|LL BE! .3... ts???» ivy-LXI! Q. tn waft?! 2‘26:- Cj. .3. ,w n!(!. Jul/n: .. c_s‘!‘..l.o.oflrlkflh.nfigp , $.35 :41; Jifiluvsi¢r 0’. mt): Val‘finuwfiw 1.05013 LI! fig _ .‘ ail-its“: ul- 0 ‘ 01%.. .zcflipi am3.§s .9455». Manny... 2...... 3:41.}. 3 $54!? it 1019 ._._— LIBRARY Michigan State University_____ This is to certify that the dissertation entitled FUNCTIONAL ANALYSIS OF CYTOPLASMIC y—ACTIN MUTATIONS CAUSING NON-SYNDROMIC, PROGRESSIVE AUTOSOMAL DOMINANT HEARING LOSS presented by Soumya Korrapati has been accepted towards fulfillment of the requirements for the Doctoral degree in Genetics aKwW/W Major Professor’s SignatureCJ—w ;/ 0d 2‘00? Date MSU is an Affirmative Action/Equal Opportunity Employer PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DAIEDUE DAIEDUE DAIEDUE 5108 K:IProj/Aoc&Pres/CIRCIDateDue.indd 7—* FUNCTIONAL ANALYSIS OF CYTOPLASMIC y-ACTIN MUTATIONS CAUSING NON-SYNDROMIC, PROGRESSIVE AUTOSOMAL DOMINANT HEARING LOSS By Soumya Korrapati A DISSERTATION Submitted to Michigan State University In partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Genetics 2009 FUNCTIONAL ANALYSIS OF CYTOPLASMIC y-ACTIN MUTATIONS CAUSING NON-SYNDROMIC, PROGRESSIVE AUTOSOMAL DOMINANT HEARING LOSS By Soumya Korrapati Mutations in cytoplasmic y-actin cause non-syndromic, post-lingual, autosomal- dominant, progressive sensorineural hearing loss. LLC-PK1- CL4 cells provide a model system to study the distribution of actins and the role of y-actin and its mutations in repair of damaged structures like microvilli. lmmunohistochemistry and confocal localization studies showed that B-actin was found primarily at the periphery of cells while y-actin is abundant in the perinuclear space and cytoplasm of the cell. Exogenous expression of mutant y-actins showed distribution to all the actin structures in the cell; the periphery, stress fibers and perinuclear space. In response to exogenous espin, filamentous mutant actins co-localized with espin in the microvilli. Co-transfection of espin and mutant actin resulted in each of the mutants co-localizing with filamentous actin in the microvilli. Cytochalasin D treatments of WT y-actin and mutant y-actins showed no difference in the repair of the damaged microvilli. Measurements of the lengths of microvilli however indicated that the microvilli expressing mutant actins were ~20-25% shorter than the WT y-actin microvilli. Quantitative FRAP assays heat shock promoter, were used to over-express the mutant actins in the zebrafish. Confocal images of hair cells of cristae and maculae from fish at 4-day post fertilization (dpf) showed that five out of six mutants are expressed in hair did not reveal any differences in the recovery rates between WT and mutants actins. Our data suggest that mutations in y-actin exhibit subtle phenotypes and might interfere with basic actin assembly dynamics. To determine the physiological relevance of the cell culture data, a multi-site Gateway system based EGFP tagged WT and mutant y-actin constructs were made to create a transgenic zebrafish model for the y-actin mutants. These constructs, under a heat shock promoter, were used to over-express the mutant actins in the zebrafish. Confocal images of hair cells of cristae and maculae from fish at 4-day post fertilization (dpf) showed that five out of six mutants are expressed in hair cells and stereocilia. Fish harboring these mutations did not Show any morphological defects and appeared healthy like the WT counterparts. This dissertation is dedicated to my beloved grandmothers: Sitaramamma Kurada and Seshamma Korrapati iv ACKNOWLEDGMENTS The credit for the completion of my doctoral degree goes to many wonderful people I have known in my life. I am firstly very grateful to my mentor, Dr. Karen Friderici. She has always shown tremendous faith in my scientific abilities and given me freedom to pursue various avenues during my years as her graduate student. I especially appreciate this trait in her as this has helped me believe in myself. I have learnt a great deal from her. Patience, generosity, honesty and open mindedness are few of the virtues I believe I have inherited from her. She has been a wonderful mentor in every sense. I couldn’t have asked for a better scientific advisor. I am very thankful to my committee members. Dr. Pat Venta, Dr. Steven Heidemann, Dr. David Kohrman and Dr. Laura McCabe for their guidance and insightful suggestions throughout my graduate career. Many thanks to our collaborators Dr. Pete Rubenstein, Dr. Suzanne Thiem and Dr. Jarema Malicki for their guidance and scientific help. Hard work, sincerity and perseverance are the key to success in life. My parents have always instilled these values in me. They are my support system and have encouraged and supported me all my life in all my endeavors. I am eternally grateful to them for being such wonderful parents. I am also thankful to my brother for his continued love and support. I am lucky to have a sibling who always stands by me and is proud of my achievements. My Sincere thanks to my mother-in-Iaw and father-in-Iaw who have loved me and supported me all along my graduate career. My heartfelt gratitude to all the lab members of the Friderici lab. They have been wonderful colleagues who made working in the lab a delight. I am especially thankful to Meghan, Ellen, Mei, Kathy and Donna. They all are bright scientists and great friends whom I can always count on. I am very lucky to have found a home in the Friderici lab. I have met wonderful people during my graduate career and formed lasting friendships. A big thanks to Meghan and Gabe for being great friends. Special thanks to Nitin and Akanksha for their continued love and support. A big thank you to all my friends, (Sireesha, Mukta, Sachin, Amit, Prerna, Lavanya, Kiran, Bhaskar, Priya, Chetan, Rahul, Gauri, Sundari, Sudhakar, Sandhya, Erin, Gabby, Rachel, Rabeah, Tuddow), for being there for me during my graduate career. I greatly appreciate their presence in my life. Part of the credit for the completion of my PhD goes to my husband and best friend, TejaS. He is my support system. He believed in me more than I believed in myself, never let me give up and always pushed me to become a better scientist. His constant love and support have been instrumental in attaining the degree. I am extremely lucky to have found such a wonderful life partner. Lastly I am thankful to Turbo Taq, my cat for having found me-because he a constant source of joy in my life. vi TABLE OF CONTENTS LIST OF TABLES .................................................................................. ix LIST OF FIGURES ................................................................................. x LIST OF ABBREVIATIONS ..................................................................... xii INTRODUCTION ................................................................................... 2 REFERENCES .................................................................................... 39 CHAPTER 1 LITERATURE REVIEW 1. Introduction ...................................................................................... 2 1 Anatomy of the ear .......................................................................... 2 2. Inner Ear and Hearing ......................................................................... 5 2.1 Inner ear physiology ..................................................................... 5 2.2 Sensory hair cells ........................................................................ 8 2.3 Stereocilia ................................................................................. 11 2.4 Hearing .................................................................................... 14 3. Hearing Loss ................................................................................... 16 3.1 Characterization of hearing loss .................................................... 16 3.2 Age-related hearing loss (ARHL) or Presbycusis ............................... 17 3.3 Genetics of hearing loss ............................................................... 19 4. Actin .............................................................................................. 25 4.1 Cytoplasmic actins ....................................................................... 25 4.2 Regulation of actin cytoskeleton ..................................................... 29 4.3 Actin in the ear ........................................................................... 30 4.4 y-actin mutations and hearing loss .................................................. 31 5. Espin ............................................................................................ 33 5.1 Distribution of espin isoforms ........................................................ 33 5.2 ESPN mutations and hearing loss .................................................. 34 6. LLC-PK1-CL4 transfection model ........................................................ 36 7. Zebrafish as a model to study human hearing loss ................................. 37 References ..................................................................................... 39 CHAPTER 2 MUTATIONS IN CYTOPLASMIC y-ACTIN INTERFERE WITH ACTIN FILAMENT DYNAMICS Abstract ......................................................................................... 48 Introduction .................................................................................... 49 Materials and Methods ...................................................................... 53 Results .......................................................................................... 56 Discussion ...................................................................................... 83 vii References ..................................................................................... 87 CHAPTER 3 TRANSGENIC ZEBRAFISH MODEL OF y-ACTIN MUTATIONS Abstract ........................................................................................ 92 Introduction ................................................................................... 93 Materials and Methods ..................................................................... 97 Results ......................................................................................... 99 Discussion and future directions ........................................................ 113 References ................................................................................... 1 16 CHAPTER 4 SUMMARY, DISCUSSION AND FUTURE DIRECTIONS ............................ 118 viii LIST OF TABLES Table 1-1 Genes underlying non-syndromic deafness .............................. 21 Appendix 1: List of primers used in this chapter ....................................... 115 ix LIST OF FIGURES F igure1-1 Schematic representation of human ear ........................................ 4 Figure1-2 Schematic of cross section of inner ear ......................................... 6 Figure1-3A Cross section of organ of Corti ................................................ 1O Figure1-3B Scanning electron microscopy images of OHC and IHC ............... 10 Figure1-4 Stereocilia bundle ................................................................... 13 F igure1-5 Various interstereociliary cross-links ........................................... 13 i F igure1-6 Schematic of stereocilia deflection .............................................. 15 Figure1-7 Treadmilling of actin filaments ................................................... 28 Figure2-1 Ribbon structure model representing the location of y-actin mutations .......................................................................................... 52 Figure2-2 Spatial distribution of endogenous cytoplasmic actins in LLC-PK1-CL4 cells .................................................................................... 57 Figure2-3 Exogenous-expression of cytoplasmic actins in LLC-PK1-CL4 cells .......................................................................................... 59 Figure2-4 Over-expression of y-actin mutants in LLC-PK1-CL4 cells ............... 60 Figure2-5 Exogenous expression of espin results in lengthening of brush border microvilli actins in LLC-PK1-CL4 cells .......................................... 62 F igure2-6 Localization of WT or mutant y—actins with espin in LLC-PK1-CL4 cells .......................................................................................... 64 Figure2-7 Over-expression of mutant y-actin results in shorter microvilli ............................................................................... 67 Figure2-8 Microvilli lengths in cells before and after cytochalasin D treatment...69 Figure2-9 A, B Fluorescence recovery after photobleaching of EGFP-WT y-actin ........................................................................ 73,74 F igure2-QC-H FRAP of EGFP-tagged y-actin mutants ................................. 76 Figure2-10A-F Recovery graphs of y-actin mutants following FRAP ........ 78,79,80 Figure 2-11 Expression of EGFP and DSRed-espin in CL4 cells ..................... 82 Figure3-1 Sketch of an adult zebrafish ear ................................................. 94 F igure3-2 Multi-site gateway-based construction of expression vectors ............................................................................. 102,103,104 Figure3-3 EGF P expression in the zebrafish embryo following heat-Shock ...... 106 F igure3-4 Expression of y-actin mutants in the crista of zebrafish ear ............. 109 Figure3-5 Expression of y-actin mutants in the macula of zebrafish ear .......... 111 “Images in this dissertation are presented in color” xi ABM ACTB ACTG1 ADP ARHL ATP Att BB CD CMV dpf EGFP EnaNASP ENU ESPN EVH1 ,EVH2 F-actin FRAP G-actin hpf IHC LIST OF ABBREVIATIONS Actin-bundling module B-actin y-actin Adenosine di-phosphate Age-related hearing loss Adenosine tri-phosphate Attach Sites Brush border Cytochalasin D Cytomegalovirus Days post fertilization Enhanced green fluorescent protein Vasodilator-stimulated phosphoprotein N-ethyl N-nitrosourea mutagenesis Espin EnaNASP homology domain1,2 Filamentous actin Fluoresence Recovery After Photobleaching Globular actin Hours post fertilization Inner hair cell xii IRES KO LLC-PK1-CL4 OHC PFA pDest pDONR RR SEM siRNA SNHL UTR WASP WH2 WT Internal ribosomal entry site Knock-out Pig proximal kidney epithelial clone 4 cells Outer hair cell Para formaldehyde Destination vector Donor plasmid Proline rich Scanning elctron microscopy Small interference RNA Sensorineural hearing loss Untranslated region Wiskott-aldrich syndrome protein Wiskott-aldrich syndrome protein homology domain Wild type xiii CHAPTER 1 Literature Review Introduction The goal of my research is to evaluate how cytoplasmic y-actin mutations cause autosomal dominant, non-syndromic, late-onset progressive hearing loss. The approach I use relies primarily on a cell culture model but I have also explored a zebrafish model to examine the effect of mutations in y-actin. To evaluate the appropriateness of these models, I will provide a brief review of the mammalian ear, cell biology and physiology of inner ear, the process of hearing and the genetics of hearing loss. 1 Anatomy of the ear The mammalian ear is compartmentalized into outer ear, middle ear and inner ear (Figure 1-1). Mechanical stimulus is converted into electrical Signal by the orchestrated activity of these three compartments. The outer ear is made up of an auricle or pinna and a 2 cm auditory canal [1]. The primary function of the pinna is to concentrate and direct the sound waves into the auditory canal. It also serves as a Shield to protect the middle ear structures from damage. Sound waves, which are in the form of pressure waves, are collected by the ear canal and channelized to the middle ear. A stretched membrane known as the eardrum or tympanic membrane, and three interconnected ossicles-incus, malleus and stapes, constitute the middle ear [1]. The pressure waves from the ear canal are received by the tympanic membrane causing it to vibrate. These vibrations set the ossicles in motion, which amplify the sound waves. These waves are then transmitted to the inner ear via the stapes. The cochlea in the inner ear is responsible for converting the sound waves into an electrical Signal while the vestibule maintains a sense of gravity, acceleration and balance. Both cochlea and the vestibule send the electrical signal to the brain via the auditory nerve. In the following section I will describe the inner ear components in detail. ' ' .Ve‘stlbule...‘ . So'mieircuu .. I:: ...... ....... Figure 1-1: Schematic representation of the human ear. It is divided into outer, middle and inner ear. Auricle and auditory canal constitute the outer ear, while the tympanic membrane and the three middle ear ossicles (incus, malleus and stapes) form the middle ear. The vestibule and the cochlea constitute the inner ear. Semicircular canals, utricle, saccule and the endolymphatic canal form the vestibular system while organ of Corti and membranous labyrinth constitute the cochlea (modified from [1]). 2 Inner Ear and Hearing 2.1 Inner ear physiology The vestibule consists of five-end organs namely saccule, the utricle, and three semicircular canals, which respond to linear and angular accelerations, and are responsible for maintaining balance [1] (Figure 1-1). The cochlea, the auditory organ, is a coiled snail-like structure. Both vestibule and cochlea are derived from the otic placode and hence share many structural and functional aspects [1]. The cochlea is a complex structure that includes three fluid filled chambers (also known as membranous labyrinth) and a sensory epithelium, the organ of Corti. Specialized structures separate these three chambers (Figure 2). Scala vestibuli is separated from scala media by a sheet of cells called Reissner’s membrane while scala media is separated from scala tympani by basilar membrane (reviewed in [1-3]). The lateral boundary of Scala media is the stria vascularis, a complex cellular structure, which is a key player in the generation of endocochlear potential and is home to many Na-K-ATPases, which play a critical role in K+ recycling [4] (Figure 1-2). Scala vestibuli and scala tympani are actually contiguous and are filled with perilymph, while scala media is filled with endolymph. A cross-section through a turn of the snail shaped cochlea shows that the apical side of the organ of Corti is bathed in the endolymph while the baso-Iateral side is in the perilymph (reviewed in [1-3]) (Figure 1-2). Figure 1-2: Schematic of the cross-section of inner ear. It consists of three fluid filled compartments namely scala vestibuli, scala media and scala tympani. Scala vestibuli and scala media are separated by Reissner's membrane, while the basilar membrane separates scala media and scala tympani. Scala media on one end is lined by stria vascularis. Organ of Corti consists of inner and outer hair cells and non-sensory supporting cells and rests on the basilar membrane. Tectorial membrane lies on the hair cells, contacting a subset of the stereociliary bundles (modified from [5]). Perilymph and endolymph differ from each other in their ion concentrations. Where perilymph is high in Na+ (140mM) and low in K+ (3.5mM), endolymph has a rare composition of very high K+ (about 150mM) and very low Na” (1mM) (reviewed in [1, 3, 4]). The difference in ion concentrations results in an endocochlear potential of +80 mV, believed to be the largest in the body [2]. There are advantages to having K+ as the major charge carrier for sensory transduction in the inner ear. K+ being the most abundant ion in the cytosol, influx of K+ results in the least change in the cytosol and, influx and extrusion of K+ ions are energetically inexpensive to the cell [4]. Various cell types like the ion-transporting epithelia, sensory epithelia, and relatively unspecialized epithelia add an additional layer of complexity to the cochlear structure [6, 7]. The ion-transporting epithelia are the stria vascularis and regions of the vestibular system, which as mentioned earlier play an active role in recycling the K“ ions to maintain the endolymph composition. Reissner’s membrane and the roofs of the semicircular canals form the unspecialized epithelia, which help in separating the two fluid compartments [3]. Organ of Corti is the sensory epithelium of the cochlea while the vestibular system consists of five sensory epithelial sheets, namely the maculae of the utricle and saccule and the three cristae, one in each semi-circular canal [3]. The sensory epithelia are composed of sensory hair cells and supporting cells. Various supporting cells like pillar cells, Deiters cells, and cells of Hensen, provide support to and help maintain the hair cells. They surround the hair cells such that no two hair cells are in contact with each other [6] (Figure 1-3A). Tight junctions seal the apical boundaries of hair cells and surrounding supporting cells [1]. The supporting cells are inter-connected by gap junction proteins, which play a critical role in the homeOstaSis of the cochlear fluids [6]. Gap junctions are sites of direct communication between adjacent cells via continuous aqueous hemichannel pores known as connexons. These pores allow the passage of small metabolites (glucose, ATP,~1200Da), messengers (CAMP, inositol 1,4,5 triphosphate-IPS) and ions thereby acting as sites of electrical and chemical coupling [3]. Connexin (Cx) protein family members form the hemichannel pore or connexon. Five of the twenty-one connexin isoforms are expressed in the mammalian cochlea [8]. Of these, mutations in the OX 26 isoform account for 50% of non-syndromic autosomal recessive hearing loss [8]. 2.2 Sensory hair cells In the organ of Corti, inner hair cells (IHCS) and outer hair cells (OHCS) are the two kinds of sensory hair cells, while type 1 and type 2 are the sensory hair cells found in the vestibular system (reviewed in [1, 3]). These two kinds of hair cells Show unique innervation patterns. In the organ of Corti, IHCs are innervated exclusively with afferent nerves and are considered the true sensory cells, as they send the impulses to the brain via the afferent auditory nerve [7]. OHCs, on the other hand, act as amplifiers of the auditory Signal and are innervated with efferent nerves [7]. The sensory hair cells are overlaid by an extracellular matrix known as tectorial membrane in cochlea, otolithic membranes in macular organs and cupulae in cristae [3].On the bottom, the sensory epithelia rest on a basement membrane known as basilar membrane (Figure 1-2). IHCS are pear shaped cells with centrally located nucleus [3] (Figure 1— 3A). A single row of IHCs runs along the cochlear duct. OHCS are cylindrical in shape with the nucleus at the bottom of the cylinder [3, 7] (Figure 1-3A). Three rows of OHCS are present in the organ of Corti, but in some mammals a fourth or fifth row has also been observed [7]. Each hair cell has 30-300 apical projections on its surface, known as stereocilia [1, 3, 7, 9-11]. The stereocilia on the IHCS are arranged in a W shaped pattern, while those on OHCS have a V shaped arrangement [3] (Figure 3B). Unlike true ciliary structures, stereocilia are in fact, derivatives of actin-based microvilli [1, 3, 7, 9-11]. Because y-actin plays a central role in maintaining cellular structure, it is conceivable that any of the cells described so far could be adversely affected by mutations in this protein. However, the pivotal role of stereocilia in hearing and the exceptional enrichment of cytoplasmic actin in these structures make them attractive targets for study of the potential affect of the mutations in y-actin. Therefore the following sections are especially pertinent to my thesis work. Figure 1-3A: Cross section of organ of Corti. There are three rows of outer hair cells that are cylindrical in shape and the nucleus is towards the bottom of the cell body. On the other hand, inner hair cells are pear-shaped and the nucleus is the center of the cell body. IHC: inner hair cell, OHCS: outer hair cells, TM: tectorial membrane, BM: basilar membrane (modified from [12]). --,,-.-.-;.\ :29 Figure 1-3B: Scanning electron microscopy images of OHC and IHC. Stereocilia on outer hair cells are arranged in a V shape and those on inner hair cells have a W shaped pattern (modified from[13]). 2.3 Stereocilia Stereocilia are arranged in precisely specified rows of increasing heights forming a characteristic staircase-like structure on the surface of the cell [1, 3, 7, 9-11] (Figure 1-4). The longer stereocilia of all hair cells are in contact with the tectorial membrane [3]. The arrangement of stereocilia on the hair cells iS tightly regulated; length and width, position and orientation are all critical to proper hearing. The position of the stereocilia is determined by the true cilium, the kinocilium. Stereocilia development is well studied in the avian auditory hair cells and mammalian vestibular hair cells [10, 14]. Barring a few inter-species and inter-organ differences, the overall features of stereocilia development are as follows. A single kinocilium appears on the surface of the hair cell, surrounded by numerous precursor microvillus projections. The microvilli then stop elongating and widen instead. Eventually the kinocilium migrates to the edge of the cell surface, thus determining the polarity of the stereociliary bundle. The microvilli then elongate sequentially to form the staircase—like bundle, where the longest stereocilia is next to the kinocilium. The stereociliary rootlets project into the cuticular plates to anchor the stereocilia. The microvilli that fail to become part of the staircase regress into the cell. The kinocilium in the cochlear hair cells also regresses into the cell and is reduced to a basal body. Thus the auditory hair cell gains the stereociliary staircase structure (Figure 1-4) and is devoid of the kinocilium. In addition to detecting sound amplitude, hair cells also detect sound frequencies based on their position. Interestingly, the frequencies are distributed across the stereocilia such that higher frequencies are detected by stereocilia on 11 the basal turn of the cochlea while lower frequencies are detected by the stereocilia on the apical turn. Thus the stereociliary structure, along with the lengths, numbers and bundle shape are important for the sensitivity of the hair cell. Tilney & Tilney very accurately note in their paper that the hair cell in the cochlea “is perhaps the most extraordinary example of precision engineering seen anywhere in the vertebrate organism [10]”. Individual stereocilia are held together at various points along their entire length such that they are deflected together, in response to sound waves [1, 3]. They are held together by ankle links at the base, by side links along the sides and by tip links at the top (Figure 1-5) [1, 7, 11, 15, 16]. 12 Figure 1-4: Stereocilia bundle. The stereocilia form a staircase-like structure consisting ofiincreasing heights of individual stereocilia. The tallest of the stereocilia lies next to the kinocilium-which decides the polarity of the bundle. The bundle is held together at different locations along the entire length (modified from [15]). tip link 5” - shaft connectors Figure 1-5: Various interstereociliary cross-links. Tip links connect the top of a short stereocilium and shaft of the adjacent taller stereocilium. Tip links contain the gated ion channels. Shaft connectors connect Shafts of adjacent strereocilia and the ankle links connect adjacent stereociliary ankles. Modified from [3]. 2.4 Hearing In response to stimuli/vibrations from the middle ear, the basilar and tectorial membranes are set in motion. This motion causes the stereocilia to deflect as a whole, towards the longest stereocilia (Figure 1-6). This deflection creates tension at the tip links, which are filaments that connect the apical surface of the Shorter stereocilium to the lateral wall of the adjacent taller stereocilium [1]. This tension causes the gated channels to open letting the K“ ions from the endolymph into the hair cells, causing them to depolarize. The voltage channels open and Ca 2+ions enter the cell. This triggers the release of neurotransmitter onto the afferent fibers. The nerves then transmit the electrical Signal to the brain. Recycling of K+ ions occurs basolaterally from hair cells and via the gap junctions of the supporting cells. The gated channels close, the cells recover their resting potential and the stereocilia regain their shape [3]. 14 f” M crlr'i” ’ ""’~ r I rrfffftp :3 . 3‘ r(”'," r- '1 ' , t.fff'fff ' ,vr' f A v MUM/[”1355 Figure 1-6: Schematic of stereocilia deflection. Sound-induced excitation results in nanometer-scale deflections of the hair bundle towards the longest stereocilia. This opens mechanically gated ion channels (shown in red)(modified from [16]). 3 Hearing Loss 3.1 Characterization of hearing loss Humans hear in the range of 20-20,000 hertz and 28-35 million Americans suffer from a hearing impairment, that affects their quality of life [1, 17, 18]. Seventeen out of 1000 children under the age of 18, and 34 out of 1000 people by the age of 65 suffer from hearing loss, the statistics increase to 40-50% in individuals over 75 (NIDCD, [17, 19]). Hearing loss can be divided into various categories. It is classified into mild (20-40dB), moderate (40-60dB), severe (60-80dB) or profound (80 and more) based on the loudness required to produce a response at various frequencies [20]. Hearing loss due to mutations in the y-actin gene begins as mild loss at high frequencies which progresses to profound loss with advancing age [21, 22]. Hearing loss can be conductive, sensorineural or mixed. Conductive hearing loss relates to hearing impairment resulting from the malfunction of the outer or middle ear structures. Tympanic membrane perforation, ossicular discontinuity or fixation, earwax build-up, otitis externa and otitis media (external ear and middle ear infections, respectively) or reactions to certain drugs like chloroquine, may result in conductive hearing loss [23]. This kind of hearing loss is reversible following wax clean up or treatment of the infections [23]. Sensorineural hearing loss (SNHL) results from the pathologies of the inner ear and the auditory nerve. It is believed to be the cause of 70% of all hearing loss 16 [24]. SNHL can result from aging, noise exposure, toxic drug exposure, viral infections (rubella virus-from mother to fetus), and genetic factors. Hair cells of the mammalian organ of Corti are terminally differentiated and they do not regenerate following damage [2]. Hence SNHL is a permanent condition and there is no cure/treatment. Hearing loss occurring from the malfunction of outer or middle ear along with inner ear or auditory nerve is known as mixed hearing loss. All y-actin mutations described to date cause SNHL. Hearing loss is often accompanied as a phenotype in many human syndromes. Over 400 syndromes with hearing loss have been described [25]. Syndromic hearing loss accounts for 30% of all cases of hearing loss [1]. Non- syndromic hearing loss occurs due to environmental factors and/or genetic factors. The hereditary forms of non-syndromic hearing loss can be Y-Iinked, X- linked, autosomal dominant, autosomal recessive and mitochondrial DNA (mtDNA) linked [20]. Mutations in the y-actin gene cause autosomal dominant, non-syndromic, late onset hearing loss [21, 22]. 3.2 Age related hearing loss (ARHL) or Presbycusis ARHL affects the quality of life of adults, a condition that is a polygenic and/or multifactorial in nature [18]. It results in bilateral hearing loss and is the most common cause of hearing loss in adults [26]. ARHL starts initially as high frequency hearing loss that has adverse affects on communication. ARHL is believed to be a result of age dependent atrophy of cochlea and age related 17 accumulation of noise insults [18]. Based on the region of cochlea that degenerates, ARHL can be sensory, neural, strial or metabolic [18, 26]. Degeneration of the lateral wall of stria vascularis is believed to be one of the contributors to ARHL [18]. Audiograms of such persons show hearing loss across all frequencies, thereby exhibiting a flat loss [26]. This view is supported by evidence of age-related loss of expression of Na-K-ATPaseS in the stria vascularis, followed by dramatic loss of the endocochlear potential [18]. Sensory presbycusis is the most common type of ARHL in adults, whose cochlear pathologies and audiograms are very Similar to those of non-syndromic late-onset hearing loss individuals [18]. Though a significant portion of AHRL (~40-50%) iS genetically determined, the molecular etiology of presbycusis still remains unknown [22]. Barring the age of onset, phenotype and progression of AHRL and late-onset hearing loss are similar. Hence, genes involved in non-syndromic late-onset hearing loss might be excellent candidates to study the pathophysiology of ARHL. y-Actin mutations, we believe is one such promising candidate. All the y-actin mutations identified so far result in hearing loss, like in presbycusis, initially in high frequencies. With age the hearing loss becomes progressive and extends into all frequencies [21, 22, 27-29]. Besides shedding light on the role of y—actin in hair cells, research on these mutations might provide us with clues about the molecular pathology of ARHL. 18 3.3 Genetics of hearing loss The role of genetics has been well established in the pathogenesis of hearing loss. Nearly 130 genetic loci have been identified as the cause of non- syndromic hearing loss, while many of the underlying genes still remain to be determined [19, 25]. Mouse models have been pivotal in dissecting the genetics and pathophysiology of hearing loss [19, 30]. Almost identical ear architecture and physiology between mice and humans, the close genetic relationship (~99% of mice genes have a human orthologue) and the occurrence of spontaneous deaf mice have been invaluable to researchers in unveiling the genetic components of the hearing process [19]. In addition, phenotype-oriented and gene-oriented mutagenesis screens have been other clever techniques used to delineate the various gene products involved in hearing [19]. SNHL can occur due to mutations in genes expressed in any region of the inner ear. A list of proteins expressed in different cell types of the inner ear, mutations in which cause either sensory or vestibular defects in humans and mice, is Shown in Table 1 [20]. The best known and most studied are the connexin family members of which GJBZ, coding for connexin 26 protein, is responsible for causing the most prevalent genetic form of autosomal recessive hearing loss [8]. It is evident from these examples that the inner ear is a very specialized and complex structure, which houses a myriad of proteins. It iS also true that most of these proteins are not ear-specific yet the expression of these genes is critical for normal functioning of this structure. 19 I am particularly interested in the bundling protein espin and the mutations in the actin gene because two of the six initially identified actin mutations are in a predicted protein-binding domain of actin. In addition, one of those (T89l) is identical to a yeast actin mutation found in a complementation screen for an actin-bundling protein [31]. In the following sections, I will describe the role and function of actin and espin proteins in the hair cells and hearing. 20 Table 1-1. Genes underlying isolated deafness as a result of primary defects in hair cells, non-sensory cells and the tectorial membrane or unknown cell type and corresponding mouse mutants (modified from [20]). Primary Forms of human Mouse defect Gene Gene product deafness mutants Hair MYO7 A Myosin VIIA DFNB2iretinopathy (Usher Shaker- cells (motor protein) 1 B) DFNA11 1 (SM) Myosin XV Shaker- MYO15 (motor protein) DFNB3 2 (Sh2) . Snell's M06 My°s.'“ V' (”mm DFNA22icardiomyopathy waItzer proteIn) (SV) DFNB37 MYO3A My°Sin ”'A . DFNB30 (motor protein) MYO1A MV°SI” 'A(m°t°' DFNA48 proteIn) v-Actin ACTG1 (cytoskeletal DFNA20 (DFNA26) protein) Harmonin (PDZ Deaf domain- DFNB18iretinopathy . USH1C . . crrcler contaInIng (Usher 1C) ( D for) protein) Deaf circler 2 Jackson (chr-2J) Whirlin (PDZ domain- Whirler WHRN containing DFNB31 (WI) protein) 3 . CDH23 (cell-adhesion 3mg: figet'mpathy Yvwa'tze' protein) 21 Table 1-1 contd Primary Forms of human Mouse defect Gene Gene product deafness mutants Protocadherin-15 . Ames PCDH15 cell-adhesion DFNBZBEretInopathy waItzer . (Usher 1 F) proteIn) (Av) TMIE (transmembrane S inner TMIE domain- DFNB6 p . . (Sr) contaInIng protein) STRC Stereocilin DFNB16 Prestin (anion SLCZ6A5 transporter) DFNB61 8/02635 Espin (actin- Jerker ESPN bundling protein) DFNB36' DFNA‘ (Je) KCNQ4 (K* KCNQ4 channel subunit) DFNA2 TMC1 Deafnes TMC1 (transmembrane DFNB7 (DFNB11), 3 (Dn) channel-like DFNA36 Beethov protein) en (Bth) Otoferlin OTOF (putative vesicle DFNBQ traffic protein) POU4F3 Brn3c POU4F3 (transcription DFNA15 Dreidl factor) (Ddl) DFNB1, Non- Connexin-26 DFNA3ztkeratodermia sensory GJBZ (gap junction (Vohwinkel, palmoplantar Cx26 cells protein) keratoderma, KID, Bart- Pumphrey) Connexin-30 . . DFNB1, GJB6 (gap I“"°t'°” DFNA3:l:keratodermia (KID) 0X30 proteIn) Cx26+"/ Cx30 22 Table 1-1 contd Primary Forms of human Mouse defect Gene Gene product deafness mutants Connexin-31 (gap junction GJB3 Pram”) DFNA Pendrin (l'—CI’ DFNB4ztthyroid goiter -,_ SLCZ6A4 transporter) (Pendred) Pds u-Cristallin CRYM (thym'd . . DFNA hormone-binding protein?) Otoancorin (cell- OTOA surface protein) DFNBZZ Claudin-14 (tight- Cldn14" CLDN14 junction protein) DFNB29 " Cochlin (extracellular COCH matrix DFNA9 component) TMPRSSS TMPRSS3 (transmembrane DFNB8(DFNB10) serine protease) Myosin llA DFNA17tgiant platelets MYHQ (motor protein) (Fechtner) Myosin IIC MYH14 (motor protein) DFNA4 EYA4 E YA4 (transcriptional DFNA10 coactivator) . POU3F4 POU3F4 (targtrgsfnptlon DFN3 fidget (sIf) Bm4"’ 23 Table 1-1 contd Primary Forms of human Mouse defect Gene Gene product deafness mutants Collagen XI (02- Tectorial chain) membra COL11A2 (extracellular dDFSNQLZiEEEfiSfQSm‘ 9.0’1732 ne matrix y p component) a-Tectorin (extracellular DFNA8 (DFNA12), _,_ TECTA matrix DFNB21 TeCta component) Diaphanous-1 Unknow HD I A 1 (cytoskeleton DFNA1 n regulatory protein) DFNA5 Unidentified DFNA5 Dfna5'" Wolframin (endoplasmic- DFNA6 (DFNA14, WFS1 reticulum DFNA38);tdiabetes and membrane optic atrophy (Wolfram) protein) TFCP2L3 TFCP2L3 (transcription DFNA28 factor) Mitochondrial MTRNR1 12$ rRNA ND Mitochondrial MTTS1 tRNAW‘UCN) ND Abbreviation: ND, not defined nomenclature. 4 Actin 4.1 Cytoplasmic actins Actins are highly conserved, abundant proteins found in nucleated cells of the eukaryotic genome, which play pivotal roles in many important cellular processes [32, 33]. In vertebrates, six isoforms namely skeletal muscle a-actin, cardiac muscle a-actin, smooth muscle or-actin, smooth muscle y-actin, cytoplasmic B-actin and cytoplasmic y-actin of actin have been isolated based on their tissue specificity [33]. Though individual genes encode each isoform, the cytoplasmic isoactins are only ~10% different from any of the muscle actins. Interestingly, the cytoplasmic isoactins [3 and y differ at only 4 amino acid positions in the N-terminus of their 375 amino acids [33]. At the nucleic acid level, the cytoplasmic actins differ from each other at the 3’untranslated regions (UTRS) that are believed to be central to mRNA transcript localization in the cell [34]. In most cells the cytoplasmic B and y isoactins are found in the constant ratio of 2:1 [33]. Studies have Shown a spatial and temporal segregation of the cytoplasmic isoforms in different cell types, and until recently it was thought that the distribution of y-actln was ubiquitous and not obviously related to function. B-actin distribution is dependent on the physiological function of the cell and is primarily found localized in the more dynamic structures of the cell such as the growth cone of differentiating neurons and the dendrites and axons of mature neurons [33,35,36] 25 The traditional functions of the ~43 kDa protein family include muscle contraction, cell motility, cell Shape determination, exocytosis, endocytosis, cytokinesis and organelle transport. Recent research confirms the presence and importance of actin in the nucleus. Actin and actin-binding proteins are actively involved in gene expression, chromatin remodeling and pre-mRNA Splicing of many nuclear genes [32]. In the cytoplasm, actin interacts with an unusually wide variety of interacting proteins to form higher order structures such as bundles or networks to perform various functions in the cell [32, 33]. Globular actin (G-actin), which is bound to ATP and/or ADP, polymerizes to form filamentous actin (F- actin) in the cell [33, 37, 38]. The F- actin in combination with myosins forms contractile fibrils (e.g. myofibrils of muscle), along with espin, fimbrin, or fascin forms densely packed non-contractile bundles (e.g. core bundles of stereocilia and microvilli) or forms gel-like networks (e.g. cuticular plate of hair cells, leukocytes cortex) in combination with a-actinin [39]. Actin polymerization has been extensively studied in the filopodia, Iammelipodia and the dendritic spines of neurons [37, 38]. Actin filaments are helical polymers of G-actin subunits. These helical polymers consist of a barbed end and a pointed end. ATP-bound actin monomers are added to the barbed end while ADP-bound actin monomers dissociate from the pointed end. The actin polymer hence is polar in nature [38]. Actin binding and actin sequestering proteins like profilin bind to the ATP-G-actin pool in the cell, preventing spontaneous polymerization events [37, 38]. The 26 initial steps of monomer addition are unfavorable and occur slowly, while addition of subsequent monomers become favorable and the process proceeds rapidly [37, 38]. The growing actin filaments are characterized by continuous assembly (at the barbed end) and disassembly (at the pointed end), a process brought about by ATP-dependent hydrolysis of ATP in the filament [37]. Disassembly results from the disassociation of y—phosphate (following ATP hydrolysis), which activates actin severing and debranching proteins, ultimately resulting in the disassociation of ADP-actin at the pointed end [37, 38]. This process of actin assembly and disassembly along the filaments is referred to as treadmilling or retrograde flow [37, 40] (Figure 1-7). Treadmilling is observed in the actin bundles of nerve growth cones, microvilli and newly formed stereocilia (reviewed in [41]). These rates vary from being ~1.5/s'1 in microvillar parallel actin bundles to ~25 times slower in stationary filopodia to ~ 5-50 times slower in Listeria comet tails [42]. For the actin bundle to continue treadmilling and maintain constant length, assembly of actin monomers at the barbed end and the disassembly at the pointed end of the filament should be tightly regulated [41]. Imbalance in these two processes can result in shorter or complete disappearance of the filaments. Though treadmilling is primarily regulated by two key proteins-ADF /cofilin, an actin depolymerizing factor and barbed end capping proteins (as shown in growth cones and Iamellipodia), it is possible to imagine the pivotal roles of other actin binding proteins like myosins, and cross-linking proteins like espin [40, 41]. 27 barbed and pointed end wEEDQEEED) EWZQ‘L ADP ATP Actln treadmllllng Figure 1-7: Treadmilling of actin filaments. ATP bound actin monomers are added at the barbed end, while ADP bound actin monomers dissociate from the pointed end. ATP-actin monomers move along the filament, with ATP moieties hydrolyzing to ADP and dissociation of y—phosphate. This activates actin severing and debranching proteins which eventually cause the dissociation of ADP from pointed end (modified from [40]). This process of simultaneous addition and dissociation of actin monomers is termed treadmilling. 28 4.2 Regulation of actin cytoskeleton Rho GTPases, members of the Ras super family of GTPases, are key regulators of actin cytoskeleton in many cell types [43, 44]. However proteins of the EnaNASP family have been Shown to be pivotal in actin assembly and cell motility. In various cell types these proteins were Shown to localize to the tips of filopodia and edges of Iammelipodia, which are regions of actin reorganization [45]. EnaNASP family proteins contain an amino-terminal homology domain (EVH1), a proline-rich central region and a carboxy-terminal homology domain (EVH2). The proline- rich domain of all the family members contain binding sites for monomeric actin binding protein, profilin [45]. EVH2 domain contains a monomeric-actin-binding Site, filamentous actin-binding Site and an oligomerization facilitating coiled-coil site. EnaNASP family members act as barbed end de-capping proteins, thereby positively regulating actin filament lengths. In organisms like Listeria, these proteins bind directly to profilin at sites of actin organization and hence affect the rates of cell motility. They also perform the role of reducing actin filament branching events by possibly interfering with Arp2/3 complex [45]. Based on the data from my experiments (Chapter 2), it is possible that EnaNASP proteins are critical in regulation of actin filament lengths in LLC-PK1- CL4 cells. 29 4.3 Actin in the ear Three forms of actin filament cytoskeleton are found in the hair cells of vertebrates [39, 46]. These different actin filament assemblies include the extensively cross-linked actin bundles with uniform polarities of the stereocilia, gel-like actin networks of the cuticular plate, into which the base of stereocilia are anchored and the zonula adherens junction, that form a circumferential array of anti parallel actin filaments [10, 39, 46]. Actin filaments in stereocilia are cross- linked by fimbrin and espin, those in cuticular plate are cross-linked by spectrin while the filaments in zonula adherens are cross linked by a-actinin [39, 46]. Various studies have Shown a spatial and temporal segregation of the cytoplasmic isoforms 6 and y-actins in different cell types such as gastric parietal cells, osteoblasts, and neurons [33]. 2D gel electrophoresis followed by immunoblotting displayed, for the first time, the presence of both the cytoplasmic actin isoforms- [3 and y in isolated sensory epithelium of chicken auditory cells [39]. Later studies showed similar results in the guinea pig hair cells [46]. The two cytoplasmic actin isoforms [3 and y—actins are expressed in the stereocilia of hair cells and microvilli of the supporting cells [46]. In the chicken auditory hair cells B-actin expression is found only in the stereocilia and zonula adherens while y-actin expression is seen in all the three actin rich structures of the inner ear: stereocilia, cuticular plate and zonula 30 adherens [39]. The current data suggest that y-actin is the predominant isoform in the cochlea, unlike other major tissues [33, 39, 46]. In the guinea pig auditory cells, B-actin expression is detected in the cuticular plate but the levels are much less compared to the stereocilia and their rootlets [46]. y-Actin expression on the other hand is more uniform along the stereocilia and the cuticular plate [46]. Recent work has shown that treadmilling occurs in the actin bundles of stereocilia as well, Showing for the first time, that these actin bundles are highly dynamic structures [47, 48]. Though the specific function of each of the isoforms is not completely understood, data from rodents suggest that B-actin plays a role in maintaining the hair bundle structure [47]. 4.4 y-actin mutations and hearing loss In 2003, novel mutations in the y-isoform of cytoplasmic actin were identified by our alb and others as the cause of non-syndromic, autosomal progressive, late-onset, progressive hearing loss [21, 22, 49]. Recently, four more mutations (l122V, K118N, E241K and D51 N) in y-actin gene have been . identified to cause non-syndromic progressive hearing loss [27-29]. The Six missense mutations originally discovered were T89l, K118M, P264L, P332A, T278l and V370A [21, 22]. The mutations are found in evolutionarily conserved regions of the y-actin gene and are distributed in 3 of 4 sub-domains of the protein (Figure 8). Based on their location, it is postulated 31 that these mutations might affect key actin processes like filament formation/polymerization (T278I, P264L), interaction with myosin (P332A, T89l, K118M), and bundling or gelation (T89I, K118M). Persons harboring these mutations exhibit similar phenotype and progression of hearing loss, with differences only in onset times [21]. In addition, mutations in this ubiquitously expressed protein cause only hearing loss. These observations indicate that y- actin performs a very specific and pivotal role in hearing and suggests hair cell as the key target. To investigate the affect of these mutations in-vivo and in-vitro, the six mutants were introduced into yeast actin [50]. Yeast cells harboring each of the mutant actins as the only copy of actin were found viable, but four mutants (K118M, T278l, P332A, and V370A) exhibited abnormal mitochondrial morphology. In addition, except T89l, the rest of the five mutants displayed abnormal vacuole formation. Biochemical analysis using purified mutant actins indicated that T89l, K118M and V370A were more susceptible to cofilin-induced disassembly, while P332A was more resistant [51]. In addition, V370A displayed abnormal actin polymerization [50]. The data from the yeast confirm that the missense mutations have a functional consequence for actin. However though informative, these findings do not provide a clear correlation with the deafness phenotype in humans [50]. Hence, investigation of these mutants in a vertebrate and mammalian model 32 system will be important to delineate their physiological effect. In this dissertation, a zebrafish model and a mammalian cell culture model system were explored to study the affect of the y-actin mutations. 5 Espin I 5.1 Distribution of espin isoforms Espin (ESPN) was first identified in the ectoplasmic specializations of Sertoli cells of rat testis [52]. Immunogold labeling localized this novel actin- binding protein to the parallel actin bundles of ectoplasmic specializations [52]. A 2.9kb mRNA transcript encodes this 837-amino acid, 110kDa protein that showed specific localization to rat testis [52]. Later, multiple espin isoforms were detected in the brush border microvilli of rat intestine and kidney, dendritic spines of cerebellar purkinje fibers, and the actin bundles of many sensory cells like the hair cell stereocilia [42, 53-55]. Encoded by a single gene, the multiple isoforms arise from alternative transcriptional start sites and differential splicing [54, 55]. The longer 837-amino acid isoform contains 8 ankyrin-like repeats in its N- terminus, the 116 amino acid COOH-terminus carries an actin bundling module (ABM) [52, 55]. ABM has been shown to be necessary and sufficient to bundle actin filaments in vitro [56]. Besides ABM, there is an additional actin-binding (AB) site in the N-terminus, which makes espin two and half times more efficient at bundling actin filaments than other known actin-bundling proteins [57]. Unlike other actin-bundling proteins, espin is Ca2+ insensitive, another feature that makes espin an ideal actin-bundling protein in the hair cells [55]. The longer 33 isoform also contains two proline rich (PR1, PR2) regions and a WASP (Wiskott- Aldrich syndrome protein) homology (WH2) domain [55]. Espin binds to profilin (ATP-actin binding protein) via its proline rich domains and directly to monomeric actin via its WH2 domain. Recently, it was shown that the WH2 domain is capable of building actin bundles, when targeted to specific sites in the cells [58]. These data hint at a novel mechanism of actin bundling. All the isoforms namely espin 1(found in Sertoli cells), espins 2A, 2B (expressed in Purkinje cells), espins 3A, 33 (found in sensory epithelium and retina) and 4 (expressed in the brush border of intestine and kidney) contain the ABM, encoded by 167 amino acid COOH-terminal peptide [53, 55]. Espins 2A, 2B, espins 3A, 3B and small espin 4, all lack the 8 ankyrin-Iike repeats in their N- terminal, while espins 3A, 3B and espin 4 do not contain a proline rich site and the additional actin binding site in the N-terminus [55]. When espin 3A and 3B isoforms were expressed in cell culture, they were found to localize to the actin bundles of microvilli and increased the average microvilli length from 1.33 :t 0.04 to 6.28 i 0.09 and 6.04 :l: 0.12 pm respectively [55]. These in-vitro data were validated when hair cell stereocilia of espin deficient jerker mice were studied [59]. 5.2 ESPN mutations and hearing loss Deaf jerker mice carry a recessive mutation in the espin gene that results in hair cell stereocilia degeneration resulting in hearing loss and vestibular dysfunction [60]. This spontaneously arising mutation results in a frame-shift in 34 the ABM of the C-terminal of espin protein, thereby creating an espin null. Fluorescence scanning confocal and electron microscopy studies were performed on the stereocilia of deaf jerker mice [59]. Espin deficient actin bundles of stereocilia were found to be shorter than the WT control stereocilia, which subsequently degenerated starting at post natal day 11, coinciding with the onset of hearing in mice [59, 60]. In addition, when espin was over-expressed in the neuroepithelial cells of organ of Corti cultures, lengthening of the actin bundles of stereocilia and microvilli was observed [59]. The above data establish espin as a pivotal protein for the growth and maintenance of stereocilia. Recently, mutations in the espin gene (ESPN) were identified as the cause of autosomal recessive and dominant hearing loss in human populations [61-64]. Two frame-shift mutations, 1988deIAGAG and 2469delGTCA, mapped to the actin-bundling module of the espin gene were shown to co-segregate with recessively inherited hearing loss and vestibular dysfunction in two consanguineous families [61]. A novel mutation, 1757insG, in the WH2 domain of the espin gene was determined to be the cause of recessively inherited congenital deafness without vestibular dysfunction [63]. Autosomal dominant mutations like D744N, R774Q and delK848 were mapped to the actin bundling module of espin protein [62]. Using an LLC-PK1—CL4 cell transfection model, it was shown that while D744N causes elongation of microvilli confined to small patches of the apical surface, delK848 severely impairs the elongation of microvilli, and R774Q does not Show any effect [62]. Interestingly, the mutant 35 phenotypes observed in the cell culture model corresponded to the phenotype and severity of hearing loss in patients [62]. These data highlight the pivotal role of cell culture models in understanding the molecular biology of disease causing human mutant proteins. 6 LLC-PK1-CL4 transfection model The pig proximal kidney epithelial cell line, LLC-PK—CL4 (CL4) forms a well-ordered brush order (BB) microvilli on its apical surface. CL4 cells have been successfully used to study the dynamics of cytoskeletal proteins like MYOl A in these BB microvilli [65]. The parallel actin bundles in the microvilli are cross- linked by actin-bundling proteins like fimbrin/plastin and villin [66]. These cells lack endogenous expression of espin, the third actin-bundling protein. However, when differentiated CL4 cells are transfected with espin, the BB microvilli elongate to form long, spiky microvilli, without causing any other apparent morphological change in the cells [42]. These microvilli have been routinely used to study the biological properties/functions of the various espin isoforms and espin mutations causing hearing loss [55, 59, 62, 67, 68]. One such study showed that in espin expressing cells, the lengths of microvilli were dependent on the levels of espin expression [42]. Similar studies were performed on the deaf jerker mouse neuroepithelial cells of the organ of Corti cultures [59]. It was observed that in the absence of espin, stereocilia remain short while exogenous expression of espin results in lengthening of the stereocilia and microvilli. 36 In the context of studying hearing loss and genes involved in hearing loss, CL4 cells have become a valuable tool, especially in the absence of an animal model, in unraveling the role of important proteins that contribute to the structure of stereocilia, like actin and espin. In this thesis, the microvilli of CL4 cells have been used a model system to study the expression and functional properties of the y-actin mutations. 7 Zebrafish as a model to study human hearing loss Zebrafish has becOme an excellent model system to study human diseases [69]. Besides exhibiting rapid development and transparent embryos, many key factors that have made zebrafish a favorite model system include the ease of performing in-vivo imaging, transgenic knock-down, transgenic over- expression, small molecule screening, and chemical mutagenesis [69]. Thus human diseases like muscular dystrophies, myopathies, neurodegenerative diseases and cardiovascular disease have been vastly modeled in zebrafish [69- 71]. For all the above-mentioned features, zebrafish has been an attractive model to study the function of genes involved in human hearing loss. N-Ethyl-N- nitrosourea (ENU) mutagenesis, has been successfully used to identify ~30 genes required for ear development [71]. Using zebrafish, it was shown that myosin Vl is required for structural integrity of the apical surface of sensory hair 37 cells [72]. Zebrafish harboring mutations in many of these genes display ‘circler’ behavior, similar to the ‘Shaker’ mice. There are zebrafish models of many human syndromic deafness disorders as well [71]. In addition to designing a cell culture model for the y-actin mutations, we were successful in over-expressing the y-actins in zebrafish. The details about the transgenic model will be elaborated in Chapter 3 of this thesis. 38 REFERENCES 1. 10. Petit, C., J. Levilliers, and JP. Hardelin, Molecular genetics of hearing loss. Annual Review of Genetics, 2001. 35: p. 589-646. Vrijens, K., L. Van Laer, and G. Van Camp, Human hereditary hearing impairment: mouse models can help to solve the puzzle. Human Genetics, 2008. 124(4): p. 325-348. Forge, A. and T. Wright, The molecular architecture of the inner ear. British Medical Bulletin, 2002. 63: p. 5-24. Wangemann, P., K+ cycling and the endocochlear potential. Hearing Research, 2002. 165(1-2): p. 1-9. Vlastarakos, P.V., et al., Novel approaches to treating sensorineural hearing loss. Auditory genetics and necessary factors for stem cell transplant. Medical Science Monitor, 2008. 14(8): p. RA114-RA125. Couloigner, V., O. Sterkers, and E. Ferrary, What's new in ion transports in the cochlea? Pflugers Archiv-European Journal Of Physiology, 2006. 453(1): p. 11-22. Raphael, Y. and RA Altschuler, Structure and innervation of the cochlea. Brain Research Bulletin, 2003. 60(5-6): p. 397-422. Nickel, R. and A. Forge, Gap junctions and connexins in the inner ear: their roles in homeostasis and deafness. Current Opinion In Otolaryngology & Head And Neck Surgery, 2008. 16(5): p. 452-457. DeRosier, DJ. and LG. Tilney, F-actin bundles are derivatives of microvilli: What does this tell us about how bundles might form? Journal of Cell Biology, 2000. 148(1): p. 1-6. Tilney, L.G., M.S. Tilney, and DJ. Derosier, Actin-Filaments, Stereocilia, and Hair-Cells - How Cells Count and Measure. Annual Review of Cell Biology, 1992. 8: p. 257-274. 39 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. Belyantseva, l.A., et al., Stereocilia: the long and the short of it. Trends in Molecular Medicine, 2003. 9(11): p. 458-461. Dallos, P., Cochlear amplification, outer hair cells and prestin. Current Opinion In Neurobiology, 2008. 18(4): p. 370-376. Belyantseva, l.A., et al., gamma-Actin is required for cytoskeletal maintenance but not development. Proceedings Of The National Academy Of Sciences Of The United States Of America, 2009. 106(24): p. 9703- 9708. Bryant, J., R.J. Goodyear, and GP. Richardson, Sensory organ development in the inner ear: molecular and cellular mechanisms. British Medical Bulletin, 2002. 63: p. 39-57. Rzadzinska, A.K., et al., An actin molecular treadmill and myosins maintain stereocilia functional architecture and self-renewal. Journal of Cell Biology. 2004. 164(6): p. 887-897. F rolenkov, G.I., et al., Genetic insights into the morphogenesis of inner ear hair cells. Nature Reviews Genetics, 2004. 5(7): p. 489-498. Cotanche, D.A., Genetic and pharmacological intervention for treatmenflprevention of hearing loss. Journal Of Communication Disorders, 2008. 41(5): p. 421-443. Liu, X2. and D. Yan, Ageing and hearing loss. Journal Of Pathology, 2007. 211(2): p. 188-197. Brown, S.D.M., R.E. Hardisty-Hughes, and P. Mburu, Quiet as a mouse: dissecting the molecular and genetic basis of hearing. Nature Reviews Genetics, 2008. 9(4): p. 277-290. Petit, C., From deafness genes to hearing mechanisms: harmony and counterpoint. Trends In Molecular Medicine, 2006. 12(2): p. 57-64. van Wijk, E., et al., A mutation in the gamma actin 1 (ACTG 1) gene causes autosomal dominant hearing loss (DFNA20/26). Journal of Medical Genetics, 2003. 40(12): p. 879-884. 40 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. Zhu, M., et al., Mutations in the gamma-actin gene (ACTG 1) are associated with dominant progressive deafness (DFNA20/26). American Journal of Human Genetics, 2003. 73(5): p. 1082-1091. Eisen, MD. and D.K. Ryugo, Hearing molecules: contributions from genetic deafness. Cellular And Molecular Life Sciences, 2007. 64(5): p. 566-580. Strenzke, N., et al., Update on physiology and pathophysiology of the inner ear. Pathomechanisms of sensorineural hearing loss. Hno, 2008. 56(1): p. 27-36. Nance, W.E., The genetics of deafness. Mental Retardation and Developmental Disabilities Research Reviews, 2003. 9(2): p. 109-119. Nelson, BC. and R. Hinojosa, Presbycusis: A human temporal bone study of individuals with downward sloping audiometric patterns of hearing loss and review of the literature. Laryngoscope, 2006. 116(9): p. 1-12. Morin, M., et al., In vivo and in vitro effects of two novel gamma-actin (ACTG 1) mutations that cause DFNA20/26 hearing impairment. Human Molecular Genetics, 2009. 18(16): p. 3075-3089. de Heer, A.M.R., et al., Audiometric and Vestibular Features in a Second Dutch DFNA20/26 Family With a Novel Mutation in ACTG1. Annals Of Otology Rhinology And Laryngology, 2009. 118(5): p. 382-390. Liu, P., et al., Novel ACTG1 mutation causing autosomal dominant non- syndromic hearing impairment in a Chinese family. Journal Of Genetics And Genomics, 2008. 35(9): p. 553-558. Leibovici, M., S. Safieddine, and C. Petit, Mouse Models Of Human Hereditary Deafness, in Mouse Models Of Developmental Genetic Disease. 2008. p. 385-+. Adams, A.E.M., D. Botstein, and D.G. Drubin, A Yeast Actin-Binding Protein Is Encoded By Sac6, A Gene Found By Suppression Of An Actin Mutation. Science, 1989. 243(4888): p. 231-233. 41 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. Vartiainen, M.K., Nuclear actin dynamics - From form to function. Febs Letters, 2008. 582(14): p. 2033-2040. Khaitlina, S.Y., Functional specificity of actin isoforms, in International Review of Cytology - a Survey of Cell Biology, Vol. 202. 2001. p. 35-98. Hoock, T.C., P.M. Newcomb, and I.M. Herman, Beta Actin and Its Messenger-Rna Are Localized at the Plasma-Membrane and the Regions of Moving Cytoplasm During the Cellular-Response to Injury. Journal of Cell Biology, 1991. 112(4): p. 653-664. Otey, C.A., et al., Immunolocalization of the Gamma-lsoform of Nonmuscle Actin in Cultured-Cells. Journal of Cell Biology, 1986. 102(5): p. 1726-1737. Hill, MA and P. Gunning, Beta-Actin and Gamma-Actin Messenger-Rnas Are Differentially Located within Myoblasts. Journal of Cell Biology, 1993. 122(4): p. 825-832. Pollard, TD. and G.G. Borisy, Cellular motility driven by assembly and disassembly of actin filaments (_ vol 112, pg 453, 2002). Cell, 2003. 113(4): p. 549-549. Nicholson-Dykstra, 8., HM. Higgs, and ES. Harris, Actin dynamics: Growth from dendritic branches. Current Biology, 2005. 15(9): p. R346- R357. Hofer, D., W. Ness, and D. Drenckhahn, Sorting of actin isoforms in chicken auditory hair cells. Journal of Cell Science, 1997. 110: p. 765-770. Pantaloni, D., C. Le Clainche, and MP. Carlier, Mechanism of actin-based motility (vol 292, pg 1502, 2001). Science, 2001. 292(5524): p. 2012-2012. Tilney, LG. and DJ. DeRosier, How to make a curved Drosophila bristle using straight actin bundles. Proceedings Of The National Academy Of Sciences Of The United States Of America, 2005. 102(52): p. 18785- 18792. 42 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. Loomis, RA, et al., Espin cross-links cause the elongation of microvillus- type parallel actin bundles in vivo. Journal of Cell Biology, 2003. 163(5): p. 1 045-1055. Hall, A., Rho GTPases and the actin cytoskeleton. Science, 1998. 279(5350): p. 509-514. Jaffe, AB. and A. Hall, Rho GTPases: Biochemistry and biology. Annual Review Of Cell And Developmental Biology, 2005. 21: p. 247-269. Krause, M., et al., ENA/VASP proteins: Regulators of the actin cytoskeleton and cell migration. Annual Review Of Cell And Developmental Biology, 2003. 19: p. 541-564. Furness, D.N., et al., Differential distribution of beta- and gamma-actin in guinea-pig cochlear sensory and supporting cells. Hearing Research, 2005. 207(1-2): p. 22-34. Schneider, M.E., et al., Structural cell biology: Rapid renewal of auditory hair bundles - The recovery time after noise-induced hearing loss is in step with a molecular treadmill. Nature, 2002. 418(6900): p. 837-838. Manor, U. and B. Kachar, Dynamic length regulation of sensory stereocilia. Seminars In Cell 8 Developmental Biology, 2008. 19(6): p. 502-510. Rendtorff, MD, et al., A novel missense mutation in ACTG1 causes dominant deafness in a Norwegian DFNA20/26 family, but ACTG1 mutations are not frequent among families with hereditary hearing impairment. European Journal Of Human Genetics, 2006. 14(10): p. 1097- 1105. Bryan, K.E., et al., Efiects of human deafness gamma-actin mutations (DFNA20/26) on actin function. Journal Of Biological Chemistry, 2006. 281(29): p. 20129-20139. Bryan, KB and PA Rubenstein, Allele-Specific Effects of Human Deafness gamma-Actin Mutations (DFNA 20/26) on the Actin/Cofilin Interaction. Journal Of Biological Chemistry, 2009. 284(27): p. 18260- 1 8269. 43 52. 53. 55. 56. 57. 58. 59. 60. Bartles, J.R., A. Wierda, and LL. Zheng, Identification and characterization of espin, an actin-binding protein localized to the F-actin- rich junctional plaques of Sertoli cell ectoplasmic Specializations. Journal of Cell Science, 1996. 109: p. 1229-1239. Bartles, JR, et al., Small espin: A third actin-bundling protein and possible forked protein homolog in brush border microvilli. Molecular Biology Of The Cell, 1998. 9: p. 783. Sekerkova, G., et al., Novel espin actin-bundling proteins are localized to Purkinje cell dendritic spines and bind the Src homology 3 adapter protein insulin receptor substrate p53. Journal Of Neuroscience, 2003. 23(4): p. 1 31 0-1 31 9. Sekerkova, G., et al., Espins are multifunctional actin cytoskeletal regulatory proteins in the microvilli of chemosensory and mechanosensory cells. Journal of Neuroscience, 2004. 24(23): p. 5445-5456. Bartles, JR, et al., Espin is the missing cross/inker in the actin bundles of hair cell stereocilia. Molecular Biology of the Cell, 1998. 9: p. 135A—135A. Chen, B., et al., Espin contains an additional actin-binding site in its N terminus and is a major actin-bundling protein of the Sertoli cell-spermatid ectoplasmic specialization junctional plaque. Molecular Biology of the Cell, 1999. 10(12): p. 4327-4339. Loomis, RA, et al., Targeted wild-type and jerker espins reveal a novel, WH2-domain-dependent way to make actin bundles in cells. Journal Of Cell Science, 2006. 119(8): p. 1655-1665. Rzadzinska, A., et al., Balanced levels of espin are critical for stereociliary growth and length maintenance. Cell Motility and the Cytoskeleton, 2005. 62(3): p. 157-165. Zheng, L.L., et al., The deaf jerker mouse has a mutation in the gene encoding the espin actin—bundling proteins of hair cell stereocilia and lacks espins. Cell, 2000. 102(3): p. 377-385. 44 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. Naz, S., et al., Mutations of ESPN cause autosomal recessive deafness and vestibular dysfunction. Journal Of Medical Genetics, 2004. 41(8): p. 591-595. Donaudy, F., et al., Espin gene (ESPN) mutations associated with autosomal dominant hearing loss cause defects in microvillar elongation or organisation. Journal of Medical Genetics, 2006. 43(2): p. 157-161. Boulouiz, R., et al., A Novel Mutation in the Espin Gene Causes Autosomal Recessive Nonsyndromic Hearing Loss But No Apparent Vestibular Dysfunction in a Moroccan Family. American Journal Of Medical Genetics Part A, 2008. 146A(23): p. 3086-3089. Naz, S., et al., A mutation of ESPN causes autosomal recessive nonsyndromic hearing loss, DFNB36. American Journal Of Human Genetics, 2002. 71(4): p. 1995. Tyska, M.J. and MS. Mooseker, MYO1A (brush border myosin I) dynamics in the brush border of LLC-PK1-CL4 cells. Biophysical Journal, 2002. 82(4): p. 1869-1883. Bartles, J.R., Parallel actin bundles and their multiple actin-bundling proteins. Current Opinion In Cell Biology, 2000. 12(1): p. 72-78. Li, H.W., et al., Correlation of expression of the actin filament-bundling protein espin with stereociliary bundle formation in the developing inner ear. Journal of Comparative Neurology, 2004. 468(1): p. 125-134. Sekerkova, G., et al., Espins and the actin cytoskeleton of hair cell stereocilia and sensory cell microvilli. Cellular And Molecular Life Sciences, 2006. 63(19-20): p. 2329-2341. Ingham, P.W., The power of the zebrafish for disease analysis. Human Molecular Genetics, 2009. 18: p. R107-R112. Chico, T.J.A., P.W. Ingham, and DC. Crossman, Modeling cardiovascular disease in the zebrafish. Trends In Cardiovascular Medicine, 2008. 18(4): p. 150-155. 45 71. 72. Whitfield, T.T., Zebrafish as a model for hearing and deafness. Journal Of Neurobiology, 2002. 53(2): p. 157-171. Seller, 0., et al., Myosin VI is required for structural integrity of the apical surface of sensory hair cells in zebrafish. Developmental Biology, 2004. 272(2): p. 328-338. 46 CHAPTER 2 Mutations in cytoplasmic y-actin interfere with actin assembly dynamics 47 Abstract Mutations in cytoplasmic y-actin cause non-syndromic, post-lingual, autosomal- dominant, progressive sensorineural hearing loss. LLC-PK1- CL4 cells provide a model system to study the distribution of actins and the role of y-actin and its mutations in repair of damaged structures like microvilli. lmmunohistochemistry and confocal localization studies showed that B-actin was found primarily at the periphery of cells while y-actin is abundant in the perinuclear space and cytoplasm of the cell. Exogenous expression of mutant y-actins showed distribution to all the actin structures in the cell; the periphery, stress fibers and perinuclear space. In response to exogenous espin, filamentous mutant actins co-localized with espin in the microvilli. Co-transfection of espin and mutant actin resulted in each of the mutants co-localizing with filamentous actin in the microvilli. Cytochalasin D treatments of WT y-actin and mutant y-actins showed no difference in the repair of the damaged microvilli. Measurements of the lengths of microvilli however indicated that the microvilli expressing mutant actins were ~20-25% shorter than the WT y-actin microvilli. Quantitative FRAP assays did not reveal any differences in the recovery rates between WT and mutants actins. Our data suggest that mutations in y-actin exhibit subtle phenotypes and might interfere with basic actin assembly dynamics. 48 Introduction Actin is a ubiquitously expressed eukaryotic protein which plays a key role in muscle contraction, cell motility, cell shape determination, exocytosis, endocytosis, cytokinesis and organelle transport [1, 2]. The distribution and function of actin in the cell is primarily governed by the requirements of the cell and various actin-binding proteins. Actin- filament binding proteins control processes like nucleation, assembly, disassembly and cross-linking, while actin- monomer binding proteins like profilin, ADF/cofilin etc regulate the pool of monomeric actin in the cell. Three subtypes of actin ((1.13 and 7) have been described in humans, based on their mobility on 2D gels [1]. Six isoforms of actin have been identified based on their relative abundance in specific tissues. The two cytoplasmic isoactins B and y, encoded by ACTB and ACTG1 respectively, differ at only 4 amino acid positions in the N-terminus of their 375 amino acids. In most cells, B and y isoactins are found in the constant ratio of 2:1. However, cells such as human erythrocytes primarily contain B-actin while in brush border cells from intestinal epithelium, as well as hair cells of the inner ear, y-actin predominates. Studies from various laboratories have shown a spatial and temporal segregation of the cytoplasmic isoforms in different specialized cell types such as gastric parietal cells, hair cells, osteoblasts and neurons [1, 3, 4]. 49 The only mutations in cytoplasmic y-actin described to date are missense mutations that cause non-syndromic autosomal-dominant hearing loss. Our lab and others described the first mutations in y-actin to cause autosomal-dominant, progressive, post-lingual sensorineural hearing loss (SNHL) [5, 6], and subsequently more mutations have been identified [7, 8] . This finding, though not surprising considering the central role of y-actin in the inner ear, is nevertheless perplexing. Hearing loss is the only phenotype in individuals carrying mutations in y-actin, which is a highly conserved and ubiquitously expressed protein. We hypothesize that y-actin plays a very specific role in the innereah Stereocilia, which are key players in the process of hearing, are filamentous actin filled structures found on the apical surface of hair cells [9, 10]. In mammals, these actin filaments are bundled into paracrystalline structures by actin-bundling proteins fimbrin and espin [9]. Espin isoforms, encoded by a single gene, are actin-bundling proteins that are found in brush border microvilli, hair cell stereocilia and the actin rich dendritic spines of Purkinje cells [11, 12]. In chicken hair cells, espin expression is observed as early as E8. Espin continues to be expressed in matured stereocilia-suggesting a role of espin in maintaining the hair bundle structure [13]. Spontaneous mutations in espin cause shortening of stereocilia resulting in vestibular dysfunction and deafness-further indicating a pivotal role played by espin in stereociliary length maintenance [14]. 50 Stereocilia are derivatives of actin-filled microvilli found on the surface of hair cells [10]. The pig proximal kidney epithelial cell line, LLC-PK1-CL4 (CL4), upon cell-cell contact differentiates to form well-ordered brush border (BB) microvilli on its apical surface [15]. Owing to the highly ordered filamentous actin array of microvilli, CL4 cells are an excellent model system for studying the dynamics of BB microvilli and the cytoskeletal proteins. CL4 cells do not express espin endogenously, but make long spiky actin rich microvilli in response to exogenous espin. Of the six initially identified y-actin mutations (T89l, K118M, P264L, P332A, T278l, V370A), two of them (T89l, K118M) are located in predicted interaction sites for actin bundling proteins (Figure 2-1). Hence we chose CL4 cells as the model system. In addition, due to the structural similarities between stereocilia and microvilli, CL4 cells were an ideal model system to study the effect of mutations in y-actin. Here we show that exogenous expression of all the mutant actins resulted in their localization to cell body, stress fibers and cell periphery. Mutant actins co- localized with espin in the microvillar structures. However, expression of the mutant proteins results in the shorter microvilli. Fluorescence Recovery After Photobleaching (FRAP) of the microvilli expressing mutant y-actins showed that two of the six mutants have a higher proportion of mobile fractions. This finding suggests that y-actin mutants possibly interfere with basic actin filament processes. 51 Figure 2-1: Ribbon structure model representing the location of y- actin mutations. y—Actin consists of two domains (represented by two different ribbon colors), which are further divided into 4 sub-domains (1 ,2,3,4 in blue). The six initially identified mutations of y—actin are located throughout the molecule. T89| and K118M are located in sub-domain 1, in the predicted actin-bundling domain. Modified from [16]. 52 Materials and Methods Cell line: Pig kidney proximal tubule epithelial cell line LLC-PK1-CL4 (CL4) cells were a kind gift from James R. Bartles at Northwestern University, Chicago. The cells were maintained in DMEM, supplemented with 10% serum and 100mM L- glutamine. Plasmid Constructs: The coding sequence of human WT and the six mutant 7- actins were cloned into Xho1/BamH1 sites of pEGFP-Actin vector (Clontech Inc). pEGFP-CZ-espinZB construct was a kind gift from James R. Bartles at Northwestern University, Chicago. Thom Friedman at NIDCD kindly provided stRed-Monomer C1 vector. Espin, WT and mutant y-actins were cloned into the Xho1/BamH1 sites of stRed Monomer C1 vector. WT and mutants actins were under a CMV promoter in all the plasmid constructs. lmmgnohistochemistrv: CL4 cells were plated on coverslips on day 0. On day 1, cells were fixed in cold 100% methanol at -20°C for 15 min followed by 3 washes in 1XPBS. Cells were blocked in 5% goat serum + 2% BSA at RT for 20 min. Cells were then incubated at RT for 2 hrs in 1:500 and 1:100 dilutions of y-actin and -actin antibody respectively (y antibody developed by our lab, [3 antibody from AbCam). Cells were washed with 1XPBS, 3 times for 5 min each. Cells were then incubated in 1:200 dilutions of FITC and cy3 tagged secondary antibodies, with shaking in dark, at RT for 30 min. Following a wash in 1XPBS for 10 min, coverslips were air dried and mounted. Transient Transfections and_Stable Line ngergtiron: ~8X106 cells were plated on a cover slip on clay 0.1ug DNA each of GFP and DsRed tagged plasmids were 53 transfected on day 2 using fugene transfection reagent (lnvitrogen). 24 hours post transfection, cells were fixed in 4% paraformaldehyde (PFA) for 20 min at RT. Cells were then washed in 1XPBS and permeabilized with 0.1%Triton-X100 for 5 min. Cells were blocked in 1% BSA for 30 min at RT, stained with DAPI (1:10,000 dilution) for 5 min at RT, and washed in ddH20 for 10 min. The cover slip was air dried and mounted on a slide using fluoroguard anti-fade reagent (Bio-Rad). For stable cell line generation, cells were similarly plated and transfected. Following day of transfection, the media was replaced with media containing 100mg/ml G418 selection drug. The media was changed everyday, replacing with fresh media containing G418. After ~1.5 weeks, sterile cloning cylinders were used to pick the clones that were resistant to G418. These clones were transferred into a 12 well plate and grown in G418 medium from that point onwards. Qonfoca_l MicroscogL: Cells were viewed under Olympus Fluoview 1500 confocal microscope. Images were taken under 60X PlanApo objective and 3.5 digital magnification. The lengths of the microvilli were measured using the Olympus 1500 software. FRAP assays: Cells were plated on glass-bottom dishes (MatTek Inc) on day 0. On day 1, cells were transfected with 1ug each of EGFP-actin (WT/mutant) and DsRed-espin. FRAP was performed on the microvilli of the transfected cells on day 2 using Olympus 1500 confocal microscope. Cells were placed in a 37°C chamber, with phenol red free 10% DMEM media containing 25 mM HEPES to maintain pH. The cell was scanned 10 times (at 1% 488 argon laser intensity), 54 and the 10th scan was considered as pre-bleach. Region of interest (ROI) of microvilli were bleached at 70% laser intensity for 3 seconds. ROls showing 55- 70% bleach were used for further analysis. Fluorescence recovery scans were performed 30 seconds apart, with the first scan immediately following the bleach period, over a period of 4.5 minutes, at 1% laser intensity. Metamorph software (Molecular Devices) was used to calculate background intensity for each cell over each time point. This software was used to calculate the intensity of an adjacent fluorescent cell at each time point, to confirm that the cell undergoing FRAP was not bleaching as a consequence of scanning. A stringent method was developed where cells that bleached only 55-70% were considered. Mobile fractions were then calculated using Metamorph and KaleidaGraph softwares [32]. Mstical Analysis: Students t-test, using one-tailed distribution and two-sample equal variance, was used to determine the statistical significance between the microvilli lengths of WT and mutant y-actins. Similar conditions were used to determine statistical significance for FRAP assays. Data from 3 independent experiments were combined together to obtain the final n for the microvilli length experiments. Due to the large number (~192) of n in individual experiments, it was possible to combine the 3 experiments. 55 Results Endogenous actin distrMon The distribution and localization of endogenoUs cytoplasmic actin isoforms (B,y) in CL4 cells is unknown. We investigated the isoform specific distribution of endogenous actin before over-expressing exogenous actin. Antibodies specific to B-and y-actin isoform were used to perform immunocytochemistry. As shown in Figure 2-2A&B, in spreading cells, B-actin is primarily observed in the stress fibers, and along the periphery. In contrast, y-actin (Figure 2-2B&C) is enriched in the cell body (perinuclear space) and the stress fibers. Our data suggest a possible differential role of [3 and y-actins in CL4 cells similar to that described in other cell lines and tissues [1, 17, 18]. In confluent cells, I do not observe any difference in the distribution and localization of B and y-actins in the cell periphery and perinuclear space, but there is more y-actin in the cell body like in spreading cells (Figure 2-2D-F). 56 Figure 2-2: Spatial distribution of endogenous cytoplasmic actins in LLC- PK1-CL4 cells. B-Actin specific antibody (AbCam) and y—actin specific primary antibodies were used to stain for the actins. Cells were fixed in 100% methanol at —200 and imaged under Olympus 1500 confocal microscope, 60X PlanApo objective. 2-2A,D: B-Actin secondary antibody is FITC tagged, while y—actin secondary antibody is rhodamine tagged (2-2B,E). In spreading cells, B-Actin is found along the periphery and in stress fibers, y—actin is primarily found in the perinuclear space and stress fibers (2-2C). In confluent cells, there is no observable difference in the actin isoform distributions along cell periphery. but more y—actin is observed in the cell body compared to B-actin (2-2D-F). Cells here are counter stained with DAPI for the nucleus. Bar: 10pm 57 _E_)Qgenous expression of wild-type (WT) and mutant actins Human WT [3 and y-actin and all the original six mutants described for y- actin were cloned into N-terminal tagged pEGFP vector and individually expressed in CL4 cells, under the control of the CMV promoter. Over expression of cytoplasmic actins has been reported to cause gross morphological changes in many cell types [17]. This observation is not surprising as actin plays a critical role in many physiologically important processes like motility, intracellular and organelle trafficking. Exogenous over expression could disrupt these critical processes. Hence it was important to determine if over expression of WT actins in CL4 cells would be detrimental to the cells. As seen in Figure 2-3A and 2-3B, exogenous expression of either B-actin or y-actin did not affect the gross cell morphology. Both the proteins were seen in stress fibers, along the cell periphery and in all the expected actin rich structures. Cells expressing these actins continued to look healthy at least 72 hrs post transfection (data not shown). To determine the effect of hearing loss mutations on the distribution of y- actin, the Six y-actin mutants were individually expressed in the CL4 cells. All the six mutants (Figure 2-4B-G) localized to all actin rich structures in the cell and were indistinguishable from the WT control (Figure 2-4A). 58 Figure 2-3: Exogenous-expression of cytoplasmic actins in LLC-PK1-CL4 cells. N-terminal EGFP tagged [3 and y—actins were over-expressed using a CMV promoter in the CL4 cells. 24hrs post transfection, the cells were fixed in 4% paraformaldehyde (PFA), counter stained with DAPI for nucleus and imaged under Olympus 1500 confocal microscope, 60X PlanApo objective. 2-3A: EGFP- B—actin localized to the periphery and stress fibers. 2-38: EGFPJy-actin also localized to the periphery and stress fibers. No morphological changes were observed in the cells over-expressing cytoplasmic actins. Bar: 20pm 59 Figure 2-4: Over-expression of y—actin mutants in LLC-PK1-CL4 cells. N- terminal EGFP- tagged y—actin mutations were over-expressed, under the CMV promoter, to determine their distribution and localization in CL4 cells. 24 hrs post transfection, the cells were fixed in 4% PFA and counter stained with DAPI (blue) to label DNA. They were imaged under Olympus 1500 confocal microscope, 60X PlanApo objective. All the six mutants (2-4B-G) localized to actin-rich stress fibers, were found along the periphery of the cell and in the perinuclear space, and were indistinguishable from the WT control (2-4A). 2-4A: EGFP-WT y—actin, 2-43: EGFP-T89l, 2-4C: EGFP-K118M, 2-40: EGFP-P264L, 2-4E: EGFP- P332A, 2-4F: EGFP-T278I, 2-4G: EGFP-V37OA 60 Expression and localization of espin, WT, and mutant actins The stereociliary bundles found on the surface of hair cells contain parallel bundles of filamentous actin. In the mammalian stereocilia, actin-bundling proteins like espin and fimbrin maintain these paracrystalline bundles [10]. Considering the pivotal role of espin in actin- bundling and in hearing, we set out to determine if y-actin mutants were defective in localizing/binding to espin, thereby possibly resulting in weaker filaments. To verify the CL4/espin response, we demonstrated that expression of espin results in the lengthening of brush border microvilli in these epithelial cells [19] (Figure 2-5). Hence, CL4 cells are an excellent system to study microvillar localization properties of WT and y-actin mutants. Co- transfection of DsRed-espin and GFP-WT y-actin resulted in co- Iocalization of actin and espin in the microvilli with concomitant lengthening (Figure 2-6A). Co-transfection of DsRed-espin and each of the mutant y-actins also resulted in the mutant actins co-localizing with espin in the microvilli (Figure 2-68-G). 61 Figure 2-5: Exogenous expression of espin results in lengthening of brush border microvilli actins in LLC-PK1-CL4 cells. N-terminal EGFP-tagged espin plasmid was transfected into ~90 % confluent cells. 24 hrs post transfection, the cells were fixed in 4% PFA and imaged under Olympus 1500 confocal microscope, 60X PlanApo objective. 2-5A: Over-expression results in the formation of long spiky microvilli. 2-53: The cells were counter stained with rhodamine phalloidin to label filamentous actin. 2-5C: A merge of the images confirms co-Iocalization of espin and actin. Bar: 10pm 62 Figure 2-6: Localization of WT or mutant y-actins with espin in LLC-PK1- CL4 cells. Confluent cells were transfected with N-terminal EGFP-tagged espin and either DsRed- tagged WT or each of the six mutant y-actins. 24 hrs post transfection, cells were fixed in 4% PFA and imaged under Olympus 1500 confocal microscope, 60X PlanApo objective. 2-6A: WT y-actin co-localizes with espin in the microvilli of CL4 cells. 2-68-56: All the six mutants also co-Iocalize with espin in the microvilli. Compared to WT-y-actin, mutant actins do not show any difference in the distribution or co-localization with espin. Bar: 10pm 63 DsRed- T89l DsRed- K118M 64 Expression obi-actin mutants results in shorter microvilli The lengths of espin induced microvilli are proportional to the level of espin expression in the CL4 cells [19]. To determine the effects of actin mutations on microvilli length, a cell line constitutively expressing EGFP-espin was established. This stable cell line was then transfected with DsRed-actin mutations expression constructs. On careful analysis of the microvilli, we observed that the cells expressing y-actin mutations had shorter microvilli compared to the cells expressing WT y-actin (Figure 2-7A). Briefly, 12 transfected cells for each mutation were imaged and 16 microvilli per cell were measured (n=~195). Each experiment was repeated at least 3 times (T278I was done twice). To further confirm that the microvilli lengths were shorter due to mutant actin over expression, microvilli of neighboring untransfected cells in each case were measured and compared to the microvilli lengths of neighboring untransfected cells of WT y-actin transfection. As expected, the lengths of microvilli of the mutant untransfected cells were comparable to that of WT untransfected cells (Figure 2-7B) and longer than their mutant actin transfected neighbors. Also, the microvilli in cells expressing WT y—actin and those of neighboring untransfected cells were n_ot significantly different, thereby demonstrating that the expression of exogenous actin itself does not cause microvilli shortening (Figure 2-7C). 65 Figure 2-7: Over-expression of mutant y-actin results in shorter microvilli. DsRed-WT or DsRed-mutant y-actins were individually over-expressed in EGFP- espin stable cell line. The cells were fixed in 4% PFA and imaged under Olympus 1500 confocal microscope, 60X PlanApo objective. In each case, 12 cells expressing the actin construct were imaged. 16 microvilli per cell were measured. The mean lengths of WT and mutant actins were calculated and students t-test was performed to determine if the difference in the lengths between mutants and WT was significant. Standard deviations are shown for each data set. 2-7A: n=3 (12 cells x 16 microvilli x _3 = 576 microvilli). The means from 3 independent experiments were pooled and the graph was plotted. The microvilli expressing mutant actins were ~20-25% shorter than their WT control. 2-7B: The maximum intensity projection image (of a Z series) of DsRed- WT-y-actin expressing cell in EGFP-espin stable cell line. 2-7C: The maximum intensity projection (of a Z series) of one of the mutants (Dsred-T89l-y-actin) expressing cell in EGFP-espin stable cell line. Microvilli of cells expressing the mutant actin are shorter than the microvilli of cells expressing WT -y-actin. 2-7D: To further confirm this data, untransfected cells in each set were imaged and their microvilli were measured. n=1,WT-transfected: mean of 3 experiments. The lengths of microvilli of cells not expressing the mutant actins were no different than the lengths of microvilli of cells not expressing the WT-actin. In addition, the lengths of microvilli expressing WT-actin were same as the lengths of microvilli of cells not expressing the mutant actins further confirming that over-expression of WT actin does not result in shorter microvilli. 66 2-7A Microvilli lengths 5 , * * * * * * .2 1 I .E. 3. ‘ II 4: S 2 . C 2 I I: 1 I E o l . . WT T89l K118M P264L P332A T278| V370A * p< 0 .0001 '.° N U Microvilli lengths i tutti K118M T89I T278I V370A O I & U Mean lengths(microns) N transfected 67 Cytochalasin D treatments of y-actin expressing microvilli Persons harboring y-actin mutations develop hearing loss in the 2nd and 3rd decade [5, 6]. Recent data from a y-actin knock-out mouse model indicates that y-actin is not involved in development, but might play a critical role in the maintenance and /or repair of actin structures [20]. To test whether mutant y- actins are impaired in repair, low concentrations of the actin depolymerizing agent cytochalasin D (CD) was used to simulate damage and subsequent repair. Two mutations, T89I and K118M were chosen for this experiment since they lie in the predicted actin-bundling interaction domain of actin. Thus any defects, if observed in the repair process, could be attributed to poor binding of the mutants to espin. Briefly, CL4 cells stably expressing espin and transiently expressing EGFP tagged actins (WT or T89l or K118M) were treated with 100 nM of CD for 13 hrs. Low concentrations of CD result in shortening of actin filaments [19] but the affect is reversible. Following cytochalasin D treatment, the cells were allowed to recover for 4 1/2 hrs. The lengths of the microvilli of untreated and recovered microvilli were measured using the Olympus 1500 confocal microscope. I did not observe any difference in the lengths of untreated and recovered microvilli in the mutants (Figure 2-8). The recovered microvilli of cells expressing the mutants were not any shorter than the untreated microvilli. In addition, the recovered microvilli of the mutants remained shorter than the recovered microvilli 68 Microvilli lengths of untreated and recovered cells 17, 6 = I .5 5. E ‘ =5; 4 ‘ * * luntrsated I a 3 I urecovered 5 c 2 I 8 z 1 +r _ i T89l K118M * p<.0001 * p<_ooo1 Figure 2-8: Microvilli lengths In cells before and after cytochalasin D treatment. EGFP-espin stable cell line was transfected with either DsRed-WTy- actin or T89l or K118My—actin mutants. Following transfections, cells were treated with 100 nM cytochalasin D for 13hrs. Cytochalasin D was removed from the media and cells were allowed to recover for 4.5 hrs. Microvillar lengths were measured before cytochalasin D treatment and after 4.5 hrs of recovery. In WT and mutant cells, there was no difference in the microvillar lengths between the untreated and recovered cells. The mutant microvilli were as long as their untreated counter parts. Like the untreated microvilli (Figure 2-6A), mutant recovered microvilli were also Significantly shorter than WT recovered microvilli. 69 of WT y-actin (like the untreated microvilli lengths). Since no difference in the recovery rates were observed in two mutants, this assay was not performed on the other 4 mutants. FRA_P assays to determine protein mgtailiw It is evident from various experiments that the physiological role of actin in the cell is dependent on numerous actin-binding proteins [2, 21]. Actin treadmilling has been shown to occur in microvilli and young stereocilia [19, 22, 23]. This phenomenon is dependent on the polymerization rates of monomeric actin and depolymerization of filamentous filaments. Monomeric actin-binding protein profilin and filament severing protein cofilin are pivotal for filament polymerization and depolymerization rates [21, 24]. Espin and myosin XVa, actin- binding proteins, are critical for lengthening of stereociliary bundles [25, 26]. Recent biochemical experiments performed on purified mutant y-actins showed that the mutant actins exhibited various degrees of sensitivities to cofilin binding [27]. Based on our data and what is reported in literature, we hypothesized that the cause of shorter microvilli could be due to impaired actin protein trafficking or filament incorporation. To this end, I performed fluorescence recovery after photobleaching assays on the microvilli of cells expressing mutant actins. Briefly, cells are scanned 10 times, before an area of microvilli is bleached. The 10th 70 scan is considered the prebleach scan. Microvilli are bleached for ~3 seconds at 70% laser intensity. Recovery scans were performed every 30 seconds for ~4.5 minutes. Figure 2-9A Shows an entire series of WT y-actin cell undergoing FRAP. Using Metamorph imaging software (Molecular Devices) and KaleidaGraph software (Synergy Software), the mobile fraction for WT and each of the mutants was calculated. For each scan, a background value (region not expressing EGFP) and a value from an adjacent cell expressing EGFP is calculated. These values are essential for calculating the mobile fraction. Recovery graphs of WT y-actin and the Six mutants were plotted. Mobile fraction is an indicator of the availability of the protein being studied. No significant difference between the mobile fractions of WT and mutant actins was observed (Figure 2-9C-9H). Figures 2-10A-F are representative images of pre- bleach, bleach and recovered cells (after 4.5’) of each of the mutants following FRAP. 71 Figure 2-9A-B: Fluorescence recovery after photobleaching of EGFP-WT y- actin. 2-9A:The cell was scanned 10 times, without any interval between the scans, to obtain the prebleach image (image following the 10th scan). A region of the microvilli is then bleached for ~3 seconds at 70% laser intensity. This is the bleach image. Following the bleach, the entire cell is scanned every 30 seconds to obtain the recovery images. The recovery was performed for ~4.5 minutes. 2- 9B: A representative graph is shown, where recovery of the region of interest (ROI), adjoining cell (CF) and background (BG) fluorescence are plotted. 72 €85 snow 5: Tapas—p.65 ooom suun'u pezueuuou m2-~ c m. w .u n m. n r 8' 363”. o c _..N 88.2.: R" new as“ 8. a... R. 8 8 R n e F p F b b _ — io ION w m uiElll .3 n East a m. w U .1... m 8. mean. b66061 3:8 ".2233. 2:9". 79 68:5: nhN 9N MFN n2. n9. nNr mm mm mm m o _ _ P . o . ION u .1 m .2 _, u>ouom ”3 TN .2 2 .8325 Ru 9a 2... n2 8. R. 8 8 8 n e i. Il_ ,iri I _I Ii ,1 i F _i LII m C a m ill .3 m 55+ 2 a. BEEREIOI HIM H r8 W .n , “- HIIHIH H a 4 es. 922 E9006”. ma ...N 3:00 "5 Fr