Alsaai Udel 0060D 14351.Pdf

INVESTIGATION OF TDRD7 FUNCTION

IN OCULAR LENS DEVELOPMENT

Salma Mohammed Al Saai

A dissertation submitted to the Faculty of the University of Delaware in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biological Sciences

Summer 2020

INVESTIGATION OF TDRD7 FUNCTION

IN OCULAR LENS DEVELOPMENT

Salma Mohammed Al Saai

Approved: ______Velia M. Fowler, Ph.D. Chair of the Department of Biological Sciences

Approved: ______John A. Pelesko, Ph.D. Dean of the College of Arts & Sciences

Approved: ______Douglas J. Doren, Ph.D. Interim Vice Provost for Graduate and Professional Education and Dean of the Graduate College

I certify that I have read this dissertation and that in my opinion it meets the academic and professional standard required by the University as a dissertation for the degree of Doctor of Philosophy.

Signed: ______Salil A. Lachke, Ph.D. Professor in charge of dissertation

Signed: ______Melinda K. Duncan, Ph.D. Member of dissertation committee

Signed: ______Donna Woulfe, Ph.D. Member of dissertation committee

Signed: ______Shawn Polson, Ph.D. Member of dissertation committee

Signed: ______Jeffrey Caplan, Ph.D. Member of dissertation committee

ACKNOWLEDGMENTS To my father, who has instilled in me the passion for knowledge and science. My father always believed in me, and whenever I come across hard tasks he says: “Salma I know that if you want something you can do it! No matter how challenging it is!” To my husband who has been my best supporter/ friend/ mentor/ harshest critique of my work, best colleague to discuss science with……and the list goes on!

Thank you!

To my mother and all my family who have accompanied me throughout this joyful journey and have been relentlessly encouraging and praying for me, despite being miles away... Thank you all!

To Dr. Lachke, my advisor, who accepted me to his lab and gave me the privilege to work on all the amazing cutting-edge projects. I learned A LOT!!! and I will continue to seek knowledge and learn. While being in your lab, you have supported me all the time. You gave me a great freedom in conducting research which I always appreciated. The freedom you provided helped me build on more skills and become an independent thinker. Under your mentorship I became a scientist who sets up the bar so high, cares about the big question and picture but yet never dismisses any minute details because ultimately everything contributes to a well-constructed image of the whole story. Many, Many thanks to you!

To all my committee members who, despite their other commitments, have provided me with advice, critiqued my work and asked me hard question because they wanted me to become a better scientist. I appreciate all the time you have spent to support me during this journey. Many thanks to you all.

To all the past and current lab members, to all my students whom I have mentored, to all my friends and colleagues from other labs: you made the PhD journey a special experience! Together we shared moments of joy and laughter… moments of stress and frustration which would sporadically be healed with an invaluable moment of success. Thank you!

To the animal facility, who has always helped us in maintaining our experimental animals, so that we can focus on producing great science. Thank you!

To Betty, thanks for your smile which you greet me with every time I walk to your desk. Thanks for providing me with an answer/help for everything I ever asked you for. Thank you!

To all my friends who stayed by my side, kept checking on me, encouraged me, helped me, and always spent quality time with me. Thank you!

TABLE OF CONTENTS

LIST OF TABLES ...... x LIST OF FIGURES ...... xi ABSTRACT ...... xiv

Chapter

1 INTRODUCTION ...... 1

1.1 The ocular lens ...... 1 1.2 Lens fiber cell architecture ...... 3 1.3 Lens cytoskeleton ...... 7

1.3.1 F-actin dynamics ...... 11

1.4 Cataract ...... 12 1.5 Tdrd7 gene and protein ...... 14 1.6 Tdrd7 function in development ...... 17 1.7 Loss of Tdrd7 causes fully penetrant cataracts in mouse ...... 20 1.8 Tdrd7-/- mice exhibit severe lens fiber cell defects prior to overt cataract formation ...... 21

2 MATERIALS AND METHODS ...... 24

2.1 Animals studies ...... 24 2.2 DNA isolation from mouse tails ...... 25 2.3 Mouse genotyping ...... 26 2.4 Western Blot Analysis ...... 27 2.5 Cryopreservation of eye tissue and cryosectioning ...... 28 2.6 Immunofluorescence ...... 28 2.7 Fluorescence in situ hybridization coupled with Immunostaining ...... 30 2.8 RNA sequencing for Postnatal stage 15 in Tdrd7-/- mouse lens ...... 31 2.9 Analysis of RNA sequencing data ...... 31 2.10 Tandem Mass Tag (TMT) Mass spectrometry for Postnatal stage 15 in Tdrd7-/- mouse lens ...... 32 2.11 Cluster heatmap for normal expression pattern of misregulated proteins and genes in P15 Tdrd7-/- lens ...... 32

vii

2.12 Gene Set Enrichment Analysis (GESA) ...... 33 2.13 Immunoprecipitation for Tdrd7 protein coupled with mass spectrometry ...... 33 2.14 RNA immunoprecipitation assay for Tdrd7 protein ...... 35 2.15 RNA Isolation ...... 36 2.16 Reverse Transcriptase for Quantitative PCR ...... 36 2.17 Scanning electron microscopy for Cap2-/- mouse lens ...... 37 2.18 Correlation analysis for Tdrd7 and downstream misexpressed genes ..... 38 2.19 Immunofluorescence Quantification analysis ...... 39

3 TDRD7 CONTROLS THE LEVEL OF HEAT SHOCK PROTEIN HSPB1 (HSP27) AND LENS FIBER CELL MORPHOLOGY ...... 40

3.1 Introduction ...... 40 3.2 Results ...... 44

3.2.1 Tdrd7-/- mice exhibit morphological defects in lens fiber cells that have undergone nuclear degradation ...... 44 3.2.2 Hspb1 is a top priority candidate down-regulated gene in Tdrd7-/- lens ...... 48 3.2.3 Hspb1 is expressed in normal lens development and is reduced in Tdrd7-/- lens ...... 50 3.2.4 Tdrd7 protein associates with Hspb1 mRNA ...... 52 3.2.5 Single molecule fluorescence in situ hybridization shows that Tdrd7 protein co-localizes with Hspb1 mRNA in differentiating lens fiber cells ...... 53

3.3 Discussion ...... 56

4 MULTI-OMICS INVESTIGATION OF TDRD7 IN MOUSE LENS ...... 60

4.1 Introduction ...... 60 4.2 Results ...... 61

4.2.1 Analysis workflow ...... 61 4.2.2 Tdrd7-/- mouse lens RNA/proteome profiling at P15 identifies misexpressed genes and proteins ...... 63 4.2.3 Identification of Tdrd7 downstream non-coding genes ...... 63 4.2.1 Characteristic expression pattern for high-priority differentially expressed candidates in Tdrd7-/- lens ...... 73 4.2.2 Snhg12 gene and Cap2 protein are among top priority candidates in Tdrd7-/- lens ...... 76 4.2.3 Gene Set Enrichment Analysis (GESA) identifies processes that are implicated by the absence of Tdrd7 in the lens ...... 80

viii

4.3 Discussion ...... 84

5 CHARACTERIZATION OF LENS DEFECTS IN CAP2 KNOCKOUT MICE ...... 88

5.1 Introduction ...... 88 5.2 Results ...... 91

5.2.1 iSyTE (integrated Systems Tool for Eye gene discovery) Identifies Cap2 as a Potential Key Factor in the Lens ...... 91 5.2.2 Cap2 expression highly correlates with Tdrd7 and Pax6 through developmental stages in mouse lens ...... 93 5.2.3 Cap2-/- lens fiber cell exhibits morphological defects ...... 95

5.3 Discussion ...... 100

6 DISCUSSION AND FUTURE DIRECTIONS...... 104

6.1 Introduction ...... 104 6.2 Identification of Tdrd7 RNA targets in the lens by RNA- immunoprecipitation coupled with RNA-sequencing (RIP-seq) ...... 105 6.3 Validation of Tdrd7-protein binding partners in the lens identified by Tdrd7 pull-down followed by mass spectrometry ...... 106 6.4 Investigate the mechanistic basis of Tdrd7 regulation of misregulated downstream candidates identified in Tdrd7-/- lens ...... 108 6.5 Generating double knockouts mice to test genetic interactions of Tdrd7 ...... 109 6.6 Study the biomechanical properties of Tdrd7-/- lens ...... 109

REFERENCES ...... 111

Appendix

A RNA-SEQ RESULTS IN TDRD7-/- P15 ...... 129 B TMT/MS PROTEOME RESULTS ...... 155 C LIST OF TDRD7 PULLDOWN PROTEINS ...... 162 D ANIMAL APPROVED PROTOCOL ...... 166 E REPRINT PERMISSION ...... 167

LIST OF TABLES

Table 2.1: Antibodies and immunofluorescence staining conditions ...... 29

Table 4.1 Summary for Ensemble biotype classes identified in P15 Tdrd7-/- lens RNA-seq data ...... 64

Table 4.2 A list of misregulated non-protein coding genes in Tdrd7-/- lens ...... 66

Table 4.3 List of differentially expressed genes in P15 Tdrd7-/- mouse lens ...... 68

Table 4.4 List of differentially expressed proteins in P15 Tdrd7-/- mouse lens ..... 71

Table-A. 1 List of all significantly missed expressed genes from Tdrd7-/- lens transcriptome with a cut of FDR <0.5...... 129

Table-B. 1 List of all significantly missed expressed proteins from Tdrd7-/- lens transcriptome with a cut of P-value <0.5 ...... 155

Table-C. 1 Tdrd7 pulldown list ...... 162

LIST OF FIGURES

Figure 1.1 The lens is a transparent tissue in the eye ...... 1

Figure 1.2 Organization of cells in the lens ...... 3

Figure 1.3 Schematic representation of lens fiber cell architecture ...... 5

Figure 1.4 Lens fiber cell architecture ...... 6

Figure 1.5 F-actin dynamics ...... 12

Figure 1.6 A schematic representation of the TDRD7 Protein ...... 15

Figure 1.7 RNA Binding Proteins Function in Post-transcriptional Gene Expression Control ...... 16

Figure 1.8 Tdrd7 is expressed in the lens and its deficiency causes cataract ..... 18

Figure 1.9 Phenotypic characterization of Tdrd7-/- lens and eye defects...... 22

Figure 1.10 Scanning electron microscopy (SEM) of Tdrd7-/- lenses demonstrates abnormal cortical fiber cell morphology ...... 23

Figure 2.1 Schematic for Immunoprecipitation assay ...... 35

Figure 3.1 iSyTE-based analysis of Tdrd7-/- lens RNA-Seq data identifies Hspb1 among high-priority candidates...... 41

Figure 3.2 2D-DIGE proteome screen identifies HSPB1 protein to be down- regulated in Tdrd7-/- lens ...... 42

Figure 3.3 Abnormal F-actin staining in maturing lens fiber cells in Tdrd7-/- mice...... 46

Figure 3.4 WGA staining indicates membrane abnormalities in maturing Tdrd7-/- lens fiber cells ...... 47

Figure 3.5 Hspb1 is a top candidate gene among the differentially expressed genes in the Tdrd7-/- lens ...... 49

Figure 3.6 Hspb1 protein expression in normal mouse lens development ...... 51

Figure 3.7 Hspb1 protein is abnormally reduced in Tdrd7-/- lens fiber cells ..... 52

Figure 3.8 Single RNA fluorescence in situ hybridization (smFISH) coupled with immunostaining demonstrates close association of TDRD7 protein-Hspb1 mRNA...... 55

Figure 3.9 Models for Tdrd7 function in lens development and cataractogenesis ...... 58

Figure 4.1 Analysis work flow for Tdrd7-/- lens proteome and transcriptome . 62

Figure 4.2 Distribution of differentially expressed transcripts in Tdrd7-/- lens 65

Figure 4.3 Cluster heatmap for lens normal expression from iSyTE for the candidate misregulated genes in P15 Tdrd7-/- lens ...... 74

Figure 4.4 Cluster heatmap for lens normal expression from iSyTE for the candidate misregulated proteins in P15 Tdrd7-/- lens ...... 75

Figure 4.5 Top potential candidate misregulated genes and proteins in Tdrd7-/- lens ...... 78

Figure 4.6 Functional classification for high-priority candidates whose expression correlates with Tdrd7 in normal lens development ...... 79

Figure 4.7 Misregulation of MIP in P15 Tdrd7-/- lens ...... 80

Figure 4.8 Scatter plots for differentially expressed (DE) genes enriched GO terms ...... 82

Figure 4.9 Scatter plots for differentially expressed (DE) proteins enriched GO terms ...... 83

Figure 5.1 Cap2-/- Mice Exhibit Microphthalmia ...... 89

Figure 5.2 Cap2 Protein is abnormally elevated in Tdrd7-/- Mouse Lens ...... 90

Figure 5.3 iSyTE predicts Cap2 as a high-priority candidate in the lens ...... 92

Figure 5.4 Cap2 is Mis-regulated in Several Cataract Mouse ...... 93

Figure 5.5 Cap2 expression correlates with Tdrd7 across nine developmental stages ...... 94

xii

Figure 5.6 Scanning electron microscopy of Cap2-/- lens ...... 95

Figure 5.7 Cap2-/- mouse lens exhibits abnormal cellular morphology and F- actin distribution in fiber cell ...... 97

Figure 5.8 Abnormal cellular F-actin distribution in Cap2-/- mature fiber cells ...... 98

Figure 5.9 Hexagonal appearance alterations in Cap2-/- mouse lens fiber cells 99

Figure 5.10 Measurement of uniformity of fiber cell organization in Cap2-/- mouse lens fiber cells ...... 100

Figure 6.1 Tdrd7 protein is associated with Epn1 mRNA in the lens ...... 106

Figure 6.2 Tdrd7 pull down from from wild type mouse lense ...... 107

xiii

ABSTRACT

Clear eyesight requires maintenance of a transparent ocular lens tissue throughout the life of an individual. Loss of lens transparency, termed cataract, is the leading cause of blindness worldwide. Depending on the age of their onset, cataracts can be age related (common above the age of 70) or congenital (present at birth or within the first year of life). The current cataract treatment option is surgery, which is costly but is successful in restoring vision, especially in adults. Treatment of congenital cataract in children is clinically challenging due to long-term complications of surgery such as retinal detachment and secondary glaucoma. Additionally, timely restoration of vision is critical because an opaque lens in early childhood may cause amblyopia, sensory deprivation, which can prevent normal eye development and cause permanent blindness. About 25%-50% of congenital cataracts are estimated to be caused by an underlying genetic alteration and over 40 genes have been linked to congenital cataract. Mutations in the TDRD7 (Tudor domain containing 7) gene are linked to human congenital cataract, and TDRD7 polymorphisms are associated with age related cataract. Additionally, multiple mouse models that are Tdrd7 deficient have been described and they exhibit fully penetrant cataract, with defects similar to that observed in human. Tdrd7 encodes a ribonucleoprotein / RNA granule component protein that is predicted to bind to other proteins through its Tudor domains and OST-HTH/LOTUS domains and closely associate with RNA. The identification of Tdrd7 mutations linked to congenital cataract suggests that regulators of post-transcriptional gene expression control are critical for lens development and maintenance of transparency in vertebrates.

xiv

However, the pathobiology of the Tdrd7-deficienct cataract defects and mechanistic basis of Tdrd7 function in the lens is not fully understood. My dissertation work sought to address this knowledge gap by studying Tdrd7 targeted knockout mouse (Tdrd7-/-) as a model of this human disease. Phenotypic characterization of Tdrd7-/- mouse lens identified fully penetrant cataracts by postnatal age 22 (P22). Analysis by scanning electron microscopy revealed that Tdrd7-/- lenses at P18, which are non-cataractous, exhibit fiber cell defects.

However, the nature of the fiber cell defect was undefined. By detailed characterization of Tdrd7-/- lenses using molecular and cellular approaches, I have uncovered a novel function for Tdrd7 in regulating fiber cell morphology. My confocal microscopy data on Tdrd7-/- lens cross-sections stained for Phalloidin (to visualize F-actin) shows that severe F-actin cytoskeletal defects specifically in fiber cells in late stages of post-nuclear degradation. Integrated bioinformatics analysis of RNA-seq data coupled with protein two-dimensional fluorescence difference in-gel electrophoresis and mass spectrometry screen on Tdrd7-/- lens 14 days prior to the onset of cataract, identified Tdrd7 to be upstream of the heat shock protein Hspb1. My experiments confirmed the reduction of Hspb1 mRNA and protein in the Tdrd7-/- lens. Further, I also performed confocal microscopy to characterize Hspb1 spatiotemporal expression in normal lens development. I also identified Tdrd7 protein to be associated with the mRNA of Hspb1 in the lens by both biochemical (i.e. RNA-immunoprecipitation coupled with RT-PCR) and fluorescence microscopy (i.e. small RNA imaging coupled with immunostaining) approaches. Together, these findings in my dissertation have led to a model where Tdrd7 sustains optimal high levels of the Hspb1 in maturing lens fiber cells that have undergone organelle degradation. These lens fiber cells in terminal stages of maturation

can be considered to resemble a stress-like condition. Moreover, Hspb1 is known to interact with F-actin and function in maintaining the cytoskeleton under conditions of stress. Thus, the model derived from my data further suggests that Tdrd7-based control of Hspb1 levels may be necessary for maintenance of F-actin cytoskeleton and cellular morphology of fiber cells post-nuclear degradation. Together these findings serve to expand the role of Tdrd7-family proteins in cellular differentiation during organogenesis.

Next, to gain further molecular insights into the alterations caused by the deficiency of Tdrd7, I have generated and investigated the transcriptome (by RNA-seq) and the proteome (by TMT-coupled with Mass spectrometry) of Tdrd7-/- mouse lens 7 days prior to the onset of cataract (postnatal day P15). Using a systematic integrated approach to carefully analyze both data sets I was able to identify several new high- priority candidate genes (Bfsp1, Bfsp2, Cap2) and proteins in P15 Tdrd7-/- lens, several of which may contribute to the cataract defects as well as may serve to further explain the cytoskeletal defects. These data suggests a role for Tdrd7 in maintaining several aspects of the lens cytoskeleton. In addition, these new data have identified a long noncoding RNA (Snhg12) to be severely reduced in Tdrd7-/- P15 lens transcriptome. These findings further supports the importance of Tdrd7 as a key regulatory factor important for maintaining lens transparency. Finally, I have examined a new high-priority target downstream of Tdrd7, namely Cap2 (Cyclase Associated Proteins2). The P15 Tdrd7-/- lens RNA-seq and proteome analysis identified Cap2 among the top upregulated candidates. By iSyTE analysis, I showed that Cap2 is highly expressed and enriched in the mouse lens across several embryonic and postnatal stages. I performed cellular and molecular

xvi

characterization of Cap2-/- mouse lens. Interestingly, Cap2-/- lenses exhibited lens fiber cell morphological defects, including F-actin aggregates, in the same region of lens that exhibited defects in Tdrd7-/- mice, namely the region containing maturing fiber cells, post-nuclear degradation. Thus, I have identified Cap2 as a new factor critical for lens biology. Together, these data provide new evidence that Tdrd7 functions in the control of the cellular morphology of maturing fiber cells at the beginning of the organelle degradation zone in the lens.

In sum, work in this dissertation has uncovered the cellular and molecular basis of the lens defects in Tdrd7-/- mice, in turn leading to the advancement of our understanding of the pathobiology of Tdrd7-deficiency based cataracts.

xvii

Chapter 1

INTRODUCTION

1.1 The ocular lens The ocular lens is a transparent avascular tissue which is located in the anterior chamber of the eye (Figure 1.1). It is composed of a monolayer of lens epithelial cells facing the cornea at the anterior hemisphere, and a bulk of lens fiber cells in the posterior (Bassnett et al., 2011a; Bhat, 2001). The lens tissue is fully surrounded by a basement membrane known as the lens capsule (Danysh and Duncan, 2009).

Figure 1.1 The lens is a transparent tissue in the eye (A) The eye is a multicomponent organ, with the lens located in the anterior chamber. (B) The ocular lens is a transparent tissue ((A) Adapted from Dash et al. 20116; (B) lens image: Salma Al Saai).

Besides providing a physical barrier and an extracellular matrix for the lens, the capsule also allows the passive exchange of nutrients and waste in an out of the lens (Beyer et al., 1984; Friedenwald, 1930). Epithelial cells in the germinative zone of the lens divide mitotically and get displaced toward the equatorial regions, termed the “transition zone”, where they exit the cell cycle and initiate differentiation into fiber cells (Figure 1.2) (Bhat, 2001). In a coordinated manner, transcriptionally active early differentiating lens fiber cells undergo elongation and as they terminally differentiate, they migrate towards the center of the lens. During terminal differentiation and maturation, fiber cells undergo organelle degradation. The process of organelle degradation gives rise to a central region in the lens known as the organelle free zone (Figure 1.2) (Bassnett, 2009a). This process ensures a light scatter-free zone and a high refractive index around the visual axis of the lens, owing to the lens its transparency (Bassnett, 2009a). The lens tissue has a complex architecture that is necessary for achieving its function of refracting and focusing light on to the retina. Cells in the lens are organized in a way that if one would peel one fiber at a time starting from the outermost (younger fiber cells) towards the innermost (older fiber cells) until the center (oldest fiber cells/lens nucleus), a temporal history of the molecular changes throughout a lens lifespan would be uncovered (Bhat, 2001).

Figure 1.2 Organization of cells in the lens Schematic demonstrating the organization of lens cells. In cross section, the fiber cells appear as hexagon with two broad sides and two short sides. The core of the lens is formed during embryonic development. Fiber cells in the core have irregular shape and arrangement. They are “nuclear” fiber cells formed during embryogenesis and early lens development. Tips of the elongating fibers cells meet at the anterior and posterior lens creating the lens suture. Abbreviations: (Epi) Epithelial fiber cells. Fiber cells (Fib). nuclei (N). Organelle free zone (OFZ). Anterior suture (AS). Posterior suture (PS). Primary fiber (Adapted from Bassnett et al. 2009)

1.2 Lens fiber cell architecture Lens fiber cells have a complex structure (Shi et al., 2009). A cross-section of a mouse lens reveal a profile of tightly arranged hexagonal fiber cells (Figure 1.3). The hexagonal shape ensures optimum packing which eliminates intracellular spaces, resulting in minimum light scattering and high refractive index (Kuszak, 1995; Kuszak et al., 2004). The short side of lens fiber cells is presented with membrane interdigitations that function to connect neighboring fiber cells (Figure 1.4). These interdigitations become progressively prominent as fiber cells go through

differentiation, elongation and maturation and their shape gradually transition into finger-like projections (Audette et al., 2017; Bassnett et al., 2011a). These membrane protrusions are characterized by undulations that becomes higher in amplitude as lens fiber cells move towards the core (Figure 1.4-A,B). Fiber cells in the core of the lens – termed the lens nucleus – lose these undulations and become irregular in morphology. Thus, at different radial depths of the lens, lens fiber cells have different membrane architectures (Figure 1.4-C) (Audette et al., 2017; Bassnett et al., 1999). As for the broad sides of hexagons, young fiber cells have a “ball and socket” membrane structure that are rich in gap junctions providing an inter-cellular communication between neighboring cells (Figure 1.4-A,B) (Bassnett et al., 2011a). The ball and socket junctions are observed in mature fiber cells but not obviously discernible in deeper layers (Audette et al., 2017; Kuszak, 1995).

Figure 1.3 Schematic representation of lens fiber cell architecture (A) Schematic illustrating the orientation of the ocular lens tissue in a cross section through the equator. (B) Cross section cut diagram that shows organized hexagonally packed fiber cells. Image (B) was 90° rotated with respect to image (A). (C) Shows the broad sides (red), short sides (blue) and vertices (yellow) of an individual lens fiber cell. (D) SEM for a cut across fiber cell major axis. This allows visualization of hexagonal geometry. Yellow vertices and white arrows indicate membrane protrusions along the edges. Red– broad side; blue– short side; yellow– vertices/membrane protrusions; (E) Is a cross section of the lens at the equatorial region. (Epi) represents epithelial cells that are found on the periphery. To the right of the Epi are hexagonally shaped lens fiber cells. Note that each individual fiber cell (red asterisk) is surrounded by 6 neighboring cells, colored blue, yellow or orange. ((A-C and E) Adapted from Cheng at al., 2017, (D) Adapted from Audette et al., 2017)

Figure 1.4 Lens fiber cell architecture (A) young (cortical) lens fiber cells are characterized by the presence of membrane interdigitations on the sides that function to connect neighboring fiber cells (B) lens fiber cells interdigitations change progressively into as fiber cells mature and their shape gradually develop into paddles and valleys (note: these cells have undergone organelle degradation), abbreviations: s:socket, b: ball (C-D) A

schematic for lens fiber cell membrane protrusions (C) Diagram for normal cortical, differentiating, and mature fiber cells. (D) Diagram (not to scale) of mature lens fiber cells with large paddles (light blue shading), valleys between large paddles (red lines), small protrusions (green shading), and bases of small protrusions (yellow lines) along the short sides of lens. (A-B Adapted from Bassnett et al. 2009 and Audette et al., 2017 , C-D Adapted from Cheng at al. 2016 IOVS) different stages of differentiation and maturation, from newly formed cortical fiber cells (closest to the epithelial cells) to mature fiber cells after organelle degradation.

1.3 Lens cytoskeleton In mammalian cells, the organization of the plasma membrane associated actin cytoskeleton is important for membrane morphology, cell shape, adhesion, and polarity (Fischer et al., 2003; Kuwabara, 1968; Rao and Maddala, 2006a, 2006a). Mature lens fiber cells are highly elongated and their transparency depends on cytoskeletal proteins that support the organization of cytosolic and membrane-bound proteins (Rao and Maddala, 2006a), and defects in the cytoskeletal organization can lead to cataracts

(Bennett et al., 2016; Gokhin et al., 2012; Logan et al., 2017). Lens fiber cells morphogenesis is supported by three major cytoskeletal components: 1. microtubules, 2. intermediate filaments and 3. actin filaments. Each one of these components serves a distinct function in the lens as discussed below. Microtubules: These are distributed along the axis of the elongated lens fiber cells and their expression is high in cortical fiber cells (Farnsworth et al., 1980; Kuwabara, 1968). Microtubules are implicated in lens vesicular trafficking, fiber cell elongation as well as morphogenesis (Kuwabara, 1968; Logan et al., 2018). Intermediate filaments: Filensin (also known as Beaded Filament Structural Protein 1 - Bfsp1) and phakinin (also known as Beaded Filament Structural Protein 2 - Bfsp2) are two divergent members of the intermediate filaments protein family, and in

the lens they are associated in a macromolecular complex with spectrin-actin network along with tropomodulin1 (Tmod1) to form a distinct structure known as beaded filaments (Fischer et al., 2003). Filensin and phakinin expression starts in early differentiating fiber cells and peaks in elongated fiber cells (Blankenship et al., 2001). It is suggested that beaded filaments are important for maintaining transparency and mechanical stiffness in adult lens (Fudge et al., 2011; Gokhin et al., 2012). F-actin network: I discuss the third component which is filamentous actin (F- actin) in detail due to its relevance to my work. F-actin is a polymer composed of assembled monomeric globular actin (G-actin). F-actin has two ends, the pointed (minus) end and the barbed (plus) end. Without an apparent change in the cellular ratio of G-actin to F-actin, there is continuous depolymerization taking place at the pointed end and continuous polymerization taking place at the barbed end. This turnover of F- actin is known as treadmilling (Ono, 2013). The F-actin network is dependent on a diverse group of proteins. Multiple actin-regulatory proteins modulate the polymerization and depolymerization of actin, such as Cyclase-associated protein (CAP) proteins. The dynamics of actin reorganization is crucial to a many biological processes (Ono, 2013), such as the process of lens fiber cell elongation which is coupled with cell migration along the anterior and posterior of lens. (Rao and

Maddala, 2006b). The synchrony of these processes in the lens results in a symmetric mass of aligned lens fiber cells and a suture structure at both poles (Figure 1.2). An increase in the cellular ratio of F-actin to G-actin has been linked to the process of lens fiber cell differentiation (Ramaekers et al., 1981). Indeed, disruption of F-actin to G- actin ratio was shown to cause cell migration defects (Maddala et al., 2011a). Lens fiber cell morphology also depends highly on regulated cell movement (Piatigorsky,

1981), a process that depends on the formation of a cellular structure termed “lamellipodium” which is also regulated by F-actin polymerization and actin capping proteins (Rao and Maddala, 2006b).

Unlike microtubules and beaded filaments, which degrade along with other organelles in the center of the lens, F-actin does not get degraded in mature lens fiber cells (Lee et al., 2000). F-actin is found to be associated with lens fiber cell membranes beyond the organelle free zone (Fudge et al., 2011; Kuwabara, 1968; Lee et al., 2000). These findings suggest that F-actin and its associated proteins are critical for the stability and integrity of mature lens fiber cells (Cheng et al., 2017). F-actin is assembled and arranged in bundles along the entire length of young lens fiber cell lateral membrane and is enriched at the tricellular junctions of hexagonally packed lens fiber cells (Lo et al., 1994). A subset of specialized F-actin cross-linking proteins have been described in the lens (Sjöblom et al., 2008). These proteins associate with F-actin bundles such as alpha-actinin, which is a member of the spectrin protein superfamily (Lo et al., 1994, 1997). Another F-actin cross-linker is fimbrin, (also known as plastin) and it bundles F-actin in a parallel orientation (Delanote et al., 2005). Fimbrin is expressed in lens epithelial cells and is found in membranes of differentiating and mature lens fiber cells, appearing as puncta at the base of small protrusions (Cheng et al., 2016a). It has been suggested that membrane protrusions between neighboring lens fiber cells are maintained by coordinated pulling and pushing force generated by F-actin branching filaments and the membrane protein clathrin along with AP-2 adaptor protein (Logan et al., 2017). Aquaporin 0 (AQP0), also known as major intrinsic protein (MIP), is also found along these protrusions. MIP dually acts as

a water channel and an adhesion molecule (Lo et al., 2014; Varadaraj et al., 1999). Mutations in MIP are linked to cataract in human (Senthil Kumar et al., 2013). Mice lacking MIP exhibit defective lens fiber cell packing, reduced focusing power and develop cataract (Al-Ghoul et al., 2003; Lo et al., 2014; Shiels et al., 2001). N-cadherin also contributes to lens fiber cell morphogenesis through its influence on F-actin dynamics (Leonard et al., 2011). It was shown that N-cadherin associates with actin- related protein-3 (Arp3) in mature lens fiber cells (Leonard et al., 2011). Arp3 is a component of Arp2/3 complex that acts a nucleator for F-actin assembly (Higgs and Pollard, 2001). Single lens fiber cell staining demonstrated that Arp3 is abundantly expressed as puncta at the base of the small F-actin protrusions (Cheng et al., 2016a). Arp3, N-cadherin and F-actin colocalize at the vertices of lens fiber cell hexagons, while only N-cadherin and F-actin colocalizes at the lateral cell borders (Leonard et al., 2011). Additionally, fiber cell elongation is defective in N-cadherin lens-specific conditional knockout mouse (Logan et al., 2017). These mice also exhibit upregulation of Ephrin type-A receptor 2 (Epha2), and interestingly EphA2 knockout mouse fiber cells exhibit disrupted or absent N-cadherin (Cheng and Gong, 2011; Logan et al., 2017). These findings on Epha2 and N-cadherin suggests that both may coordinately function in lens fiber cell elongation (Logan et al., 2017). EPHA2 mutations are linked to both congenital and age-related cataract in human (Cheng et al., 2013; Jun et al., 2009). Ezrin–radixin–moesin (ERM) proteins bind actin filament networks to the plasma membrane (Bretscher et al., 2000). In the lens, Ezrin, for example, is uniformly expressed along the membrane of mature fiber cells (Bagchi et al., 2004). In lens cross sections, ezrin can be seen colocalizing to F-actin, with its expression being enriched in at the base of small protrusions and valleys between large paddles of mature fiber cells

(Cheng et al., 2016a; Maddala et al., 2011b). Genetic variations in Ezrin gene have been lined with age related cataracts in human (Lin et al., 2013). The F-actin-spectrin network is essential for lens fiber cells stability and packing (Fischer et al., 2003). In fiber cell cross sections, β2-spectrin appears as puncta along the broad sides and along the border of the paddle domains (Cheng et al., 2016a). Periaxin (Prx), a PDZ domain protein, is also a component of adherens junctions in fiber cells (Maddala et al., 2011b). PDZ domains play a key role in anchoring receptor proteins in the membrane to cytoskeletal components (Lee and Zheng, 2010). In lens, Prx protein localization is dynamic and exhibits enriched expression at the tricellular junctions of mature fiber cells (Maddala et al., 2011b). Prx is involved in membrane organization by forming a macromolecular complex with other proteins such spectrin, filensin and ezrin. Mice lacking Prx exhibit softer lens and disruption in lens fiber cell hexagonal packing (Maddala et al., 2011b).

1.3.1 F-actin dynamics F-actin is a polar polymers and is characterized by a pointed end (-) and a barbed end (+) (Figure 1.5). Each of these ends have different dynamic properties (Blanchoin et al., 2014; Ono, 2013). ATP bound monomeric G-actin monomers have high affinity to bind to the barbed end (+) of F-actin than to the pointed end(-). A term known as treadmilling describes the status where G-actin continuously polymerizes at the barbed and depolymerizes from the pointed end without an actual change in the ratio of G- and F-actin. Due to the stability of F-actin, its disassembly requires actin depolymerizing factors such as cofilin 1 (Cfl1) and Cap2. These factors act by severing F-actin and promoting its turnover (Ono, 2013).

Figure 1.5 F-actin dynamics A schematic illustrating F-actin polymer, indicating the pointed end (-) and the barbed end (+). Free G-actin have a higher affinity to be polymerized into F-actin via the barbed end (+). Depolymerization and polymerization of F-actin is a constant process in the cell that is maintained and controlled by many factors

1.4 Cataract Loss of ocular lens transparency is termed cataract (Graw, 2009). The lens must be transparent to transmit light sharply on the retina, which in turn converts this photon based information into signal impulses and sends them to the brain, which results in optimal visual perception (Bazan, 1989). Depending on the age of onset, cataract cases can be classified as age related or pediatric (McCarty and Taylor, 2001). According to the NIH-eye, by the age of 75, half of the elderly in the US develop age related cataracts. Unfortunately, the number of current cataract cases (24.4 million) in the US is expected to double (reaching ~50 million cases) by 2050. Age related cataracts are complex in nature because they result from the interaction of multiple risk factors such as genetic mutations, environmental insults, or disease state like diabetes. Most of these risk factors are not well characterized (McCarty and Taylor, 2001; Shiels and Hejtmancik, 2015, 2019). Pediatric cataracts are termed congenital when they are present at birth or

develop during the first year of life, or are termed juvenile when they develop within the first 10 years of life (Shiels and Hejtmancik, 2013). If congenital cataract is not treated by surgical intervention in a timely manner, it might cause complications such as amblyopia - a sensory deprivation that can result in permanent blindness (Kuhli- Hattenbach et al., 2008). However, postoperative care for cataract surgery in young children (below 18 months) is clinically challenging. In young children an immediate vision correction after the removal of the cataractous lens is critical (Kuhli-Hattenbach et al., 2008). One of the options for correcting aphakia (lens removal) is using intra ocular lenses. In children younger than two years old the use of intra ocular lenses is complicated with a rapid eye growth. This complication makes it hard to insert a lens with a correct power (Yorston, 2004). Thus, on top of the intraocular lenses, spectacles are required to achieve the best vision possible. In addition, routine frequent follow-ups are a must, to ensure proper eye development and to prevent or manage surgery related complications. Failure to comply with close medical follow-ups can be detrimental, causing secondary glaucoma or retinal detachment, both of which can lead to blindness (Kuhli-Hattenbach et al., 2008; Yorston, 2004). About 25%-50% of pediatric cataracts are estimated to be caused by an underlying genetic alteration (Shiels and Hejtmancik, 2015). Genetic mutations can be hereditary or can result from an intrauterine insult such as exposure to rubella infections or ionizing radiations like X-rays (Bardram and Brændsdrup, 1947; Lim et al., 2015; Mets, 2001) . New cataract loci are being identified in a rapid rate. Currently, approximately 60 mapped loci for isolated congenital cataracts have been linked to alterations in over 40 genes (Hejtmancik, 2008; Messina-Baas and Cuevas-Covarrubias, 2017; Shiels and Hejtmancik, 2019). Congenital cataract-linked genes fall within the following functionally classified families: 1) structural proteins

such as crystallins, 2) gap junction proteins such as connexins, 3) membrane proteins such as aquaporins and 4) cytoskeletal proteins such as beaded filaments (Hejtmancik, 2008). Recently, mutations in post-transcriptional regulators and RNA-binding proteins were added to the list of cataract linked genes with the discovery of TDRD7, deficiency of which causes congenital cataract in human and animal models (Chen et al., 2017a; Lachke et al., 2011a, 2012; Tan et al., 2017), as well as other proteins such as Caprin2, Celf1 and Rbm24 (Aryal et al., 2020a; Dash et al., 2015, 2020; Siddam et al., 2018).

1.5 Tdrd7 gene and protein TDRD7 (Tudor domain containing 7; also known as TRAP (tudor repeat associated with PCTAIRE 2) gene (OMIM: 611258) is located on chromosome 9 in human and on chromosome 4 in mouse. It was first described in a screen among proteins that interacted with the Cdc2-related kinase, PCTAIRE (Hirose et al., 2000). Tdrd7 belongs to Tudor domain family proteins that act as molecular adaptors (Pek et al.,

2012). Tudor domains (Figure 1.6) consists of approximately 60 amino acids making up five anti-parallel β-strands that give rise to a β-barrel structure (Côté and Richard, 2005). The barrel structure is covered with a pocket lined with aromatic residues and it has a special ability to bind proteins with methylated lysine or methylated arginine residues (Chen et al., 2011; Côté and Richard, 2005). It is suggested that Tudor domain- mediated binding of proteins can promote assembly of macromolecular complexes (Pek et al., 2012). Tdrd7 also has three copies of a novel conserved domain at the N-termini termed OST-HTH (Oskar-TDRD5/TDRD7-Helix-Turn-Helix)/LOTUS (Limkain, Oskar, and Tudor containing proteins 5 and 7) domain, which are predicted to bind to RNA (Figure 1.6) (Anantharaman et al., 2010; Callebaut and Mornon, 2010; Cui et al., 2013).

Recently, the crystal structures of the OST-HTH/LOTUS domains of Tdrd7 were shown to bind and stimulates the DEAD-box RNA helicase (Jeske et al., 2015, 2017), which function in regulating multiple aspects of RNA metabolic processes (Linder and Jankowsky, 2011). This recent finding suggests that Tdrd7 is closely associated with RNA.

Figure 1.6 A schematic representation of the TDRD7 Protein TDRD7 contains three OST-HTH/LOTUS domains and three Tudor domains. The Tudor domains are predicted to bind methylated arginine residues within other proteins. The OST-HTH/LOTUS domains are predicted to bind to RNA, and have recently shown to bind to dead-box helicase proteins.

RNA-binding proteins (RBPs) and post-transcriptional regulators that can interact with RNA to form ribonucleoprotein complexes, are involved in distinct aspects of post- transcriptional gene expression control (Anderson and Kedersha, 2009; Kishore et al., 2010). RBPs can stabilize a target mRNA, transport an mRNA to a specific subcellular localization, or channel an mRNA to degradation (Figure 1.7) (Anderson and Kedersha,

2009; Dash et al., 2016; Jiang et al., 2014). RBPs and post-transcriptional regulators can be part of cytoplasmic RNA granules that are classified based on their composition and function as: 1) stress granules, 2) processing bodies, 3) ribonucleoprotein (transport) complexes, or 4) germ cell granules such as chromatoid bodies (Anderson and Kedersha, 2006, 2009). Different RNA granules, and their associated proteins, function in executing a distinct post-transcriptional level of gene expression control on their

target mRNA to determine their fate. The direct or indirect interactions of RBPs and post-transcriptional regulators with other proteins and factors determine the specificity of the post-transcriptional control outcome (Aulas et al., 2017).

Figure 1.7 RNA Binding Proteins Function in Post-transcriptional Gene Expression Control RNA binding proteins execute distinct level of gene expression control on their mRNA targets. The level of control determines the fate of mRNA. Depending on the RNA binding protein, mRNA targets can be stabilized, transported to a specific subcellular localization or can be destined to degradation (Adapted from Dash et al. 2016)

1.6 Tdrd7 function in development Tdrd7 was first described to be expressed in rat brain and testis (Hirose et al., 2000). In this early study, Tdrd7 was found as a protein interacting partner of PCTAIRE2 in terminally differentiated neurons (Hirose et al., 2000). In addition to Tdrd7’s implication in post-transcriptional control, it has recently been shown to be an interferon stimulated gene (Subramanian et al., 2018). Tdrd7 exerts antiviral activity via inhibiting the autophagy that is required by the virus replication machinery

(Subramanian et al., 2018, 2020). It was also shown that Tdrd7 is implicated in the process of miosis (Pandey et al., 2019). Another recent study has shown that Tdrd7 is among the eye-regulatory elements where each contributes by small effect mutations collectively leading to a polygenic basis for eye degeneration in the blind mole rat

(Roscito et al., 2020). TDRD7 has been shown to be a component of ribonucleoprotein complexes and RNA granules found in the cytoplasm of differentiating sperm (Kotaja and Sassone-

Corsi, 2007; Lachke et al., 2011b; Tanaka et al., 2011a; Updike and Strome, 2010). These granules are key in developing the identity and properties of germ cells (Marcello and Singson, 2011). Chromatoid bodies are germline granules specific to male germ cells (Kotaja et al., 2006), that are expressed in late differentiating spermatocytes and are crucial for RNA processing (Hosokawa et al., 2007a; Kotaja and Sassone-Corsi,

2007). It was shown that Tdrd7 along with Tdrd1 and Tdrd6 localize to chromatoid bodies (Hosokawa et al., 2007a). Tdrd7 orchestrates a programmed remodeling of the chromatoid body by dynamically changing its composition across several developmental stages (Tanaka et al., 2011a).

Figure 1.8 Tdrd7 is expressed in the lens and its deficiency causes cataract (A) In situ hybridization indicates highly enriched expression of Tdrd7 transcripts in mouse and (B) chicken lens fiber cells. (C) Tdrd7 knockdown in chicken lenses causes cataract. Deficiency of Tdrd7 in (D) human or (E-G) mouse causes cataract and dislocation of lens mass

Unlike Tdrd1 and Tdrd9, Tdrd7 participates in maintaining the genomic integrity of male germline by suppressing LINE1 retrotransposons in a pathway that is independent of Piwi-interacting RNA biogenesis (Tanaka et al., 2011a; Wang, 2017). Mutations in Tdrd7 result in azoospermia and male sterility in human and mouse (Lachke et al., 2011a; Tanaka et al., 2011a; Tan et al., 2017). Interestingly, in addition to its role in male germ cell differentiation, Tdrd7 was the first Tudor domain protein family member, whose deficiency is linked to cataract in human, chicken and mouse (Figure 1.8) (Chen et al., 2017a; Lachke et al., 2011a; Tan et al., 2017; Tanaka et al., 2011a). Indeed, multiple distinct mutations in TDRD7 are causative of congenital cataract in human. Initially, two pediatric cataract cases were reported: 1) De novo balanced paracentric inversion of chromosome 9 that disrupts one allele of TDRD7, and 2) an in frame deletion of 3 base pairs resulting in the loss of a

highly conserved amino acid Valine at position 618. Recently, several new TDRD7 mutations were reported: 3) a novel homozygous (c.1129delG), p.(Ala377Profs*2) TDRD7 frameshift mutation which possibly results in nonsense- mediated decay of the TDRD7 mRNA was identified in a consanguineous family in Pakistan 4) Two novel loss-of-function mutations (c.324_325insA (T110Nfs*30) and (c.688_689insA) (p.Y230X) were identified in two patients from unrelated consanguineous families (Chen et al., 2017a; Tan et al., 2017). While congenital cataract was penetrant in both male and female affected individuals, only males were diagnosed with sterility (Tan et al., 2017). Furthermore, single nucleotide polymorphisms of TDRD7 have been associated with age related cataracts in Han Chinese populations (Zheng et al., 2014). Together, these findings demonstrate that TDRD7 is necessary for lens transparency in human.

To further understand the implication of Tdrd7 mutations, the following three germline Tdrd7 deletion mouse models have been studied: 1) A mouse carrying a Tdrd7 nonsense mutation c.2187C>T (Q723X) that was identified in an N-ethyl-N-nitrosurea (ENU) screen (Lachke et al., 2011a), 2) a mouse targeted Tdrd7 knockout (Tdrd7-/-) model that removes exons 8-12 (Tanaka et al., 2011a), and 3) Tdrd7-/- mouse in which exon 3 is targeted by the clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9 approach (Tan et al., 2017). All three Tdrd7 homozygous null mouse models develop both the defects observed in humans, namely they exhibit fully penetrant cataract and male-specific sterility. This indicates that Tdrd7-/- mice can be used as a model to study the human cataract defects (Lachke et al., 2011a; Tan et al., 2017).

Thus far, the characterization of Tdrd7-/- mouse lenses shows a rupture of the lens capsule and extrusion of the lens fiber cell mass to the vitreous humor by P30 (Lachke et al., 2011a). In wild-type mice, in situ hybridization and immunostaining shows that Tdrd7 transcripts and protein are robustly expressed in lens fiber cells at embryonic day (E)12.5 and later stages (Figure 1.8) (Lachke et al., 2011a). Co-staining of Tdrd7 and Pyronin Y, a dye that preferentially binds RNA, suggests that Tdrd7 colocalizes with RNA in the lens (Lachke et al., 2011a). Further, Tdrd7 variably colocalizes with processing bodies and transport ribonucleoproteins (Lachke et al., 2011a). It is also suggested that Tdrd7 affects stress granules assembly, because Tdrd7 knockdown in human lens cell lines show a significant reduction of stress granules under conditions of induced oxidative stress (Lachke et al., 2011a). Microarray analysis of Tdrd7-/- lenses as well as Tdrd7 knockdown in 21EM15 mouse lens cells, identified among other key factors, a heat shock protein encoding gene, Hspb1, to be significantly downregulated (Lachke et al., 2011a).

1.7 Loss of Tdrd7 causes fully penetrant cataracts in mouse To gain initial insight into the nature of the cataract, a detailed phenotypic characterization was carried out on Tdrd7-/- mice lens (Barnum, 2015). These mice were generated as a targeted germline Tdrd7 knockout with a deletion spanning exons

8-12, as described previously (Tanaka et al., 2011b). Light microscopy and histology revealed that the Tdrd7-/- mice exhibit profound ocular and lens defects by age 3 months

(Figure 1.9-A). While histological analyses of Tdrd7-/- mice lens did not show clear defects at embryonic (E16.5) or early postnatal (P4) stages, it did identify profound defects by P30 (1 month) (Figure 1.9-B). Interestingly, brightfield microscopy and grid imaging have identified fully penetrant cataract in Tdrd7-/- mice by P22, while there

were no apparent defects detected just four days prior (at P18) (Figure 1.9-C). This suggests that the absence of Tdrd7 leads to a gradual accumulation of pathological changes in lens cells, which manifests as cataracts within a few days (Barnum, 2015).

1.8 Tdrd7-/- mice exhibit severe lens fiber cell defects prior to overt cataract formation Scanning electron microscopy (SEM) indicated that at stage P18, the organization of lens fiber cells is abnormal in Tdrd7-/- lenses (Figure 1.10) These data suggest that although Tdrd7-/- lenses have apparently normal morphology under light microscopy at stage P18 and do not exhibit severe cataracts until stage P22, their fiber cells show early signs of severe abnormalities. These defects accumulate progressively as SEM at P28 indicates, where cell membrane protrusions of lens fiber cells were further disturbed. These data indicate that the onset of fiber cell morphology and organizational defects in Tdrd7-/- lens is initiated prior to manifestation of severe opacities that are discernable by light microscopy and anatomical phenotyping approaches, such as histology (Barnum, 2015).

In summary, the mechanism by which Tdrd7 functions in mammalian lens and how its deficiency results in cataract is not well understood. Therefore, through the detailed, cellular and molecular characterization and multi-omics data analysis of Tdrd7-/- mouse lens, and by characterization of some high priority candidates that are downstream of Tdrd7, I propose to gain new insights into the function of Tdrd7 in the normal lens and into the mechanism of cataract resulting from its deficiency.

Figure 1.9 Phenotypic characterization of Tdrd7-/- lens and eye defects. (A) Examination of ocular defects in 3-month (P90) Tdrd7-/- mouse mutant lenses by light microscopy (left panels). Control eye exhibits a normal iris and a clear lens, while a flattened iris and a smaller lens with posterior rupture and cataract is observed in Tdrd7-/- eye (right panel indicated with asterisks). Note: In the bottom panel, dissected lenses are balanced on their equator. (B) Histological characterization of control (Tdrd7+/-; left) and mutant (Tdrd7-/-; right) lenses at stages embryonic (E)16.5, and postnatal (P)4, and P30. Control lenses show no defects through P30. However, Tdrd7-/- lenses exhibit profound fully penetrant lens defects. (C) Tdrd7-/- mouse exhibit eye and lens defects by P22. Light microscopy of control (Tdrd7+/-; left) and mutant (Tdrd7-/-; right) eye and lens indicates no difference at P18. However, at P22 Tdrd7-/-

mice exhibit cataract (asterisk, right image middle panel) and a flattened iris (asterisk, left image middle panel). At P22, histological analyses also detected defects in Tdrd7-/- lens. (Adapted from Barnum, Al Saai et al., 2020)

Figure 1.10 Scanning electron microscopy (SEM) of Tdrd7-/- lenses demonstrates abnormal cortical fiber cell morphology SEM was performed to visualize cortical fiber cells for control and Tdrd7-/- lenses at P18 and P28. abnormal fiber cell morphology (asterisk) was observed in Tdrd7-/- lenses at both P18 and P28. . (Adapted from Barnum, AlSaai et al., 2020)

Chapter 2

MATERIALS AND METHODS

2.1 Animals studies

Animal studies were conducted according to the statement issued by the Association for Research in Vision and Ophthalmology (ARVO) for the recommended use of animals in ophthalmic and vision research. Experimental protocols for animal studies were approved by the University of Delaware Institutional Animal Care and Use Committee (IACUC). Tdrd7 targeted germline knockout (KO) mouse line used in this dissertation work was developed in Dr. Chuma’s laboratory in Japan and is described in detail by (Tanaka et al., 2011b). In summary, to generate Tdrd7 targeted KO mouse line,

LoxP sites were inserted in Tdrd7 gene of C57BL/6 mice upstream and downstream of exon 8 and exon 12, respectively. C57BL/6 CMV-Cre transgenic mice were used to excise LoxP sites resulting in the removal of exon 8 to exons 12. Mice were housed in a 14 hour light to 10 hour dark cycle at the university of Delaware animal care facility. Tdrd7 homozygous null (Tdrd7-/-) male mice are infertile (Tanaka et al., 2011a). To efficiently obtain Tdrd7-/- mice, Tdrd7-/- females, which are fertile, were bred to Tdrd7+/- male mice. This breeding scheme did not produce wild-type (Tdrd7+/+ ) litter-mate mice to be used as litter-mate controls in experiments. Since Tdrd7+/- mice did not exhibit an obvious eye defects or cataract, Tdrd7+/- were used as litter-mate controls in some of the data described in this disseration. To generate Tdrd7+/+ controls, Tdrd7+/- males and females were bred and the offspring with Tdrd7+/+

genotype males and females were bred to generate Tdrd7+/+ control for some of the experiment. Wild-type ICR outbred mice (Harlan Laboratories, Frederick, MD) were used for collecting lens tissue in immunoprecipitation assays (IP) and RNA- immunoprecipitation (RIP) assays Eyes dissected from C57BL/6J germline Cap2-/- mice were kindly provided by Dr. Jeffery Field’s lab from the department of Systems Pharmacology and Translational

Therapeutics at the University of Pennsylvania Perelman School of Medicine Philadelphia. The generation of Cap2-/- is described in detail by (Field et al., 2015). In summary, a construct that disrupts Cap2 gene flanked with FRT was inserted, which results in a truncated Cap2 fusion protein (Field et al., 2015).

2.2 DNA isolation from mouse tails Mouse-tail tissue genomic DNA was used for genotyping. DNA was isolated using Gentra Puregene Mouse Tail Kit (QIAGEN, Cat No. 158267). The tails were digested with 300µl cell lysis buffer and 1.5µl of Protinase K 10 mg/ml (Invitrogen,Grand Island, NewYork) at 57°C overnight. Digested tails were treated with 1.5µl RNase A solution (QIAGEN, Cat No. 158922) at 37 °C for 45 minutes . Before the protein precipitation step, digested tails were brought to room temperature for 15 minutes then digested tails were placed on ice for 1 minutes . 100 µl of protein precipitation solution (supplied in the kit) were added to the digested tails, followed by

20 seconds of vigorous vortexing then centrifugation at 130000g for 5 minutes . For DNA precipitation, the supernatant (~300µl) was transferred to new 1.7 ml Eppendorf tubes containing 300ul of absolute isopropanol. These Eppendorf tubes were gently inverted several times then centrifuged at 13,0000g for 5-15 minutes . After DNA

precipitation, the supernatant was carefully removed and residual isopropanol was allowed to air dry. To wash the precipitated DNA pellet, 300µl of 70% ethanol was added and the samples were centrifuged at 13,0000g for 5-15 minutes . The ethanol was carefully drained, and the Eppendorf tubes containing the DNA pellet were allowed to completely air dry at room temperature or at 37 °C. Dried DNA pellets were rehydrated using 20µl-50µl of DNA hydration solution and placed on a heating block at 65°C for 1 hour. Isolated DNA Samples were left on a rocker overnight for the DNA to completely dissolve. The DNA concentration was determined using NanoDrop™ One/OneC Microvolume UV-Vis Spectrophotometer (Thermo fisher scientific). Isolated DNA samples were stored at -20°C until used for PCR.

2.3 Mouse genotyping Genomic DNA isolated from mouse tail tissue was diluted into (200 ng/µl) and 1 µl of diluted isolated genomic DNA was added to 25 µl polymerase chain reaction

(PCR) in 0.2 ml tubes (VWR: Cat No.490003-692). The PCR reaction was prepared as follow (for one PCR reaction): 2.5µl of 10x CoralLoad PCR Buffer and 0.2 µl Taq DNA Polymerase from (QIAGEN: Cat No. 201205); 0.5 µl of prepared 10 mM dNTP set mix (QIAGEN: Cat No.201913); 0.5 µl of 100 μm of each forward and reverse primers; 1.25 µl of Dimethyl sulfoxide (Sigma-Aldrich Cat No. D2650); 18 µl nuclease free water.

The following are the primer sequences used for genotyping: Tdrd7-WT-g-F 5'-GAG TAA CTC TGG GCG CAG TC-3', Tdrd7-WT-g-R 5'-GCC ATA GCA ATC AGT GAG

CA-3', expected product size 250 bp; and Tdrd7-KO-g-F 5' GTC TAA CCC ATT CAG GGA TGA AGA 3', Tdrd7-KO-g-R 5' GAA TCC TCA CCA GTT AGC CTC ACC 3', expected product size 500 bp, as previously described (Tanaka et al., 2011b). The PCR reaction was prepared on ice, mixed well then centrifuged briefly before being placed

on BioRad® Thermal Cycler PCR machine. The following are cycling conditions used for Tdrd7 genotyping PCR: 95°C for 5 minutes ., 35 cycles of ( 95 °C for 30seconds, 56°C for 1 minutes ., 42°C for 30 sec.). The PCR product was stored at 4°C for short term or at -20°C for long term. Gel electrophoresis was performed using 1.5% agarose gel (The Lab Depot Cat No. TLDA1705) mixed with 15 µl of 1% Ethidium Bromide Solution (Thermo Fisher Scientific Cat No. BP1302-10). To determine the amplicon size of the PCR product 8 µl of 100 base pair ladder (Thermo Fisher Scientific Cat No.

SM0242) was loaded on the agarose gel.

2.4 Western Blot Analysis Ocular lens tissue was dissected from Tdrd7-/- and control mice. To prepare lens protein lysate, collected lens tissue were homogenized on ice in RIPA lysis buffer (48 mM Tris-HCL at pH 8.0, 150 mM NaCl, 0.1% SDS and 0.4% sodium deoxycholate), 1X phosphatase inhibitor cocktail and 1X protease cocktail (both from Thermo Fisher scientific: Cat No PI-78420 and 87785, respectively). Lysed lens tissue was centrifuged for 30 minutes at 14,000g at 4°C. The supernatant (protein lysate) was collected and its concentration was determined by Pierce BCA Protein Assay Kit following the manufacturer’s instructions (Thermo Fisher Scientific Cat No. 23228). Equal concentrations of the protein lysate for Tdrd7-/- and control lenses were denatured with

1X laemmli buffer on a heating block at 95°C for 5 minutes. Denatured protein lysates were run on a 7% SDS-polyacrylamide gel for 90 minutes at 90 volts. The proteins from the resolved 7% SDS-polyacrylamide were then transferred onto PVDF Transfer Membrane (Thermo Fisher Scientific Cat No. 88518) for 1 hour at 100 volts 7% SDS- polyacrylamide 4°C. The membrane was then blocked with 5% non-fat milk in Tris Buffered Saline with 0.1% Tween 20 (Thermo Fisher Scientific Cat No. BP337500)

(TBST) for 1 hour at room temperature, followed by overnight incubation at 4°C with primary antibody for HSPB1 (Abcam, Cat No. ab12351, used at 1:200 dilution) or for HPRT (loading control) (Abcam, Cat No. ab10479) in blocking buffer or Tdrd7 (affinity purified (Lachke et al., 2011a), used at 1:200 dilution). The membrane was then washed 3 times in TBST for 10 minutes each and probed with a secondary HRP conjugated antibody (Cell Signaling, Catalog No. 7074S, used at 1:1000 dilution) for 1 hour at room temperature, followed by washes as described above and imaging.

2.5 Cryopreservation of eye tissue and cryosectioning Eye tissue from test and control mice were enucleated and fixed for 30 minutes. on ice with 4% paraformaldehyde Aqueous Solution (PFA) (Electron Microscopy Sciences Cat No. 15710) in 1X phosphate-buffered salt solution (PBS) (Thermo Fisher Scientific Cat No. MT-20-031-CV). This was followed by a brief wash with 1X PBS and incubation in 30% D-sucrose (Thermo Fisher Scientific Cat No. BP220-10) in 1 X

PBS overnight at 4°C. The following day, eye tissue was embedded in Tissue-Tek O.C.T. Compound Medium on dry ice and stored at -80°C until needed. Cryosections (12 μm thick) were then obtained by using cryostat (Leica CM3050) and mounted on microscope charged slides (Thermo Fisher Scientific Cat No. 12-550-18). Slides were stored at -80°C until needed.

2.6 Immunofluorescence

For immunofluorescence staining, sections were incubated with blocking buffer consisting of serum in PBS or PBS with Triton X-100 (PBST) for 1 hour at room temperature. Sections were then incubated with primary antibody in the blocking buffer overnight at 4°C. The next day, sections were washed in 1X PBS or PBST three times

for 10 minutes each, at room temperature. Washed slides were then incubated with a fluorescently conjugated Alexa secondary antibody and Alexa Fluor™ conjugated Phalloidin in blocking buffer for 1 hour at room temperature. This was followed by 1X PBS or PBST washes three times for 10 minutes each and samples were mounted with VECTASHIELD Antifade Mounting Media (Vector Laboratories, Cat No. H-1000). When sections were stained for only phalloidin and wheat germ agglutinin 488 conjugate (WGA) (Thermo Fisher Scientific Cat No. W11261) sections were incubated in blocking buffer for 1 hour at room temperature. Followed by 1hour incubation at room temperature with Alexa Fluor™ conjugated Phalloidin and WGA (both used at, 1:90 dilution). This was followed by washes and mounting the slides as described above and slides were stored in -20°C until imaging. Fluorescence images were collected using a Zeiss LSM780 and LSM880 airyscan confocal microscope. Details on specific primary antibody staining conditions can be found in Table 2.1.

Table 2.1: Antibodies and immunofluorescence staining conditions

Primary Dilution Blocking buffer Washes antibody 5% horse (Sigma-Aldrich Cat no. H0146) HSPB1 5% goat serum (Jackson ImmunoResearch Cat no. Abcam 1:50 PBS 005-000-121) ab1426 PBST-0.1% TDRD7 Affinity purified 5% normal chicken serum (Abcam Cat no. ab7477) 1:100 PBS (Lachke PBST-0.1% et al., 2011a) MIP EMD 10% horse (Sigma-Aldrich Cat no. H0146) 1:100 PBS Millipore PBST-0.1% AB3071

2.7 Fluorescence in situ hybridization coupled with Immunostaining Eye lens sections from three biological replicates of Tdrd7-/- and control at stage P15 were prepared (as described in Section 2.62.5) in the cross-orientation at 10 μm thickness to visualize the hexagonal appearance of fiber cells. Slides were stored at - 80°C until needed. Sections were fixed with 4% PFA in 1X PBS, permeabilized using 70% ethanol, and hybridized with RNA specific probes. These probes (commercially called as steallaris probes), were a mix of about forty 20-nucleotide long oligomers, each having the same single fluorophore tag conjugated to its 3’end. These probes bind in a sequence-specific manner to the target RNA (in this case, mouse Hspb1) giving diffraction limited spots that can be visualized using fluorescence microscopy (Batish et al., 2011). In these assays, probes tagged with the Texas red fluorophore were used. After overnight hybridization with Hspb1-specific RNA-probes at 37°C, the sections were washed with 2x SSC buffer and then blocked using BSA. Sections were then incubated with Tdrd7 primary antibody (rabbit) (Lachke et al., 2011b) for 1 hour at 37°C. Sections were blocked again and then incubated with anti-rabbit secondary antibody tagged with Cy5. The samples were washed with 1X PBS, stained with DAPI and mounted for imaging as described (Bayer et al., 2015). Imaging was performed using a Nikon TiE inverted fluorescence microscope with 100x oil objective, equipped with a fully automated stage, a cooled CCD back illuminated PIXIS 1024B camera and a Metamorph (Biovision, Inc) image acquisition software. The acquired images were analyzed and processed using a custom written MATLAB program as described (Batish et al., 2012; Markey et al., 2014).

2.8 RNA sequencing for Postnatal stage 15 in Tdrd7-/- mouse lens Eye tissue from three biological replicates for mice at postnatal stage 15 (P15) of Tdrd7-/- and Tdrd7+/+ were collected. Lens tissue were dissected, immediately flash frozen (two lens tissues/one animal per biological replicate) placed in 1.7 ml Eppendorf RNASE free tubes on dry ice) and stored at -80°C. Total RNA was extracted from lens tissue using mirVana™ miRNA isolation kit, with phenol (Thermo fisher scientific, Cat No. AM1560) according to the manufacturer’s instructions. Total RNA samples were shipped to DNA LINK for RNA quality assessment and sequencing. All 6 samples passed the quality assessment (RIN >7, total RNA amount >1µg). Samples then were subjected to library preparation using Truseq Stranded Total RNA H/M/R Prep kit. Paired-end RNA-seq was performed on Illumina Novaseq6000 platform. The samples were sequenced at 50 million reads depth.

2.9 Analysis of RNA sequencing data Raw FASTQ files were downloaded from DNA LINK server. Trim Galore version 0.6.5 (https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/) was used for adaptor trimming and removal low quality bases for paired-end reads. Trimmed reads were mapped against Mus musculus reference genome (GRCm38) using HISAT2 version 2.2.0 (Kim et al., 2015, 2019). BAM files were then used to obtain raw counts of mapped reads with HTSeq version 0.11.2 (Anders et al., 2015) . Differential expression analysis R/Bioconductor version. 3.6.2 using the EdgeR package version

3.28.0 (Robinson et al., 2010) and p-values were adjusted for multiple comparisons using FDR method of Benjamini & Hochberg (1995).

2.10 Tandem Mass Tag (TMT) Mass spectrometry for Postnatal stage 15 in Tdrd7- /- mouse lens Eye tissue from five biological replicates for P15 mice of Tdrd7-/- and Tdrd7+/+ were collected (two lens tissues per biological replicate). Dissected lens tissue were immediately flash frozen (lens tissue placed in 1.7 ml Eppendorf RNase free tubes on dry ice) and stored at -80°C. Flash frozen lens tissue were shipped to Oregon Health & Science University for TMT Mass spectrometry and data analysis. Samples were labeled with 10-plex tandem mass tag (TMT) isobaric labeling reagents (Thermo

Fisher Scientific). The tryptic digests were fractionated to get sufficient proteome profiling depth. Data was acquired using the SPS MS3 method on the Thermo Orbitrap Fusion instrument to reduce interference and increase reporter ion dynamic range. The PAW pipeline (https://github.com/pwilmart/PAW_pipeline) was used for database searching and results processing. Typical TMT search parameters in Comet were used with a canonical mouse reference proteome (2018.01_UP000000589_10090_Mus- musculus_canonical_for. fasta). Statistical analysis for protein differential expression

(DE) was done using the Bioconductor R package edgeR. An exact test two-sample comparison was performed in Jupyter notebook.

2.11 Cluster heatmap for normal expression pattern of misregulated proteins and genes in P15 Tdrd7-/- lens Normal gene expression data was downloaded from iSyTE 2.0 for 9 developmental stages corresponding to each differentially expressed genes and proteins in Tdrd7-/- P15 data (significance <0.05, fold change > 1.2). For each data set, (the transcriptome and the proteome) a cluster heatmap clustered by rows was generated in in Jupyter notebook using python Seaborn library in Python (v2.7).

2.12 Gene Set Enrichment Analysis (GESA) Gene set enrichment analysis (GESA) was performed on the data sets obtained from Tdrd7-/- lens P15 differentially expressed gene ( Table 4.3) and differentially expressed proteins (Table 4.4). GSEA was performed in R studio using the following packages: clusterProfiler, enrichplot and org.Mm.eg.db for mouse database. The gene list was ranked based on log2 fold change and analysis for all Gene Ontology (GO) terms were considered significant at P-value<0.05.

2.13 Immunoprecipitation for Tdrd7 protein coupled with mass spectrometry Tdrd7 protein was pulled down via immunoprecipitation technique using Tdrd7 antibody, along with IgG pull-down as negative control. See (Figure 2.1) for a schematic summarizing the experiment. Pierce™ Protein G Plus Agarose beads (Thermo Fisher Scientific, Cat No. 22851) were centrifuged at 2000g for 1 minutes (supernatant discarded) and washed twice with 1X PBS. Washed beads were resuspended with 25µl of 1XPBS. To form beads-antibody conjugation complex, 10ug of Tdrd7 mouse monoclonal antibody (Milipore-Sigma, Cat No. MABE64) and 10ug of normal mouse IgG (Santa Cruz, Cat No. sc2025) were mixed with washed beads in separate 1.7ml Eppendorf tubes. The volume was brought up to 250µl with 1X PBS and beads-antibody complex mix was incubated on a tube rotator. After one hour of incubation at room temperature, excess antibody from the beads-antibody mix were washed once with 750µl of 1X PBS. To minimize background signals on downstream mass spectrometry resulting from antibody leakage during protein final elution step, the antibody-bead complex formed was crosslinked with BS3(bis(sulfosuccinimidyl)suberate) (Thermo Fisher, Cat No. A39266). This was achieved by adding freshly prepared 9µl of 2.5mM BS3 and 41µl of

1X PBS to the washed beads-antibody mix which was incubated on a tube rotator for 1 hour at room temperature. The crosslinking reaction was quenched by adding 2.5µl of 1M Tris and 15 minutes incubation. The quenched reaction was washed twice with 50mM Tris (PH=7.4). The crosslinked beads-antibody complex was washed twice with NP40 cell lysis buffer (Thermo Fisher Scientific, Cat No. FNN0021) and stored at 4°C (for 1 day) in 200µl NP40 cell lysis buffer. The next day, previously flash frozen 20 lenses collected from wild type ICR mice at postnatal age day 8 (P8) were homogenized on ice in 300µl NP40 treated with 1X phosphatase inhibitor cocktail and 1X protease cocktail (both from Thermo Fisher scientific: Cat No PI-78420 and 87785 respectively). Lysed lens tissue was centrifuged for 30 minutes at 14,000g at 4°C. The supernatant (Protein lysate) was collected and its concentration was determined by NanoDrop™ One/OneC Microvolume UV-Vis Spectrophotometer (Thermo fisher scientific). Equal quantities from the protein lysate (2000ug) were used for the immunoprecipitation reaction. The pre-conjugated beads-antibody crosslinked mix, previously prepared, was washed once with NP40 containing 1X phosphatase inhibitor cocktail and 1X protease cocktail. Then protein lysate and 250µl of NP40 containing 1X phosphatase inhibitor cocktail and 1X protease cocktail were added to each pull-down reaction. Tubes containing the pull-down reaction were placed on a tube rotator for 1 hour at room temperature. The pull-down reactions were then centrifuged at 2500g for 2 minutes and the supernatant was discarded. Each pull-down reaction was washed with 500µl of NP40 containing 1X phosphatase inhibitor cocktail and 1X protease cocktail. After discarding the supernatant, the pull-down reaction was eluted with 2X laemmli buffer at 95°C for 5 minutes. ~25ul of the eluted pull-down reaction was transferred into a fresh 1.7ml Eppendorf tubes and stored at -80°C. Part of the stored pull-down was used

for western blot analysis and the rest was sent for mass spectrometry facility at the University of Delaware.

Figure 2.1 Schematic for Immunoprecipitation assay

2.14 RNA immunoprecipitation assay for Tdrd7 protein

Postnatal day 15 (P15) wild-type mouse ICR lens tissue was used (10 individual lens tissues per biological replicates) in three biological replicates. Lens tissue was freshly dissected on ice and used with the 17-700 | Magna RIP™ RNA-Binding Protein Immunoprecipitation Kit (EMD Millipore) according to the manufacturer’s protocol.

This was followed by semiquantitative PCR for mRNA targets as described in section (2.16).

2.15 RNA Isolation Dissected lens tissue from mouse eyes from 3 biological replicates of Tdrd7-/- and control were collected, flash frozen and stored at -80°C. Total RNA extraction was performed using RNeasy Mini Kit (Qiagen, Cat No. 74104) according to the manufacture’s instruction. In summary, frozen lens tissue was homogenized in 350µl RLT (supplied with the kit) on ice. Lens tissue lysate was then transferred to RNeasy spin column with equal volume of 70% ethanol and centrifuged for 15 seconds at 8000g. To wash the spin membrane, after discarding the flow through, 700µl of RW1 buffer was added to RNeasy spin column and centrifuged for 15 seconds at 8000g. To ensure that no residual ethanol is carried over during the final RNA elution step, two rounds of centrifugation with 500µl of RPE buffer at 8000g were performed for 15 seconds and two minutes, respectively. RNeasy spin column was centrifuged again at full speed with new 2ml collection tube for 1 minute. Finally, RNA was eluted from RNeasy spin column with 25µl RNase-free water and centrifugation at 8000g. RNA concentration was determined by NanoDrop™ One/OneC Microvolume UV-Vis Spectrophotometer (Thermo fisher scientific). The RNA was then aliquoted and stored at -80°C.

2.16 Reverse Transcriptase for Quantitative PCR

Complementary DNA (cDNA) was synthesized by reverse transcriptase reaction from 1μl of (250ng/μl) total RNA that was extracted from mouse lens tissue using iScript cDNA synthesis kit (Bio-Rad, Cat No. 1708890EDU) according to manufacturer’s instruction. The concentration of the cDNA was determined and

subsequently used to perform quantitative PCR on QuantStudio 6 (Thermo Fisher Scientific) or semi quantitative PCR on BioRad® Thermal Cycler PCR machine. For quantitative PCR, 1μl of (70ng/μl) of cDNA was added to 10μl Power Syber Green PCR master mix (Life technologies, Cat No. 4367659), 0.5μl each from 10μM of forward and reverse primers and 8μl of nuclease free water. The reaction mix were prepared in 96 MicroAmp® Optical 96-Well Reaction Plate (Life technologies, Cat No. N8010560) on ice block and the plate was briefly pulsed before loading onto the QuantStudio 6. All quantitative PCR reactions were performed in 3 biological replicates and 3 technical replicate per sample. An appropriate housekeeping gene was used with all quantitative PCR assays to normalize mRNA levels between different samples. The differential expression was determined using ΔΔCT method on Microsoft Excel and statistical significance was determined by using t-test on JMP pro14 software. For semi quantitative PCR, cDNA was added to a reaction mix consisting of : 2.5µl of 10x CoralLoad PCR Buffer and 0.2µl Taq DNA Polymerase from (QIAGEN: Cat

No. 201205); 0.5µl of prepared10mM dNTP set mix (QIAGEN: Cat No.201913); 0.5µl from 10μm of each forward and reverse primers; and nuclease free water. The PCR reaction mix was prepared on ice, mixed well by gently pipetting up and down then briefly centrifuged before placing them on BioRad® Thermal Cycler PCR machine.

2.17 Scanning electron microscopy for Cap2-/- mouse lens Scanning electron microscopy was performed as described (Duncan et al., 2000;

Agrawal et al., 2015). Eye tissue from 3 biological replicates of Cap2-/- and Cap2+/+ were used. Eyes were enucleated and incubated in Kimble 20 mL Glass Screw-Thread Scintillation Vials (Thermo fisher scientific, Cat No. 03-340-4C) with a fixative solution at room temperature for 5 hours. The fixative solution contained: 0.08M sodium

cacodylate, 1.25% glutaraldehyde, 1% PFA. The reagents used above were obtained from Electron Microscopy Sciences, with the following Catalog no. 11652, 16100 and 15710, respectively. Lens tissue was dissected and returned to the fixative solution for 48 hours. Fixation was followed with 1X PBS wash and careful lens peeling with forceps to reveal different depths of lens fiber cells. Peeled lenses were then dehydrated in a series of ethanol dilutions (25%, 50%, 70%) with exposure to each dilution for 10 minutes. Lens tissue was incubated twice for 2.5 hours in 100% ethanol then dried in gradient dilution of ethanol and hexamethyldisilazane reagent (HMDS) (Electron Microscopy Sciences Cat No. 999-97-3) as follow: one hour in 1 part HMDS to 2 parts of absolute ethanol, another 1 hour in 2 part HMDS to 1 part of absolute ethanol ,and two rounds of 30 minutes in HMDS only. Lens tissue was placed onto filter paper in a 12-well plate container and left in a vacuum desiccator at room temperature overnight. Dried lens tissue were mounted on aluminum stubs, covered with carbon adhesive tabs, and painted with silver conductive paint. The above three items were from Electron

Microscopy Sciences, with the following Catalog no. 75620, 76762-01 and 12630, respectively. Lens mounting was followed by sputter coating with gold/palladium ,and Scanning electron microscopy images were obtained by Hitachi S-4700 (Tokyo, Japan) Field-Emission Scanning Electron Microscope.

2.18 Correlation analysis for Tdrd7 and downstream misexpressed genes Lens expression data for analyzed genes in normal lens development were downloaded from iSyTE database (Kakrana et al., 2018; Lachke et al., 2011b). The heatmap showing Pearson correlation for genes expression across multiple stages in normal lens development was generated using Seaborn library in Python (v2.7).

2.19 Immunofluorescence Quantification analysis Image J was used to Measure F-actin signal across fiber cell meridonial rows of Cap2-/- lens tissue and control (Figure 5.10). Images obtained from ZEN software were initially split into individual channels then was followed by the removal of pseudocolor from each channel. All the images were analyzed in grey scale. The membrane staining with WGA was used to measure the length between distant vertices as well as the length between the middle point of the broad sides within lens fiber cell hexagons. I used line scan which provides the length in pixels, then I converted the pixels into micrometers. Both measurements were performed on 50 cells in the mid-mature fiber cell zone and 50 cells in the mature fiber cell zone. JPMpro14 software was used to perform t-test for statistical significance and the results were displayed in a violin plot using JMP software. Image J was used to measure the uniformity of lens fiber cell organization in Cap2-/-. For this purpose, a line scan of a fixed length was horizontally placed across a row of lens fiber cells with phalloidin. This was performed for the mid-mature fiber cell zone and in the mature fiber cell zone. The line scan generates an arbitrary number that reflects the intensity of the staining. Results from the line scan were displayed in a line plot using JPMpro14 software.

Chapter 3

TDRD7 CONTROLS THE LEVEL OF HEAT SHOCK PROTEIN HSPB1 (HSP27) AND LENS FIBER CELL MORPHOLOGY

3.1 Introduction

Mutations or deficiency of Tdrd7, encoding an RNA granule component, cause cataracts in human, mouse, and chicken. Beside indicating the importance of Tdrd7 in lens development (Chen et al., 2017b; Dash et al., 2015, 2016; Lachke and Maas, 2011; Lachke et al., 2011b; Siddam et al., 2018; Tan et al., 2017), these data also suggested that post-transcriptional control of gene expression is critical in vertebrate lens development (section 1.5 and 1.6). Thus, characterization of the cataract phenotype in Tdrd7-/- mice would provide critical insights into the pathobiology of this human developmental defect (Lachke et al., 2011b; Tan et al., 2019; Tanaka et al., 2011b). Initial characterization of Tdrd7-/- mouse lens indicated the presence of lens fiber cells disorganization (Section 1.8) (Barnum, 2015). Analysis of postnatal stage P4 of Tdrd7-/- lens, which is 14 days prior to the earliest stage that lens cellular morphological changes can be detected by SEM in Tdrd7-/- mice, was selected for detailed molecular characterization. Unbiased omics level approach of RNA-Seq followed by bioinformatics analysis (Figure 3.1) in conjunction with Fluorescence two-dimensional difference in gel electrophoresis (2D- DIGE) 2D-DIGE analysis (Figure 3.2) led to the identification of Hspb1 as a high- priority Tdrd7 candidate in the lens (Barnum, Al Saai et al., 2020).

Figure 3.1 iSyTE-based analysis of Tdrd7-/- lens RNA-Seq data identifies Hspb1 among high-priority candidates. (B) An integrated analysis using iSyTE lens expression and enrichment data along with functional annotation and GO enrichment of significant (FC ±1.5, P-value ≤ 0.05) RNA-seq DEGs. On the left, GO categories are indicated for DEGs. The heatmap diagram is expression and enrichment data for DEGs from iSyTE-microarray data across embryonic and postnatal lens-development stages (E10.5, E12.5, E16.5 and P0). Tdrd7-/- lens DEGs enrichment and expression are represented in the heatmap using ‘red-white-green’ and ‘red-white-white’ color gradients, respectively. Lens expression of DEGs postnatal stage P0 over that at early embryonic stage E10.5 (expression in late vs. early lens) is represented using ‘blue-white-yellow’ color gradient. (Adapted from Barnum, Al Saai et al., 2020)

Figure 3.2 2D-DIGE proteome screen identifies HSPB1 protein to be down- regulated in Tdrd7-/- lens

2D-DIGE detected a single protein to be significantly reduced in Tdrd7- /- lens compared to control. (A, B) 2D DIGE analysis showed that protein coinciding with “spot 951” was down-regulated in Tdrd7-/- lens. (C) Significantly reduced abundance of spot 951 (which was identified as HSPB1 protein) in P15 Tdrd7-/- lens (asterisk represents P-value of 0.039). (D) Spot 951 and a reference spot (which was identified as CRYBA1) was picked up for LC-MS/MS-based identification. (E) LC- MS/MS identified spot 951 as mouse HSPB1 with a 100% Protein CI and 100% Best Ion CI, while the reference spot was identified as CRYBA1. (Adapted from Barnum, Al Saai et al., 2020) Hspb1, also called Hsp27, is closely related in structure to α-crystallins, and it belongs to the small heat shock protein family (Taylor and Benjamin, 2005). These proteins are known for their small molecular weight and are characterized by conserved

α-crystallin domain at the C-terminal, which promotes small heat shock proteins assembly by oligomer/hetero-oligomer formation (Baranova et al., 2011; Haslbeck et al., 2005; Mearow, 2013; Weeks et al., 2018). Hspb1 functions as an ATP-independent chaperone (Jakob et al., 1993). It can bind to unfolded proteins and prevent their (Agrawal et al., 2015)until they get properly refolded by the ATP-dependent heat shock proteins (Haslbeck et al., 2005). Hspb1 also participates in cell differentiation, and has anti-apoptotic and anti-oxidant functions, and has been shown to interact with F-actin

(Hoffman et al., 2017; Mearow, 2013). Under conditions of stress, it has been demonstrated that Hspb1 protein, in both, its phosphorylated and non-phosphorylated forms, complexes with F-actin (Hoffman et al., 2017; Mearow, 2013). In contrast to other heat shock proteins, Hspb1 was shown to be steadily present in all regions fetal and aging human lens (Bagchi et al., 2002). Mice deficient of MIP (Major Intrinsic Protein) , which is critical for lens fiber cell architecture and is linked to inherited and age related cataract in human, were shown to have significant reduction in Hspb1(Zhou et al., 2016). Mutations in Hspb1 in human is linked to a neuromuscular disease in human known as Charcot-Marie-Tooth disease. Additionally, Hspb1 was shown to have the capacity to effectively protect human lens epithelial cells against ultraviolet B induced apoptosis (Liu et al., 2014). Due to its involvement with many cellular pathways, understanding the mechanism of Hspb1 function in the Tdrd7-/- lens is important.

In this chapter I will investigate the cellular basis of the morphological abnormalities that contribute to cataract development in Tdrd7-/- mice. I will also

investigate Hspb1 in the lens to gain insights into the relationship between Tdrd7 and Hspb1 in the lens.

3.2 Results

3.2.1 Tdrd7-/- mice exhibit morphological defects in lens fiber cells that have undergone nuclear degradation To follow up on the morphological defects initially identified by SEM, I characterized lens fiber cell cytoskeleton in Tdrd7-/- lens in collaboration with Drs. Velia Fowler and Catherine Chen at the University of Delaware and Indiana University, respectively. To allow for examination of fiber cell hexagonal architecture and packing at different stages of differentiation and maturation, I obtained Tdrd7-/- and control lenses in equatorial cross sections. Such sections allow the simultaneous visualization of fiber cells in different stages of differentiation and maturation. Thus, young fiber cells (closest to the epithelial cells) to mature fiber cells which are located beyond the organelle degradation and are marked by nuclear degradation (see Figure 1.3 for schematic illustration of lens cross section) are discernable by this approach. Phalloidin staining for F-actin indicated abnormalities in the hexagonal cell morphology and packing in Tdrd7-/- lenses when compared to control lenses that exhibit normal hexagonal cellular morphology. Abnormal hexagonal fiber cells were specifically observed in Tdrd7-/- lens fiber cells that have undergone or are undergoing nuclear degradation and are entering late stages of maturation (i.e. organelle free zone, OFZ) (Figure 3.3). In normal lens development, unlike other cytoskeleton components which are degraded in mature fiber cells, F-actin is found to be associated with fiber cell membranes beyond the organelle free zone (Fudge et al., 2011; Kuwabara, 1968; Lee et al., 2000). This suggests that normal F-

actin distribution is key for the stability and integrity of mature lens fiber cells (Cheng et al., 2017), which is found to be disrupted in Tdrd7-/- lens maturing fiber cells. In addition to Phalloidin staining for F-actin, I stained fiber cell membranes with wheat germ agglutinin (WGA), which is a lectin that binds to sialic acid and N-acetyl- D-glucosaminyl residues on cell membranes (Kistler et al., 1986). WGA staining showed that Tdrd7-/- lenses exhibited abnormal lens fiber cell membrane morphology that was also apparent at the beginning of the nuclear degradation zone (Figure 3.4).

This indicates that membrane morphological defects in lens fiber cell contributes to the abnormal appearance of F-actin staining in Tdrd7-/- lenses. Together, these data indicate that Tdrd7 is needed to maintain proper morphology and regular hexagonal packing pattern in mature lens fiber cells that are undergoing nuclear degradation.

Figure 3.3 Abnormal F-actin staining in maturing lens fiber cells in Tdrd7-/- mice. (A) P15 lens were stained with phalloidin which specifically binds to F- actin. Abnormal fiber cell morphology and F-actin distribution are observed. The asterisk indicates the zone from where onward the defects are visualized. Notice that there is no difference between Tdrd7-/- and control in early differentiating fiber cells until the asterisk. (B) Arrowheads indicates the begging of the organelle free zone in the lens. Distorted fiber cell morphology and abnormal F-actin distribution starts sharply at the begging of the organelle free zone. (Adapted from Barnum, Al Saai et al., 2020)

Figure 3.4 WGA staining indicates membrane abnormalities in maturing Tdrd7-/- lens fiber cells WGA staining demonstrates abnormal membrane appearance specifically in fiber cells with nuclear degradation in Tdrd7-/- lenses. Control and Tdrd7-/- lens sections at mouse stage P15 were stained with wheat germ agglutinin (WGA) and phalloidin to visualize cellular membranes and F-actin, respectively. Hoechst stain was used to visualize DNA. Images from comparable areas representative of early to late fiber differentiation and maturation (left to right) are shown for both control and Tdrd7-/- lenses. While there is no discernable difference in young differentiation and maturing lens fiber cells at the lens periphery (left-most panel and left-half of the second panel) between control and Tdrd7-/- lens, fiber cells showed abnormal WGA staining and abnormal fiber cell morphology in the Tdrd7-/- lens, coinciding with nuclear degradation (arrowheads), compared to control (second panel). (Adapted from Barnum, Al Saai et al., 2020)

3.2.2 Hspb1 is a top priority candidate down-regulated gene in Tdrd7-/- lens Based on iSyTE expression data, initial analysis of high-priority candidates of DEGs in Tdrd7-/- lens indicated that the expression dynamics of Hspb1 as the lens transits from embryonic to postnatal stages (i.e. E10.5 and P0) is comparable to that of Tdrd7. To extend this analysis, expression of the high-priority candidate genes across four stages in normal lens development were obtained from iSyTE to performed Pearson correlation. Pearson correlation analysis showed that Hspb1 was among the top two genes with a high positive correlation with Tdrd7 expression in normal progression of lens development (Figure 3.5-A). Next, I performed RT-qPCR analyses and validated the observation that Hspb1 mRNA is significantly reduced in Tdrd7-/- lenses (Figure 3.5-B). Additionally, I have validated the reduction in Hspb1 protein in Tdrd7-/- lens, that was independently in 2D-DIGE screen, by Western blotting (Figure 3.5-C). Together these analyses identifies Hspb1 as a top high-priority candidate gene that is likely relevant to explain the manifestation of the cellular and cataract defects in Tdrd7-

/- lens.

Figure 3.5 Hspb1 is a top candidate gene among the differentially expressed genes in the Tdrd7-/- lens A) Pearson correlation heat-map of candidate genes based on iSyTE wild-type lens expression data across the four developmental stages. Results indicates Hspb1 to have very strong expression Pearson correlation with a value of 1 (rounded to the next decimal), bringing Hspb1 as a top candidate gene with as it met the following criteria: expression score in normal lens >2000, enrichment score in normal lens >4.0, differential expression in Tdrd7-/- lens >1.5-fold change, P-value < 0.05). (B) RT-qPCR analysis of Hspb1 validates its down-regulation in Tdrd7-/- lenses at stages P4 and P15. Asterisk indicates P-value < 0.01. (C) Western blot confirmed the reduction of HSPB1 protein in Tdrd7-/- lenses compared to control. (D) TDRD7 protein associates with Hspb1 mRNA in the lens. RNA-immunoprecipitation (RIP) analysis using Tdrd7 antibody was performed for wild-type mouse lenses at stage P15.

3.2.3 Hspb1 is expressed in normal lens development and is reduced in Tdrd7-/- lens Hspb1 was identified to be reduced through two high-throughput analytical screens (RNA-seq combined with downstream analyses with iSyTE; 2D-DIGE combined with mass spectrometric identification of protein) on both the gene and protein level in Tdrd7-/- lenses. Although RT-qPCR and Western blotting indicated reduced Hspb1 RNA and protein levels, respectively, in Tdrd7-/- whole lenses (Figure 3.5-D), these assays do not provide spatiotemporal information on its distribution in the lens. Further, there is no systematic time-course data on Hspb1 expression in the lens. Additionally, Hspb1 is known to interact with F-actin which may aid the explanation of observed lens fiber cell morphological defects in Tdrd7-/- lenses. Thus, my next focus was to characterize Hspb1 expression and its spatiotemporal distribution with respect to lens development, fiber cell differentiation and maturation. First, I investigated Hspb1 protein expression in normal lens development by immunostaining of lens sections from E12.5 to P25. Immunostaining shows that at E12.5, Hspb1 staining is detected weakly in the primary differentiating fiber cells, while at E14.5, Hspb1 staining increases beginning with secondary fiber cell differentiation (Figure 3.6). Elevated Hspb1 staining at P0 appears to be in the fiber cells in the center of the lens, possibly indicating relatively increased protein expression in late fetal lens fiber cells, which form the centrally located fiber cells of the lens. However, by P15, Hspb1 staining is present fairly uniformly throughout the lens, both in the differentiating fiber cells at the lens periphery and in the inner, more mature fiber cells after they lose their nuclei (Figure 3.6). Analysis of P15 lens cross sections at higher magnification indicates elevated that Hspb1 protein staining in the inner cortical fiber cells where it is associated with fiber cell membranes, and remains associated with the fiber cell membranes even after they

have lost their nuclei (Figure 3.6). However, in Tdrd7-/- lens cross sections, Hspb1 protein staining appears reduced in the mature lens fiber cells after they have lost their nuclei (Figure 3.7).

Figure 3.6 Hspb1 protein expression in normal mouse lens development Immunostaining with HSPB1 antibody was performed on sagittal sections of wild-type mouse embryonic lenses E12.5 and E14.5 and postnatal lenses at P0, P2 and P15. Abbreviations: e, anterior epithelium of the lens; f, fiber cells. The scale bar is 20 m

Figure 3.7 Hspb1 protein is abnormally reduced in Tdrd7-/- lens fiber cells Immunostaining of Hspb1 in Cross-sections of control and Tdrd7-/- P15 lenses. Note the reduction in Hspb1 protein staining in Tdrd7-/- fiber cells that have lost their nuclei (indicated by asterisk). DNA (blue) is stained by DAPI. Scale bar indicates 20 m. Abbreviations: e, anterior epithelium of the lens; f, fiber cells.

3.2.4 Tdrd7 protein associates with Hspb1 mRNA Tdrd7 is known to closely associate with RNA in ribonucleoprotein complexes such as RNA granules, and was shown to closely associate with RNA (Hosokawa et al., 2007b; Jeske et al., 2017; Lachke et al., 2011b; Tanaka et al., 2011b). Next, I wanted to test the hypothesis that there is a potential close association between Tdrd7 protein and Hspb1 mRNA in a ribonucleoprotein complex in normal lens. Therefore, I first took a biochemical approach and performed RNA Immunoprecipitation (RIP) assays on wild- type ICR mouse lens lysates at P15 using Tdrd7-specific antibody. Results from the RIP assays showed a significant enrichment of Hspb1 mRNA in Tdrd7-RIP pulldown but not in the IgG-only negative control (Figure 3.5-D). Result from this analyses suggest that Tdrd7 is associated with Hspb1 mRNA in the lens. This is a novel finding as prior

experiments suggesting Tdrd7 protein-Hspb1 mRNA association were performed in cell lines and not in wild-type lenses.

3.2.5 Single molecule fluorescence in situ hybridization shows that Tdrd7 protein co-localizes with Hspb1 mRNA in differentiating lens fiber cells The RIP assay provides biochemical evidence supporting a close association of Tdrd7 protein with Hspb1 mRNA in the lens. However, it does not inform us on whether these associations occur in vivo and on their spatiotemporal location in the lens.

Therefore, I sought to examine the relevance of these biochemical protein-RNA interactions in the context of lens fiber cells. In collaboration with Dr. Mona Batish, University of Delaware, I used state-of-the-art approach to single-molecule RNA imaging, termed smFISH (single molecule fluorescence in situ hybridization) (Batish et al., 2011; Bayer et al., 2015), coupled with immunostaining with Tdrd7-specific antibody to image these interactions at single molecule resolution. This technique allows a direct visualization of association of Tdrd7 protein with its potential target mRNA, in this case Hspb1. Usually for smFISH a probe that is designed for one target contains 48 oligos while excellent signal can be achieved with only 30 per target. The probes are also designed to bind in close proximity to each other along the target mRNA length. Thus binding of large number of probes highly improves the fluorescent diffraction spot. For this analysis, I used wild-type C57BL/6 mouse P15 lens sections. Results demonstrated that Tdrd7 protein co-localized with Hspb1 mRNA in differentiating lens fiber cells (Figure 3.8). Image analysis for co-localization between protein and RNA using custom written programs in MATLAB (Batish et al., 2012; Markey et al., 2014) indicated that about 10% of the TDRD7 protein co-localizes with Hspb1 mRNA in the examined area of late differentiating fiber cells prior to nuclear

degradation (Figure 3.8). These data indicate that Tdrd7-ribonucleoprotein complexes involving specific target mRNAs, such as Hspb1, occur in specific spatiotemporal locations in fiber cells and may be necessary for maintaining optimal mRNA levels to thereby achieve optimal protein levels in the lens fiber cells primed for nuclear degradation. Optimal Hspb1 may in turn be important for interaction with F-actin, which is necessary for maintenance of normal cellular morphology in lens fiber cells that are in late maturation stages (i.e. post-nuclear degradation).

Figure 3.8 Single RNA fluorescence in situ hybridization (smFISH) coupled with immunostaining demonstrates close association of TDRD7 protein-Hspb1 mRNA. smFISH assays indicates that Tdrd7 protein to co-localize with Hspb1 mRNA in lens fiber cells. (A) Wild-type C57BL/6 mouse lens at stage P15 was stained with complementary RNA-probes specific to Hspb1 mRNA (green), (B) along with immunostaining for TDRD7 protein (red). (C) Merged image of the co-staining (yellow) and (D) analysis of significant co-localization of Hspb1 mRNA and TDRD7 protein using custom written MATLAB program as described in Methods (colored open circles). (A’-D’’) shows zoom-in of regions indicated by broken- line boxes in A-D. Arrows indicate co-localizing elements (yellow circles) scored by the MATLAB analysis.

3.3 Discussion In the lens tissue, cells of the lens epithelium terminally differentiate into lens fiber cells (Cvekl and Zhang, 2017). During differentiation, fiber cells undergo organelle and nuclear degradation which ensures a light scatter-free zone with a high refractive index around the visual axis of the lens, rendering the lens its transparency (Bassnett et al., 2011b; Wride, 2011). Fiber cells that are located in the “organelle free zone” of the lens are transcriptionally and translationally “silent” (Bassnett, 2009b;

Shestopalov and Bassnett, 1999). These transcriptionally silent lens fiber cells, which have lost their nuclei and organelles, can be considered to resemble stress-like condition. Thus, post-transcriptional gene expression control mechanisms may be necessary to ensure proper expression of the proteome of maturing fiber cells. Indeed, recent data indicates that several post-transcriptional regulatory factors are important for controlling distinct regulatory events in fiber cell differentiation, and the mis-regulation of these factors is implicated in the development of lens defects and cataract (Dash et al., 2015, 2016; Lachke and Maas, 2011; Lachke et al., 2011b; Li and Piatigorsky, 2009; Shaham et al., 2013; Siddam et al., 2018; Wolf et al., 2013). Previous work shows that Hspb1 protein is highly expressed in all human tissue (Mymrikov et al., 2011; Taylor and Benjamin, 2005). In the past, Hspb1 was shown to be involved in cell differentiation, anti-apoptotic and anti-oxidative activities, and the regulation of cytoskeleton under stress conditions, and is linked to various pathological conditions (Arrigo, 2017; Hoffman et al., 2017; Mearow, 2013). However, Hspb1’s significance to eye and lens development was not examined. My work suggests that Tdrd7, a Tudor-family protein and a posttranscriptional regulator, has a novel function for regulating fiber cell morphology, specifically in cells post-nuclear degradation. The data also indicated that Tdrd7 is upstream of Hspb1. I

propose a model (Figure 3.9) wherein fiber cells that have undergone organelle degradation⎯ and can be considered to face cellular challenges resembling a stress-like condition need Tdrd7 to sustain optimally high levels of the chaperone protein Hspb1, which may in turn function in maintaining F-actin cytoskeleton and cellular morphology. Together these findings serve to expand the role of Tdrd7-family proteins in cellular differentiation during organogenesis.

Figure 3.9 Models for Tdrd7 function in lens development and cataractogenesis

The following models are proposed to explain the temporal, the cellular and molecular basis of the lens defects and cataract observed in Tdrd7- /- mice. (A) Overlay of temporal aspects of Tdrd7-/- lens defects, at a molecular and phenotypic level, corresponding to wild-type expression of the Tdrd7 target mRNA Hspb1, a heat shock protein associated with F-actin under stress conditions, in normal mouse lens development. iSyTE microarray-based expression analysis of wild-type lenses shows that mRNA expression of Hspb1 up-regulates before slight downregulation, but yet maintaining relatively high levels in postnatal wild- type lens development. Interestingly, Tdrd7-/- lenses show reduced levels of Hspb1 mRNA at early post-natal stages (RNA-Seq, RT-qPCR data), several days prior to morphological detection of the lens cellular defects. Furthermore, Hspb1 protein levels remain low in Tdrd7-/- lenses in later post-natal stages at P15 (proteome, Western, immunofluorescence data), just prior to the initial detection of the defect at the cellular level (SEM at P18) or the gross morphological level (overt cataract at P22). The timing of expression of the lens defects in Tdrd7- /- mice coincides with the timing of normal up-regulation of Hspb1 in wild-type post-natal lens development, and therefore may reflect disruption of the expression dynamics of this F-actin associated protein upon Tdrd7 deficiency. (B) Based on current molecular and phenotypic data, including the critical observation that Tdrd7-/- mice show abnormal cellular morphology particularly in fiber cells after nuclear degradation, the following model is proposed. The association of TDRD7 protein with Hspb1 mRNA ensures its elevated levels, which translate into high levels of Hspb1protein necessary for F-actin stability in cells subjected to stress. Tdrd7 based elevated levels of Hspb1 protein may function in the maintenance of normal F-actin distribution and cellular morphology in late maturation stage fiber cells, which can be considered a stress-like condition as they undergo nuclear degradation. Source: Barnum, Al Saai et al. 2020.

Chapter 4

MULTI-OMICS INVESTIGATION OF TDRD7 IN MOUSE LENS

4.1 Introduction The unbiased detailed molecular characterization of Tdrd7-/- mouse lens at postnatal stage P4, described in Chapter 3, led to the identification of Hspb1 as a high- priority Tdrd7 downstream target candidate in the lens (Barnum, Al Saai et al., 2020). While P4 Tdrd7-/- mouse lens transcriptomic data identified other misregulated key genes besides Hspb1, the Tdrd7-/- mouse lens 2D-DIGE proteome data was not informative beyond identifying Hspb1, likely due to the high abundance of crystallin proteins, which has been identified as a challenge in such analyses (Elsobky et al., 2014; Semba et al., 2013). Further, the control used in this P4 analyses were

Tdrd7+/- lenses that while non-cataractous, were considered to (1) likely have subtle defects and (2) likely not comprehensively informative over a wild-type control. Thus, the goal of my work in this chapter was to address these concerns and gain greater insights into the progression of the molecular changes that build up and lead to the onset of cataract in Tdrd7-/- mouse. Thus, I have chosen to investigate postnatal day

15 (P15), where Tdrd7-/- lens is apparently normal, i.e. non-cataractous, under light microscope, but exhibits lens fiber cell morphological defects detected by phalloidin staining and confocal microscopy imaging as described in Chapter 3 (3.2.1). Due to the low correlation between the transcriptome and proteome in a given tissue, studying either one alone may not yield a comprehensive understanding of the molecular changes associated with the underlying defects (Haider and Pal, 2013). For

all the above reasons, I have generated high throughput RNA-sequencing data and TMT- Mass spectrometry proteomic data for Tdrd7-/- mouse lens at stage P15 (Methods described in (section 2.10). Thus, it is expected that an integrated analysis of the transcriptomic and proteomic data performed in this chapter will provide new molecular insights into the pathobiology of Tdrd7-/- lens.

4.2 Results

4.2.1 Analysis workflow I have designed a systematic workflow to process both data sets (transcriptome and proteome) with the aim of identifying high-priority candidates among the number of misexpressed genes and/or proteins in the Tdrd7-/- mouse lens. This workflow will also allow integrated analysis of both data sets for downstream analysis. The workflow consists of sequential filters to identify biologically significant candidates (Figure 4.1). The workflow starts with identification of all significantly misexpressed genes/proteins, followed by a fold change filter with a cut off > 1.2 absolute fold-change. This is followed by prioritizing candidates of potential significance in lens biology, which was achieved based on microarray expression and enriched expression data on normal lens in the iSyTE 2.0 resource as described by (Kakrana et al., 2018). This strategy included candidates that have lens enrichment fold change that is >1.5 in at least one stage among the nine stages considered for this analysis. The nine stages spanned early to late embryonic development and early postnatal to adult stages (E10.5, E11.5, E12.5, E16.5, E17.5, E19.5, P0, P2 and P56). Candidates that passed these filters (Table 4.3 , list of filtered genes and Table 4.4 list of filtered proteins) were subjected to global Pearson correlation analysis to identify candidates that have a pattern of expression that is similar

to Tdrd7. This analysis is expected to uncover candidate clusters that correlate with each other. Additionally, candidates identified from the fold change filter, at the first step before passing the data through iSyTE filter, can be subjected to different downstream analysis such as gene ontology (GO) enrichment analysis or can be used to build a Tdrd7 regulatory co-expression network for differentially expressed candidates.

Figure 4.1 Analysis work flow for Tdrd7-/- lens proteome and transcriptome This is the workflow that was used for both Tdrd7-/- transcriptome and proteome data sets. First, differential expression analysis identified 1033 genes and 234 proteins that are significantly misexpressed. After applying a fold change filter with a cut off > 1.2, each list was refined into 167 genes and 111 proteins. Data obtained until this point can be used for downstream analysis such as enrichment analysis. Further, upon examining each list for lens enrichment fold change that is >1.5 in at least one stage among the nine stages the list of genes was reduced to 82 while the list of proteins was reduced to 69 proteins. Both lists obtained retaining lens enriched genes and proteins were used for Pearson

correlation analysis. Abbreviations: DEG, Differentially Expressed Genes. DEP, Differentially Expressed Proteins

4.2.2 Tdrd7-/- mouse lens RNA/proteome profiling at P15 identifies misexpressed genes and proteins Differential expression analysis of Tdrd7-/- mouse lens RNA-seq data has identified 1033 significantly (FDR < 0.05) misexpressed genes (Figure 4.2) (A list for the1033 genes can be found in Error! Reference source not found.). Among these candidates, 635 were downregulated while 398 genes were upregulated in the Tdrd7-/- lens. On the other hand, differential expression analysis of the proteome data identified 234 significantly (P-value < 0.05) misexpressed proteins in the Tdrd7-/- mouse lens (A list for the 234 proteins can be found in Appendix Table 1 ). Among those misexpressed proteins, 116 proteins were downregulated, and 118 proteins were upregulated in the Tdrd7-/- lens. Candidates identified at this point from both data sets vary in their fold change values, although their misregulation is significant but the fold change if very small. These candidates will be further filtered as described above by setting a cut of for the fold change (cut off > 1.2 absolute fold-change).

4.2.3 Identification of Tdrd7 downstream non-coding genes I applied the Ensemble database in the analysis of RNA-seq data that classifies genes and transcripts into different biotypes. Within the Tdrd7-/- lens transcriptome, transcripts belonging to multiple biotype classes were found among the significantly misexpressed candidates. While 986 (~95%) of the misexpressed transcripts belong to protein coding genes class, the remainder 47 (~5%) were among long non coding RNA or small noncoding RNA related biotypes (Figure 4.2B). Table 4.1 below has a brief description of the significantly misexpressed transcript biotype classes identified in Tdrd7-/- lens.

Table 4.1 Summary for Ensemble biotype classes identified in P15 Tdrd7-/- lens RNA-seq data

These are transcripts that contains an open reading frame and 1. Protein coding they code for proteins. These are transcripts that lack introns and are thought to arise 2. Processed pseudogene from reverse transcription of mRNA followed by reinsertion of DNA into the genome. These are long intergenic non coding RNA transcripts that are 3. lincRNA larger than 200bp. They belong to long noncoding transcripts. They overlap the 4. Antisense genomic span (i.e. exon or introns) of a protein-coding locus on the opposite strand They belong to long non-coding transcripts. They are found in 5. Sense intronic introns of a coding gene that does not overlap any exons They belong to long non-coding transcript. These transcripts 6. Sense overlapping contain a coding gene in its intron on the same strand 7. Processed transcript These are transcript lacking an open reading frame 8. Ribozymes These are small RNA capable of catalytic activity like enzymes These are Small RNA molecules found in the cell nucleolus and 9. snoRNA are involved in the post-transcriptional modification of other RNAs. They belong to small nucleolar RNAs (snoRNAs) and known as Small Cajal body-specific RNAs. scaRNA localize to Cajal body 10. scaRNA which is a nuclear organelle involved in the biogenesis of small nuclear ribonucleoproteins. These are pseudogene that can contain introns since produced by 11. Unprocessed pseudogene gene duplication Known as Miscellaneous RNA and they are non-coding RNA 12. miscRNA that cannot be classified

Figure 4.2 Distribution of differentially expressed transcripts in Tdrd7-/- lens (A) Smear plot for differentially expressed transcripts in Tdrd7-/- lens. Each dot on the smear plot depicts the log2 fold change (y-axis) of a transcript verses average log CPM (count per million) for each gene. Any transcript that does not meet the cut off (FDR < 0.05) is black in color. Transcripts that meet the cut off cut off (FDR < 0.05) are colored either blue for down regulated or red for upregulated. Note that many significantly missexpressed genes have log2 fold change below 1 and above -1 and these will be filtered as described in the work flow (4.2.1) by setting a fold change cut off (B) Biotype classes distribution among differentially expressed transcripts in Tdrd7-/- lens. The pie chart shows the distribution of biotype classes among Tdrd7-/- misexpressed transcriptome. Majority of misregulated transcripts are protein coding genes 986 (~95%).

Table 4.2 A list of misregulated non-protein coding genes in Tdrd7-/- lens

The table contains a list of non-protein coding transcripts belonging to several biotype classification groups. The table is sorted in ascending order within each biotype class Symbol Log2 Fold Change FDR Biotype classification Snhg12 -2.2 1.46E-48 lincRNA Snhg6 -1.2 7.34E-16 lincRNA 2410004N09Rik -1.4 4.73E-08 lincRNA Gm4131 1.0 1.83E-05 lincRNA Gm27042 1.0 0.000925 lincRNA 2310015A10Rik -0.6 0.004193 lincRNA Gm26870 1.8 0.015525 lincRNA Gm26847 0.4 0.017103 lincRNA 1600020E01Rik -0.7 0.020326 lincRNA Gm12750 -1.4 3.74E-06 antisense Gm20631 0.8 0.00113 antisense Gm10138 -0.7 0.00121 antisense Gm6548 -1.4 2.84E-09 processed_pseudogene Gm4850 -1.2 0.000845 processed_pseudogene Gm5776 0.7 0.002922 processed_pseudogene Gm1821 0.4 0.003887 processed_pseudogene Gm13456 -0.4 0.007823 processed_pseudogene Gm5345 0.7 0.009831 processed_pseudogene Gm14680 0.4 0.010778 processed_pseudogene Gm10131 0.6 0.015625 processed_pseudogene Gm6170 0.7 0.020391 processed_pseudogene Gm6863 0.3 0.028625 processed_pseudogene Gm7335 0.7 0.03153 processed_pseudogene Gm10698 0.4 0.03445 processed_pseudogene Gm14681 0.4 0.035717 processed_pseudogene Gm5454 0.7 0.040853 processed_pseudogene Gm7336 0.7 0.041903 processed_pseudogene Gm21399 0.4 0.044869 processed_pseudogene 2410006H16Rik -1.7 8.76E-24 processed_transcript 1810026B05Rik -0.9 1.10E-07 processed_transcript Snhg8 -0.8 2.06E-05 processed_transcript

E130218I03Rik -0.6 0.004212 processed_transcript 1110038B12Rik -0.7 0.007971 processed_transcript 5530601H04Rik -1.1 0.018829 processed_transcript B130055M24Rik -0.7 0.034693 processed_transcript AC245272.2 1.1 0.033215 miRNA Rn7s2 -2.9 0.042003 misc_RNA Rmrp -0.4 0.007502 ribozyme Rpph1 -0.5 0.015281 ribozyme Scarna6 -0.4 0.007663 scaRNA Gm22009 -0.4 0.040853 scaRNA Gm15513 -1.1 0.00126 sense_intronic Gm5607 -0.3 0.017192 sense_overlapping Scarna10 -0.5 0.000502 snoRNA U3 -0.4 0.025502 snoRNA Gm37938 1.1 6.55E-06 TEC H2-K2 -0.8 0.001077 unprocessed_pseudogene

Table 4.3 List of differentially expressed genes in P15 Tdrd7-/- mouse lens A Table for 82 differentially expressed genes from RNA-seq data. All genes in this list have passed the following cut off: FDR<0.05, absolute fold change > 1.2, lens enrichment > 1.5 fold-change for at least one stage across nine lens developmental stage. The gene list is presented in the ascending FDR values. Genes in this table were used for the correlation and GSEA Symbol Log2 Fold Change Absolute Fold Change FDR 1. Bfsp2 3.7 13.4 1.77E-86 2. Snhg12 -2.2 -4.3 1.46E-48 3. Tdrd7 -4.2 -16.9 1.52E-44 4. Ints3 -1.6 -2.9 1.89E-40 5. Ackr4 -1.9 -3.6 6.93E-27 6. 2410006H16Rik -1.7 -3.2 8.76E-24 7. Lpin1 -1.2 -2.3 4.20E-23 8. Ppp1r15a -1.4 -2.5 5.40E-20 9. Snhg6 -1.2 -2.2 7.34E-16 10. 1810041L15Rik -1.8 -3.4 7.53E-16 11. Tmprss11e 1.5 2.9 7.53E-16 12. Chrna4 -1.8 -3.5 7.16E-13 13. Tob2 -1.3 -2.4 8.82E-13 14. Ceacam10 -1.3 -2.4 1.17E-12 15. Syt1 -3.5 -12.5 6.89E-12 16. Sphkap -3.2 -10.0 8.51E-12 17. Fam189a2 -1.3 -2.5 1.06E-11 18. Eno2 -2.6 -6.3 1.26E-11 19. Stambpl1 -1.3 -2.3 1.29E-11 20. H2-Eb1 -2.1 -4.1 1.98E-11 21. Rp1 -4.7 -29.5 4.40E-11 22. Rbp3 -4.1 -18.9 1.39E-10 23. Rpgrip1 -2.3 -5.1 4.70E-10 24. Rorb -3.3 -11.0 7.51E-10 25. Sag -4.4 -24.6 7.92E-10 26. Stxbp1 -2.1 -4.5 1.32E-09 27. Pde6b -4.1 -18.6 1.74E-09 28. Gnat1 -4.6 -27.1 1.75E-09 29. Pde6a -4.0 -18.0 2.71E-09 30. Doc2b -1.3 -2.5 3.95E-09 31. Rho -4.6 -28.0 4.45E-09 32. Nav1 -1.3 -2.4 9.43E-09 33. Pbx4 -1.5 -2.8 1.14E-08

34. Plekhb1 -3.4 -12.0 1.86E-08 35. Ahdc1 -1.3 -2.5 2.17E-08 36. Pcdh17 -1.3 -2.5 2.25E-08 37. Nfasc -1.1 -2.3 2.25E-08 38. Hyal3 1.5 2.8 5.25E-08 39. Acad11 -1.2 -2.1 6.09E-08 40. Rgs9bp -2.9 -7.9 6.62E-08 41. Rom1 -4.0 -18.2 8.29E-08 42. Gnao1 -1.8 -3.5 1.33E-07 43. Dlg2 -1.9 -3.9 1.40E-07 44. Rfx8 1.4 2.7 1.51E-07 45. 6430548M08Rik -1.8 -3.7 1.83E-07 46. Pvrl1 -1.3 -2.4 2.76E-07 47. Rlbp1 -2.4 -5.6 3.19E-07 48. Zfp354b -1.7 -3.1 5.90E-07 49. Ccdc69 1.8 3.4 7.36E-07 50. Pex5l -3.0 -8.6 1.08E-06 51. Rnf157 -1.4 -2.6 1.17E-06 52. Sv2b -2.5 -6.0 1.95E-06 53. Lin7a -1.1 -2.2 1.95E-06 54. Synj1 -1.3 -2.5 9.34E-06 55. Pcbp3 -1.6 -3.1 1.14E-05 56. Syngr1 -2.0 -4.3 1.29E-05 57. Myrip -1.1 -2.3 1.61E-05 58. Mef2d -1.6 -3.2 1.83E-05 59. Brsk2 -1.4 -2.5 2.07E-05 60. Frmpd1 -1.7 -3.3 3.78E-05 61. Pfkfb2 -1.8 -3.7 3.94E-05 62. Kcnab2 -1.6 -3.1 4.68E-05 63. Stx3 -1.5 -3.0 5.07E-05 64. Impdh1 -2.0 -4.3 0.0002 65. Plekhb2 -1.2 -2.5 0.0002 66. Abca4 -1.3 -2.7 0.0002 67. Kcnj14 -2.0 -4.4 0.0003 68. Cadm2 -1.2 -2.3 0.0003 69. Reep6 -1.6 -3.2 0.0004 70. Sv2a -1.2 -2.3 0.0004 71. Cngb1 -1.5 -3.1 0.0005 72. Thrb -1.1 -2.2 0.0005 73. Herc3 -1.5 -3.0 0.0006 74. Gm4850 -1.2 -2.3 0.0008

75. Fam161a -1.5 -3.0 0.0019 76. Madd -1.5 -3.0 0.0022 77. Ccdc126 -1.6 -3.2 0.0026 78. Cecr2 -1.3 -2.6 0.0104 79. 5530601H04Rik -1.1 -2.1 0.0188 80. Ntrk2 1.6 3.1 0.0379 81. Tyrp1 1.3 2.7 0.0425 82. Fam129a 1.6 3.2 0.0481

Table 4.4 List of differentially expressed proteins in P15 Tdrd7-/- mouse lens A Table for 62 differentially expressed proteins from Tandem Mass Tag (TMT) Mass spectrometry proteome data. All proteins in this list have passed the following cut off: P-value < 0.05, absolute fold change > 1.2, lens enrichment > 1.5 fold-change for at least one stage across the nine lens developmental stage. The gene list is presented based on ascending FDR values. Genes in this table were for the correlation and GSEA Symbol Log2 Fold Change Fold Change P-Value 1. Tdrd7 -3.5 -11.2 0 2. Prkca -0.9 -1.9 0 3. Pstpip2 -0.7 -1.6 0 4. Stambpl1 -0.6 -1.5 0 5. Kctd12 0.5 1.4 0 6. Cap2 0.3 1.2 5.00E-06 7. Gstp1 0.4 1.3 7.00E-06 8. Casp7 0.4 1.3 8.00E-06 9. Rbp3 1.3 2.4 8.00E-06 10. Map4 0.5 1.4 1.00E-05 11. Eml2 0.3 1.2 1.10E-05 12. Ddx19a 0.3 1.2 1.30E-05 13. Akap2 -0.4 -1.3 1.50E-05 14. Sparc -0.3 -1.3 2.70E-05 15. Fdft1 0.6 1.5 5.10E-05 16. Irs1 -0.6 -1.5 6.40E-05 17. Hagh 0.3 1.2 7.20E-05 18. Pdlim5 -0.3 -1.3 9.20E-05 19. Dst 0.3 1.2 0.000149 20. Pacsin3 -0.3 -1.2 0.000177 21. Mtor -0.3 -1.3 0.000183 22. Ptpru -0.3 -1.3 0.000312 23. Lama2 -0.6 -1.6 0.000452 24. Ppl 0.3 1.2 0.000517 25. Pgk1 0.3 1.2 0.000653 26. Tmx1 0.3 1.2 0.000936 27. Rho 1.5 2.8 0.00128 28. Vps16 -0.6 -1.6 0.001471 29. Fam114a1 0.3 1.3 0.001677 30. Dsp 0.5 1.4 0.001956 31. Cpt1a -0.3 -1.2 0.00203 32. Gstm1 -0.3 -1.2 0.002054 33. Tspan14 0.3 1.2 0.002134

34. Acad11 -0.3 -1.2 0.002493 35. Trafd1 -0.3 -1.2 0.00405 36. Faf1 -0.6 -1.5 0.004332 37. Gstm2 -0.3 -1.2 0.004979 38. Rlbp1 0.5 1.5 0.006018 39. Efhd2 -0.3 -1.2 0.006235 40. Pcyox1 -0.3 -1.2 0.006461 41. Fyco1 -0.4 -1.3 0.009786 42. Echdc1 0.3 1.2 0.010075 43. Adam10 0.6 1.5 0.010928 44. Hspb1 -0.4 -1.3 0.010985 45. Bfsp1 0.3 1.2 0.011156 46. Lamc1 -0.8 -1.7 0.01127 47. Mtmr3 -1.3 -2.4 0.011988 48. Bag2 -0.3 -1.3 0.012639 49. Mgarp 0.4 1.3 0.013502 50. Itga6 0.3 1.2 0.014987 51. Mapk8ip3 0.4 1.3 0.016157 52. Rab3b -0.3 -1.2 0.017043 53. Susd2 -0.3 -1.2 0.018172 54. Dpcd -0.7 -1.7 0.018892 55. Gnat1 0.8 1.8 0.018959 56. Mtr -0.3 -1.2 0.020336 57. Folr1 -0.3 -1.2 0.024191 58. Col18a1 -0.5 -1.4 0.027145 59. Sag 1.4 2.6 0.02738 60. Gnao1 0.9 1.8 0.031649 61. Lamb1 -0.7 -1.6 0.03258 62. Pcbd2 -0.3 -1.2 0.033133 63. Prkg1 -0.5 -1.4 0.035056 64. S100a4 0.3 1.3 0.035597 65. Lama5 -0.7 -1.6 0.040381 66. Pon2 0.4 1.3 0.040505 67. Nid1 -0.6 -1.5 0.041518 68. Eea1 0.4 1.3 0.043609 69. Hsd17b10 -0.3 -1.2 0.043691

4.2.1 Characteristic expression pattern for high-priority differentially expressed candidates in Tdrd7-/- lens Next, I sought to examine the normal spatiotemporal lens gene expression microarray data for the differentially expressed high-priority candidate transcripts (Table 4.3) and differentially expressed high-priority candidate proteins (Table 4.4), by using the iSyTE 2.0 database (Kakrana et al., 2018). To explore the normal pattern of expression among Tdrd7-/- differentially expressed high-priority candidates I used nine developmental stages spanning early embryonic and postnatal stages. To visualize the pattern of expression I generated two cluster heatmap (clustered by row) to display the normal expression across different stages for each of the high priority candidates. Both heatmaps (red for the transcriptome and blue for the proteome) (Figure 4.3 and Figure 4.4) reveals different spatiotemporal dynamic patterns of expression of the candidate genes in normal lens development. A group of candidates exhibit their highest lens expression during early embryonic development (E10.5, E11.5 and E16.5). The expression of these candidates continues to taper gradually as development progresses.

Examples of candidates with this pattern of expression are NAV1 and Cecr2 genes or Dsp and Ddx19 proteins. Another group of candidates show an expression pattern at its highest during late embryonic development until early postnatal stages (P2), examples are Lin7a and Stx3 genes as well as Dst and Prkca proteins. Finally, a group of candidates exhibit their highest expression during adulthood (P56), with less expression observed during postnatal or early embryonic stages. Examples for candidates in this group are Brsk2 and Doc2b genes as well as Sag and Pcbd2 proteins. These analyses demonstrate that Tdrd7 downstream target genes are potentially important in distinct stages in lens development.

Figure 4.3 Cluster heatmap for lens normal expression from iSyTE for the candidate misregulated genes in P15 Tdrd7-/- lens Cluster heatmap displays a wide span of normal expression pattern across early and late developmental stages for Tdrd7-/- high priority candidate genes

Figure 4.4 Cluster heatmap for lens normal expression from iSyTE for the candidate misregulated proteins in P15 Tdrd7-/- lens Cluster heatmap displays a wide span of normal expression pattern across early and late developmental stages for Tdrd7-/- high priority candidate proteins

4.2.2 Snhg12 gene and Cap2 protein are among top priority candidates in Tdrd7-/- lens Next, I performed correlation analysis for lens normal expression across nine developmental stages for the top 82 candidate transcripts (from the RNA-seq data) and for the 69 top candidate proteins (from the proteome data). The aim of this analysis was to further prioritize the top priority candidates by identifying the ones with a normal expression pattern that is similar to the normal expression pattern of Tdrd7 across the different lens stages. This analysis revealed 9 top candidate genes and 23 top candidate proteins. All candidates had a Pearson’s correlation values >0.8 . To refine the results, the list of candidates was sorted by their significance value and the top 9 candidates are shown (Figure 4.5). The results uncover Beaded Filament specific Protein 2 gene (Bfsp2) discussed in (section1.3), whose mutation is linked to cataract (OMIM : 603212) (Cui et al., 2007; Jakobs et al., 2000; Sandilands et al., 2004), as the top candidate gene, which is upregulated in Tdrd7-/- by a log2 fold change of 3.7. Interestingly, the second top candidate identifies a Snhg12 (Small Nucleolar RNA Host Gene 12), which is a long non coding RNA (lncRNA) that is recognized as a biomarker and potential therapeutic target in cancers. Snhg12 had a downregulation fold change of log2 of -2.2 in Tdrd7-/- lens. As for the proteome data, the analysis uncovered Cyclase Associated Proteins2 (Cap2) as the top candidate protein, which is upregulated by a log2 fold change of 0.3 in Tdrd7-/- lens. Cap2 is known to modulate F-actin dynamics and was linked to eye developmental defects in flies and to microphthalmia mouse (Benlali et al., 2000; Field et al., 2015). The second top candidate protein is A-Kinase Anchoring Protein 2 (Akap2), which is downregulated by a log2 fold change of -0.4 in Tdrd7-/- lens. Akap2 gene products were shown to form a stable complex with Major Intrinsic Protein MIP, a process that is necessary for maintaining lens transparency (Gold et al., 2012).

Interestingly, by the candidate gene approach, I have identified the downregulation of MIP in the lens on P15 in Tdrd7-/- by RT-qPCR and by immunofluorescence (Figure 4.7). To gain a broad insight into the potential biological significance of the high- priority candidate genes and proteins, I have performed a functional classification analysis using Panther classification system (PANTHER v.14.0) (Mi et al., 2019). Results for each list containing the top candidates from the transcriptome or the proteome, yielded 13 GO terms for the transcriptome and 18 GO terms for the proteome. Interestingly, among all the GO terms identified, 9 GO terms were commonly found (Figure 4.6). Some of the common GO terms are: biological regulation (GO:0065007), cellular process (GO:0009987), catalytic activity (GO:0003824), membrane (GO:0016020), protein-containing complex (GO:0032991), among others.

Figure 4.5 Top potential candidate misregulated genes and proteins in Tdrd7- /- lens (A-B) Top potential candidate genes in red and (C-D) Top potential candidate proteins. (A and C) are the heatmap on the for Pearson’s correlation among top candidates genes (red) and proteins (blue) respectivly. The Yellow rectangle highlightes the correlation values of the normal expression of thoese candidates to the normal expression of Tdrd7 the red heat map is for the transcriptomre (A), while the blue is for the proteoome data (B). Log2 fold change of the top candidates obtained from their diffrential expression analysis for (B) trancripts and (D) proteins.

Figure 4.6 Functional classification for high-priority candidates whose expression correlates with Tdrd7 in normal lens development Panther Go-slim functional classification results for top differentially expressed candidates whose expression highly correlates with Tdrd7 expression across 9 developmental stages. The top panels represent top genes while the lower panel represents top proteins. Numbers inside the pie chart reflects the number of genes that falls within that specific classification

Figure 4.7 Misregulation of MIP in P15 Tdrd7-/- lens (A) RT-qPCR for MIP in Tdrd7-/- and control lens (P-value <0.05). (B) immunofluorescence staining for MIP protein in lens tissue cross section in Tdrd7-/- and control. Results in (A) an (B) indicate reduced levels of MIP in Tdrd7-/- lens on the RNA and the protein level.

4.2.3 Gene Set Enrichment Analysis (GESA) identifies processes that are implicated by the absence of Tdrd7 in the lens Gene set enrichment analysis is an approach to identify categories of genes or proteins that are enriched in a large predefined set genes or proteins. Predefined sets contain a groups of genes/proteins that share common terms such as biological function, or regulation, among others. The implementation of statistical methods in GSEA enables the identification of significantly enriched sets for a given large profiling data of ranked genes or protein (Subramanian et al., 2005).

To gain insights into terms that are implicated in the absence of Tdrd7 on the transcriptome and the proteome level, I performed GSEA for the prioritized 82 differentially expressed genes (Table 4.3) as well as the 69 prioritized differentially expressed proteins (Table 4.4). This analysis yielded different enrichment terms that did not overlap between the proteome and the transcriptome.

Analysis of differentially expressed genes indicated that terms such as DNA binding transcription activity and transcription regulation to be highly enriched among down regulated genes. For the upregulated genes, terms such as sequestering of extracellular ligand from receptor and protein localization to extra cellular region were the most enriched terms. As for the differentially expressed proteins, the most enriched terms among down regulated proteins were for locomotion that had 17 proteins and cell motility that also that had the same 17 proteins, while for the upregulated proteins the most enriched terms were visual perception and sensory perception of light stimuli, both terms contained the same 4 proteins.

Figure 4.8 Scatter plots for differentially expressed (DE) genes enriched GO terms The y-axis represents the enriched GO categories and the x-axis represents the rich factor of the enriched GO terms. The dot size represents the gene count and the color of the dots represents the P- values. Rich factor is the ratio of differentially expressed gene number enriched in the pathway to the total gene number in a certain pathway

Figure 4.9 Scatter plots for differentially expressed (DE) proteins enriched GO terms The y-axis represents the enriched GO categories and the x-axis represents the rich factor of the enriched GO terms. The dot size represents the gene count and the color of the dots represents the P- values. Rich factor is the ratio of differentially expressed gene number enriched in the pathway to the total gene number in a certain pathway

4.3 Discussion Investigation of the transcriptome and the proteome of Tdrd7-/- lens represents an opportunity to gain new insights into the key factors that are impacted by the absence of Tdrd7 in the lens. RNA seq-data allows the discovery of misregulated protein coding genes as well as non-coding RNA, and unlike the previous microarray analyses on the lens, RNA-seq data enables the identification of novel noncoding RNA (Hutchins et al.,

2016; Zhang et al., 2017). Non-coding RNAs have been shown to be critical for maintaining lens transparency and their disruption can cause lens defects (Conte et al., 2010; Fu et al., 2017; Liu et al., 2018), but this remains an understudied area. RNA analysis of P15 in Tdrd7-/- lens have identified Snhg12, a long non coding RNA, to be among the high-priority candidates that are abnormally expressed in the absence of Tdrd7. Misexpression of Snhg12 is linked to several carcinomas, ischemia/reperfusion injuries and apoptosis. Snhg12 is also considered to be a biomarker and a potential therapeutic target for cancers in human. It has been shown to regulate miRNAs in cancer tissues, which in turn impacts the expression of genes that also participate in regulatory pathways known to be important in the lens (Chen et al., 2019; Fujimura, 2016; Li et al., 2019; Tamang et al., 2019; Wang et al., 2017; Yao et al., 2019; Zhou et al., 2018a, 2018b). The present study represents the first report for Snhg12 in the lens which suggests that a long noncoding RNA is downstream of Tdrd7. It is important to complement the RNA-seq analysis performed in Tdrd7-/- lens here with a comprehensive analysis for long non coding RNAs as well as miRNAs. This would best be achieved by using databases that are well annotated for such analysis like GNECODE and Mirbase databases (Derrien et al., 2012; Kozomara and Griffiths-Jones, 2014;

Kozomara et al., 2019) and will allow through identification of all impacted noncoding RNA in Tdrd7-/- lens. Together, these analyses will advance our knowledge of Tdrd7- based posttranscriptional control in the lens. To identify high-priority misexpressed candidates in Tdrd7-/- lens transcriptome and proteome, I used iSyTE 2.0 lens enrichment fold-change values (which informs on the ratio of gene expression in the lens compared to its expression in the whole body) across 9 developmental stages as a condition (Kakrana et al., 2018; Lachke et al., 2012).

Although overlaying this filter over differential expression significance value and fold change values might be stringent, this strategy has been shown to be efficient in predicting candidates with strong relevance to lens biology. Further, candidates with lens expression that highly correlates with the expression of Tdrd7 across development might be the ones that are impacted the most by the absence of Tdrd7. Due to the assumption that these orchestrate cellular functions together across several time points and the absence of a key regulator upstream can affect their abundance. For example,

Hspb1 described in Chapter 3 (Section 3.2.2) was identified by a similar integrated analysis and further investigation performed on Hspb1 by another graduate student, Shaili Patel, have demonstrated for the first time, that knocking down Hspb1 in Xenopous causes eye and lens defects (Barnum, Al Saai, Patel et al., 2020). Use of a similar approach, but with more developmental stages added (9 stages instead of using only four), lead to the identification of new promising candidates such as Cap2, (discussed in detail in Chapter 5 ), where I demonstrate, for the first time, that deficiency of Cap2 in mouse causes lens fiber cell defects. It was also interesting to identify both Bfsp2 and Bfsp1, by the transcriptome and proteome data, respectively, among the high- priority candidates in the Tdrd7-/- lens. Bfsp2 and Bfsp1 encode intermediate filament

proteins that are critical lens cytoskeleton components essential for maintaining transparency and mechanical stiffness. Indeed, misregulation or mutations of beaded filaments are linked to lens defects (Alizadeh et al., 2003; Blankenship et al., 2001; Fischer et al., 2003; Fudge et al., 2011, 2011; Perng et al., 2007; Ramachandran et al., 2007; Song et al., 2009; Wang et al., 2013). Identifying both beaded filament proteins to be downstream of Tdrd7 suggest a strong regulatory role for Tdrd7 on another lens cytoskeletal component. Additionally, Akap2 that is down regulated in Tdrd7-/- lens, and is among top candidates, has been shown to be important for lens transparency as well (Gold et al., 2012). Despite the efficiency of the analysis strategy described here in predicting these high-priority candidates for further follow up, it is also important to study and investigate the reminder 23 significantly mis-regulated 12 genes and 11 proteins that also survived these stringent filters. While the general idea derived from the central dogma assumes that studying the transcriptome will inform on the proteome, this relationship is not always true.

Indeed, low correlation between mRNA and Protein expression that can be attributed to several factors, such as half-life variation, or posttranscriptional regulation within a given tissue (Haider and Pal, 2013). My work here demonstrates that concomitant analysis of the transcriptome and the proteome lead to the identified of many potential misregulated high-priority candidates and enrichment terms as opposed to the individual use of the approaches. It may be even more insightful to perform an analysis that joins both omics levels as one joint data set. Multiple approaches for ways to combine a set of transcriptomic and proteomic profiles are available. To perform joint meta-analysis it is important to well understand the techniques that were used to generate the data sets. Thus it is necessary to carefully review and select the approach that suits the presently

generated datasets while being aware of the advantages and limitations of the selected approach verses other (Haider and Pal, 2013). Finally, the correlation data generated for both data sets can be used to build a co-expression network for Tdrd7. It was observed that a major portion of misexpressed transcriptome and proteome of Tdrd7-/- at P15 had small absolute fold change. For that reason I have chosen the 1.2 absolute fold change as a cutoff for my analysis. The small absolute fold change observed could be a result of Tdrd7 functioning as a global regulator that for many genes. Thus, in the absence of Tdrd7 affects the abundance of those genes in miniscule fold changes, while at this level the Tdrd7-/- lens is apparently normal under light microscopy (Figure 1.9-A ). And the continuous buildup of those small gene abundance changes is what leads to full impact of cataract seen seven days later. It is interesting to note than among significantly differentially expressed genes and proteins, my analysis has picked some retina specific genes, and that was also reflected in enriched terms identified by GSEA such as the term retinal binding or the term photoreceptor inner segment. Results pertaining to retinal genes within Tdrd7-/- transcriptome and proteome has to be carefully evaluated. There have been reports in retina research describing a similar outcomes where the identified lens genes in their retina work and concluded these were carried over during tissue collection (Kamphuis et al., 2007). This could possibly be one explanation for identifying retina genes and proteins in my data set. It is also possible that some of the retina-specific genes that are normally reduced in the lens are elevated due to Tdrd7 deficiency. Future studies should focus on addressing this experimentally.

Chapter 5

CHARACTERIZATION OF LENS DEFECTS IN CAP2 KNOCKOUT MICE

5.1 Introduction CAPs (Cyclase Associated Proteins) are highly conserved proteins that are found in diverse organisms such as yeast, plants, and vertebrates. The first isolated CAP protein was a yeast protein that is required for RAS activation of adenylate cyclase (Field et al., 1990; Kawamukai et al., 1992). Later studies revealed that CAP proteins act as modulators of actin filament dynamics. Actin filaments are composed of actin monomer which are a major component of the actin cytoskeleton that is important for cell membrane morphology, cell shape, adhesion, and polarity (Fischer et al., 2003; Kuwabara, 1968; Rao and Maddala, 2006a). In vertebrates, there are two Cap family genes, Cap1 and Cap2. Based on the amino acid sequence, both homologs share more than 60% similarity (Swiston et al., 1995). The homology between mouse Cap1 and Cap2 amino acid sequence is high in the C-terminal region compared to the proline-rich central domain and the N-terminal domain (Peche et al., 2007). As for their expression, Cap1 and Cap2 have varied pattern in different tissues. Cap1 is widely expressed while

Cap2 has been shown to be more specifically expressed in testes, brain, heart, skeletal muscle and skin (Bertling et al., 2004; Peche et al., 2007; Swiston et al., 1995). This variation in their expression suggests that Cap1 and Cap2 modulate actin filament dynamics in a tissue specific manner (Swiston et al., 1995). (Benlali et al., 2000). Cap2 mutation in humans cause cardiac myopathy (Kim et al., 2019), is linked to human 6p22.3 deletion syndrome (heart defects is one of the disorders of this syndrome)

(Celestino-Soper et al., 2012) and is associated with cancer progression and poor prognosis in patients (Saker et al., 2020). Several Cap2 deficient mouse models exhibited one or more of the following: sudden death due to cardiomyopathy (similar to human), skeletal muscle architectural defects and altered wound healing (Aspit et al., 2019; Field et al., 2015, 2016; Kepser et al., 2019; Kosmas et al., 2015). Expression of Cap2 in the lens was first described in Xenopous embryos (Wolanski et al., 2009). Additionally, disruption of act up (acu) gene, a Cap homolog in drosophila, was shown to cause developmental eye defects (Benlali et al., 2000). Field et al., 2016, also reported eye developmental defects in Cap2 targeted knockout (Cap2-/-) mice that exhibited in microphthalmia and reduced iris diameter (Figure 5.1). In addition, Cap2- /- mice were prone to eye infection and inflammation (Field et al., 2016). In this study, Cap2 expression in the mouse eye was detected in the cornea, lens, retina and retinal pigment epithelium (RPE), with highest expression being in the retina and the lens (Field et al., 2015).

Figure 5.1 Cap2-/- Mice Exhibit Microphthalmia An image for cap2-/- mice exhibiting microphthalmia. These mice are also prone to eye infections, notice the yellow secretions in the Cap2-/- eye, while the eye in the control mouse is clear from secretions (Adapted from Field et al., 2016)

Further, the Cvekl lab has identified Cap2 as a positively controlled direct target of Pax6 in lens cells (Sun et al., 2016). Interestingly, Cap2 was identified among top abnormally elevated proteins in Tdrd7-/- mouse lens proteome at postnatal stage day 15 (P15) (Figure 5.2), suggesting that Cap2 is downstream of Tdrd7. Tdrd7-/- lens exhibit abnormal F-actin distribution and lens fiber cell morphological defects at the beginning of the organelle free zone (3.2.1). Moreover, while Cap2 has been demonstrated to function in maintaining cell morphology and as actin dynamic modulator in other tissues such as the heart and the skin (Aspit et al., 2019; Kepser et al., 2019; Kosmas et al., 2015; Wolanski et al., 2009), its function in the lens is unexplored. Therefore, I sought to perform an initial characterization of Cap2-/- mouse lens to gain insights on Cap2 function in the lens. This study will also add to our knowledge on the pathophysiology of Tdrd7-/- and the impact of Cap2 abnormal elevation in the Tdrd7-/- proteome.

Figure 5.2 Cap2 Protein is abnormally elevated in Tdrd7-/- Mouse Lens Analysis of Tdrd7-/- mouse lens proteome at postnatal stage 15 (P15) identified Cap2 among top upregulated differentially expressed proteins. Cap2 fold change of 1.2 and FDR < 0.000653

5.2 Results

5.2.1 iSyTE (integrated Systems Tool for Eye gene discovery) Identifies Cap2 as a Potential Key Factor in the Lens To investigate Cap2 expression in the lens, I used iSyTE database that is a resource for lens expression based on microarray, RNA-seq and proteome datasets and which has facilitated the identification of cataract linked genes (Anand et al., 2018; Aryal et al., 2020b; Kakrana et al., 2018; Lachke et al.,

2012). iSyTE analysis based on microarray data as well RNA-seq data indicates the high expression and enrichment of Cap2 mRNA in the lens during embryonic stages through adulthood. Cap2 lens-enrichment begins as early as embryonic (E) day 11.5 at the lens vesicle stage in mouse. Cap2 is among the top 150 proteins expressed in the mouse E14.5 lens (Aryal et al., 2020b). iSyTE analysis also indicates that the actin depolymerizing factor (ADP) Cofilin 1 which associates with Cap2 C-terminal domain (Kumar et al., 2016) to be highly

expressed in the lens (Figure 5.3). Interestingly, iSyTE analysis shows that Cap2 is significantly misexpressed in lenses of various gene-specific knockout mouse models (Notch2 or Hsf4 deficient) that exhibit lens defects (Figure 5.4). Together, these findings suggest that Cap2, enriched and highly expressed in mouse lens is under control of several different important regulators in the lens,

and therefore represents a promising candidate in the lens.

Figure 5.3 iSyTE predicts Cap2 as a high-priority candidate in the lens iSyTE microarray and RNA seq meta-analysis data identifies Cap2 to be highly expressed and enriched lens gene. The expression is found across embryonic and postnatal stages. iSyTE also identifies the expression of cofilin 1 ,which associates with Cap2, to be highly expressed as well

Figure 5.4 Cap2 is Mis-regulated in Several Cataract Mouse Model iSyTE meta-analysis indicates that Cap2 is down-regulated in lenses of Pax6+/-, Notch2-/- and E2f1/2/3-/- mice (yellow dotted squares). iSyTE also indicates that Cap2 is up-regulated in lenses of Hsf4-/- and Brg1-/- mice (green dotted squares). Among those, Cap2 missregulation is significant in Notch2-/- and in Hsf4-/- (P-value<0.05) (shaded with red).

5.2.2 Cap2 expression highly correlates with Tdrd7 and Pax6 through developmental stages in mouse lens To identify the strength of association between Cap2 and its upstream regulators described above (Figure 5.4), I performed Pearson correlation analysis of Cap2 across nine developmental stages (Figure 5.5). The expression data for these time points were downloaded iSyTE microarray data set. Tdrd7 showed the strongest positive correlation with Cap2 (0.8) while Pax6 showed a strongest negative

correlation with Cap2 (-0.8). These results suggest that Cap2 might be expressed and will function in fiber cells, because Pax6 is largely expressed in the epithelium and Tdrd7 is largely expressed in the fiber cells (Koroma et al., 1997; Lachke et al., 2011b).

Figure 5.5 Cap2 expression correlates with Tdrd7 across nine developmental stages The yellow frame indicates the correlation values of Cap2 against the genes on the left. Pearson correlation heat-map indicates that Tdrd7 had the strongest positive correlation to the expression of Cap2 while Px6 had the strongest negative correlation. Also note the strong negative correlation between Tdrd7 and Pax6

5.2.3 Cap2-/- lens fiber cell exhibits morphological defects Lens fiber cell architectural defects were observed on the scanning electron microscopy (Figure 5.6).

Figure 5.6 Scanning electron microscopy of Cap2-/- lens Scanning electron microscopy of cortical region of the lens showed fiber cells defects. Images for the wild-type littermate (control) and Cap2-/- mouse lens (left) and high-magnification of fiber cells in the cortical region (right) are shown.

To investigate these defects further and gain insights into the hexagonal lens fiber cell architecture, I obtained lens tissue cross sections from ~4 weeks old Cap2-/- and performed immunofluorescence staining for F-actin using phalloidin staining and lens fiber cells membrane using Wheat Germ Agglutinin. When Cap2-/- sections were compared to the wild-type control, the membrane staining revealed distorted architecture of lens fiber cells hexagonal shape which became progressively prominent at the beginning of the organelle free zone. The hexagonal shape distortion exhibited in the misorientation of some early differentiating fiber cells resulting in some degree of disorganization of the lens characteristic meridonial rows (Figure 1.1 and Figure 5.7). I

have also observed discontinuous patches of the membrane staining in Cap2-/- lens i.e. the staining fades and disappears in some areas of the section. when compared to the control which displayed smooth membrane staining along the lens section. Phalloidin staining revealed abnormal pattern of distribution for F-actin, especially along lens fiber cells hexagonal borders. This abnormal distribution exhibited in accumulation of more F-actin along the hexagonal short sides (Figure 5.7 and Figure 5.8). Analysis of Cap2- /- lens fiber cells hexagonal appearance (described above in section 2.19) revealed significant increase in the length between the distant vertices as well as decrease in the length between the broad sides for mid-fiber cells zone (Figure 5.9). As for the mature fiber cell zone there were a significant decrease in the length between the broad sides (Figure 5.9). Measurement of F-actin signal across the mid-region fiber cell meridonial rows shows periodic appearance of hexagons in Cap2-/- and control lens (Figure 5.10). However, similar measurement across the mature-region fiber cell meridonial rows shows irregular or sporadic appearance of hexagons and elevated F-actin signals in

Cap2-/- mouse lens (Figure 5.10).

Figure 5.7 Cap2-/- mouse lens exhibits abnormal cellular morphology and F- actin distribution in fiber cell (A) Phalloidin staining shows abnormal pattern of distribution for filamentous actin, especially along lens fiber cells hexagonal borders. Elevated F-actin staining is observed on the hexagonal short sides in Cap2-/- lens fiber cells, particularly in the post-nuclear degradation zone, coinciding with the defective lens fiber cell region in Tdrd7-/- lens. (B) WGA membrane staining indicates lens fiber cells have altered the hexagonal appearance and disorganization of the lens fiber cells

meridonial rows. (C) Image for merged channels of phalloidin (red), WGA (green) and DAPI (blue, nuclear staining)

Figure 5.8 Abnormal cellular F-actin distribution in Cap2-/- mature fiber cells The displayed panels are high-magnification view of the mature fiber cell zone of Cap2-/- and control lens. Top two panels are phalloidin staining exhibiting abnormal distribution and accumulation filamentous actin along the hexagonal short sides. The accumulation is not observed with the membrane staining WGA. Instead, discontinuous patches of membrane staining are observed in Cap2-/-

Figure 5.9 Hexagonal appearance alterations in Cap2-/- mouse lens fiber cells (A) In Cap2-/- mouse lens, while mid-region fiber cells retain a general “hexagonal” morphology, the length between the distant vertices is abnormally increased. (B) In Cap2-/- mouse lens, while mid-region fiber cells retain a general “hexagonal” morphology, the length between the broad sides is abnormally decreased (C) In Cap2-/- mouse lens, while mature-region fiber cells retain a general “hexagonal” morphology, the length between the broad sides is abnormally decreased

Figure 5.10 Measurement of uniformity of fiber cell organization in Cap2-/- mouse lens fiber cells (A) Phalloidin staining for Cap2-/- and control, this image illustrates how F-actin signal was measured using line scan in imageJ, each line depicts the a line scan by imageJ and the colors correspond to the results in (B,C,D,E) below. (B,C) Measurement of F-actin signal across the mid- region fiber cell meridonial rows shows periodic appearance of hexagons in Cap2-/- and control lens. (D, E), similar measurement across the mature-region fiber cell meridonial rows shows irregular appearance of hexagons and elevated F-actin signals in Cap2-/- mouse lens.

5.3 Discussion It has been shown in Xenopous that Cap2 is expressed during early development, and its expression was abundant in cells that are undergoing cytoskeletal remodeling.

100

Cap2 was particularly abundant in elongating cells such as myotubes, axons, dendrites and lens fiber cells. (Wolanski et al., 2009). Cap2 can regulate actin dynamics in distinct mechanisms: 1. Cap2 contains WASP homology (WH2) domain, a small motif consisting of ~25 amino acids, which was shown to binds to G-actin (sequester G-actin), and was shown to have the capacity to sever actin filament (Chereau et al., 2005; Peche et al., 2013). 2. Cap2 was shown to interact, via its C-terminal domain, to the ADP/Cofilin (nonmuscle-cofilin) (Kumar et al., 2016). Cofilin 1 modulate actin dynamics also by severing and by promoting the depolymerization of actin filaments’ pointed and barbed end (Wioland et al., 2017). In Cap2 deficient mice, it was shown that brain cells and primary fibroblasts exhibited increased F-actin (Kosmas et al., 2015; Kumar et al., 2016). In addition, Cap2 deficient fibroblast were able to rearrange F-actin faster without a change in the total actin content (Kosmas et al., 2015). Furthermore, Cap2 deficient fibroblast had extended cellular protrusions and increased focal adhesions resulting in decreased cell motility (Kosmas et al., 2015). Elevated β-actin were also reported in Cap2 deficient patient-derived fibroblasts (Aspit et al., 2019). The findings in patient samples in addition to the findings in Cap2 deficient mouse models suggests impaired actin dynamics in absence of Cap2 (Aspit et al., 2019; Bertling et al., 2004; Field et al., 2015; Kosmas et al., 2015; Peche et al., 2013). In the lens tissue, the process of lens cells differentiation and elongation have been critically linked to many factors such as the reorganization of actin cytoskeleton. For example, during differentiation, the F-actin to G-actin ratio increases due to the increase in filamentous actin (Ramaekers et al., 1981). The increase in filamentous actin is found to be highest in young differentiating fiber cells, and is followed by a decrease in older differentiating and mature lens fiber cells (Ramaekers et al., 1981). Contrary to normal lens, in Cap2-

101

/- lens, I have observed a progressive increase in F-actin staining in the form of aggregates in maturing fiber cells, especially in cells that are undergoing organelle degradation (mid-mature zone) and beyond (Figure 5.7-A). This increase in F-actin staining was coupled with morphological defects in Cap2-/- lens fiber cells. This is similar to previous studies on Cap2 deficient mice, where it was shown that F-actin abnormally increases in absence on Cap2 (Kosmas et al., 2015). Further, I found that F- actin distribution was increasingly abnormal in appearance along the short sides of lens fiber cells hexagonal architecture (Figure 5.7). These observations suggest that Cap2 is a factor that is important for lens biology and it suggests that Cap2 modulates actin dynamics that controls fiber cell morphology. To gain more insights into how Cap2 regulates these aspects of actin dynamics in the lens, it would be interesting to investigate the total actin content and the ratio of F-actin to G-actin in Cap2-/- lens fiber cells. To further characterize the cytoskeletal defects in Cap2-/- lens it would be also interesting to investigate the spatiotemporal expression of Cap2 in wild-type lens, Cap2 interacting partner cofilin 1 as well as other actin binding proteins in the Cap2-/- mouse lens. It is interesting to note that the area where the morphological defects are apparent in Cap2-/- mouse lens, coincides with the area of lens fiber cells defects in

Tdrd7-/- mouse lens. This suggests that Cap2 expression level is important for the morphology of lens fiber cells that are undergoing organelle degradation. In Tdrd7-/- mouse lens, Cap2 was significantly upregulates in the proteome data (Figure 5.2). In other tissues, over-expression of Cap2 is linked to cancer and poor prognosis in patients. Cancer cells are known to have special cell motility characteristics that promote metastasis. Those characteristics modulates cell migration, adhesion and

102

morphological changes in metastasized cancer cells. It is suggested that cytoskeleton changes that are required for metastasized cells are being met by increasing the expression of Cap2. (Kirfel, 2004; Masugi et al., 2015; Zhou et al., 2014). Cell migration and actin reorganization are two factors that are coordinately controlled during lens fiber cell differentiation. The upregulation of Cap2 in Tdrd7-/- lens could cause lens fiber cells actin dynamics to shift in a manner similar to that of metastasized cancer cells ultimately resulting in the disorganization observed in

Tdrd7-/- lens. It is also possible that elevation of Cap2 expression in Tdrd7-/- lens is a response to correct the F-actin dynamic defects that arises in Tdrd7 deficient mice.

In summary, this study has identified Cap2 as a new factor important for lens biology. Cap2 is downstream of several factors that control lens development and are linked to cataract. Along with the Tdrd7 data, these new findings on Cap2 offer strong support indicating that distinct mechanisms are functional in the control of the cellular morphology of maturing fiber cells at the beginning of the organelle degradation zone in the lens.

103

Chapter 6

DISCUSSION AND FUTURE DIRECTIONS

6.1 Introduction Deficiency of the gene TDRD7, encoding a conserved ribonucleoprotein/RNA granule component, causes congenital cataracts, which suggests its critical role in mediating post-transcriptional gene expression control in normal lens development and maintenance of transparency (Lachke and Maas, 2011; Lachke et al., 2011b; Tanaka et al., 2011a). My work in this dissertation has uncovered a novel role of Tdrd7 in sustaining optimal mRNA and protein expression levels of a stress-responsive chaperone, Hspb1, which may be play a role in F-actin cytoskeletal maintenance of lens fiber cells post-nuclear degradation. Further, I have performing high-throughput analyses of the transcriptome and the proteome of Tdrd7-/- lens at stage P15, which is near the manifestation of the cellular defects. This has allowed me to identify, among several new candidates, a long noncoding RNA Snhg12 (Yoon et al., 2013) that is significantly misexpressed in the Tdrd7-/- lens, and is therefore downstream in the Tdrd7 pathway. This finding provides further evidence that supports a role for post- transcriptional regulatory factors in lens development and whose misregulation results in early onset of cataract. In addition, this analyses also identified multiple cataract- linked genes to be downstream of Tdrd7, including both members of the intermediate filaments protein superfamily that forms beaded filaments in the lens, Bfsp1 and Bfsp2, thus suggesting a role for Tdrd7 in maintaining another aspect of lens cytoskeleton components. In my dissertation, I also examined one of the top candidate misregulated

104

in the Tdrd7-/- lens proteome, Cap2, by characterizing the lens of Cap2-/- mouse lens. Through this analysis, I have identified morphological defects in Cap2-/- mouse lens fiber cells that coincides with the area of lens fiber cell defects that were also described in the Tdrd7-/- mouse lens. This suggests that distinct factors and regulatory mechanisms function in the maintenance of the F-actin cytoskeleton in late stages of lens fiber cell maturation. I have included a detailed discussion of the significance of my findings at the end of each Chapter. Therefore, in this chapter, I will discuss some of my preliminary findings and propose future experiments and approaches that would likely enrich our understanding of Tdrd7 function in the lens.

6.2 Identification of Tdrd7 RNA targets in the lens by RNA- immunoprecipitation coupled with RNA-sequencing (RIP-seq) Co-staining of Tdrd7 and Pyronin Y, a dye that preferentially binds to RNA, suggests that Tdrd7 protein closely localizes with RNA in the lens (Lachke et al., 2011b). In Chapter 3 (section 3.2.4) I have shown that Tdrd7 protein associates with

Hspb1 mRNA in the lens by both RIP assay and single RNA FISH coupled with immunofluorescence. Additionally, I have identified a new Tdrd7 target, Epsin (Epn1), using a candidate gene approach (Figure 6.1). Epn1, which I found to be reduced in the Tdrd7-/- lens, is important for creating membrane curvatures in cells (Lee et al., 2000; Lo et al., 1997). To gain further insights into the repertoire of Tdrd7 RNA targets on the global level in the lens, a Tdrd7-specific RIP assay coupled with high-throughput RNA- sequencing (RIP-seq) screen should be undertaken in the future. This experiment is especially important as it is currently not known whether Tdrd7 only associates with mRNA in the lens or whether it also associates with other classes of RNA such as non-

105

coding RNA. RIP-seq will address this critical knowledge gap on a whole-genome level and further advance our understanding of the regulatory role of Tdrd7 in the lens.

Figure 6.1 Tdrd7 protein is associated with Epn1 mRNA in the lens A. RNA-immunoprecipitation (RIP) coupled with RT-PCR using Tdrd7 antibody was performed for wild-type mouse lenses at stage P15. This shows enrichment of Epn1 transcripts in the Tdrd7-specific pulldown but not the IgG control. (B) RT-qPCR analysis, that indicates the down regulation of Epn2 in Tdrd7-/- lens. Asterisk indicates Pvalue < 0.05.

6.3 Validation of Tdrd7-protein binding partners in the lens identified by Tdrd7 pull-down followed by mass spectrometry The tudor domains in Tdrd7 are predicted to interact with methylated arginine or lysine residues within other proteins. However, the identity of the Tdrd7 binding

106

protein partners in the lens is presently not known. To uncover the potential Tdrd7 protein binding partners in the lens, I have successfully optimized and performed a Tdrd7 pull-down in wild type ICR mice (P8) lenses using Tdrd7-specific antibody, followed by mass spectrometry as described in (section 2.13).

Figure 6.2 Tdrd7 pull down from from wild type mouse lense Lysates from ICR mice lens at Postnatal stage (P8) were used to perform Tdrd7 pull-down. The western blot indicates a succsfull pulldown. Tdrd7 band pears in Tdrd7-IP and the lens lysate 10% input but it is not present in the negative control pulldown of IgG and the no antibody/beads only

This experiment has yielded a list of proteins (The List of proteins identified in Tdrd7 pull-down can be found in Error! Reference source not found.) that were pulled down specifically with Tdrd7 protein but not in the negative control (the IgG antibody pulldown). Future experiments should be aimed at validating these Tdrd7 protein binding partners in the lens. This can be achieved by confirming their association with Tdrd7 in lens lysates by performing an independent co-immunoprecipitation assay with Tdrd7-specific antibody to identify the target, followed by reciprocal assay using a specific antibody against the target to identify Tdrd7. This approach would confirm the association of the proteins in the lens lysate. Additionally, a co-expression network based on the Tdrd7 binding partners in the lens can be derived. Such a network can be constructed by obtaining the normal

107

expression values during lens development from iSyTE 2.0 and calculating the correlation of the expression of these candidate proteins to the expression of Tdrd7. This network will reveal the closely associated protein “neighbors” of Tdrd7 (i.e. those connected by direct edges) as well as its secondary protein “neighbors” (i.e. those connected in an indirect manner).

6.4 Investigate the mechanistic basis of Tdrd7 regulation of misregulated downstream candidates identified in Tdrd7-/- lens

Tdrd7, a ribonucleoprotein/RNA granule component, can likely mediate its downstream posttranscriptional regulation through a variety of mechanisms. For example it is possible that Tdrd7 may, depending on its protein partners, stabilize its target mRNAs or may even channel it to degradation (Figure 1.7) (Anderson and Kedersha, 2009). To understand the specific mechanism of Tdrd7 control on its mRNA targets, shRNA based Tdrd7 known-down in lens epithelium-derived cell lines (e.g. SRA01/04, 21EM15), can be generated for performing a decay assay. I have demonstrated that Tdrd7 can be successfully knocked down in the 21EM15 cell line, which is a suitable model for studying specific aspects of lens biology as it is known to endogenously express several key lens genes (Terrell et al., 2015). The RNA-decay assay can be performed using Actinomycin D, a global transcription inhibitor (Koba and Konopa, 2005). In this assay the rate of decay of Tdrd7 mRNA targets in control, Tdrd7-KD or Tdrd7 overexpression conditions will be quantified. This will be achieved by collecting total RNA from a temporal series of cells in the above different conditions that are subjected to Actinomycin D treatment. The amount of Tdrd7 mRNA target remaining over time after Actinomycin D treatment will be estimated by RT-qPCR in

108

treatment and control and compared to a housekeeping mRNA (e.g. Gapdh, Hprt, etc.) (Chen et al., 2008).

6.5 Generating double knockouts mice to test genetic interactions of Tdrd7 Targeted knockout mouse models are an effective model that aids in understanding complex molecular genetic interactions in vivo, which is not possible by in vitro assays. Several Tdrd7-/- misregulated candidates have knockout mouse models available. Thus future experiments can be aimed at generating double knockout mouse models by breeding Tdrd7-/- mice to other mouse models such as Cap2-/- mice or Bfsp1-/- mice (Field et al., 2015; Fischer et al., 2003; Sandilands et al., 2004). These studies will expand our knowledge on the mechanism of Tdrd7 function and its potential genetic interaction with its downstream targets. This will contribute to building the Tdrd7 regulatory network in the lens. For example, if I hypothesize that the upregulation of Cap2 protein in Tdrd7-/- lens is a response to correct F-actin in Tdrd7-/- lens, then probably a Tdrd7-/- and Cap2-/- double knockout mouse would exhibit sever lens fiber cell defect. Conversely, it possible that the double knockout could normalize (rescue) the phenotype via compensatory mechanism of upregulating genes with similar functions.

6.6 Study the biomechanical properties of Tdrd7-/- lens It has been demonstrated that lens fiber cell organization is important for biomechanical properties of the lens, such as stiffness and resilience. Beaded filament have also been shown to be an important determinant of these biomechanical properties (Fudge et al., 2011; Gokhin et al., 2012). My data shows that Tdrd7-/- lenses exhibit, in addition to abnormal lens fiber cell architecture (Barnum et al. Al Saai, 2020), the

109

misregulation of intermediate filaments (Section 4.2.2). I hypothesize that these factors would affect the biomechanical properties of Tdrd7-/- lenses. Therefore, future studies can test this hypothesis by applying sequential glass coverslips on Tdrd7-/- lens as described by (Cheng et al., 2016b) to assess the proposed change in their biomechanical properties. These assays will lead to the understanding of how Tdrd7 deficiency impacts the stiffness and other biomechanical properties of the lens.

110

REFERENCES

Agrawal, S.A., Anand, D., Siddam, A.D., Kakrana, A., Dash, S., Scheiblin, D.A., Dang, C.A., Terrell, A.M., Waters, S.M., Singh, A., et al. (2015). Compound mouse mutants of bZIP transcription factors Mafg and Mafk reveal a regulatory network of non-crystallin genes associated with cataract. Hum. Genet. 134, 717–735.

Al-Ghoul, K.J., Kirk, T., Kuszak, A.J., Zoltoski, R.K., Shiels, A., and Kuszak, J.R. (2003). Lens structure in MIP-deficient mice. Anat. Rec. A. Discov. Mol. Cell. Evol. Biol. 273, 714–730.

Alizadeh, A., Clark, J., Seeberger, T., Hess, J., Blankenship, T., and FitzGerald, P.G. (2003). Targeted Deletion of the Lens Fiber Cell–Specific Intermediate Filament Protein Filensin. Invest. Ophthalmol. Vis. Sci. 44, 5252–5258.

Anand, D., Kakrana, A., Siddam, A.D., Huang, H., Saadi, I., and Lachke, S.A. (2018). RNA sequencing-based transcriptomic profiles of embryonic lens development for cataract gene discovery. Hum. Genet. 137, 941–954.

Anantharaman, V., Zhang, D., and Aravind, L. (2010). OST-HTH: a novel predicted RNA-binding domain. Biol. Direct 5, 13.

Anders, S., Pyl, P.T., and Huber, W. (2015). HTSeq--a Python framework to work with high-throughput sequencing data. Bioinforma. Oxf. Engl. 31, 166–169.

Anderson, P., and Kedersha, N. (2006). RNA granules. J. Cell Biol. 172, 803–808.

Anderson, P., and Kedersha, N. (2009). RNA granules: post-transcriptional and epigenetic modulators of gene expression. Nat. Rev. Mol. Cell Biol. 10, 430–436.

Arrigo, A.-P. (2017). Mammalian HspB1 (Hsp27) is a molecular sensor linked to the physiology and environment of the cell. Cell Stress Chaperones 22, 517–529.

Aryal, S., Viet, J., Weatherbee, B.A.T., Siddam, A.D., Hernandez, F.G., Gautier- Courteille, C., Paillard, L., and Lachke, S.A. (2020a). The cataract-linked RNA- binding protein Celf1 post-transcriptionally controls the spatiotemporal expression of the key homeodomain transcription factors Pax6 and Prox1 in lens development. Hum. Genet.

111

Aryal, S., Anand, D., Hernandez, F.G., Weatherbee, B.A.T., Huang, H., Reddy, A.P., Wilmarth, P.A., David, L.L., and Lachke, S.A. (2020b). MS/MS in silico subtraction- based proteomic profiling as an approach to facilitate disease gene discovery: application to lens development and cataract. Hum. Genet. 139, 151–184.

Aspit, L., Levitas, A., Etzion, S., Krymko, H., Slanovic, L., Zarivach, R., Etzion, Y., and Parvari, R. (2019). CAP2 mutation leads to impaired actin dynamics and associates with supraventricular tachycardia and dilated cardiomyopathy. J. Med. Genet. 56, 228–235.

Audette, D.S., Scheiblin, D.A., and Duncan, M.K. (2017). The molecular mechanisms underlying lens fiber elongation. Exp. Eye Res. 156, 41–49.

Aulas, A., Fay, M.M., Lyons, S.M., Achorn, C.A., Kedersha, N., Anderson, P., and Ivanov, P. (2017). Stress-specific differences in assembly and composition of stress granules and related foci. J Cell Sci jcs.199240.

Bagchi, M., Katar, M., and Maisel, H. (2002). Heat shock proteins of adult and embryonic human ocular lenses. J. Cell. Biochem. 84, 278–284.

Bagchi, M., Katar, M., Lo, W.-K., Yost, R., Hill, C., and Maisel, H. (2004). ERM proteins of the lens. J. Cell. Biochem. 92, 626–630.

Baranova, E.V., Weeks, S.D., Beelen, S., Bukach, O.V., Gusev, N.B., and Strelkov, S.V. (2011). Three-dimensional structure of α-crystallin domain dimers of human small heat shock proteins HSPB1 and HSPB6. J. Mol. Biol. 411, 110–122.

Bardram, M., and Brændsdrup, P. (1947). MATERNAL RUBELLA DURING PREGNANCY AS CAUSE OF CONGENITAL CATARACT AND OTHER CONGENITAL MALFORMATIONS. Acta Ophthalmol. (Copenh.) 25, 353–367.

Barnum, C.E. (2015). Phenotypic and molecular characterization of cataract in the Tdrd7 knockout mouse mutant. M.S. University of Delaware.

Barnum, C.E., Al Saai, S., Patel, S.D., Cheng, C., Anand, D., Xu, X., Dash, S., Siddam, A.D., Glazewski, L., Paglione, E., et al. (2020). The Tudor-domain protein TDRD7, mutated in congenital cataract, controls the heat shock protein HSPB1 (HSP27) and lens fiber cell morphology. Hum. Mol. Genet.

Bassnett, S. (2009a). On the mechanism of organelle degradation in the vertebrate lens. Exp. Eye Res. 88, 133–139.

Bassnett, S. (2009b). On the mechanism of organelle degradation in the vertebrate lens. Exp. Eye Res. 88, 133–139.

112

Bassnett, S., Missey, H., and Vucemilo, I. (1999). Molecular architecture of the lens fiber cell basal membrane complex. J. Cell Sci. 112, 2155–2165.

Bassnett, S., Shi, Y., and Vrensen, G.F.J.M. (2011a). Biological glass: structural determinants of eye lens transparency. Philos. Trans. R. Soc. Lond. B Biol. Sci. 366, 1250–1264.

Bassnett, S., Shi, Y., and Vrensen, G.F.J.M. (2011b). Biological glass: structural determinants of eye lens transparency. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 366, 1250–1264.

Batish, M., Raj, A., and Tyagi, S. (2011). Single molecule imaging of RNA in situ. Methods Mol. Biol. Clifton NJ 714, 3–13.

Batish, M., van den Bogaard, P., Kramer, F.R., and Tyagi, S. (2012). Neuronal mRNAs travel singly into dendrites. Proc. Natl. Acad. Sci. U. S. A. 109, 4645–4650.

Bayer, L.V., Batish, M., Formel, S.K., and Bratu, D.P. (2015). Single-Molecule RNA In Situ Hybridization (smFISH) and Immunofluorescence (IF) in the Drosophila Egg Chamber. Methods Mol. Biol. Clifton NJ 1328, 125–136.

Bazan, H.E. (1989). The synthesis and effects of eicosanoids in avascular ocular tissues. Prog. Clin. Biol. Res. 312, 73–84.

Benlali, A., Draskovic, I., Hazelett, D.J., and Treisman, J.E. (2000). act up Controls Actin Polymerization to Alter Cell Shape and Restrict Hedgehog Signaling in the Drosophila Eye Disc. Cell 101, 271–281.

Bennett, T.M., Zhou, Y., and Shiels, A. (2016). Lens transcriptome profile during cataract development in Mip-null mice. Biochem. Biophys. Res. Commun. 478, 988– 993.

Bertling, E., Hotulainen, P., Mattila, P.K., Matilainen, T., Salminen, M., and Lappalainen, P. (2004). Cyclase-associated Protein 1 (CAP1) Promotes Cofilin- induced Actin Dynamics in Mammalian Nonmuscle Cells. Mol. Biol. Cell 15, 2324– 2334.

Beyer, T.L., Vogler, G., Sharma, D., and O’Donnell, F.E. (1984). Protective barrier effect of the posterior lens capsule in exogenous bacterial endophthalmitis--an experimental primate study. Invest. Ophthalmol. Vis. Sci. 25, 108–112.

Bhat, S.P. (2001). The Ocular Lens Epithelium. Biosci. Rep. 21, 537–563.

113

Blanchoin, L., Boujemaa-Paterski, R., Sykes, C., and Plastino, J. (2014). Actin Dynamics, Architecture, and Mechanics in Cell Motility. Physiol. Rev. 94, 235–263.

Blankenship, T.N., Hess, J.F., and FitzGerald, P.G. (2001). Development- and differentiation-dependent reorganization of intermediate filaments in fiber cells. Invest. Ophthalmol. Vis. Sci. 42, 735–742.

Bretscher, A., Chambers, D., Nguyen, R., and Reczek, D. (2000). ERM-Merlin and EBP50 protein families in plasma membrane organization and function. Annu. Rev. Cell Dev. Biol. 16, 113–143.

Callebaut, I., and Mornon, J.-P. (2010). LOTUS, a new domain associated with small RNA pathways in the germline. Bioinforma. Oxf. Engl. 26, 1140–1144.

Celestino-Soper, P.B., Skinner, C., Schroer, R., Eng, P., Shenai, J., Nowaczyk, M.M., Terespolsky, D., Cushing, D., Patel, G.S., Immken, L., et al. (2012). Deletions in chromosome 6p22.3-p24.3, including ATXN1, are associated with developmental delay and autism spectrum disorders. Mol. Cytogenet. 5, 17.

Chen, C., Nott, T.J., Jin, J., and Pawson, T. (2011). Deciphering arginine methylation: Tudor tells the tale. Nat. Rev. Mol. Cell Biol. 12, 629–642.

Chen, C.-Y.A., Ezzeddine, N., and Shyu, A.-B. (2008). Messenger RNA Half-Life Measurements in Mammalian Cells. Methods Enzymol. 448, 335.

Chen, J., Wang, Q., Cabrera, P.E., Zhong, Z., Sun, W., Jiao, X., Chen, Y., Govindarajan, G., Naeem, M.A., Khan, S.N., et al. (2017a). Molecular Genetic Analysis of Pakistani Families With Autosomal Recessive Congenital Cataracts by Homozygosity Screening. Invest. Ophthalmol. Vis. Sci. 58, 2207–2217.

Chen, J., Wang, Q., Cabrera, P.E., Zhong, Z., Sun, W., Jiao, X., Chen, Y., Govindarajan, G., Naeem, M.A., Khan, S.N., et al. (2017b). Molecular Genetic Analysis of Pakistani Families With Autosomal Recessive Congenital Cataracts by Homozygosity Screening. Invest. Ophthalmol. Vis. Sci. 58, 2207–2217.

Chen, Q., Zhou, W., Du, S.-Q., Gong, D.-X., Li, J., Bi, J.-B., Li, Z.-H., Zhang, Z., Li, Z.-L., Liu, X.-K., et al. (2019). Overexpression of SNHG12 regulates the viability and invasion of renal cell carcinoma cells through modulation of HIF1α. Cancer Cell Int. 19, 128.

Cheng, C., and Gong, X. (2011). Diverse roles of Eph/ephrin signaling in the mouse lens. PloS One 6, e28147.

114

Cheng, C., Ansari, M.M., Cooper, J.A., and Gong, X. (2013). EphA2 and Src regulate equatorial cell morphogenesis during lens development. Dev. Camb. Engl. 140, 4237– 4245.

Cheng, C., Nowak, R.B., Biswas, S.K., Lo, W.-K., FitzGerald, P.G., and Fowler, V.M. (2016a). Tropomodulin 1 Regulation of Actin Is Required for the Formation of Large Paddle Protrusions Between Mature Lens Fiber Cells. Invest. Ophthalmol. Vis. Sci. 57, 4084–4099.

Cheng, C., Gokhin, D.S., Nowak, R.B., and Fowler, V.M. (2016b). Sequential Application of Glass Coverslips to Assess the Compressive Stiffness of the Mouse Lens: Strain and Morphometric Analyses. J. Vis. Exp. JoVE.

Cheng, C., Nowak, R.B., and Fowler, V.M. (2017). The lens actin filament cytoskeleton: Diverse structures for complex functions. Exp. Eye Res. 156, 58–71.

Chereau, D., Kerff, F., Graceffa, P., Grabarek, Z., Langsetmo, K., and Dominguez, R. (2005). Actin-bound structures of Wiskott-Aldrich syndrome protein (WASP)- homology domain 2 and the implications for filament assembly. Proc. Natl. Acad. Sci. U. S. A. 102, 16644–16649.

Conte, I., Carrella, S., Avellino, R., Karali, M., Marco-Ferreres, R., Bovolenta, P., and Banfi, S. (2010). miR-204 is required for lens and retinal development via Meis2 targeting. Proc. Natl. Acad. Sci. U. S. A. 107, 15491–15496.

Côté, J., and Richard, S. (2005). Tudor Domains Bind Symmetrical Dimethylated Arginines. J. Biol. Chem. 280, 28476–28483.

Cui, G., Botuyan, M.V., and Mer, G. (2013). (1)H, (15)N and (13)C resonance assignments for the three LOTUS RNA binding domains of Tudor domain-containing protein TDRD7. Biomol. NMR Assign. 7, 79–83.

Cui, X., Gao, L., Jin, Y., Zhang, Y., Bai, J., Feng, G., Gao, W., Liu, P., He, L., and Fu, S. (2007). The E233del mutation in BFSP2 causes a progressive autosomal dominant congenital cataract in a Chinese family. Mol. Vis. 13, 2023–2029.

Cvekl, A., and Zhang, X. (2017). Signaling and Gene Regulatory Networks in Mammalian Lens Development. Trends Genet. TIG 33, 677–702.

Danysh, B.P., and Duncan, M.K. (2009). The Lens Capsule. Exp. Eye Res. 88, 151– 164.

115

Dash, S., Dang, C.A., Beebe, D.C., and Lachke, S.A. (2015). Deficiency of the RNA binding protein caprin2 causes lens defects and features of peters anomaly. Dev. Dyn. Off. Publ. Am. Assoc. Anat. 244, 1313–1327.

Dash, S., Siddam, A.D., Barnum, C.E., Janga, S.C., and Lachke, S.A. (2016). RNA- binding proteins in eye development and disease: implication of conserved RNA granule components. Wiley Interdiscip. Rev. RNA 7, 527–557.

Dash, S., Brastrom, L.K., Patel, S.D., Scott, C.A., Slusarski, D.C., and Lachke, S.A. (2020). The master transcription factor SOX2, mutated in anophthalmia/microphthalmia, is post-transcriptionally regulated by the conserved RNA-binding protein RBM24 in vertebrate eye development. Hum. Mol. Genet. 29, 591–604.

Delanote, V., Vandekerckhove, J., and Gettemans, J. (2005). Plastins: versatile modulators of actin organization in (patho)physiological cellular processes. Acta Pharmacol. Sin. 26, 769–779.

Derrien, T., Johnson, R., Bussotti, G., Tanzer, A., Djebali, S., Tilgner, H., Guernec, G., Martin, D., Merkel, A., Knowles, D.G., et al. (2012). The GENCODE v7 catalog of human long noncoding RNAs: Analysis of their gene structure, evolution, and expression. Genome Res. 22, 1775–1789.

Duncan, M.K., Cvekl, A., Li, X., and Piatigorsky, J. (2000). Truncated Forms of Pax-6 Disrupt Lens Morphology in Transgenic Mice. Invest. Ophthalmol. Vis. Sci. 41, 464– 473.

Elsobky, S., Crane, A.M., Margolis, M., Carreon, T.A., and Bhattacharya, S.K. (2014). Review of application of mass spectrometry for analyses of anterior eye proteome. World J. Biol. Chem. 5, 106–114.

Farnsworth, P.N., Shyne, S.E., Caputo, S.J., Fasano, A.V., and Spector, A. (1980). Microtubules: a major cytoskeletal component of the human lens. Exp. Eye Res. 30, 611–615.

Field, J., Vojtek, A., Ballester, R., Bolger, G., Colicelli, J., Ferguson, K., Gerst, J., Kataoka, T., Michaeli, T., Powers, S., et al. (1990). Cloning and characterization of CAP, the S. cerevisiae gene encoding the 70 kd adenylyl cyclase-associated protein. Cell 61, 319–327.

Field, J., Ye, D.Z., Shinde, M., Liu, F., Schillinger, K.J., Lu, M., Wang, T., Skettini, M., Xiong, Y., Brice, A.K., et al. (2015). CAP2 in cardiac conduction, sudden cardiac death and eye development. Sci. Rep. 5, 17256.

116

Field, J., Ye, D.Z., Shinde, M., Liu, F., Schillinger, K.J., Lu, M., Wang, T., Skettini, M., Xiong, Y., Brice, A.K., et al. (2016). Erratum: CAP2 in cardiac conduction, sudden cardiac death and eye development. Sci. Rep. 6, 26640.

Fischer, R.S., Quinlan, R.A., and Fowler, V.M. (2003). Tropomodulin binds to filensin intermediate filaments. FEBS Lett. 547, 228–232.

Friedenwald, J.S. (1930). The Permeability of the Lens Capsule to Water, Dextrose, and Other Sugars. Trans. Am. Ophthalmol. Soc. 28, 195–211.

Fu, Q., Qin, Z., Zhang, L., and Yao, K. (2017). The Long noncoding RNA p10540 regulates autophagy through LC3B during human lens development. Invest. Ophthalmol. Vis. Sci. 58, 1211–1211.

Fudge, D.S., McCuaig, J.V., Stralen, S.V., Hess, J.F., Wang, H., Mathias, R.T., and FitzGerald, P.G. (2011). Intermediate Filaments Regulate Tissue Size and Stiffness in the Murine Lens. Invest. Ophthalmol. Vis. Sci. 52, 3860–3867.

Fujimura, N. (2016). WNT/β-Catenin Signaling in Vertebrate Eye Development. Front. Cell Dev. Biol. 4.

Gokhin, D.S., Nowak, R.B., Kim, N.E., Arnett, E.E., Chen, A.C., Sah, R.L., Clark, J.I., and Fowler, V.M. (2012). Tmod1 and CP49 synergize to control the fiber cell geometry, transparency, and mechanical stiffness of the mouse lens. PloS One 7, e48734.

Gold, M.G., Reichow, S.L., O’Neill, S.E., Weisbrod, C.R., Langeberg, L.K., Bruce, J.E., Gonen, T., and Scott, J.D. (2012). AKAP2 anchors PKA with aquaporin-0 to support ocular lens transparency. EMBO Mol. Med. 4, 15–26.

Graw, J. (2009). Genetics of crystallins: cataract and beyond. Exp. Eye Res. 88, 173– 189.

Haider, S., and Pal, R. (2013). Integrated Analysis of Transcriptomic and Proteomic Data. Curr. Genomics 14, 91–110.

Haslbeck, M., Franzmann, T., Weinfurtner, D., and Buchner, J. (2005). Some like it hot: the structure and function of small heat-shock proteins. Nat. Struct. Mol. Biol. 12, 842–846.

Hejtmancik, J.F. (2008). Congenital Cataracts and their Molecular Genetics. Semin. Cell Dev. Biol. 19, 134–149.

117

Higgs, H.N., and Pollard, T.D. (2001). Regulation of actin filament network formation through ARP2/3 complex: activation by a diverse array of proteins. Annu. Rev. Biochem. 70, 649–676.

Hirose, T., Kawabuchi, M., Tamaru, T., Okumura, N., Nagai, K., and Okada, M. (2000). Identification of tudor repeat associator with PCTAIRE 2 (Trap). A novel protein that interacts with the N-terminal domain of PCTAIRE 2 in rat brain. Eur. J. Biochem. 267, 2113–2121.

Hoffman, L., Jensen, C.C., Yoshigi, M., and Beckerle, M. (2017). Mechanical signals activate p38 MAPK pathway-dependent reinforcement of actin via mechanosensitive HspB1. Mol. Biol. Cell 28, 2661–2675.

Hosokawa, M., Shoji, M., Kitamura, K., Tanaka, T., Noce, T., Chuma, S., and Nakatsuji, N. (2007a). Tudor-related proteins TDRD1/MTR-1, TDRD6 and TDRD7/TRAP: Domain composition, intracellular localization, and function in male germ cells in mice. Dev. Biol. 301, 38–52.

Hosokawa, M., Shoji, M., Kitamura, K., Tanaka, T., Noce, T., Chuma, S., and Nakatsuji, N. (2007b). Tudor-related proteins TDRD1/MTR-1, TDRD6 and TDRD7/TRAP: domain composition, intracellular localization, and function in male germ cells in mice. Dev. Biol. 301, 38–52.

Hutchins, E.D., Eckalbar, W.L., Wolter, J.M., Mangone, M., and Kusumi, K. (2016). Differential expression of conserved and novel microRNAs during tail regeneration in the lizard Anolis carolinensis. BMC Genomics 17.

Jakob, U., Gaestel, M., Engel, K., and Buchner, J. (1993). Small heat shock proteins are molecular chaperones. J. Biol. Chem. 268, 1517–1520.

Jakobs, P.M., Hess, J.F., FitzGerald, P.G., Kramer, P., Weleber, R.G., and Litt, M. (2000). Autosomal-Dominant Congenital Cataract Associated with a Deletion Mutation in the Human Beaded Filament Protein Gene BFSP2. Am. J. Hum. Genet. 66, 1432–1436.

Jeske, M., Bordi, M., Glatt, S., Müller, S., Rybin, V., Müller, C.W., and Ephrussi, A. (2015). The Crystal Structure of the Drosophila Germline Inducer Oskar Identifies Two Domains with Distinct Vasa Helicase- and RNA-Binding Activities. Cell Rep. 12, 587–598.

Jeske, M., Müller, C.W., and Ephrussi, A. (2017). The LOTUS domain is a conserved DEAD-box RNA helicase regulator essential for the recruitment of Vasa to the germ plasm and nuage. Genes Dev. 31, 939–952.

118

Jiang, H., Xu, L., Wang, Z., Keene, J., and Gu, Z. (2014). Coordinating expression of RNA binding proteins with their mRNA targets. Sci. Rep. 4, 7175.

Jun, G., Guo, H., Klein, B.E.K., Klein, R., Wang, J.J., Mitchell, P., Miao, H., Lee, K.E., Joshi, T., Buck, M., et al. (2009). EPHA2 Is Associated with Age-Related Cortical Cataract in Mice and Humans. PLoS Genet. 5.

Kakrana, A., Yang, A., Anand, D., Djordjevic, D., Ramachandruni, D., Singh, A., Huang, H., Ho, J.W.K., and Lachke, S.A. (2018). iSyTE 2.0: a database for expression-based gene discovery in the eye. Nucleic Acids Res. 46, D875–D885.

Kamphuis, W., Dijk, F., Kraan, W., and Bergen, A.A.B. (2007). Transfer of lens- specific transcripts to retinal RNA samples may underlie observed changes in crystallin-gene transcript levels after ischemia. Mol. Vis. 13, 220–228.

Kawamukai, M., Gerst, J., Field, J., Riggs, M., Rodgers, L., Wigler, M., and Young, D. (1992). Genetic and biochemical analysis of the adenylyl cyclase-associated protein, cap, in Schizosaccharomyces pombe. Mol. Biol. Cell 3, 167–180.

Kepser, L.-J., Damar, F., Cicco, T.D., Chaponnier, C., Prószyński, T.J., Pagenstecher, A., and Rust, M.B. (2019). CAP2 deficiency delays myofibril actin cytoskeleton differentiation and disturbs skeletal muscle architecture and function. Proc. Natl. Acad. Sci. 116, 8397–8402.

Kim, D., Langmead, B., and Salzberg, S.L. (2015). HISAT: a fast spliced aligner with low memory requirements. Nat. Methods 12, 357–360.

Kim, D., Paggi, J.M., Park, C., Bennett, C., and Salzberg, S.L. (2019). Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat. Biotechnol. 37, 907–915.

Kirfel, G. (2004). Cell migration: mechanisms of rear detachment and the formation of migration tracks. Eur. J. Cell Biol. 83, 717–724.

Kishore, S., Luber, S., and Zavolan, M. (2010). Deciphering the role of RNA-binding proteins in the post-transcriptional control of gene expression. Brief. Funct. Genomics 9, 391–404.

Kistler, J., Gilbert, K., Brooks, H.V., Jolly, R.D., Hopcroft, D.H., and Bullivant, S. (1986). Membrane interlocking domains in the lens. Invest. Ophthalmol. Vis. Sci. 27, 1527–1534.

Koba, M., and Konopa, J. (2005). [Actinomycin D and its mechanisms of action]. Postepy Hig. Med. Doswiadczalnej Online 59, 290–298.

119

Koroma, B.M., Yang, J.M., and Sundin, O.H. (1997). The Pax-6 homeobox gene is expressed throughout the corneal and conjunctival epithelia. Invest. Ophthalmol. Vis. Sci. 38, 108–120.

Kosmas, K., Eskandarnaz, A., Khorsandi, A.B., Kumar, A., Ranjan, R., Eming, S.A., Noegel, A.A., and Peche, V.S. (2015). CAP2 is a regulator of the actin cytoskeleton and its absence changes infiltration of inflammatory cells and contraction of wounds. Eur. J. Cell Biol. 94, 32–45.

Kotaja, N., and Sassone-Corsi, P. (2007). The chromatoid body: a germ-cell-specific RNA-processing centre. Nat. Rev. Mol. Cell Biol. 8, 85–90.

Kotaja, N., Bhattacharyya, S.N., Jaskiewicz, L., Kimmins, S., Parvinen, M., Filipowicz, W., and Sassone-Corsi, P. (2006). The chromatoid body of male germ cells: Similarity with processing bodies and presence of Dicer and microRNA pathway components. Proc. Natl. Acad. Sci. U. S. A. 103, 2647–2652.

Kozomara, A., and Griffiths-Jones, S. (2014). miRBase: annotating high confidence microRNAs using deep sequencing data. Nucleic Acids Res. 42, D68-73.

Kozomara, A., Birgaoanu, M., and Griffiths-Jones, S. (2019). miRBase: from microRNA sequences to function. Nucleic Acids Res. 47, D155–D162.

Kuhli-Hattenbach, C., Lüchtenberg, M., Kohnen, T., and Hattenbach, L.-O. (2008). Risk factors for complications after congenital cataract surgery without intraocular lens implantation in the first 18 months of life. Am. J. Ophthalmol. 146, 1–7.

Kumar, A., Paeger, L., Kosmas, K., Kloppenburg, P., Noegel, A.A., and Peche, V.S. (2016). Neuronal Actin Dynamics, Spine Density and Neuronal Dendritic Complexity Are Regulated by CAP2. Front. Cell. Neurosci. 10, 180.

Kuszak, J.R. (1995). The ultrastructure of epithelial and fiber cells in the crystalline lens. Int. Rev. Cytol. 163, 305–350.

Kuszak, J.R., Zoltoski, R.K., and Sivertson, C. (2004). Fibre cell organization in crystalline lenses. Exp. Eye Res. 78, 673–687.

Kuwabara, T. (1968). Microtubules in the Lens. Arch. Ophthalmol. 79, 189–195.

Lachke, S.A., and Maas, R.L. (2011). RNA Granules and Cataract. Expert Rev. Ophthalmol. 6, 497–500.

120

Lachke, S.A., Alkuraya, F.S., Kneeland, S.C., Ohn, T., Aboukhalil, A., Howell, G.R., Saadi, I., Cavallesco, R., Yue, Y., Tsai, A.C.-H., et al. (2011a). Mutations in the RNA granule component TDRD7 cause cataract and glaucoma. Science 331, 1571–1576.

Lachke, S.A., Alkuraya, F.S., Kneeland, S.C., Ohn, T., Aboukhalil, A., Howell, G.R., Saadi, I., Cavallesco, R., Yue, Y., Tsai, A.C.-H., et al. (2011b). Mutations in the RNA granule component TDRD7 cause cataract and glaucoma. Science 331, 1571–1576.

Lachke, S.A., Ho, J.W.K., Kryukov, G.V., O’Connell, D.J., Aboukhalil, A., Bulyk, M.L., Park, P.J., and Maas, R.L. (2012). iSyTE: integrated Systems Tool for Eye gene discovery. Invest. Ophthalmol. Vis. Sci. 53, 1617–1627.

Lee, H.-J., and Zheng, J.J. (2010). PDZ domains and their binding partners: structure, specificity, and modification. Cell Commun. Signal. CCS 8, 8.

Lee, A., Fischer, R.S., and Fowler, V.M. (2000). Stabilization and remodeling of the membrane skeleton during lens fiber cell differentiation and maturation. Dev. Dyn. 217, 257–270.

Leonard, M., Zhang, L., Zhai, N., Cader, A., Chan, Y., Nowak, R.B., Fowler, V.M., and Menko, A.S. (2011). Modulation of N-cadherin junctions and their role as epicenters of differentiation-specific actin regulation in the developing lens. Dev. Biol. 349, 363–377.

Li, Y., and Piatigorsky, J. (2009). Targeted deletion of Dicer disrupts lens morphogenesis, corneal epithelium stratification, and whole eye development. Dev. Dyn. Off. Publ. Am. Assoc. Anat. 238, 2388–2400.

Li, Y., Guo, S., Liu, W., Jin, T., Li, X., He, X., Zhang, X., Su, H., Zhang, N., and Duan, C. (2019). Silencing of SNHG12 Enhanced the Effectiveness of MSCs in Alleviating Ischemia/Reperfusion Injuries via the PI3K/AKT/mTOR Signaling Pathway. Front. Neurosci. 13, 645.

Lim, H., Agopian, A.J., Whitehead, L.W., Beasley, C.W., Langlois, P.H., Emery, R.J., and Waller, D.K. (2015). Maternal Occupational Exposure to Ionizing Radiation and Major Structural Birth Defects. Birt. Defects Res. A. Clin. Mol. Teratol. 103, 243– 254.

Lin, Q., Zhou, N., Zhang, N., Zhu, B., Hu, S., Zhou, Z., and Qi, Y. (2013). Genetic variations and polymorphisms in the ezrin gene are associated with age-related cataract. Mol. Vis. 19, 1572–1579.

Linder, P., and Jankowsky, E. (2011). From unwinding to clamping - the DEAD box RNA helicase family. Nat. Rev. Mol. Cell Biol. 12, 505–516.

121

Liu, L., Yu, R., Shi, Y., Dai, Y., Zeng, Z., Guo, X., Ji, Q., Wang, G., and Zhong, J. (2014). Transduced protein transduction domain linked HSP27 protected LECs against UVB radiation-induced damage. Exp. Eye Res. 120, 36–42.

Liu, X., Liu, C., Shan, K., Zhang, S., Lu, Y., Yan, B., and Luo, Y. (2018). Long Non- Coding RNA H19 Regulates Human Lens Epithelial Cells Function. Cell. Physiol. Biochem. Int. J. Exp. Cell. Physiol. Biochem. Pharmacol. 50, 246–260.

Lo, W.K., Mills, A., and Kuck, J.F. (1994). Actin filament bundles are associated with fiber gap junctions in the primate lens. Exp. Eye Res. 58, 189–196.

Lo, W.-K., Shaw, A.P., and Wen, X.-J. (1997). Actin Filament Bundles in Cortical Fiber Cells of the Rat Lens. Exp. Eye Res. 65, 691–701.

Lo, W.K., Biswas, S.K., Brako, L., Shiels, A., Gu, S., and Jiang, J.X. (2014). Aquaporin-0 targets interlocking domains to control the integrity and transparency of the eye lens., Aquaporin-0 Targets Interlocking Domains to Control the Integrity and Transparency of the Eye Lens. Investig. Ophthalmol. Vis. Sci. Investig. Ophthalmol. Vis. Sci. 55, 55, 1202, 1202–1212.

Logan, C.M., Rajakaruna, S., Bowen, C., Radice, G.L., Robinson, M.L., and Menko, A.S. (2017). N-cadherin regulates signaling mechanisms required for lens fiber cell elongation and lens morphogenesis. Dev. Biol. 428, 118–134.

Logan, C.M., Bowen, C.J., and Menko, A.S. (2018). Functional role for stable microtubules in lens fiber cell elongation. Exp. Cell Res. 362, 477–488.

Maddala, R., Chauhan, B.K., Walker, C., Zheng, Y., Robinson, M.L., Lang, R.A., and Rao, P.V. (2011a). Rac1 GTPase -deficient mouse lens exhibits defects in shape, suture formation, fiber cell migration and survival. Dev. Biol. 360, 30–43.

Maddala, R., Skiba, N.P., Lalane, R., Sherman, D.L., Brophy, P.J., and Rao, P.V. (2011b). Periaxin is required for hexagonal geometry and membrane organization of mature lens fibers. Dev. Biol. 357, 179–190.

Marcello, M.R., and Singson, A. (2011). Germline Determination: Don’t Mind the P Granules. Curr. Biol. 21, R155–R157.

Markey, F.B., Ruezinsky, W., Tyagi, S., and Batish, M. (2014). Fusion FISH imaging: single-molecule detection of gene fusion transcripts in situ. PloS One 9, e93488.

Masugi, Y., Tanese, K., Emoto, K., Yamazaki, K., Effendi, K., Funakoshi, T., Mori, M., and Sakamoto, M. (2015). Overexpression of adenylate cyclase-associated protein 2 is a novel prognostic marker in malignant melanoma. Pathol. Int. 65, 627–634.

122

McCarty, C.A., and Taylor, H.R. (2001). The genetics of cataract. Invest. Ophthalmol. Vis. Sci. 42, 1677–1678.

Mearow, K.M. (2013). Cell stress promotes the association of phosphorylated HspB1 with F-actin. PloS One 8, e68978.

Messina-Baas, O., and Cuevas-Covarrubias, S.A. (2017). Inherited Congenital Cataract: A Guide to Suspect the Genetic Etiology in the Cataract Genesis. Mol. Syndromol. 8, 58–78.

Mets, M.B. (2001). Eye manifestations of intrauterine infections. Ophthalmol. Clin. N. Am. 14, 521–531.

Mi, H., Muruganujan, A., Huang, X., Ebert, D., Mills, C., Guo, X., and Thomas, P.D. (2019). Protocol Update for large-scale genome and gene function analysis with the PANTHER classification system (v.14.0). Nat. Protoc. 14, 703–721.

Mymrikov, E.V., Seit-Nebi, A.S., and Gusev, N.B. (2011). Large potentials of small heat shock proteins. Physiol. Rev. 91, 1123–1159.

Ono, S. (2013). The role of cyclase-associated protein in regulating actin filament dynamics – more than a monomer-sequestration factor. J. Cell Sci. 126, 3249–3258.

Pandey, A., Yadav, S.K., Vishvkarma, R., Singh, B., Maikhuri, J.P., Rajender, S., and Gupta, G. (2019). The dynamics of gene expression during and post meiosis sets the sperm agenda. Mol. Reprod. Dev. 86, 1921–1939.

Peche, V., Shekar, S., Leichter, M., Korte, H., Schröder, R., Schleicher, M., Holak, T.A., Clemen, C.S., Ramanath-Y, B., Pfitzer, G., et al. (2007). CAP2, cyclase- associated protein 2, is a dual compartment protein. Cell. Mol. Life Sci. CMLS 64, 2702–2715.

Peche, V.S., Holak, T.A., Burgute, B.D., Kosmas, K., Kale, S.P., Wunderlich, F.T., Elhamine, F., Stehle, R., Pfitzer, G., Nohroudi, K., et al. (2013). Ablation of cyclase- associated protein 2 (CAP2) leads to cardiomyopathy. Cell. Mol. Life Sci. 70, 527– 543.

Pek, J.W., Anand, A., and Kai, T. (2012). Tudor domain proteins in development. Dev. Camb. Engl. 139, 2255–2266.

Perng, M.-D., Zhang, Q., and Quinlan, R.A. (2007). Insights into the beaded filament of the eye lens. Exp. Cell Res. 313, 2180–2188.

123

Piatigorsky, J. (1981). Lens Differentiation in Vertebrates. Differentiation 19, 134– 153.

Ramachandran, R.D., Perumalsamy, V., and Hejtmancik, J.F. (2007). Autosomal recessive juvenile onset cataract associated with mutation in BFSP1. Hum. Genet. 121, 475–482.

Ramaekers, F.C.S., Boomkens, T.R., and Bloemendal, H. (1981). Cytoskeletal and contractile structures in bovine lens cell differentiation. Exp. Cell Res. 135, 454–461.

Rao, P.V., and Maddala, R. (2006a). The Role of the Lens Actin Cytoskeleton in Fiber Cell Elongation and Differentiation. Semin. Cell Dev. Biol. 17, 698–711.

Rao, P.V., and Maddala, R. (2006b). The Role of the Lens Actin Cytoskeleton in Fiber Cell Elongation and Differentiation. Semin. Cell Dev. Biol. 17, 698–711.

Robinson, M.D., McCarthy, D.J., and Smyth, G.K. (2010). edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinforma. Oxf. Engl. 26, 139–140.

Roscito, J.G., Subramanian, K., Naumann, R., Sarov, M., Shevchenko, A., Bogdanova, A., Kurth, T., Foerster, L., Kreysing, M., and Hiller, M. (2020). Recapitulating evolutionary divergence in a single cis-regulatory element is sufficient to cause expression changes of the lens gene Tdrd7. BioRxiv 2020.03.22.002535.

Saker, Z., Bahmad, H.F., Fares, Y., Al Najjar, Z., Saad, M., Harati, H., and Nabha, S. (2020). Prognostic impact of adenylyl cyclase-associated protein 2 (CAP2) in glioma: A clinicopathological study. Heliyon 6.

Sandilands, A., Wang, X., Hutcheson, A.M., James, J., Prescott, A.R., Wegener, A., Pekny, M., Gong, X., and Quinlan, R.A. (2004). Bfsp2 mutation found in mouse 129 strains causes the loss of CP49’ and induces vimentin-dependent changes in the lens fibre cell cytoskeleton. Exp. Eye Res. 78, 875–889.

Semba, R.D., Enghild, J.J., Venkatraman, V., Dyrlund, T.F., and Van Eyk, J.E. (2013). The Human Eye Proteome Project: Perspectives on an emerging proteome. Proteomics 13, 2500–2511.

Senthil Kumar, G., Kyle, J.W., Minogue, P.J., Dinesh Kumar, K., Vasantha, K., Berthoud, V.M., Beyer, E.C., and Santhiya, S.T. (2013). An MIP/AQP0 mutation with impaired trafficking and function underlies an autosomal dominant congenital lamellar cataract. Exp. Eye Res. 110, 136–141.

124

Shaham, O., Gueta, K., Mor, E., Oren-Giladi, P., Grinberg, D., Xie, Q., Cvekl, A., Shomron, N., Davis, N., Keydar-Prizant, M., et al. (2013). Pax6 regulates gene expression in the vertebrate lens through miR-204. PLoS Genet. 9, e1003357.

Shestopalov, V.I., and Bassnett, S. (1999). Exogenous gene expression and protein targeting in lens fiber cells. Invest. Ophthalmol. Vis. Sci. 40, 1435–1443.

Shi, Y., Barton, K., De Maria, A., Petrash, J.M., Shiels, A., and Bassnett, S. (2009). The stratified syncytium of the vertebrate lens. J. Cell Sci. 122, 1607–1615.

Shiels, A., and Hejtmancik, J. (2013). Genetics of human cataract. Clin. Genet. 84, 120–127.

Shiels, A., and Hejtmancik, J.F. (2015). Chapter Twelve - Molecular Genetics of Cataract. In Progress in Molecular Biology and Translational Science, J.F. Hejtmancik, and J.M. Nickerson, eds. (Academic Press), pp. 203–218.

Shiels, A., and Hejtmancik, J.F. (2019). Biology of Inherited Cataracts and Opportunities for Treatment. Annu. Rev. Vis. Sci. 5, 123–149.

Shiels, A., Bassnett, S., Varadaraj, K., Mathias, R., Al-Ghoul, K., Kuszak, J., Donoviel, D., Lilleberg, S., Friedrich, G., and Zambrowicz, B. (2001). Optical dysfunction of the crystalline lens in aquaporin-0-deficient mice. Physiol. Genomics 7, 179–186.

Siddam, A.D., Gautier-Courteille, C., Perez-Campos, L., Anand, D., Kakrana, A., Dang, C.A., Legagneux, V., Méreau, A., Viet, J., Gross, J.M., et al. (2018). The RNA- binding protein Celf1 post-transcriptionally regulates p27Kip1 and Dnase2b to control fiber cell nuclear degradation in lens development. PLoS Genet. 14, e1007278.

Sjöblom, B., Salmazo, A., and Djinović-Carugo, K. (2008). α-Actinin structure and regulation. Cell. Mol. Life Sci. 65, 2688.

Song, S., Landsbury, A., Dahm, R., Liu, Y., Zhang, Q., and Quinlan, R.A. (2009). Functions of the intermediate filament cytoskeleton in the eye lens. J. Clin. Invest. 119, 1837–1848.

Subramanian, A., Tamayo, P., Mootha, V.K., Mukherjee, S., Ebert, B.L., Gillette, M.A., Paulovich, A., Pomeroy, S.L., Golub, T.R., Lander, E.S., et al. (2005). Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. U. S. A. 102, 15545–15550.

Subramanian, G., Kuzmanovic, T., Zhang, Y., Peter, C.B., Veleeparambil, M., Chakravarti, R., Sen, G.C., and Chattopadhyay, S. (2018). A new mechanism of

125 interferon’s antiviral action: Induction of autophagy, essential for paramyxovirus replication, is inhibited by the interferon stimulated gene, TDRD7. PLoS Pathog. 14, e1006877.

Subramanian, G., Popli, S., Chakravarty, S., Taylor, R.T., Chakravarti, R., and Chattopadhyay, S. (2020). The interferon-inducible protein TDRD7 inhibits AMP- activated protein kinase and thereby restricts autophagy-independent virus replication. J. Biol. Chem. 295, 6811–6822.

Sun, J., Zhao, Y., McGreal, R., Cohen-Tayar, Y., Rockowitz, S., Wilczek, C., Ashery- Padan, R., Shechter, D., Zheng, D., and Cvekl, A. (2016). Pax6 associates with H3K4- specific histone methyltransferases Mll1, Mll2, and Set1a and regulates H3K4 methylation at promoters and enhancers. Epigenetics Chromatin 9, 37.

Swiston, J., Hubberstey, A., Yu, G., and Young, D. (1995). Differential expression of CAP and CAP2 in adult rat tissues. Gene 165, 273–277.

Tamang, S., Acharya, V., Roy, D., Sharma, R., Aryaa, A., Sharma, U., Khandelwal, A., Prakash, H., Vasquez, K.M., and Jain, A. (2019). SNHG12: An LncRNA as a Potential Therapeutic Target and Biomarker for Human Cancer. Front. Oncol. 9.

Tan, Y.-Q., Tu, C., Meng, L., Yuan, S., Sjaarda, C., Luo, A., Du, J., Li, W., Gong, F., Zhong, C., et al. (2017). Loss-of-function mutations in TDRD7 lead to a rare novel syndrome combining congenital cataract and nonobstructive azoospermia in humans. Genet. Med. Off. J. Am. Coll. Med. Genet.

Tan, Y.-Q., Tu, C., Meng, L., Yuan, S., Sjaarda, C., Luo, A., Du, J., Li, W., Gong, F., Zhong, C., et al. (2019). Loss-of-function mutations in TDRD7 lead to a rare novel syndrome combining congenital cataract and nonobstructive azoospermia in humans. Genet. Med. Off. J. Am. Coll. Med. Genet. 21, 1209–1217.

Tanaka, T., Hosokawa, M., Vagin, V.V., Reuter, M., Hayashi, E., Mochizuki, A.L., Kitamura, K., Yamanaka, H., Kondoh, G., Okawa, K., et al. (2011a). Tudor domain containing 7 (Tdrd7) is essential for dynamic ribonucleoprotein (RNP) remodeling of chromatoid bodies during spermatogenesis. Proc. Natl. Acad. Sci. U. S. A. 108, 10579–10584.

Tanaka, T., Hosokawa, M., Vagin, V.V., Reuter, M., Hayashi, E., Mochizuki, A.L., Kitamura, K., Yamanaka, H., Kondoh, G., Okawa, K., et al. (2011b). Tudor domain containing 7 (Tdrd7) is essential for dynamic ribonucleoprotein (RNP) remodeling of chromatoid bodies during spermatogenesis. Proc. Natl. Acad. Sci. U. S. A. 108, 10579–10584.

126

Taylor, R.P., and Benjamin, I.J. (2005). Small heat shock proteins: a new classification scheme in mammals. J. Mol. Cell. Cardiol. 38, 433–444.

Terrell, A.M., Anand, D., Smith, S.F., Dang, C.A., Waters, S.M., Pathania, M., Beebe, D.C., and Lachke, S.A. (2015). Molecular characterization of mouse lens epithelial cell lines and their suitability to study RNA granules and cataract associated genes. Exp. Eye Res. 131, 42–55.

Updike, D., and Strome, S. (2010). P granule assembly and function in Caenorhabditis elegans germ cells. J. Androl. 31, 53–60.

Varadaraj, K., Kushmerick, C., Baldo, G.J., Bassnett, S., Shiels, A., and Mathias, R.T. (1999). The role of MIP in lens fiber cell membrane transport. J. Membr. Biol. 170, 191–203.

Wang, P.J. (2017). Tracking LINE1 retrotransposition in the germline. Proc. Natl. Acad. Sci. U. S. A. 114, 7194–7196.

Wang, H., Zhang, T., Wu, D., and Zhang, J. (2013). A novel beaded filament structural protein 1 (BFSP1) gene mutation associated with autosomal dominant congenital cataract in a Chinese family. Mol. Vis. 19, 2590–2595.

Wang, J.Z., Xu, C.L., Wu, H., and Shen, S.J. (2017). LncRNA SNHG12 promotes cell growth and inhibits cell apoptosis in colorectal cancer cells. Braz. J. Med. Biol. Res. Rev. Bras. Pesqui. Medicas E Biol. 50, e6079.

Weeks, S.D., Muranova, L.K., Heirbaut, M., Beelen, S., Strelkov, S.V., and Gusev, N.B. (2018). Characterization of human small heat shock protein HSPB1 α-crystallin domain localized mutants associated with hereditary motor neuron diseases. Sci. Rep. 8.

Wioland, H., Guichard, B., Senju, Y., Myram, S., Lappalainen, P., Jégou, A., and Romet-Lemonne, G. (2017). ADF/Cofilin Accelerates Actin Dynamics by Severing Filaments and Promoting Their Depolymerization at Both Ends. Curr. Biol. 27, 1956- 1967.e7.

Wolanski, M., Khosrowshahian, F., Jerant, L., Jap, I.-S., Brockman, J., and Crawford, M.J. (2009). Expression of CAP2 during early Xenopus embryogenesis. Int. J. Dev. Biol. 53, 1063–1067.

Wolf, L., Gao, C.S., Gueta, K., Xie, Q., Chevallier, T., Podduturi, N.R., Sun, J., Conte, I., Zelenka, P.S., Ashery-Padan, R., et al. (2013). Identification and characterization of FGF2-dependent mRNA: microRNA networks during lens fiber cell differentiation. G3 Bethesda Md 3, 2239–2255.

127

Wride, M.A. (2011). Lens fibre cell differentiation and organelle loss: many paths lead to clarity. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 366, 1219–1233.

Yao, X., Yao, R., Huang, F., and Yi, J. (2019). LncRNA SNHG12 as a potent autophagy inducer exerts neuroprotective effects against cerebral ischemia/reperfusion injury. Biochem. Biophys. Res. Commun. 514, 490–496.

Yoon, J.-H., Abdelmohsen, K., and Gorospe, M. (2013). Posttranscriptional gene regulation by long noncoding RNA. J. Mol. Biol. 425, 3723–3730.

Yorston, D. (2004). Surgery for Congenital Cataract. Community Eye Health 17, 23– 25.

Zhang, N., Zhang, C., Wang, X., and Qi, Y. (2017). High-throughput sequencing reveals novel lincRNA in age-related cataract. Int. J. Mol. Med. 40, 1829–1839.

Zheng, C., Wu, M., He, C.-Y., An, X.-J., Sun, M., Chen, C.-L., and Ye, J. (2014). RNA granule component TDRD7 gene polymorphisms in a Han Chinese population with age-related cataract. J. Int. Med. Res. 42, 153–163.

Zhou, B., Li, L., Li, Y., Sun, H., and Zeng, C. (2018a). Long noncoding RNA SNHG12 mediates doxorubicin resistance of osteosarcoma via miR-320a/MCL1 axis. Biomed. Pharmacother. Biomedecine Pharmacother. 106, 850–857.

Zhou, G.-L., Zhang, H., and Field, J. (2014). Mammalian CAP (Cyclase-associated protein) in the world of cell migration. Cell Adhes. Migr. 8, 55–59.

Zhou, S., Yu, L., Xiong, M., and Dai, G. (2018b). LncRNA SNHG12 promotes tumorigenesis and metastasis in osteosarcoma by upregulating Notch2 by sponging miR-195-5p. Biochem. Biophys. Res. Commun. 495, 1822–1832.

Zhou, Y., Bennett, T.M., and Shiels, A. (2016). Lens ER-stress response during cataract development in Mip-mutant mice. Biochim. Biophys. Acta BBA - Mol. Basis Dis. 1862, 1433–1442.

RNA granules: post-transcriptional and epigenetic modulators of gene expression | Nature Reviews Molecular Cell Biology.

128

Appendix A

RNA-SEQ RESULTS IN TDRD7-/- P15

Table-A. 1 List of all significantly missed expressed genes from Tdrd7-/- lens transcriptome with a cut of FDR <0.5.

Symbol Log2 Fold Change FDR Bfsp2 3.67191 1.77E-86 Snhg12 -2.15158 1.46E-48 Tdrd7 -4.21381 1.52E-44 Ints3 -1.562 1.89E-40 Ackr4 -1.89688 6.93E-27 2410006H16Rik -1.73322 8.76E-24 Lpin1 -1.19703 4.20E-23 Sash1 -1.16312 2.78E-20 Ppp1r15a -1.35991 5.40E-20 Ppm1e -1.36466 4.29E-18 Slc18a2 -1.74773 1.80E-17 Isy1 -1.68618 3.27E-16 Snhg6 -1.1912 7.34E-16 Atp1a3 -2.88882 7.34E-16 1810041L15Rik -1.80562 7.53E-16 Tmprss11e 1.514326 7.53E-16 Chrna4 -1.8369 7.16E-13 Tob2 -1.27573 8.82E-13 Aven -1.42483 9.07E-13 Ceacam10 -1.33217 1.17E-12 Syt1 -3.51591 6.89E-12 Sphkap -3.20862 8.51E-12 Fam189a2 -1.32822 1.06E-11 Eno2 -2.58182 1.26E-11 Stambpl1 -1.25514 1.29E-11 H2-Eb1 -2.10392 1.98E-11 Dner -1.06973 3.76E-11 Sntb2 -0.8668 3.86E-11

129

Rp1 -4.70715 4.40E-11 Nlgn2 -1.59968 5.38E-11 Fdft1 0.800024 5.47E-11 Rassf3 0.86994 6.76E-11 Rbp3 -4.07604 1.39E-10 Mylk -1.03192 1.75E-10 Rpgrip1 -2.29181 4.70E-10 Hspb1 -1.06135 6.32E-10 Rorb -3.31611 7.51E-10 Sag -4.43699 7.92E-10 Gpsm2 -0.96893 8.95E-10 Stxbp1 -2.09561 1.32E-09 Klhdc8b -0.88141 1.33E-09 Pde6b -4.05242 1.74E-09 Gnat1 -4.57778 1.75E-09 Impg2 -4.15531 2.71E-09 Pde6a -3.99907 2.71E-09 Npr1 -1.44269 2.74E-09 Gm6548 -1.35929 2.84E-09 Kif5a -2.36585 2.99E-09 Doc2b -1.31501 3.95E-09 Lrrc3 -1.58243 4.36E-09 Rho -4.6173 4.45E-09 Tnrc18 -0.98332 5.53E-09 Ush2a -2.68219 8.54E-09 Dst 0.770567 8.68E-09 Tmem108 -0.9716 8.68E-09 Nav1 -1.2624 9.43E-09 Pbx4 -1.49936 1.14E-08 Plekhb1 -3.42558 1.86E-08 1500012F01Rik -1.00109 1.88E-08 Idi1 0.990535 1.94E-08 Ahdc1 -1.33152 2.17E-08 Pcdh17 -1.33543 2.25E-08 Nfasc -1.14672 2.25E-08 Rab6b -1.21137 3.09E-08 Slc35d1 -1.54373 3.38E-08 Prom1 -3.50887 3.38E-08 2410004N09Rik -1.36028 4.73E-08 Rgs9 -2.44143 4.73E-08 Atxn1 -1.32872 4.87E-08

130

Mrvi1 -1.3671 5.00E-08 Cflar 0.686111 5.25E-08 Hyal3 1.450637 5.25E-08 Fam114a1 0.684264 5.64E-08 Acad11 -1.15014 6.09E-08 Eif4e3 0.833596 6.35E-08 Rgs9bp -2.85553 6.62E-08 Hmgcs1 0.690834 7.66E-08 Cacna2d4 -2.97748 8.11E-08 Rom1 -4.00777 8.29E-08 1810026B05Rik -0.9209 1.10E-07 Med25 -0.92066 1.21E-07 Gnao1 -1.75856 1.33E-07 Dlg2 -1.89135 1.40E-07 Rfx8 1.401817 1.51E-07 6430548M08Rik -1.82065 1.83E-07 Insig1 0.66621 2.46E-07 Adcy5 -0.9059 2.66E-07 Pvrl1 -1.27458 2.76E-07 Metrnl -0.87446 3.01E-07 Ttr -1.91735 3.10E-07 Rlbp1 -2.39199 3.19E-07 Amotl1 -0.6573 3.95E-07 Rab43 -1.00881 4.20E-07 Dync1li2 0.620721 5.84E-07 Zfp354b -1.74089 5.90E-07 Trpc6 0.920925 6.11E-07 Ap2a1 -0.62051 6.11E-07 Enpp2 -1.01805 6.80E-07 Golga4 0.613102 6.85E-07 Ccdc69 1.751377 7.36E-07 Hspa12a -2.00481 9.38E-07 Zeb1 0.60844 1.03E-06 Pex5l -2.95146 1.08E-06 Dmrta2 -0.74048 1.11E-06 Rnf157 -1.39355 1.17E-06 Prkar1b -1.16155 1.22E-06 Ankrd46 -0.70361 1.36E-06 Klhl22 -0.77159 1.37E-06 Ngef -0.78765 1.45E-06 Slc26a2 -1.14084 1.95E-06

131

Lin7a -1.1368 1.95E-06 Sv2b -2.46244 1.95E-06 Pgm2l1 -1.72868 2.35E-06 Ammecr1 0.864813 2.89E-06 Kcnq2 -1.28081 3.38E-06 Gm12750 -1.42174 3.74E-06 Cttn 0.589688 3.79E-06 Tmem9b 0.556657 4.30E-06 Zfand5 -0.83167 4.54E-06 Gltp -0.9951 4.84E-06 Cdr2 -1.76305 5.48E-06 Trak1 -0.65436 6.44E-06 Gm37938 1.090458 6.55E-06 Tmem229b -2.01992 8.38E-06 Spag9 0.581309 8.44E-06 Synj1 -1.28745 9.34E-06 Ttc39b -0.92183 9.42E-06 Igfn1 1.231461 9.77E-06 Mgea5 -0.64012 1.09E-05 Pcbp3 -1.58232 1.14E-05 Syngr1 -1.99575 1.29E-05 Rcor1 -0.80024 1.31E-05 Zfp385a -0.8032 1.31E-05 Itgb1bp1 -0.97717 1.47E-05 Usp46 0.549685 1.53E-05 Myrip -1.14433 1.61E-05 Pkdcc -0.86902 1.61E-05 Nfic -0.87037 1.80E-05 Gm4131 1.002528 1.83E-05 Mef2d -1.62672 1.83E-05 Sh3bgr -0.77241 1.85E-05 Tbc1d30 -1.01634 1.93E-05 Rai14 0.596222 1.98E-05 Snhg8 -0.75228 2.06E-05 Brsk2 -1.4227 2.07E-05 Slc1a2 -2.69249 2.10E-05 Col6a3 0.726311 2.23E-05 Dpysl3 -1.98409 2.26E-05 Ksr2 -1.3965 2.37E-05 Mypn -0.70688 2.74E-05 Rnf169 -0.71554 3.10E-05

132

Selk 0.605601 3.27E-05 Wdtc1 -0.79849 3.38E-05 Epas1 -1.08955 3.52E-05 Frmpd1 -1.66266 3.78E-05 Pfkfb2 -1.78436 3.94E-05 Plcxd2 -2.32108 4.19E-05 Csrp1 -0.83005 4.42E-05 Bcar1 -0.64762 4.49E-05 Fus 1.001207 4.62E-05 Kcnab2 -1.55097 4.68E-05 Ndrg2 -1.15515 4.80E-05 Stx3 -1.48097 5.07E-05 Poln 0.75725 5.59E-05 Susd3 -0.82768 5.81E-05 Zscan29 -0.72905 6.07E-05 Wdfy3 -0.67617 6.07E-05 Tbc1d24 -1.42397 6.15E-05 Scoc 0.519812 6.30E-05 Msi1 -0.68103 6.78E-05 Casp7 0.584283 6.88E-05 Cald1 0.511592 7.16E-05 Dnm3 -0.95096 7.25E-05 Morf4l1 0.6473 7.84E-05 Taf5 -0.94042 7.84E-05 Tshz1 -0.8864 8.03E-05 Abcc6 -0.96625 8.17E-05 Nacad -1.12712 9.18E-05 Ttyh1 -1.81011 0.000101 Sec13 0.501438 0.000102 Rora -0.63185 0.000108 Sfxn1 0.540485 0.000108 Cgnl1 0.641232 0.000112 Taf1d -0.82249 0.000125 Polr2g 0.585688 0.000128 Syne1 -0.89049 0.000133 Sh3bp4 -0.66301 0.000137 Zfp106 0.566873 0.000138 Fam126b -0.7055 0.000138 Mdm1 -2.05392 0.000139 Efr3b -0.95881 0.000141 Arl2bp 0.511434 0.000144

133

Inf2 -0.84444 0.000144 Chchd10 -0.88406 0.000147 Snapc3 -0.5225 0.000147 Unc119 -1.86263 0.000147 Usp45 -0.77358 0.000149 Rai2 -0.74074 0.00015 Sirpa -0.66966 0.000151 Itfg3 -0.73236 0.000151 Ret -1.22613 0.000152 BC005537 -0.6597 0.000155 Ppp1r14b 0.533 0.000155 D8Ertd82e -0.94713 0.000156 Coa5 0.505245 0.000158 Mcf2l -2.04712 0.000158 Plekhg5 -0.79314 0.000158 Epn1 -0.58769 0.000158 Stk35 -0.55155 0.000158 Syt7 -0.96301 0.000158 Psmd11 0.697883 0.000167 Zfyve28 -0.63296 0.000173 Sf3a1 -0.73591 0.000174 Ufm1 0.537963 0.000174 Dzank1 -1.09306 0.000175 Mki67 0.842063 0.000175 Cyb5r3 -0.54391 0.000183 Trak2 0.744215 0.000189 Arhgap23 -0.66635 0.00019 Cacna1a -1.09421 0.000199 Sphk2 -0.83693 0.000201 Impdh1 -1.98381 0.000202 Uox 0.853059 0.000202 Mapt -1.31272 0.000205 Sacs -0.55898 0.000206 Myo7b 0.549772 0.000207 Slc4a7 -1.61111 0.000209 Ednrb -0.76713 0.000209 Yipf5 0.641845 0.000211 Nol4l -1.57 0.000222 Rsbn1 -0.7051 0.000227 Plekhb2 -1.23199 0.000231 Dgat2 -0.52845 0.000236

134

3110002H16Rik -0.58986 0.000238 Gpd1 0.547504 0.00024 Abca4 -1.33692 0.000246 Kcnma1 -0.7733 0.000249 Hist1h4c -0.72772 0.000249 Zmiz2 -0.93502 0.000254 Rnf113a1 0.738078 0.000254 Rnf165 -0.71338 0.000261 Dlg4 -1.37798 0.000263 Epha4 -1.15793 0.000266 Hap1 -1.32093 0.000279 Cdk18 -0.97032 0.000279 Kcnj14 -2.01109 0.000281 Hdac4 -0.64497 0.000286 Prrc2a -0.94745 0.000289 Cadm2 -1.16382 0.000292 Snrk -0.92916 0.000292 Mark4 -0.95818 0.000316 Dhx32 0.699246 0.000338 2310036O22Rik -0.5822 0.000347 Fam3c -1.23397 0.000352 Slc45a4 -0.92364 0.000354 Sptbn2 -0.56371 0.000379 Reep6 -1.60669 0.000382 Ccnd2 0.629043 0.000399 Sv2a -1.18607 0.000404 Zfyve16 0.694482 0.000423 Ntn1 -1.21548 0.000424 Haus7 -1.03552 0.000432 Nsdhl 0.71594 0.000439 Dmwd 0.532337 0.000448 Tubb3 -1.50692 0.00046 Aplp1 -0.62738 0.000464 Ncaph -0.81916 0.000467 Gm10800 2.413215 0.000467 Ndrg3 -0.96249 0.000474 Cngb1 -1.54475 0.000499 Foxk1 -0.92544 0.000499 Prrc2b -0.55763 0.000502 Scarna10 -0.50026 0.000502 Gdpd5 -1.29855 0.000513

135

Thrb -1.10405 0.000527 Atg9a -0.61091 0.000561 Atp6v0a1 -0.67015 0.000612 Nfkb2 -1.04854 0.000618 Herc3 -1.49488 0.000629 Mtpn -0.67504 0.00063 Camta1 -0.91976 0.000633 Ddhd1 -1.19297 0.000633 Fam171b -1.35264 0.000633 Nbeal1 0.515535 0.000635 Myl12a 0.509339 0.000635 Tomm70a -0.50093 0.000638 Wipi1 0.484046 0.000653 Eif3a 0.581541 0.000669 Pkp4 0.464971 0.000676 Pmp22 0.435421 0.00069 Itpr1 -0.50162 0.000697 Eva1a -0.77415 0.000725 2610507B11Rik 0.459759 0.000725 Esyt2 -0.54314 0.00074 Rap1gap -1.09412 0.000753 Dhcr7 -0.8219 0.000764 Ppl 0.478066 0.000797 Adcy6 -0.90634 0.000802 Tmem161b -0.69579 0.000812 Sqle 0.500603 0.000822 Ntm -1.72496 0.00083 Actn4 0.62254 0.00083 Tubgcp4 -0.63048 0.000831 Mturn 0.470052 0.000831 Gpsm3 -0.88439 0.000841 Slc8a1 -0.51172 0.000845 Gm4850 -1.21775 0.000845 Nsf -1.39425 0.000845 Slc25a22 -1.05539 0.00085 Anapc2 0.750652 0.000874 Tsc22d4 -0.82433 0.000908 Gm27042 1.020098 0.000925 Gusb 0.675682 0.000943 Adgrl1 -0.55319 0.000943 Aldh6a1 0.525876 0.000951

136

Plxna2 -1.54508 0.001013 Tet3 -0.72548 0.001023 Ptpn11 0.454513 0.001038 Scgn 0.638849 0.001047 5730409E04Rik 0.765737 0.001061 Atxn2 -0.66709 0.001065 H2-K2 -0.7838 0.001077 Map7d1 -1.0115 0.001107 Htt -0.54589 0.001113 Pfkp -1.01868 0.001129 Gm20631 0.785415 0.00113 Socs2 -1.33627 0.00115 Plekha3 0.481309 0.001166 Gm10138 -0.66418 0.00121 Arid1a -0.60094 0.00126 Hmcn1 0.797306 0.00126 Gm15513 -1.1 0.00126 Ptpn4 -0.44172 0.00126 Peak1 0.516002 0.00126 Ppp4r1 -0.46897 0.00126 Trim62 -1.17515 0.001343 Ttc27 0.556819 0.001404 Med30 0.546996 0.001407 Nasp -0.49935 0.00143 Sparcl1 -0.93863 0.00143 Islr 0.456811 0.001441 Plaa 0.43588 0.00145 Tet2 -0.75882 0.001452 Fryl -0.58414 0.001457 Copb2 0.440902 0.00147 Samd14 -0.78184 0.00149 Myh7b -0.7852 0.00149 Snx22 -0.60104 0.001498 Prr14 -1.06706 0.001506 Plekhs1 0.858601 0.001533 Magt1 0.443439 0.001533 Mxd4 -0.45106 0.001693 Usp16 0.490291 0.001712 Ccni -0.44269 0.001712 Hsp90b1 0.430145 0.001749 Gadd45a -0.71912 0.001749

137

Atp1b1 -0.66736 0.001749 Herpud1 -0.51151 0.001865 Scap -0.62707 0.001865 Gpr179 -1.37067 0.001871 Pgk1 0.4553 0.001871 Rassf5 -0.63249 0.001873 Msn 0.528638 0.001873 Tns3 0.620329 0.001876 Srgap3 -0.9001 0.001876 Fam161a -1.48258 0.001931 Fstl1 0.46605 0.001946 Ccdc117 -0.54239 0.001946 Csmd2 0.636782 0.001948 Supt3 -0.69007 0.001959 Zfp592 -0.8089 0.001968 Idh2 0.408228 0.00199 Aff1 -0.45919 0.001992 Cyfip1 0.441702 0.002029 Iqck 0.768443 0.002092 Prune2 -0.70093 0.002147 Smad1 -0.45693 0.002163 Canx 0.401412 0.002178 Trappc9 -0.82589 0.002181 Bicd2 0.609141 0.002191 Madd -1.46721 0.002207 Nras 0.435811 0.002215 Gnpat 0.442875 0.002224 Rbfox1 -1.20033 0.002319 Chmp5 0.472974 0.002319 Pcmtd2 -0.7015 0.002329 Ap3d1 -0.4372 0.00235 Ube2o -0.41941 0.00239 Osbpl6 -0.65348 0.00239 Lamc2 0.866693 0.002435 Anxa2 0.534705 0.002447 Grina -0.69596 0.002459 2-Sep 0.416645 0.00252 Stk39 0.41743 0.00258 Npr2 -0.75562 0.002603 Ccdc126 -1.55699 0.002622 Serp1 0.400812 0.002733

138

Lgmn -0.50126 0.002733 Nup210l 0.950633 0.002733 6-Mar -0.45222 0.002799 Opn3 -0.56479 0.002813 Zcchc14 -0.7721 0.002823 Myl12b 0.414778 0.002865 Gtf2h1 0.473484 0.002922 Gm5776 0.66383 0.002922 Zbtb11 -0.66298 0.002922 Ercc6l2 0.521673 0.002961 Cenpf 1.189068 0.002974 Lats2 -0.63169 0.003011 Haus8 -0.55038 0.003011 Ptms -0.68875 0.003011 Adamts17 -0.89722 0.003011 Ttll7 -0.49487 0.003051 Atg2b -0.70783 0.003061 Cpeb4 -0.43957 0.003088 Pmaip1 -0.97328 0.003101 Uso1 0.624895 0.003138 Greb1l -0.92094 0.00317 Scg5 -1.0815 0.003185 Arhgap32 -0.83032 0.003212 Dbp -0.72863 0.003354 Gpm6a -1.74198 0.003354 Crebbp -0.51857 0.003354 Ncoa2 -0.4925 0.003354 Slc22a2 0.890403 0.003385 Lamc1 0.482492 0.003385 Creg1 -0.55357 0.003385 Klhl8 -0.62753 0.003412 Gnaz -1.47541 0.003464 Npm3 0.501406 0.003566 Sulf2 -0.90245 0.003566 Nek9 0.394341 0.003566 Fgd4 -0.91447 0.003608 Kmt2c -0.44286 0.003689 Lgalsl -0.7317 0.003689 Surf4 0.40438 0.003689 2410131K14Rik 0.481429 0.003697 Homer2 -0.77582 0.003713

139

Aqp1 0.764477 0.003719 Reep3 0.48592 0.003787 Fscn1 -0.53834 0.003787 Sec23a 0.401058 0.003818 Hdac5 -0.69739 0.003818 Fam53b -0.80874 0.003825 Tmem132b 0.643247 0.003825 Slc7a11 0.688621 0.003829 Ipmk -1.03425 0.00383 Rffl 0.573262 0.003867 Gje1 0.498117 0.003867 Gm1821 0.432317 0.003887 Cxxc4 -1.02742 0.003916 Las1l 0.469288 0.003921 1810022K09Rik 0.456428 0.003921 Sbk1 -1.10901 0.003923 Smim13 -0.91977 0.003952 Akap14 0.870579 0.003998 Pigq 0.441306 0.003999 Slx1b -0.93294 0.00405 2310015A10Rik -0.6401 0.004193 Syt11 0.561839 0.004197 Ppp4r2 -0.49541 0.004197 Rhot1 -0.68319 0.004199 E130218I03Rik -0.56542 0.004212 Eno3 0.38822 0.004249 2510003E04Rik 0.431097 0.004324 Lbhd1 -1.28723 0.004368 Bhlhe41 -0.60028 0.004423 Ttyh3 -0.59418 0.004475 Ankrd13a -0.47589 0.004475 Srp68 0.514165 0.004497 Ankrd12 -0.41561 0.004509 Pink1 -0.39602 0.004588 Sae1 0.378831 0.004598 Itfg2 -0.80387 0.004646 Pcolce 0.55781 0.004648 Ctnnal1 0.431797 0.004678 Wipf2 -0.77871 0.004741 Sh3bp1 -0.70764 0.004754 Cp 0.436551 0.004754

140

Cyp51 0.399281 0.004819 Erbb4 -0.5688 0.004875 Ppp3cc -0.95992 0.004879 Dmd -0.67319 0.00504 Vma21 0.507677 0.00504 Aig1 -0.44225 0.005055 Unc119b -0.52793 0.005188 Rad54l2 -0.45339 0.005193 Cnn3 0.387319 0.005209 Tnk2 -0.96866 0.00521 Ptn 0.630019 0.00521 4930402H24Rik -0.71832 0.00521 Chrdl1 0.705148 0.005247 Efcab2 0.504462 0.005578 Snap47 0.377011 0.005629 Ank3 -0.61192 0.005682 Ric8 0.457194 0.005812 Bgn -1.13673 0.005812 Sptssa 0.489715 0.005934 Unc13b 0.394814 0.005934 Cry1 -0.68712 0.005934 Lonrf2 -0.7286 0.005934 Atp13a5 0.989717 0.006081 Gm21738 2.027961 0.006086 Ipo5 0.490607 0.006109 Dusp6 0.774723 0.006196 Bicd1 0.432134 0.006374 Gnb5 -1.27654 0.006536 Mxd1 -0.48992 0.00659 Tspan9 -0.4447 0.006616 Eif2a 0.471468 0.006634 Ghitm 0.360624 0.006634 Mapk14 0.49657 0.006679 Prex2 -0.89957 0.006782 Znf512b -0.61009 0.006946 Pik3c2b -0.67667 0.006961 Spry1 -0.59277 0.007063 Cela1 -0.50107 0.007073 6030419C18Rik -0.81961 0.007073 BC029214 -0.63132 0.007073 Rptor -0.53571 0.007073

141

Atp2b1 -1.11769 0.007073 Ermap 0.59074 0.007075 Tnr -1.35658 0.007132 Cic -0.59056 0.00717 Hspd1 0.37604 0.007301 Cfh -1.0735 0.007482 Rmrp -0.42124 0.007502 Serpind1 -0.58993 0.00753 Maea 0.366115 0.007537 Scarna6 -0.43786 0.007663 Zfp319 -0.71775 0.007668 Banf1 0.601218 0.007702 Actg1 0.468209 0.007744 Gm13456 -0.39505 0.007823 D630045J12Rik -0.46181 0.007824 Sox1 -0.45462 0.007824 Stt3a 0.40466 0.007941 Ccnd1 0.38695 0.007971 1110038B12Rik -0.74935 0.007971 Mecom 0.971149 0.008038 Usp33 -1.03512 0.008068 BC005624 0.434859 0.008068 Ppp2r2b -0.56705 0.008085 Map1lc3a -0.80046 0.008085 Rabgef1 -1.0813 0.008101 Sphk1 -0.92743 0.008136 Sfpq -0.62958 0.008297 Tmem26 -0.88146 0.008418 Pex5 -0.48222 0.008516 Slit3 -0.47764 0.008782 Etnk1 -0.44543 0.008821 Casz1 -0.93386 0.00885 Setd8 0.427108 0.00886 Slc6a17 -0.43063 0.008897 Cdc42ep4 -0.36968 0.008897 Ssr3 0.377897 0.008933 Uckl1 -0.97072 0.00914 Ggt1 0.616851 0.0092 Capzb 0.428427 0.009304 Kmt2d -0.79278 0.009409 Plekhf2 -1.07539 0.009442

142

Egflam -1.24269 0.009559 Gtpbp4 0.415195 0.00958 Xpo7 0.390907 0.009785 Zfp521 -0.82717 0.009785 Gm5345 0.747981 0.009831 Caskin1 -0.85551 0.010009 Parva 0.403417 0.010009 Lrrc8b -1.13907 0.010232 Arntl -0.75377 0.010233 Mapre1 0.356993 0.010294 Gle1 0.423793 0.010294 Gucd1 -0.70383 0.010309 Tcerg1 -0.58202 0.010353 Timp4 -0.78935 0.010376 Manf 0.427193 0.010387 Cecr2 -1.30207 0.0104 Tspyl4 -0.65773 0.010406 Scarb1 -0.76663 0.010436 Fnbp4 -0.52099 0.010454 Gfap 0.63771 0.010567 Pla2g7 -0.52925 0.010567 Btbd1 0.386587 0.010567 Hopx 0.429684 0.01063 Gm15800 -0.71887 0.010692 Nomo1 -0.51622 0.010692 Glmn -0.65748 0.010692 Sspn 0.645405 0.010692 Plekha4 -0.47573 0.01071 Gm14680 0.374396 0.010778 Ruvbl1 0.406398 0.010778 Klhl32 -0.65208 0.010778 Dctd 0.688337 0.010949 Pelp1 -0.76981 0.011018 Rnf24 -0.67894 0.011018 Gpx8 0.54783 0.011018 Hdac9 -0.65011 0.011196 Igfbp2 2.334832 0.011196 Smarcd2 -0.46531 0.011212 6030458C11Rik 0.431395 0.011364 Prkce -0.61845 0.011497 Nudt4 0.499188 0.01155

143

Net1 0.579678 0.011743 Kdm4c -0.61187 0.011759 Bbs9 -0.56913 0.011969 Cplx2 -0.56418 0.01232 Zfp329 -0.47509 0.01232 Pcnxl3 -0.64283 0.01232 Gsn 0.377976 0.012401 Ddost 0.360899 0.012401 Srrm2 -0.37223 0.012431 Nid1 0.409354 0.012458 Nenf -0.40929 0.012556 Ctnna2 0.354205 0.012601 Capn5 0.552707 0.012601 Sik1 -0.69789 0.012777 Ddit3 -0.66404 0.012834 Fgfrl1 -0.43309 0.012834 Endod1 -0.64692 0.012834 Zfp423 -0.86409 0.012852 Chrng -0.4603 0.012852 Cap2 0.364739 0.012989 Ssfa2 0.454255 0.013293 Thsd4 0.437061 0.013408 Gpcpd1 -0.55128 0.01352 Dlg5 -0.67791 0.01379 Rbp1 -0.75051 0.01379 Eml2 0.456401 0.01379 Npnt 0.593565 0.013855 Sugp2 -0.39911 0.01392 Fam222b -0.7017 0.013942 Eea1 0.406179 0.014088 Becn1 0.353919 0.014088 Rnpepl1 -0.60296 0.014113 Ostc 0.429379 0.014114 Sobp -0.78406 0.014435 Slc19a3 0.569003 0.014474 Icmt -0.70966 0.014586 Lca5 -0.74836 0.014705 Tuba4a -0.70309 0.014849 Nfix -0.81758 0.014868 Myo1b 0.358238 0.014944 Cep104 0.573923 0.014944

144

Nox4 0.654844 0.014981 Nrxn3 -0.94951 0.015106 Eif3b 0.600118 0.015113 Tm9sf3 0.351416 0.015193 Rfwd2 0.410646 0.015193 Sar1a 0.355879 0.015224 Pkp2 0.335179 0.015235 Rpph1 -0.48095 0.015281 Bahd1 -0.85964 0.015505 Gm26870 1.815625 0.015525 Rasgrp1 -0.49374 0.015583 Gm10131 0.555602 0.015625 Fam193a -0.43417 0.015904 Znfx1 -0.57725 0.016004 S100a4 0.373371 0.016362 Pten -0.37181 0.016416 Gm26847 0.448992 0.017103 Pclo -0.70242 0.017103 Tmem131 -0.52816 0.017105 Parp6 -0.81533 0.017178 Gm5607 -0.33974 0.017192 Eno1b 0.439085 0.017192 Rad51c -0.77477 0.017197 Nid2 1.455903 0.017197 L2hgdh -0.57049 0.017197 Cirbp -0.56645 0.017197 Ncald -0.99014 0.017223 Ptbp1 0.383759 0.017223 Fam46c -0.42442 0.01745 Rnf10 -0.46523 0.017483 Hsd17b12 0.444441 0.017512 Enpp5 -0.44762 0.01752 Celf2 -0.60822 0.018044 Ints8 0.429902 0.018508 Sike1 0.374297 0.018535 Fkbp1a 0.543657 0.018561 Amer1 -0.45643 0.018561 Zfp870 -0.59442 0.018636 Strip1 0.382029 0.018636 Chd7 -0.55067 0.018764 Ndufaf5 -0.57842 0.018764

145

Dusp8 -0.81323 0.018764 5530601H04Rik -1.10365 0.018829 Ep300 -0.43729 0.018829 Ep400 -0.47507 0.018936 Nucb1 -0.41153 0.01897 Gpd1l -0.4158 0.018984 Mdm2 -0.50692 0.019023 Vopp1 -1.41647 0.019023 Aatk -0.52664 0.019046 Nudt2 0.413778 0.019053 Pak3 -0.89562 0.019194 Egln1 0.415782 0.019204 Vamp1 -0.54395 0.019299 Tppp3 0.385429 0.019309 Tmem189 0.415074 0.019322 Tgfb1 -0.68705 0.019378 Nim1k -0.83516 0.019455 Tmem167 0.368912 0.019578 Wiz -0.68913 0.019592 Ints6 -0.35372 0.019617 Gabbr1 -1.108 0.019889 Wdr35 0.415357 0.019953 Ctsc 0.758752 0.02007 Huwe1 0.324355 0.020215 1600020E01Rik -0.73357 0.020326 Cdon 0.580251 0.020365 Gm6170 0.736331 0.020391 Ptcd3 0.340811 0.020499 Ybx3 -0.45063 0.020914 Il4 -0.65286 0.020921 Ids 0.394068 0.020922 Ypel1 -0.65908 0.020928 mt-Nd2 0.550108 0.021003 Scaf1 -0.55296 0.021095 Arhgdia 0.342129 0.02145 Cd24a 0.444951 0.021451 Lss 0.457681 0.021451 Gja1 1.034095 0.02153 Pdzd2 -0.75722 0.02153 Hist2h2be 0.569733 0.021737 Rybp -0.64549 0.02179

146

Fas -0.67721 0.021902 Nedd9 0.652971 0.021902 Mn1 -0.68283 0.021902 Ppip5k1 -0.60289 0.022073 Ulk1 -0.70996 0.022247 Pja2 0.329351 0.022435 Fn3krp 0.328665 0.022435 Rnf19a -0.38676 0.022437 Slmap -0.37224 0.022508 Tgfbr3 0.771548 0.022508 Armcx4 0.485395 0.022602 Anapc1 0.332729 0.022669 Arhgef10 0.43 0.022684 Ywhae 0.577127 0.022762 Copg2 0.340183 0.022762 Scn2a1 -0.76894 0.022825 Zswim8 -0.49818 0.022975 Clip3 -0.49259 0.023122 Slc3a2 0.347794 0.023122 Hmgn1 -0.65379 0.023122 Ap1m1 0.548259 0.023122 Adarb1 -0.47088 0.023145 Fam214a -0.46798 0.023163 Arid1b -0.41537 0.023163 Mcm2 0.536042 0.02318 Ralb 0.441782 0.023196 Tmem40 0.398758 0.02322 Tmed10 0.470356 0.023339 Mta1 -0.63893 0.023413 Fbn1 1.637057 0.023526 Zeb2 0.322912 0.023528 Zfp365 0.392996 0.023586 mt-Nd5 0.487923 0.023586 Tmem30a 0.346158 0.023694 Dhx9 0.52765 0.023731 Pptc7 -0.59745 0.023778 Atrnl1 -0.40106 0.024044 Actb 0.383338 0.024044 Tmed9 0.491837 0.024044 Plat 0.635988 0.024106 Nap1l4 0.378818 0.024131

147

Trim37 0.34575 0.024156 Psmd7 0.466291 0.024265 Sppl2a 0.321991 0.024758 Ncs1 -0.68506 0.024806 Scamp5 -0.43709 0.024806 Eno1 0.397425 0.024806 Hexb 0.415061 0.024829 Sybu -0.56086 0.024859 Fahd2a -0.45253 0.024895 Cxxc5 0.406904 0.024942 Pcsk5 0.484448 0.025391 Git1 -0.61625 0.025391 Gja8 0.367235 0.025427 Rnf5 0.642852 0.025454 C77080 -0.51565 0.025468 Pitx3 -0.48474 0.025468 Wdfy4 0.489311 0.025468 U3 -0.42065 0.025502 Zc3h13 0.572877 0.025644 Sdk2 0.395656 0.025707 Kif1c 0.385951 0.02577 Vim 0.370634 0.025773 Rufy3 -0.37443 0.025818 Nsun2 0.345261 0.025819 Ccnh 0.434612 0.0259 Hras -0.36754 0.026016 Wasf3 -0.59333 0.026016 Nars2 -0.61966 0.026124 Hk1 -0.91066 0.026159 Timeless 0.590892 0.026581 Gan -0.58211 0.026581 Cdk1 0.793837 0.026581 Baz2a -0.48365 0.026808 Gnb1 -0.68983 0.026808 Pdia3 0.427514 0.026917 Litaf 0.444573 0.026935 Rab5c 0.327811 0.027186 Fam84b -0.69183 0.027346 Agtpbp1 0.317807 0.027346 Pcbd1 -0.44026 0.027346 Itga8 0.66998 0.027382

148

Mapk3 0.436187 0.027577 Arhgap39 -0.49706 0.027614 Cog6 0.313663 0.027625 Rbm17 0.380144 0.027777 Tacc1 -0.4556 0.027777 mt-Nd6 0.492297 0.027925 Wasl -0.58471 0.028058 Bbs5 -0.64307 0.028135 Dync1i1 0.325486 0.028261 Rhbdf2 -0.59416 0.028261 Ift88 0.498855 0.028326 Ets2 -0.7088 0.028326 Bbs1 -0.71538 0.028478 Itgb1 0.343861 0.028488 Gm6863 0.326842 0.028625 Usp8 0.437028 0.028837 Sik3 -0.60274 0.028944 Galnt12 -0.33957 0.029119 Arpc3 0.330285 0.029388 Tpm4 -0.57182 0.029508 Olfm1 -0.40236 0.029772 Fech 0.432305 0.029993 1700052N19Rik -0.4884 0.029993 Top2a 0.786162 0.030149 Zdhhc2 -0.67797 0.030149 Epb4.1l5 -0.76294 0.030336 Lmtk2 -0.54531 0.030336 Elovl4 -0.90069 0.030564 Igsf3 0.533243 0.030699 Rbm38 -0.49139 0.030784 Lman1 0.397267 0.03097 Prkab1 -0.72616 0.031235 Hs6st2 0.684315 0.031244 Gm7335 0.684378 0.03153 Eaf1 -0.44349 0.031685 Hs3st3b1 -1.15704 0.03198 1300017J02Rik -0.55034 0.032048 Samd8 0.407573 0.03223 Krcc1 0.382569 0.03223 Kifc3 -0.33114 0.03223 Lig1 0.615873 0.032525

149

Agap1 -0.41803 0.032663 Klf9 -0.69506 0.032695 Prkca -0.50101 0.032723 Dyrk1b -0.82594 0.03273 Camk2d -0.46049 0.03273 Igf2 -0.60625 0.03273 Metap1 0.396106 0.032818 Dazap2 0.343704 0.032818 Dym -0.59975 0.032874 Itpr2 0.530736 0.032888 Ezh1 -0.54969 0.032903 Ksr1 -0.70821 0.033137 Grik5 -0.61341 0.033137 AC245272.2 1.067322 0.033215 Usp2 -0.8156 0.033462 Prnp -0.75304 0.033462 Actr2 0.566397 0.033462 Zfp551 -0.68402 0.033502 Osbp2 0.36049 0.033613 Cct8 0.307601 0.034273 Gm10698 0.365719 0.03445 Clasrp -0.72375 0.034461 Acsl4 -0.66288 0.034483 B130055M24Rik -0.65768 0.034693 Gatad1 0.322326 0.034693 Anks6 -0.77149 0.0348 Osbpl3 -0.53389 0.034877 Nxn 0.296722 0.034956 mt-Cytb 0.467459 0.034991 Nap1l1 0.439935 0.034995 Tnfaip6 -0.57471 0.035174 Sox12 -0.52669 0.035179 Tecrl -0.43074 0.035179 Tanc1 0.403459 0.035717 Mfsd6 -0.55738 0.035717 Gm14681 0.404351 0.035717 Tmco3 0.380715 0.035791 Arcn1 0.307469 0.035924 Tsnax 0.350778 0.036255 Col4a5 0.444169 0.036527 Tpp2 0.368218 0.036527

150

Lpgat1 -0.48501 0.036839 Nphp1 0.425974 0.037045 Mtcl1 -0.70364 0.037045 Msmo1 0.370065 0.037045 Ccdc88c 0.629464 0.037045 Cetn2 0.407177 0.037091 Crocc -0.73629 0.037173 Arhgap10 -0.69644 0.037524 Cep162 0.52098 0.037524 Glb1l 0.712733 0.037524 Zswim5 -0.40837 0.037524 Nab1 0.496793 0.037524 Zfp189 -0.62327 0.037524 Kcne1 0.729585 0.037524 Kat2b 0.307708 0.037711 Dgke -0.93188 0.037827 2300009A05Rik 0.472244 0.037827 Ptch1 -0.37339 0.037827 Ntrk2 1.624502 0.037916 PISD 0.589481 0.037916 Iqsec1 -0.41128 0.038165 Cd82 0.646873 0.038681 Ogdhl -0.77006 0.038682 Tm9sf4 0.352141 0.038925 Cdk14 -0.50279 0.039226 Ebf1 -0.68586 0.039274 Ccdc66 0.422598 0.039274 Desi1 -0.68133 0.039408 Gprasp2 -0.67392 0.039801 Upf3b 0.437883 0.03986 Gdi1 -0.4857 0.03986 Mybbp1a 0.411621 0.040684 Atxn2l -0.59076 0.040724 Gm5454 0.702677 0.040853 Gm22009 -0.41242 0.040853 Unc13c -0.86714 0.040887 Pfdn2 0.5113 0.040887 Rsl1d1 0.325708 0.040911 Larp1 -0.36226 0.041004 Tuba1a 0.344231 0.041004 Slc7a5 -0.43222 0.041004

151

Sh3pxd2b -0.54216 0.041062 Col9a3 1.612112 0.041267 Lrrc4 -0.34394 0.041323 Esco1 -0.37814 0.041494 Hist1h2be 0.428966 0.041494 Ncoa1 -0.41893 0.041761 Osgin2 0.306403 0.041771 Adam9 0.311577 0.041799 Camk1d -0.41514 0.041903 Mgat3 -0.50221 0.041903 Gm7336 0.747407 0.041903 Cul1 0.31395 0.041977 Cep250 -0.92694 0.041977 Cramp1l -0.49818 0.041977 Nnt 0.371817 0.041977 Ankrd24 -0.60728 0.041977 Rn7s2 -2.91714 0.042003 Atp5l -0.38999 0.042118 Cables1 -0.57213 0.042118 Pls3 0.495145 0.042118 Rhob -0.40212 0.042118 Calu 0.328384 0.042158 Sort1 -0.29656 0.042248 Fzd2 0.700731 0.042318 Map3k12 -0.64233 0.042318 Tyrp1 1.315591 0.042507 Bcr -0.47146 0.042507 Tex14 0.752272 0.042583 Rogdi -0.63075 0.042722 Wdr66 -0.46775 0.043044 Siah1a -0.43726 0.043198 Map2k2 -0.33538 0.043198 Slmo1 -0.51078 0.043756 Tril -0.82842 0.043756 Dip2a -0.38558 0.04383 Taf15 0.812277 0.043963 Zfp654 -0.4992 0.043992 Csrnp2 -0.7662 0.044153 Ap3m1 0.3483 0.04433 Mafk 0.50655 0.044771 Atp8a2 -0.37653 0.044815

152

Gm21399 0.40284 0.044869 Cdadc1 0.854424 0.044869 Vldlr -0.48819 0.044869 Helz -0.52013 0.044889 Fam219a -0.67219 0.044889 Spop 0.344382 0.044889 Card6 0.465052 0.04497 Tmsb4x 0.304309 0.045003 Bzw2 -0.61455 0.045089 Kif15 0.729786 0.045264 Raph1 -0.58093 0.04527 Irs1 -0.31999 0.045293 Htatsf1 0.314523 0.045293 Zfp462 -0.82253 0.045392 Ltbp1 0.725935 0.045602 Igbp1 0.344995 0.045856 Smyd4 -0.46295 0.045946 Oxt 0.606614 0.046029 Gstcd 0.551125 0.046029 Bend7 -0.47668 0.046029 C330027C09Rik 0.664629 0.046065 Nckap1l 0.620094 0.046065 Rsu1 0.488053 0.046077 Ralgds 0.342826 0.046109 Got1 -0.94274 0.046109 Gcg 0.806661 0.046109 Maz -0.41792 0.046109 Spns2 -0.32203 0.046258 Sulf1 0.527606 0.046258 Mmadhc 0.336255 0.046258 Celsr1 -0.5145 0.046258 Poc5 -0.55293 0.04638 Clip1 -0.76472 0.046411 Ankrd52 -0.54175 0.046418 Arrdc3 -0.53895 0.046924 Ergic1 0.333002 0.047228 Morc4 -0.54859 0.047228 Triobp 0.334493 0.047273 Pgd 0.576999 0.047321 Cebpz 0.32339 0.047503 Glb1 0.366388 0.047583

153

Znhit6 0.709142 0.047647 Capn15 -0.78765 0.047922 Serping1 0.824908 0.048084 Tbca 0.329183 0.048084 Fras1 0.721617 0.048084 Fam129a 1.635588 0.048084 Dnajb11 0.33219 0.048147 Cdcp1 0.473618 0.048166 Rgp1 -0.46134 0.04842 Grpel2 -0.3209 0.04842 Aldh1a1 0.359498 0.048472 Gabrb2 -0.86221 0.048728 Atf4 -0.32407 0.048728 Dse 0.42388 0.048728 Eml3 -0.99541 0.048768 Nrip1 -0.49022 0.048822 Mnt -0.55261 0.048837 Snrpg 0.516909 0.049114 Slc23a2 -0.69851 0.049169 Mical2 -0.62912 0.049774 Eif4e 0.335844 0.049896

154

Appendix B

TMT/MS PROTEOME RESULTS

Table-B. 1 List of all significantly missed expressed proteins from Tdrd7-/- lens transcriptome with a cut of P-value <0.5

155

156

sp|Q920E5|FPPS_MOUSE 0.262 0.000757 sp|Q91ZU6|DYST_MOUSE 0.271 0.000149 sp|Q3ULJ0|GPD1L_MOUSE -0.141 0.049574 sp|P48193|41_MOUSE 0.213 0.001866 sp|Q9CQH7|BT3L4_MOUSE 0.205 0.002368 sp|P46638|RB11B_MOUSE 0.219 0.006878 sp|Q9EQQ9|OGA_MOUSE -0.280 0.000018 sp|P29758|OAT_MOUSE -0.195 0.031904 sp|Q9DCN2|NB5R3_MOUSE -0.209 0.003734 sp|P58044|IDI1_MOUSE 0.601 0.000000 sp|P15626|GSTM2_MOUSE -0.282 0.004979 sp|P00920|CAH2_MOUSE 0.636 0.012560 sp|Q8BJW6|EIF2A_MOUSE 0.129 0.044484 sp|Q61655|DD19A_MOUSE 0.281 0.000013 sp|P52825|CPT2_MOUSE -0.187 0.007793 sp|Q9Z1G3|VATC1_MOUSE 0.150 0.014856 sp|Q91W90|TXND5_MOUSE -0.134 0.045826 sp|Q9Z1Q5|CLIC1_MOUSE -0.161 0.011388 sp|Q9CQC9|SAR1B_MOUSE 0.230 0.009115 sp|Q6ZQ73|CAND2_MOUSE -0.220 0.002321 sp|P59017|B2L13_MOUSE -0.240 0.006278 sp|P29391|FRIL1_MOUSE 0.881 0.000000 sp|Q5EBG6|HSPB6_MOUSE -0.205 0.037950 sp|Q64737|PUR2_MOUSE 0.144 0.045252 sp|Q9JI91|ACTN2_MOUSE -0.167 0.023829 sp|Q8BGD9|IF4B_MOUSE -0.158 0.026477 sp|Q60737|CSK21_MOUSE 0.134 0.022630 sp|P97461|RS5_MOUSE -0.173 0.010002 sp|Q9D8B4|NDUAB_MOUSE -0.273 0.012474 sp|O54931|AKAP2_MOUSE -0.379 0.000015 sp|Q9R269|PEPL_MOUSE 0.271 0.000517 sp|O08795|GLU2B_MOUSE -0.191 0.030302 sp|P06801|MAOX_MOUSE 0.171 0.025919 sp|Q9R0N0|GALK1_MOUSE -0.159 0.028267 sp|Q9CY64|BIEA_MOUSE 0.155 0.023963 sp|Q9CZ04|CSN7A_MOUSE 0.172 0.004356 sp|P98083|SHC1_MOUSE -0.159 0.036234 sp|Q9D1J1|NECP2_MOUSE 0.136 0.045272

157

sp|P14873|MAP1B_MOUSE 0.361 0.004116 sp|Q9D1D4|TMEDA_MOUSE 0.171 0.021858 sp|Q9QY48|DNS2B_MOUSE -0.203 0.013532 sp|P42669|PURA_MOUSE 0.200 0.030310 sp|P62874|GBB1_MOUSE 0.542 0.044905 tr|Q6DFZ1|Q6DFZ1_MOUSE -0.171 0.010512 sp|P07214|SPRC_MOUSE -0.332 0.000027 sp|O55013|TPPC3_MOUSE 0.152 0.028200 sp|P58389|PTPA_MOUSE -0.144 0.046924 sp|Q8BHL5|ELMO2_MOUSE -0.178 0.024904 sp|Q99KB8|GLO2_MOUSE 0.280 0.000072 sp|P35293|RAB18_MOUSE 0.160 0.014915 sp|Q921J2|RHEB_MOUSE 0.296 0.000001 sp|P97355|SPSY_MOUSE 0.293 0.000195 sp|Q76N33|STALP_MOUSE -0.559 0.000000 sp|O35226|PSMD4_MOUSE -0.166 0.019550 sp|Q99JB8|PACN3_MOUSE -0.272 0.000177 sp|Q8K3C3|LZIC_MOUSE -0.228 0.001574 sp|P97792|CXAR_MOUSE -0.351 0.000001 sp|P36536|SAR1A_MOUSE 0.239 0.001078 sp|P26231|CTNA1_MOUSE -0.524 0.000000 sp|Q8K003|TMA7_MOUSE 0.457 0.034893 sp|P17563|SBP1_MOUSE -0.240 0.037622 sp|Q99KP6|PRP19_MOUSE 0.168 0.028564 sp|P39061|COIA1_MOUSE -0.459 0.027145 sp|Q60675|LAMA2_MOUSE -0.632 0.000452 sp|P20612|GNAT1_MOUSE 0.820 0.018959 sp|Q61584|FXR1_MOUSE 0.237 0.004392 sp|Q61001|LAMA5_MOUSE -0.660 0.040381 sp|Q69ZR2|HECD1_MOUSE -0.199 0.015040 sp|O08663|MAP2_MOUSE 0.278 0.000171 sp|Q9D394|RUFY3_MOUSE -0.189 0.016692 sp|B1AUH1|PTPRU_MOUSE -0.327 0.000312 sp|P97864|CASP7_MOUSE 0.430 0.000008 sp|Q8QZY6|TSN14_MOUSE 0.312 0.002134 sp|P35846|FOLR1_MOUSE -0.298 0.024191 sp|P16125|LDHB_MOUSE 0.228 0.047584 sp|Q6WVG3|KCD12_MOUSE 0.502 0.000000

158

159

160

161

Appendix C

LIST OF TDRD7 PULLDOWN PROTEINS

Table-C. 1 Tdrd7 pulldown list A list of proteins returned from the mass spectrometry of Tdrd7 pulldown. None of these proteins were identified in the IgG pulldown (negative control). The pulldown was performed on wildtype Postnatal age P8 ICR proteins

Protein Gene Gene Name Q8K1H1uniprot ID TDRD7Symbol tudor domain containing 7(Tdrd7) F8WGN6 KIF21A kinesin family member 21A(Kif21a) Q9CQV8 YWHAB tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation A6X919 DPY19L1 dpyprotein,-19- likebeta 1 polypeptide(Ywhab) (C. elegans)(Dpy19l1) P53026 RPL10A ribosomal protein L10A(Rpl10a) E9Q9M0 CEP290 centrosomal protein 290(Cep290) P18760 CFL1 cofilin 1, non-muscle(Cfl1) S4R1W1 Gm3839 glyceraldehyde-3-phosphate dehydrogenase pseudogene(Gm3839) Q91ZX7 LRP1 low density lipoprotein receptor-related protein 1(Lrp1) G3UZJ4 PRDX5 peroxiredoxin 5(Prdx5) E0CX99 NWD1 NACHT and WD repeat domain containing 1(Nwd1) Q61171 PRDX2 peroxiredoxin 2(Prdx2) Q8R3F5 MCAT malonyl CoA:ACP acyltransferase (mitochondrial)(Mcat) E9Q1K0 SNX25 sorting nexin 25(Snx25) Q921M7 FAM49B family with sequence similarity 49, member B(Fam49b) P14131 RPS16 ribosomal protein S16(Rps16) P62264 RPS14 ribosomal protein S14(Rps14) E9Q459 Zfp160 zinc finger protein 160(Zfp160) Q8VI93 OAS3 2'-5' oligoadenylate synthetase 3(Oas3) E9Q341 STK36 serine/threonine kinase 36(Stk36) Q922R8 PDIA6 protein disulfide isomerase associated 6(Pdia6) P97351 Rps3a1 ribosomal protein S3A1(Rps3a1) P14901 HMOX1 heme oxygenase 1(Hmox1) P48774 GSTM5 glutathione S-transferase, mu 5(Gstm5) G3UWY0 GUF1 GUF1 homolog, GTPase(Guf1)

162

Q62348 TSN translin(Tsn) H3BKD0 HNRNPK heterogeneous nuclear ribonucleoprotein K(Hnrnpk) Q61206 PAFAH1B2 platelet-activating factor acetylhydrolase, isoform 1b, subunit Q3UGX2 SPTB spectrin2(Pafah1b2) beta, erythrocytic(Sptb) Q8BMK4 CKAP4 cytoskeleton-associated protein 4(Ckap4) Q3U2A8 VARS2 valyl-tRNA synthetase 2, mitochondrial(Vars2) F7APL9 Zfp872 zinc finger protein 872(Zfp872) Q62504 SPEN SPEN homolog, transcriptional regulator (Drosophila)(Spen) Q8BIJ6 iars2 isoleucine-tRNA synthetase 2, mitochondrial(Iars2) O88704 HCN1 hyperpolarization-activated, cyclic nucleotide-gated K+ 1(Hcn1) E9Q1R5 CACNA1A calcium channel, voltage-dependent, P/Q type, alpha 1A E9PWX8 MAST4 microtubulesubunit(Cacna1a) associated serine/threonine kinase family member Q01768 NME2 NME/N4(Mast4)M23 nucleoside diphosphate kinase 2(Nme2) E9Q0E7 Vmn2r115 vomeronasal 2, receptor 115(Vmn2r115) A0A0A6YW TRAF3IP3 TRAF3 interacting protein 3(Traf3ip3) Q69Z98R4 BRSK2 BR serine/threonine kinase 2(Brsk2) P63276 RPS17 ribosomal protein S17(Rps17) Q08189 TGM3 transglutaminase 3, E polypeptide(Tgm3) Q9CQ60 PGLS 6-phosphogluconolactonase(Pgls) E9PZ12 MAGI1 membrane associated guanylate kinase, WW and PDZ domain Q5F293 ZBTB4 zinccontaining finger 1(Magi1)and BTB domain containing 4(Zbtb4) Q80YR6 RBBP8 retinoblastoma binding protein 8(Rbbp8) D3Z6I7 CALD1 caldesmon 1(Cald1) P01863 Ighg2a immunoglobulin heavy constant gamma 2A(Ighg2a) Q9JKF7 MRPL39 mitochondrial ribosomal protein L39(Mrpl39) P61087 UBE2K ubiquitin-conjugating enzyme E2K(Ube2k) Q70FJ1 AKAP9 A kinase (PRKA) anchor protein (yotiao) 9(Akap9) A0A075B5K Igkv1-99 immunoglobulin kappa variable 1-99(Igkv1-99) Q9CZD38 GARS glycyl-tRNA synthetase(Gars) Q8VCE1 DNAJC28 DnaJ heat shock protein family (Hsp40) member C28(Dnajc28) Q76MZ3 PPP2R1A protein phosphatase 2, regulatory subunit A, alpha(Ppp2r1a) A0A0N4SV3 SERBP1 serpine1 mRNA binding protein 1(Serbp1) P629832 RPS27A ribosomal protein S27A(Rps27a) F7C528 CHD3 chromodomain helicase DNA binding protein 3(Chd3) P21550 ENO3 enolase 3, beta muscle(Eno3) E9Q9X4 NOS3 nitric oxide synthase 3, endothelial cell(Nos3) P70166 CPEB1 cytoplasmic polyadenylation element binding protein 1(Cpeb1) E9Q903 CTNND1 catenin (cadherin associated protein), delta 1(Ctnnd1) Q9CQE8 2700060E02 RIKEN cDNA 2700060E02 gene(2700060E02Rik) E9Q8I0 CFHRik complement component factor h(Cfh) P63073 EIF4E eukaryotic translation initiation factor 4E(Eif4e)

163

A0A0G2JD ANKRD17 ankyrin repeat domain 17(Ankrd17) B7ZNL2Z9 NAP1L4 nucleosome assembly protein 1-like 4(Nap1l4) Z4YJE4 OBSCN obscurin, cytoskeletal calmodulin and titin-interacting P63325 RPS10 ribosomalRhoGEF(Ob proteinscn) S10(Rps10) Q8QZX2 HAUS3 HAUS augmin-like complex, subunit 3(Haus3) P48036 ANXA5 annexin A5(Anxa5) E9Q638 ADAM9 a disintegrin and metallopeptidase domain 9 (meltrin A0A140T8I9 PI4KA phosphatidylinositolgamma)(Adam9) 4-kinase, catalytic, alpha polypeptide(Pi4ka) Q62261 SPTBN1 spectrin beta, non-erythrocytic 1(Sptbn1) E9Q5A8 CEP250 centrosomal protein 250(Cep250) O88623 USP2 ubiquitin specific peptidase 2(Usp2) E9Q3P9 RAB11A RAB11A, member RAS oncogene family(Rab11a) Q66JQ7 KNL1 kinetochore scaffold 1(Knl1) E9Q414 APOB apolipoprotein B(Apob) A0A0J9YV7 NSUN7 NOL1/NOP2/Sun domain family, member 7(Nsun7) Q3UV111 KRT6B keratin 6B(Krt6b) P17751 TPI1 triosephosphate isomerase 1(Tpi1) P06745 Gpi1 glucose phosphate isomerase 1(Gpi1) Q99N32 KLB klotho beta(Klb) E9QA63 MACF1 microtubule-actin crosslinking factor 1(Macf1) P10852 SLC3A2 solute carrier family 3 (activators of dibasic and neutral amino acid B7ZCJ0 ARHGAP21 Rhotransport), GTPase member activati 2(Slc3a2)ng protein 21(Arhgap21) E9PVZ8 GOLGB1 golgi autoantigen, golgin subfamily b, macrogolgin 1(Golgb1) Q8BQM9 MED12L mediator complex subunit 12-like(Med12l) Q8BQM8 EML5 echinoderm microtubule associated protein like 5(Eml5) Q9WV71 ASB4 ankyrin repeat and SOCS box-containing 4(Asb4) Q4VWZ5 DBI diazepam binding inhibitor(Dbi) D3Z1M2 AGBL1 ATP/GTP binding protein-like 1(Agbl1) Q8BLJ6 KLHL4 kelch-like 4(Klhl4) Q7TQH0 ATXN2L ataxin 2-like(Atxn2l) P58389 PTPA protein phosphatase 2 protein activator(Ptpa) A0A0A6YX DCLK2 doublecortin-like kinase 2(Dclk2) P1486971 RPLP0 ribosomal protein, large, P0(Rplp0) A0A087WQ KTN1 kinectin 1(Ktn1) Q3UL97F8 atf7ip2 activating transcription factor 7 interacting protein 2(Atf7ip2) P14206 RPSA ribosomal protein SA(Rpsa) D3YVN7 Gm9755 Tu translation elongation factor pseudogene(Gm9755) F6W2Q5 EPS15 epidermal growth factor receptor pathway substrate 15(Eps15) B1ART2 VPS13D vacuolar protein sorting 13D(Vps13d) E9QM99 dock10 dedicator of cytokinesis 10(Dock10) Q5SXJ3 BRIP1 BRCA1 interacting protein C-terminal helicase 1(Brip1)

164

Q02788 COL6A2 collagen, type VI, alpha 2(Col6a2) Q9EQZ7 RIMS2 regulating synaptic membrane exocytosis 2(Rims2) A0A087WR PFKFB2 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 2(Pfkfb2) O354G0 52 TNXB tenascin XB(Tnxb) B7ZNR9 map4k4 mitogen-activated protein kinase kinase kinase kinase 4(Map4k4) D3Z763 BOC biregional cell adhesion molecule-related/down-regulated by E9Q7S4 SPATS2L spermatogenesisoncogenes (Cdon) associate bindingd, protein(Boc) serine-rich 2-like(Spats2l) Q96DY5 RNF112 ring finger protein 112(Rnf112) O54786 DFFA DNA fragmentation factor, alpha subunit(Dffa) Q9DB20 ATP5O ATP synthase, H+ transporting, mitochondrial F1 complex, O F6YVP7 Gm10260 ribosomalsubunit(Atp5o) prote in S18 pseudogene(Gm10260) Q6NZH9 RASGRP3 RAS, guanyl releasing protein 3(Rasgrp3) A0A0R4J0T IKBKB inhibitor of kappaB kinase beta(Ikbkb) 4

165

Appendix D

ANIMAL APPROVED PROTOCOL

166

Appendix E

REPRINT PERMISSION

167

168

169

170

171

172

173