase HDA6 is required for maintenance of transcriptional gene silencing and determines nuclear organization of rDNA repeats. The Plant cell 2004, 16(4):1021-1034.
109. Tessadori F, van Zanten M, Pavlova P, Clifton R, Pontvianne F, Snoek LB, Millenaar FF, Schulkes RK, van Driel R, Voesenek LA et al: Phytochrome B and histone deacetylase 6 control light-induced chromatin compaction in Arabidopsis thaliana. PLoS genetics 2009, 5(9):e1000638.
110. Pecinka A, Dinh HQ, Baubec T, Rosa M, Lettner N, Mittelsten Scheid O: Epigenetic regulation of repetitive elements is attenuated by prolonged heat stress in Arabidopsis. The Plant cell 2010, 22(9):3118-3129.
47

CHAPTER TWO
ARABIDOPSIS CROWDED NUCLEI (CRWN) PROTEINS ARE REQUIRED FOR NUCLEAR SIZE CONTROL AND HETEROCHROMATIN ORGANIZATION
Abstract
Background Plant nuclei superficially resemble animal and fungal nuclei, but the machinery and processes that underlie nuclear organization in these eukaryotic lineages appear to be evolutionarily distinct. Among the candidates for nuclear architectural elements in plants are coiled-coil proteins in the NMCP (Nuclear Matrix Constituent Protein) family. Using genetic and cytological approaches, we dissect the function of the four NMCP family proteins in Arabidopsis encoded by the CRWN genes, which were originally named LINC (LITTLE NUCLEI).
Results CRWN proteins are essential for viability as evidenced by the inability to recover mutants that have disruptions in all four CRWN genes. Mutants deficient in different combinations of the four CRWN paralogs exhibit altered nuclear organization, including reduced nuclear size, aberrant nuclear shape and abnormal spatial organization of constitutive heterochromatin. Our results demonstrate functional diversification among CRWN paralogs; CRWN1 plays the predominant role in control
48

of nuclear size and shape followed by CRWN4. Proper chromocenter organization is most sensitive to the deficiency of CRWN4. The reduction in nuclear volume in crwn mutants in the absence of a commensurate reduction in endoreduplication levels leads to an increase in average nuclear DNA density.
Conclusions Our findings indicate that CRWN proteins are important architectural components of plant nuclei that play diverse roles in both heterochromatin organization and the control of nuclear morphology.
Background
The cellular components and processes that specify nuclear size, shape and internal organization are poorly understood, particularly in flowering plants. Despite similarities at the gross morphological level among all eukaryotic nuclei, such as a double-membrane boundary perforated with nuclear pores, most of the proteins known to affect nuclear structure in animals are not evolutionarily conserved and are consequently difficult to recognize or absent entirely in plant proteomes [111-113]. These observations indicate that the machinery, and perhaps the principles, specifying nuclear organization in flowering plants are distinct from those operating in animals and represent a convergent evolutionary path to a canonical nuclear organization in eukaryotic cells [114].
49

We demonstrated previously [78, 115] that two paralogous Arabidopsis coiled-coil proteins, originally named LITTLE NUCLEI 1 and 2 (LINC1 & 2), play important roles in specifying nuclear shape and size. Supporting this conclusion, Sakamoto and Takagi recently reported that disruption of LINC4, another of the four paralogous genes in this family, leads to reduced nuclear size and loss of elongated nuclear shape in differentiated cells, mirroring the phenotype of linc1 mutants [19]. These proteins are closely related to NMCP1, Nuclear Matrix Constituent Protein 1, originally identified as a protein residing on the periphery of carrot nuclei and a component of the salt-resistant nuclear matrix [116]. Although NMCP1 and related proteins are plant-specific and share no significant amino acid similarity to lamins, their tripartite structure with an extensive central coiled-coil domain and their localization at the nuclear periphery suggest that NMCP1-related plant proteins might be functional analogs of this core component of the animal nuclear lamina [76, 77]. More recent computational analysis [80], however, has suggested that the NMCP class of plant proteins shares more structural similarities to myosins or paramyosins than to lamins.
Here, we extend our reverse genetic analysis to encompass all four members of the Arabidopsis NMCP-related protein family, which we have renamed CRWN (CROWDED NUCLEI) to avoid confusion with the acronym LINC (LINKER of NUCLEOSKELETON and CYTOSKELETON) that refers to SUN-KASH protein linkages that bridge the inner and outer nuclear membrane [65, 117-120]. Our findings demonstrate that CRWN proteins are essential for viability, and our analyses uncover complex functional diversification among CRWN proteins with regards to
50

their effects on whole-plant morphology, nuclear size, and the spatial organization of constitutive heterochromatin aggregates (chromocenters) in interphase nuclei. We found that CRWN1 plays the most prominent role among CRWN paralogs in controlling nuclear size, while CRWN4 has the most important role in controlling the distribution and number of heterochromatic chromocenters. The reduced nuclear size in crwn mutants is not matched by a commensurate reduction in endopolyploid levels, resulting in increased nuclear DNA densities (mass per unit volume) up to four-fold higher than wild type levels.
Results
We performed a phylogenetic analysis of Arabidopsis CRWN proteins and their homologues in other species to begin our investigation of the potential diversification within this family. The predicted Arabidopsis proteome contains four closely related CRWN proteins (CRWN1 through 4) that share 30-40% amino acid identity; no other Arabidopsis proteins with extended regions of significant amino acid identity to CRWN proteins were found. Similar proteins were found in other plant species, but interestingly, no fungi or animal CRWN homologues were identified from searches of protein databases. In addition, CRWN homologues were absent in the predicted proteome of the green algae Chlamydomonas and Volvox.
We constructed a phylogram of CRWN proteins and related plant homologs using a maximum likelihood algorithm (Figure 2.1 and Supplementary Table 2.1). The tree
51

features two major clades distinct from CRWN homologues in two basal plants, Selaginella moellendorffii and Physcomitrella patens. One clade includes three of the Arabidopsis paralogs, CRWN1, CRWN2 and CRWN3, while CRWN4 belongs to the other clade. Within each clade, the monocot proteins, represented by maize, sorghum and rice, group independently from the dicot proteins. Only two CRWN paralogs exist in these monocots – one CRWN1-like and one CRWN4-like. However, certain dicot species, such as Arabidopsis, poplar, grape, and castor bean, contain multiple copies of CRWN1-like proteins. The dicot CRWN4-like proteins are also distinct from their monocot counterparts in lacking a conserved amino acid motif at the extreme C-terminus (yellow inset in Figure 2.1 and Supplementary Figure 2.1).
Genetic redundancy in the CRWN family The inference that CRWN4 and related proteins are divergent from members of the CRWN1-containing clade was supported by genetic analyses to dissect the functions of the CRWN paralogs. We used Agrobacterium T-DNA insertion alleles to study the effects of inactivating different combinations of CRWN genes [121]. Previously, we demonstrated that the crwn1-1 and crwn2-1 T-DNA alleles severely reduce or eliminate transcription downstream of the T-DNA insertion [78]. Here, we performed transcript analysis by RT-PCR for the crwn3-1 and crwn4-1 alleles used in this study (Supplementary Figure 2.2). For crwn3-1, some transcript was detected downstream of the insertion; however, no transcript could be detected using primers that flanked the insertion. The crwn4-1 insertion blocked transcription downstream of the T-DNA.
52

Oryza 1 Zea 1
Sorghum 1

998

Daucus NMCP1

Apium NMCP1

Ricinus 1 Vitis 1

Populus 1 A. lyrata 1

741 992 505 778 1000 863 896

CRWN1 Vitis 2

792 999 506

Ricinus 2

1000 998

Populus 2

1000 1000

CRWN2 997 CRWN2S A. lyrata 2
A. lyrata 3 CRWN3

A. lyrata 4 CRWN4

1000

Populus 3

Vitis 3

Ricinus 3

420 861

994 651
1000

Apium NMCP2
Daucus NMCP2

1000 970

1000

Zea 2 979
Sorghum 2
Oryza 2

1000

Selaginella

Physco 1 Physco 2

Figure 2.1 Phylogenetic relationships among CRWN proteins. Figure 2.1 A maximum likelihood tree of CRWN homologs constructed from an alignment of amino acid sequences that correspond to the coiled-coil domains. Bootstrap values (of 1000 replicates) are indicated on each branch. The A. thaliana CRWN proteins are indicated in bold, and homologs are labeled with the genus name and an assigned number (Supplementary Table 2.1). Two major clades are marked by blue and green; the yellow inset oval indicates a subgroup of CRWN4-like proteins from dicots that lack the conserved C-terminal domain (Supplementary Figure 2.1).

53

Supplementary Table 2.1 CRWN-like proteins used in this study
Table S1

Protein Physco 1 Physco 2
Selaginella
Apium NMCP1 Apium NMCP2 Daucus NMCP1 Daucus NMCP2 Ricinus 1 Ricinus 2 Ricinus 3 Vitis 1 Vitis 2 Vitis 3 Populus 1 Populus 2 Populus 3 CRWN1 CRWN2 CRWN2S CRWN3 CRWN4 A_lyrata 1 A_lyrata 2 A_lyrata 3 A_lyrata 4 Zea 1 Zea 2 Oryza 1 Oryza 2 Sorghum 1 Sorghum 2

Species name Physcomitrella patens Selaginella moellendorffii Apium graveolens
Daucus carota
Ricinus communis
Vitis vinifera
Populus trichocarpa
Arabidopsis thaliana
Arabidopsis lyrata
Zea mays Oryza sativa (Japonica) Sorghum bicolor

Locus name or Sequence ID Pp1s200_64V6.1 Pp1s76_81V6.1
XP_002993584.1
BAF64421.1 BAI67716.1 BAA20407.1 BAI67718.1 XP_002525969.1 XP_002524388.1 XP_002530596.1 CAO49297.1 (GSVIVG01031076001) CAN74873.1 (GSVIVT01011972001) CAO17747.1 (GSVIVT01007428001) XP_002329317.1 (Potri.017G111400.2) XP_002312375.1 (Potri.008G114800.1) XP_002317738.1 (Potri.012G034300.1) At1g67230.1 At1g13220.2 At1g13220.1 At1g68790.1 At5g65770.1 scaffold_2:12,299,939..12,304,385 471477 476006 scaffold_803322.1.1 ZEAMMB73_827243 (AFW63577.1) ZEAMMB73_204423 (DAA57458.1) Os02g0709900 (NP_001047893.1) Os01g0767000 (NP_001044359.1) Sb04g030240.1 Sb03g035670.2

Database source* JGI
GenBank
GenBank
GenBank
GenBank & (JGI)
GenBank & (JGI)
TAIR
JGI JGI JGI Ensembl Genomes Genbank Genbank JGI

* JGI: www.phytozome.net GenBank: www.ncbi.nlm.nih.gov/genbank/ TAIR: www.arabidopsis.org Ensembl Genomes: plants.ensembl.org/Arabidopsis_lyrata/

Supplementary Table 2.1. The first column shows the abbreviated name of the protein used in alignment to construct the tree shown in Figure 2.1. The remaining columns indicate the identity and the source of each protein sequence.

54

basal
CRWN1 clade proteins
Monocot CRWN4 like proteins Dicot CRWN4 like proteins
Supplementary Figure 2.1 Amino acid sequences comprising the exreme C termini of 28 CRWN-like proteins, including ten CRWN4-like proteins
Supplementary Figure 2.1 The similarity in this region, which falls outside of the coiled-coil domains, reinforces the topology of the tree shown in Figure 2.1. All of the proteins within the CRWN1-like clade, as well as the Physcomitrella homologs, contain a conserved C-terminal motif and a group of acidic residues approximately 25 amino acids from the end of the protein. Monocot CRWN4-like proteins contain a region with similar features but these conserved motifs are absent in CRWN4-like proteins from dicots (denoted by the yellow oval in Figure 2.1).
55

wild type crwn3 H 2O wild type crwn3 H 2O H 2O SSwwiillAAddLLttKKyy__pp00cceewwrr09iill99ww92ddnn28tt8333yy3ppee SSwwiillAAddLLttKKyy__pp00cceewwrr99iill99ww22ddnn88tt3333yyppee

wild type crwn3 H 2O wild type crwn3 H 2O H 2O SSwwiillAAddLLttKKyy__pp00ee779922SSww99iiccllww66AArriiddllLLwwttddKKyynntt__pp44yy00eepp77ee99229966 SSwwiillAAddLLttKKyy__pp00ee779922SSww99iiccllww66AArriiddllLLwwttddKKyynntt__pp44yy00eepp77ee99229966

A
Cyclophilin F & R primers M +RT +RT +RT -RT -RT -RT

B AActin2F & 2R primers -RT -RT +RT +RT

At1g68790 F & R primers -RT -RT +RT +RT

Actin2 F & R primers

At1g68790 F & R primers

Actin2F & 2R primers

At1g68790 F & R primers

-R-TRT-R-TRT +R+TRT+R+TRT M --RRTT --RRTT ++RRTT ++RRTT

At1g68790 F2 & R2 primers, spanning the insertion site M +RT +RT +RT -RT -RT -RT

B C
B

AAcctitnin22FF && R2Rprpimrimeresrs --RARTcTtin-2-RRFTT& 2++RRRTTprim++RReTTrs
-RT -RT +RT +RT

AtA5tg56g56757700FF&&RRpprirmimeerrss M At-5Rg-TR65T7-R7-0TRTF +&R+RRTTp+riRm+TReTrs
-RT -RT +RT +RT

Gene/GI
CRWN3 At1g68790
CRWN4 At5g65770 Actin 2
Cyclophilin

Allele/T-DNA
crwn3-1 SALK_099283
crwn4-1 SALK_079296

Insertion Site

RT-PCR Primers

Position of primers

6th exon

F 5 -AGTGAACAGGCAGCTGGTGATAGT-3 R 5 -ACTTCCAACTGCGGATCTTCGACT-3
F2 5 –TCTCCTTCACGGTTTTGAGC–3 R2 5 –GAGAAGCACATGAGGCAGTGT–3

the 6th exon the 8th exon
the 6th exon the 4th exon

6th exon

F 5 -TCGCTAAACCGAGAGCGTGAAGAA-3 R 5 -TTGGTCACCTCTGTCTCACACGTT-3
F 5'-TGATATTCAACCAATCGTGTGTGAC-3' R 5'-AAGCAAGAATGGAACCACCGATCC-3'
F 5 –CGATAAGACTCCCAGGACTGCCGA–3 R 5 –TCGGCTTTCCAGATGATGATCCAACC–3

the 6th exon the 7th exon
the 1st exon the 2nd exon
the only exon the only exon

Supplementary Figure 2.2 Transcript analysis of the crwn3-1 and crwn4-1 alleles used in this study

56

Supplementary Figure 2.2 Reverse transcription-PCR results investigating the effect of T-DNA insertions on the transcription of CRWN3 and CRWN4. Panel A shows that a CRWN3 transcript is produced from the wild-type allele but not from the crwn3-1 allele using primers spanning the T-DNA insertion site. Panel B demonstrates that some transcription can be detected downstream of the insertion site from the crwn3-1 allele using RT-PCR and a primer set recognizing sequences 3’ of the insertion site. Panel C indicates that the T-DNA insertion in the crwn4-1 allele blocks transcription. Amplification of cDNA from cyclophilin and Actin2 were used as positive controls. M, marker lanes; + RT (plus reverse transcriptase); - RT (no reverse transcriptase). Information on the oligonucleotide primers used in these experiments is shown at the bottom of the figure. Our previous data [78] indicated that the crwn1-1 and crwn2-1 alleles block transcription downstream of the T-DNA insertion site in the sixth exon of both genes.
57

The lack of full-length CRWN transcripts from homozygous mutant lines indicates that all four mutations used in this study are likely to be loss-of-function alleles. We note that the CRWN genes have similar developmental gene expression patterns: the steadystate abundance of transcripts for all four paralogs peak in proliferating tissues (http://bar.utoronto.ca/efp/cgi-bin/efpWeb.cgi) [122].
Mutant plants carrying single insertions were intercrossed and progeny carrying homozygous insertions in different combinations were recovered. Figure 2.2 shows the whole-plant phenotype of the viable mutants at the rosette stage just after the transition to flowering. Plants carrying a mutation in any single CRWN gene had phenotypes similar to wild-type Columbia plants, as did the double crwn2 crwn3 and crwn3 crwn4 mutants. The crwn2 crwn4 and crwn1 crwn4 double mutants exhibited slightly smaller rosettes, while the remaining double mutants, crwn1 crwn2 and crwn1 crwn3, displayed markedly smaller rosette sizes. We were able to recover only two of the four triple mutants - crwn1 crwn2 crwn4 and crwn1 crwn3 crwn4, both of which were extremely stunted and set few seed.
58

crwn1

crwn2

wild type

crwn1 crwn2

crwn1 crwn3

crwn1 crwn4

crwn1 crwn2 crwn4

crwn3

crwn4

crwn2 crwn3

crwn2 crwn4

crwn3 crwn4

crwn1 crwn3 crwn4

Figure 2.2 Whole plant phenotypes of crwn mutants. Figure 2.2 Leaf rosette structure of one month-old plants imaged just after initiation of flowering. Representative plants for the various genotypes are compared to a wild type (WT) Columbia plant. All plants were grown in parallel and photographed at the same magnification; the diameter of a WT rosette at flowering is ca. 8 cm.

59

Our inability to isolate a mutant combining alleles in all four CRWN genes indicates that at least one functional CRWN protein is required for viability. Triple mutant plants carrying only CRWN2 or CRWN3 were extremely stunted, but still viable. This result suggests that CRWN2 or CRWN3 alone can cover the minimum requirements for the entire CRWN protein family. However, plants carrying only CRWN1 or only CRWN4 were not recovered, suggesting that CRWN1 and CRWN4 are specialized and that neither protein alone can express the full range of functions of the CRWN protein family.
CRWN proteins are required to maintain proper nuclear size and shape We next observed crwn mutant nuclei from adult leaf tissue to determine the role of different CRWN proteins in specifying nuclear size and shape. Among the single mutants, a deficiency of CRWN1 or CRWN4 reduced nuclear size (Figures 2.3 and 2.4A; Supplementary Table 2.2), while loss of CRWN2 or CRWN3 had no effect. Combining a crwn1 mutation with a crwn2 or crwn3 mutation had a synergistic effect on nuclear size, suggesting that CRWN1 function overlaps, at least partially, with those of CRWN2 and CRWN3. Double mutant combinations containing crwn4 and either crwn2 or crwn3 did not show additive phenotypes but rather resembled crwn4. In contrast, combination of a crwn1 with a crwn4 mutation had an additive effect on nuclear size. These findings indicate that CRWN1 and CRWN4 are the major determinants of nuclear size among the CRWN paralogs. Further, the additive effects of crwn1 and crwn4 mutations suggest CRWN4 acts independently from CRWN1,
60

consistent with their distinct phylogenetic grouping (Figure 2.1) and the genetic analysis shown in Figure 2.2.
CRWN proteins are required for development or maintenance of the elongated spindle shapes which characterize larger nuclei in differentiated wild-type cells [123]. We previously reported that a deficiency of CRWN1 causes nuclei in all cells to adopt the spherical shape characteristic of proliferating tissue at root and shoot apices [78]. The present study confirmed the importance of CRWN1 for nuclear shape differentiation and also uncovered a similar role for CRWN4 (Figure 2.3), a conclusion also reached recently by Sakamoto and Takagi [81]. Nuclei from crwn4 leaf tissue often have irregular margins and are more spherically shaped, compared to wild-type nuclei (Supplementary Figure 2.3). However, crwn4 nuclei are less uniformly round in comparison to crwn1 nuclei, particularly larger crwn4 nuclei. Further, crwn4 nuclei occasionally contain thin projections that appear to be drawn from the nuclear surface (arrowheads in Figure 2.3).
61

Supplementary Table 2.2 The nuclear phenotype data for crwn mutants used to construct Figure 2.4

ave. ploidy level fraction of wt ave. nuclear size (µm2) fraction of wt

wild type crwn1 crwn2 crwn3 crwn4 crwn1 crwn2 crwn1 crwn3 crwn1 crwn4 crwn2 crwn3 crwn2 crwn4 crwn3 crwn4 crwn1 crwn2 crwn4 crwn1 crwn3 crwn4

9.74 9.52 8.42 8.41 9.52 6.91 7.92 8.40 8.85 7.43 8.66 6.31 5.16

1.00 0.98 0.86 0.86 0.98 0.71 0.81 0.86 0.91 0.76 0.89 0.65 0.59

73.0 + 1.18 35.7 + 1.40 72.1 + 2.71 69.8 + 3.43 39.3 + 2.20 11.8 + 0.78 26.5 + 1.14 27.8 + 1.09 72.9 + 3.60 42.5 + 1.43 38.1 + 1.24 10.3 + 0.31 11.2 + 0.51

1.00 0.49 0.99 0.96 0.54 0.16 0.36 0.38 1.00 0.58 0.52 0.14 0.15

Supplementary Table 2.2 The nuclear phenotype data for crwn mutants used to construct Figure 2.4 is displayed in tabular form. The average endopolyploid level (ave. ploidy level) was determined by flow cytometry as described in Methods. The actual measurements were converted to relative measurements (fraction of wt, third column) using the wild type (wt) values for normalization. The average nuclear size (ave. nuclear size ± standard error of the mean) corresponds to data from Figure 2.4A. The fifth column normalizes these size values to the wild type values.

62

wild type

crwn2

crwn3

crwn2 crwn3

5 µm
crwn1

crwn1 crwn2

crwn1 crwn3 crwn1 crwn4

crwn4

crwn2 crwn4

crwn3 crwn4

crwn1 crwn2 crwn4

crwn1 crwn3 crwn4

Figure 2.3 Nuclear phenotypes of crwn mutants. Figure 2.3 Images of representative DAPI-stained adult leaf cell nuclei from wild type and the twelve viable crwn mutants. The top row contains the wild type control and crwn2, crwn3 and crwn2 crwn3 mutants with normal nuclear morphology phenotypes. The middle row shows nuclear phenotypes of crwn1 and double mutants containing a crwn1 mutation. The bottom row displays nuclear phenotypes from crwn4 mutants, as well as higher-order mutants containing a crwn4 mutation. The arrowheads highlight thin projections from crwn4 nuclei. A 5 µm size bar is shown in the upper left inset; all images are shown at the same magnification.
63

Loss of CRWN proteins affects nuclear DNA packing density The direct correlation between endopolyploidy and nuclear size in wild-type Arabidopsis cells [124] prompted us to examine this relationship within the crwn mutants. We measured the average endopolyploidy level of nuclei from the same adult leaves harvested for the nuclear size analysis shown in Figure 2.4A (see also Supplementary Table 2.2).
Some crwn mutants showed a decrease in endopolyploid levels, particularly the crwn triple mutants and the crwn1 crwn2 double mutant, but the remaining crwn genotypes had average endopolyploidy levels that approached wild-type levels (Figure 2.4B). The dashed line in Figure 2.4B depicts the expected nuclear size change in response to a reduction in endopolyploidy based on the established one-to-one relationship between nuclear volume (approximated by nuclear area in our measurements of isolated and flattened leaf cell nuclei, Supplementary Figure 2.4) and DNA content in wild type plants. With the exception of crwn2 and crwn3, the crwn mutations caused a more pronounced reduction in nuclear size than predicted from the observed endopolyploidy level. As a consequence, crwn mutants display a spectrum of nuclear DNA densities, ranging from wild-type values in crwn2 and crwn3 mutants to fourfold higher densities in crwn1 crwn2 double mutants and the two viable crwn triple mutants.
64

Average nuclear size in crwn leaf cells ( m 2) wild type crwn1
crwn1 crwn3 crwn4 Average ploidy level (fraction of wild type)
Figure 2.4 The effects of crwn mutations on nuclear size and nuclear DNA density in leaf cells.
65

A
90 80 70 60 50 40 30 20 10 0

crwn2 crwn3 ccrrwwnn41 ccrrwwnn12 ccrrwwnn31 ccrrwwnn42 ccrrwwnn32 cccrrrwwwnnn431ccrrwwnn42 crwn4

B

1 0.9 0.8 0.7 12 0.6 134 0.5 124 0.4

14

14 13

34 24

wild-type

wild type 23
32

0.3

0.2

0.1

0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Average nuclear area (fraction of wild type)

Figure 2.4 (A) Developmentally matched rosette leaves from approximately one month-old plants were fixed, cells isolated and stained with DAPI, and nuclei imaged using epifluorescence microscopy. The average areas of randomly-selected individual nuclei (n = 32-108) were determined for each genotype. Error bars indicated standard error of the mean. (B) Nuclear area measurements from panel A (see also, Supplementary Table 2.2) were converted to relative values and were plotted against average endopolyploidy level for each genotype expressed as a fraction of the wildtype value. The average endopolyploidy levels (Supplementary Table 2.2) were measured by flow cytometry using corresponding leaf samples. The diagonal dashed line indicates the expected linear relationship between nuclear area and endopolyploid level observed in wild-type plants. Note that panel A measures nuclear area but the isolated nuclei are flattened under a coverslip to a uniform thickness (see Supplementary Figure 2.4) and therefore nuclear area is proportional to and therefore a proxy for nuclear volume. The numbers next to the symbols indicate the corresponding crwn genotype. Data points above the dashed line indicate an elevated nuclear DNA density relative to wild type. The solid line corresponds to a nuclear DNA density four-times that of wild type.
66

circularity 4 π Area / (perimeter)2
shape index perimeter2 / π2 major axis2

wild type
1 0.8 0.6 0.4 0.2
0 0 100 200 300
crwn1
1 0.8 0.6 0.4 0.2
0 0 100 200 300
crwn4
1 0.8 0.6 0.4 0.2
0 0 100 200 300
Nuclear Area (μm2)

400 400 400

wild type
1 0.8
0.6
0.4
0.2 0 0 100 200 300 400

crwn1
2
1.6
1.2
0.8
0.4
0 0 100 200 300 400

crwn4
2
1.6
1.2
0.8
0.4
0 0 100 200 300
Nuclear Area (μm2)

400

Supplementary Figure 2.3 Nuclear shape changes in crwn1 and crwn4

67

Supplementary Figure 2.3 Images of representative DAPI-stained adult leaf cell nuclei from wild type, crwn1 and crwn4 mutants were processed by ImageJ software to determine the circularity index (4π · Area / (perimeter)2), as well as a shape index (perimeter/π · major axis)2. Nuclei that deviate from a perfect circle (1.0) show a lower circularity index. The shape index highlights different types of deviations from the round shape. Nuclei in the crwn1 sample show a shape index close to 1.0, indicating consistently round nuclei. The reduced (relative to 1.0) shape indices in the wild-type sample across all nuclear sizes indicate uniformly elongated nuclear shapes. The elevated shape indices characteristic of larger crwn4 nuclei result from the presence of thin projections from the surface of otherwise round nuclei.
68

nuclear thickness (µm)

4
3
2
1
0 0 20 40 60 80 100 120 140 160 180 nuclear area (µm2)
Supplementary Figure 2.4 Leaf nuclear preparation and confocal imaging reveals a consistent nuclear thickness across a range of nuclear sizes Supplementary Figure 2.4. Mature leaves were harvested from five individuals in a F2 population segregating both crwn1 and crwn2 mutations (F2 plants of a crwn1 crwn2 x wild type cross). Consequently, the sample captured a range of nuclear shapes and sizes. The nuclei were fixed, isolated, and prepared for imaging as described for Figure 2.4. Following DAPI staining, the three-dimensional signal of different nuclei were recorded and reconstructed using a Leica SP5 confocal microscope. The area of each nucleus was measured using ImageJ, while the thickness of each nucleus was determined by the number and thickness of steps on the z-axis necessary to move from the top to the bottom of each nucleus. The different colored dots on the graph correspond to different slides imaged in this experiment. The results indicate that our preparation and imaging procedure generates nuclei with a relatively uniform thickness, mostly in the 2-3 micrometer range, regardless of the size and shape of the nuclei. Further, this thickness is consistent across individual slides.
69

We then investigated the relationship between nuclear size and DNA content by examining the effects of different crwn genotypes on nuclear size in leaf guard cells, a diploid cell type where endopolyploidy is not a factor [125]. crwn1 mutant guard cell nuclei were smaller than nuclei in wild type cells with an area approximately one-half of the wild type value, corresponding to a volume difference of approximately threefold assuming a roughly spherical shape to nuclei in the cell (Figure 2.5). Double and triple mutants lacking CRWN1 displayed nuclear sizes similar to the crwn1 single mutant. Consistent with their effects on nuclear size shown in Figure 2.4A, neither the crwn2 nor crwn3 mutation affected in nuclear size in guard cells. Interestingly, the size of nuclei in crwn4 guard cells was also unaffected, in contrast to the effect seen in a population of adult leaf cells (Figure 2.4A). However, crwn2 crwn3, crwn2 crwn4, and crwn3 crwn4 double mutants had nuclei approximately 20% smaller than those seen in wild-type guard cells, suggesting some functional redundancy among CRWN2, CRWN3 and CRWN4 proteins. Overall, our results indicate that CRWN1 plays the major role in affecting nuclear size in the absence of changes in endopolyploidy.
CRWN4 maintains interphase chromocenter integrity and organization Considering the dramatic effects of crwn mutations on nuclear size and morphology, we turned our attention to role of CRWN proteins on the internal organization of the nucleus. A conspicuous feature of Arabidopsis interphase nuclei are discrete foci of heterochromatin, or chromocenters, visualized as bright spots after staining with fluorescent DNA-intercalating dyes [126]. A typical interphase nucleus contains
70

approximately ten chromocenters corresponding to the number of diploid chromosomes (2n=10) [106]. Chromocenter number remains fairly constant over a wide range of nuclear sizes and endopolyploid levels (2n to 16n), most likely as a result of lateral association of sister chromatids after endoreduplication [101, 127]. We found that the average chromocenter number in crwn1, crwn2 and crwn3 leaf cell nuclei was similar to that seen in wild-type leaf cell nuclei (Figure 2.6A) and did not change dramatically as a function of nuclear size. In crwn4 nuclei, however, chromocenter number was strongly correlated with nuclear size (Figure 2.6A): smaller nuclei contained fewer chromocenters than the wild-type value of ~9, while larger, presumably endopolyploid, crwn4 nuclei exhibited a wide range of chromocenter numbers (2-27). A similar pattern was observed in double mutants containing the crwn4 mutation (Supplementary Figure 2.5). In contrast, double mutants containing the crwn1 allele paired with another crwn mutation displayed a reduced average chromocenter number with a weaker association with nuclear size (Figure 2.6A and Supplementary Figure 2.5).
71

Average guard cell nuclear size (µm2)

crwn4 crwn2 crwn3 crwn4 crwn3 crwn4 crwn4 crwn4 crwn4

crwn3

crwn2

crwn1

wild type

crwn1 crwn1 crwn1 crwn2 crwn2 crwn3 crwn2 crwn3

crwn1 crwn1

'1$4" '1#2" '1!"0
&8 " %6 " $4" #2" !0"
Figure 2.5 Average leaf guard cell nuclear sizes in crwn mutants. Figure 2.5 Two week-old plants were harvested, fixed, stained with DAPI, and leaf guard cell nuclei were imaged using confocal laser scanning microscopy. The area of randomly-selected individual nuclei (n = 19-29) were determined for each genotype. Error bars indicated standard error of the mean.
72

Chromocenter number in leaf cell nuclei Aggregation Index (AI)

A 30
25

wild type

20

15

10

5

0 0 50 100 150 200 250

30 25 20 15 10 5 0
0
30 25 20 15 10 5 0
0

crwn1
50 100 150 200 250 crwn2
50 100 150 200 250

30 25 20 15 10 5 0
0

crwn3 50 100 150 200 250

30 25 20 15 10 5 0
0

crwn4 50 100 150 200 250

30 25 20 15 10 5 0
0

crwn1crwn2 50 100 150 200 250

Nuclear area (µm2)

B 0.8
0.6

wild type

0.4

0.2

0 0 5 10 15 20 25 30
0.8 crwn1 0.6

0.4

0.2

0 0
0.8
0.6

5 10 15 20 25 30 crwn2

0.4

0.2

0 0 5 10 15 20 25 30
0.8 crwn3 0.6

0.4

0.2

0 0 5 10 15 20 25 30

0.8 crwn4 0.6

0.4

0.2

0 0 5 10 15 20 25 30
0.8 crwn1crwn2 0.6

0.4

0.2

0 0 5 10 15 20 25 30

Total chromocenter area (µm2)

Figure 2.6 Chromocenter morphology changes in crwn mutants.

73

Figure 2.6 (A) Nuclei were imaged from developmentally matched rosette leaves from approximately one month-old plants, cells isolated and stained with DAPI, and nuclei imaged using epifluorescence microscopy. The area and chromocenter number of randomly-selected individual nuclei (n = 47-132) were determined for each genotype, and chromocenter number was plotted against nuclear area. A linear regression line showing the relationship between chromocenter number and nuclear size (as a proxy for endopolyploidy level) was plotted for each genotype. (B) Nuclei were imaged fully expanded rosette leaves from approximately one month-old plants and the chromocenter Aggregation Index (AI) was plotted against the total chromocenter area (µm2).
74

Chromocenter number in leaf cell nuclei

30 25 20 15 10
5 0
0
25 30 25 20 15 10
5 0
0
25 30 25 20 15 10
5 0
0

wild type

50 100 150 200 250

crwn1 crwn2

30 25

20

15

10

5

50 100 150 200 250

0 0

crwn1 crwn3

30 25

20

15

10

5

50 100 150 200 250

0 0

crwn2 crwn3
50 100 150 200 250
crwn2 crwn4
50 100 150 200 250

30 25 20 15 10
5 0
0

30
crwn1 crwn4
25 20 15 10
5 0 50 100 150 200 250

30 25 20 15 10
5 0
0

crwn3 crwn4
50 100 150 200 250

Nuclear area (µm2)

Supplementary Figure 2.5 Chromocenter changes in crwn double mutants

75

Supplementary Figure 2.5 Nuclei were harvested from developmentally matched rosette leaves from approximately one month-old plants, stained with DAPI, and imaged using epifluorescence microscopy. The area and chromocenter number of randomly-selected individual nuclei (n = 41-52) were determined for each genotype, and chromocenter number was plotted against nuclear area. A linear regression line showing the relationship between chromocenter number and nuclear size (as a proxy for endopolyploidy level) was plotted for each genotype.
76

To explore the chromocenter phenotype in more detail, we developed a statistic, referred to as an aggregation index (AI) (see Methods), to characterize the distribution of visible DAPI-bright spots within interphase nuclei. The value of this index ranges from 0 to 1, reflecting both the number of distinct chromocenter spots and the uniformity of their size distribution. The expected AI for wild-type nuclei containing 10 equally sized chromocenters is 0.1, while clustering of chromocenters into fewer but larger aggregates will lead to a higher AI value. A dispersal of chromocenters into smaller heterochromatic satellites will push the AI lower. For a given chromocenter number, a skewed CC size distribution is associated with a larger AI compared to when each CC is equally sized. As shown in Figure 2.6B, the AI index of wild-type nuclei averaged close to 0.1 and was not affected significantly by nuclear size. The absence of a significant correlation between AI and nuclear size indicates that chromocenter organization remains constant across different endopolyploidy levels in wild-type nuclei. A similar pattern was observed for the crwn1, crwn2, and crwn3 mutant samples. In contrast, combining crwn1 and crwn2 mutations led to an approximately two-fold higher AI over a range of nuclear sizes, consistent with the two-fold reduction in chromocenters number via aggregation in crwn1 crwn2 mutants. A different pattern was displayed in the crwn4 sample, which displayed a negative correlation between AI and nuclear size. This result suggests a tendency for chromocenters to aggregate in smaller crwn4 nuclei and to become dispersed in larger crwn4 nuclei. The reduction in chromocenter number in crwn1 crwn2 and crwn4 mutants with smaller nuclei could reflect the aggregation of individual chromocenters in the limited confines of these nuclei, but a similar clustering does not occur in small
77

wild-type nuclei, arguing that small nuclear dimensions alone are insufficient to cause clustering. The variability in chromocenter size and number in crwn mutant nuclei suggests that CRWN proteins are required for proper organization of heterochromatin in interphase nuclei.
We tested this hypothesis by visualizing the spatial arrangement of chromocenterassociated genomic regions in crwn1 crwn2 and crwn4 mutants. Arabidopsis chromocenters are comprised of large segments of repetitive DNA such as the tandemly-arrayed centromeric and 5S RNA repeats located within pericentromeric regions [106]. Using fluorescent in situ hybridization (FISH), we examined the spatial organization of the major 180-bp centromeric tandem repeat and the 5S RNA gene arrays in both large and small nuclei from wild-type, crwn1 crwn2 and crwn4 plants (Figure 2.7A, B). The centromeric and 5S RNA repeats were co-localized with the DAPI-bright spots in both small and large wild-type nuclei, confirming previous reports that these sequences are normally compartmentalized within chromocenters at the nuclear periphery [128] (Supplementary Movie 2.1 - 2.3, http://www.biomedcentral.com/1471-2229/13/200/additional). It was common to find a decondensed centromere signal at several chromocenters in wild-type nuclei; however, decondensed centromeric repeat clusters were infrequently observed in crwn1 crwn2 nuclei and the total number of clusters was reduced (Figure 2.7C) (also see Supplementary Movie 2.1 - 2.3). These findings indicate that there is a compaction of the centromere repeat arrays within coalesced chromocenters in crwn1 crwn2 nuclei. In contrast, the number of discrete centromere repeat clusters visible in
78

crwn4 nuclei was more variable, and decondensed signals were often seen in nuclei with numerous clusters. This pattern is consistent with the hypothesis that chromocenters become dispersed in larger crwn4 nuclei. A similar but more pronounced trend was seen for the 5S RNA gene arrays (Figure 2.7B, D), which were dispersed outside chromocenter aggregates in roughly one-half of the crwn4 nuclei. We note that the dispersed 5S RNA gene signal remained localized to the nuclear periphery (see Supplementary Movie 2.1 - 2.3). The apparent dispersal of chromocenters in larger crwn4 nuclei and the mis-positioning of centromeric and 5S RNA repeats outside of the chromocenter indicates that higher-order organization of heterochromatin breaks down in interphase in the absence of CRWN4.
79

A wild type

crwn1 crwn2

crwn4

B

5 m DAPI

cen180 5S

Number of cen180 clusters Number of 5S clusters

C 20
18 16 14 12 10 8 6 4 2 0
wild type crwn1 crwn2 crwn4

D All
dispersed
6 5 4 3 2 1 0
wild type crwn1 crwn2 crwn4

54 3 2 1 0 # decondensed cen180 clusters

nuclei with all 5S signal associated with CC nuclei with dispersed 5S signal not associated with CC

Figure 2.7 Chromocenter organization is altered in crwn1 crwn2 and crwn4 mutants. (A and B) Fluorescence in situ hybridization of representative small and large nuclei from wild type, crwn1 crwn2 and crwn4 cells prepared from leaves of adult plants after bolting. Blue: DAPI, pink: 180-bp centromere repeats, green: 5S RNA genes, bar, 5 µm. (C) The distribution of nuclei based on the number of centromere repeat clusters in different genotypes. Each circle represents an individual nucleus in the FISH experiment. The color indicates the level of decondensation of the centromere signal. (D) The distribution of nuclei based on the number of 5S RNA repeat clusters in different genotypes. Each circle represents an individual nucleus. The color indicates the level of dispersion of the 5S repeat signal.

80

Discussion
Our results demonstrate that the CRWN family is essential for viability in Arabidopsis and required for proper nuclear organization. Redundancy and diversification exists among the four CRWN paralogs, which belong to a plant-specific family of nuclear coiled-coil proteins that forms two clades: one including CRWN1, CRWN2 and CRWN3 and the other containing CRWN4 (Figure 2.1). The divergence of CRWN4 relative to the other CRWN paralogs, also reported by Kimura et al. (2010) [76] and Ciska et al. (2013) [77], is supported by our genetic analysis. First, the inviability of triple mutants containing either CRWN1 or CRWN4 alone indicates that CRWN1 and CRWN4 possess non-overlapping functions (Figure 2.2). Second, loss-of-function mutations in CRWN1 and CRWN4 have distinct phenotypes; for example, crwn1 is the only crwn mutation that affects nuclear size in diploid guard cells (Figure 2.5), while the dispersed chromocenter phenotype is unique to the crwn4 mutant (Figures 2.3 and 2.7). Finally, the functional distinction between the two CRWN clades is supported by taxa, such as rice and maize, which contain only two CRWN-like proteins: one CRWN1-like and one CRWN4-like.
The phenotypic effects of combining mutations in different CRWN genes demonstrates that all four paralogs are involved in specifying nuclear size in adult cells. A deficiency in CRWN1 led to a dramatic reduction in nuclear size independent of an effect on endopolyploidy (Figure 2.4). Loss of CRWN4 also reduced nuclear size in leaf cells (Figure 2.4) as was recently shown by Sakamoto and Takagi (2013) [81].
81

Interestingly, crwn4 nuclei in leaf guard cells were not reduced in size (Figure 2.5), despite the fact that CRWN4 is expressed in this cell type (http://bar.utoronto.ca/efp/cgi-bin/efpWeb.cgi) [122], suggesting cell-type specific requirements exist for different CRWN proteins. Combining a crwn1 loss-of-function mutation with a deficiency in any of the remaining three paralogs causes a further reduction in nuclear size in leaf cells. The additive effect on nuclear size indicates that CRWN2 and CRWN3 share overlapping functions with CRWN1, although loss of CRWN2 and/or CRWN3 has no effect on nuclear size or endopolyploidy in leaf cells when a functional CRWN1 allele is present. We previously reported that the crwn2-1 allele caused a reduction in nuclear size in leaf cells [78]; the reason for the different behavior displayed in this study is unknown, but we have noted variability in nuclear phenotypes among different crwn2 mutant lines. Regardless, it is clear that CRWN1 plays the major role in adult leaf tissue among the three paralogs in the CRWN1-like clade. This situation might reflect the higher level of expression of CRWN1 in leaf tissue compared to CRWN2 and CRWN3 – ca. 30 FPKM (fragment per kilobase/million mapped RNA-seq reads) for CRWN1, with 4x and 2x less expression from CRWN2 and CRWN3, respectively (data not shown). Alternatively, the different contributions of CRWN1-like genes could also result from structural differences among the encoded proteins.
Our data indicate that the primary phenotype of crwn mutants is a reduction in nuclear size and that the reduction in endopolyploidy observed in double and triple crwn mutants is a secondary effect in response to reduced nuclear size. First, mutation of
82

either CRWN1 or CRWN4 had an effect on nuclear size in leaf cells without affecting endopolyploidy (Figure 2.4) (see also [81]). Second, effects on endopolyploid levels were only seen in mutants that contained two or more crwn mutations and exhibited severely reduced nuclear size. These considerations indicate that loss of CRWN activity alters the relationship between DNA content and nuclear volume, leading to a higher than normal nuclear DNA density. In the most severely affected mutants, the nuclear DNA density reaches four times the level seen in wild-type cells. It is intriguing that the crwn mutants with the most abnormal whole-plant dwarfing phenotypes, including crwn1 crwn2 and the viable triple crwn mutants, are the ones with the highest average nuclear DNA density.
We hypothesize that the CRWN proteins are required for nuclear expansion after nuclear reformation in telophase [42, 129, 130]_ENREF_26, and that the loss of these proteins, especially in combination, results in an elevated nuclear DNA density. Evidence supporting this mechanism comes from a recent report demonstrating that crwn1 crwn2 cotyledon nuclei expand more slowly than their wild type counterparts during the first 72 hours of seed germination, a period normally characterized by a ca. 10-fold expansion of nuclear size in the absence of endoreduplication [104]. Interestingly, the rate of contraction of crwn1 crwn2 embryonic cotyledon nuclei during seed maturation is also reduced relative to wild type. These observations suggest that CRWN proteins are involved in remodeling nuclear structure during interphase in response to developmental and environmental cues. Similarly, a
83

constraint in nuclear expansion associated with endoreduplication [124, 131] could explain the limit on endopolyploid levels observed in high-order crwn mutants.
The large coiled-coil domains that comprise the central region of all four CRWN proteins point toward a structural role in the nucleus. Recent analysis based on secondary structure for analogues of Arabidopsis coiled-coil proteins indicates that CRWN proteins are candidates for paramyosin homologues [80]. Paramyosin is a structural protein found in invertebrate muscles, where it forms the core of thick filaments and bundles with myosin motors. Deficiency of paramyosin, encoded by the unc-15 gene in C. elegans, leads to shortened, mis-formed and apparently fragile thick filaments in the nematode’s muscles [132]. The structural similarity to paramyosin suggests models for CRWN action as architectural components of the nucleoskeleton. Models of this type predict that crwn mutant nuclei would have a less sound structural foundation and be more prone to breakage or distortion by intracellular forces (e.g., exerted by the cytoskeleton or at programmed nuclear expansion/contraction transitions), as seen in animal cells when lamin is disrupted [45] or down-regulated [133]. The irregular margins and thin projections seen among crwn4 nuclei are consistent with this structural integrity model. These predictions, however, are not consistent with the smaller, round nuclei in crwn1 mutants that do not adopt the elongated spindle shapes typical of wild-type nuclei in many cell types. These considerations suggest alternative models wherein the CRWN1 protein might establish flexible hinge regions or expansion zones in the nucleoskeleton, facilitating nuclear size changes that accompany endopolyploidy and other developmental transitions. We
84

note that both nucleoporin 136 [134, 135] and LINC complex mutants [65] in Arabidopsis also lead to more spherical nuclear shapes, suggesting that CRWN proteins might interact with these complexes at the nuclear periphery.
Deficiency of CRWN proteins also affects chromocenter organization (Figure 2.6 and 2.7). We previously reported that chromocenter number decreases approximately twofold in crwn1 crwn2 nuclei [78], and we confirmed these results here while extending our analysis to the entire CRWN gene family. One unexpected finding was the wide variation in chromocenter number in crwn4 nuclei and the direct correlation between chromocenter number and nuclear size. The reduction in chromocenter number in crwn1 crwn2 mutants, as well as crwn4 mutants with smaller nuclei, is consistent with aggregation of individual chromocenters. The in situ hybridization data shown in Figure 2.7 further support this conclusion. Further, we demonstrated that chromocenter organization was disrupted in larger crwn4 nuclei as evidenced by the dispersed signals seen for the 5S RNA genes, and to a less extent, the centromere repeat arrays. Considering that chromocenters are localized primarily to the nuclear periphery [106, 128] (see Supplementary Movie 2.1 - 2.3) where CRWN proteins are also located [78, 81, 115], CRWN proteins might play a direct role in ensuring proper heterochromatin organization. In such models, CRWN1 and CRWN2 would act to prevent chromocenter aggregation and CRWN4 would exert a complementary effect to maintain chromocenter integrity.
85

The four distinct phenotypes displayed by crwn mutants – reduced nuclear size, altered nuclear shape, elevated nuclear DNA density and abnormal organization of constitutive heterochromatin – highlight a functional connection among these different aspects of nuclear architecture. The lack of whole-plant phenotypes of Arabidopsis crwn1 and crwn4 mutants is remarkable in light of the dramatic nuclear changes occurring in these mutants and underscores plants’ plasticity in their ability to execute an apparently normal developmental program in spite of these nuclear changes. The diversity of CRWN proteins and the ability to work with viable crwn mutants that exhibit dramatic nuclear phenotypes will facilitate the elucidation of the mechanisms through which these essential proteins exert their effects on nuclear organization.
Conclusions
This study addresses fundamental questions about how plant cells specify and control the morphology of their nuclei and its relationship with internal chromatin organization. We conducted a comprehensive reverse genetics study of the CRWN gene family in Arabidopsis, which encode NMCP-class proteins implicated in nuclear morphology and organization. We demonstrated that CRWN proteins are essential for viability, and in the process, uncovered a surprisingly high degree of functional diversity among the CRWN proteins. CRWN1 and CRWN4 are the major determinants of nuclear size and shape, and we hypothesize that deficiency in CRWN proteins leads to defects in nuclear expansion and remodeling. One consequence of
86

this deficiency is an increase in nuclear DNA density as endoreduplication levels are not affected except in the most extreme cases (e.g., crwn1 crwn2 and the viable triple mutants). Our findings also demonstrated that CRWN4 plays a role in maintenance of heterochromatin organization in interphase nuclei. The specificity of the nuclear morphological and higher-order chromatin organization defects seen in crwn mutants reveals the interplay between nuclear morphology and the three-dimensional packaging of the genome.
Materials and Methods
Plant materials and growth conditions All T-DNA insertion alleles used in this study are from the SALK collection [121] in strain Columbia, and single mutant lines were originally obtained from the Arabidopsis Biological Resource Center (ABRC) at The Ohio State University. Plants were grown in long-day lighting conditions (16 h of light / 8 h of dark) at 23˚C on soil (Metro-Mix 360, SunGro, Vancouver) in environmental growth chambers. Genotyping of individual T-DNA alleles was performed by standard PCR using the following pairs of allele-specific primers: SALK_025347 (crwn1-1) , 5’-TGC CTT CTC CTC GCT TTT CAA-3’ and 5’-TGC GTG AAT GGG AAA GAA AGT TG-3’; SALK_076653 (crwn2-1), 5’-GAA GCT CAT TGC TAG AGA AGG GG-3’ and 5’AAC GCT GAT CGT TCA TGT TCC A-3’; SALK_099283 (crwn3-1), 5’-TTC TGC ATC TTG ACA CCA TCC AA-3’ and 5’-TCG TCG ACT AGT TAA CAA AAT CA-
87

3’; SALK_079296 (crwn4-1), 5’-CGC AAA GCC TTC GAA GAC AAA-3’ and 5’GCT TCA GCC AGC ATT TCA AGC-3’.
Phylogenetic tree construction Amino acids similar to CRWN1 were downloaded from public databases (see Supplementary Table 2.1). The program ClustalX was used to align the amino acid sequences. The tree in Figure 2.1 is based on an alignment of the region of highest conservation across all amino acid sequences, corresponding to the coiled-coil domains (amino acids 64 to 651 in CRWN1). A maximum likelihood tree was constructed using Phylip 3.69 with 1000 bootstrap replicates.
Diploid guard cell nuclear area measurement Two-week old seedlings were harvested and fixed in 3:1 acetic acid:ethanol. Nuclei in the fixed tissue were stained using DAPI (10µg/ml, 2 minutes), and guard cell nuclei were imaged using laser scanning confocal microscopy (Leica SP5). Images were taken at the focal plane with the maximum nuclear area, and the resulting images were processed using ImageJ software.
88

Leaf nuclei isolation and imaging Nuclei were isolated from the fifth true leaf of adult plants harvested after initiation of flowering stem elongation (stem height ≥ vegetative rosette diameter). Therefore, the tissues were developmentally matched across the different genotypes. Mesophyll cells predominated but other cell types were present. Each harvested leaf was bisected and one half used for the nuclear area measurement, while the remaining half leaf was processed for flow cytometry measurements (see below). Leaf tissue was fixed using a 3:1 acetic acid:ethanol solution and tissues were rehydrated in 100 mM sodium citrate buffer pH 4.8 for 15 min followed by incubation in digestion buffer (0.03% cytohelicase, 0.03% pectolyase, and 0.03% cellulase Onozuka RS in 100 mM sodium citrate buffer pH 4.8) for 2 hours at 37˚C. Digested tissue was carefully homogenized by pipetting, centrifuged briefly at low speed, and resuspended with 100 mM sodium citrate buffer pH 4.8; this cycle was repeated three times and the final pellet was resuspended in 3:1 acetic acid:ethanol. The resulting suspension of nuclei were pipetted onto microscope slides, dried for ca. 1 min, and stained with 10µg/ml 4',6diamidino-2-phenylindole (DAPI). A Leica DM 5500 epifluorescence microscope was used to image the nuclei, and the nuclear area was measured from digital images using ImageJ software after manual tracing of nuclear boundaries. Note that the chromocenter number versus leaf cell nuclear area scatter plots shown in Figure 2.6A were generated in a separate experiment from the one shown in Figure 2.4, but developmentally matched leaf tissues were harvested as described above.
Aggregation Index measurement
89

Tissue from adult leaves was harvested and prepared as described for FISH (see below). Rather than performing the in situ hybridization step, nuclei were stained with DAPI and imaged using optical sectioning microcopy. Projections of the processed images were analyzed using ImageJ software to identify chromocenters. Briefly, the images were manually manipulated when necessary to adjust the local threshold and chromocenter area and number were assigned using the Analyze Particle function of the software. The aggregation index (AI) was calculated using the following equation: AI = Σ (Si/Stotal)2; where i = 1, … , n; Si = the area of chromocenter i; Stotal = the total area of all chromocenters in the nucleus.
Flow cytometry Bisected tissue from the fifth true leaf of adult plants were harvested (see above) and immersed in magnesium sulfate buffer [136] and chopped with razor blades in a petri dish. The resulting suspension was filtered through a nylon mesh (diameter = 30 µm; Partec Cell Trics ®, Münster, Germany). The nuclear suspension was incubated with RNAse A (Ribonuclease A, from bovine pancreas, Sigma, St. Louis, MO, USA) (50 µg/ml) on ice for 15 min and stained with propidium iodide (50 µl/ml) in the dark for 6 hours. Average ploidy level for each genotype was calculated based on the peaks generated from an analytical flow cytometer (Accuri 6 model, Accuri cytometers, Ann Arbor, MI, USA).
Fluorescent in situ hybridization (FISH)
90

Fully expanded adult leaf tissue was harvested less than a week after flowering stem elongation, and fixed in Buffer A [137] with 4% formaldehyde at room temperature with agitation for >1 hour. After rinsing with Buffer A, the tissue was chopped repeatedly with razor blades until a homogenous texture was achieved. A clear nuclear suspension was pipetted from the leaf debris and used for fluorescence in situ hybridization as described by Golubovskaya et al. (2002) [137]. The centromere probe 5’- Cy5-GGTTGCGGTTTAAGTTCTTATACTCAATC -3’ was synthesized by Integrated DNA Technologies (Coralsville, IA, USA), and the 5S probe was amplified from genomic DNA using primers 5’-CTNCCNGGNAGNTCACCC-3’ and 5’-CCTNGTGNTGNANCCCTC-3’, followed by labeling using a nick translation protocol and Rhodamine labeled dCTP (Roche; Indianapolis, IN, USA).
91

REFERENCES
1. Meier I: The plant nuclear envelope. Cell Mol Life Sci 2001, 58(12-13):17741780.
2. Gruenbaum Y, Goldman RD, Meyuhas R, Mills E, Margalit A, Fridkin A, Dayani Y, Prokocimer M, Enosh A: The nuclear lamina and its functions in the nucleus. Int Rev Cytol 2003, 226:1-62.
3. Graumann K, Evans DE: The plant nuclear envelope in focus. Biochem Soc Trans 2010, 38(Pt 1):307-311.
4. Rose A: Open Mitosis: Nuclear Envelope Dynamics. In: Cell Division Control in Plants. Edited by Verma DPS, Hong Z. Berlin Heidelberg: Springer-Verlag; 2007.
5. Dittmer TA, Stacey NJ, Sugimoto-Shirasu K, Richards EJ: LITTLE NUCLEI genes affecting nuclear morphology in Arabidopsis thaliana. Plant Cell 2007, 19(9):2793-2803.
6. Dittmer TA, Richards EJ: Role of LINC proteins in plant nuclear morphology. Plant Signal Behav 2008, 3(7):485-487.
7. Masuda K, Xu ZJ, Takahashi S, Ito A, Ono M, Nomura K, Inoue M: Peripheral framework of carrot cell nucleus contains a novel protein predicted to exhibit a long alpha-helical domain. Experimental Cell Research 1997, 232(1):173-181.
92

8. Ciska M, Masuda K, Moreno Diaz de la Espina S: Lamin-like analogues in plants: the characterization of NMCP1 in Allium cepa. J Exp Bot 2013, 64(6):1553-1564.
9. Kimura Y, Kuroda C, Masuda K: Differential nuclear envelope assembly at the end of mitosis in suspension-cultured Apium graveolens cells. Chromosoma 2010, 119(2):195-204.
10. Gardiner J, Overall R, Marc J: Putative Arabidopsis homologues of metazoan coiled-coil cytoskeletal proteins. Cell Biol Int 2011, 35(8):767774.
11. Razafsky D, Hodzic D: Bringing KASH under the SUN: the many faces of nucleo-cytoskeletal connections. J Cell Biol 2009, 186(4):461-472.
12. Sosa BA, Rothballer A, Kutay U, Schwartz TU: LINC complexes form by binding of three KASH peptides to domain interfaces of trimeric SUN proteins. Cell 2012, 149(5):1035-1047.
13. Tapley EC, Starr DA: Connecting the nucleus to the cytoskeleton by SUNKASH bridges across the nuclear envelope. Curr Opin Cell Biol 2013, 25(1):57-62.
14. Zhou X, Graumann K, Evans DE, Meier I: Novel plant SUN-KASH bridges are involved in RanGAP anchoring and nuclear shape determination. J Cell Biol 2012, 196(2):203-211.
93

15. Graumann K, Evans DE: Plant SUN domain proteins: components of putative plant LINC complexes? Plant Signal Behav 2010, 5(2):154-156.
16. Alonso JM, Stepanova AN, Leisse TJ, Kim CJ, Chen H, Shinn P, Stevenson DK, Zimmerman J, Barajas P, Cheuk R et al: Genome-wide insertional mutagenesis of Arabidopsis thaliana. Science 2003, 301(5633):653-657.
17. Winter D, Vinegar B, Nahal H, Ammar R, Wilson GV, Provart NJ: An "Electronic Fluorescent Pictograph" browser for exploring and analyzing large-scale biological data sets. PLoS One 2007, 2(8):e718.
18. Chytilova E, Macas J, Sliwinska E, Rafelski SM, Lambert GM, Galbraith DW: Nuclear dynamics in Arabidopsis thaliana. Mol Biol Cell 2000, 11(8):27332741.
19. Sakamoto Y, Takagi S: LITTLE NUCLEI 1 and 4 regulate nuclear morphology in Arabidopsis thaliana. Plant Cell Physiol 2013, 54(4):622-633.
20. Jovtchev G, Schubert V, Meister A, Barow M, Schubert I: Nuclear DNA content and nuclear and cell volume are positively correlated in angiosperms. Cytogenet Genome Res 2006, 114(1):77-82.
21. Melaragno JE, Mehrotra B, Coleman AW: Relationship between endopolyploidy and cell size in epidermal tissue of Arabidopsis. Plant Cell 1993, 5(11):1661-1668.
22. Fransz P, Soppe W, Schubert I: Heterochromatin in interphase nuclei of Arabidopsis thaliana. Chromosome Res 2003, 11(3):227-240.
94

23. Fransz P, De Jong JH, Lysak M, Castiglione MR, Schubert I: Interphase chromosomes in Arabidopsis are organized as well defined chromocenters from which euchromatin loops emanate. Proc Natl Acad Sci U S A 2002, 99(22):14584-14589.
24. Schubert V, Berr A, Meister A: Interphase chromatin organisation in Arabidopsis nuclei: constraints versus randomness. Chromosoma 2012, 121(4):369-387.
25. Schubert V, Klatte M, Pecinka A, Meister A, Jasencakova Z, Schubert I: Sister chromatids are often incompletely aligned in meristematic and endopolyploid interphase nuclei of Arabidopsis thaliana. Genetics 2006, 172(1):467-475.
26. Fang Y, Spector DL: Centromere positioning and dynamics in living Arabidopsis plants. Mol Biol Cell 2005, 16(12):5710-5718.
27. Bradley MV: Cell and nuclear size in relation to polysomaty and the nuclear cycle. American Journal of Botany 1954, 41(5):398-402.
28. Burke B: The nuclear envelope: filling in gaps. Nat Cell Biol 2001, 3(12):E273-274.
29. Levy DL, Heald R: Nuclear size is regulated by importin alpha and Ntf2 in Xenopus. Cell 2010, 143(2):288-298.
30. van Zanten M, Koini MA, Geyer R, Liu Y, Brambilla V, Bartels D, Koornneef M, Fransz P, Soppe WJ: Seed maturation in Arabidopsis thaliana is
95

characterized by nuclear size reduction and increased chromatin condensation. Proc Natl Acad Sci U S A 2011, 108(50):20219-20224. 31. Sugimoto-Shirasu K, Roberts K: "Big it up": endoreduplication and cellsize control in plants. Curr Opin Plant Biol 2003, 6(6):544-553. 32. Epstein HF, Casey DL, Ortiz I: Myosin and paramyosin of Caenorhabditis elegans embryos assemble into nascent structures distinct from thick filaments and multi-filament assemblages. J Cell Biol 1993, 122(4):845-858. 33. Dechat T, Pfleghaar K, Sengupta K, Shimi T, Shumaker DK, Solimando L, Goldman RD: Nuclear lamins: major factors in the structural organization and function of the nucleus and chromatin. Genes Dev 2008, 22(7):832-853. 34. Friedl P, Wolf K, Lammerding J: Nuclear mechanics during cell migration. Curr Opin Cell Biol 2011, 23(1):55-64. 35. Tamura K, Fukao Y, Iwamoto M, Haraguchi T, Hara-Nishimura I: Identification and characterization of nuclear pore complex components in Arabidopsis thaliana. Plant Cell 2010, 22(12):4084-4097. 36. Tamura K, Hara-Nishimura I: Involvement of the nuclear pore complex in morphology of the plant nucleus. Nucleus 2011, 2(3):168-172. 37. Arumuganathan K, Earle ED: Estimation of nuclear DNA content of plants by flow cytometry. Plant Molecular Biology Reporter 1991, 9:229-233. 38. Golubovskaya IN, Harper LC, Pawlowski WP, Schichnes D, Cande WZ: The pam1 gene is required for meiotic bouquet formation and efficient
96

homologous synapsis in maize (Zea mays L.). Genetics 2002, 162(4):19791993.
97

CHAPTER THREE
THE LOSS OF CRWN PROTEINS LEAD TO BROAD TRANSCRIPTIONAL MIS-REGULATION
Abstract
I describe significant transcriptional changes in Arabdiopsis crwn mutants, which have reduced nuclear size and altered nuclear shape. The comparison among crwn transcriptomic changes confirmed the synergistic interaction between crwn1 and crwn2 mutations evident in our previous genetic studies. In addition, my transcriptomic data uncovered a functional suppression between the crwn1 and crwn4 mutations, and demonstrated that CRWN1-like genes and CRWN4 regulate the transcription of an overlapping set of genomic loci. Many mis-expressed loci were identified as candidates for mediating the variety of phenotypic changes observed in crwn mutants. The mis-regulated transcriptomes of crwn mutants feature activation of stress response pathways and altered expression of many nuclear proteins. I propose that the loss of CRWN proteins affects the structure of the nucleoskeleton and/or the organization of the transcriptional machinery, both of which could lead to alteration of transcriptional profiles.
98

Introduction
The three-dimensional spatial organization of the eukaryotic genome in the nucleus is thought to play an important role in control of gene expression. For example, actively expressed genes are frequently positioned in the vicinity of nuclear pore complexes [138]. Another demonstration of epigenetic regulation of gene expression through spatial contextualization comes from reports of gene mis-expression after relocalization to the nuclear periphery via association with lamin proteins [139].
These experiments have been carried out in diverse eukaryotic models ranging from mammalian cells to single-cell fungi, arguing that a functional connection between nuclear organization and gene expression is a general principle in eukaryotes. Yet few studies have investigated or focused on this level of epigenetic control in plants. I have investigated a set of mutations in Arabidopsis thaliana that alter nuclear shape and size. These mutations disrupt members of a small gene family, called CRWN (CROWDED NUCLEI), which encodes four paralogous proteins that contain a large central coiled-coil domain [78, 140]. Mutation of either CRWN1 or CRWN4 leads to a reduction in nuclear size and a loss of the elongated nuclear shape characteristic of expanded, differentiated cells in this species [81, 140]. Further, crwn4 mutations lead to a dispersal of the large heterochromatic ‘chromocenter’ foci in interphase nuclei [140]. These nuclear morphology defects indicate that crwn mutations alter nuclear
99

architecture, and are therefore useful tools to probe the relationship between nuclear organization and gene expression in plants.
Here I report transcriptional changes in selected crwn mutants that advance our understanding of the functional interaction among different CRWN proteins. Our results indicate that i) crwn4 and crwn1 crwn2 mutants display the most significant transcriptional mis-regulation; ii) crwn1 and crwn4 mutations functionally suppressed each other; and iii) CRWN1-like genes and CRWN4 shared an overlapping set of genomic targets in transcriptional regulation. Many nuclear proteins were misregulated in crwn mutants, providing explanations for both the altered nuclear organization and nuclear function in these mutants.
Results
To determine if a loss of CRWN proteins causes gene expression changes, I performed mRNA-seq profiling in one-month-old adult leaf samples from six genotypes: wild type, crwn1, crwn2, crwn4, crwn1 crwn2, and crwn1 crwn4. These genotypes cover the full range of phenotypes characteristic of crwn mutants, both in terms of nuclear size and shape, as well as heterochromatin organization (summarized in Table 3.1). Strand-specific cDNA libraries of three biological replicates were constructed for each genotype, and sequence reads were generated using an Illumina Hi-Seq platform. Table 3.2 describes the number of loci identified by reads mapped to the reference
100

Columbia wild-type genome (annotation version: TAIR10). Approximately 27,000 loci in each genotype were detected, and 73% of these loci have sufficient read numbers to support a statistically-meaningful test. These ‘tested’ loci included approximately 19,000 genes and 1,000 transposable elements.
Figure 3.1A displays the fold change of the tested loci relative to the corresponding q value statistic. Transposable elements are shown in red and genes in blue. The distribution of data points shifted to different extents in the crwn mutants to form a broader base corresponding to mis-regulated genes deviating from the wild-type values with a more robust statistical significance (lower q value). This shift was clearly seen for gene loci (blue dots), but was also evident for transposons (red dots). However, most transposon loci clustered toward the top range of the distributions in Figure 3.1A with higher q values – a situation due in part to the low read numbers associated with these normally transcriptionally quiescent loci. Consistent with this explanation, transposons are over-represented among ‘non-tested’ loci for which too few reads were recovered to support a robust statistical test for expression changes (see Table 3.1). Nonetheless, transposon mis-regulation in crwn mutants was evident when all loci, tested and non-tested, were ordered by their physical position from chromosome 1 to 5 (Figure 3.2). The transposon-rich centromeric regions on each chromosome stood out with a more red color pattern indicating a general trend towards up-regulation. In addition, certain regions in the chromosome arms in the crwn4, crwn1 crwn2, and crwn1crwn4 samples also showed mis-regulation.
101

Table 3.1 Summary of phenotypic changes in crwn mutants
Table&1&Summary&of&phenotypic&changes

significantly+mis-expressed+loci whole+plant+morphology cell+size+ nuclear+shape Ave.+leaf+nuclear+size

crwn1 crwn2

crwn4

crwn1(crwn2

crwn1(crwn4

1327

466

5607

4815

2126

normal normal

normal

dwarf,+short+internode,+ slightly+dwarf,+other+

smaller+leaves,+more+ changes+follow+crwn1(

branches,+early+flowering,+ crwn2,(but+not+as+

longer+life+span

visible+or+severe

normal round

wild+type

round,+with+ fragile+margin,+ become+irregular+ during+harsh+

smaller+cells round

round

50% 75%+-+100%

50%

25%

30%

Ave.+guard+cell+nuclear+size

50%

Ave.+endopoly+ploidy+level

98%

Sister+chromatid+cohision+and+

heterochromatin++++++++++++++++++++++n++o+++r+mal

(chromocenters,+5S+and+180bp+

100% 86%
normal

80% 98%
dispersion

50% 71%
aggregation

50% 86%
aggregation+and+ dispersion

Table 3.1 This table summarizes different phenotypic alterations observed in crwn mutants.

102

Table 3.2 Total number of tested genes and TEs
Table&2&&

All#loci Tested#loci
TE gene Significantly#mis>regulated#loci TE#(transposable#elements) gene Non>tested#loci TE#(transposable#elements) gene

crwn1
26976 19878
973 18905 1327
12 1315 7098 2449 4649

crwn2
27072 20053
977 19076
466 7
459 7019 2419 4600

crwn4
27505 20193
925 19268 5607
53 5554 7312 2667 4645

crwn1crwn2
27860 20371 1034 19337 4815
77 4738 7489 2835 4654

crwn1crwn4
27521 19957
941 19016 2126
23 2103 7564 2757 4807

Table 3.2 Summary of my mRNA-seq analysis. The total number of loci aligned with reads was counted for each crwn mutant (orange) and broken down into statistically tested loci (blue) and non-tested loci (green). Among the tested loci, significantly mis-expressed ones are displayed independently (purple). The number of gene and transposable element targets is indicated for each category.

103

To determine if this pattern was due to transposon blocks within the chromosome arms, I arranged all loci in sequential order by gene locus ID number followed by transposon locus ID number for each of the five chromosomes (Figure 3.2). The regions corresponding to transposable elements were over-expressed in the crwn mutants, particularly crwn4 and crwn1 crwn2. Table 3.1 and Figure 3.1B display significantly up-regulated and down-regulated loci using a q statistic cut-off of 0.01, almost all of which were genes rather than transposons. The number of mis-regulated loci varied across crwn genotypes: crwn4 had over 5500 mis-regulated loci, followed by crwn1 crwn2 (4800); next in abundance were crwn1 crwn4 (2300) and crwn1 (1300), while crwn2 had no more than 500 mis-expressed loci. As illustrated in Figure 3.1B, there was a skew toward higher fold changes in the up-regulated targets, especially in the crwn1 crwn2 mutant.
104

q value ( FDR adjusted P value )

Figure 1 Global pattern of transcriptional mis-regulation in crwn mutants

crwn1

crwn1

1 300

0.8 200
0.6

0.4
0.2
0 - -4 -3 -2 -1 0 1 2 3 4
crwn2
1

100
0 - -4 -3 -2 -1 0 1 2 3 4 5
crwn2
300

0.8 200
0.6

0.4 100 0.2

0 - -4 -3 -2 -1 0 1 2 3 4
crwn4
1

0 - -4 -3 -2 -1 0 1 2 3 4 5
crwn4
300

0.8 200
0.6

0.4 100 0.2

0 - -4 -3 -2 -1 0 1 2 3 4
crwn1 crwn2
1

0 - -4 -3 -2 -1 0 1 2 3 4 5
crwn1 crwn2
300

0.8 200
0.6

0.4 100 0.2

0 - -4 -3 -2 -1 0 1 2 3 4
crwn1 crwn4
1

0 - -4 -3 -2 -1 0 1 2 3 4 5
crwn1 crwn4
300

0.8 200
0.6

0.4 0.2
0 - -4 -3 -2 -1 0 1 2 3 4

100 0 - -4 -3 -2 -1 0 1 2 3 4 5

log 2 ( fold change )

log 2 ( fold change )

Figure 3.1 The global pattern of transcriptional mis-regulation in crwn mutants

105

Figure 3.1 The global pattern of transcriptional mis-regulation in crwn mutants The scatter plot in panel A (left) displays the log2 (fold change) value of all tested loci plotted against their corresponding statistic q value for each crwn mutant sample. Genes are shown in blue, and transposable elements in red. The histograms in panel B (right) show the distribution of significantly mis-regulated loci in each crwn mutant, according to number of loci (Y axis) with log2 (fold change) value falling into each interval (X axis).
106

Figure 2

Chr 1

Chr 2

Chr 3

crwn1 crwn4 crwn1 crwn2
crwn4 crwn1 crwn2
all tested loci ordered by chromosomal position

Chr 4

Chr 5

Figure 3.2 Transposon activation in heterochromatic regions 107

crwn1 crwn4 crwn1 crwn2
crwn4 crwn1 crwn2
all tested loci ordered by ID (e.g., AT1G followed by AT1TE)

Red: up regulation Green: down regulation

Chromosome arms Centromere region

Genes Transposable Elements

Figure 3.2 The heatmap displays the log2 (fold change) of all statistically tested loci in five crwn mutants in this study. A color gradient from red to green was employed: the extremely up-regulated loci (+∞) were colored in red, while the extremely downregulated loci (-∞) were colored in green. If no transcriptional change occurs, that locus will be shown in black. In the upper panel, all the loci were ordered by their chromosomal position from left to right. In the bottom panel, the TAIR ID was used as the sorting index; therefore the transposable elements (AtTE#) were distinguished from the genes (AtG#) on each chromosome.
108

Genetic interactions among the crwn mutations To investigate further the relationship of the altered transcriptomes among different crwn mutants, Venn diagrams were used to display the number and identity of the misexpressed loci from different combinations of multiple crwn mutants (Figure 3.3). The up-regulated (left) and down-regulated (right) targets are shown separately, but the patterns of overlap in the left and the right column were consistent for all three combinations. On the top row of Figure 3.3, crwn1, crwn2, and crwn1 crwn2 mutants are displayed. crwn1 and crwn2 mutants have a relatively small number of significantly mis-regulated loci, represented by small circles, and the majority of crwn2 mis-regulated targets were shared by the crwn1 genotype. However, the crwn1 crwn2 combination led to a much larger number of mis-regulated loci, indicating a synergistic relationship between the effects of CRWN1 and CRWN2 deficiencies. This pattern is consistent with our previous results that showed that a combination of crwn1 and crwn2 mutations causes a non-additive reduction in nuclear size, as well as an aggregation of heterochromatin and a dwarfing that was absent in either single mutant. Taken together, these data support the conclusion that CRWN1 and CRWN2 possess overlapping functions.
109

Figure 3

up-regulated

crwn1
crwn2
crwn4
crwn1 crwn4
crwn1 crwn2

down-regulated

Figure 3.3 The relationship among significantly mis-regulated loci in crwn mutants Figure 3.3 The Venn diagrams illustrate the number of significantly mis-regulated loci among different crwn mutants. Each circle with a distinct color represents a particular crwn mutant genotype. The size of the circle reflects the number of the loci in each genotype and the overlap between two circles reflects the number of shared mis-regulated loci.
110

The middle row in Figure 3.3 displays the relationship among mis-regulated loci in crwn1, crwn4, and crwn1 crwn4 mutants. As shown in Table 2 and Figure 3.1, the crwn4 mutant had the largest number of mis-regulated targets among all mutants tested. Interestingly, the majority of crwn1 mis-regulated loci were also aberrantly expressed in the crwn4 mutant, indicating that CRWN1 targets a subgroup of genes affected by CRWN4. However, the circle representing crwn1 crwn4 does not overlap significantly with either the crwn1 or crwn4 circle. The number of mis-regulated loci in crwn1 crwn4 is slightly larger than the number in crwn1, but much smaller than that in crwn4. This pattern indicates that the crwn1 mutation partially suppresses the effects of the crwn4 mutation. This unexpected suppression explains our previous puzzling observation that crwn1 crwn4 double mutants do not exhibit severe defects on either the nuclear or whole plant level, despite losing the major gene in the CRWN1-like clade and the only gene in the CRWN4 clade.
The bottom row of Figure 3.3 displays the large overlap between crwn4 and crwn1 crwn2 mis-regulated targets. Our previous phylogenetic and genetic analyses indicated that CRWN1 and CRWN4 are structurally and functionally diverged. CRWN2 is structurally similar to CRWN1 and these close paralogs possess overlapping functions (see above). The significant overlap of the crwn1 and crwn1 crwn2 domains with the crwn4 circle in the Venn diagram indicates that the apparently specialized CRWN1/2 and CRWN4 functions converge on a common set of transcriptional targets.
111

To extend the comparison of gene expression patterns among crwn mutants to the whole genome, I generated a heatmap that included all statistically tested loci (Figure 3.4). For each locus, a relative expression level among wild type and the five crwn mutants was calculated for every genotype, where the denominator was the sum of RPKM (Reads Per Kilobase of exon per Million fragments mapped) values for all genotypes and the numerator is the RPKM value for a specific genotype. The results were sorted by this ratio in the wild type sample and displayed in a color gradient where the lowest ratio is shown in blue and the highest in red. In this heatmap, crwn1 crwn2 showed the most contrasting color distribution compared to wild type, followed by crwn4. The similarity of these patterns indicates that the transcriptome in crwn1 crwn2 and crwn4 mutants are disturbed in a similar way. In contrast, crwn1 had a milder color shift relative to wild type, but the pattern showed some similarity to that of crwn4, indicating that crwn4 and crwn1 share common targets. The crwn2 pattern looked most similar to wild type, confirming that the crwn2 genotype had less transcriptional alteration compared to other crwn mutants. These observations are consistent with the genetic relationship among crwn mutants revealed by the Venn diagrams shown in Figure 3.3.
112

Figure 4
crwn1 crwn4 crwn1 crwn2
crwn4 crwn1 crwn2 wild type
Blue: Lowest expression among six genotypes Red: Highest expression among six genotypes
Figure 3.4 Relative expression levels of all statistically tested loci in crwn mutants
Figure 3.4 All statistically tested loci from wild type and crwn mutants were compared.For each locus, a red-blue color gradient was used to display the relative expression level among all genotypes. The genotype exhibiting the highest columnnormalized RPKM value is displayed in a red color, and the lowest in a blue color. Intermediate colors on the red-blue gradient were assigned to other genotypes corresponding to their relative RPKM value at that locus. The arrays of relative expression values were sorted by values for the wild type sample from low (blue) to high (red).
113

Epigenetic pathways are affected in crwn mutants The observation that many transposons are induced in crwn mutants prompted us to investigate whether epigenetic regulatory pathways are affected. I examined the significantly mis-expressed protein coding genes in crwn mutants, and determined if there was an enrichment of epigenetically-regulated loci, defined by their misexpression in met1, vim1 vim2 vim3, ddc, and rdd mutants where DNA methylation (met, vim, ddc) or de-methylation (rdd) pathways were impaired [141]. First, a baseline percentage of the mis-regulated loci in the whole genome was obtained for these epigenetic mutants. The ratio of epigenetically-regulated genes to total misexpressed loci in crwn mutant was then compared to the baseline percentage. Table 3.3 describes the enrichment of these epigenetically-regulated loci in each pathway for crwn mutants. Both down-regulated targets of the hypomethylation mutants met1, ddc, and vim1 vim2 vim3, and the up-regulated targets of the hypermethylation rdd mutant are enriched in crwn mutant mis-regulated targets. These observations suggest that CRWN proteins are involved in regulating the transcription of epigenetically-regulated loci. Whether the preferentially affected genes are primary or secondary targets, remain unknown.
114

Table&3&Epigenetic&pathways&3&DNA&methylation&and&small&RNA

Table 3.3 Epigenetically controlled loci are affected in crwn mutants 115

TAIR%7%%%%%%%%%%%%%%%%%%%%%#% %of%

(coding%genes)

loci

wild%type

26994

met1%up

308

met1%down

347

ddc%up

273

ddc%down

83

rdd%up

80

rdd%down

66

vim1*vim2*vim3%up

104

vim1*vim2*vim3%down 68

crwn*4*

5409

met1%up

43

met1%down

123

ddc%up

54

ddc%down

30

rdd%up

28

rdd%down

17

vim1*vim2*vim3%up

29

vim1*vim2*vim3%down 35

crwn1*crwn2

4583

met1%up

52

met1%down

94

ddc%up

53

ddc%down

40

rdd%up

22

rdd%down

15

vim1*vim2*vim3%up

35

vim1*vim2*vim3%down 26

crwn1*crwn4*

2039

met1%up

32

met1%down

70

ddc%up

38

ddc%down

24

rdd%up

14

rdd%down

11

vim1*vim2*vim3%up

19

vim1*vim2*vim3%down 14

crwn1*

1271

met1%up

15

met1%down

32

ddc%up

15

ddc%down

14

rdd%up

7

rdd%down

4

vim1*vim2*vim3%up

18

vim1*vim2*vim3%down 18

ep%i%#%%%%/%%%%%%%%%%c%%r%%w%f%o%n%%l%%d%%%%%%%/%%%%%%%%%%%n%%u%%%m%%%%%%b%%e%%r%%%%o%%f%%%%loci

up same down

1.14%

7 113 188

1.29%

17 208 122

1.01%

5 217 51

0.31%

1 67 15

0.30%

3 66 11

0.24%

7 51 8

0.39%

0.25%

up same down

0.79% 0.70 4 32 7

2.27% 1.77 8 83 32

1.00% 0.99 1 48 5

0.55% 1.80 1 24 5

0.52% 1.75 0 24 4

0.31% 1.29 3 13 1

0.54%

1.39

0.65%

2.57

up same down

1.13% 0.99 4 34 14

2.05% 1.60 7 61 26

1.16% 1.14 1 45 7

0.87% 2.84 1 32 7

0.48% 1.62 1 19 2

0.33% 1.34 4 10 1

0.76%

1.98

0.57%

2.25

up same down

1.57% 1.38 3 17 12

3.43% 2.67 5 41 24

1.86% 1.84 0 34 4

1.18% 3.83 1 21 2

0.69% 2.32 0 10 4

0.54% 2.21 2 8 1

0.93%

2.42

0.69%

2.73

up same down

1.18% 1.03 2 8 5

2.52% 1.96 3 18 11

1.18% 1.17 0 12 3

1.10% 3.58 0 13 1

0.55% 1.86 0 7 0

0.31% 1.29 1 3 0

1.42%

3.68

1.42%

5.62

epi%#%/%ge%n%%o%%t%y%%p%%%e%%%%#%%%%%%%%%%%%%%%%%c%%r%w%%%%n%%%f%%%o%%l%%d/%%%w%%%t%%%%%%%%%%%%%%%%%%%%%%%%%n%%u%%%m%%%ber%of%loci

up same down

up same down

0.03% 0.42% 0.70%

7 192 109

0.06% 0.77% 0.45%

3 328 16

0.02% 0.80% 0.19% =1%%no%enrichment

3 253 17

0.00% 0.25% 0.06% >1%%enriched

0 75 8

0.01% 0.24% 0.04% <1%%under%represented 0 80 0

0.03% 0.19% 0.03%

3 62 1

up same down up same down up same down 0.07% 0.59% 0.13% 2.85 1.41 0.19 0 42 1 0.15% 1.53% 0.59% 2.35 1.99 1.31 0 117 6 0.02% 0.89% 0.09% 1.00 1.10 0.49 0 53 1 0.02% 0.44% 0.09% 4.99 1.79 1.66 0 25 5 0.00% 0.44% 0.07% 0.00 1.81 1.81 0 28 0 0.06% 0.24% 0.02% 2.14 1.27 0.62 1 16 0

up same down up same down up same down 0.09% 0.74% 0.31% 3.37 1.77 0.44 0 48 4 0.15% 1.33% 0.57% 2.43 1.73 1.26 0 91 3 0.02% 0.98% 0.15% 1.18 1.22 0.81 0 51 2 0.02% 0.70% 0.15% 5.89 2.81 2.75 0 35 5 0.02% 0.41% 0.04% 1.96 1.70 1.07 0 22 0 0.09% 0.22% 0.02% 3.37 1.15 0.74 1 14 0

up same down up same down up same down 0.15% 0.83% 0.59% 5.67 1.99 0.85 0 29 3 0.25% 2.01% 1.18% 3.89 2.61 2.60 1 68 1 0.00% 1.67% 0.20% 0.00 2.07 1.04 0 37 1 0.05% 1.03% 0.10% 13.24 4.15 1.77 0 24 0 0.00% 0.49% 0.20% 0.00 2.01 4.81 0 14 0 0.10% 0.39% 0.05% 3.78 2.08 1.65 0 11 0

up same down up same down up same down 0.16% 0.63% 0.39% 6.07 1.50 0.56 0 14 1 0.24% 1.42% 0.87% 3.75 1.84 1.91 0 28 4 0.00% 0.94% 0.24% 0.00 1.17 1.25 0 15 0 0.00% 1.02% 0.08% 0.00 4.12 1.42 0 12 2 0.00% 0.55% 0.00% 0.00 2.25 0.00 0 7 0 0.08% 0.24% 0.00% 3.03 1.25 0.00 0 4 0

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%f%o%%%ld%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

epi%#%/%genotype%#

crwn%%%/%wt%%%

up same down

0.03% 0.71% 0.40%

0.01% 1.22% 0.06%

0.01% 0.94% 0.06% =1%%no%enrichment

0.00% 0.28% 0.03% >1%%enriched

0.00% 0.30% 0.00% <1%%under%represented

0.01% 0.23% 0.00%

up 0.00% 0.00% 0.00% 0.00% 0.00% 0.02%

same down 0.78% 0.02% 2.16% 0.11% 0.98% 0.02% 0.46% 0.09% 0.52% 0.00% 0.30% 0.00%

up 1.66

same 1.09 1.78 1.05 1.66 1.75 1.29

down 0.05 1.87 0.29 3.12

up 0.00% 0.00% 0.00% 0.00% 0.00% 0.02%

same down 1.05% 0.09% 1.99% 0.07% 1.11% 0.04% 0.76% 0.11% 0.48% 0.00% 0.31% 0.00%

up 1.96

same 1.47 1.63 1.19 2.75 1.62 1.33

down 0.22 1.10 0.69 3.68

up 0.00% 0.05% 0.00% 0.00% 0.00% 0.00%

same down 1.42% 0.15% 3.33% 0.05% 1.81% 0.05% 1.18% 0.00% 0.69% 0.00% 0.54% 0.00%

up 4.41

same 2.00 2.74 1.94 4.24 2.32 2.35

down 0.36 0.83 0.78

up same down 0.00% 1.10% 0.08% 0.00% 2.20% 0.31% 0.00% 1.18% 0.00% 0.00% 0.94% 0.16% 0.00% 0.55% 0.00% 0.00% 0.31% 0.00%

up

same 1.55 1.81 1.26 3.40 1.86 1.37

down 0.19 5.31 0.00 5.31

Table 3.3 The statistics are shown for the enrichment of epigenetically controlled loci corresponding to significantly mis-expressed loci in crwn mutants. The epigenetic pathways examined include: CpG DNA methylation guided by VIM1, VIM2, VIM3 [142, 143] and MET1 [144]; CHG/CHH DNA methylation guided by DRM1, DRM2, and CMT3 [144]; and DNA de methylation guided by ROS1, DML2 and DML3 [144]. In four crwn mutants, the total number of mis-regulated genes belonging to met1, ddc, rdd or vim1 vim2 vim3 target lists are displayed in Column B (in blue), the ratios of epigenetically regulated targets over total mis-regulated targets within each genotype were calculated as percentage in Column C (in blue), and the enrichment fold (percentage in crwn mutant / percentage in wild type) were shown in Column D (in blue). Genes mis-regulated in met1, ddc and rdd mutants background are further categorized by three classes of alterations in DNA methylation (in orange, Column E M) or small RNA expression (in purple, Column N – V) at corresponding genomic loci. Similarly, in each class of alteration, the total number, percentage and fold enrichment of epigenetically regulated loci were calculated for these four crwn mutants.
116

Not all epigenetically-regulated targets are affected in crwn mutants, however. I broke down the genes affected by the met1, ddc, and rdd pathway into sub-categories according to whether changes in small RNA accumulation or DNA methylation occurred at each locus. The most commonly enriched epigenetic targets among the mis-expressed loci in crwn mutants were the ones with little or no changes in small RNA expression and/or DNA methylation. Interestingly, a major pool of up-regulated targets in the met1 mutant (109 loci, 35%) included loci for which the transcriptional activation is coupled with reduced small RNA expression, and almost none of these loci are affected by crwn mutations. Also, the genes whose up-regulation is associated with CpG hypo-methylation due to the loss of MET1 were not over-represented in crwn mutants. A similar bias also applied to primary ddc targets, where the CHG or CHH methylation was lost; while an opposite pattern existed among the hypomethylated rdd targets, for which the de-methylation function was impaired. rdd mutations also lead to hyper-methylation, and these type of loci, despite their small number, were significantly enriched among mis-regulated loci in all crwn mutants. Taken together, this analysis suggests that crwn mutations affect a subset of epigenetically-controlled loci.
Categories of loci mis-expressed in crwn mutants
Next I explored the functional categorization of genes whose expression was disturbed in crwn mutants. Figure 3.5 displays the percentage of significantly mis-regulated loci in different sub-categories (e.g., based on cellular location, molecular function, and
117

biological process). A genome-wide baseline percentage was calculated for each category using all the annotated loci in TAIR10. I then compared the mis-expressed loci in different crwn mutants relative to this default ratio, and found varying levels of under- or over-representation among different categories. Among the different biological processes, many loci involved in response to biotic/abiotic stimuli and stress were significantly over-represented among the genes up-regulated in crwn mutants, as were signal transduction and transport loci (Figure 3.5A). Among different molecular functions, transcription factors and kinases were up-regulated (Figure 3.5B). Further, from the perspective of different cellular compartments, the proteins located in plasma membrane were, as a group, up-regulated (Figure 3.5A). These observations are consistent with the activation of signaling transduction proteins at the cell surface that are involved in stress response pathways. In contrast, genes involved in biogenesis were over-represented in the down-regulation group, but under-represented in the upregulation group. Additionally, proteins targeted to the cell wall and extracellular region were significantly down-regulated. These data suggest that homeostatic functions (e.g., development, biogenesis) are down-regulated in crwn mutants, while the defense response cascades are activated.
118

Up-regulated

Down-regulated

Up-regulated

40 30 20 10 0 40 30 20 10 0
40 30 20 10 0 30 20 10
0
Figure 3.5A Functional categorization of mis-expressed loci in crwn mutants
119

Down-regulated

Figure 5B
70 60 50 40 30 20 10 0
70 60 50 40 30 20 10 0

biological process

all crwn2 crwn1 crwn4 crwn1 crwn1

crwn2

crwn4

Figure 3.5B Functional categorization of mis-expressed loci in crwn mutants

120

Figure 3.5 Three panels display the functional categorization of genes significantly mis-regulated loci in crwn mutants depending on cellular compartment, molecular function, and biological process. In each panel, the up-regulated loci and downregulated loci are displayed separately. The Y-axis represented the percentage of misexpressed loci for each sub-category displayed on X-axis. The wild type and the five crwn mutant samples are displayed using different colors. Red rectangles were used to highlight those sub-categories that had over- or under-representation of mis-expressed loci among crwn mutants in comparison to wild type.
121

To investigate these provisional conclusions further, I looked more closely at transcription factors that were mis-regulated in crwn mutants. The major gene families affected include ERF/AP2, WRKY, NAC, MYB, and bHLH. Among them, the overrepresentation of ERF/AP2 and WRKY family proteins was the most striking (Figure 3.6A). The Arabidopsis predicted proteome contains 117 ERF/AP2 members [145], belonging to twelve classes (A1-6, B1-6) constituting 7% of all transcription factors. The enrichment of ERF/AP2 proteins was biased to the extremely up-regulated transcription factors in crwn1 (25%), crwn1 crwn2 (17%), and crwn4 (18%), but not in crwn1 crwn4 (Figure 3.6A). The loci affected encode ERF/AP2 proteins in the A1, A5, B1, and B3 groups predominantly, and involve members that operate in the response to cold, pathogen, and mechanical stresses (Supplementary Table 3.1). The WRKY family consists of 75 members [146], which comprise 4.5% of all the transcription factors genome wide. These WRKY members are enriched among the up-regulated transcription factors in all crwn mutants, especially in crwn1 crwn4 (26%) and crwn1 crwn2 (17%) (Figure 3.6A; Supplementary Table 3.1).
122

Figure 6A

100% 90% 80% 70% 60% 50% 40% 30% 20% 10%
0 whole gnome

crwn1 crwn4 crwn1 crwn1 crwn2 crwn4
up regulated transcription factors

crwn1

crwn4

crwn1 crwn2

crwn1 crwn4

down regulated transcription factors

Figure 6B

3 hr Cold

Mech stress

WRKY Others ERF/AP2

TuMV

crwn4

crwn4 Flg

3 hr Cold
crwn4
24 hr Cold
Up regulated loci

6 hr Cold Mech stress
TuMV

crwn4

Down regulated loci

Figure 3.6 The activation of stress response pathways in crwn mutants

123

Figure 3.6 Panel A uses a percentage bar graph to exhibit the enrichment of ERF/AP2 and WRKY transcriptional factors among mis-expressed loci in crwn1, crwn4, crwn1 crwn2, and crwn1 crwn4 mutants. A baseline percentage was calculated as follows: total number of ERF/AP2 family transcription factors / total number of transcription factors in the whole genome. This calculation was also applied to WRKY family. Panel B uses Venn diagrams to display the overlap of significantly misregulated loci among crwn4 and various stressed wild type plants. Three comparisons among up-regulated loci were made; and a fourth one in the green box displays several multi sample comparisons of down-regulated loci between crwn4 and several stressed wild type plants. There were no overlapping loci between the three-hour cold treatment experiment and the crwn4 sample (not shown); a few loci overlapped among the six-hour cold treatment, the mechanical stress treatment and TuMV infection.
124

Supplementary Table 3.1A The activation of stress response pathways
Supplemental*Table*1A*Stress*response*in*crwn%&%summary

TXF$total ERF/AP2
WRKY other$TXF

whole$ genome$
total 1677 117
75 1485

crwn1 93 23 11 59

up$regulation

crwn1' crwn1'

crwn4 251 43 23

crwn2 195 35 33

crwn4 76 5 20

185 127 51

down$regulation

crwn1' crwn1'

crwn1 crwn4 crwn2 crwn4

32 189 194

69

3 6 16 8

3 53 0

26 178 175

61

ERF/AP2$//$ TXF$total WRKY$/$$$ TXF$total

6.98% 4.47%

24.73% 17.13% 17.95% 6.58% 9.38% 3.17% 8.25% 11.59% 11.83% 9.16% 16.92% 26.32% 9.38% 2.65% 1.55% 0.00%

125

Supplementary Table 3.1B The activation of stress response pathways

Supplemental*Table*1B*Stress*response*in*crwn%&%ERF/AP2*family

ERF/AP2(
loci
AT1G63030 AT4G25480 AT4G25470 AT4G25490 AT5G51990 AT1G12610 AT3G11020 AT5G05410 AT2G38340 AT2G40350 AT5G18450 AT2G40340 AT1G75490 AT3G57600 AT2G40220 AT1G01250 AT1G12630 AT1G33760 At1G63040 AT1G71450 AT1G77200 AT2G25820 AT2G35700 AT2G36450 AT2G44940 AT3G16280 AT3G60490 AT4G16750 AT4G32800 AT5G11590 AT5G25810 AT5G52020 AT1G19210 AT1G21910 AT1G22810 AT1G44830 AT1G46768 AT1G71520 AT1G74930 AT1G77640 AT2G23340 AT3G50260 AT4G06746 AT4G31060 AT4G36900 AT5G21960 AT5G67190 AT4G28140 AT1G78080 AT4G39780 AT1G36060 AT4G13620 AT1G64380 AT2G22200 AT5G65130

type
DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB DREB

class
A=1 A=1 A=1 A=1 A=1 A=1 A=2 A=2 A=2 A=2 A=2 A=2 A=2 A=2 A=3 A=4 A=4 A=4 A=4 A=4 A=4 A=4 A=4 A=4 A=4 A=4 A=4 A=4 A=4 A=4 A=4 A=4 A=5 A=5 A=5 A=5 A=5 A=5 A=5 A=5 A=5 A=5 A=5 A=5 A=5 A=5 A=5 A=6 A=6 A=6 A=6 A=6 A=6 A=6 A=6

wild(
type(
0.14 1.14 0.85 0.89 0.45 2.68 3.09 3.84 0.06 0.29 0.00 0.89 1.71 8.93 0.00 9.69 0.12 0.78 2.08 0.00 0.14 5.12 2.64 0.00 15.50 1.13 3.42 5.69 121.90 4.35 16.86 0.45 0.99 2.87 1.57 1.13 14.80 0.09 6.49 0.91 51.15 5.77 16.37 3.94 16.72 1.13 16.94 34.71 223.67 12.91 0.24 0.02 51.12 15.26 0.39

crwn1&
crwn2
3.94 2.73 5.74 5.71 7.03 4.65 0.85 3.80

crwn4
3.78 2.99 4.46 3.34 4.03 3.93
1.95

crwn1 3.57
2.47 0.97

1.24 =0.97 =2.22 7.24

=2.22 4.19

=1.28 1.83

=1.50 =2.78
=1.72
=1.80
=2.48 =3.58 4.18 5.22 3.30
=0.83
5.14 4.91 =1.10 3.73

=1.55
0.65 =2.80 3.93 2.27 1.79 2.63
3.34 2.91 3.75

=1.40 2.61
1.72 1.49 2.16

2.70 2.12 4.63 2.12
0.76 2.55 0.87
0.97 0.84

=2.64 =0.81

ERF/AP2(
loci
AT1G03800 AT1G12980 AT1G24590 AT1G28160 AT1G28360 AT1G28370 AT1G50640 AT1G53170 AT1G80580 AT3G15210 AT3G20310 AT5G13910 AT5G18560 AT5G44210 AT1G72360 AT1G53910 AT2G47520 AT3G14230 AT3G16770 AT1G04370 AT1G06160 AT2G31230 AT2G44840 AT3G23220 AT3G23230 AT3G23240 AT4G17490 AT4G17500 AT4G18450 AT4G34410 AT5G07580 AT5G43410 AT5G47220 AT5G47230 AT5G51190 AT5G61590 AT5G61600 AT5G07310 AT2G33710 AT5G61890 AT5G64750 AT5G50080 AT1G43160 AT5G13330 AT4G27950 AT5G53290 AT4G23750 AT1G22985 AT1G71130 AT4G11140 AT1G15360 AT1G25470 AT1G49120 AT1G68550 AT2G20350 AT3G25890 AT5G11190 AT5G19790 AT5G25190 AT5G25390 AT5G67000 AT5G67010

type
ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF ERF

class
B=1 B=1 B=1 B=1 B=1 B=1 B=1 B=1 B=1 B=1 B=1 B=1 B=1 B=1 B=2 B=2 B=2 B=2 B=2 B=3 B=3 B=3 B=3 B=3 B=3 B=3 B=3 B=3 B=3 B=3 B=3 B=3 B=3 B=3 B=3 B=3 B=3 B=4 B=4 B=4 B=4 B=4 B=4 B=4 B=5 B=5 B=5 B=5 B=5 B=5 B=6 B=6 B=6 B=6 B=6 B=6 B=6 B=6 B=6 B=6 B=6 B=6

wild(
type(
0.06 0.00 0.00 0.00 0.14 4.04 52.37 17.09 0.00 90.93 22.76 0.04 0.00 3.68 3.26 108.09 0.04 74.20 158.64 0.00 1.71 4.79 4.47 0.00 0.54 0.48 1.40 9.75 0.00 7.83 59.75 0.00 15.54 3.23 2.93 73.06 9.82 1.22 0.28 16.67 44.74 0.00 141.77 6.82 3.22 1.28 3.26 4.74 18.19 0.81 0.00 6.72 0.00 27.91 0.26 22.64 0.17 0.30 6.04 0.13 0.07 0.16

crwn1& crwn2

crwn4

crwn1

inf

6.34 4.77 2.40 0.73 0.87 2.15 1.98 0.86
2.40 1.41 1.57 1.34
=1.19 0.98 1.35 1.26

1.21 0.96
2.99 3.35 2.57
3.57 3.10 1.79
2.72 2.99 4.17 2.60 4.79 2.28 1.42 3.28 2.84 2.62
4.91 4.96 3.66

3.15 3.02 3.18
2.86 =1.90 2.22 =2.66
=2.77
=1.36

3.92 2.14 1.55
=1.55 1.74
1.37
=1.58

3.59 1.36 0.97
=0.84
=1.34

1.28 0.98

126

Supplementary Table 3.1C The activation of stress response pathways

Supplemental*Table*1C*Stress*response*in*crwn%&%WRKY*family

gene& Locus&ID name AT2G04880 WRKY01 AT5G56270 WRKY02 AT2G03340 WRKY03 AT1G13960 WRKY04 AT5G46310 WRKY05 AT1G62300 WRKY06 AT4G24240 WRKY07 AT5G46350 WRKY08 AT1G68150 WRKY09 AT1G55600 WRKY10 AT4G31550 WRKY11 AT2G44745 WRKY12 AT4G39410 WRKY13 AT1G30650 WRKY14 AT2G23320 WRKY15 AT5G45050 WRKY16 AT2G24570 WRKY17 AT4G31800 WRKY18 AT4G12020 WRKY19 AT4G26640 WRKY20 AT2G30590 WRKY21 AT4G01250 WRKY22 AT2G47260 WRKY23 AT5G41570 WRKY24 AT2G30250 WRKY25 AT5G07100 WRKY26 AT5G52830 WRKY27 AT4G18170 WRKY28 AT4G23550 WRKY29 AT5G24110 WRKY30 AT4G22070 WRKY31 AT4G30935 WRKY32 AT2G38470 WRKY33 AT4G26440 WRKY34 AT2G34830 WRKY35 AT1G69810 WRKY36 AT5G22570 WRKY38 AT3G04670 WRKY39 AT1G80840 WRKY40 AT3G32090 WRKY40 AT4G11070 WRKY41

crwn1' crwn1'

crwn1' crwn1'

crwn1 crwn4 crwn2 crwn4 crwn1 crwn4 crwn2 crwn4

WRKY02

WRKY03

WRKY06 WRKY06 WRKY06 WRKY07 WRKY08

WRKY11 WRKY11

WRKY15 WRKY15 WRKY15 WRKY15

WRKY18 WRKY18 WRKY18 WRKY18

WRKY21 WRKY22 WRKY22 WRKY22

WRKY25 WRKY26
WRKY30

WRKY25 WRKY25 WRKY26 WRKY26
WRKY28 WRKY29 WRKY30 WRKY30

WRKY25 WRKY26

WRKY27

WRKY27

WRKY27 WRKY28

WRKY33 WRKY33 WRKY33 WRKY33

WRKY32 WRKY32

WRKY36 WRKY36 WRKY38 WRKY38
WRKY40 WRKY40 WRKY40 WRKY40
WRKY41

127

Supplementary Table 3.1D The activation of stress response pathways

Supplemental*Table*1D*Stress*response*in*crwn%&%pathway*overlap

wild% type 27416 161 60 127 194 580 536 326 1193 2859 3109 785 600 489

crwn1&
781 47 25 61 74 75 127 42 91 420 166 106 92 97

crwn4&

crwn1& crwn2&

crwn1&

crwn4&

2922 2459 534 2632

96 93 4

26

42 47 1

5

107 102 2

3

134 124 4

6

211 159 21 66

328 405 10 33

117 137 9

63

328 350 20 61

974 919 88 221

533 481 89 343

296 257 60 152

251 192 36 98

291 282 18 38

crwn1 crwn2& 2279
19 3 3 22 86 35 44 48 201 285 142 97 34

TAIR%10
crwn%mutants mechanical mec%top%60
3%h%cold 6%h%cold 24%h%cold
virus drought virus%+%drought
heat virus%+%drought%+%heat
DC3000H6 DC3000H8 E%coli%H%flg

crwn1& crwn4&
2.85% 10.66% 29.19% 59.63% 41.67% 70.00% 48.03% 84.25% 38.14% 69.07% 12.93% 36.38% 23.69% 61.19% 12.88% 35.89% 7.63% 27.49% 14.69% 34.07% 5.34% 17.14% 13.50% 37.71% 15.33% 41.83% 19.84% 59.51%

crwn1& crwn2& 8.97% 57.76% 78.33% 80.31% 63.92% 27.41% 75.56% 42.02% 29.34% 32.14% 15.47% 32.74% 32.00% 57.67%

crwn1&

crwn4&

crwn1& crwn2&

1.95% 9.60% 8.31%

2.48% 16.15% 11.80%

1.67% 8.33% 5.00%

1.57% 2.36% 2.36%

2.06% 3.09% 11.34%

3.62% 11.38% 14.83%

1.87% 6.16% 6.53%

2.76% 19.33% 13.50%

1.68% 5.11% 4.02%

3.08% 7.73% 7.03%

2.86% 11.03% 9.17%

7.64% 19.36% 18.09%

6.00% 16.33% 16.17%

3.68% 7.77% 6.95%

wild% type 27416 44
4 21 265 125 118 482 1623 3178 893 1056 621

crwn1&
534 5 0 0 21 4 6 10 88
149 64 78 59

crwn4&

crwn1c rwn2&

crwn1&

crwn4&

2632 2279 781 2922

10 7 1 10

0 00 0

4 20 0

89 68 7

19

21 16 2

9

19 31 13 26

116 100 16 71

393 341 52 173

733 637 161 472

308 274 50 135

329 256 79 203

263 213 19 55

crwn1& crwn2& 2459
4 0 0 24 13 11 55 124 403 117 196 49

TAIR%10
crwn%mutants mechanical
3%h%cold 6%h%cold 24%h%cold
virus drought virus%+%drought
heat virus%+%drought%+%heat
DC3000H6 DC3000H8 E%coli%H%flg

crwn1& crwn4&
2.85% 10.66% 2.27% 22.73% 0.00% 0.00% 0.00% 0.00% 2.64% 7.17% 1.60% 7.20% 11.02% 22.03% 3.32% 14.73% 3.20% 10.66% 5.07% 14.85% 5.60% 15.12% 7.48% 19.22% 3.06% 8.86%

crwn1& crwn2& 8.97% 9.09% 0.00% 0.00% 9.06% 10.40% 9.32% 11.41% 7.64% 12.68% 13.10% 18.56% 7.89%

crwn1&

crwn4&

crwn1& crwn2&

1.95% 9.60% 8.31%

11.36% 22.73% 15.91%

0.00% 0.00% 0.00%

0.00% 19.05% 9.52%

7.92% 33.58% 25.66%

3.20% 16.80% 12.80%

5.08% 16.10% 26.27%

2.07% 24.07% 20.75%

5.42% 24.21% 21.01%

4.69% 23.06% 20.04%

7.17% 34.49% 30.68%

7.39% 31.16% 24.24%

9.50% 42.35% 34.30%

Supplementary Table 3.1 Supplementary Table 3.1A displays the statistics of misexpressed ERF/AP2 and WRKY transcription factors in crwn mutants. The locus ID and RPKM value are listed for these enriched loci in Supplementary Table 3.1B (ERF/AP2) and Supplementary Table 3.1C (WRKY). Supplementary Table 3.1D shows the enrichment of crwn mis-regulated loci in various experimentally stressed wild type plants.

128

To investigate the activation of stress response pathways further, I compared the transcriptomic data between crwn mutants and wild type plants that were experimentally subjected to either abiotic or biotic stresses, including cold stress, mechanical stress, virus infection, bacteria treatment, heat stress and drought treatment [147-150] (Supplementary Table 3.1). The Venn diagrams in Figure 3.6B summarize the common transcriptional targets among different treatments and the crwn4 mutant. A very significant percentage (>75%) of up-regulated targets in the response to either three hours of cold treatment or TuMV (Turnip mosaic virus) infection (10 minutes post infection) belong to the up-regulated loci in crwn4 mutants, consistent with the previously noted up-regulation of ERF/AP2 and WRKY transcription factor families under these conditions. A slightly lower percentage (~60%) of up-regulated loci in the experiment involving brief mechanical stress were shared with the crwn4 mutant (Supplementary Table 3.1D; an 80% overlap of the top 60 extremely over expressed loci). However, a lower percentage of overlap was seen between crwn4 mutants and stressed wild type plants among the down-regulated loci (bottom panel of Figure 3.6, and Supplementary Table 3.1D). These data indicate that crwn mutants preferentially activate some stress response pathways.
Related to defense responses, several members of the wall associated protein kinase family [151] were up-regulated, and cell wall and extra-cellular proteins were enriched among down-regulated loci, including receptor proteins sensing various environment stimuli and enzymatic proteins regulating cell wall stiffness and expansion. These changes were most notable in crwn1 and crwn2 single mutants,
129

indicating that genes ecnoding cell wall and extra-cellular proteins are transcriptionally sensitive to even mild disturbances in the nuclei due to the loss of CRWN proteins.
Nuclear proteins mis-expressed in crwn mutants
I also attempted to understand how the transcriptional mis-regulation of loci might relate to the altered nuclear morphology seen in crwn mutants. I focused my examination on nuclear-localized proteins in three categories. First, I considered proteins in the nuclear envelope, nuclear pore complex, and putative nuclear matrix. In this category, one notable locus is the nucleoporin auto-peptidase encoded by At1g59660, which is up-regulated by five fold and three fold in crwn1 crwn2 and crwn1 crwn4 mutants, respectively. This up-regulation is unusual as transcripts for other nucleoporins were unaffected, with the exception of At2g05120 (Nup133/Nup155-like), which is mildly down-regulated (less than two fold) in the crwn4 mutant. Transcripts encoding two other nuclear membrane components, At3g13360 (WIP3) and At3g63000 (NPL41, Nuclear Pore Localization Protein 4-like protein 1) were also slightly down-regulated (less than two fold) in the crwn4 mutant.
The second class of nuclear proteins I examined are ones important in chromosome organization, including SMC (Structural Maintenance of Chromosomes) complex components and proteins involved in chromatin condensation and sister chromatid cohesion [100], as well as nuclear-localized actin-related proteins (AtARP), and
130

putative chromosome-interacting AT hook family proteins. Among SMC members crwn4 and crwn1 crwn2 mutants, while transcription for the SMC1 and SMC3 were not disturbed. The SMC5/6 complexes, which promote sister chromatid cohesion for DNA repair through homologous recombination in somatic cells [102], were severely disrupted. The SMC6A subunit was significantly under-expressed in crwn1, crwn4, and crwn1 crwn2 mutants, as were non-SMC proteins necessary for sister chromatin cohesion (SYN3, ETG1 and CTF18, Nse1) particularly in crwn4. These changes, together with the alterations in condensin and cohesin formation, might cause the chromocenter diffusion seen in crwn4 nuclei ([140], Chapter2, Table 3.4A). On the other hand, SMC6B and non-SMC members in the complex Nse4 and HEB2, were up-regulated in crwn1 crwn2 and crwn1 crwn4 mutants (Table 3.4B), which could be related to the chromocenter aggregation seen in these mutants. Among the actinrelated proteins, nuclear localized AtARP8 (encoded by At5g56180) was up-regulated in crwn4. Among the approximately 50 members in the AT hook protein family, four members were down-regulated in the crwn4 mutant, including At1g63470, At1g63480, At2g45850, and At3g04590 (Table 3.4A) .
Finally, I considered proteins serving basic nuclear functions in DNA and RNA metabolism. In total, there are 29 genes up-regulated and 54 genes down-regulated genes in this class in crwn1 crwn2 mutants. In crwn4 mutants, there are 34 genes upregulated and 120 genes down-regulated belonging to this class. Noticeably, the loss of CRWN1 protein alone leads to many changes in this category, although small and round nuclei are the only obvious phenotypes observed in crwn1 mutant. First, a set of
131

endo-reduplication regulators was significantly mis-expressed, including LGO and TCP15, which were down-regulated, and KRP1, which was up-regulated. These alterations should prohibit the cells from entering the endo-cycles [152]. Meanwhile, the negative regulator of endo reduplication, cycle2,3A is significantly downregulated, possibly as a compensation for mis-expression of LGO, TCP15, and KRP1. (Table 3.4C) The more severe crwn1 crwn2 and crwn4 mutants also exhibited a significantly reduced expression level of SIM and KRP2, the key positive regulators of endo cycles (Table 3.4C) [153]. Second, several key epigenetic modifiers (MET1, VIM1, SUVR2) were expressed at a lower level in crwn1, and this down-regulation was significantly expanded to many other proteins in severe crwn mutants, including a 50% reduction of DDM1 expression in crwn4 mutants (Table 3.4E). The misexpression of the epigenetic modifiers for DNA methylation was coupled with transcriptional changes of histone modification regulators. These changes are correlated with visible changes in epigenetic mis-regulation in crwn mutants (Figure 3.2 and Table 3.3). A few examples of nucleosome assembly components (histone H2B, H2A) were under-expressed in crwn1, and roughly 10 to 20 of histone subunits of H2A/B, H3, H4 were added into this down-regulated list in crwn1 crwn2 and crwn4 mutants (Table 3.4G and Table 3.4H). These results suggest that crwn mutations might cause open chromatin configurations. In addition, DNA polymerase and many other subunits of the DNA replication machinery (DNA pol alpha, MCM2, MCM5, ANP3, SMC6A, TPX2, RNR1, RPA70D) are expressed at a lower level in crwn1 mutants, and this list grew in crwn1 crwn2 (e.g., cyclinB2, involved in mitotic cell cycles) and crwn4 mutants (ATX1, ORC complexes, DNA pol A, I, III, alpha, gamma, Y family,
132

etc.), consistent with hindered endo-cycles, decreased endopolyploidy and reduced nuclear area. These changes might also affect regular mitotic cycles (Table 3.4D). Lastly, an over-expression of DNA repair enzymes (DNA glycosylase, SNM1, BARD1, BBX2) participating in base excision, non-homologous and homologous recombination mediated DNA repair was observed in crwn1 mutants. In one particularly striking example, a DNA glycosylase encoded by At1g75230 was upregulated 10 fold in crwn4, indicating an activation of base excision DNA repair. These changes were associated with a broad activation of other DNA repair factors (DAN1, RAD21, RAD23B, PARP2, CEN2, REV1, RAD51, XRI1, RECQ2) in more severe crwn mutants (Table 3.4F). Taken together, these changes suggest that the loss of CRWN proteins interrupt various basic nuclear processes, including DNA replication, endo-cycle regulation, epigenetic modification, nucleosome assembly and DNA repair.
133

Table&4A&&Nuclear&envelope&&proteins,&AT&hook&containing&proteins,&Actin&related&proteins

Locus&ID

gene&name

Nucleoporin&

AT1G59660 autopeptidase

AT1G79280 NUA

Nup133/Nup155F

AT2G05120 like&nucleoporin

AT3G13360 WIP3

AT3G63000 NPL41

AT1G63470

AT1G63480

&AT2G45850

&AT3G04590

AT1G13180

AT5G56180

categories Nuclear& envelope
ATFhook At&ARP

wild&type
2.09 19.23
3.61 33.38 32.75 5.59 5.63 17.72 27.88 7.85 9.96

crwn1

crwn4
2.38 21.43 23.31 2.46 2.14 10.60 19.50 5.24 16.40

crwn1' crwn2
11.57

crwn1' crwn4
6.97

crwn2 12.38

Table&4B&&SMC&complexes&subunit&and&interacting&proteins

Locus&ID AT3G47460 AT5G61460 AT3G16730 AT1G51130 AT5G15920 AT2G27170 AT3G54670 AT3G57060 AT5G21140 AT5G62410 AT3G23890 AT5G48600 AT1G04730 AT2G40550 AT3G49250 AT5G07660 AT3G59550

gene&name SMC2 SMC6B HEB2 Nse4 SMC5 SMC3 SMC1
Nse1 SMC2 TOPII SMC4 CTF18 ETG1 DMS3/IDN1 SMC6A SYN3

complex codensin& SMC5/6& condensin& SMC5/6& SMC5/6& cohesin& cohesin& condensin& SMC5/6& codensin& cell&division codensin& cohesion& cohesion&
DDR SMsisCt5e/r&6 chromatid&

wild&type 0.53 2.03 0.93 1.22 3.60 13.13 7.68 1.12 1.70 1.01 0.55 1.14 0.57 1.84 3.57 1.11 0.43

crwn1 0.45 2.25 0.91 1.04 3.60 12.13 7.67 1.07 1.61 0.64 0.23 0.58 0.32 0.79 2.23 0.27 0.09

crwn4 0.54 2.05 0.94 1.17 3.44 11.54 6.62 0.85 1.16 0.54 0.25 0.50 0.24 0.74 1.40 0.15 0.05

crwn1' crwn2 0.40 3.15 1.77 1.71 3.58 11.99 7.45 0.81 1.40 0.63 0.20 0.67 0.48 2.03 1.90 0.34 0.10

crwn1' crwn4 0.53 2.91 1.28 1.52 4.17 16.37 10.70 1.05 1.75 1.03 0.33 0.92 0.66 1.58 2.80 0.88 0.26

crwn2 0.48 1.93 1.04 1.39 3.01 10.13 6.32 0.96 1.53 0.72 0.30 0.68 0.53 1.39 3.38 0.62 0.23

134

Table&4C&Endo-reduplication&regulators

locus&ID AT3G59550 AT1G15570 AT2G42260 AT3G10525 AT5G67100 AT3G60840 AT1G03780 AT3G48160 AT3G50630 AT1G69690 AT1G33240 AT2G40550 AT5G04470 AT2G32710 AT1G20330 AT2G23430 AT4G22910 AT1G77390 AT3G19150 AT1G75950 AT3G08690 AT5G10440 AT2G21550

gene&name RAD21,&SYN3 cyclin&A2;3
PYM,&UVI4 LGO
ICU2,&DPA MAP65 TPX2 DEL1 KRP2 TCP15 GTL1 ETG1 SIM KRP4 CVP1 KRP1 CCS52A1 CYCA1 KRP6 ASK1 UBC11 CYCD4

widlt&type 0.44 1.58 0.48 131.39 1.08 0.71 0.42 0.56 10.68 34.54 39.15 1.87 18.15 12.40 99.75 38.05 8.25 0.95 17.74 266.31 51.64 0.69 0.60

crwn4 0.05 0.34 0.11 31.52 0.34
0.21 4.07 13.38 15.50 0.75 9.24 7.34 61.97 24.35 5.30 3.77 56.15 378.61 70.77

crwn1'crwn2' crwn1'crwn4 0.10 0.28
42.41
0.24 0.15
5.24 9.43 24.22
6.46 9.41
59.85

2.88 52.34
79.98 1.63 1.42

27.92

crwn1 0.48 60.51
21.07
37.91

Table&4D&DNA&replication&machinery

locus&ID

gene&name wild&type crwn4&

AT4G38680

CSP2

AT5G37630 EMB2656

AT5G43990

SUVR2

1.12

0.32

AT5G45720 DNA&Pol&III

1.52

0.78

AT5G07660

SMC6A

1.13

0.15

AT1G20720 RAD3Llike

0.43

0.06

AT3G02820

1.82

0.32

AT5G61000

RPA70D

1.84

0.32

AT3G14890

4.87

1.06

AT2G20980

MCM10

0.72

0.16

AT3G06030

ANP3

2.27

0.52

AT5G41880 DNA&Pol&A

1.42

0.33

AT1G75150

0.64

0.16

AT5G44635

MCM6

1.00

0.25

AT3G121C7h0aperone&DnaJLdomain 0.76

0.19

AT4G12620

ORC1B

0.50

0.13

AT5G15510TPX2&protein&family 1.48

0.43

AT5G13060

ABAP1

0.49

0.15

AT3G23740

0.68

0.22

AT1G67630 DNA&Pol&alpha&2 1.13

0.36

AT5G52950

0.54

0.18

AT1G44900

MCM2

1.50

0.51

AT2G24970

1.75

0.60

AT2G07690

MCM5

1.81

0.63

AT2G37560

ORC2

0.49

0.18

AT2G16440

MCM4

1.22

0.46

AT3G49250

DMS3

3.62

1.42

AT4G14770

TCX2

0.95

0.39

AT1G04730

CTF18

0.58

0.24

AT5G16690

ORC3

0.66

0.28

AT5G48600 AT1G63470

SMC3 AT&hook&

1.16 5.59

135

0.50 2.46

AT3G23890

TOPII

0.55

0.26

AT2G31650

ATX1

1.25

0.60

AT5G23420

HMGB6

5.82

2.82

crwn1& crwn2
0.52 0.23

crwn1& crwn4 48.86 0.63 0.58 0.88

crwn1 0.27

0.70 1.18 0.24 0.28

0.55 0.89

0.43 0.19

0.52
0.61 0.75

0.21

AT5G44635

MCM6

1.00

AT3G121C7h0aperone&DnaJLdomain 0.76

AT4G12620

ORC1B

0.50

AT5G15510TPX2&protein&family 1.48

AT5G13060

ABAP1

0.49

AT3G23740

0.68

AT1G67630 DNA&Pol&alpha&2 1.13

AT5G52950

0.54

Table 4DAATTc12oGGn4244ti99n0700ue MCM2

1.50 1.75

AT2G07690

MCM5

1.81

AT2G37560

ORC2

0.49

AT2G16440

MCM4

1.22

AT3G49250

DMS3

3.62

AT4G14770

TCX2

0.95

AT1G04730

CTF18

0.58

AT5G16690

ORC3

0.66

AT5G48600

SMC3

1.16

AT1G63470 AT&hook&

5.59

AT3G23890

TOPII

0.55

AT2G31650

ATX1

1.25

AT5G23420

HMGB6

5.82

AT2G21790

RNR1

9.54

AT3G42660

1.54

AT4G18820 DNA&Pol&III

9.49

AT5G13960

SUVH4

3.50

AT2G36200

0.67

AT5G62410

SMC4

1.03

AT4G24790 DNA&Pol&III

1.88

AT4G36180

2.46

AT3G54750

3.06

AT2G16390

DRD1

2.45

AT1G14460 DNA&pol&III

3.86

AT3G20540 DNA&Pol&I

10.32

AT3G22780

TSO1

9.96

AT1G50840 DNA&Pol&gama2 16.43

AT1G76310 CYCLIN&B2

0.99

AT3G20150

0.44

AT4G33400

2.71

AT2G06510

RPA1A

AT3G27060

TSO2

19.89

AT4G02070

MSH6

AT4G19130 RPA1&related

AT4G37490

CYCB1

AT1G33420

2.55

AT1G80190

PSF1

4.03

AT2G21660

GRP7

1729.44

AT4G36020

CSDP1

16.98

AT5G09790 ATXR5/SDG15

2.40

AT5G49570

PNG1

10.94

0.25 0.19 0.13 0.43 0.15 0.22 0.36 0.18 0.51 0.60 0.63 0.18 0.46 1.42 0.39 0.24 0.28 0.50 2.46 0.26 0.60 2.82 4.70 0.76 4.74
1.79 0.35 0.54 1.00 1.39 1.75 1.45 2.41 6.64 6.61 11.68

0.43 0.19
0.21 6.09 0.23
0.27 0.12 1.34 48.25

4.52 9.56 3321.58 28.78
16.21

7.14
29.99 4.94

6.25 36.61 2.59 0.88 1.56

0.52 0.61 0.75
4.68

Table&4E&DNA&methylation&modifiers

locus&ID

gene&name

AT5G43990

SUVR2

AT4G36180 LLR&kinase&family

AT5G25590

AT4G37750

CKC1

AT2G28290

CHR3

AT3G12710 DNA&glycosylase

AT2G36490

ROS1

AT1G04020

BARD1

AT5G08020

RPA70B

AT5G57970 DNA&glycosylase

AT5G07660

SMC6A

AT3G06030

ANP3

AT5G15510 TPX2&protein&family

AT3G23740

AT1G67630 DNA&Pol&alpha&2

AT5G52950

AT1G44900

MCM2

AT2G07690

MCM5

wild&type 1.12 2.46 3.27 1.31 12.71
16.17 1.11 0.88 2.40 1.13 2.27 1.48 0.68 1.13 0.54 1.50 1.81

crwn4 0.32 1.39 0.72 0.55 8.79
11.83 0.21 0.18 0.90 0.15 0.52 0.43
13600..2326
0.18 0.51 0.63

crwn1' crwn2 0.52
0.71
8.15 0.35
0.96 0.35 0.28 0.43

crwn1' crwn4 0.58 1.56 1.21 0.51
1.61

cwn1 0.42 1.17
0.27 0.89 0.52

0.61 0.75

Table&4E&DNA&methylation&modifiers

locus&ID

gene&name

AT5G43990

SUVR2

AT4G36180 LLR&kinase&family

AT5G25590

Table 4EAATTc42GGo32n7872ti59n00 ue

CKC1 CHR3

AT3G12710 DNA&glycosylase

AT2G36490

ROS1

AT1G04020

BARD1

AT5G08020

RPA70B

AT5G57970 DNA&glycosylase

AT5G07660

SMC6A

AT3G06030

ANP3

AT5G15510 TPX2&protein&family

AT3G23740

AT1G67630 DNA&Pol&alpha&2

AT5G52950

AT1G44900

MCM2

AT2G07690

MCM5

AT2G16440

MCM4

AT3G49250

DMS3

AT5G48600

SMC3

AT1G63470 AT&hook&motif

AT2G31650

ATX1

AT3G42660

AT5G13960

SUVH4

AT2G36200

AT5G62410

SMC4

AT3G54750

AT2G16390 CHR35,&DRD1

AT3G22780

TSO1

AT2G40550

ETG1

AT3G14980 ROS4,&IDM1

AT4G29360

AT3G12550

FDM3

AT1G57820 ORTH2,&VIM1

AT5G49160

MET1

AT2G23380

CLF

AT5G04290

KTF1

AT4G19020 AT3G23890 AT1G76310 AT3G20150 AT4G33400 AT1G03780 AT4G14200 AT5G48360 AT5G66750 AT3G57300 AT4G37490 AT5G04290 AT5G04560 AT5G20850 AT1G15910 AT1G63020 AT1G80420 AT2G30280 AT3G10010 AT5G09790 AT5G20320

CMT2 TOPII CYCLIN&B2
TPX2
DDM1 INO80 CYCB1 SPT5L DME RAD51 FDM1,&IDP1 NRPD1 XRCC1 DMS4,&RDDM4 DML2 ATXR5 DCL4

wild&type 1.12 2.46 3.27 1.31 12.71
16.17 1.11 0.88 2.40 1.13 2.27 1.48 0.68 1.13 0.54 1.50 1.81 1.22 3.62 1.16 5.59 1.25 1.54 3.50 0.67 1.03 3.06 2.45 9.96 1.87 7.80 1.70 5.73 1.82 2.18 4.14 6.45
5.61 0.56 0.99 0.44 2.71 0.42 1.47 2.41 0.88
3.50 0.80 10.32 6.81 2.86 2.39 4.70

crwn4 0.32 1.39 0.72 0.55 8.79
11.83 0.21 0.18 0.90 0.15 0.52 0.43 0.22 0.36 0.18 0.51 0.63 0.46 1.42 0.50 2.46 0.60 0.76 1.79 0.35 0.54 1.75 1.45 6.61 0.75 3.85 0.93 3.16 0.49 0.76 2.23 3.65
3.76
0.36
5.33
18.45 10.76
4.97 7.21

crwn1' crwn2 0.52
0.71

crwn1' crwn4 0.58 1.56 1.21 0.51

cwn1 0.42
1.17

1.61 8.15 0.35

0.96 0.35 0.27 0.28 0.89 0.43 0.52

0.61 0.75
1.94

0.23

0.51
0.62 0.74 0.98

0.21 0.27 0.12 1.34 0.15 0.35 1.25
5.99 1.48 18.57
5.60
7.60

30.66 1.56 9.30 15.88 1.32

137

Table&4F&DNA&repair&pathways

wild&

locus&ID

gene&name

type crwn4

AT1G04020

BARD1

1.11 0.21

AT5G08020

RPA70B

0.88 0.18

AT4G17760 damaged&DNA&binding 0.75 0.17

AT5G24280

GMI1

1.82 0.64

AT5G57970 DNA&glycosylase

2.40 0.90

AT5G64630

FAS2

1.46 0.55

AT5G44680 DNA&glycosylase 61.65 24.78

AT3G46940

DUT1

4.19 1.71

AT4G00020

BRCA2A

0.88 0.39

AT2G01750

MAP70L3

3.70 2.33

AT1G31360

RECQ2

6.06 4.02

AT3G23580

RNR2A

28.67 19.23

AT5G66050 wounding&responsive 29.63 21.42

AT2G36490

ROS1

16.17 11.83

AT1G57820

ORTH2,&VIM1

1.84

AT3G49250

DMS3,&IDN1

3.65

AT5G07660

SMC6A

1.14 0.15

AT1G60930

RECQ4B

1.86

AT1G02730

CSLD6,&SOS6

1.77

AT5G43080

Cyclin&A3

1.55

AT1G11190

BFN1,&ENDO1

2.41

AT1G02730

SOS6

AT2G06510

RPA1A

3.97

AT4G19130

RPA1&related

0.37

AT4G21070

BRCA1

0.54

AT4G37490

CYCB1

0.27

AT5G03780

TRFL10

5.38

AT5G04560

DME

11.03

AT5G20850

RAD51

0.52

AT5G24280

GMI&1

1.79

AT5G49570

PNG1

10.78

AT3G27060

TSO2

19.89

AT4G02390

PARP2

0.98

AT4G37010

CEN2

2.53

AT5G44750

REV1

11.16

AT5G48720

XRI

5.32

AT1G78870

UBC13A

107.642 149.71

AT5G58720

26.2747 73.254

AT1G75230 DNA&glycosylase

11.16 108.16

AT3G26680

SNM1

2.45 9.31

AT5G15850

BBX2

12.23 36.94

AT5G58720

26.27 73.25

AT3G04620

DAN1

2.50 4.99

AT1G80420

XRCC1

10.24 18.45

AT1G79650

RAD23B

17.57 28.05

AT1G78870

UBC13A

107.64 149.71

AT3G10010

DML2

2.86

AT3G12040

6.10

AT5G20320

DCL4

4.70

crwn1' crwn2 0.35
0.96 21.35 0.34
8.15 0.62 1.94 0.35 0.90 0.52 0.48 0.57
48.25 4.60 11.56 20.57 10.54
30.06 4.83 68.49 18.57
5.60 14.42 7.60

crwn1' crwn4 crwn1
0.38
0.27
1.00 0.88 6.25 0.88 1.23 1.56 10.84 15.88 1.32 3.45 16.00 36.61 7.79 8.71 15.75 14.07
69.76 5.12 24.38

138

Table 3.4 This series of tables summarize the categories of nuclear proteins significantly mis-regulated in crwn mutants, including: nuclear envelope, AT hook and ARP (actin related protein) (Table 3.4A); SMC complexes relevant (Table 3.4B); endo-reduplication cycle regulators (Table 3.4C); DNA replication machinery (Table 3.4D); DNA methylation modifiers (Table 3.4E); DNA repair (Table 3.4F); histones (Table 3.4G), and histone modifiers (Table 3.4H). RPKM values were displayed for each locus in different genotypes. Red boxes contain up-regulated loci and green boxes contain down-regulated loci.
139

Transcriptional variation among crwn1 crwn2 replicates To investigate the quality of these transcriptome datasets, I also checked the mRNA expression of CRWN genes, and evaluated the within-genotype variation through comparisons of each mutant sample to the three wild type samples individually. I found that in one of the three crwn1 crwn2 replicates, both CRWN3 and CRWN4 mRNA expression level dropped significantly. Interestingly, examination of the distribution of sequence reads (Figure 3.7A) showed a dramatic reduction of CRWN3 and CRWN4 mRNA in the 5’ region of the gene to the beginning of the sixth exon. This apparent post-transcriptional silencing of CRWN3 and CRWN4 is associated with variation of the crwn1 crwn2 transcriptome (Figure 3.7B): the mis-regulated profile of the particular crwn1 crwn2 replicate with partially silenced of CRWN3 and CRWN4 (12-1187), differs from the profiles of the other two crwn1 crwn2 replicates (12-1221, and 12-1222). Despite this variability within the crwn1 crwn2 samples, the relationships among the crwn1 crwn2 genotype and other mutants were not altered. As represented in Figure 3.7B, all three replicates in crwn1 crwn2 had a similar number of mis-expressed loci. Also, the average profile of crwn1 crwn2 displayed a coherent group of targets among replicates (Supplementary Figure 3.1A). Moreover, the Venn diagram pattern of significantly mis-regulated loci between representative crwn1 crwn2 replicates (12-1187 or 12-1222) and other crwn mutants (crwn1, crwn2, and crwn4) remained similar compared to the previous Venn diagram comparisons using the average crwn1 crwn2 profile (Supplementary Figure 3.1B, C, D, and Figure 3.3). Therefore, the transcriptional variation induced by the silencing of CRWN3 and
140

CRWN4 loci in one of the crwn1 crwn2 replicates does not change the overall relationship between crwn1 crwn2 and other crwn mutants on the transcriptomic level.
141

Figure 3.7 Partial silencing of CRWN3 and CRWN4 in one of the crwn1 crwn2 replicates
142

Figure 3.7 Panel A displays the alignments of reads to the four CRWN loci in the three crwn1 crwn2 replicates. In the CRWN1 and CRWN2 loci, the T-DNA insertions are positioned at the beginning of the sixth and longest exon. As expected, very few reads are present downstream of the insertion point, while the mRNA expression upstream of the T-DNA insertion is present. In two replicates, reads align with comparable depth across all exons of CRWN3 and CRWN4. However, in replicate C, the mRNA upstream of the T- DNA insertion is largely reduced, and this reduction extends to the largest exon. In Panel B, The Venn diagrams illustrate significantly misregulated loci among the three crwn1 crwn2 replicates. Replicate A, B, and C of crwn1 crwn2 is represented in red, yellow and green, respectively. The size of the circle reflects the number of the loci, and the overlap between two circles reflects the number of shared mis-regulated loci between two crwn1 crwn2 replicates.
143

Supplementary Figure 3.1 Partial silencing of CRWN3 and CRWN4 proteins in one of the three crwn1 crwn2 replicates
144

Supplementary Figure 3.1 The Venn diagrams illustrate the significantly misregulated loci among representative crwn1 crwn2 replicates and the average profile of different crwn mutants. Each circle with a distinct color represents a crwn mutant genotype or a crwn1 crwn2 replicate. The size of the circle reflects the number of the loci, and the overlap between two circles reflects the number of shared mis-regulated loci between two crwn samples.
145

Discussion
In this study, I performed mRNA-seq profiling of selected crwn mutants in A. thaliana, and revealed significantly altered transcriptomes due to the loss of CRWN proteins. The comparison of the transcriptomic data from different crwn mutants (Figure 3.3) confirmed the synergistic interaction between crwn1 and crwn2 mutations, and uncovered the functional suppression between crwn1 and crwn4 mutations - both suggested by our initial genetic analysis [140]. Further, the transcription profiling indicated that CRWN1-like and CRWN4-like genes affect the expression of similar sets of genomic loci. It is likely that CRWN1 and CRWN4 proteins form a functional complex in the cell (see Chapter 4), providing a mechanistic explanation for the overlapping sets of transcriptional targets in crwn1 crwn2 and crwn4 mutants. Our previous interpretation of the phenotypic suppression between crwn1 and crwn4 mutations rested on the model that these proteins had at least partially antagonistic functions. However, my transcriptomic results point to a more straightforward hypothesis that CRWN1 and CRWN4 are working in concert and that the counterintuitive genetic suppression result is possibly due to compensation by CRWN2 and/or CRWN3, which are up-regulated in crwn1 crwn4 mutants (see Chapter 4). These considerations demonstrate the importance of combining transcriptomic data with traditional genetics analysis to aid in the interpretation of genetic interactions.
To understand how CRWN proteins affect transcription, I first examined whether the previously described phenotypic changes (Table 3.1) were directly correlated with the
146

severity of global transcriptional mis-regulation. First, crwn1 crwn2 mutants had the most dwarfed stature and the smallest nuclei among the five tested crwn mutants, consistent with the most severe transcriptional changes observed in this double mutant. On the other hand, crwn2 mutants, which resemble wild type morphologically and display only mild and variable effects on nuclear size, exhibited very limited changes in gene expression. At these extremes on the phenotypic spectrum, these two crwn genotypes illustrated an overall consistency between morphological and transcriptional alterations. However, this simple correlation does not always hold true. crwn4 mutants resemble wild type plants morphologically, but their nuclear size is decreased by 50%; while crwn1 crwn4 plants are slightly dwarfed, with shorter internodes, more branches, and a more extreme reduction in nuclear area to 25% of the wild type value (see Chapter 2). These facts contradict the observation that crwn4 mutants had more widespread transcriptional changes that were largely suppressed in crwn1 crwn4 mutants. Moreover, our previous cytological investigation of crwn mutants revealed that chromocenters were diffuse in crwn4 nuclei but aggregated in crwn1 crwn2 nuclei, yet these two genotypes displayed highly similar transcriptomes. Thus, the dispersion and aggregation of heterochromatin themselves are not associated with transcriptional activation or silencing, respectively. I note, however, that our analysis was limited to poly adenylated transcripts and that I was unable to detect transcription from heterochromatic repeats. Nonetheless, the loci mis-regulated in crwn mutants are well dispersed in the genome (data not shown), suggesting that an underlying disruption of nuclear organization in general is more likely responsible for these genome-wide transcriptional alterations. Although the overall morphological
147

changes seen in particular crwn mutants roughly correlate with the level of transcriptional disruption observed, no single phenotypic alteration predicts the extent of gene mis-expression at the global level.
I then tried to associate mis-expressed loci with corresponding alterations of nuclear function in crwn mutants, and considered which were likely primary changes versus downstream effects. Our transcriptomic data fits the original idea that CRWN proteins function as structural components of the nuclei because many of the alterations I noted affected nuclear proteins. I hypothesize that the loss of CRWN proteins primarily alters nuclear organization. These spatial changes could directly affect gene expression due to the mis-positioning of either chromosomes or various transcriptional machineries. The disruption of a scaffolding system in the nuclei could also impair many other nuclear functions, and result in transcriptional mis-regulation indirectly.
One category of nuclear functional change I investigated was epigenetic alteration in crwn mutants in an attempt to examine potential mechanistic connections between altered heterochromatin organization and genome-wide transcriptional changes. Our data uncovered a mild suppression of transposon silencing concentrated in the heterochromatic regions in crwn mutants (Figure 3.2). In addition, protein-coding genes regulated by epigenetic pathways are enriched among the significantly misexpressed loci (Table 3.2), suggesting an impairment of epigenetic regulation in crwn mutants. The majority of the enrichment involved loci that are down-regulated in met1 and ddc, or up-regulated in rdd mutations, rather than the more typical examples of
148

coding genes activated by a loss of DNA methylation. The small RNA-regulated genes in the MET1 pathway were mostly undisturbed, implying that the RNA-directed CpG DNA methylation is not visibly affected by crwn mutations. Consistent with these observations, I observed a down-regulation of many nuclear proteins important for maintaining a dynamic balance of DNA methylation (Table 3.3). The mis-regulation of various epigenetic modification pathways provides a straightforward explanation for the broad but subtle reduction in epigenetic silencing in both transposon activation and enrichment of epigenetically-targeted coding genes. Another aspect reflecting open chromatin is the down-regulation of histone modifiers and under-expression of many histone proteins in crwn mutants, especially crwn4. These lowered activities may also attenuate the self-reinforced feedback loop for the dynamic maintenance of epigenetic modification. All of these changes could lead to downstream transcriptional mis-regulation. In the future, genome-wide profiling of different epigenetic marks will be needed to determine which genomic regions are primary targets for epigenetic regulation sensitive to a loss of CRWN protein function.
Overall, the mis-expression of epigenetic regulators is most significant in crwn4 mutants, consistent with the dispersion of heterochromatic regions in this mutant. However, the aggregation of chromocenters and pericentromeric repeats suggests that heterochromatin is more condensed in crwn1 crwn2 nuclei, despite the fact that these double mutants shared a significant portion of the mis-regulated epigenetic modifiers observed in crwn4 mutants. This result suggests that the apparent diametrically
149

opposed conformation of heterochromatin in crwn4 and crwn1 crwn2 mutants are built upon similarly disrupted epigenetic pathways.
Another major observation in this study concerns the over-expression in crwn mutants of transcripts responding to biotic and abiotic stresses, featuring the cascades executed by the ERF/AP2 and WRKY transcription factors. The up-regulation profiles in crwn mutants covered the majority of over-expressed transcripts observed in other studies for plants subjected to a short cold treatment, a transient virus infection, or a brief mechanical stress. However, the up-regulated loci after a prolonged cold treatment, or down-regulated loci in all the studies mentioned above, did not match well with crwn transcriptional profiles. Therefore, crwn cells do not respond as if subjected tochronic external stresses. Rather, the mRNA-seq profile of crwn mutants exhibited a bias toward a transcriptional activation of pathways highly sensitive to acute environmental stress.
I hypothesize that crwn mutations mis-regulate a set of molecular switches for prompt stress response, via conformational changes of either the chromosomes or their local nuclear environments, directly affecting the transcription of corresponding loci. Future investigation should look at the conformation and positioning of the genomic loci responsible for activation of ERF/AP2 and WRKY pathways. For example, release from the nuclear periphery, where CRWN proteins are concentrated, might be involved in activation of these loci.
150

I also described the enrichment of cell wall and extra-cellular components in downregulated targets in crwn mutants. It is known that during biotic infection, the increased production of salicylic acid (SA) is often associated with the loosening of pectin matrix [154], and down-regulation of lignin synthesis or pectin matrix production will also induce SA-mediated stress responses [155]. Consistent with this interpretation, key regulators of SA production are all up-regulated in crwn4 and crwn1 crwn2 mutants, including NPR1, EDS1, ALD1 and SID2 [156]. However, the activation of defense responses in crwn mutants does not fully explain the enrichment of cell wall and extra-cellular components among down-regulated loci. One reason is these proteins are also among up-regulated loci in response to stress, such as Pseudomonas syringae DC3000 and TuMV-treated plants (Supplementary Table 3.1D). Yet, these loci are not up-regulated significantly in crwn mutants. Also, during cold acclimation, the cell wall rearranges its structure to protect the internal membrane system, and mechanical stress usually enhances stiffness of cell walls. Both responses require extra metabolism of cell wall and extra-cellular components. Other potential mechanisms might also contribute to the enrichment of down-regulated cell wall proteins. For example, cell wall metabolism is highly active during cell division and expansion, and the lowered metabolism is consistent with the reduced DNA replication activity and endo-reduplication cycles in crwn mutants. It is also possible that the cell wall metabolism is directly blocked by transcriptional landscapes in crwn nuclei that contribute to the activation of various stress response pathways.
151

Finally, I focused on structural nuclear proteins to understand the mechanisms through which nuclear organization might be altered in crwn mutants. Despite the nuclear periphery localization of all CRWN proteins [78, 81], nuclear envelope components, including SUN-domain proteins and most nucleoporin subunits, generally retain their normal transcript expression levels in crwn mutants. These data suggest that these structures at the nuclear periphery are not communicating with CRWN proteins. However, I elicited both over- and under-expression of transcripts encoding SMC complex proteins in different crwn mutants. In yeast, mouse, and human, these complexes are involved in chromosomal condensation and sister-chromatin association in interphase, and localize to heterochromatic chromosomes during mitosis and meiosis [52]. In yeast, SMC5 and SMC6 proteins were enriched at ribosomal RNA genes and at some telomeres. smc5 and smc6 cells exhibit mitotic aberrations involving impaired chromosome segregation of repetitive regions [157]. The disrupted SMC5/6 complexes lead to defects in HR repair, increase DNA damage, and ectopic activation of non-homologous end joining (NHEJ) repair pathways [158]. Similarly, both SMC complexes and CRWN1 proteins localize to mitotic chromosomes in Arabidopsis.. It has also been shown that the SMC5/6 complex in Arabidopsis is important for homologous recombination-mediated double-strand break (DSB) repair via sister chromatin cohesion [102]. It is possible that the SMC complexes interact with CRWN proteins functionally and physically, and adjust their expression level in response to the loss of CRWN proteins.
152

Our analysis of nuclear proteins also uncovered a general down-regulation of both positive and negative regulator proteins of endo-reduplication cycles. The combination of these changes predicts reduced endopolyploidy levels especially in severe crwn mutants, because the down-regulation of positive regulators, such as LGO and SIM [152, 153], is more prominent. Indeed, the severe crwn mutants do exhibit a reduction in endopolyploidy (see Chapter 2). The hindered endo-cycles appear to be coupled with a broad down-regulation of DNA replication machinery, with very few upregulated loci in this class in severely impaired crwn mutants. A reduced demand for DNA replication could be associated with a lower requirement for the deposition and maintenance of epigenetic marks on newly replicated DNA strands.
The crwn mutations also caused broad mis-expression of various DNA repair pathways. Several different defects in crwn mutants could be related to alterations in DNA repair. For example, a down-regulation of DNA replication machinery may result in more frequent replication errors. Also, the mis-expression of SMC complexes, especially SMC6A, might disrupt homologous recombination-mediated DNA repair [102], and therefore enhance the activity of a compensating repair pathway, such as base excision repair. Further, the release of transposon silencing in crwn mutants could create novel genomic polymorphisms (e.g., DSB, copy number variations), which could induce DNA repair. In addition, the whole genome undergoes constitutive activation of defense responses, which could be coupled to a higher incidence of DNA damage [159]. These problems could concurrently slow down the progression of cell cycles and engage the cells in continuous DNA repair tasks to recover the fidelity of
153

genetic information. To dissect this mechanism, functional assays of DNA repair are needed. Scans to assess mutation rates in crwn lines should also be pursued by whole genome sequencing to detect the accumulation of DNA polymorphisms.
In summary, CRWN proteins appear to employ different mechanisms to regulate or alter gene expression. I proposed that the loss of nuclear structural CRWN proteins primarily alters nuclear organization, leading to a change in the nuclear landscape of various functional machineries (e.g., transcription, DNA repair), which in turn leads to downstream effects. Specifically, I propose that the loss of CRWN proteins, especially CRWN4, expands to affect CRWN-interacting proteins, including the chromosomal scaffolding system (e.g., SMC complexes), thus impairing basic nuclear processes including DNA replication, nucleosome assembly, and epigenetic modification. These defects consequentially activate DNA repair pathways to protect genomic integrity. In addition, rearrangements in nuclear organization might also alter the “transcriptional flavor” of the local environment of the chromosome, especially those regions that are co-localized with CRWN proteins on the nuclear periphery, thereby altering their transcription. The broadly activated stress response pathways might be altered in crwn mutants through this mechanism.
The most critical question of how the loss of CRWN proteins disrupts nuclear organization remains to be addressed. It is possible that CRWN proteins, especially CRWN4, work closely with SMC complexes, and the loss of CRWN directly downregulates the quality and quantity of these basic components of chromosomal
154

organization. CRWN proteins localize to the nuclear periphery, therefore SUN-domain proteins, the only clear plant homologs of animal nuclear envelope proteins, are another candidate for CRWN-interacting proteins. Different CRWN paralogs could also form working complexes. In the next chapter, I will start to address these questions using biochemical approaches.
Materials and Methods
Plant materials and growth conditions All T-DNA insertion alleles used in this study were obtained from the SALK collection in strain Columbia [121], and single mutant lines were originally obtained from the Arabidopsis Biological Resource Center (ABRC) at The Ohio State University. Plants were grown in long-day conditions (16 h of light / 8 h of dark) at 23˚C on soil (Metro-Mix 360, SunGro, Vancouver) in environmental growth chambers. Each genotype profiled includes three biological replicates from independent genetic lineages, and tissue from twenty individuals was pooled within each biological replicate. Mature rosette leaves were harvested from 4-week-old adult plants for each sample, and frozen in liquid nitrogen immediately for subsequent RNA purification/n.
155

RNA extraction and library construction Total RNA was extracted from each biological replicate sample using the traditional Trizol® method (http://tools.lifetechnologies.com/content/sfs/manuals/trizol_reagent.pdf) and further purified using a Qiagen RNeasy kit. The total RNA sample was sent to the Genomic Core Facility at the Weill Medical School (http://corefacilities.weill.cornell.edu/genomics.html) , and the standard Illumina protocol TruSeq RNA Sample Prep Guide (15008136 A) (http://support.illumina.com/sequencing/documentation.ilmn) was used for singlestranded mRNA library construction.
Sequencing, alignment and comparison of mRNA transcriptome 18 samples from 6 genotypes were barcoded and pooled in three lanes for sequencing using an Illumina Hi seq 2000 platform. Each sample recovered 20-30 million reads 51 base pair in length. The rRNA reads were removed from the data set after alignment using Bowtie2. A Tophat-Cufflink pipeline was employed to map the reads to the genome and to perform between-sample comparisons, and a q value for each locus was calculated using the default settings in Cuffdiff 2. [160]. The DEseq pipeline in the R package was also used to conduct a parallel analysis (data not shown) to confirm that the results were consistent with the analysis that the Tophat-Cufflink pipeline produced.
156

Other bioinformatics analyses Venn diagrams were generated by the online tool BioVenn (http://www.cmbi.ru.nl/cdd/biovenn/). The heatmaps were generated using HeatMapImage V6 on the public server Gene Pattern (http://genepattern.broadinstitute.org). The functional categorization was assigned to the mis-regulated loci using the gene ontology (GO) annotation tools on the TAIR website (http://www.arabidopsis.org). All other analysis was conducted using a standard spreadsheet program (Microsoft Excel).
157

REFERENCES
1. Egecioglu D, Brickner JH: Gene positioning and expression. Current opinion in cell biology 2011, 23(3):338-345.
2. Kind J, van Steensel B: Genome-nuclear lamina interactions and gene regulation. Current opinion in cell biology 2010, 22(3):320-325.
3. Dittmer TA, Stacey NJ, Sugimoto-Shirasu K, Richards EJ: LITTLE NUCLEI genes affecting nuclear morphology in Arabidopsis thaliana. The Plant cell 2007, 19(9):2793-2803.
4. Wang H, Dittmer TA, Richards EJ: Arabidopsis CROWDED NUCLEI (CRWN) proteins are required for nuclear size control and heterochromatin organization. BMC plant biology 2013, 13(1):200.
5. Sakamoto Y, Takagi S: LITTLE NUCLEI 1 and 4 regulate nuclear morphology in Arabidopsis thaliana. Plant & cell physiology 2013, 54(4):622-633.
6. Zhang X, Yazaki J, Sundaresan A, Cokus S, Chan SW, Chen H, Henderson IR, Shinn P, Pellegrini M, Jacobsen SE et al: Genome-wide high-resolution mapping and functional analysis of DNA methylation in arabidopsis. Cell 2006, 126(6):1189-1201.
7. Nakano T, Suzuki K, Fujimura T, Shinshi H: Genome-wide analysis of the ERF gene family in Arabidopsis and rice. Plant physiology 2006, 140(2):411-432.
158

8. Eulgem T, Rushton PJ, Robatzek S, Somssich IE: The WRKY superfamily of plant transcription factors. Trends in plant science 2000, 5(5):199-206.
9. Lee BH, Henderson DA, Zhu JK: The Arabidopsis cold-responsive transcriptome and its regulation by ICE1. The Plant cell 2005, 17(11):31553175.
10. Prasch CM, Sonnewald U: Simultaneous application of heat, drought, and virus to Arabidopsis plants reveals significant shifts in signaling networks. Plant physiology 2013, 162(4):1849-1866.
11. Walley JW, Coughlan S, Hudson ME, Covington MF, Kaspi R, Banu G, Harmer SL, Dehesh K: Mechanical stress induces biotic and abiotic stress responses via a novel cis-element. PLoS genetics 2007, 3(10):1800-1812.
12. Thilmony R, Underwood W, He SY: Genome-wide transcriptional analysis of the Arabidopsis thaliana interaction with the plant pathogen Pseudomonas syringae pv. tomato DC3000 and the human pathogen Escherichia coli O157:H7. The Plant journal : for cell and molecular biology 2006, 46(1):34-53.
13. Kohorn BD: WAKs; cell wall associated kinases. Current opinion in cell biology 2001, 13(5):529-533.
14. Schubert V: SMC proteins and their multiple functions in higher plants. Cytogenetic and genome research 2009, 124(3-4):202-214.
15. Watanabe K, Pacher M, Dukowic S, Schubert V, Puchta H, Schubert I: The STRUCTURAL MAINTENANCE OF CHROMOSOMES 5/6 complex promotes sister chromatid alignment and homologous recombination after
159

DNA damage in Arabidopsis thaliana. The Plant cell 2009, 21(9):26882699. 16. Roeder AH, Chickarmane V, Cunha A, Obara B, Manjunath BS, Meyerowitz EM: Variability in the control of cell division underlies sepal epidermal patterning in Arabidopsis thaliana. PLoS biology 2010, 8(5):e1000367. 17. Churchman ML, Brown ML, Kato N, Kirik V, Hulskamp M, Inze D, De Veylder L, Walker JD, Zheng Z, Oppenheimer DG et al: SIAMESE, a plantspecific cell cycle regulator, controls endoreplication onset in Arabidopsis thaliana. The Plant cell 2006, 18(11):3145-3157. 18. Bethke G, Grundman RE, Sreekanta S, Truman W, Katagiri F, Glazebrook J: Arabidopsis PECTIN METHYLESTERASEs Contribute to Immunity against Pseudomonas syringae. Plant physiology 2014, 164(2):1093-1107. 19. Gallego-Giraldo L, Escamilla-Trevino L, Jackson LA, Dixon RA: Salicylic acid mediates the reduced growth of lignin down-regulated plants. Proceedings of the National Academy of Sciences of the United States of America 2011, 108(51):20814-20819. 20. Ng G, Seabolt S, Zhang C, Salimian S, Watkins TA, Lu H: Genetic dissection of salicylic acid-mediated defense signaling networks in Arabidopsis. Genetics 2011, 189(3):851-859. 21. Hirano T: At the heart of the chromosome: SMC proteins in action. Nature reviews Molecular cell biology 2006, 7(5):311-322.
160

22. Torres-Rosell J, Machin F, Farmer S, Jarmuz A, Eydmann T, Dalgaard JZ, Aragon L: SMC5 and SMC6 genes are required for the segregation of repetitive chromosome regions. Nature cell biology 2005, 7(4):412-419.
23. Kegel A, Sjogren C: The Smc5/6 complex: more than repair? Cold Spring Harbor symposia on quantitative biology 2010, 75:179-187.
24. Yan S, Wang W, Marques J, Mohan R, Saleh A, Durrant WE, Song J, Dong X: Salicylic Acid activates DNA damage responses to potentiate plant immunity. Molecular cell 2013, 52(4):602-610.
25. Alonso JM, Stepanova AN, Leisse TJ, Kim CJ, Chen H, Shinn P, Stevenson DK, Zimmerman J, Barajas P, Cheuk R et al: Genome-wide insertional mutagenesis of Arabidopsis thaliana. Science 2003, 301(5633):653-657.
26. Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL: TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome biology 2013, 14(4):R36.
27. Woo HR, Dittmer TA, Richards EJ: Three SRA-domain methylcytosinebinding proteins cooperate to maintain global CpG methylation and epigenetic silencing in Arabidopsis. PLoS genetics 2008, 4(8):e1000156.
28. Stroud H, Greenberg MV, Feng S, Bernatavichute YV, Jacobsen SE: Comprehensive analysis of silencing mutants reveals complex regulation of the Arabidopsis methylome. Cell 2013, 152(1-2):352-364.
29. Lister R, O'Malley RC, Tonti-Filippini J, Gregory BD, Berry CC, Millar AH, Ecker JR: Highly integrated single-base resolution maps of the epigenome in Arabidopsis. Cell 2008, 133(3):523-536.
161

CHAPTER FOUR
THE NUCLEAR COILED-COIL PROTEINS CRWN1 AND CRWN4 PHYSICALLY INTERACT TO REGULATE NUCLEAR ORGANIZATION IN
ARABIDOPSIS THALIANA
Abstract
In this study I describe the biochemical characterization of the nuclear coiled-coil proteins CRWN1 and CRWN4, which are required for proper maintenance of nuclear organization in the flowering plant Arabidopsis thaliana. Polyclonal antisera against CRWN1 and CRWN4 were used as immunological probes to study the localization and interaction of these proteins. I demonstrate that both CRWN1 and CRWN4 are enriched in nuclear extracts and are resistant to salt extraction. Through a coimmunoprecipitation assay, I provide evidence of a physical interaction between CRWN1 and CRWN4 proteins, supporting the hypothesis that CRWN1 and CRWN4 form functional complexes in vivo. The abundance of CRWN4 is significantly reduced in crwn1 mutant backgrounds, while a complex feedback regulation on the mRNA level exist among different CRWN genes. My results suggest that a balance between CRWN1-like and CRWN4-like proteins is important to accomplish the proper regulation of nuclear organization.
162

Introduction
Nuclei are the cellular settings for essential functions, such as DNA replication and repair, epigenetic modification, and transcription. Dynamic nuclear organization plays an important role in regulating these molecular activities [45]. An important regulator of nuclear organization is the nuclear lamina, a lattice-like structure underlying the inner nuclear envelope in animal cells composed of the type V intermediate filament protein, lamin, and its interacting proteins (i.e., ‘lamin-associated proteins’) [45]. Mutations in the human genes encoding lamin A/C, LMNA, and lamin-associated proteins cause a complex set of clinical syndromes called laminopathies [58]. The nuclei of cells from laminopathy patients are irregular looking and fragile, and contain redistributed histone modification marks [45, 58].
Lamins have been reported to exist only in metazoans, although NUP1 in Trypanosomes [60] and NE81 in Dictyostelium have been proposed to be candidate lamin analogs [161]. There are no plant homologs resembling lamin proteins [78], and it is unclear how plants organize their nuclei. In the 1990s, Masuda and colleagues discovered the coiled-coil domain protein NMCP1 (Nuclear Matrix Constituent Protein 1), using a monoclonal antibody screen for antigens in nucleoskeleton preparations from carrot cells [75]. Since then, NMCP1 and its paralogs have been considered to be prime candidates for the central components of a plant nuclear lamina [78, 79, 120, 162] .
163

NMCP proteins harbor nuclear localization signals and a variety of cytological observations, including immunolocalization and fluorescent protein tagging experiments, demonstrate the NMCPs are nuclear proteins [78, 79, 81, 115]. The Arabidopsis NMCP proteins, called CRWN [140] (and previously known as LINC [78]), also contain long coiled-coil domains in their central regions, and are localized primarily at the nuclear periphery. Our previous studies showed that CRWN1 and CRWN2 proteins concentrate at the nuclear periphery, but over-expression of these two proteins, particularly CRWN2, led to their distribution throughout the nucleoplasm [78]. Localization of this class of protein at the nuclear periphery is supported by electron microscope immunostaining experiments in onion (Allium cepa) that show AcNMCP1 proteins were concentrated around the nuclear rim, with some speckled signals in the nucleoplasm [77]. A recent study also reported that LINC4 (CRWN4) proteins are located at the nuclear periphery, and further showed that LINC1 proteins co-localize with mitotic chromosomes during the cell cycle [81]. Similarly, AgNMCP1 in celery (Apium graveolens) co-localizes with the mitotic spindle and segregating chromosomes, while AgNMCP2 is dispersed throughout the mitotic cytoplasm. These two proteins accumulate at the nuclear periphery in anaphase at the end of the cell cycle [76].
NMCP proteins are characterized by insolubility during biochemical treatment of isolated nuclei. Both the carrot (Daucus carota) prototypic protein DcNMCP1 and its onion (Allium cepa) ortholog AcNMCP1 fractionate with insoluble nuclear extracts
164

resistant to salt and non-ionic detergent extraction [75, 79]. This behavior is not simply associated with the hydrophobicity typical of nuclear membrane proteins, because NMCP proteins do not possess any conventional transmembrane domains. An immunoprecipitation-based proteomic study of GFP-tagged nucleoporins in Arabidopsis failed to detect any NMCP family proteins interacting with the several tested nucleoporins [134]. It is more likely that NMCP proteins connect to the inner nuclear membrane via other interacting partners, and the biochemical fractionation characteristics could be due to the long coiled-coil domain in the center of NMCP proteins. The coiled-coil structure is a widespread motif involved in oligomerization and interaction with DNA or other proteins [163]. One classical example of a single coiled-coil is the leucine zipper structural motif in various transcription factors, which consist of parallel, left-handed homodimers for DNA binding [164]. It is common for proteins containing long coiled-coil domains to polymerize with each other to form insoluble networks and serve structural functions in cells. Examples include SMC (Structural Maintenance of Chromosomes) proteins [100], tropomyosins [53], and lamin proteins [48, 51].
In this study, I investigate biochemical aspects of CRWN proteins. My results show that CRWN1 and CRWN4 proteins fractionated with the insoluble nuclear residue, which was resistant to high salt and non-ionic detergent treatment. CRWN4 mRNA was up-regulated in crwn1 mutants; however, CRWN4 protein abundance was greatly reduced in the absence of CRWN1. Moreover, CRWN1 and CRWN4 proteins interacted with each other in vivo. The complex regulation among CRWN genes on
165

both the transcriptional and protein levels supports a model wherein the balance between CRWN1-like and CRWN4 functions are important for the regulation of nuclear organization.
Results
The structure and natural variation of CRWN family proteins
Figure 3.1 depicts the structure of the Arabidopsis CRWN proteins. Each protein is approximately 1100 amino acids in length with a long coiled-coil domain starting from the second or third exon and extending to the middle of the largest sixth exon (black bar). In addition, CRWN1, 2 and 3 contain a well-conserved C-terminus (purple bar), which is absent in CRWN4. Polymorphic amino acid sites of CRWN proteins within species Arabidopsis thaliana were identified by comparing predicted protein sequences of various natural accessions from the 1001 Genome Project [165, 166]; Figure 4.1 shows the distribution of sites exhibiting low and high levels of polymorphisms (Figure 4.1). Close to 10% of the amino acid residues of each CRWN protein were identified as polymorphic sites (Table 4.1), and they are randomly distributed across the protein, with no obvious clustering in any particular domain. Polymorphism levels are similar among the four CRWN proteins (Table 4.1), and polymorphic patterns among the four CRWN proteins were not correlated.
166

To broaden the scope of this investigation of natural variation, I extended the comparison beyond Arabidopsis thaliana to its close relative Arabidopsis lyrata. There are four CRWN paralogs in A. lyrata, corresponding to CRWN1, CRWN2, CRWN3 and CRWN4 in A. thaliana. Specifically, I compared the CRWN protein sequences of the A. lyrata genome (version 1.0, one genotype) with the consensus sequence of the corresponding CRWN proteins in A. thaliana. A few insertions and deletions were detected (pink gaps) between the A. thaliana and A. lyrata proteins, and dozens of sites (light green) in A. lyrata differed from those seen in A. thaliana. The majority of these variations were A. lyrata-specific, suggesting independent evolutionary forces are exerted on these two species, or that the variation occurred randomly after divergence of a common ancestor. Among these A. lyrata-A.thaliana polymorphisms, the abundance of these sites in the AtCRWN4 protein was the lowest among the four CRWN paralogs. This pattern contradicts the typical situation in which the divergence patterns for different gene families between A. lyrata and A. thaliana mirror within-A thaliana polymorphism levels [167]. It is possible that a selective pressure in A. lyrata constrained the accumulation of polymorphisms in the CRWN4 protein among natural strains, while the presence of three copies of CRWN1-like proteins [140] releases that pressure through redundancy. Nonetheless, due to the fact that only one A. lyrata genome was available for comparison, it is difficult to assess the significance of these polymorphism patterns.
167

Figure 1
CRWN1
A. thaliana A. lyrata

CRWN2
A. thaliana
A. lyrata

CRWN3
A. thaliana
A. lyrata

CRWN4
A. thaliana
A. lyrata

1000 a.a.

Neighboring exons

Mild polymorphism

Antigen peptide Coiled-coil domain

Severe polymorphism

Gap

Conserved C terminal domain

Figure 4.1 CRWN protein domains and sequence polymorphism

168

Figure 4.1 The domain composition of CRWN proteins is illustrated. The amino acid sequences of CRWN proteins are displayed in gray boxes, and the neighboring exons are indicated by alternated light and dark gray colors. The conserved C-termini, predicted long coiled-coil domain [78], and N-terminal sequences used for antigen design, are shown by separate bars in purple, black, and blue, respectively, for each CRWN protein. Polymorphic sites within Arabidopsis thaliana are displayed in yellow (mild polymorphisms with similar biochemical properties) or red (severe polymorphisms with distinct biochemical properties) determined using a Gonnet PAM 250 matrix with a cut off value of 0.5 [168]. The predicted CRWN protein sequences from the Arabidopsis lyrata 1.0 genome were displayed in a similar manner, with polymorphic sites divergent from the A. thaliana consensus sequence colored in light green. Pink gaps indicate the insertions and deletions in the alignments between the two species.
169

polymorphism CRWN1 CRWN2 CRWN3 CRWN4

Arabidopsis0thaliana010010genome severe mild total protein0size
8 96 104 1132 14 101 115 1128 10 94 104 1085 13 73 86 1010

% 9.19% 10.20% 9.59% 8.51%

polymorphism CRWN1 CRWN2 CRWN3 CRWN4

insertion 3 5 2 1

A0lyrata0v.s.0A0thaliana deletion changes
0 81 14 73 2 93 6 44

total 84 92 97 51

protein0size 1135 1119 1085 1005

% 7.40% 8.22% 8.94% 5.07%

000000000000000000000000000000000000overlap0with0thaliana0polymorphic0sites

polymorphism lyrata0only

severe

mild

CRWN1

77

25

CRWN2

81

1 10

CRWN3

77

5 15

CRWN4

44

52

Table 4.1 Summary of CRWN protein sequence polymorphisms Table 4.1 The statistics for Figure 4.1. Numbers of polymorphic sites in each CRWN protein within A. thaliana, and between A. thaliana and A. lyrata are counted and compared.

170

Development of antibody probes recognizing CRWN1 and CRWN4
My previous genetic analysis demonstrated that CRWN1 and CRWN4 are the two major proteins in the CRWN family, possessing phylogenetically distinct amino acid sequences and exhibiting different nuclear functions (see Chapter 2). However, the loss of CRWN1-like and CRWN4 functions mis-regulates common loci at the transcription level (see Chapter 3), suggesting that CRWN1 and CRWN4 might work together to fulfill their apparently divergent functions. Thus, I focused on these proteins as representatives of this protein family. To develop immunological probes to study CRWN function, peptide antigens were designed to raise antisera specific for either CRWN1 or CRWN4. Two ~20 amino acid sequences specific for the targeted proteins were chosen from the N-termini of the protein sequences (blue bar in Figure 4.1) for three reasons. First, terminal sequences are more likely to be exposed on the surface of well-folded proteins in vivo [169]; therefore, these regions might be more readily available for antibody binding. Second, the C-termini of CRWN family proteins are relatively conserved compared to the N-termini [78], suggesting that the C-terminus might have more important functional constraints, for example, by participating in protein-protein interactions. Consequently, I focused on a less conserved region for antisera recognition and binding in hopes of avoiding disruption of interactions among CRWN proteins and any relevant working partners. Third, the N-termini of CRWN proteins are more variable, thus it is more likely to find a region specific to either CRWN1 or CRWN4 proteins.
171

The peptide antigens were synthesized, injected into rabbits, and corresponding antisera were affinity purified (see Materials and Methods). I screened antisera raised against CRWN1 or CRWN4 peptides using western blots of nuclear extracts from seedlings of wild type and corresponding crwn mutants, and I identified specific antisera for the two CRWN proteins (Figure 4.2A). The antibody against CRWN1 detected a series of four distinct bands between 100 kDa and 150 kDa, while no signal was detected in crwn1 mutant samples, indicating that all the bands detected by this antibody are specific to CRWN1. Among the four bands, the brightest signal shifted from the larger bands toward the smaller 110 kDa band as extraction time was extended (Figure 4.2B). The 110 kDa band was dominant and reproducible compared to other bands across all our experiments (data not shown), and the 110 kDa size is consistent with the molecular weight of CRWN1 protein. Thus, I used the 110 kDa as a signature of the presence of CRWN1 proteins in our studies. The antibody against CRWN4 protein detected three equally bright bands ranging in size from 110 kDa to 150 kDa in wild type samples. All of these bands were missing from crwn4 mutant samples, indicating that these three bands specifically recognize CRWN4 proteins. It is likely that the distinct bands detected by the antisera represent different isoforms and/or modifications of these two proteins in vivo. Additional characterization, for example, using mass spectrometry, will be necessary to determine the full spectrum of the protein isoforms and modifications. With these validated immunological probes in hand, I turned my attention to the following biochemical investigations.
172

150 KDa

wild type

cwn1

wild type cwn4

150 KDa

100 KDa

100 KDa

Figure 4.2 Specificity test for antisera against CRWN1 and CRWN4 proteins Figure 4.2 Detection of CRWN1- and CRWN4- specific bands using antisera raised for this study. Proteins from nuclear extracts of Col wild type, crwn1, and crwn4 seedlings were resolved by SDS-PAGE protein electrophoresis, and proteins were detected using western immunoblot techniques.

173

CRWN1 and CRWN4 fractionate with the salt-resistant insoluble nucleoskeleton
I first investigated the sub-nuclear fractionation characteristics of CRWN1 and CRWN4. Many investigators have used high salt and detergent to perform nuclear extractions to generate an insoluble ‘nuclear matrix’ fraction resistant to these harsh biochemical conditions [170]. This residual framework is thought to represent the network of fibers that exist throughout the nucleus. To test the hypothesis that CRWN proteins are structural components in plant nuclei, I followed the above criteria and applied various salt and detergent treatments to Arabidopsis nuclear extracts to solubilize proteins, and determined whether CRWN proteins remained in the insoluble pellet after separation by centrifugation.
Crude nuclear preparations from seedling tissue were extracted using various concentrations of NaCl. The results displayed in Figure 4.3A show a control chromatin protein, histone H2B, was solubilized to a significant extent by NaCl treatment at a concentration of 0.5M and higher. In contrast, only faint CRWN1 bands were recovered in the salt-treated supernatant samples, while strong signals were detected in all pellet samples. This partitioning indicates that most of the CRWN1 protein remained insoluble after extraction, resulting in a pattern resembling that seen in the control sample to which no salt was applied. The salt treatment only solubilized a very small portion of CRWN1 proteins regardless of the NaCl concentration in the buffer. Similarly, no significant CRWN4 signal was detected among the solubilized
174

samples; the three CRWN4 bands were recovered in the pellet samples with equal signal strength, indicating that CRWN4 is also resistant to salt extraction.
In addition to salt treatment, various concentrations of detergents, including Triton X100, CHAPS, and SDS, were applied to assess the solubility of CRWN proteins. Triton X-100 by itself did not affect the solubility of CRWN1 or CRWN4 (Figure 4.3A, Supplementary Figure 4.1A). Different concentrations of CHAPS also did not significantly affect CRWN1 or CRWN4 solubility (Figure 4.3B), although in some cases, a low concentration of CHAPS slightly solubilized CRWN4 (data not shown). In contrast, SDS efficiently solubilized both CRWN1 and CRWN4 (Supplementary Figure 4.1), even at a low (0.2%) concentration. Nonetheless, a portion of the proteins remained in the nuclear pellet after extraction with SDS (Supplementary Figure 4.1). These data demonstrate that CRWN proteins were resistant to high salt and mild detergent extraction, suggesting that CRWN1 and CRWN4 are components of the nuclear matrix.
175

A Salt and Triton treatments on CRWN proteins

NaCl

Triton

NaCl

Triton

No 0.5M 1M 1.5M 2M 1% 5% No 0.5M 1M 1.5M 2M 1% 5%

H2B

CRWN1

CRWN4

Pellet

Supernatant

B CHAPS and sonication treatments on CRWN proteins

CHAPS SC SC SC +

CHAPS SC SC SC +

No 1% 2% 4% Short Long 4% No 1% 2% 4% Short Long 4%

CRWN1

CRWN4 Pellet
SC + 4%: Short sonication + 4% CHAPS

Supernatant

Figure 4.3 CRWN1 and CRWN4 proteins are insoluble under mild detergent and high salt extraction conditions

176

Figure 4.3 Detection of CRWN1 and CRWN4 proteins in salt- or detergent- treated nuclear extracts of wild type samples. A concentration series of NaCl (0M, 0.5M, 1M, 1.5M, 2M), Triton X-100 (1%, 2%) and CHAPS (1%, 2%, 4%) were applied. The left panel represents the insoluble pellets, while the right panel represents the solubilized supernatants. The signal against CRWN1 proteins (the major band is a little bit above 100 kDa) was displayed in the upper panel, and CRWN4 proteins (triplet bands between 100 -150 kDa) in the bottom panel. Histone H2B (17 kDa) was used as loading control. SC indicates sonication treatment.
177

A Sonication and SDS solubilize CRWN1 protein

SC SC 0.2% 1% 2% No Short Long SDS Triton Triton 150 KDa

SC SC 0.2% 1% 2% No Short Long SDS Triton Triton

100 KDa

Pellet

Supernatant

B Salt precipitates CRWN1 protein

SC SC

SC SC SDS SDS

SC SC SDS SDS

No SC Salt SDS Salt Salt SDS No SC Salt SDS Salt Salt SDS

100 KDa

Pellet

Supernatant

C Condition test for CRWN1 - CRWN4 immunoprecipitation

No
CRWN1 100 KDa

SC in SC in SC

SC 4% 0.05% +

No

CHAPS SDS Bezonase

SC in SC in SC SC 4% 0.05% +
CHAPS SDS Bezonase

CRWN4 100 KDa
H2B 17 KDa

Pellet

Supernatant

1A, 1B and 1C: SC: Sonication 1B: Salt: 2M NaCl

Supplementary Figure 4.1 The optimization of conditions for CRWN1-CRWN4 co-immunoprecipitation

178

Supplementary Figure 4.1 Detection of CRWN1 and CRWN4 proteins in sonication, salt-, and detergent-treated nuclear extracts of wild type samples. In Panel A, B and C, Short (5 sec x 10 times) and long (5 sec x 20 times) sonications, NaCl (2M), Triton X-100 (1%, 2%), CHAPS (4%), SDS (0.05%, 0.2%) and Benzonase ® (Sigma-Aldrich) were applied individually or in combination. The left panel represented the insoluble pellet, while the right panel represented the solubilized supernatants. The signal against CRWN1 proteins (the major band is a little bit above 100 kDa) and CRWN4 proteins (triplet bands between 100 -150 kDa) were displayed. In Panel C, histone H2B (17 kDa) was used as loading control.
179

The abundance of CRWN4 protein is dependent on CRWN1 protein expression
Phylogenetic analyses place CRWN family proteins into two major clades [140]: the CRWN1-like clade (CRWN1, 2 and 3 in Arabidopsis) and the CRWN4 clade. The genetic analysis presented in Chapter 2 further supports this classification by showing non-redundant functions for CRWN1-like and CRWN4 proteins in controlling wholeplant phenotypes and nuclear morphology. However, my mRNA-seq analysis (see Chapter 3) demonstrate that despite the structural and functional divergence between CRWN1-like and CRWN4 proteins, these proteins share common genomic targets in transcriptional regulation. Based on these observations, I propose that CRWN1 and CRWN4 work coordinately to carry out their functions. I further hypothesize that the cooperation of CRWN1 and CRWN4 occurs because the proteins physically interact in a functional complex within the nucleus.
To test this hypothesis, I first used protein immunoblots to examine whether the abundance of CRWN1 or CRWN4 protein was altered in crwn single and double mutants. Figure 4.4 displays the distinct bands of the CRWN1 protein pool in wild type, crwn2, crwn3, and crwn4 single mutants, as well as crwn2 crwn3, crwn2 crwn4, and crwn3 crwn4 double mutants, in which the crwn1 mutation was not present. Interestingly, the triplet bands representing CRWN4 proteins were not only absent from the crwn4 single mutant, but were significantly reduced in the crwn1 single mutant. Among the double mutants, crwn2 crwn3 was the only genotype exhibiting a strong signal for the CRWN4 triplet bands. These signals were severely diminished in
180

crwn1 crwn2 and crwn1 crwn3 samples, and not detectable in the crwn1 crwn4, crwn2 crwn4, and crwn3 crwn4 mutant backgrounds. Any plant that carried a crwn1 mutation lost the CRWN1 protein and suffered a dramatic reduction in the abundance of the CRWN4 protein. However, the loss of CRWN4 did not have a significant effect on the abundance of CRWN1. These results suggest that the presence of CRWN1 is required for the proper production or stability of CRWN4.
These observations prompted me to look in my mRNA-seq dataset at the abundance of CRWN mRNA in different crwn mutants, especially CRWN4 transcripts in crwn1 single mutants. Figure 4.5 summarizes the average RPKM (Reads Per Kilobase of exon per Million fragments mapped) values of the mRNA for the four CRWN genes and a control gene, cyclophilin. The abundance of CRWN4 transcripts was elevated approximately 30% in crwn1 mutants, indicating that loss of CRWN1 leads to a modest boost in CRWN4 expression at the mRNA level. Therefore, down-regulation at the transcript level cannot explain the significant reduction in CRWN4 protein expression I observed in Figure 4.4. Rather, this observation suggests that a compensatory feedback up-regulation of CRWN4 mRNA occurs in response to either the loss of CRWN1 or the reduction in CRWN4 protein abundance.
181

Figure 4.4 CRWN4 abundance is reduced in crwn1 backgrounds Regulation of CRWN expression at the mRNA level Figure 4.4 Detection of CRWN1 and CRWN4 proteins in crwn single and double mutants using protein immunoblots. The signals against CRWN1 protein (the major band is slightly above 100 kDa) were displayed in the upper panel, and the signals against CRWN4 proteins (triplet bands between 100 – 150 kDa) in the bottom panel. Histone H2B (17kDa) was used as loading control. In each of the three columns, all lanes shown are from the same experiment. Each row within a column is from the same film exposure, but some lanes were rearranged for clarity.
182

Figure 5 CRWN mRNA expression in crwn mutants

RPKM value (normalized to cyclophilin)

25 20 15 10 5 0
wild type crwn1

crwn4

crwn1 crwn4

CRWN1

CRWN2

25 20 15 10 5 0 wild type

a bc crwn1 crwn2

CRWN3

CRWN4

Figure 4.5 The expression of CRWN mRNAs in various crwn mutants Figure 4.5 The left panel shows the normalized average RPKM value of each CRWN mRNA in wild type, crwn1, crwn4, and crwn1 crwn4 mutants. The right panel showed the normalized RPKM value of each replicate in the crwn1 crwn2 mutant sample in comparison to wild type. The RPKM value of the cyclophilin locus was used as internal control for normalization.

183

Analysis of my mRNA-seq datasets for crwn mutants also revealed a complicated regulatory relationship among CRWN transcripts. First, I observed an up-regulation of CRWN2 and CRWN3 mRNA in crwn1 mutant backgrounds, suggesting that a transcriptional compensation occurs among CRWN1-like genes. Second, CRWN1 mRNA was down-regulated in crwn4 mutants, although a significant change in CRWN1 protein abundance was not observed by protein immunoblots. Nonetheless, the down-regulation of CRWN1 transcripts in crwn4 mutants indicates that the presence of the CRWN4 protein is required for normal production of transcripts encoding CRWN1-like proteins.. Third, in crwn1 crwn4 double mutants, CRWN2 and CRWN3 transcripts were expressed at a higher level, especially CRWN3, which nearly doubles in abundance. This elevation of CRWN2 and CRWN3 transcripts might be responsible for the apparent functional suppression between crwn1 and crwn4 mutations. Taken together, these observations highlight the complex interactions and regulation among different CRWN genes on the mRNA level, and aides the interpretation of the phenotypes of plants carrying different crwn mutations. In addition, these interactions further support the hypothesis that the CRWN family of coiled-coil proteins function together on the biochemical level.
CRWN1 and CRWN4 proteins function in the same complexes
To test this hypothesis, I used immunoprecipitation to examine whether CRWN1 and CRWN4 proteins physically interact with each other in vivo. There were several technical challenges I needed to overcome for this experiment, and I began by
184

adapting an established protocol for Arabidopsis protein co-immunoprecipitation [171]. First, to concentrate CRWN proteins, nuclear extracts from large amounts of seedlings tissues were used. Public microarray database compilations (eFP browser; http://bar.utoronto.ca/efp/cgi-bin/efpWeb.cgi) showed that the CRWN mRNA accumulation is highest among proliferating tissues (e.g., root tips and shoot meristems). Further, our lab’s unpublished translational fusion reporter data shows that CRWN1 expression at the protein level is concentrated in meristematic regions and that the expression drops dramatically in mature tissues.
A second obstacle was the insolubility of CRWN proteins (Figure 4.3). Efficient solubilization with strong ionic detergents, such as SDS, might disrupt native proteinprotein interactions. To maximize the recovery of potential CRWN complexes from the nuclear extracts, lysis conditions needed to be optimized to achieve a balance between solubilizing CRWN complexes and preserving their native interactions. As shown in Supplementary Figure 4.1A-C, different conditions were applied to wild type nuclear extracts, and the solubility of CRWN proteins were assessed by western blotting. In Panel 1A, a short sonication in a buffer containing 0.2% SDS solubilized CRWN1 proteins well, with weak bands left in pellet samples and strong signals present in supernatant samples. Longer sonication treatment resulted in a loss of CRWN1 signals in both pellet and supernatant samples, indicating protein degradation. Panel 1B confirmed that sonication solubilized CRWN1 proteins but also led to degradation; while SDS treatment by itself or in combination with sonication efficiently solubilized CRWN1 proteins. However, high salt (2M NaCl) treatment, by
185

itself or in combination with either SDS or sonication, did not solubilize CRWN1 proteins. This result indicated that high salt precipitates rather than solubilizes CRWN1 proteins, even in the presence of SDS detergent and sonication. Based on this information, I tested the conditions for CRWN1-CRWN4 co-immunoprecipitation. As shown in Supplementary Figure 4.1 Panel 1C, I piloted a sonication approach, in combination with detergent treatment (e.g., CHAPS (4%), low concentration SDS (0.05%)) or application of the chromatin depletion enzyme, Bezonase [171]. Ultimately, I determined that the 0.05% SDS + sonication treatment solubilized both CRWN1 and CRWN4 best, and these conditions were used in the following investigations.
A third concern was that the antisera I raised might not have enough specificity and binding affinity to recognize and pull out the CRWN complexes from the protein extracts in the relatively harsh immunoprecipitation buffer conditions. To maximize the yield, protein extracts were incubated with antisera thoroughly, before the antibody affinity beads were added for the recovery. In addition, I used antibody affinity beads with higher specificity (i.e., Dynabeads versus Agrobeads, see Materials and Methods) to reduce the loss of recovered complexes during the washing steps.
Using these modifications, I conducted reciprocal co-immunoprecipitation experiments in wild type, crwn1 and crwn4 mutant samples. Figure 4.6A shows the detection of CRWN1 protein. In wild type samples, as expected, CRWN1 bands were present in the anti-CRWN1 precipitate pellets but absent from the no-antiserum
186

control pellets. A strong signal of CRWN1 protein was detected in the anti-CRWN4 pellets, indicating that CRWN1 protein could be precipitated by anti-CRWN4 antisera. The same sets of experiments were performed in the crwn1 and crwn4 mutant backgrounds. As expected, no CRWN1 signals were present in the crwn1 mutant sample pellets. In the crwn4 background, CRWN1 protein was still present in the anti-CRWN1 precipitates, but absent from the anti-CRWN4 precipitates, suggesting that CRWN1-CRWN4 complexes no longer exist due to the loss of CRWN4 protein. Panel B of Figure 4.6 shows the detection of the CRWN4 protein on the same membrane. In the wild type samples, CRWN4 bands were detected in the antiCRWN4 precipitates but absent from the no-antiserum control. At the same time, a faint signal of the CRWN4 triplet bands were present in the anti-CRWN1 pellets, suggesting that CRWN4 protein could be precipitated by anti-CRWN1 antisera, although at a lower efficiency. This weak binding from the reciprocal coimmunoprecipitation has been replicated several times (data not shown). In crwn4 mutants, no CRWN4 signal was detected as expected. In the crwn1 background, CRWN4 protein was absent from not only the anti-CRWN1 pellets but also from the anti-CRWN4 precipitates, indicating that the abundance of CRWN4 protein in the crwn1 background is too low for it to be precipitated or detected. These results demonstrate that CRWN1 and CRWN4 proteins are components of the same complex(es) in vivo.
187

Figure 4.6 CRWN1 and CRWN4 proteins interact with each other in vivo Figure 4.6 Detection of CRWN1 protein (upper panel) and CRWN4 protein (bottom panel) in three parallel immunoprecipitation experiments (no antisera, anti-CRWN1 antisera, and anti-CRWN4 antisera) are displayed for wild type, crwn1, and crwn4 mutants. The 110kDa band represents CRWN1 proteins, and the 100kDa-150kDa triplet bands correspond to CRWN4 proteins. All lanes shown are from the same experiment and the same film exposure; but some lanes were rearranged for clarity.
188

Discussion
In this chapter I undertook a biochemical characterization of the putative nuclear architectural proteins, CRWN1 and CRWN4. First, I developed native antisera against CRWN1 and CRWN4 proteins. I fractionated Arabidopsis cells and used western blots to reveal that CRWN1 and CRWN4 proteins belong to the insoluble nuclear fraction resistant to different concentration of salt and detergent treatment. A similar fractionation method to isolate plant nuclear matrices led to the initial discovery of NMCP1 protein from carrot suspension cells [75]. Also, Ciska et al. showed that onion NMCP1 was only observed in the insoluble fraction after a stepwise fractionation of nucleoskeleton preparations in onion tissues, where the extracted nuclei were sequentially treated with detergent (0.5% TritonX-100), a chromatin removal reagent (1M (NH4)2SO4), and high salt (4M NaCl) [77]. The presence of onion NMCP1 in the residual nucleoskeleton was also demonstrated by the immunogold staining using electron microscopy [77].
Nonetheless, these findings may not be enough to demonstrate the existence of a nuclear matrix system in vivo. The primary concern is that the extraction procedures themselves can lead to protein precipitation and form insoluble nuclear residues{Jackson, 1992 #167;Martelli, 1994 #168}. Non-ionic detergents such as Triton X-100 solubilize lipids and may deplete the proteins attached to the outer membrane of the nuclei. However, they do not break nuclear membranes, and Triton treatment had little effect on releasing the nuclear proteins into the supernatant, such
189

as histone H2B as well as CRWN1 (Figure 4.3A). CHAPS, another detergent frequently used for membrane protein extraction, had little effect in solubilizing CRWN proteins; rather, CHAPS helped preserve CRWN1 and CRWN4 in the insoluble pellet from being degraded during sonication (Figure 4.3B). Finally, high salt treatment itself, which is often used to break the nuclei, had an effect in precipitating the CRWN1 protein (Supplementary Figure 4.1B). This increased insolubility could be due to the biochemical properties of NMCP (CRWN) proteins in reaction to particular extraction conditions. In that case, the apparent rigidity of the NMCP (CRWN)-associated residual framework could be more of an artifact than a true reflection of the nature of the nucleoskeleton in vivo. With these caveats in mind, I interpret my extraction experiments results as an indication that Arabidopsis CRWN family proteins are located in the nucleus, and further that their unconventional biochemical properties argue for a potential structural function in nuclei.
My subsequent immunoblot experiments revealed that CRWN4 protein is significantly reduced in crwn1 mutant backgrounds, indicating that the phenotypes observed in crwn1 mutants reflect the combined effect of a loss of CRWN1 and a partial loss of CRWN4. Therefore, the observation that crwn1 induced mis-expressed loci belonging to the pool of genes mis-regulated in the crwn4 mutant could be due to reduced CRWN4 protein in the crwn1 mutant, rather than the loss of CRWN1 itself. In this interpretation, CRWN1-like and CRWN4 proteins target similar genomic regions primarily via CRWN4 function. However, several points of evidence support our original hypothesis that CRWN1-like proteins possess independent functions
190

divergent from the CRWN4 protein, and that CRWN proteins affect transcriptional regulation through a complex set of interactions. First, guard cell nuclear size is reduced in crwn1 mutants but not in crwn4 mutants, and CRWN1 plays the major role in controlling nuclear size in leaves [140] . Second, the aggregation phenotype of heterochromatin in the crwn nuclei and dwarfism at the whole plant level is exclusively associated with the loss of CRWN1-like function [140] . Third, crwn1 crwn2 double mutants show more severe whole plant and nuclear morphology phenotypes, as well as more profound mis-expression genome wide, in comparison to crwn1 mutants (Chapter 3) [78, 140]. This difference is attributable to the addition of the crwn2 mutation, as our protein immunoblot data did not detect any further reduction of CRWN4 abundance in crwn1 crwn2 mutants. Therefore, it is likely that CRWN1 and CRWN4 functions in the same complex, collectively regulating the transcription of common loci in Arabidopsis. However, other complexes might also exist which contains either CRWN1 or CRWN4 and are responsible for crwn1- or crwn4-specific morphological changes.
I used co-immunoprecipitation experiments to demonstrate that CRWN1 and CRWN4 physically interact in vivo. Although this approach cannot distinguish between direct and an indirect interactions, the structure of CRWN proteins suggests a simple explanation for a direct interaction via the long coiled-coil domain. This type of oligomerization has been shown in many cases for other proteins containing long coiled-coil domains, such as SMC (Structural Maintenance of Chromosome) proteins, tropomyosins, and lamins. For example, each SMC protein contains a long coiled-coil
191

motif in the middle for self-folding in an antiparallel manner to form the arm of the V shape dimer [52]. Tropomyosins are a large family of integral components of actin filaments. It consists of rod-shaped coiled-coil hetero- or homo-dimers that lie along the α-helical groove of actin filaments to regulate their function [53, 172]. Lamin proteins are the basic elements of the animal nuclear lamina. In vitro experiments showed that lamin A proteins form dimers and polymerize through their long coiledcoil domain. In the case of CRWN proteins, the long coiled-coil region of CRWN1 and CRWN4 monomers could wrap around each other in parallel or anti-parallel fashion, and these heterodimers could form a basic unit to execute CRWN function. Alternatively, homo-dimer or polymerized CRWN1 and CRWN4 proteins could interact with each other to form higher-order functional structures. As noted above, it is also possible that other molecules are involved in mediating an indirect interaction between CRWN1 and CRWN4. To address this question thoroughly, a domain analysis of truncated CRWN proteins will be needed to determine which part of the protein is necessary for the interaction. In addition, an immuno-affinity purification coupled with mass spectrometry could be used to examine other molecular components of CRWN1-CRWN4 complexes.
The interaction between CRWN1 and CRWN4 proteins prompted me to examine the mRNA expression level of these CRWN genes in different crwn mutants. Collectively, the mRNA expression level of CRWN4 gene is up-regulated in crwn1 backgrounds, although the amount of CRWN4 protein is significantly reduced in crwn1 mutants. Another demonstration of the interdependence of CRWN1 and CRWN4 functions is
192

my observation that CRWN1 mRNA was reduced to half the normal level in crwn4 mutant backgrounds. Based on these and other interactions among CRWN genes, I propose a model (Figure 4.7) to account for the apparent feedback mechanisms that maintain a balance of CRWN1-like and CRWN4-like functions. The model postulates an equilibrium in complementary CRWN functions via regulation of the quantity of CRWN1-like and CRWN4 mRNA and proteins. In wild type, CRWN1 and CRWN4 play specialized roles; yet both are needed to fulfill the function of CRWN complexes in regulating nuclear morphology and transcription. Our genetic analysis indicates that CRWN2 and CRWN3 possess both CRWN1 and CRWN4 functions, with a bias toward CRWN1-like function. In crwn1 mutants, where the CRWN1 function is absent, this situation is compensated by a mild up-regulation of CRWN2 and CRWN3 mRNA, as well as a reduced amount of CRWN4 protein. Therefore, a balance of CRWN1-like and CRWN4 function is be restored, resulting in relatively few misexpressed loci in crwn1 mutants. In crwn4 mutants, CRWN4 function is lost, with a compensatory reduction in mRNA expression of the CRWN1 gene. However, this compensation only partially alleviates the excess of CRWN1-like proteins in the nuclei. A similar scenario happens in crwn1 crwn2 double mutants, where a more complete loss of CRWN1-like function push the ratio between CRWN1-like and CRWN4 functions toward an extreme. A reduction of CRWN4 protein also exist in crwn1 crwn2 mutants, however, there is no obvious up-regulation of the CRWN3 gene, indicating that this extreme condition impairs the compensating mechanism. The low quantity and unbalanced pool of CRWN1-like and CRWN4 proteins resulted in dramatic changes of nuclear function, reflected by the mRNA-seq analysis discussed
193

in Chapter 3. In crwn1 crwn4 double mutants, however, CRWN2 and CRWN3 genes were significantly over-expressed in the absence of both CRWN1 and CRWN4 proteins, suggesting an active compensation to restore the quantity and balance of CRWN1-like and CRWN4 functions. This model is consistent with the morphological and transcription suppression observed in crwn1 crwn4 double mutants. Further work to test the predictions of this working model will require a more complete understanding of CRWN2 and CRWN3 protein regulation and activity.
194

Figure 4.7 A balancing model of CRWN1-like and CRWN4-like functions 195

Figure 4.7 A series of cartoons describe the genetic interaction (top) as well as the balance of CRWN1-like and CRWN4 functions (bottom) in wild type and representative crwn mutants. The rectangle boxes in the models (top) illustrated divergent and redundant functions of CRWN genes. The width of the grey area represents the minimum coverage of CRWN function that is necessary for the survival of a plant. The size of each colored rectangle reflects the portion of the function that each CRWN gene contributes to nuclear organization and transcriptional regulation. Filled rectangles reflect transcription levels of wild type copy CRWN genes, while outlined empty rectangles reflect CRWN genes with null function in crwn mutant backgrounds. In particular, the size of dark green rectangle is reduced in crwn1 and crwn1 crwn2 mutant backgrounds against the background of an enlarged light green rectangle, indicating the down-regulation of CRWN4 protein and up-regulation of CRWN4 mRNA when crwn1 mutation is present. The potential compensatory activation or repression among different CRWN genes is displayed by solid red arrows or dashed blue lines. Other potential regulation is shown by question marks. The cartoons (bottom) illustrate the hypothesized balance of CRWN1-like and CRWN4 functions, depending on CRWN mRNA and protein expression levels in each crwn mutant. The loss of balance between CRWN1-like and CRWN4 functions is postulated to lead to extreme morphological and transcriptional changes.
196

Materials and Methods
Protein sequence analysis Predicted CRWN proteins sequences of various Arabidopsis thaliana ecotypes were downloaded from the publicly available genomic sequences of 856 natural strains on the 1001 genome website (http://signal.salk.edu/atg1001/3.0/gebrowser.php). Predicted CRWN protein sequences of Arabidopsis lyrata were obtained from the A. lyrata 1.0 genome from the JGI website (http://genome.jgipsf.org/Araly1/Araly1.home.html). Polymorphic amino acid sites were identified using the multiple protein sequence alignment webserver Clustal Omega [25] (https://www.ebi.ac.uk/Tools/msa/clustalo/).
Plant materials and growth conditions Seeds from wild type, crwn1, and crwn4 genotypes were planted on MS plates, coldtreated for 3 days in 4°C, and then germinated under long-day lighting conditions (16 h of light / 8 h of dark) at 23˚C in environmental growth chambers.
Nuclear extract preparation 2-week-old seedlings of wild type and crwn1 and crwn4 mutants were harvested, and fresh tissue homogenized thoroughly in Honda buffer [173] (0.44 M sucrose,1.25% Ficoll, 2.5% Dextran T40, 20 mM Hepes-KOH pH 7.4,10 mM MgCl2, 0.5% Triton X100) on ice, and filtered through Calbiochem® Miracloth (EMD Millipore Bioscience) twice. The clear liquid phase was centrifuged at 7000g for 30 minutes in 4°C, and
197

nuclear pellets were collected and washed three times using Honda buffer until the pellets turned gray. If needed, an additional wash using sucrose gradient nuclear extraction buffer (NEB3 in original protocol) was performed [171].
Antisera development and test Peptide antigens against the N-terminal regions of CRWN1 and CRWN4 proteins were synthesized by a commercial antisera production company (Proteintech): (CRWN1: MSTPLKVWQRWSTPTKATN; CRWN4: RVLKSPLTEEIMWKRLKD.) These peptides were injected to rabbits to raise the corresponding antisera, which were affinity purified for this study by Proteintech. The nuclear extracts from wild type and crwn mutants were boiled in Laemmli buffer for 5 min and resolved by SDS-PAGE gel electrophoresis. The specificity and efficacy of antisera were tested at different concentrations and washing stringency using protein immunoblots (see protocol below).
Salt, detergent and sonication treatment on nuclear extracts A mock nuclear lysis buffer was made as control (40mM HEPES pH 7.9, 3mM MgCl2, 0.4mM EDTA, 40% Glycerol, add protease inhibitor before use), and each condition was prepared by adding one or a combination of a few of these treatments to the lysis buffer: a particular concentrations of NaCl (0M, 0.5M, 1M, 1.5M, 2M), one or more detergents: Triton X-100 (1%, 2%), CHAPS (1%, 2%, 4%), or SDS (0.05%, 0.2%), a short (5 sec x 10) or long (5 sec x 20) sonication. The extracted nuclei were treated and gently rocked for 30 minutes at 4°C, followed by centrifugation to separate the
198

supernatant from the insoluble pellet. Proteins were harvested by boiling the supernatants and pellets separately in Laemmli buffer (with 5% beta-mercaptoethanol) for 5 min and resolved on SDS-PAGE gels. Protein immunoblots (see below) were used to determine whether CRWN proteins were solubilized and present in the supernatant samples in each treatment condition.
Protein immunoprecipitation and protein immunoblots The extracted nuclei were sonicated in immunoprecipitation (IP) buffer (1 mM EDTA, 100 mM Tris-Cl pH 7.4, 10% v/v glycerol, 75 mM NaCl, 0.05% w/v SDS, 0.1% vol/vol Triton X-100), and centrifuged to collect the supernatant. Different CRWN antisera was added into the IP buffer at 100ug/ml and incubated with gentle rotation overnight at 4°C. The next day, to recover the antisera-protein complexes, protein-G Dynabeads® (Life Technologies) were added into each sample with continued gentle rotation for two more hours at 4 °C . Then, the beads were collected using magnetic capture and washed 3 times using IP buffer. Proteins were harvested by boiling the beads in Laemmli buffer [174] (2% SDS, 10% glycerol, 0.01% Bromophenol blue, with 5% beta-mercaptoethanol) for 5 min, and resolved by SDS-PAGE gel electrophoresis. Proteins were semi-dry transferred onto a PVDF (polyvinylidene difluoride) membrane (GE Health Care) and the signals of CRWN proteins were detected using ECL (Enhanced Chemiluminescence) western blotting detection system (GE Health Care).
199

REFERENCES 1. Dechat T, Pfleghaar K, Sengupta K, Shimi T, Shumaker DK, Solimando L,
Goldman RD: Nuclear lamins: major factors in the structural organization and function of the nucleus and chromatin. Genes & development 2008, 22(7):832-853. 2. Butin-Israeli V, Adam SA, Goldman AE, Goldman RD: Nuclear lamin functions and disease. Trends in genetics : TIG 2012, 28(9):464-471. 3. Rout MP, Field MC: Isolation and characterization of subnuclear compartments from Trypanosoma brucei. Identification of a major repetitive nuclear lamina component. The Journal of biological chemistry 2001, 276(41):38261-38271. 4. Batsios P, Peter T, Baumann O, Stick R, Meyer I, Graf R: A lamin in lower eukaryotes? Nucleus 2012, 3(3):237-243. 5. Dittmer TA, Stacey NJ, Sugimoto-Shirasu K, Richards EJ: LITTLE NUCLEI genes affecting nuclear morphology in Arabidopsis thaliana. The Plant cell 2007, 19(9):2793-2803. 6. Masuda K, Xu ZJ, Takahashi S, Ito A, Ono M, Nomura K, Inoue M: Peripheral framework of carrot cell nucleus contains a novel protein predicted to exhibit a long alpha-helical domain. Experimental cell research 1997, 232(1):173-181. 7. Ciska M, Moreno S: NMCP/LINC proteins: Putative lamin analogs in plants? Plant signaling & behavior 2013, 8(12).
200

8. Zhou X, Meier I: How plants LINC the SUN to KASH. Nucleus 2013, 4(3):206-215.
9. Graumann K, Evans DE: Plant SUN domain proteins: components of putative plant LINC complexes? Plant signaling & behavior 2010, 5(2):154156.
10. Sakamoto Y, Takagi S: LITTLE NUCLEI 1 and 4 regulate nuclear morphology in Arabidopsis thaliana. Plant & cell physiology 2013, 54(4):622-633.
11. Dittmer TA, Richards EJ: Role of LINC proteins in plant nuclear morphology. Plant signaling & behavior 2008, 3(7):485-487.
12. Wang H, Dittmer TA, Richards EJ: Arabidopsis CROWDED NUCLEI (CRWN) proteins are required for nuclear size control and heterochromatin organization. BMC plant biology 2013, 13(1):200.
13. Ciska M, Masuda K, Moreno Diaz de la Espina S: Lamin-like analogues in plants: the characterization of NMCP1 in Allium cepa. Journal of experimental botany 2013, 64(6):1553-1564.
14. Kimura Y, Kuroda C, Masuda K: Differential nuclear envelope assembly at the end of mitosis in suspension-cultured Apium graveolens cells. Chromosoma 2010, 119(2):195-204.
15. Tamura K, Fukao Y, Iwamoto M, Haraguchi T, Hara-Nishimura I: Identification and characterization of nuclear pore complex components in Arabidopsis thaliana. The Plant cell 2010, 22(12):4084-4097.
201

16. Burkhard P, Stetefeld J, Strelkov SV: Coiled coils: a highly versatile protein folding motif. Trends in cell biology 2001, 11(2):82-88.
17. Harbury PB, Zhang T, Kim PS, Alber T: A switch between two-, three-, and four-stranded coiled coils in GCN4 leucine zipper mutants. Science 1993, 262(5138):1401-1407.
18. Schubert V: SMC proteins and their multiple functions in higher plants. Cytogenetic and genome research 2009, 124(3-4):202-214.
19. McLachlan AD, Stewart M: Tropomyosin coiled-coil interactions: evidence for an unstaggered structure. Journal of molecular biology 1975, 98(2):293304.
20. Dechat T, Adam SA, Taimen P, Shimi T, Goldman RD: Nuclear lamins. Cold Spring Harbor perspectives in biology 2010, 2(11):a000547.
21. Strelkov SV, Schumacher J, Burkhard P, Aebi U, Herrmann H: Crystal structure of the human lamin A coil 2B dimer: implications for the headto-tail association of nuclear lamins. Journal of molecular biology 2004, 343(4):1067-1080.
22. Weigel D, Mott R: The 1001 genomes project for Arabidopsis thaliana. Genome biology 2009, 10(5):107.
23. Atwell S, Huang YS, Vilhjalmsson BJ, Willems G, Horton M, Li Y, Meng D, Platt A, Tarone AM, Hu TT et al: Genome-wide association study of 107 phenotypes in Arabidopsis thaliana inbred lines. Nature 2010, 465(7298):627-631.
202

24. Hu TT, Pattyn P, Bakker EG, Cao J, Cheng JF, Clark RM, Fahlgren N, Fawcett JA, Grimwood J, Gundlach H et al: The Arabidopsis lyrata genome sequence and the basis of rapid genome size change. Nature genetics 2011, 43(5):476-481.
25. Jacob E, Unger R: A tale of two tails: why are terminal residues of proteins exposed? Bioinformatics 2007, 23(2):e225-230.
26. Pederson T: Half a century of "the nuclear matrix". Molecular biology of the cell 2000, 11(3):799-805.
27. Fiil BK, Qiu JL, Petersen K, Petersen M, Mundy J: Coimmunoprecipitation (co-IP) of Nuclear Proteins and Chromatin Immunoprecipitation (ChIP) from Arabidopsis. CSH protocols 2008, 2008:pdb prot5049.
28. Hirano T: At the heart of the chromosome: SMC proteins in action. Nature reviews Molecular cell biology 2006, 7(5):311-322.
29. Kalyva A, Schmidtmann A, Geeves MA: In vitro formation and characterization of the skeletal muscle alpha.beta tropomyosin heterodimers. Biochemistry 2012, 51(32):6388-6399.
30. Gaudino RJ, Pikaard CS: Cytokinin induction of RNA polymerase I transcription in Arabidopsis thaliana. The Journal of biological chemistry 1997, 272(10):6799-6804.
31. Laemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970, 227(5259):680-685.
32. Sievers F, Higgins DG: Clustal Omega, accurate alignment of very large numbers of sequences. Methods in molecular biology 2014, 1079:105-116.
203

CHAPTER 5
FUTURE DIRECTIONS
In this thesis, I used genetic, molecular and biochemical tools to study the long coiledcoil domain CRWN family proteins in flowering plants Arabidopsis thaliana. I showed that different combinations of crwn null mutations lead to a series of morphological changes in mutant plants, such as rounded nuclear shape, reduced nuclear area, disturbed endo-reduplication, increased nuclear DNA density, and disrupted heterochromatin organization. My transcriptomic profiling further uncovered various mis-regulated loci, which potentially contribute to these phenotypic alterations.
Phylogenetic analysis revealed the sub-classification of CRWN1-like and CRWN4-like genes, and this grouping was supported by the divergent morphological changes caused by mutation of CRWN1-like genes versus CRWN4. Surprisingly, these two clades of CRWN proteins regulate common genomic loci during transcription. At present, I do not know if the transcriptional mis-regulation in crwn mutants occurs via direct or indirect mechanisms. In any case, the convergence on common transcriptional target is possibly fostered by the physical interaction between CRWN1 and CRWN4 proteins, which I demonstrated in subsequent biochemical characterization. These data suggested a complementary relationship between
204

CRWN1-like and CRWN4 functions. I propose a a balancing model of these two complementary functions in organizing the nuclei.
CRWN1 and CRWN4 proteins are both nuclear proteins resistant to high salt and mild detergent extraction, and are candidates for structural components of the nucleoskeleton. To test this hypothesis, more cell biology and biochemistry investigations need to be carried out to understand the structural status of CRWN proteins in vivo, and the identity of working partners for CRWN proteins. FRAP (Fluorescence Recovery After Photobleaching), immuno-gold localization of CRWN proteins using EM (Electronic Microscopy), and more sophisticated cell fractionation could help illustrate the physical properties of CRWN proteins in the nuclei. Immunoaffinity purification and mass spectrometry techniques could be used to identify interacting partners in CRWN complexes, as well as reveal possible post-translational modifications. A candidate approach by immunoprecipitation could also be taken to explore CRWN-interacting proteins, such as SMC subunits, CRWN2, CRWN3, SUN1, and SUN2 proteins. Alternative systems could be used for validation, such as yeast-two-hybrid studies, split GFP or luciferase complementation in tobacco, and in vitro pull down of candidate proteins expressed in bacteria.
Another important aspect of CRWN function relevant to transcriptional regulation involves chromosome organization. Chromosome immunoprecipitation (ChIP) of CRWN1 and CRWN4 proteins could help understand whether any CRWN complexes interact with chromatin directly, and could determine whether CRWN complexes
205

target specific genomic regions. In addition, whole genome bisulfite sequencing (WGBS) of crwn mutants, as well as profiling of various histone modifications, would also provide information about the potential interaction between CRWN proteins and the epigenome. It would also be interesting to compare interacting loci from these profiling results with genomic regions recognized by other proteins (e.g., SMC subunits). Any shared profiles would aid in identification of partner proteins functionally interacting with CRWN, and help dissect the mechanisms through which CRWN proteins regulate nuclear organization.
Other interesting directions to explore include determining whether the physical properties of crwn nuclei differ from wild type nuclei, and if so, whether these changes are primary or secondary, and are associated with alteration in nuclear function. A second avenue to pursue is the long-term consequences of crwn-mediated nuclear changes. For instance, do epigenetic and genetic variation accumulate in crwn mutant backgrounds due to the mis-expressed epigenetic modifiers and DNA replication machinery. A third set of possible experiments include study of genomic level reorganization in crwn mutants. To understand whether crwn mutations alter the organization of chromosomes, chromatin conformation capture (3C) and related techniques could be utilized to study the chromosomal interactome in crwn mutant backgrounds.
206