Molecular Motor Forces and Nuclear Positioning in Living Cells

MISSING IMAGE

Material Information

Title:
Molecular Motor Forces and Nuclear Positioning in Living Cells
Physical Description:
1 online resource (179 p.)
Language:
english
Creator:
Wu, Jun
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Chemical Engineering
Committee Chair:
Lele, Tanmay
Committee Co-Chair:
Dickinson, Richard B
Committee Members:
Tseng, Yiider
Khargonekar, Pramod

Subjects

Subjects / Keywords:
actomyosin -- centrosome -- dynein -- force -- frap -- microtubule -- nuclear -- positioning -- rotation
Chemical Engineering -- Dissertations, Academic -- UF
Genre:
Chemical Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
The nucleus is the largest and heaviest organelle in a cell and its proper positioning is crucial for basic cell functions, such as cell migration, division, fertilization and establishment of polarity. Irregularities in nuclear movement (or nuclear positioning) are associated with various serious diseases. Nuclear movement in the cell is a complex process that involves interactions with all three cytoskeletal systems - actin, intermediate filaments and microtubules. The interactions occur through nuclear-embedded molecular tethers that link to the cytoskeleton. The molecular linkage between the nucleus and the cytoskeleton can be also established through molecular motors, proteins that convert chemical energy to mechanical forces. How molecular motors drive nuclear movement remains poorly understood. Nuclear rotation and nuclear translation are the two types of nuclear motions commonly observed in the cell. We investigated the physical mechanism for nuclear rotation in the cell. We found that the nuclear rotation angle is directionally persistent on a time scale of tens of minutes, but rotationally diffusive on longer time scales, and rotation required the activity of the microtubule motor dynein. Based on these results, a mechanical model for torque generation on the nucleus was proposed. To investigate nuclear translation, we designed experiments utilizing two different techniques- protein photo activation and cell micromanipulation. The results from these experiments point to a tug of war between forward pulling and rearward pulling forces on the nuclear surface generated by actomyosin contraction. Net nuclear motion occurs when the forward pulling force increases during lamellipodial formation. We also investigated the mechanisms of positioning of the centrosome, which is in physical proximity with the nucleus. The positioning of centrosome is very important in cell migration and cell division. Whether the centrosome is positioned by pulling forces or pushing forces originating in dynamic microtubules remains a controversy. By severing a single microtubule with femtosecond laser ablation, we found that microtubules are under tension generated by dynein. We also show that dynein-mediated pulling forces are sufficient to center the centrosome.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Jun Wu.
Thesis:
Thesis (Ph.D.)--University of Florida, 2012.
Local:
Adviser: Lele, Tanmay.
Local:
Co-adviser: Dickinson, Richard B.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2014-05-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2012
System ID:
UFE0044035:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 MOLECULAR MOTOR FORCES AND NUCLEAR POSITIONING IN LIVING CELL S By JUN WU A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Jun Wu

PAGE 3

3 To my Parents

PAGE 4

4 ACKNOWLEDGMENTS I would like to acknowledge all the people who support ed me during this very important time of my life. First I would like to acknowledge the guidance and support of my advisor, Dr. Tanmay Lele. He brought me into a new world of science. He not only taught me a lot of new things, but also taught me how to learn and how to think. He is always supportive and extremely patient. His expertise helped me to select problems and complete projects, and grow to be an independent researcher Without him, I d like to thank my other committee members. Dr. Richard Dickinson is such a great teacher. I learned so much from him in and out of the class. We also had very good collaborations in most of my publications. I always enjoyed talking to Dr. Yiider Tseng. To me, he is a teacher, but also a friend. We had a lot of wonderful discussions about both research and life. Professor Pramod Khargonekar has been supportive and offered helpful advice I would also like to thank Professor Kyle Roux for the valuable discus sion and support on the nuclear movement project. I want to thank Professor Daniel Purich for holding the Cytoskeleton Journal Club through which I gained so much beyond my own research by interacting with other scientists Professor Scott Grieshaber kindl y provided some plasmids, which are of great help for my research. I would also thank all my labmates and other students who were always helpful throughout my time in University of Florida. Jiyeon Lee taught me the basic biology lab skills at the beginning of my PhD study. I also learned from her the carefulness and lot. Discussions with him always gave me good idea to improve my research, and we had an enjoyable time work ing together on the microtubule project. He is also a great

PAGE 5

5 tennis partner. T.J. Chancellor is a good friend, and he was always helpful with the experiments and the life. David Lovett was supportive at the end of my time in the lab. Shen Hsiu Hung in Dr. Y also would like to thank the master students and undergraduates who worked with me over the years, most spec ifically Nandini Shekhar, Agnes Mendonca and Kristen Lee. I want to specifically thank my roommate and best friend here, Robert Colmyer. He was and enjoy my life here s o much, and his family is full of nice people. Finally, I would like to thank all my family members who supported me throughout so many years and made my dream come true. My parents have always been encouraging, loving and have motivated me to achieve my g oals. They gave a healthy body, smart brain and most importantly, they taught me how to be a better person. My uncle Daiyi Wu always gave me good advice and helped me come here to pursue my PhD

PAGE 6

6 TABLE OF CONTENTS page ACKNO WLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ .......... 10 LIST OF FIGURES ................................ ................................ ................................ ........ 11 LIST OF ABBREVIA TIONS ................................ ................................ ........................... 14 ABSTRACT ................................ ................................ ................................ ................... 16 C H A P T E R 1 INTRODUCTION ................................ ................................ ................................ .... 18 Nuclear Positioning is an Important Cell Function ................................ .................. 18 Dynein Based Torque Generation on the Nucleus ................................ ........... 19 Actomyosin Forces on the Nucleus ................................ ................................ .. 22 Dynein Force Generation on Microtubules ................................ .............................. 24 Fluorescence Recovery after Photobleaching (FRAP) ................................ ........... 27 2 DYNEIN MOTOR FORCES ON THE NUCLEUS ................................ ................... 31 Materials and Methods ................................ ................................ ............................ 31 Cell Culture, Plasmids and Transfection ................................ .......................... 31 Time Lapse Imaging and Analysis ................................ ................................ ... 32 Immunofluorescence ................................ ................................ ........................ 33 Cell Shape Patterning ................................ ................................ ...................... 33 Results ................................ ................................ ................................ .................... 34 Nuclear Rotation is a Persistent Random Walk ................................ ................ 34 Cytoplasmic Dynein is Required for Nuclear Rotation in Stationary Fibroblasts ................................ ................................ ................................ ..... 35 Model Formulation ................................ ................................ ............................ 35 Model Results ................................ ................................ ................................ ... 40 Nuclear Rotation Depends on Nuclear Centrosomal Distance ......................... 41 Discussion ................................ ................................ ................................ .............. 42 Summary of Findings ................................ ................................ .............................. 46 3 THE NUCLEUS IS IN A TUG OF WAR BETWEEN A CTOMYOSIN PULLING FORCES IN A CRAWLING CELL ................................ ................................ ........... 57 Materials and Methods ................................ ................................ ............................ 57 Cell Culture, Plasmids and T ransfection, Drug Treatment ............................... 57 Time Lapse I maging an d A nalysis ................................ ................................ ... 5 8 Confocal Microscopy and Protein Photo Activation ................................ .......... 58

PAGE 7

7 Image Analysis ................................ ................................ ................................ 59 Micromanipulation by Microinjector ................................ ................................ .. 60 Results ................................ ................................ ................................ .................... 60 Myosin Inhibition In the Trailing Edge Causes Nuclear Motion Toward the Leading Edge without Change in Cell Shape. ................................ ............... 60 Solid Like Coupling Between the Nucleus and the Trailing Edge. .................... 61 Forward Nuclear Movement does not Require Trailing Edge Detachment ....... 62 Forward Motion of the Nucleus Occurs due to Actomyosin Contraction Between the Leading Edge and the Nucleus ................................ ................ 62 Discussion ................................ ................................ ................................ .............. 64 Summary of Findings ................................ ................................ .............................. 65 4 EFFECTS OF DYNEIN MOTOR ON MICROTUBULE MECHANICS AND CENTROSOME CENTERING ................................ ................................ ................ 75 Materials and Methods ................................ ................................ ............................ 76 Cell Culture, Plasmids and Transfection ................................ .......................... 76 Time Lapse Imaging and Analysis ................................ ................................ ... 76 Cell Shape Patterning ................................ ................................ ...................... 77 Laser Ablation ................................ ................................ ................................ .. 77 Results ................................ ................................ ................................ .................... 81 Dynamics of Severed Micr otubules ................................ ................................ .. 81 Model for Dynein Generated Microtubule Forces ................................ ............. 83 Simulations of Microtubule Buckling Dynamics ................................ ................ 85 Centrosome Centering by Motor Driven Microtubules ................................ ...... 86 Discussion ................................ ................................ ................................ .............. 89 Summary of Findings ................................ ................................ .............................. 92 5 MODELING OF FLUORESCENCE RECOVERY AFTER PHOTOBLEACHING .. 106 Materials and Methods ................................ ................................ .......................... 106 Cell Culture, Plasmids and T ransfection ................................ ........................ 106 Confocal Microscopy and FRAP ................................ ................................ ..... 106 Results ................................ ................................ ................................ .................. 107 Modeling Bleaching D uring Image Capture. ................................ ................... 107 FRAP Model to Account for Photobleaching due to Image Capture ............... 109 Accounting for Bleaching of Free Protein ................................ ....................... 112 FRAP Model to Account for an Immobile Fraction ................................ .......... 113 Calculations of Normalized Recovery: the Behavior of Equation 5 6A and 6B ................................ ................................ ................................ ................ 114 Effect of the Immobile Fraction on FRAP Recovery Without Bleaching of Free Protein) ................................ ................................ ............................... 115 Analysis of Focal Adhesion Protein Exchange ................................ ............... 116 Discussion ................................ ................................ ................................ ............ 117 Summary of the Model ................................ ................................ .......................... 121 6 CONCLUSIONS ................................ ................................ ................................ ... 128

PAGE 8

8 Summary of Findings ................................ ................................ ............................ 128 Nuclear Rotation in Living Cells ................................ ................................ ...... 128 Nuclear Translation in Living Cells ................................ ................................ 128 Effects of Dynein on Microtubule Mechanics and Centrosome Centering ...... 129 FRAP Model Accounting for Immobile Molecules and Bleaching due to Im age Capture ................................ ................................ ............................ 129 Future Work ................................ ................................ ................................ .......... 130 Further Investigatio n on Nuclear Movement ................................ ................... 130 Microtubule motor kinesin in nuclear rotation ................................ ........... 130 Actomyosin contraction and nuclear positioning ................................ ...... 131 The role of intermediate filament in nuclear movement ........................... 132 Nuclear Positioning under the Influence of Extracellular Forces .................... 133 Cell Migration During Wound Healing ................................ ............................ 133 Effects of Other Motor Forces on Microtubule Mechanics .............................. 134 APPENDIX A MATLAB CODE FOR NUCLEAR ROTATION ANALYSIS ................................ .... 140 B MATLAB CODE FOR NUCLEAR TRANSLATION ANALYSIS ............................. 143 C MATLAB CODE FOR CENTROSOME POSITION ANALYSIS ............................ 145 D MODEL FOR MICROTUBULE MECHANIC S AND CENTROSOME CENTERING ................................ ................................ ................................ ......... 148 Microtubule Mechanics ................................ ................................ ......................... 148 Model for Dynein Forces ................................ ................................ ....................... 149 Simulati on Methods. ................................ ................................ ............................. 150 Simulations of Microtubule Buckling ................................ ................................ ..... 152 Simulations of Centrosome Centering ................................ ................................ .. 152 Centrosome Relaxation Time ................................ ................................ ............... 154 E for BLEACHING OF FREE PROTEIN IN THE CYTOPLASM ........................... 156 F MATLAB CODE FOR FRAP MODELING AND SIMULATION .............................. 157 Simulation of Bleaching Dynamics in Image Capture Process ............................. 157 Comparison of Observed Dynamics and Actu al Dynamics ............................ 157 Modeling of Normalized Bleaching Dynamics ................................ ................ 158 Simulation of FRAP ................................ ................................ .............................. 159 Model Fitting of FRAP ................................ ................................ ........................... 159 Define Fitting Function of FRAP Model ................................ .......................... 159 Fitting Focal Adhesion FRAP Experiment Data to the Model ......................... 160 Define Fitting Function for Free Molecule Intensity ................................ ........ 161 Fitting Free Molecule Intensity in Focal Adhesion FRAP Experiments ........... 161 Simulation of FRAP with Multiple Time Intervals ................................ .................. 162

PAGE 9

9 LIST OF REFERENCES ................................ ................................ ............................. 164 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 179

PAGE 10

10 LIST OF TABLES Table page 2 1 Nuclear Rotation Model Parameters ................................ ................................ ... 56 4 1 Microtubule Mechanics and Centrosome Centering Model Parameters ........... 105

PAGE 11

11 LIST OF FIGURES Figure page 1 1 Cartoon of dynein structure and walking on the microtubule. ............................. 30 2 1 Nuclear rotation is a biased random walk in NIH 3T3 fibroblasts. ....................... 47 2 2 Dynein inhibition significantly reduces nuclear rotation. ................................ ..... 48 2 3 Microtubules drive nuclear rotation. ................................ ................................ .... 49 2 4 The centrosome does not rotate with the nucleus. ................................ ............. 49 2 5 Schematic of the nuclear rotation model. ................................ ........................... 50 2 6 Simulations of nuclear rotation in a circular cell. ................................ ................. 51 2 7 Distance between the centrosome and the nucleus decreased in patterned square NIH 3T3 cells. ................................ ................................ ......................... 53 2 8 Th e centrosome is underneath the nucleus ................................ ....................... 54 2 9 Nuclear rotation in un patterned cells is significantly larger than that in patterned cells ................................ ................................ ................................ .... 54 2 10 Nesprin 1 knock down affects nuclear rota tion. ................................ .................. 55 3 1 Schematic of how collective trajectories of nuclear translation were generated. ................................ ................................ ................................ .......... 66 3 2 Nuclear movement upon local introduction of blebbistatin. 67 3 3 Micromanipulation reveals that the nucleus is under tension between the leading edge and trailing edge. ................................ ................................ ........... 68 3 4 Nuclear movement in motile fibroblasts does not required detachment of the trailing edge. ................................ ................................ ................................ ....... 69 3 5 Directional nuclear translation upon Rac 1 photo activation. .............................. 70 3 6 Kymograph of photo activation experiments. ................................ ...................... 72 3 7 Nuclear retraction due to lamellipodia release. ................................ ................... 73 3 8 Effect of KASH on trailing edge detachment. ................................ ..................... 74 4 1 A comparison of centrosome centering (simulation) with the autocorrelation function of fluctuations in centrosome position (experiment). ............................. 94

PAGE 12

12 4 2 Deploymerization rates of microtubules severed by laser ablation. .................... 95 4 3 Minus end microtubules underneath the nucleus increase in bending after laser severing. ................................ ................................ ................................ .... 96 4 4 Minus end microtubules at the cell periphery increase in bending after laser severing. ................................ ................................ ................................ ............. 98 4 5 Chan ge in curvature after severing is not correlated with the initial curvature, nor with the spatial location of the cut. ................................ ................................ 98 4 6 Inhibi tion of dynein causes dispersion of the Golgi complex. ............................. 99 4 7 Microtubules radiate from the centrosome in dynein inhibited cells. ................... 99 4 8 Simulations predict dynein induced buckling of microtubules. .......................... 100 4 9 Dynein forces are sufficient to center the centrosome. ................................ ..... 101 4 10 Microtubules undergo three distinct behaviors upon reaching the periphery .... 102 4 11 Simulations of centrosome centering. ................................ ............................... 103 4 12 Minus end microtubules underneath the nucleus does not traighten after laser severing. ................................ ................................ ................................ .. 104 5 1 Hypothetical effect of photobleaching due to imaging capture on a whole cell imaging experiment. ................................ ................................ ......................... 122 5 2 Solutions to Equation 5 6 showing how photobleaching during image capture ........................... 123 5 3 Solutions to Equation 5 10A and 10B that account for the presence of an actual immobile fraction. ................................ ................................ ................... 124 5 4 Example of the application of Equation 5 10A and 10B for fitting a GFP VASP FRAP experiment. ................................ ................................ .................. 125 5 5 Typical fitting for a GFP VASP FRAP experiment. ................................ ........... 127 5 6 Illustration of typical fitting failed to estimate in experiments with different time interval ................................ ................................ ................... 127 6 1 Image of intermediate filament in living cells. ................................ ................... 137 6 2 Cartoon of a dynein motor indicating how the minus directed motor bound to the cytomatrix exerts a force towards the microtubule plus end. ...................... 138

PAGE 13

13 6 3 Formation of acto myosin retrograde flow and stress fiber remodeling upon photo activation of Rac 1. ................................ ................................ ................. 139

PAGE 14

14 LIST OF ABBREVIATION S AAA ATPase associated with various cellular activities ATP Adenosine Triphosphate Nucleotide BCE Bovine Capillary Endothelial CCD Charge Coupled Device DBS Donor Bovine Serum DMEM EGFP Enhanced Green Fluorescent Protein FRAP Fluorescence Recovery After Photobleaching GTP Guanosine 5' triphosphate IF Intermediate Filaments INM Inner Nuclear Memberane KASH Klar si ch t ANC 1 Syne Homolog y LINC Li nker of N ucleoskele ton to C ytoskeleton LOV Light Oxygen Voltage MT Microtubule MTOC Microtubule Organizing Center NA Numerical Aperture ONM Outter Nuclear Membrane PCM PeriCentriolar Material PDMS Polydimethylsiloxane Rac1 Ras related C3 botulinum toxin substrate 1 Rho Ras homolog gene family SF Stress Fiber SUN Sad1p, UNC 84

PAGE 15

15 TAN Transmembrane Actin associated N uclear VASP Vasodilator Stimulated Phosphoprotein YFP Yellow Fluorescent Protein

PAGE 16

16 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MOTOR FORCE S AND NUCLEAR POSITIONING IN LIVING CELL S By Jun Wu May 2012 Chair: Tanmay Lele Coc hair: Richard Dickinson Major: Chemical Engineering The nucleus is the largest an d heaviest organelle in a cell and its proper positioning is crucial for basic cell functions, such as cell migration, division, fertilization and es tablishment of polarity. Irregularities in nuclea r movement (or nuclear positioning) are associated with various serious diseases. Nuclear movement in the cell is a complex process that involves interactions with all three cytoskeletal systems actin, intermediate filaments and microtubules The intera ctions occur through nuclear embedded molecular tethers that link to the cytoskeleton. The molecular linkage between the nucleus and the cytoskeleton can be also established through molecular motors, proteins that convert chemical energy to mechanical forc es. How molecular motors drive nuclear movement remains poorly understood Nuclear rotation and nuclear translation are the two types of nuclear motions commonly observed in the cell We investigated the physical mechanism for nuclear rotation in the cell W e found that the nuclear rotation angle is directionally persistent on a time scale of tens of minutes, but rotationally diffusive on longer time scale s and r otation required the activity of the microtubule motor dynein. Based on these results, a

PAGE 17

17 mechan ical model for torque generation on the nucleus was proposed. To investigate nuclear translation, we designed experiments utilizing two different te chniques protein photo activation and cell micromanipulation. The results from these experiments point to a tug of war between forward pulling and rearward pullin g forces on the nuclear surface generated by actomyosin contraction. Net nuclear motion occurs when the forward pulling force increases during lamellipodial formation. We also investigated the mechani sms of positioning of the centrosome, which is in physical proximity with the nucleus. The positioning of centrosome is very important in cell migration and cell division. W hether the centrosome is positioned by pulling force s or pushing force s originating in dynamic microtubules remains a controversy By severing a single microtub u le with femtosecond laser ablation we found that microtubules are under tension generated by dynein We also show that dynein mediated pulling force s are suffi cient to center th e centrosome

PAGE 18

18 CHAPTER 1 INTRODUCTION Nuclear P ositioning is an Important Cell F unction Cell and developmental processes like fertilization, cell migration and division and establishment of polarity require specific positioning of the nucleus within the cell (1 5) R ecent studies suggested that nuclear migration is a critical process for neuronal development, including interkinetic nuclear migration and nuclear translocation during retinogenesis (6) D uring mitosis in S. cerevisiae the nucleus needs to be moved into the bud neck for proper cell division (7) Defects in th e nuclear force generating system are related to disorders of the nervous system (8) and the musculo skeletal system (1) such as cardiac and skeletal myopathies, partial lipodystrophy, peripheral neuropathy (9) Recently, it h as also been linked to premature aging diseases like Hutchison Gilford progeria syndrome (HGPS) a t ypical Werner s syndrome and mandibulocaral dysplasia (MAD) (10) Despite the importance of nuclear positioning as a critical cellular function and its relevance to disease, our understanding of how forces are generated on the nuclear surface in living cells is surprisingly limited. At the molecular level, n uclear movement in the cell is a complex process that involves interactions with all three cytoskeletal systems actin, intermediate filaments and microtubules (2) On the cytoplasmic side, these interactions are mediated by molecular tethers suc h as nesprin family proteins that link the nuclear surface to the actin and intermediate filament cytoskeleton (11) and molecular motors such as nuclear bound dynein (12 14) The force transfer between the nucleus and the cytoskeleton are mediated by the LINC ( li nker of n ucleoskeleton to c ytoskele ton ) (15, 16) complex which includes t wo

PAGE 19

19 families of proteins containing KASH ( Klarsicht ANC 1 Syne Homology ) domains and SUN ( Sad1p, UNC 84 ) domains (17 20) L arge nesprin isoforms (m embers of KASH proteins) located on the outer nuclear membrane can bind to cytoskeletal structures with their amino terminal cytoplasmic domain. At the same time, these nesprins can interact with SUN proteins on the inner nuclear membrane via their carboxy l terminal perinuclear domain. SUN proteins in turn bind to lamins in the nuclear lamina chromatin, and other, yet unknown nuclear envelope proteins, thus completing the link from the cytoskeleton to the nucleus (15, 21 23) R ecent ly, it has been shown that nesprin 1G is involved in recruiting nuclei to, or anchoring nuclei at, the neuromuscular junction (3) O ther studies suggest that the mispositioning of t he nucleus may be caused by mutations or deficiency of nuclear lamina proteins and related proteins, such as lamin A/C (24, 25) Molecular motors are biological molecular machines that are the essential agents of movement in living organisms. These motors harness the chemical free energy released by the hydrolysis of ATP in orde r to perform mechanical work For example, the mo tor protein myosin is responsible for muscle contraction (26) and microtubule motors dynein and kinesin are responsible for intracel lular transportation (27 30) D yn ein has been shown to pull the nuclear surface in NIH 3T3 fibroblasts and cause nuclear rotation (12) and nuclear translation in other cell ty pes (22, 31, 32) Similarly, kinesin 1also mediates cell polarization (31) and involves in microtubule dependent nuc lear migration in C. elegans along with dynein (14, 16, 33) Dynein B ased T orque G eneration on the N ucleus Cytoplasmic dyne in was isolated from both C. elegans and calf brain white matter in 1987 by Lye et al. (34) and Paschal et al. (35, 36) Dynein is a large protein complex

PAGE 20

20 (1.2 MDa) composed of two iden tical heavy chains and several associ ated chains (37) The heavy chain contains si x AAA domains arranged in a ring and has the motor machinery that is responsible for transducing chemical energy into directed mechanical force applied to the microtubule surface. The intermediate and light chains help specify the intracellular location o f the dynein and regulate its motor activity (38) It s MT binding domain (MTBD) is located at the end of a 15 nm long coiled coil stalk that emerges from the AAA domains (39) Dynein walks on microt u bule toward the minus end (centrosome ) in a hand over hand like fashion ( (40) and Figure 1 1). There are some studies which explored how dynein connects to the nucleus Mosley Bishop et al (41) ha ve shown that microtubule based movement of nuclei during eye development in Drosophila requires K la rsicht, a prototype KASH domain protein that binds cytoplasmic dynein. Klarsicht is tethered in the ONM via a translumenal interaction with Klaroid in the INM to form a KASH/SUN protein pair, and then Klaroid binds to the lamin (42) Another dynein binding KASH domain protein, Zyg 12, has also been identified in C. elegans (43) It is believed to be related to the mammalian Hook protein family. N uclear anchoring has been shown to be mediated by the engagement of Zyg 12 associated dynein with nonc entrosomal microtubules that were nuclea ted at the plasma membrane (13) Nuclear rotation has been observed in many cell types (12, 44 47) One study suggested th at the rotation may be due to a transient bo nd between the centrosome and the nuclear membrane mediated by dynein and Hook/SUN family proteins, such that the nuclear rotation is coupled with the movement of the centrosome itself (48) Levy and Ho lzbaur (12) showed that the migration of fibroblasts into a newly created

PAGE 21

21 wound triggered nuclear translocation and coupled rotation, and both were decreased in dyne in null fibroblasts; interestingly these authors suggested that the centrosome is not bound to the nucleus. In non wounded cells, another interesting feature of nuclear rotation is that the rotation angle fluctuates (49) and rotation can occur over several hours of observation (46) The cause of fluctuations and long time persistence in the rotation angle is currently unknown. A physical explanation for how torque could arise through interactions between nuclear bound dynein and microtubules emanating from a stationary centrosome to create fluctu ating, persistent nuclear rotation is needed In Chapter 2 we show that in NIH 3T3 fibroblasts, nuclear rotation is a persistent random walk that requires dynein. The centrosome does not rotate with the nucleus. We formulated a model based on microtubules undergoing dynamic instability, with tensional forces between a stationary centrosome and the nuclear surface mediated by dynein. The model predicts that the fluctuations and persistence in nuclear rotation are due to the dynamic instability of microtubul es. A key model prediction is that the rotation should decrease with decreasing distance between the nucleus and the centrosome. We experimentally tested this prediction by showing that rotation in patterned cells (where the centrosome overlaps with the nu cleus and is close to the nuclear centroid) is considerably reduced compared to unpatterned stationary cells with larger nuclear centrosomal distance. Together, these results show that force generation by dynein on microtubules undergoing dynamic instabil ity is sufficient to explain the key features of nuclear rotation.

PAGE 22

2 2 Actomyosin F orces on the Nucleus As mentioned before, t he cell accomplishes nuclear motion by transferring active cytoskeletal forces onto the nuclear surface through molecular connections between the nuclear lamina and the cytoskeleton established by LINC complex proteins (15, 16) One of t he main source of the active forces are non muscle myosin II (N MMII) based contraction of F actin filaments (50, 51) Myosin refers to a family of F actin associated motors. It was first isolated as a complex with actin filaments by Kuhne and coworke r (1864) (52) though it was not until the 1940s that the complex was dissociated into the separate proteins, myosin and actin (Straub, 1941 1942: Szent Gyorgyi, 1941 1942) (52) The protein complex compo sed of actin a nd myosin is referred to as "actomyosin The discovery of the myosin crossbridge by H.E. Huxley in 1957 provided a molecular basis for the contraction of muscle: t he bending or rotation of these crossbridges causes the actin containing thin filaments to s lide relative to the myosin containing thick filaments, and the sliding of these filaments, in turn, leads to the shortening of the muscle (26, 53 55) T here are contras ting views on how the force generated by actomyosin contraction acts on the nucleus. For actomyosin based nuclear force generation, much of the research has focused on initial polarization mechanisms in a wounded NIH 3T3 fibrobalsts monolayer where the nuc leus is observed to move away from the leading edge while the centrosome stays in the same position within the first few hours after wounding (5, 56, 57) The polarization was abolished when myosin II activity was inhibited by blebbistatin treatment. M yosin activity was also found to be regulated by Cd c42 in these wounded cells W hen dynein was inhibited, the centrosome moved with the nucleus, resulting in failure of polarization. Based on these experimental results,

PAGE 23

23 Gunderson and coworker s proposed that the actomyosin retrograde flow pushes the nucleus possibly through specific linkages between the nucleus and actin filaments such as Syne/ANC 1 (5) Recently, Gundersen and co workers have shown that the re ar ward nuclear movement in wound ed cells is inhibited upon overexpression of KASH domains (57) Depletion of nesprin2G with small interference RNA (siRNA) also blocked re ar ward nuclear movement and could be rescued by expression of mini nesprin2G, which confirmed that nuclear movement required nesprin2G (57) N uclear envelope proteins SUN2, mini nesprin 2G and nesprin2G all colo calize d with dorsal actin cables to form linear arrays referred as TAN (transmembrane actin associated nuclear) lines which transmit force from retrograde actin flo w to the nucleus (57) Furthermore, d efects in the TAN line structure due to protein mutation s may be associated with striated muscle dis ease (56) In addition, there is evidence for a squeezing (pushing) force due to actomyosin contraction in the trailing edge (58) which may move the n ucleus toward the leading edge. O ther studies have suggest ed that the nucleus is normally under a state of tension (50, 59 62) Thus t into position by compressive cytoskeletal forces (due to retrograde actomyosin flow or pushing forces from the trailing edge) actomyosin forces, still remains controversial Both pushing and pulling forces may simultaneously operate on the nucleus, but of interest in this thesis is the net direction of the force balance (i.e. pushing ver sus pulling) and the dominant cytoskeleton origin of these forces in a single crawling NIH 3T3 fibroblast.

PAGE 24

24 In Chapter 3 we discuss two approaches to perturb the force balance on the nucleus. First, we inhibited myosin activity in the trailing edge locally T he nucleus move d toward the leading edge in a LINC complex dependent fashion. We next detached the trailing edge of motile fibroblasts and recorded nuclear motion and deformation in response to detachment as well as subsequent pulling of the trailing edge. We found that the nucleus underwent elastic deformations on manipulation of the trail ing edge in a myosin dependent and LINC complex dependent fashion. Collectively, these experiments suggest the presence of pulling forces on the nucleus from the trailing edge. We next used the Rac1 photoactivation assay to trigger the formation of new lam ellipodia (63) ; the nucleus was observed to move in the direction of newly formed lamellipodia in a myosin and LINC complex dependent but microtubule independent manner. Collectively, our results suggest that the nucleus is pulled on both sides, resulting in a tug of war between act omyosin forces. Consistent with this picture, trailing edge detachment is significantly reduced on disruption of nuclear cytoskeletal linkages, suggesting that the nucleus acts as an integrator of tensile actomyosin forces in a motile cell. Dynein Force Ge neration on Microtubules The c entrosome is an organelle that serves as the main microtubule organizing center (MTOC) where microtubules are produced. In animal cells, centrosome s are composed of two perpendicular arranged centrioles surrounded by an amorp hous mass of protein termed the pericentriolar material (PCM) (64) During mitosis, thses two centrioles duplicate to form two new centrosomes. The two new centrosome s then move to opposite ends of the nucleus and form spindles to separate chromosome into the two daughter cells (65, 66) During interphase, the centrosome normally stays at the

PAGE 25

25 center of a cell and close to the nucleus. The position of the centrosome is very important in that microtubules originating from the centrosome are responsible for the intracellular transportation of different organelles (67) Therefore, if the centrosome is mislocated, it can disru pt intracellular traffic Centrosome positioning involve s force generation by microtubules so do m any essential eukaryotic cell functions, including migr ation and mitosis (68 70) M icrotu bules have a large bending stiffness, with a thermal persistence length of the order of several millimeters Yet they are nearly always bent or buckled in cells, implying that they are being subjected to substantial lateral forces along their lengths (71, 72) or compressive forces at their tips (73) Compressive forces have been proposed as the means by which the centrosome and spindle bodies are centered (74 76) In this mechanism, microtubules spanning the shorter distance b etween the centrosome and the cell boundary exert a larger force because the critical buckling force is a strong function of length ( L 2 ). Consistent with this view, in vitro experiments (77) showed that microtubule organizing centers could be centered by elongating microtubules pushing on the boundaries of a microfabricated chamber On the other hand, there is increasing evidence that molecular motors play a key role in microtubule based force generation. Tensile forces generated by cytoplasmic dynein, a molecular motor that walks towards microtubule minus ends while bound to the cortex, have been proposed as the driver for centrosome centering (78 80) A similar dynein driven mechanism has been proposed for spindle body p ositioning in yeast cells (81) Dynein has also been implicated in the buckling of microtubules by pushing segments towards the plus end (82)

PAGE 26

26 The apparently contradictory results of previous studies leave basic questions regarding in vivo microtubule mechanics unresolved. What is the contribution of dynein to the force balance on individual microtubules? Do microtubules exert tensile or compressive forces on the centrosome? How does a radial array of microtubules cause centrosome cent ering? In Chapter 4, we discuss our investigation on dynein force generation on microtubules. We first performed experiments in which individual microtubules in BCE cells were severed by laser ablation. The goal was to remove a key element in the overall f orce balance, which is thought to be a combination of elastic forces from microtubule bending and elastic deformation of the surrounding cytomatrix (73, 83) Surprisingly, our experiments showed that microtubules do not straighten after severing, suggesting a large frictional resistance to lateral motion. Instead, segments near a newly created min us end usually increase in curvature following severing, as if the end of the segment was under a compressive load. By contrast, in dynein inhibited cells microtubule segments near the cut always straightened and did so much more rapidly than in normal cel ls. To explain these observations we propose a model for dynein force generation that accounts for stochastic binding and unbinding of dynein motors from the microtubules. An ensemble of these motors develops a steady force in the direction of the tangent to the microtubule and a frictional resistance transverse to the microtubule. Numerical simulations of individual microtubules show that the model can explain the concentration of microtubule buckles near the cell periphery (73) Simulations of centrosome centering by a radial array of microtubules are consistent with tensile forces on the centrosome; in the absence of motor forc es the centrosome

PAGE 27

27 remains off center, which is consistent with observations in dynein inhibited cells (5, 78, 80, 84) Fluorescence Recovery after Photobleaching (FRAP) Fluorescence recovery after photobleaching (FRAP) is an optical technique to study intracellular protein dynamics and binding kinetics in living cells (85 92) FRAP experiments involve bleaching o f fluorescently labeled proteins in a pre chosen location inside the cell with a high intensity laser pulse. When proteins are transiently bound to structures in the photobleached spot, the fluorescence recovers owing to exchange between fluorescently labe led diffusing molecules in the cytoplasm or membrane with the bound photobleached molecules in the bleached spot. The recovery curve can be fit to models to estimate transport and binding parameters. The accurate modeling of FRAP experiments, issues with p arameter estimation and validation of estimated parameters are active areas of interest (87, 93 101) The approach to fit FRAP experiments to mathematical models involves a suitable normalization of the experimental data (102) For example, if is the fluorescence in a spot in the cytoplasm, and bleaching occurs at then one way to normalize the signal is Here, the denominator represents the amount of fluorescence that should be theoretically recovered after photobleaching assuming one waits long enough in the experiment (i.e. while the numerator represents fluorescence that has recovered at any time. The assumption can be made in most cases that the bleaching pulse at itself does not alter the total fluorescence significantly. If the experiment i s then stopped at time (when the

PAGE 28

28 fluorescence appears to visually plateau), in many cases it is found that i.e. complete fluorescence recovery does not occur. If the usual procedure is to calculate the so called immobile fraction ; the hypothesis is that there is a sub population of fluorescent molecules in the bleached spot that do not recover to any measurable extent over the time While this a pproach is widely followed in the literature and may be applicable for many situations, it is obvious that if there was significant bleaching as a result of the image capture process itself, then even though there is no real immobile fraction. Of all the different experimental complications that make FRAP analysis difficult, the undesirable decay of the fluorescence due to the image capture process itself has he observed signal by the overall signal in the cell. This procedure can potentially invalidate the fitting of mathematical models to FRAP data owing to the arbitrary correction of experimental data with another time varying curve. If the effect of bleachi ng during image capture is significant and no correction to the data is applied, then this can invalidate the fitting because the mathematical models do not include the effect of photobleaching during image capture. Either way, neglecting the effect of pho tobleaching during image capture has the potential to render serious errors in the estimation of kinetic or transport parameters from th e FRAP experiment. In Chapter 5 we take the view that mathematical models for FRAP analysis should explicitly account f or the effects of bleaching during image capture instead of relying on corrections to data, or on the perfect experiment that does not suffer from the effects of photobleaching. We develop models that should be generally applicable and provide an experimen tal demonstration

PAGE 29

29 on how to use the models. The analysis discussed here can help bring greater clarity into the interpretation of FRAP experiments.

PAGE 30

30 Figure 1 1. Cartoon of dynein structure and walking on the microtubule. Dynein has two heavy chains which contai n six AAA domains arranged in a ring (red circles). The two coiled coil stalks emerging from the AAA ring contain MT binding domain. The cargo end binding site bind to dynactin to form a complex which could bind to cargo or cytomatrix. Individu al dynein molecules walk towards the microtubule minus end (centrosome).

PAGE 31

31 CHAPTER 2 DYNEIN MOTOR FORCES ON THE NUCLEUS In living cells, a fluctuating torque is exerted on the nuclear surface but the origin of the torque is unclear We found that t he nuclear rotation angle is directionally persistent on a time scale of tens of minutes, but rotationally diffusive on longer time scales. Rotation required the activity of the microtubule motor dynein. We formulated a model based on microtubules undergoi ng dynamic instability, with tensional forces between a stationary centrosome and the nuclear surface mediated by dynein. Model simulations suggest that the persistence in rotation angle is due to the transient asymmetric configuration of microtubules exer ting a net torque in one direction until the configuration is again randomized by dynamic instability. The model predicts that the rotational magnitude must depend on the distance between the nucleus and the centrosome. To test this prediction, rotation wa s quantified in patterned cells in which prediction, the angular displacement was found to decrease in these cells relative to unpatterned cells. This work provides the first mechanistic explanation for how nuclear dynein interactions with discrete microtubules emanating from a stationary centrosome cause rotational torque on the nucleus. Materials and Methods Cell Culture, Plasmids and T ransfection NIH 3T3 fibroblasts were cultured in DMEM (Mediatech Manassas, VA ) with 10% donor bovine serum (Gibco Grand Island, NY ) and 1% Penn Strep (Mediatech). For This chapter is reproduced from a paper previously published in Journal of Cellular Physiology with permission from John Wiley & Sons. Modeling and simulations are attributed to Dr. Richard B. Dickinson in the Departme nt of Chemical Engineering, University of Florida.

PAGE 32

32 microscopy, cells were cultured on glass bottomed dishes (MatTek Corp Ashland, TX ) DsRed CC1 plasmid was kindly provided by Prof. Trina A. Schroer from Johns Hopkins University, pDsRed plasmid was kindly provided by Prof. Scott S. Grieshaber from the University of Florida. YFP tubulin was prepared from the MBA 91 AfCS Se t of Subcellular Localization Markers (ATCC Manassas, VA ). Transient transfection of plasmids into NIH 3T3 fibroblasts was performed with Effectene Transfection Reagent ( Qiagen, Valencia, VA ). Time Lapse Imaging and A nalysis Time lapse imaging was performed on a Nikon TE2000 inverted fluorescent microscope with a 60X/1.49NA objective and CCD camera ( CoolSNAP, HQ 2 Photometrics, Tucson, AZ temperature, CO 2 and humidity controlled environmental chamber. Images were imported into MATLAB (Mathwork, Natick, MA ) and two nucleoli in the nucleus were tracked with time to calculate the rotation angle. The positions of the nucleoli ( and ) at time point were calculated to sub pixel resolution using previously published image correlation methods (103) and the angle between lines joining nucleoli in successive images was calculated as where is the angular displacement between time point and The

PAGE 33

33 autocorrelation function was calculated as The mean squared angular displacement was calculated as Immunofluorescence Immunostaining studies were carried out as previously described (104) with mouse monoclonal anti tubulin (Sigma Aldrich, St. Louis, MO ) and polyclonal rabbit anti tubulin (Abcam Cambridge, MA ) antibodies, and Hoechst 33342 in 4% paraformaldehyde fixed cells permeabilized with 0.1% Triton X 100 in PBS. Goat anti mouse and goat anti rabbit antibodies conjugated with Alexa Fluor 488 and Alexa Fluo r 594 fluorescent dyes (Invitrogen Carlsbad, CA ) were used as secondary antibodies. Cell Shape P atterning Cell shape patterning was done by using the micro contact printing technique described in (105) Molds for the stamps were produced with the UV lithography techniqu e by illuminating a positive photoresist through a chrome photomask on which micropatterns were designed (Photo Sciences, Inc. Torrance, CA ). PDMS (Sylgard 184 kit, Dow Corning Midland, MI ) was cast on the resist mold using a 10:1 ratio (w/w) of elastome (BD Biocoat TM Franklin Lakes, NJ ). The stamp was then dried and placed onto the substrate onto which the cells were plated. Ibidi dishes (Ibidi Verona, WI ) were chosen as the substrate. After 5 min, the stamp was removed and the remaining area was blocked with PLL g Poly ethylene glycol (SuSoS AG Dbendorf, Switzerland ),

PAGE 34

34 preventing protein adsorp tion and cell attachment. After treatment the surface was washed and cells were plated. Result s Nuclear Rotation is a Persistent Random Walk Nuclear rotation has been typically studied by time lapse microscopy; however, a quantitative analysis of fluctuat ing nuclear rotation has not been previously reported. We measured nuclear rotation angle with high accuracy using image correlation methods. The nucleus (and its contents) is known to rotate as a solid object (106) therefore tracking fiduci ary markers on the nucleus (nucleoli) allowed us to calculate the angular displacement between successive images. By tracking the non moving nucleus in cells fixed with paraformaldehyde which cross links the cellular contents and ensures zero rotation, we calculated the error in the image correlation as being less than 1%. As shown by the trajectories in Fig ure 2 1, the net angle through which the nucleus rotates exhibited short time fluctuations and typically long time persistence in the direction of rotat ion. Most trajectories displayed a significant angle of rotatio n in around two hours, while some fluctuated in position without achieving much net displacement. To quantitatively characterize the angular trajectories, the average mean squared angular disp lacement (MSD) was estimated (see methods for how MSD was calculated) from data p ooled from several cells ( Figure 2 1 C ). The MSD showed characteristics of a persistent random walk, with a parabolic shape at short time approachin g linearity at long time ( Fi gure 2 1 C ). To further characterize the directional persistence, we estimated the autocorrelation function of angular displacements ( Figure 2 1 C inset). This autocorrelation function resembled a double exponential, reflecting the relaxation

PAGE 35

35 of short time f luctuations in directional rotation and longer time relaxation of persistent directional rotation. Cytoplasmic Dynein is Required for Nuclear Rotation in Stationary F ibroblasts To determine the role of dynein in nuclear rotation in stationary fibroblasts we transfected cells with DsRed CC1, which inhibits dynein by competitive binding (107, 108) pDsRed plasmid transfected cells were used as the control. Transfection with DsRed CC1 significantly decreased nuclear r otation compared with cells expressed the control pDsRed plasmid ( Figure 2 2). We also found that rotation was abolished upon depolymerizing microtubules ( Figure 2 3 ). Also, the centrosome was observed not to rotate with the nucleus, but rather occupied a relatively fixed position in space even as the nucleus rotated through a significant angle ( Figure 2 4 ). These results are consistent with recent observations by Levy et al (12) in wounded fibroblasts, and Gerashchenko et al (49) in non wounded murine cells. Model Formulation We developed a mechanistic model for dyn ein forces on the nucleus. The purpose of this model is to demonstrate that retrograde MT associated motors pulling on the nucleus are sufficient to explain nuclear rotation with observed statistical characteristics. transiently bind and pull the nucleus toward the microtubule ( ) ends ( Figure 2 5 ). As is typically assumed for molecular motors, the motor speed depends on the pulling force (a vector) that the dynein molecule exerts on the nucleus. For simplicity, the force speed relation for dynein motors is assumed approximately linear (109) similar to other recent studies on dy nein mediated nuclear translation (81)

PAGE 36

36 ( 2 1) where is a unit vector directed toward the (+) end of the MT. Here represents the speed at zero load and represents the stall force required of the motor, such that the motor stalls when the component of the pulling force in the ( ) end direction is equal to MT nucleus linkages are assumed to have a finite lifetime, such that the force accumulated by translation of the nuclear surface at velocity relative to the MT (assumed stationary), or by the motor walking along the MT, is re lieved upon dissociation of the linkage. The accumulation of the force depends on the stiffness of the linkage, and it is simplest to assume that the linkages behave as linear springs with stiffness Thus, accounting for the changing load, the force on the nuclear surface changes with time as ( 2 2) Upon substitution of Equation 2 1 for and integration, this become ( 2 3) Linkages are assumed to dissociate with a first order rate constant (s 1 ) such that the probability of a linkage existing at time is The mean force over the bond lifetime is thus ( 2 4)

PAGE 37

37 where is the average force per dynein molecule on a stationary nucleus For dynein density ( number /length) the net mean force per unit length, is then ( 2 5) where is the mean longitudinal force/MT length for a stationary nucleus, and is the friction coefficient ( ( force/speed ) /MT length) to lateral motion. Mechanics and kinematics of the n ucleus. The net force and torque exerted on the nucleus by an MT depends on the length span of the MT that is close enough to interact with the nucleus. Let be the position vector on the MT parameterized by contour length between the centrosome position ( ) and the (+) end ( ). Further, let and repr esent the beginning and end of the interaction range with the nucleus in which the distance between the MT and nuclear surface is within Based on the geometry of a line passing near a sphere, MT will interact with the nucleus provide d where is the center of rotation of the nucleus The endpoints of the interacting span of the MT are the n (2 6) ( 2 7 ) where is the contour position of the (+) end. The net MT force and torque are thus ( 2 8 )

PAGE 38

38 ( 2 9 ) where the local velocity generally depends on translation and rotation as ( 2 10 ) Neglecting other cytoskeletal (frictional) contributions to force and torque on the nucleus and neglecting inertia, the net force and torque balances are ( 2 10) ( 2 11) which yield six equations for the six unknown components of vectors and The model can be simplified substantially by neglecting translation (assuming that the distance between the centrosome and the nucleus does n ot cha nge during rotation; see Figure 2 4 ) and allowing rotation only about the vertical z axis, consistent with experimental observations. A possible explanation for rotation about only the z axis is that the nucleus is compressed into an oblate spheroid by cell spreading (104, 110) thus rotation in other directions is resisted as it requires substantially more deformation of the cytoskeleton. Hence, rather than six degrees of freedom, we need only consider the compone nt of the rotational velocity normal to the image plane ( ) determined by solving such that ( 2 12)

PAGE 39

39 Lastly, although dynein may uniformly coat the nuclear surface, we speculate that s run parallel to the nuclear surface and are inhibited when MTs impinge normal to the surface To account for this, we assume that the effective local linkage density varies with angle of incidence of the MT with the surface as Upon canceling from numerator and denominator, Equation 2 12 becomes ( 2 13) Simulations of dynamic instability and nuclear r otation. The rotati onal velocity obtained from Equation 2 1 3 depends on the current configuration of microtubules In the simulations, MTs are assumed to elongate at constant speed from the centrosome until c atastrophe occurs. When catastrophe occurs, MTs begin to shrink at a constant speed until recovery. On recovery, they begin to grow again. Catastrophe and recovery have constant probabilities per unit time and respectively The parameters are shown in Table 2 1. If an MT shrinks completely to the centrosome, a new MT immediately nucleates in another random direction on the unit sphere thereby maintaining a constant number N Any MT which impinges on the cell or nuclear boundaries is assumed to stop growing and remain a constant length until catastrophe straight and rigid in this treatment Although actual MT's are not straight in vivo, MT's typically do not change much in direction over the distance from the centrosome to the nucleus; therefore treating them as straight is a reasonable simplification. Another complication is that some MT's may wrap trace along

PAGE 40

40 the nuclear surface, thereby increasing the effective in teraction length of the MT. However, increasing the span of interaction is similar to changing the dynein density, which has no effect on the predicted long time dynamics. This is because changing the interaction span changes both the pulling force and opp osing friction force by equal proportions (see Equation 2 13 which shows that the rotational velocity does not depend on the density of dynein). Model Results We hypothesized that the rotation was due to an imbalance of torque from MT associated dynein pulling forces on the nucleus. In this model, an instantaneous imbalance in the net dynein pulling force would arise when, as the MT configuration evolves due to dynamic instability, more microtubules contact one side of the nucleus t han the other. Angular trajectories simulated based on this model reproduced motions that were qualitatively similar to those observed experimentally ( Figure 2 6 A and B ). Like the experimental data, the autocorrelation function of the simulated trajectorie s exhibited short time fluctuations due to stochastic contacts between the MTs and the nucleus, and a long time directional persistence in rotation that relaxed as the MT network relaxed due to dynamic instability ( Figure 2 6C ), as evident by the long time non zero tail of the autocorrelation plot. The time scale of the fluctuations, which is evident in the autocorrelation function o f ten minute displacements, depends primarily on the parameter group of which a value of 56 pN s/ m yie lded good quantitative agreement to the experimental data This value also yielded excellent agreement for the mean squared displacement( Figure 2 6D ). Based on recent FRAP measurements of

PAGE 41

41 cytoplasmic dynein (111) a rough estimate of Then pN / m which is not unreasonable for proteins (76) A testable prediction of this model is that the amount of rotation caused by an ome directly underlies the nucleus. As shown in Figure 2 6E the mean squared rotation angle measured from the simulations decreased as the centrosome was placed at positions closer to the nuclear centroid. This prediction can be explained by the fact that the dynein generated torque is smaller when the lever arm length (i.e., the distance from the nuclear centroid to the MT nucleus contact position) is smaller ( Figure 2 5 B ). We tested this model prediction by patterning cells into symmetrical shapes as dis cussed below. Nuclear Rotation Depends on Nuclear Centrosomal D istance In stationary, unpatterned cells, the centrosome is typically observed in two dimensional images at one side of the nucleus ( Figure 2 7A ), and below the mid plane of the nucleus. It i s known that the centrosome is positioned at the cell centroid due to centering by dynamic microtubules interacting with the cell periphery (78) We reasoned that in patterned cells of square shape, the centrosome should be located at the geometrical center of the square; this was indeed found to be the case ( Figure 2 7A ). The nucleus was also observed close to the center of the square, such that the centrosome overlapped with the nucleus in two dimensional images ( Figure 2 7A ). Confocal imaging showed that the centrosome was underneath the nucleus in patterned cells ( Figure 2 8 ). This arrangement was rarely observed in unpatterned cells; the

PAGE 42

42 projected distance between the centrosome and the nuclear centroid was significantly higher in unpatterned cells ( Figure 2 7B and C ). Our model predicts that a decrease in nuclear centrosomal distance should result in decreased torque on the nucleus for a given MT configuration. To test this prediction, we tracked rotation of the nucleus in square cells. The nucleus in square cells was observed to rotate significantly less than unpatterned cells ( Figure 2 9 ), confirming a key prediction of the model Indeed, as indicated by the solid lines in Figure 2 9 the MSD generated by simulations performed with the centrosome located at the experimentally measured average position in patterned cells agreed quantitatively with experimental measurements Discussion In this study we quantified the persistent directional rotation of cell nuclei in fibroblasts and used the data to propose a mechanistic model that can predict nuclear motion based on tensional forces on perinuclear microtubules generated by the minus ended motor dynein In this model, the mean force per dynein molecule depends on the local velocity of the nucleus rel ative to the microtubule, and the net force and torque on the nucleus depends on the instantaneous configuration of perinuclear microtubules. The model qualitatively captures the dynamic behavior of nuclear rotation, and agrees quantitatively upon fixing o ne unknown parameter group ( ). We experimentally tested a key prediction of the model that the rotation should depend on the nuclear centrosomal distance. Together, this work provides the first mechanistic explanation for how nuclear dynein interactions with discrete microtubules emanating from a stationary centrosome can cause nuclear rotation.

PAGE 43

43 The finding that the centrosome is close to the nuclear centroid in patterned cells can be explained by the nearly isotropic inward flow of a ctomyosin from the cell periphery (112) As rearward flow continuously occurs from the corners of the squares of the patterned cells (113) causes the nucleus to be centered (although not perfectly depending on the local shape of the lamellipodial protrusions). Because the centrosome is centered at the cell centroid due to microtubule cortical actin interactions mediated by dynein (78) this forces the nucleus and t he centrosome to overlap. In unpatterned cells, the centrosome is always observed on one side of the nucleus, probably due to asymmetrical positioning of the nucleus by rearward actin flow only from the leading edge of the (polarized) cell. The decrease i n nuclear rotation in patterned cells can be interpreted in the light of the model shown in Figure 2 5 When the centrosome is under the nucleus, then torque is generated primarily by microtubules oriented parallel to the lower nuclear surface. When the ce ntrosome is not underneath the nucleus but rather to the side, then the net torque generated is expected to be higher when the MT configuration evolves by dynamic instability to become spontaneously asymmetric. Mathematically, the z component of the torque decreases to zero when the centrosome is positioned at the nuclear centroid because the radial position vector drawn from the centroid to the point of dynein forces becomes parallel to the microtubule(and mean dynein force) direction ( Figure 2 5B ). Also, the lever arm (the distance from the nuclear centroid to the MT nucleus contact position) is smaller ( Figure 2 5B )

PAGE 44

44 That the centrosome is essentially stationary during rotation is consistent with similar findings reported in two other studies (12, 49) Cons istent with these findings, the model does not require the centrosome to rotate for the nucleus to rotate rather, the origin of the torque is due to the asymmetric spatial distribution of nuclear associated microtubules. Thus, the centrosome is actually relevant to the rotation process because its position controls the degree of spatial asymmetry thereby influencing the rate of rotation. The model qualitatively reproduces our experimental finding that nuclear rotation is a random walk with directional pe rsistence. The fluctuations are a natural consequence of the fact that the microtubule configuration fluctuates due to dynamic instability The time over which the nuclear rotation persists on average in a given direction is the time it takes for an initial asymmetry in the microtubule configuration to be reversed by dynamic instability. Thus, the model nicely explains a number of features of nuclear rotation that have been observed by us and other researchers. The model assumes a linear force speed relationship for dynein, similar to another recent paper on dynein mediated nuclear movement (81) Given the level of coarse graining in the model, small deviations from linearity would not alter the main conclusions of the paper. Although Equation 2 1 allows a load on the motor in the ( ) end direction ( > 0) to enhance the motor speed above because tangential speeds (~0.001 m/s) are much less than this situation is rarely encountered in the application of the model Therefore, any non the force velocity relation for dynein that is not accounted for in the model should not impact the model predictions.

PAGE 45

45 Dynein pulling has been proposed previously as a mechanism for pulling of microtubule organi zing centers (78) A recent model for oscillatory spindle pole body translation in S. pombe assumed that directional translation arises from breakage of dynein linkages by the "winning" side in a tug o war of competing pulling mic rotubules (81) In that model, the spindle body translates at nearly constant maximum motor speed until it reaches one end of the cell, consistent with the exp erimental observations in that system In contrast, nucleus rotational speeds are stochastic, fluctuating in magnitude and direction, but with values much smaller than the dynein motor speed suggesting no winning side of the dynein tug o war. We also note that there is no direct experimental evidence of dynein force induced breakage of dynein linkages in vivo, and our model can explain the relevant observations without invoking this untested assumption. In fact, for observed maximum tangential speed of of = 0.001 m/s, the force over the bond lifetime (given by ) is significantly smaller than 1 pN for our estimates of (see model results). This force is much smaller than forces required for bond breakage that are typically several pN (114) There is evidence that nuclear rotation can be influenced by other molecules. For example, Lammerding and co workers (46) showed that over expression of nesprin 1, a protein that binds the nucleus to the F actin cytoskeleton, reduce s the extent of rotation. This suggests that rotation ma y at least in part be influenced by frictional drag due to connections with other members of the cytoskeleton. However we did not observe an increase in the rotation in NIH3T3 fibroblasts on nesprin 1 depletion ( Figure 2 10 ). Similarly, myosin inhibition may increase rotation (12) which again suggests a role for actomyosin in opposing the rotation. Drag in our model is due to dynein motors bound

PAGE 46

46 to th ose microtubules that oppose the rotation. We cannot rule out drag due to frictional connections with actin or intermediate filaments. While including this effect would add a term similar to the drag due to dynein, it is not expected to change the main con clusions of the model. Summary of Findings In this chapter, a mechanical model describing torque generation on the nucleus was proposed. A key prediction of the model is that the magnitude of nuclear rotation depends on the distance between the centrosome and nuclear centroid, which was confirmed by nuclear rotation experiments in patterne d cells. Our model provides new insight into how nuclear dynein interactions with discrete microtubules emanating from a stationary centrosome cause rotational torque on the nucleus which may further help us understand the biological function of nuclear rotation in the near future.

PAGE 47

47 Figure 2 1. Nuclear rotation is a biased random walk in NIH 3T3 fibroblasts. (A) Captured images of a rotating nucleus. Scale bar is 5 m. (B) Time dependence of the rotation angle; time between successive data points is two minutes. The angle fluctuates randomly. (C ) Time dependence of the pooled angular mean squared displacement (n=25 cells). The MSD shows a parabolic shape at short times followed by a linear dependence at longer times which indicates a persistent random walk. Inset shows the averaged autocorrelatio n of angular displacements over 10 min intervals ( ) indicating a fast decay followed by long time decay, again consistent with the conclusion that the rotation is a persistent random walk Error bars indicate st andard error (SE ).

PAGE 48

48 Figure 2 2 Dynein inhibition significantly reduces nuclear rotation. NIH 3T3 fibroblasts were transfected with pDsRed as mock control (A), and pDsRed CC1 for dynein inhibition (B; inset is re scaled plot). A clear decrease in the rotation is observed i n dynein inhibited cells. This is confirmed from the MSD plots(C), with MSD in control cells ( squares n=14 cells) found to be significantly larger than that in dynein inhibited cells ( circles n=11 ce lls). Error bars indicate SE

PAGE 49

49 Figure 2 3. Microtubules drive nuclear rotation. (A) Nuclear rotation in NIH 3T3 was tracked for three hours, followed by treatment with nocodazole, and then tracking of nuclear rotation for about 10 hours post treatment. The dash line shows the time when the microtub ule polymerization inhibitor, nocodazole, was added (10 M). The nucleus did not rotate significantly after the addition of nocodazole. (B ) Time dependence of the pooled angular mean squared displacement (black, control; red, nocodazole treated, N>5). Err or bars indicate SE. Figure 2 4. The centrosome does not rotate with the nucleus. The top panel is captured DIC images of nuclear rotation in NIH 3T3 fibroblast from movie 1. The time unit is minute. The bottom panel is YFP tubulin labeled centrosom es. Circles in the fluorescence image mark nucleoli to aid visualization. Scale bars, 5 m.

PAGE 50

50 Figure 2 5 Schematic of the nuclear rotation model. (A) Dynein molecules walking on microtubules (straight lines) generate forces ( ) on the nuclear surface directed toward the centrosome (intersection of straight lines) The resulting mean net force from the microtubule and the lever arm (vector where is the position on the contour, is a unit vector directed towards the MT plus end, and is a unit vector directed from the centrosome to the center of the nucleus) create a torque on the nucleus (B) The magnitu de of the torque depends on the centrosome position, because the lever arm length is smaller when the centrosome is closer to the nucleus centroid

PAGE 51

51 Figure 2 6 Simulations of nuclear rotation in a circular cell (A) Simulation snapshots of microtubule configuration and nucleus orientation at 30 min intervals, for two different centrosome positions: a distance of eight microns from the nucleus center (upper sequence), and a distance of three micron s from nuc leus center (lower sequence ) Diame ter of the cell is 50 microns. (B ) Example simulation trajectory of nuclear rotation (default parameters are shown in Table 2 1),

PAGE 52

52 exhibiting short term fluctuations and long time directional persistence. The same trajectory sampled at 10 min intervals ( red as done in the experiments) is also shown. (C ) The autocorrelation function of rotational displacements is plotted versus time increment (solid lines), as calculated by the same method used for Figure 1C (inset), for three values of (from upper to lower: = 20 pN s/ m, 55 pN s/ m, and 200 pN s/ m, respectively. Red c ircles indicate the calculated average autocorrelation function fr om experimental trajectories. (D ) Model predictions of mean squared angular displ acement versus time for the same parameters in (C ) (solid lines, from upper to lower: = 20 pN s/ m, 55 pN s/ m, and 200 pN s/ m, respectively; red circles are experimental data points). (E) Model predictions of mean squared angular displacement versus time for three distances between the centrosome position and the nuclear centroid (solid lines, from upper to lower: black, 8 m, blue, 3 m, and red, 0.5 m, respectively ).

PAGE 53

53 Figure 2 7 Distance between the centrosome and the nuc leus decreased in patterned square NIH 3T3 cells. (A) Positions of the centrosome (green) in unpatterned (top panel) and patterned cells (bottom panel). Nucleus is stained blue and microtubules in red. The last two pictures on the right are overlay images. All scale bars are 5 m. (B) Distributions of nuclear centrosomal distance in unpatterned (black) and patterned cells (red). (C) Nuclear centrosomal distance in patterned square cells is significantly smaller than that in unpatterned c ells. Error bars ind icate SE ; p < 0.01.

PAGE 54

54 Figure 2 8. The centrosome (arrow, green dot) is underneath the nucleus (blue). Cross section view of centrosome position in square cells acquired with confocal imaging. NIH 3T3 fibroblasts were patterned into square shapes and stained tubulin antiobody and Hoechst 33342. Scale bar, 5 m Figure 2 9 Nuclear rotation in unpatterned cells ( circles n=25) is significantly larger than that in patterned cells ( squares n=24), which agrees with the MSD generated by simulati on using experimentally measured average centrosome nucleus distances of 8 m and 3 m, respectively (solid li nes). Error bars indicate SE

PAGE 55

55 Figure 2 10. Nesprin 1 knock down affects nuclear rotation. Cells were transfected with 100 nM of SMARTpool siRNAs (Dharmacon, Lafayette, CO) against human nesprin 1 using siLentFect lipid transfection reagent (BioRad, Hercules, CA). The siRNA oligonucleotide target sequences used were as follows: GAAAUUGUCCCUAUUGAUU, GCAAAGCCCUGGAUGAUAG, GAAGAGACGUGGCGAUUGU and CCAAACGGCUGGUGUGAUU. Nontargeting SMARTpool siRNAs served as controls. (A) Western blot analysis of nesprin 1expression in NIH 3T3 fibroblasts. NIH 3T3 fibroblasts transfected with siRNA targeting nesprin 1 (Nes 1) show a significant reduction in nesprin 1 expression as compared to nontransfected cells (Control) and cells transfected with control siRNA (Scramble). (B) Nuclear rotation in control NIH 3T3 fibroblasts (black square, n=27) is not increased in nesprin 1 knock down cells (red circles, n=35) Int erestingly, a slight decrease is observed in nesprin 1 depleted cells; the p value for the difference at 40 minutes is p=0.08; so this is not statistically significant. MSD was calculated using overlapping time intervals. Error bars indicate SE.

PAGE 56

56 Table 2 1. Nuclear Rotation Model Parameters Symbol Parameter Range Source Value Used Maximum dynein force 5 8 pN (115) 7 pN Speed of unstressed dynein 0.8 m/s (109) 0.7 m/s Dynein spring constant No measured value k off = 56 pN s/ m Dynein nucleus off rate No measured value Dynein density (number /length) No measured value Not needed N Number of microtubules 200 500 (116) 250 MT polymerization speed 5 10 m/min (116, 117) 7 m/min MT depolymerization speed 15 20 m/min (116, 117) 17 m/min MT catastrophe rate constant 0.01 0.06 s 1 (116, 117) 0.05 s 1 MT recovery rate constant 0.04 0.2 s 1 (116, 117) 0.19 s 1 MT Nucleus Interaction Distance Based on length of cytoplasmic dynein 60 nm

PAGE 57

57 CHAPTER 3 THE NUCLEUS IS IN A TUG OF WAR BETWEEN ACTOMYOS IN PULLING FORCES IN A CRAWLING CELL On cellular length scales, the nucleus is massive (~10 15 microns in diameter) and stiff relative to the cytoplasm Motion of such a large object in the crowded intracellular space requires a significant expenditure of energy and hence represents a significant task fo r the motile cell. We found that the nucleus in a single, polarized crawling fibroblast wa s pulled forward toward the leading edge. The pulling force s originate from actomyosin contraction between the leading edge and the nuclear surface; these forces are opposed by actomyosin pulling in the trailing edge. Microtubules serve to damp fluctuations in nuclear position, but are not required for directional nuclear motion. Our results indicate that the nucleus is under net tension in a crawling cell due to a competition between actomyosin pulling from the front and back of the crawling cell. Materials and Methods Cell C ulture, P lasmids and T ransfection Drug Treatment NIH 3T3 fibroblasts were cultured in DMEM (Mediatech Manassas, VA ) with 10% donor bovine serum (Gibc o Grand Island, NY ) For microscopy, cells were cultured on glass bottomed dishes ( WPI Sarasota FL ( BD Bi ocoat TM Franklin Lakes, NJ ) For photoactivation experiments, cells were serum starved for two days in DMEM with 1% BSA (Sigma Aldrich, St. Louis, MO ). Micromanipulation experiments and related figures are attributed to T.J. Chancellor and Agnes Mendonca in the Department of Chemical Engineering, University of Florida. Some of the experiments regarding nuclear movement in migrating cell were done by Nandini Shekhar in the Department of Chemical Engineering, University of Florida.

PAGE 58

58 YFP tubulin was prepared from the MBA 91 AfCS Set of Subcellular Localization Marker s (ATCC Manassas, VA ). GFP actin was provided by Dr. Donald E. Ingber in Harvard University EGFP KASH4 was previous described in (31) mCherry PA Rac1 (Addgene plasmid 22027). DsRed CC1was kindly provided by Prof. Trina A. Schroer from Johns Hopkins University Transient transfection of plasmids into NIH 3T3 fibroblasts was performed with Lipofactamine TM 2000 t ransfection r eagent (Life Technologies, Invitrogen, Carlsbad, CA ). For microtubule disruption experiment, cells were treated with nocodazole (Sigma Aldrich, St. Louis, MO) at the final concentration of 1.6 M for over one hour prior to the experiment. For myosin inhibition, cells were treated with blebbistatin (EMD, Gibbstown, NJ) at the final concentration of 50 M for over one hour prior to t he experiment. Time L apse I maging an d A nalysis Time lapse imaging was performed on a Nikon TE2000 inverted fluorescent microscope with a 40 X/1.4 5 NA oil immersion objective and CCD camera (CoolSNAP, HQ 2 Photometrics, Tucson, AZ ). During microscopy, cells temperature, CO 2 and humidity controlled environmental chamber. Confocal Microscopy and Protein Photo A ctivation Protein photo activation is a novel technique developed in the last decade for studying spatio temporal dynamics o f enzyme activity in living cells This technique is based on biosensors called light oxygen voltage (LOV) domains. The idea is that a light switch biosensor in its closed form was fused to the target protein to block the target protein from binding to any effectors. Then when light with certain wavelength shines on the biosensor, it changes the conformation and release the inhibition of the target protein. Therefore the target protein would be activated (118)

PAGE 59

59 Recent NMR studies revealed the mechanism of a protein light switch in Avena sativa phototropin1: a flavin binding LOV2 domain interacts with a carboxy terminal helical extension (J ) in the dark. Photon absorption leads to formation of a covalent bond between Cys 450 and the flavin chro mophore, causing conformational changes that result in dissociation and unwinding of the J helix (118, 119) Hahn and coworkers (63) used LOV2 J sequence (404 547) as the light switch of the amino terminus of a constitutively active Rac 1. They showed that locally photo activation of Rac was sufficient to generate polarized cell movement. Another study by Wang et al. (120) showed similar role of Rac in collective guidance of cell movement using the same mehthod. Machacek et al. (121) applied ph oto activation of Rho GTPase to study the spatio temporal coordination of Rho GTPase during cell protrusion. In our experiment, t he samples were imaged on a Leica SP5 DM6000 confocal microscope equipped with a 63 X oil immersion objective. For photoactivation, a region in between the nucleus and the edge of a cell, which is approximately the size of the nucleus, was chosen using the ROI (region of interests) function. 488 Argon laser was applied at 1% power to activate Rac1. Images were tak en every 10 seconds at 1024X1024 resolution. temperature, CO 2 and humidity controlled environmental chamber. Image Analysis Image series from cell migration experim ents were processed in ImageJ (NIH). Then they were imported into Matlab ( MathWorks, Natick, MA ). Programs were developed to track the nuclear centroid and the contour of cells. Image series from the photoactivation experiment were imported into Matlab ( MathWorks, Natick, MA ), and a program w as developed for nuclear position tracking.

PAGE 60

60 After the positions of nuclei in different experiments were obtained, the coordinates w ere rotated as shown in Figure 3 1. The vector pointing from nuclear centroid at time=0 to the activation center was used as trajectories were rotated following this rule. The directional movements then were Micromanipulation by Microinjector Eppendorf Femtoje t microinjection system (Eppendorf North America, Hauppauge, NY ) was used. The microneedle was inserted underneath the trailing edge of the cell, and then the trailing tail was lifted up. The same method was applied to the leading edge lamellipodia. Result s Myosin Inhibition In the Trailing Edge C auses Nuclear Motion Toward the Leading Edge without Change in Cell S hape. To locally inhibit myosin activity, we used a micropipette to introduce a flow of blebbistatin containing solution at high concentrations (500 M) over the trailing edge. In less than five minutes of local introduction of blebbistatin, the nucleus moved toward the leading edge and away from th e trailing edge (Figure 3 2 A). This occurred without any appreciable change in trailing edge shape, and without any forward motion of the cell body We found that nuclear movement toward the leading edge on myosin inhibition in the trailing edge was eliminated when KASH domains were over expressed in fibroblasts (GFP KASH4 expression competitively inhib its nuclear cytoskeletal linkages, Figure 3 2A and B, (31) ).

PAGE 61

61 Solid Like Coupling Between the Nucleus and the Trailing E dge. We next performed micromanipulation experiments in which the trailing edge was detached by introducing the micropipette tip und er the trailing edge and snapping it. Trailing edge detachment resulted in movement of the nucleus toward the leading edge (Figure 3 3 A). The forward motion of the nucleus on detachment of the trailing edge could be interpreted as due to either pushing for ces generated by forward motion of the trailing edge contents, or due to a disruption of the tensile forces on the trailing surface of the nucleus, resulting in a net forward force on the nucleus in a manner as suggested by the myosin inhibition experiment Forward nuclear motion on trailing edge detachment was eliminated in GFP KASH expressing cells suggesting a requirement for nuclear cytoskeletal linkages for the effect(Figure 3 3A, B and C). A closer examination of the dynamics of nuclear shape changes on trailing edge detachment revealed that in control cells, the nucleus underwent considerable deformation while its shape was unaltered in KASH expressing cells (Figure 3 3 D). When the trailing edge was detached, the nucleus changed shape from an elliptic al, elongated cross section to a circular cross section. This shape change again could occur either due to pushing forces as the trailing edge retracts or a dissipation of pulling forces. More interestingly, when a detached trailing edge was again pulled o n and extended, the nucleus was observed to almost instantaneously elongate again to nearly its original shape in a myos in dependent manner (Figure 3 3E ). This experiment suggests that the nucleus is hardwired with actomyosin tensile structures, because wh ile pushing forces can move the nucleus forward on detachment, elongation of the nucleus due to trailing edge extension cannot be explained by pushing forces in the reverse direction.

PAGE 62

62 The forward motion of the nucleus on local myosin inhibition and the m yosin dependent solid like coupling between the nucleus and the trailing edge suggest strongly that the nucleus is pulled toward the trailing edge by actomyosin forces. The transfer of the pulling forces to the nuclear surface occurs through molecular link ages with the cytoskeleton as suggested by the lack of nuclear motion or deformation in GFP KASH4 expresssing cells. Forward Nuclear Movement does not Require Trailing Edge D etachment As the trailing edge is mechanically coupled with the nucleus, we next e xamined the correlation between nuclear motion and trailing edge motion. Forward motion of the nucleus toward the leading edge did not necessarily require the detachment of the trailing edge (Figure 3 4 A and B). The forward motion of the nucleus correlated with forward motion of the cell centroid, but not with the trailing edge (Figure 3 4C, D and E). This is not to suggest that nuclear motion does not occur when the trailing edge detached; but significant forward motion can occur of the nucleus without lar ge changes in the shape of the trailing edge. Forward Motion of the Nucleus Occurs due to Actomyosin Contraction Between the Leading Edge and the N ucleus If the nucleus is pulled on from the trailing edge, then what causes its motion forward in the absenc e of significant trailing edge detachment? One hypothesis (originally proposed by Lauffenburger and Horwitz (122) ) is that the nucleus is pulled for ward by actomyosin contraction occurring in the leading edge. Owing to its large area, myosin inhibition locally and/or detachment of the lamellipodium appeared to present experimental difficulties. To circumvent these problems, we instead adapted the Rac1 photoactivation assay recently introduced by Hahn and coworkers (63, 120, 121)

PAGE 63

63 The idea is to cause local polymerization of F actin; this newly created F actin is expected to combine with myosin and result in increased loca l contraction. On the creation of local lamellipodium with Rac1 photoactivation, the nucleus was observed to move toward the direction of th e la mellipodium (Figure 3 5A and B). The motion of the nucleus was similar to the motion observed in motile cells a bove. We first examined the role of microtubules in the motion. The trajectories of the nucleus did correlate with centrosomal trajectories (both moved in the general direction of the newly created lamellipodium, Figure 3 5 C). However, depolymerization of microtubules with nocodazole did not eliminate the directional motion of the nucleus; on the contrary, nuclear movement increased in the direction of motion and the fluctuations in nuclear position also increased (Figure 3 5D through G, also see Figure 3 1 and Figure 3 6 ). These results suggest that fluctuations in motion of the nucleus are damped because the nucleus is bound to microtubules (through nuclear envelope embedded motors such as dynein and/or kinesin). But the directional motion of the nucleus o n the formation of new lamellipodia does not depend on microtubule motor activity. Consistent with this picture, we found that the nucleus can move without requiring trailing edge detachment in dynein inhibited cells; the motion is similar to that in contr ol cells because the nucleus tr acks the cell centroid (Figure 3 5H, I and J). On inhibition of myosin activity through blebbistatin treatment, the nucleus did not move directionally in the photoactivation experiment; nor did it move when KASH4 was overexp ressed in the cells (Figure 3 5D through G also see Figure 3 6 ). Taken together with the microtubule disruption experiments, the results point to actomyosin contraction as the dominant pulling force on the nucleus which moves it forward during motility of

PAGE 64

64 NIH 3T3 fibroblasts. Consistent with this picture, when the lamella was severed with a micropipette, the nucleus was observed to move back; this motion was reduced significantly on over expression of GFP KASH in the cell (Figure 3 7 ). Discussion Our results suggest that there is a tug of war between forward pulling and rearward pulling forces on the nuclear surface. The dominant contribution to these pulling forces is from actomyosin contraction. Given that F actin continuously polymerizes at the leading edge, there is a continuous source of newly polymerizing actin that can contract to pull on the nucleus. The trailing edge is relatively stable in shape (until it detaches), and hence it is reasonable to surmise that the tensile pulling forces in t he trailing edge are relatively constant in magnitude. Net forward motion of the nucleus would be predicted to occur when pulling forces at the front exceed those at the back. An actomyosin tug of war is thus a simple positioning mechanism for the nucleus in a crawling cell. This mechanism suggests an intriguing possibility that contractile forces transmitted through the nucleus to the trailing edge cause detachment of the trailing edge. To test this, we measured the trailing edge detachment frequency in control versus KASH overexpressing cells. There was a significant reduction in the trailing edge detachment frequency in KASH ex pressing cells (Figure 3 8A, B and C). KASH expressing cells move forward through a sliding of the trailing edge rather t han det achment (Figure 3 8D, E and F). As expected, KASH expressing cells migrate poorly (due to a lack of trailing edge detachment) (123) Thus, the nucleus acts as a long transmitter of forces between the front and the back of the cell. While nuclear positioning is clearly important for motion of the nucleus in the direction of the motile

PAGE 65

65 cell, our results also suggest that the positioning mechanism is crucial for normal cell motility. Summary of Findings In s ummary, using two different methods, we were able to show that in single craling fibroblasts, the nucle ar movement is the integrative result of actomyosin pulling force s originating in the leading edge and the trailing edge.

PAGE 66

66 Figure 3 1. Schematic of how collective trajectories of nuclear trans la tion were generated. Red circle represents the activiation site. (*) is the center of the The vector pointing from nuclear centroid to the center of activation site is used as the zero degree in the polar coordinates.

PAGE 67

67 Figure 3 2 Nuclear movem ent upon local introduction of blebbistatin. (A) Blebbistatin was introduced by micropipette at the trailing edge of the cell. The nucleus moved towards the leading edge, this motion was abolished in cells over expressing GFP KASH. (B) Average movement of nuclei upon local introduction of blebbistatin (control, N=10; KASH, N=10, p <0.01). (C) Epi flourescent image of blebbistatin spray with 4 kDa Fitc Dextran. The micro pippette tip was used to localize blebbistatin to the tail region of the cell.

PAGE 68

68 F igure 3 3 Micromanipulation reveals that the nucleus is under tension between the leading edge and trailing edge. (A) Removal of cell trailing edge results in forward nuclear movement in control cells The forward movement is inhibited in cells transfected with GFP KASH4. Quantification of the forward movement reveals that both the trailing (B) and leading (C) edges of the nucleus traveled further in control cells than in KASH cells Scale Bar (10 m). E rror bars P < 0.01.(D) Nuclear Deformation in response to tail release. The nucleus deformed in control cells (N=8), while remained the same shape in KASH4 transfected cells (N=8).

PAGE 69

69 Figure 3 4 Nuclear movement in motile fibroblasts does not required detachment of the trailing edge. (A) Nucleus moved towards leading edge as the lamellipodia formed.(B) The trailing edge retraction is not necessary for moving the nucleus. Superposition of cell outline at zero minut e (black) and 30 minute (pink). The nucleus moved forward upon the formation of lamellipodia, while the trailing edge did not retract. (C) The nucleus moved more than the cell centroid did in control (n=13) cells. The red line is y=x line. (D) Comparison o f mean movement of the nucleus and cell centroid in 30 minutes shows that they move similar distances. (E) Average displacement of the nucleus and trailing edge in control cells (n=13) in 30 minutes, the trailing edge did not move appreciably, yet the nucl eus moves several microns. Error bars indicate p < 0.01.

PAGE 70

70 Figure 3 5 Directional nuclear translation upon Rac 1 photo activation. (A) Frames of a time lapse recording of a nucleus (red eclipse) moving towards t he photo activation site. Scale bar, 10 m. (B) Kymograph of images corresponding to

PAGE 71

71 the box in (A). Scale bar, 5 m. (C) Examples of trajectories of the nucleus (blue) and the centrosome (red) in photoactivation experiments. (D) Trajectories of nuclear mo vement were overlaid with a common starting point. The zero degree is the direction of activation. The nucleus moved towards the activation site in control and nocodazole treated cells while it did not move much or moved away from the activation sites in b lebbistatin treated or KASH4 transfected cells. The unit is micron. (E) Average directional displacement of the nucleus towards the activation site. Positive value means the nucleus moved towards the activation site. (F) Average trajectories of the nucleus and (G) variance of the nuclear displacment in control photoactivation (black, n=11), nocodazole treated cells (red, n=11), blebbistatin treated cells (blue, n=10) and KASH4 transfected cells (green, n=10) show that there is larger fluctuation in nocodazo le treated cells. (H) Cells are transfected with DsRed CC1 to inhibited dynein. The nucleus appeared to move more that the cell centroid does upon the inhibition of dynein (n=11) in most cells. Red line is y=x line. (I) Average movement of the nucleus an d cell centroid in dynein inhibited cells in 30 minutes. Unlike in control cells, the nucleus moved slightly more than the cell centroid. (J) Average movement of the nucleus and trailing edge in control cells (n=11) in 30 minutes, the trailing edge did not p < 0.01.

PAGE 72

72 Figure 3 6. Kymograph of photo activation experiments. (A) Nocodazole treated cells. (B) Blebbistatin treated cells. (C) GFP KASH4 transfected cells. Scale bars, 10 m.

PAGE 73

73 Figure 3 7. Nuclear retraction due to lamellipodia release. (A) Captured images of cells before and after the release of leading edge lamellipodia. (B) The leading edge of the nucleus retracted more in control cells than that in K ASH4 transfected cells upon the release of cell leading edge lamellipodia.

PAGE 74

74 Figure 3 8 Effect of KASH on trailing edge detachment. (A) Captured images of the trailing edge detachment during forward protrusion of an NIH 3T3 fibroblast. (B) Superposition of cell outline at different time points. The nucleus kept moving towards the leading edg e. (C) Trailing edge detachment frequency is much higher in control (n=24) cells than in KASH4 transfected cells (n=20). (D) Nuclear movement is highly correlated with cell centroid movement in KASH4 (n=11) transfected cells. Most nuclei moved similar dist ance as or less than the cell centroid did. Red line is y=x line. (E) Average movement of the nucleus and cell centroid in KASH4 (D, n=11) transfected cells in 30 minutes also show that they moved similar distances. (G) Average movement of the nucleus and trailing edge in KASH4 transfected cells (n=11) in 30 minutes, the trailing move forward, while it was much less than the nucleus p < 0.01.

PAGE 75

75 CHAPTER 4 E FFECTS OF DYNEIN MOT OR ON MICROTUBULE MECHANIC S AND CENTROSOME CENTERING In order to determine forces on intracellular microtubules, we measured shape changes of individual microtubules following laser severing in bovine capillary endothelial (BCE) cells Surprisingly, regions nea r newly created minus ends increased in curvature following severing, while regions near new microtubule plus ends depolymerized without any observable change in shape. With dynein inhibited, regions near severed minus ends straightened rapidly following s evering. These observations suggest that dynein exerts a pulling force on the microtubule which buckles the newly created minus end. Moreover, the lack of any observable straightening suggests that dynein prevents lateral motion of microtubules. To explain these results, we developed a model for intracellular microtubule mechanics which predicts the enhanced buckling at the minus end of a severed microtubule. Our results show that microtubule shapes reflect a dynamic force balance, in which dynein motor and friction forces dominate elastic forces arising from bending moments. A centrosomal array of microtubules subjected to dynein pulling forces and resisted by dynein friction is predicted to center on the experimentally observed timescale, with or without t he pushing forces derived from microtubule buckling at the cell periphery. This is reproduced from a paper previously published on Molecular Biology of the Cell with permission from the American Soci ety for Cell Biology Modeling and simulations are attributed to Dr. Richard B. Dickinson, Dr. Anthony J.C. Ladd and Dr. Gaurav Misra in the Department of Chemical Engineering, University of Florida.

PAGE 76

76 Materials and Methods Cell C u lture, Plasmids and T ransfection Bovine capillary endothelial (BCE) cells were cultured in DMEM (Mediatech Manassas, VA ) with 10% donor bovine serum ( Gibco Grand Island, NY ), 1% HEPES (Mediatech), 1% GPS (L Glutamine penicillin streptomycin solution, SIGMA ALDRICH ) and bFGF (2 ng/ml, Sigma St. Louis, MO ). For microscopy, cells were cultured on glass bottomed dishes (MatTek Corp Ashland, TX ). These a re stationary interphase cells. They are non confluent and un synchronized. Adenoviral EGFP tubulin was provided by Prof. Donald Ingber. DsRed CC1 plasmid was provided by Prof. Trina Schroer. YFP tubulin was prepared from the MBA 91 AfCS Set of Subcell ular Localization Markers (ATCC Manassas, VA ). Transient transfection of plasmids into BCEs was done with Effectene Transfection Reagent (Qiagen, Valencia, VA ). Time Lapse Imaging and A nalysis Time lapse imaging was performed on a Nikon TE2000 inverted f luorescent microscope with a 60X/1.49NA objective and CCD camera (CoolSNAP, HQ 2 Photometrics, Tucson, AZ temperature, CO 2, and humidity controlled environmental chamber. Images of the centrosome wer e taken every two minutes and imported into MATLAB (Mathwork, Natick, MA ) The positions of the centrosome were calculated to sub pixel resolution using previously published image correlation methods (103) The autocorrelation function (Figure 4 1 ) was calculated from the x and y posi tion fluctuations in seven cells tracked for 1 2 hours.

PAGE 77

77 Cell Shape P atterning Cell shape patterning was done by using the micro contact printing technique described in (105) Molds for the stamps were produced with the UV lithography technique by illuminating a positive p hotoresist through a chrome photomask on which micropatterns were designed (Photo Sciences, Inc. Torrance, CA ). PDMS (Sylgard 184 kit, Dow Corning Midland, MI ) was cast on the resist mold and cured. For micropatterning, the PDMS stamp was treated with 50 (BD Biocoat TM Franklin Lakes, NJ ). The stamp was then dried and placed onto the substrate onto which the cells were plated. Ibidi dishes (Ibidi, Verona, WI) were chosen to be the substrate. After 5 min, the stamp was remov ed and the remaining area was blocked with PLL g Poly ethylene glycol (SuSoS AG Dbendorf, Switzerland ), preventing protein adsorption and cell attachment. After treatment the surface was washed and cells were plated. Laser A blation Laser ablation, first (124) has been recently shown to be a powerful tool for testing mechanical models of force generation in the cytoskeleton (103, 125 127) The method focuses ultrashort pulses of a laser beam on a living microtubule or actomyosin stress fiber (128, 129) A large amount of energy is generated in the focused spot resulting in vaporization of material and severing of the cytoskeleton. The short duration of the pulses (picoseconds or femtoseconds) ensures that t he amount of power (energy/time) is high but the total energy delivered is low, which minimizes damage to the cell due to energy dissipation outside the ablated spot When the energy is focused thro ugh a high resolution objective a single microtubule or a single a ctomyosin bundle can be severed (130, 131) Severing cytoskeletal filaments

PAGE 78

78 instantaneously perturbs the cytoskeletal mechanical force balance, and imaging the subsequent dynamic changes in cytoskeletal configuration provides in vivo quantitative exper imen tal data for testing mechanical models of force generation in the intact cytoskeleton. A number of recent studies have used laser ablation to measure these properties in living cells (103, 125 127, 132) Severing a single stress fiber causes a retraction of the severed edges. This retraction is due to instantaneous dissipation of all tension in the fiber. Inhibiting myosin eliminates the retraction of the severed edges (103, 125) The first observations of this phenomenon indicated that the retraction followed an exponential form such that the rate of translation of the severed edge was high immediately after severing but slowed down and eventually became zero at steady state (125) Using a Kelvin Voight mechanical model for viscoelasticity consisting of a spring and dashpot in parallel, Kumar et al (125) modeled the exponential translation of the severed edge. This work was significant because it demonstrated the use of laser ablation for testing mechanical properties of the cytoskeleton in vivo Kumar et al (125) also showed that that stresses of nearly 700 Pa could be dissipated by severing a single actomyo sin fiber in endothelial cells Thus, the stress fiber is in mechanical continuum with the underlying adhesive substrate. To understand the origin of the exponential retraction of the severed stress fiber ends, Russell et al (103) carried out laser ablation of cells expressing labeled alpha actinin which marks dense bodies and allows the quantification of stress fiber sarcomere lengths. Based on these results, they proposed a simple mechanical model for the stress fi ber sarcomere. Tanner et al (127) found that peripheral SFs have

PAGE 79

79 different viscoelasti c properties from interior fibers; the two sub populations also play distinct mechanical roles. In addition to testing mechanical models of cytoskeletal force generation, laser ablation has been a useful tool for determining the effect of force on intracel lular protein dynamics Lele et al (133) used laser ablation to sever stress fibers, and then performed fluorescence recovery after photobleaching (FRAP) experiments to measure protein exchange dynamics in the associated focal adhesion. Using this method, the authors demonstrated that dissipating stress fiber tension resulted in an increase in the dissociation rate constant of zyxin, but not vinculin, in focal adhesions of endot helial cells. The advantage of using laser ablation in these experiments was that severing a stress fiber instantaneously dissipates tension at the focal adhesion, and FRAP experiments before and after ablation allow reliable comparison of protein exchange dynamics. Another study showed that severing of individual stress fibers caused a loss of zyxin from both SFs and focal adhesions. This study also demonstrated that retraction of the severed stress fiber can actually nucleate new adhesion sites along the retracting fiber (132) T o investigate forces actin g at the junctional netw ork of disc epithelial cells, Farhadifar et al. (134) perturb ed cell monolayers by ablating cell cell junctions with a UV laser beam Measuring the contracting area and distance of the network after laser abalation gave a good estimatio n of parameters in their model which explained the packing geometry of epithelial cells in monolayers. Tinevez et al (135) performed local laser ablation of the cell cortex in L929 fibroblasts to test whether bleb formation is pressure driven. Indeed, a bleb grew from the site of ablation immediately, which

PAGE 80

80 confirmed the hypothesis that bleb growth wa s a direct consequence of cytoplasmic pressure They further investigated the relationship between cortical tension and bleb formation and expansion. In another study by Mayer et al (136) they ablated the actomyosin meshwork in polarizing C. elegans zygote at different positions (regions with or without cortical flow) and in different directions (orthogonal or parallel to the cortical flow). The results showed how cortical fl ow wa s associated with anisotropies in cortical tension and further answered the question ho w actomyosin contractility and cortical tension interact to generate large scale flow Laser ablation studies are increasingly being used in creative ways to answer fundamental questions related to microtubule function. For example, one study used laser ablation to disrupt microtubule interactions with the bud neck in budding yeast (137) This caused mitotic exit suggesting that cytoplasmic microtubules enable the monitoring of the spindle location and mitotic exit in the dividing cell in the event of positioning errors. Another study showed that ablating the centrosome does not inhibit ax on extension and growth, revealing a role for acentrosomal nucleation of microtubules in early neuronal development (138) Laser ablation has also been used to study microtubule dynamics. Colombelli et al (139) developed a new protocol to measure shrinkage rate, growth rate and rescue frequency simultaneously with high temporal and spatial specif icity in live cells. Wakida et al (140) used laser ablation to show that the microtubule dep olymerization rate in living PTK2 cells was location dependent. In our experiments, we used laser ablation to cut a single microtubule. A n inverted (Zeiss Axiovert 200M) laser scanning confocal microscope (LSM 510 NLO) was used in laser ablation experiments with a 63X, 1.4 NA Plan Approchromatic oil immersion lens

PAGE 81

81 (Zeiss) A Ti:Sapphire laser (Chameleon XR, Coherent) was used to sever the m icrotubules as described previously (103, 125, 133) The Ti:Sapphire laser was focused throu gh the objective and scanned over a thin, ~0.14 m, rectangle orthogonally crossing the width of the microtubule for 1 iteration A wavelength of 790 nm was used with a laser head power of 2W, pulse duration of 140 fs and repetition rate of 90 MHz After a blation confocal scans were collected using Zeiss LSM 510 4.2 software at 2 5 s per frame. The root mean square curvature was estimated from microtubule traces by fitting a one dimensional Gaussian approximately orthogonally across the microtubule (82, 14 1, 142) Coordinates were smoothed to eliminate short wavelength measurement error and preserve long wavelength microtubule buckles. Severing produced two microtubule ends with a nearly three fold s difference in their rates of depolyme rization (Figure 4 2 ). From experiments where the plus end was clearly visible, we found that the newly created minus end always depolymerized much slower than the newly created plus end, consistent with previous studies (139, 140, 143) The large difference in depolymerization rates allowed us to clearly identify the severed ends as plus and minus ends. Results Dynamics of Severed M icrotubules A large fraction of the microt ubules in living cells are bent, with stored elastic energy apparently arising either from cytoplasmic motion (144) or from microtubules buckling under continual polymerization against the cell periphery (73, 76) It has been suggested that the bending stresses in the microtubule are balanced by compressive forces propagating from the microtubule tip, reinforced by lateral forces arising from elastic deformation of the surrounding cytomatrix (73) We directly tested this

PAGE 82

82 assumption by severing bent microtubules in BCE cells to remove the longitudinal force at a point along the microtubule length. The two freed ends near the cut behaved differently after severing. Freed plus ends rapidly de polymerized along the original contour of the microtubule (the plus and minus ends were identified as explained in Materials and Methods). On the other hand the more slowly depolymerizing segments near freed minus ends consistently increased in cu rvature a fter severing (Figure 4 3A and Figure 4 4 ), although the increase varied from microtubule to microtubule (Figure 4 5 ). We found no correlation between the initial curvature of the microtubule and the extent to which curvature increased on severing (Figure 4 5 A), nor was there any correlation between the increase in curvature after severing and the distance of the cut from the cell p eriphery (Figure 4 5 B). Previous work has implicated dynein in anterograde transport of microtubule buckles (82) We therefore investigated the role of cytoplasmic dynein in the bending and pinning of severed microtubules. Cells were transfected with a plasmid encoding DsRed CC1 which competitively binds to dynein (107) Dynein inhibition was confirmed by dispersion of the Golgi complex (27) as shown in Figure 4 6 ; microtubules remained anchored to the ce ntrosome (Figure 4 7 ) consistent with previous reports (84) In dynein inhibited cells, segments near a free minus end did not show any increase in curvature following severing; instead, the microtubules straightened rapidly, on timescales of the order of a few seconds (Fig ure 4 3B ). An increase in curvature in normal cells and a decrease in curvature in dynein inhibited cells was consistently observed for severed minus end mic rotubules (bar graph in Figure 4 3 B).

PAGE 83

83 The dynein dependent increase in curvature of severed minus ends suggests that motor forces, directed toward the plus end, pull along the length of the microtubule. Because the increase in curvature after severing was consistently observed (Figure 4 5 A) and found to be independent of the distance f rom the p eriphery (Figure 4 5 B), we surmise that microtubules are under tension along most of their length (although tips near the cell periphery are likely to be under compression due to polymerization forces (73) How might dynein generate pulling forces along a microtubule? We hypothesize that dynein molecules linking the cytomatrix to the microtubules along their lengths pull on microtubules as they walk toward the minus end. The experimental observations also show that dynein contributes a significant fricti onal resistance to the motion of microtubules, because in normal cells we see no evidence of straightening, whereas in dynein inhibited cells microtubules straighten on timesc ales of a few seconds This observation could be explained by the transient natur e of the cytomatrix dynein microtubule linkage leading to protein friction opposing microtubule motion. Model for D ynein G enerated M icrotubule F orces We next formulated a model for dynein force generation which considers the average behavior of an ensemble of motors, transiently linking microtubules to the cytomatrix Once a cytomatrix linked motor binds to the microtubule, it exerts a force along the local tangent to the microtubule as it walks towards the minu s end (see schematic in Figure 4 8 A); the moto r also exerts a force in response to motion of the attachment point. Assuming dynein cytomatrix linkages dissociate with first order kinetics, the ensemble averaged force density (force per unit length) along the microtubule is then given by

PAGE 84

84 ( 4 1) where m 1 is the density of dynein cytomatrix linkages (number of linkages per unit length), is the average force per linkage on a stationary microtubule (~8pN) and is the speed of the force free motor (0.8 m s 1 ; Table 4 1 Model parameters). On average, a motor linked to the cytomatrix drives the microtubule in the direction of the local tangent, to compensate for the displacement of the motor towards the minus end. Motion in the transverse direction is l imited by the frictional resistance where the motor friction is the quotient of the stiffness of the dynein linkage, and the dynein dissociation rate, k off The dynein friction coefficient was chosen by matching the times scales for individual microtubule motion found in simulations (see next section) with experimental data (e.g. Figure 4 3A and Figure 4 4 ). Further details of the determination of the motor fr iction are given i n Appendix E Simulations of microtubule buckling. The value of that best matches the experimental time scales ( 10 3 pN m 1 s) can be obtained by taking the stiffness of the dynein linkage in the range 0.1 1pN nm 1 (52) and the dissociation rate k off 0.1 1s 1 Our estimate of the dynein dissociation rate is consistent with observations of long lived binding between dynein and microtubules (115) and with measurements of dynein exchange rates by photobleaching (145) The timescales for microtubule motion are insensitive to the choice of dynein density, since individual motors contribute equally to the force and the friction.

PAGE 85

85 Simulations of Mi crotubule Buckling D ynamics We next investigated whether dynein mediated forces can explain the buckling of severed microtubules. Approximating the microtubule as an elastic slender body, the force and torque balances on the microtubule (146) are (4 2) where and are the force and bending moment at contour position The motor force is given by Equation 4 1 and the bending moment by where is the flexural rigidity. Representing the microtubule as disc rete segments (Appendix E Simulation methods) and solving for the velocity at each point along the contour length we can track the motio n of a motor driven microtubule of a constant length. Our model for dynein force generation envisages a number of motors distributed uniformly along the microtubule. For clarity and simplicity we have replaced a number of individual motors with a uniform ( tangential) force and friction along the microtubule, which follows from a time (or ensemble) average over the positions of the individual motors. Simulations based on this model explain how dynein can increase the curvature of a newly created minus end. A s illust rated in the example in Figure 4 8 B, following severing, small initial bends in the microtubule are predicted to be amplified due to the pulling forces generated by the dynein motors along the microtubule length. In addition, this buckling of the n ewly created minus ended segment is predicted to occur on time scales comparable to the exper imental observations (Figure 4 3A and Figure 4 1 ).

PAGE 86

86 Centrosome C entering by M otor D riven M icrotubules The experimental and theoretical evidence presented here is c onsistent with the hypothesis that microtubules in living cells are pulled by dynein motors distributed along their length (80, 82, 147) We experimentally determined whether dynein dependent forces are essential for centering the centrosome in endothelial cells. To do this, cells were patterned into square shapes using microcontact printing, and the position of the centrosome was determined by imaging EGFP tubulin expressing cells. The centrosome was observed to be at or close to the center in normal cells, while it was substantially off center in dynein inhibited cells (Figure 4 9 A). Data from 22 different cells consistently show that the centrosome is off center in dynein inhibited cells (Figure 4 9 B). These results support the hypothesis that dynein is necessary for centrosome centering and are therefore consiste nt with a pulling mechanism (78, 80, 82, 147) Simulations of centrosome centering allowed us to compare the dynamics of an array of microtubules with and without dynein forces. Microtubules were randomly nucleated at the centrosome and allowed to grow and disassemble by dynamic instability We account for growth and shrinkage under dynamic instability (148) using the experimentally measured parameters for the polymerization velocity, depolymerization velocity as well as the rates for switching to catastrophe, and rescue, (116, 117) More details of the simulation methods ar e available in Appendix E Simulation methods. One important experimental obser vation is that although microtubules in living cells are frequently observed to pin and buckle at the cell periphery, they sometimes continue to grow by sliding along the cell boun dary (Figure 4 10 ). We incorporated both

PAGE 87

87 possibilities into a stochastic mod el of contact between the tip of a microtubule and the cell boundary, which was able to capture the main features of the pinnin g and sliding (Appendix E Simulation methods). Figure 4 9 C shows configurations of motor driven microtubules in a 40 m square cell. Some of the microtubules near the periphery buckle int o small wavelength (see Figure 4 9 C, t = 1min), consistent with experimental observations (73) The short wavelength buckles may not be immediately obvious in ( Figure 4 9 C ) because the eye is naturally drawn to the longer wavelengths in the body of the cell. Nevertheless an examination of the border regions, particularl y near the corners, shows that the motor driven microtubules have short wavelength (2 3 m) buckles (Figure 4 9 C) whereas without motors there are only long wavelength (> 10 m) buckles (Figure 4 9 D). Motor driven microtubules are predicted to drive an off center centrosome towards the center with a time constant of the order of 10 min (Figure 4 11 ), similar to the time constant measured in living cells (Appendix E Simulations of centrosome centering). Simulations also show that in the absence of motors ( Figure 4 9D ) the centrosome remains essentially fixed in place for at least 100 minutes (the duration of the simulation). This is consistent with experimental observations in dynein inhibited cells (Figure 4 9 B) where the centrosome remains off center. To determine which forces are most important for centrosome centering, we formulated a simplified model assuming that the microtubu les are rigid (Appendix E Centrosome relaxation time) The predicted time scale for centering (~24 min) remains comparable to experimental observations. Because the model only includes dynein tension and friction, the time scale is predicted to be independent of the number of

PAGE 88

88 microtubules and the density of active dynein motors along the microtubules. Therefore, a balance of tens ile and frictional forces from the dynein motors is sufficient to explain centrosome centering on timescales comparable to experimental observations. Pushing forces have been observed to cause centering in vitro (77) but not under physiological conditions. In particular, the viscosity of the buffer solution in these experiments was much smaller than the effective viscosity of the cellular fluid, as inferred from the cutting experim ents described earlier (Figure 4 3 ). In addition the fabricated cells were smaller (12 m) than the endothelial cells used in our in vivo experiments (~40 m). To see if we could resolve these apparent contradictions, we simulated the condi tions described in (77) ; a square cell of length (12 m), a low viscosity background fluid (water), and microtubules that were allowed to slip along the cell surfac e to mimic the smoothness of the glass walls. In addition, we adjusted the polymerization kinetics to allow for different steady state lengths of microtubules. Our simulations showed the same behavior as the experiments; with short microtubules, an initial ly off center centrosome moved towards the cel l center In the small chamber, shorter microtubules generate a larger buckling force than the longer microtubules in the living cells (roughly 10 fold, since buckling forces scale with L 2 ). This, combined wit h the lower viscous drag of the fluid, is sufficient for the centrosome to center. However if the polymerization kinetics were adjusted to create longer microtubules, then th e centrosome drifts off center as observed experimentally (77) Thus our simulations explain a number of apparently contradictory experimental observations in terms of the relative magnitudes of compressive, t ensile and frictional forces. A CENT ROSOME can be centered by pushing forces under conditions where microtubules are short and the

PAGE 89

89 viscosity of the fluid is similar to water, but not in animal cells where the microtubules are longer and the friction is orders of magnitude larger. However, te nsile motor forces can center the centrosome in capillary endothelial cells on the experimentally observed timescale. Further details c an be found in Appendix E Simulation of centrosome centering. Discussion The results of our investigation give new insi ght into the role of dynein in the force balance on microtubules. If individual microtubules were under compressive stress along their length, severing would result in a straightening of the microtubule, but we observed that the curvature generally increas ed near a newly created minus end. Conversely, when dynein activity was inhibited, microtubules straightened after severing. Severed plus end segments in normal cells are observed to depolymerize along the original contour. Simulations suggest that microtu bules pinned at their minus end do eventually straighten, but the timescale for the motors to push buckles towards the free plus end (~10 s) is significantly longer than the depolymerization time. These observations are consistent with dynein motor forces pulling the microtubule segments towards the cell periphery. The fact that severed microtubules in normal cells are never observed to straighten indicates that dynein contributes a large frictional resistance to lateral motion in addition to its directiona l force; this is confirmed by the rapid straightening of severed microtubules in dynein inhibited cells. We have developed a model for dynein force generation to explain the directional and frictional forces suggested by the experimental observations summ arized in the previous paragraph. In this model, frictional forces arise from the binding and dissociation of dynein motors linking microtubules to the cytomatrix, while the tangential

PAGE 90

90 forces come from the motor activity of cytomatrix bound dynein. The mod el makes several qualitative and quantitative predictions about the nature of the force balance on the microtubule. If a microtubule is pinned at its minus end at the centrosome, pulling forces from the dynein motors lead to a tensile stress along its leng th, with the maximum tension at the minus end (centrosome) If the microtubule has not reached the cell periphery, the stress at the free plus end must vanish. However when its tip impinges on the cell periphery, a compressive force develops to accommodate the excess length of the polymerizing microtubule. The microtubule is then in a state of compression near the tip but in tension near the centrosome; the balance of forces is taken up by cytoskeletal bound dynein. The model predicts that the dynein pullin g force balances the tip compressive force at a distance d = F p / f 0 (neglecting bending forces relative to dynein forces) from the tip, where F p is the compressive force at the tip, and f 0 ~16 pN/ m is the dynein force per unit length. The polymerization force is unknown in vivo but in vitro it has been estimated to be ~10pN, (149) for which d = 0.6 m using our estimate of f 0 Even for larger polymerization forces o f tens of pN, the compressive force can be balanced by only a few motors near the microtubule tip and the transition from tension to compression is predicted to occur within just a few microns. Therefore a key prediction of the model is that most of the mi crotubule is under tension despite large compressive forces at the tip This would explain the experimentally observed increase in curvature of the minus end upon severing indicating tension independent of distance from the periphery The model also helps explain what happens upon severing. Once the tension is released at the cut point, the small longitudinal extension quickly relaxes, and the force

PAGE 91

91 generated from dynein motors along the segment near the newly created minus end quickly becomes compressive This compressive force causes this segment to increase in curvature (i.e. buckle). The model predicts an increase in curvature for minus ended segments even when the plus end is not impinging on the cell boundary. This is due to the large translational res istance from protein friction, which effectively immobilizes the microtubule segments far from the cut point. Simulations predict that dynein motors pulling on a radial array of microtubules can center the centrosome in vivo consistent with our observati on that the centrosome is off center in dynein inhibited cells Moreover, the time scale for centrosome centering, calculated with the same motor parameters as in the buckling simulations, is consistent with experimental measurements. Our model predicts th at the centering time is insensitive to microtubule number and dynein density. Our simulations also explain why centering by microtubule pushing can occur in vitro without dynein (77) but only under conditions (low friction and short microtubules) where buckling does not occur. Fission yeast, for example, appears to present conditions that favor pulling during meoitic prophase but pushing during interphase, by varying the cortical localization o f dynein (149 154) However, under conditions found in animal cel ls with longer microtubules and much higher frictional resistance, both simulations and experiments suggest that compressive buckling forces are insufficient to center the centrosome. In consistence with the results discussed above, the same laser ablatio n experiment done with microtubules in U2OS ( human osteosarcoma ) also showed the same phenomenon (Figure 4 12). Thus, our model for dynein motor forces quantitatively ties the mechanics and dynamics of individual microtubules to the centering mechanism of the centrosome.

PAGE 92

92 Although our results show that dynein is the dominant contributor to the lateral friction and that dynein forces are sufficient to buckle freed minus ends, we do not exclude the possibility of other motors playing a role in microtubule forc e generation. For example, it is possible that plus end directed kinesin motors are simultaneously pulling in a direction opposite to the dynein forces If this force is significant relative to dynein pulling, then the density of dynein linkages would be l arger than our current estimate, but the qualitative predictions of the model would be unchanged Our findings have several important implications for the role of microtubules in cell mechanics. A microtubule in a living cell cannot be described by a stat ic force balance, because of the significant contribution of frictional forces from moving segments. The tangential and frictional forces generated by dynein motors dominate the elastic stresses from microtubule bending and cytomatrix deformation. Although numerical results indicate that a typical microtubule is under tension along most of its length, we do not rule out a compressive force at a microtubule tip impinging on the cell periphery. Dynein motors may in fact confine the compressive stresses to the region near the tip, consistent with a picture in which microtubule compressive forces at the periphery are transmitted to the actin cytoskeleton (83) but through transient dynein cytomatrix linkages rather than by elastic deformation (73) Our model for dynein force generation provides a unifying explanation for the shapes of individual microtubules in the cell and how these shapes are consistent with tension driven centering of the centrosome by a radial array of microtubules. Summary of Findings By using femtosecon d laser ablation, we severed bent single microtubules in living cells. The behavior of buckled microtubule after severing gave us a direct readout of

PAGE 93

93 forces on microtubules. Dynein was found to pull microtubules towards the plus end. A model for dynein for ce generation was proposed, and simulations of centrosome centering were conducted using the model. Our model for dynein force generation provides a unifying explanation for the shapes of individual microtubules in the cell and how these shapes are consist ent with tension driven centering of the centrosome by a radial array of microtubules.

PAGE 94

94 Figure 4 1. A comparison of centrosome centering (simulation) with the autocorrelation function of fluctuations in centrosome position (experiment). The microtubule network develops during the first 5 min of the simulation; subsequently the centrosome centers with a relaxation time of 12 min. Experimental measurements of the autocorrelation function of the centrosome position decay in about 8 min. The nega tive region in the autocorrelation function reflects insufficient data to obtain an accurate measurement of the mean position in each cell.

PAGE 95

95 Figure 4 2. Deploymerization rates of microtubules severed by laser ablation. The bar graphs show that depolymerization rate of the plus end is consistently larger than that of the minus end.

PAGE 96

96 Figure 4 3 Minus end microtubules underneath the nucleus increase in bending after laser severing. Representative images highlighting changes in shape after se vering a single microtubule in living cells. (A) Increased bending of minus ended microtubules after severing near the nucleus. The black arrowhead indicates the position of the cut and the severed microtubule is highlighted by small crosses. Microtubule s hapes in the images were measured in MATLAB (see plots of severed microtubule shapes; the newly created plus and minus ends are indicated) and the root mean square (RMS) curvature was calculated. The minus end microtubule (recognized as minus ended from th e lack of significant depolymerization compared to the newly exposed plus end) showed a 7 fold increase in mean curvature over the visible segment length. However, the plus end segment depolymerized but showed no measureable change in curvature. These obse rvations are consistent with the hypothesis of minus end directed motors pulling on the micr otubule. (B) Straightening of a bent microtubule in a d ynein inhibited cell The white arrow indicates the plus end of the microtubule, and the arrowhead tracks th e severed end. The microtubule straightens significantly on timescales of a few seconds,

PAGE 97

97 supporting the hypothesis that there is an additional frictional force contributed by dynein. The plot compares the change in RMS curvature near severed minus ends in control and dynein inhibited cells. Data is from at least 10 cells for each condition; the statistical significance is at p < 0.01 Error bars indicate standard error of the mean (SE M ) The pooled data strongly supports the hypothesis that dynein promotes bending of microtubules in living cells, and that in the absence of dynein, microtubules straighten upon severing. Scale bars are 2 m.

PAGE 98

98 Figure 4 4. Minus end microtubules at the cell periphery increase in bending after laser severing. A representative example of experiments where a single microtubule, near the cell periphery (white arrow), was severed. Images show increased bending of minus ended microtubules after severing (cut at black arrow) The plus end depolymerized at a rate of 0.6 07 m/s, while the minus end depolymerized at a rate of 0.164 m/s. Note that the plus end depolymerizes but does not show a change in curvature. Scale bar is 1 m. Figure 4 5. Change in curvature after severing is not correlated with the initial curva ture, nor with the spatial location of the cut. (A) The maximum change in RMS curvature for 18 experiments is plotted against the initial curvature. In all experiments that were analyzed, the change in curvature was positive and no correlation with initial curvature was found. (B) The maximum change in RMS curvature is plotted against the shortest distance from the cut to the cell periphery. No significant correlation was observed suggesting that the increase in curvature is not location dependent. In three cells the cuts were positioned under the nucleus and the distance to the periphery could not be determined. r is correlation coefficient.

PAGE 99

99 Figure 4 6. Inhibition of dynein causes dispersion of the Golgi complex. BCE cells were transfected with DsRed C C1 to inhibit dynein. The cells were then fixed and immunostained with mouse monoclonal Golgi marker (Abcam) and Hoechst 33342. Control cells (left) show a compact Golgi complex (green) near the nucleus (blue), while DsRed CC1 transfected cells (right) sho w a dispersed Golgi complex. Scale bars are 10m. Figure 4 7. Microtubules radiate from the centrosome in dynein inhibited cells. BCE cells were transfected with DsRed CC1 and infected with adenoviral GFP tubulin, and fixed and immunostained with Hoechst 33342 (blue, nucleus).

PAGE 100

100 Figure 4 8 Simulations predict dynein induced buckling of microtubules. (A) Cartoon of a dynein motor indicating how the minus directed motor bound to the cytomatrix exerts a force towards the microtubule plus end. (B) Simulations of an elastic filament based on the force balance in Equation 4 1 and 4 2 predict dynein induced buckling near the minus end of a severed microtubule. The timescale can be comp ared with t he experimental data in Figure 4 3 A.

PAGE 101

101 Figure 4 9 Dynein forces are sufficient to center the centrosome. (A) Representative images showing the centrosome in EGFP tubulin expressing square endothelial cells; control cell (left) and dynei n inhibited cell transfected with DsRed CC1 (right). Scale bar is 5 m. (B) Mean centrosome position in 42 control cells and 22 dynein inhibited cells; the statistical significance p < 0.01. Er ror bars indicate SEM. The centrosome is consistently observed t o be at or close to the center of the square in control cells, while it is substantially off center in dynein inhibited cells. Simulations of centrosome centering in square cells with (C) and wit hout (D) dynein motor activity The motor driven microtubules (C) show considerable buckling near the cell periphery (clearly visible at t = 1 min) whereas without motor activity (D) the buckling is of Euler type. Only the motor driven microtubules are observed to center an initially off center centrosome.

PAGE 102

102 Figur e 4 10. Microtubules undergo three distinct behaviors upon reaching the periphery: (1) buckling with the tip immobilized; (2) sliding along the cell periphery; (3) no growth, eventually depolymerizing. (A) Two examples of microtubules that slide along the cell periphery (dotted white line) (B) Four microtubules (colored arrows) that have reached the cell periphery and remain until they depolymerize One microtubule (green) does not buckle significantly, suggesting that it has stopped growing Scale bars are 2.5 m.

PAGE 103

103 Figure 4 11 Simulations of centrosome centering. The circles show simulated displacements of the centrosome, X ( t ), which centers on a time scale of the order of 10 mins and thereafter oscillates about the central position. The squares s how the same initial condition but with the motor forces and friction turned off. Here the microtubules experience a much reduced (100 fold) drag force from the cellular cytoplasm (as inferred from the laser severing experiments in dynein inhibited cells) but the polymerization forces are unable to push the centrosome towards the center on experimentally relevant timescales.

PAGE 104

104 Figure 4 12 Minus end microtubules underneath the nucleus does not traighten after laser severing. U2OS cells expressing RFP tu bulin was used. A representative example of experiments where a single microt ubule under neath the nucleus was severed. Images show no straightening as the minus ended microtubules after severing slowly depolymerized (cut at black arrow) Scale bar is 2 m

PAGE 105

105 Table 4 1. Microtubule Mechanics and Centrosome Centering Model Parameters Symbol Parameter Range Source Value Used f max Maximum dynein force 5 8 pN (115) 8 pN v 0 Dynein speed (no force) 0.8 m s 1 (109) 0.8 m s 1 Dynein spring constant 0.1 1 pN nm 1 (52, 155) 1 pN nm 1 k off Dynein nucleus off rate No value Filament buckling 1 s 1 Dynein density (#/length) No value Filament buckling 2 m 1 N Number of microtubules 200 500 (116) 100 v pol MT polymerization speed 0.1 0.2 m s 1 (116, 117) 0.1 m s 1 v depol Depolymerization speed 0.2 0.3 m s 1 (116, 117) 0.3 m s 1 k cat Catastrophe rate constant 0.01 0.06 s 1 (116, 117) 0.05 s 1 k rec Recovery rate constant 0.04 0.2 s 1 (116, 117) 0.2 s 1 Effective friction Filament relaxation 10 Pa s

PAGE 106

106 CHAPTER 5 MODELING OF FLUORES C ENCE RECOVERY AFTER PHOTOBLEACHING The analysis of Fluorescence Recovery After Photobleaching (FRAP) experiments requires mathematical modeling of the fluorescence recove ry process. An important issue of FRAP experiments that tends to be ignored in the modeling is that there can be a significant loss of fluorescence due to bleaching dur ing image capture. In this chap ter we explicitly include the effects of bleaching during image capture in the model for the recovery process, instead of correcting for the effects of bleaching using reference measurements. Using experimental examples, we demonstrate the usefulness of s uch an approach in FRAP analysis. The models discussed here can help bring great clarity into the interpretation of FRAP experiments. Materials and Methods Cell C ulture, P lasmids and T ransfection NIH 3T3 fibroblasts were cultured in DMEM (Mediatech Manassas, VA ) with 10% donor bovine serum (Gibc o Grand Island, NY ) For microscopy, cells were cultured on glass bottomed dishes ( WPI Sarasota FL ( BD Biocoat TM Franklin Lakes, NJ ) E GFP VASP was used in transfect cells Transient transfection of plasmids into NIH 3T3 fibroblasts was performed with Lipofactamine TM 2000 t ransfection r eagent (Life Technologies, Invitrogen, Carlsbad, CA ). Confoca l Microscopy and FRAP Cells expressing EGFP VASP was imaged on a Leica SP5 DM6000 confocal microscope equipped with a 63 X oil immersion objective. FRAP analysis was carried out by selecting the focal adhesion using the ROI (region of interests) function. A 488

PAGE 107

107 nm Argon laser was applied at 50% power to blea ch the focal adhesion adhesion for five times All images were taken at 512X512 resolution with approximately four times zoom in. For calculating immobile fraction, three images were taken after the bleaching using 4% laser power, the time interval was 40 seconds (80 seconds total). For FRAP without free protein bleaching effects, images were taken every 1.3 seconds (minimum intervals) for 60 frames using 4% laser power. For FRAP with free protein bleaching effects, i mages were taken every 1.3 seconds using 20% laser power During 2 and humidity controlled environmental chamber. Results Modeling B leaching D uring I mage C apture. We first consider the situation where fluorescence imaging is perform ed on a live cell. If an image is captured for an exposure time then the fluorescence concentration in the cell will decrease from an initial value of in this time according to the kinetic expression (156, 157) ( 5 1) where is the photobleaching rate constant (s 1 ). The precise value of will depend on imaging conditions (i.e. laser power, magnification, exposure time etc). At the end of the exposure time the concentration is Consider an experiment involving imaging of the entire cell over frames with a time interval between frames of The image capture is assumed to occur in the time interval Then applying

PAGE 108

108 Equation 5 1 for imaging at each frame, the formula for the concentration at the end of the time interval is ( 5 2) where The time evolution of the concentration predicted by equation (2) for a hypothet ical experiment is shown in Figure 5 1A. Because the imaging occurs over a time interval the measured fluorescence in the image is proportional not to but rather to the average concentration over given by where (the measured fluorescence). However, as is common practice, the fluorescence in subsequent images is normalized to the fluorescence in the first image ( ) and so the factor cancels, making the normalized fluoresc ence proportional to the ratio of concentrations Figure 5 1B illustrates how normalization scales the hypothetical data from Figure 5 1A. Noting this requirement for normalization, the normalized fluorescence in a whole cell imaging experiment obeys the equation (5 3) Equation 5 3 allows the straightforward estimation of (assuming is known). Alternatively one could capture one frame for a long enough causing significant bleaching due to image capture; this suffers from potential heating artifacts though and may not be as reliable as the procedure suggested by Equation 5 3.

PAGE 109

109 FRAP Model to A cc ount f or Photobleaching due to Image C apture When the FRAP experiment involves the selective photobleaching of bound molecules (such as molecules bound to a microtubule tip (158) or in a focal adhesio n (133, 159, 160) or at a promoter array (91, 161) ) the recovery occurs through diffusive transport of free protein molecules (in the cytoplasm, nucleoplasm or membrane) followed by exchange with bound molecules. A commonly encountered situation is where the exchange between bo und and free protein is far slower than diffusive transport into the photobleached spot and the concentration of the free protein is unaffected by the exchange process owing to the large pool of free molecules compared to bound molecules (16 0, 162) In the following discussion we develop ed the modeling approach for this situation (the approach is generally applicable as discussed later). We consider first the situation where bleaching during image capture is not significant. The equation describing the recovery process is (assuming the free concentration is well mixed and constant, and that diffusion is very fast compared to binding) ; ( 5 4 ) Where is the rate constant for binding, is the binding site concentration, is the cytoplasmic (or membranous) diffusing concentration and is the bound concentration in the photobleached spot. The initial condition reflects the fact that the photobleaching pulse reduces bound fluorescence from an initial concentration of to The solution to this equation is

PAGE 110

110 ( 5 5 ) The typical approach in the literature is to normalize the experimental data as and then fit it to The parameter estimated from the data is However, if there is bleaching during the image capture process itself, then as illustrated in Figure 5 1C t he dotted lines are the actual dynamics consisting of (unobserved) recovery interspersed by bleaching during image capture l eading to the measured recov ery (indicated by the (*) ). It is necessary then to model the unobserved dynamics, consisting of recovery between time intervals of image capture and also the bleaching due to the image capture process itself to predict the obse rved recovery dynamics. Fitting such a model to the data has the advantage of faithfully capturing the recovery process, and eliminating the need for arbitrary corrections to the data (such as correcting the recovery signal by dividing with the total cell intensity which decays due to bleaching during image capture). We now consider the time evolution of the fluorescent bound concentration under the effects of bleaching due to image capture. Consider three images: one taken just before the photobleaching p ulse corresponding to a final concentration of refers to the fluorescent bound molecule concentration at the end of the image capture), the second image immediately after the photobleaching pulse corresponding to a final conc entration of and the third image whose capture begins at a time interval of where is the time interval between successive images (similar to the logic used

PAGE 111

111 in the whole cell experiment before). T he fluorescent bound concentration just before the third image capture begins is Because ( ~milli s econds and ~second), we can approximate in Eq. 5 yielding When imaging starts, photobleaching occurs due to image capture, and the concentration at is ; as before. Here we have made the assumption that the recovery itself occurs to a negligible extent in the time interval compared to fluorescence decay due to bleaching during image capture. This is reasonable considering that and the recovery time scale is much la rger than The exchange process in the next time interval is still described by the differential equation in Eq uation 5 but now with an initial condition Extending this logic to the frame, it is possible to calculate the concentration at the end of the image capture as shown below (we note again that indicates the concentration at the end of image capture just before the photobleaching pulse, indicates the concentration at the end of image capture just after the photobleaching pulse): (5 6) The concentrations are all normalized to the concentration at the end of the image capture just before the photobleaching pulse. As before, the ratio of the measured fluorescence in the image to the fluorescence in the first image is proportional to (and should be fit to) the concentration ratio derived in Equation 6.

PAGE 112

112 An interesting point here is that the model predicts a steady state for the fluorescence recovery despite the fact that image capture results in periodic bleaching. Such a steady state will be reached when the fluorescence lost due to bleaching due to image capture is exactly balanced by recovery in between time frames. This yields the equality ( N represents any image collected in the steady state portion of the recovery curve) (5 7) At steady state, if the fluorescence intensity is known, and the bleaching function determined from experiment, it is possible to calculate using Equation 5 7. This of course requires prior knowledge of the model that describes protein exchange in the spot (in this case, Equation 5 4). Accounting for Bleaching of Free P rotein Equation 5 6 describe FRAP recovery when photobleaching during image capture is significant. In derivin g these equations, we have assumed that the free protein concentration is unaltered by the bleaching. This assumption is typically valid if we consider that the bleaching during image capture occurs predominantly at the focal plane (wh ere the laser beam is focused in a confocal microscope) and progressively less outside the bleached spot. As the space where the free molecules diffuse is well mixed on time scales of exchange with bound protein (this is the assumption underlying Equation 5 4 ), it is reasonable to expect that the free concentration of molecules will decrease much less due to image capture than proteins present in a bound spot enclosed in the thickness of the focal plane (such as a focal adhesion or a receptor

PAGE 113

113 binding to a p romoter array in the nucleus; this is discussed more in the APPENDIX E ). The assumption that the free protein is not changing in concentration due to image capture can also be checked by measuring the fluorescence of free molecules as demonstrated in the e xperimental example later. When the free molecules are also bleached during image capture, we let and be the bleaching functions for bound and free proteins (corresponding to different ; see the APPENDIX E ). Then, the free c oncentration decreases similar to Equation 5 3 (5 8 ) We continue to make the assumption that the free concentration is well mixed, and unaffected by the exchange process itself with bound protein because of the large pool of free molecules compared to bound molecules. Using Equation 5 8 with Equation 5 4 for the unobserved concentration between successive frames and accounting for bleaching, the bound concentration is ( 5 9 ) FRAP Model to Account for an Immobile F raction As discussed above, assuming that the free protein pool is unaffect ed by the imaging process, the recovery should reach a steady state. This model, however, assumed that all of the molecules in the bleached spot were able to exchange with the cytoplasmic pool of molecules on a single time scale(~ ). In many experimental

PAGE 114

114 situations, it is observed that the recovery is not complete, suggesting the presence of bleaching of the free pool) is (here is th e immobile fraction) (10A ) (10B ) H ere is the immobile fraction. is the total concentration (= still represents the fraction of fluorescent bound molecules bleached. The contribution of mobile and immobile pools to the recovery need to be separately accounted for as shown in Equation 10A and 10B (the superscript M refers to the mobile fraction, IM r efers to the immobile fraction). The fluorescence intensity in a FRAP experiment normalized to the initial fluorescence just before the photobleaching pulse should be fit to Calculations of Normalized Recovery: the B ehavior of Eq uat ion 5 6A and 6B We explored the behavior of Eq. 5 6A and 6B numerically. As seen from Equation 5 6 the recovery process depends on the parameter group the parameter and the parameter group Fixing (a typical value for bleaching in experiments) and assuming = 0.2 solutions to Equation 5 6 are plotted for different values of ( Figure 5 2). Because is kept constant (= ), the value of can be thought of

PAGE 115

115 as constant in Figure 5 2 (although its actual value is not relevant since the solution depends on the parameter group ). Figure 5 2 shows that when too many frames are collected over the characteristic time scale of the recovery process (i.e. there is a significant decrease in net recovery owing to bleaching during image capture. This is because bleaching during image capture occurs too frequently such that significant recovery due to protein exchange between successive image captures does not occur resulting in a steady state with low recovery. The extent of recovery increases with increases The photobleaching process during image capture itself can create an erroneous impression Also shown in Figure 5 2 Effect of the Immobile F raction on FRAP R ecovery Without Bleaching of Free P rotein) Figure 5 3A shows calculations of recovery in the presence of an immobile fraction. The parameter values are identical to Figure 5 2, but solutions to Equation 10A and 10B are plotted al ong with 30 % immobile fraction i.e. As seen, the recovery does not reach a steady state in comparable frame numbers (compare with Figure 5 2). Unlike the results in Figure 5 2, Figure 5 3A shows the presence of a peak in intensity s uch that the fluorescence intensity initially increases but then decreases. This is due to the fact that the immobile fraction (which by definition cannot exchange during the recovery process) continues to get bleached during the imaging process (as indica ted by the decaying grey line in Figure 5 3B). Also, due to the bleaching of the immobile fraction, there can also be parameter conditions where the total intensity decays instead

PAGE 116

116 of recovering (see decaying curve in Figure 5 3A). For Equation 10A and 10B, a steady state is only achieved when the immobile fraction is completely bleached. Figure 5 3C shows the effect of the immobile fraction itself on the recovery curve. With an increasing immobile fraction, the recovery transients show a pronounced decay, a nd for high enough values, the recovery falls below the initial fluorescence value. Figure 5 3D shows the effect of on the recovery process; the extent of photobleaching during image capture again significantly decreases the net recov ery; a maximum in fluorescence intensity is predicted Anal ysis of Focal Adhesion Protein E xchange As an example of the application of the model above, we performed FRAP analysis of the focal adhesion protein GFP VASP. As we have shown before (133) the recovery curves in the case of focal adhesion proteins yield the parameter We first measured the immobile fraction in the chosen adhesion ( Figure 5 4) by performing an initial bleach, capturing a single image immediately after bleach and a second image ~80 seconds later (when the recovery transients were determined to reach a steady state ). The immobile fraction was calculated from the formula and was found to be 0.0263 (solid circles in Figure 5 4B show the normalized concentrations immediately after bleach and after recovery ). Next, FRAP analysis was performed on the same focal adhesion in which was measured above The unknown parameters (as seen from Equation 5 10) are and First, a FRAP experiment was performed such that relatively little photobleaching occurred during image capture. The intensity of the free protein was

PAGE 117

117 approximately constant ( Figure 5 4D). The normalized recovery was fit to a model (Figure 5 4B); as seen t he model satisfactorily captures the recovery and the subsequent, slight decline in the fluorescence recovery. The recovery is substantially less than the recovery observed in the experiment to calculate the immobile fraction above, suggesting an effect of photobleaching due to image capture on the bound fluorescence in the focal adhesion. If this data was used to erroneously calculate the immobile fraction, it would yield a value of a significantly different value than the actual val ue determined above. Next the experiment was repeated in the same focal adhesion but under higher excitation laser intensities to induce more photobleaching of the cytoplasmic pool. The fluorescence recovery occurred with a marked decrease in the intensity at later times. The model again was able to describe the decrease in the intensities, and parameters could be estimated. Importantly, the determined from the two different experiments matched very well (0.17 s 1 versus 0.18 s 1 ). Also the value determined from fitting in Figure 5 4E is close to that from fitting in Figure 5 4C (0.012 versus 0.009 ). This suggests that the model is able to estimate the kinetics of dissociation accurately despite the effects of photobleaching during image capture. When the data in Fig. 4B was fit to the conventional model the value of was found to be 0.3, a clear difference in the value obtained from the above model (Figure 5 5). Discussion The analysis of FRAP experiments is an ongoing area of research. Among the many complicating factors (88) the effect of photobleaching by the image capture itself

PAGE 118

118 has not received much attention. In this paper, we propose an approach to account for this by explicitly including photobleaching into the modeling of the fluorescence recovery process. The method involves modeling the unobserved dynamics (which by definit ion are unaffected by photobleaching), and modeling the photobleaching during the period of observation. As the observation occurs at discrete time intervals (i.e. images collected at discrete time intervals), the photobleaching is modeled to occur at disc rete time intervals superimposed on the unobserved dynamics that occur continuously. A simple conclusion from the modeling is that the immobile fraction should not be calculated from the FRAP curve itself (as is common practice). Instead, the number of fr ames should be minimized, preferably to only three frames: one before bleach, one immediately after, and one when the recovery reaches a steady state (the characteristic time scale for the steady state can be established from separate FRAP experiments). As seen in Figure 5 4B, the value of would be 0.4838 if calculated from the FRAP experiment data (*) instead of 0.0263 from an independent experiment (solid circles) with only three frames collected. Another important concept is that t he immobile fraction continues to be photobleached by the image capture process, leading to decay in the total fluorescence. Therefore, the FRAP curve is a combination of dynamics due to exchange of the mobile species, bleaching of the recovered portion d ue to the imaging process and the decay due to bleaching of the immobile fraction. The main utility of this approach is when the bleaching during image capture significantly changes the FRAP dynamics. To test the extent of bleaching, the approach should b e to first estimate the immobile fraction as described above. Then when the FRAP experiment is performed, the apparent immobile fraction from the FRAP

PAGE 119

119 experiment should be compared with the measured immobile fraction. A decrease from the actual immobile fr action indicates the extent to which photobleaching during image capture is relevant in the experiment. The effect of photobleaching may be unavoidable either due to the fact that the fluorophore may be particularly susceptible to bleaching or the intensit y of the fluorophore in some cells may be lower than others requiring a higher excitation intensity leading to higher bleaching. In this situation, Equation 10A and 10B should be used to fit the FRAP experiment. The parameters that are known in these equat ions are and (measured or known directly from the experiment). The fitting should determine the values of and In situations where the diffusing cytoplasmic (or membranous) molecules can be tracked (such as in the example in Figure 5 4), it is useful to determine the value of from fitting the cytoplasmic pool, such that only two parameters need to be estimated. An interesting prediction is that when the immobile fraction is present, the fluorescence in a FRAP experiment can reach a maximum and decay subsequently. The decay is due to the bleaching of the immobile fraction. Eventually, a stead y state is reached when the bleaching due to image capture is compensated by recovery (unobserved dynamics). In the absence of the immobile fraction, the fluorescence reaches a steady state without reaching a peak. If bleaching due to image capture is so s evere that the free molecules (in the cytoplasm) are bleached with each captured image, then there will be significant decay in the FRAP experiment and no steady state will be reached. Sometimes researchers vary the time interval during FRAP experiments s uch that images are collected at a higher rate at the beginning of the recovery, and a smaller

PAGE 120

120 rate at later stages. We explored the prediction of Eq. 6 with parameters = 0.2, = 0.2 = 0.2 = 2s, 4s and 6s ( increased after every 10 images). At every increase the steady state fluorescence recovery is predicted to increase (Figure 5 6 A) y process. The increase is due to a longer time interval between image captures that allows for more recovery in between successive image capture events (the bleaching due to image capture remains the same). If a conventional model is used to fit this model generated data, the is estimated to be 0.085, which is much different from the real value = 0.2 (Figure 5 6 B). This approach is applicable to more complicated situations. The method is to substitute Equation 5 4 with the relevant model for the unobserved dynamics (for example, models that include equations for coupled transport and binding). The main concept is to replace the initial condition for the unobserved dynamics in between frames with the bleaching corrected concentration from the previous time interval. Thus, the approach is general and should work for any FRAP analysis. If the values of the bleaching par ameter are calculated not from FRAP experiments, but from whole cell imaging experiments with Equation 5 3 then it is important to ensure identical imaging conditions for the corresponding FRAP experiments. This is because depends on experimental conditions including the exposure time and the excitation intensity; changing imaging conditions will change thereby invalidating the analysis for the FRAP experiment.

PAGE 121

121 Summary of the Model Fluoresce nce recovery after photobleaching (FRAP) is a powerful technique to study protein kinetics in living cells. A lot of interesting work has been done using this technique. However, some important factors have been ignored when interpreting the results of FRA P experiment in those past study. A key factor is bleaching due to the image capture. In this chapter, w e proposed a new model which including the effect of bleaching due to the image capture. The model was tested in the analysis of our experimental result s. This model gives us better interpretation of FRAP experiments.

PAGE 122

122 Figure 5 1 Hypothetical effect of photobleaching due to imaging capture on a whole cell imaging experiment (A)The dotted lines indicate the actual decay of the fluorescence due to photobleaching during image capture. indicates the measured intensity from the resulting image. (B) Averaging involved in the imaging process is cancelled by normalization (see text for more details). The normalized average fluorescence is now equal to as seen in the plot. (C) Effect of photobleaching on hypothetical FRAP recovery The dotted curve is the actual dynamics consisting of (unobserved) recovery inte rspersed by bleaching during image capture, indicates measured intensity. The solid triangle at indicates the normalized initial intensity before photobleaching.

PAGE 123

123 Figure 5 2 Solutions to Equation 5 6 showing how photobleaching during image Recovery curves are shown with = 0.4, = 0.2 = 1 (*), 0.5 ( ), 0.25 ( ) and 0.1 ( ) ( from top to bottom). For plotting purposes, is assumed to be /10. The solid triangle at indicates the normalized initial intensity before photobleaching.

PAGE 124

124 Figure 5 3 Solutions to Equation 5 10A and 10B that account for the presence of an actual immobile fraction. (A) Observed r ecovery curves with = 0.4, = 0.3 (immobile fraction) = 0.2 << 1 (i.e. negligible photobleaching of the cytoplasmic molecules such that ) and = 1 (*), 0.5 ( ), 0.25 ( ), 0.1 ( ) (from top to bottom) (B) Illustration of behavior of mobile (dashed line) and immobile fraction s (dotted line) during recovery for = 1 (* indicates total intensity ) The immobile fraction can be seen to decay due to bleaching during image capture, resulting in a decrease in the total fluorescent intensity. (C) Effect of the immobile fraction on the observed recovery curves. = 0.4, = 0.2 << 1, = 1 and = 0.2 (*), 0.4 ( ), 0.8 ( ), 0.6 ( ) (from top to bottom) Pronounced transients are observed in the recovery. (D) Effect of the bleaching function on recovery. Observed r ecovery curves with = 0.4, = 0.3, << 1, = 1, and = 10 6 (*), 0.2 ( ), 0.46 ( ), 1.1 ( ) ( from top to bottom) Solid triangles at in all figures indicate the normalized initial intensity before the photobleaching.

PAGE 125

125 Figure 5 4. Example of the application of Equation 5 10A and 10B for fitting a GFP VASP FRAP experiment (A) Captured images from a FRAP experiment in an NIH3T3 fibroblast expressing GFP VASP The box shows the bleach spot. Scale bar, 1m. (B) Observed r ecovery with only slight apparent bleaching of the free molecules (in the cytoplasm) due to image capture. The excitation laser intensity was 4%. The solid curve is the fitting of the data to the model in Eq 9. The immobile fraction was estimated from a separate experiment (solid circles) in the same focal adhesion as described in Material and M ethods The value of was determined from the fluorescence values

PAGE 126

126 before and immediately after the bleach and = 1.3s in the experiment. The fitting yielded the parameters = 0.0902 = 0.0014 =0.17 s 1 (C) Observed recovery in the same focal adhesion from a second FRAP experiment with apparent bleaching. The excitation laser intensity was increased to 10%. The fluorescence is observed to go through a peak and then decrease due to bleaching caused by image capture. The fitting of the data to the model gave the parame ters = 0.1301 = 0.0091 =0.18 s 1 The value of is very close to that estimated from the fitting in (B) thus validating the model Solid triangles in (B) and (C) in dicate the normalized initial intensity before the photobleaching. (D) Fluorescent intensity profile in the cytoplasm (free molecules) in experiment (B) which shows there is no detectable photobleaching of cytoplasmic molecules (E) Fluorescent intensity profile of the cytoplasm (free molecules) in experiment ( C ) showing a clear decrease in the concentration due to pronounced bleaching Fitting of the cytoplasmic intensity to Eq. 3 yields the bleaching parameter = 0.0118 which is very close to the value determined from the fitting in (C). The model for the cytoplasmic intensity was fit to 70 of the 80 seconds for which the data was collected (corresponding to 53 measurements); the first 10 seconds showed a significant de viation possibly due to deviations in focus.

PAGE 127

127 Figure 5 5. Typical fitting for a GFP VASP FRAP experiment. The same FRAP experiment data as shown in Fig. 4B was fit to The fitting yielded = 0.30. Figur e 5 6. Illustration of typical fitting failed to estimate in experiments with different time interval (A) Observed recovery curve (solid line and ) calculated by Eq. 6 with parameters = 0.2, = 0.2 = 0.2 = 2s, 4s, 6s (it increases after every 10 frames). Dot lines show unobserved dynamics. (B) The solid line is fitting to the data in (A) to The fitting yielded the parameter = 0.0846.

PAGE 128

128 CHAPTER 6 CONCLUSIONS Molecular motors in living cells play very important ro les in mechanotransduction and intracellular traffic T his dissertation focus sed on molecular motor f orces and their effects on nuclear and meicrotubule mechanics Our findings significantly advance the understanding of nuclear movement and related cytoskeletal mechanics as summarized below. Summary of Findings Nuclear Rotation in Living Cells Nuclear rotation has been observed in different types of cells for decades and it is especially visible in certain cell type s like fibroblast s However, the biological function of nuclear rotation remaine s unknown. In this study, the focus was on investigating the mechanism for torque generation on the nucleus. W e found that the angle of nuclear rotation is directionally persistent on a time scale of tens of minutes, but rotationally diffusive on longer time scale s and this r otation required th e activity of the microtubule motor dynein. Based on these results, a mechanical model describing torque generation on the nucleus was proposed. A key prediction of the model is that the magnitude of nuclear rotation depends on the distance between the cen trosome and nuclear centroid, which was confirmed by nuclear rotation experiments in patterned cells. Nuclear Translation in Living Cells W toskeletal forces is still controversial In our study we designed two independent experiments utilizing two novel techniques, protein photo activation and cell micromanipulation to investigate force exerted on the nucleus.

PAGE 129

129 The results suggest that there is a tug of war between forward pulling and rearward pullin g forces on the nuclear surface and the contractile force transmits through the nucleus from cell leading e dge to the trailing edge in single migrating cells Effect s of Dynein on Microtubule Mecha nics and Centrosome Centering Contrasting m odels of centrosome centering have been proposed in the past decade and the main argument between these models is whether the cell positions the centrosome by pulling force or pushing force from dynamic microtubules. We approached this problem from single microtubule mechanics. By using femtosecond laser ablation, we showed that the freed minus end fragments generated by severing microtubules ighten as expected if the buckling was caused by compressive forces. Moreover, the curvature of the minus ended severed MT increased in most cases. W hen dynein was inhibited, the freed minus end fragment s did straighten out after severing. We proposed a mo del for dynein pulling force generation and conducted simulations of cen trosome centering using the model. According to our model m icrotubules are under tension along most of their length, while tips near the cell periphery are likely to be under compress i on due to polymerization forces T he simulation s further confirmed that d ynein dependent forces are essential for centering the centrosome in endothelial cells and a balance of tensile and frictional forces from the dynein motors is sufficient to explain centrosome centering. FRAP Model Accounting for Immobile Molecules and Bleaching due to Image Capture FRAP is a popular technique to study protein diffusion and binding kinetics in living cells. A lot of research has been done using this technique (85 92, 133) though an important fact or, the bleaching effect due to image capture, has not received attention

PAGE 130

130 The analysis of FRAP data is normally done by fitting the data to a continuous k i netic mo del. However, when the loss of fluorescence due to image capture is considered, the recovery function is affected at regular intervals due to image capture which occurs at certain time intervals. To provide a better interpret ation of the FRAP data, we proposed a new model wh ich accounts for the bleaching due to image capture. We also designed exp eriments from which we can calculate the fraction of immobile fluorescent molecules with minimum effect of the photobleaching The kinetic parameters obtained by fitting of FRAP data with bleaching due to image capture at different levels were very close t o each other, which validate s our model to be reasonable and effective. Future Work Further Investigation on Nuclear Movement Microtubule motor kinesin in nuclear r otation In Chapter 2, a nuclear rotation model based on force generated by microtubule motor dynein walking on dynamic microtubules was proposed The other microtubule motor kinesin is also known to affect intracellular transport, microtubule depolymerization and spindle formation (14, 30, 145, 163) S imilar to dynein, kinesin 1, one member of kinesin family, walks on microtubule toward the plus end. It has recently been found to form complexes with nesprin 2 by associating with and recruiting kinesin light chain1 (KLC1) to the outer nuclear membrane (164) K inesin 1/nesprin 2/SUN domain macromolecular assemblies, spanning the entire nuclear env elope (NE), function in cell polarization by anchoring cytoskeletal structures to the nuclear lamina (164) Therefore, kinesin could play a role in nuclear rotation.

PAGE 131

131 To investigate this, we could use the method s developed in Chapter 2. If kinesin 1 is shown to affect nuclear rotation, we can incorporate force s gene rated by kinesin into our model. A ctomyosin contraction and nuclear positioning We found that the nucleus is subject to contractile force s generated by actomyosin from both the leading edge and the trailing edge. While these studies gave us a good understa nding of the nature of force s driving nuclear mo tion the magnitude of this actomyosin contractile force is unknown To measure the actomyosin contractile for ce we can combine the protein photo activation technique with traction force microscopy (59, 125) The strategy is to seed cells on to soft substrate (eg. polyacrylamide gels) with fluorescent nano bead s embedded and then conduct the experiment as described in Chapter 3. By recording the bead movemen t upon photo activation, we should be able to estimate the scale of the contracti le force generated by actomyosin. A futher stud y of long t erm nuclear movement upon photo activation could also provide a good re a dout of the force generation by ac tomyosin. An issue with such experiment s is that cells are migrating and constantly changing their shape, w hich in turn introduces a lot of complexity in charateriz ing actomyosin contractile force induced nuclear translation. Micropatterning could eliminate the e ffects of both ce ll migration and shape changes which provide s a n effective system to focus on the effect of actomyosin contractile forc e on nuclear translation. It would also be interesting to observe what would happen to the nucleus and the centrosome over time by carrying out photo activation in patterned squared cells since there is no polarity in these cells as the centrosome overlaps with the nucleus (Figure 2 7).

PAGE 132

132 The role of i ntermedi ate filament in nuclear m ovement Among the three component s of cytoskeleton, the role of intermediate filament s (IF) in nuclear moveme nt is relatively unknown Fluorescent images of NIH 3T3 fibroblasts and BCE cells transfected with vi mentin plasmid show that intermediate filaments are present throughout the cell and form a very dense network around the nucleus ( Figure 6 1 ) There are no known examples of IF dependent cell movements or motor proteins that mo ve along intermediate filaments. Therefore, the basket like net work of IF around the nucleus would most likely introduce friction al forces that oppose movements of the nucleus Consist ent with this idea, Gerashchenko et al (49) showed that vimentin IFs inhibit nuclear rotation, and variant proteins of the mutated wild type gene for vimentin that lacked considerable fragments of the N and C terminal domains restored nuclear anchoring. Some proteins have been identified as the connect ors between IF and the nucleus Ralston et al (165) showed that the desmin knock out mouse muscle had mispositioned nuclei Vimentin wa s found to be closely associated with the nucleus and mutant forms of vimentin exhibit nuclear morphology defects (166) A newly identified KASH protein, nesprin 3, likely functions to connect the outer nuclear membrane to intermediate filaments (167) Plectin, a plakin family member, consists of an actin binding domain, an extended coiled c oil domain, and an intermediate filament binding domain and can crosslink actin filaments to intermediate filaments (168) .It is hypothesized that nesprin 3 and plectin together could extend from the outer nuclear membrane into the cytoplasm to interact with intermed iate filaments (167)

PAGE 133

13 3 Therefore, using the methods we developed in this thesis we should be able to investigate the role of IF in nuclear rotation and translation by interrupting the connection s between IF and the nucleus. These studies could also potentially give us better understanding of IF mechanics. Nuclear P ositioning under the Influence of E xtracellular F orces Endothelial cells form the inner wall of blood vessels. T hey are constantly subj ected to shear stress es created by blood flow. Thus, a n interesting problem is to investigate how the cell position s its nucleus under such a condition. A typical device for shear stress experiment is the parallel plate flow chameber as decribed in (169) Lee et al (48) have shown that shear stress caused cell polarization in which it appeared that the centrosome moved against the direction of flow However, it is also possible that the nucleus could have moved downstream while the centrosome stayed in the same position similar to what has been observed in nuclear movement in wounded cells (5) Some related question s to be answered are how the nuclear translation and centrosome positioning coordinate with cell migration under shear stress ? W hat are the effects of motor forces (ie., dynein and actomyosin) on the nuclear movement under shear stress? Combining shear stress flow experiment with the methods that we used to study motor forces and nuclear positioning (eg. tracking nuclear position, centrosome position and cell centroid, inhibition of dynein and myosin etc. ) we should be able to answer such questions and improve our understaning of nuclear movement in genera l. C ell Migration During W ound H ealing The scratch wound healing assay has been used as an effective method to study collective cell migration speed for decade s While this method is easy to carry out and

PAGE 134

134 gives clear results in comparing cell migration under different conditions (59) it is missing some key features that are observed during the wound healing in vivo One of these is that cells need to clean up the debris of cells in the wound to be able to move on while it is a clean wound in the scratch wound healing assay Our group has developed a new stamping wound healing assay which addressed this issue (170) The method involves the physical contact of a soft mold with raised features onto confluent epithelial cells. With t his method, we successfully created well defined wounds with dead cell debris in the wound area. Our previous study has shown that the debris clearing process was remarkably efficient with no trace of debris detectable after clearance. The rate of wound cl osure in the presence of cell debris was found to be comparable to that in the absence of cell debris. In the body, other cells like macrophage s and neutrophils might also be involved in the wound healing process. A m acrophage is supposed to phagocytose (e ngulf and then digest) cellular debris a nd neutrophils are an important part of the innate immune system. They are one of the first responders of inflammatory cells to migrate towards the site of inflammation. So it would be very interesting to see how th ese cells react to the wounding. The strategy w ould be to first co culture macrophage s or neutrophils with fibroblasts or epithelial cells, then create a wound using our stamping assay and record cell migrations. Macrophage or neutrophils can be labeled with a fluorescent tracker, so we can differentiate their movement from fibroblasts or epithelial cells. These experiments would significantly advance the in vitro wou n d healing study. Effects of Other Motor Forces on Microtubule Mechanics We di scuss ed motor protein dynein as the major force generator in shaping microtubules. However, we do not rule out the possible contribution by kinesin, which

PAGE 135

135 can also link a microtubule to the cytoskeleton. Similar laser ablation experiment s can be done to mi crotubules in kinesin inhibited cell s to verify the effect of kinesin on microtubule mechanics. If a significant effect is observed, we can incorporate kinesin force s into the motor model in a manner similar to dynein but with opposite direction. A key ass umption made in in Chapter 4 was that the underlying cytomatrix that the dynein binds to for pulling on microtubules is stationary (as t he schematic in Figure 6 2). However in m igrating cells, the underlying cytomatrix is highly dynamic due to the constant polymerization of actin filaments a nd the subsequent shape changes of the cell. Acto myosin contraction has been shown to cause MTs buckling and promote MTs breaking in motile cells (72, 73, 171) This is consistent with our model discussed before, in which the pulling force on microtubule s generated by dynein requires anchorage of dynein to the F actin network. So w hen acto myosin retrograde flow occurs it could generate force s on the dynein that binds to the F actin network and consequently transfer forces onto the MTs The protein photo activation technique would be an excellent method to study effect of acto myosin retrograde flow on microtubule mechanics. In our previous study, we notice d that when lamellipodia formed upon the photo activation of Rac 1, there was continuous acto myosin flow being generated in the region of activation (Fi gure 6 3). We also observed remodeling of the stress fiber network in this region (Figure 6 3D and Figure 6 3E) Therefore, if we track the shape of GFP tubulin labeled microtubules in the vicinity of the newly forming lamellipodia with time and compare th e curvature of bent microtubules before and after the photo activation we will be able to quantitatively test the effect of acto myosin retrograde flow on microtubule mechanics. Furthermore,

PAGE 136

136 the same experiment coul d be done in dynein inhibited cells A cc ording to our hypothesis inhibition of dynein would disrupt the linkage between F actin network and microtubule s and hence the effect of acto myosin retrograde flow on microtubules w ould be reduced.

PAGE 137

137 Figure 6 1. Image of intermediate filament in livin g cells. NIH 3T3 fibroblasts were transfected with CFP vimentin. Intermediate filaments are most concentrated around the nucleus.

PAGE 138

138 Figure 6 2. Cartoon of a dynein motor indicating how the minus directed motor bound to the cytomatrix exerts a force tow ards the microtubule plus end. Individual dynein molecules walk towards the microtubule minus end at a speed v m ( along the local tangent direction, t ) that depends on the opposing force f Each segment of the microtubule moves relative to the cytoskeleton with a velocity v

PAGE 139

139 Figure 6 3. Formation of acto myosin retrograde flow and stress fiber remodeling upon photo activation of Rac 1. A) DIC images of a NIH 3T3 fibroblast at the beginning and the end of photo activation. The nucleus moved towards the activation. Bright spots are the activation region. (B) Super imposition of the cell out line and nucleus from frame 0 minute (red; beginning of Rac1 activation) and 30 minute (blue, end of Rac1 activation. Dash line circle is the activation region. (C) G FP actin expressed NIH 3T3. No visual movement or the stress fiber was observed. (D) Overlay image of actin cytoskeleton at 0 minute (green) and 30 minute (magenda). No significant visual changes of stress fibers were observed. (E) Zoom in image of the whi te square region in (D). Actin cytoskeleton underwent remodeling at the activation site. The stress fiber at 0 minute (open arrow head, green) disappeared at 30 minute, a new stress fiber was forming (arrow head, magenta). Lots of free actin (magenta) accu mulated at the activation site. Scale bars (A), (C) and (D), 10 m; (E) 5 m.

PAGE 140

140 APPENDIX A MATLAB CODE FOR NUCL EAR ROTATION ANALYSI S The strategy of nuclear rotation tracking was previous described in Material and Methods in Chapter 2. %Start of Nuc learRotation.m clc clear all close all matfiles=dir((fullfile('H:','nuclear rotation control data','longtimedata','112709 1 11data','*.tif'))); num=5;%This is the size around the correlation peak to which we fit a Paraboloid fitting, may be changed respective to the resolution of the image template = imread(matfiles(1).name); %Crop images of two nucleoli as templates [croptemplate1, rectcroptemplate1] = imcrop(template); [croptemplate2, rectcroptemplate2] = imcrop(template); %Coordinates of the templates x1 = rectcroptemplate1(1,1); y1 = rectcroptemplate1(1,2); width1=rectcroptemplate1(1,3); height1=rectcroptemplate1(1,4); x2 = rectcroptemplate2(1,1); y2 = rectcroptemplate2(1,2); width2=rectcroptemplate2(1,3); height2=rectcroptemplate2(1,4); %Rec rop templates with roundup size. We round this because the way MATLAB program works croptemplate1=imcrop(template,[round(x1) round(y1) width1 height1]); croptemplate2=imcrop(template,[round(x2) round(y2) width1 height1]); %Image correlation give the displa cement of the features c1 = normxcorr2(croptemplate1,template); c2 = normxcorr2(croptemplate2,template); %Gaussian fitting to get the peak position in subpixels [max_c1, imax] = max(c1(:)); [ypeak1old, xpeak1old] = ind2sub(size(c1),imax(1)); image=c1(ypeak 1old num:ypeak1old+num,xpeak1old num:xpeak1old+num); [param,px,py]=Gaussianpeaknew1(image); zo = param(1); xpeaknew=param(2); ypeaknew= param(3); zn=param(4); wnx=param(5); wny=param(6); theta=param(7); x_sub=xpeak1old num 1; y_sub=ypeak1old num 1; xpeak1= xpeaknew+x_sub; ypeak1=ypeaknew+y_sub; [max_c2, imax1] = max(c2(:)); [ypeak2old, xpeak2old] = ind2sub(size(c2),imax1(1)); image=c2(ypeak2old num:ypeak2old+num,xpeak2old num:xpeak2old+num); [param,px,py]=Gaussianpeaknew1(image);

PAGE 141

141 zo = param(1); xpeaknew=param(2); ypeaknew= param(3); zn=param(4); wnx=param(5); wny=param(6); theta=param(7); x_sub=xpeak2old num 1; y_sub=ypeak2old num 1; xpeak2=xpeaknew+x_sub; ypeak2=ypeaknew+y_sub; %Calculate the actual positions corr1{1}=c1; corr2{1}=c2; py1(1)=ypeak1;%x coordinates of starting points of croptemplate1 in correlation px1(1)=xpeak1;%y coordinates of starting points of croptemplate1 in correlation py2(1)=ypeak2;%x coordinates of starting points of croptemplate2 in correlation px2(1)=xpeak2;%y coordinates of starting points of croptemplate2 in correlation imag{1}=template; magr1=sqrt((px2(1) px1(1))^2+(py2(1) py1(1))^2); mag(1)=magr1%distances between two features in each image %coordinates for the new templates in image{1} w1=px1(1) x1; h1= py1(1) y1; w2=px2(1) x2; h2=py2(1) y2; ang(1)=0; ang1(1)=0 t=50%Set how many images in the small loop, which use the same templates for n=0:28%Set how many big loops (total number of images/number of images in a small loop) a=1+t*n; xn1=px1(a) w1; yn1=py1(a) h1; xn2=px2(a) w2; yn2=py2(a) h2; croptemplaten1=imcrop(imag{a},[round(xn1) round(yn1) width1 height1]); croptemplaten2=imcrop(imag{a},[round(xn2) round(yn2) width1 height1]); croptemp1{1+n}=cropte mplaten1; croptemp2{1+n}=croptemplaten2; figure(n+1) imshow(imag{a}) hold on plot(px1(a) size(croptemp1{n+1},2)/2,py1(a) size(croptemp1{n+1},1)/2,'*') plot(px2(a) size(croptemp2{n+1},2)/2,py2(a) size(croptemp2{n+1},1)/2,'*') for i = 2:t+1 b=i+t*n; image=imread(matfiles(b).name);

PAGE 142

142 imag{b}=image; c1 = normxcorr2(croptemplaten1,image); c2 = normxcorr2(croptemplaten2,image); [max_c1, imax] = max(c1(:)); [ypeak1old, xpeak1old] = ind2sub(size(c1),imax(1)); image=c1(ypeak1old num:ypeak1old+num,xpeak1old num:xpeak1old+num); [param,px,py]=Gaussianpeaknew1(image); zo = param(1); xpeaknew=param(2); ypeaknew= param(3); zn=param(4); wnx=param(5); wny=param(6); theta=param(7); x_sub=xpeak1old num 1; y_sub=ypeak1old num 1; xpeak1=xpeaknew+x_sub; ypeak1=ypeaknew+y_sub; [max_c2, imax1] = max(c2(:)); [ypeak2old, xpeak2old] = ind2sub(size(c2),imax1(1)); image=c2(ypeak2old num:ypeak2old+num,xpeak2old num:xpeak2old+num); [param,px,py]=Gaussianpeaknew1(image); zo = param(1 ); xpeaknew=param(2); ypeaknew= param(3); zn=param(4); wnx=param(5); wny=param(6); theta=param(7); x_sub=xpeak2old num 1; y_sub=ypeak2old num 1; xpeak2=xpeaknew+x_sub; ypeak2=ypeaknew+y_sub; py1(b)=ypeak1; px1(b)=xpeak1; py2(b)=ypeak2; px2(b)=xpeak2; %Ca lculate rotation angles using positions of two nucleoli r1dotr2=(px2(a) px1(a))*(px2(b) px1(b))+(py2(a) py1(a))*(py2(b) py1(b)); magr1=sqrt((px2(a) px1(a))^2+(py2(a) py1(a))^2); magr2=sqrt((px2(b) px1(b))^2+(py2(b) py1(b))^2); angle = acosd(r1dotr2/magr1/m agr2); mag(b)=magr2;%Distance between two features in each image ang(b)=angle; one=atand((py1(a) py2(a))/(px1(a) px2(a))); two=atand((py1(b) py2(b))/(px1(b) px2(b))); angle1 = (one two); ang1(b)=angle1+ang1(a);%Accumulative rotation angles end end

PAGE 143

143 APPENDIX B MATLAB CODE FOR NUCL EAR TRANSLATION ANAL YSIS Tracking nuclear positions in photoactivation experiments described in Chapter 3. % Start of NuclearTranslation.m clc clear all close all matfiles=dir((fullfile('F:','Rac1NuclearTranslation','data','blebbistatin','1 21610 4','*.tif'))); %Position of activation center i=2;% image=imread(matfiles(i).name); imag{i}=image; imshow(imag{i}); [x1,y1]=ginput(12);%Choose a certain number of points to fit to a eclipse ellipse_c=fit_ellipse(x1,y1);%Fit to a eclipse xc1=ellipse_c.X0_in; yc1=ellipse_c.Y0_in;%X,y coordinates of the center of the eclipse imshow(imag{i}); [x2,y2]=ginput(12); ellipse_c=fit_ellipse(x2,y2); xc2=ellipse_c.X0_in; yc2=ellipse_c.Y0_in; imshow(imag{i}); [x3,y3]=ginput(12); ellipse_c=fit_ellipse(x3,y3); xc3=ellipse_c.X0_in; yc3=ellipse_c.Y0_in; xc=mean([xc1 xc2 xc3]); yc=mean([yc1 yc2 yc3]);%Use the mean of results from three fittings as the position of activation center %Position of nuclei m=31; for i=1:m%length(matfiles) image=imread(matfiles(i).name); imag{i}=image; imshow(imag{i}); [x,y]=ginput(12); ellipse_t = fit_ellipse(x,y); ellipse{i}=ellipse_t; xtrans(i)=ellipse_t.X0_in; ytrans(i)=ellipse_t.Y0_in; end %Calculate the distance between nuclear centroid and the activation center %at each time point

PAGE 144

144 for k=1:m; dl(k)=sqrt((xtrans(k) xc)^2+(ytrans(k) yc)^2); end %Calculate the changes of distance between nuclear centroid and the activation center %at each time point from time zero d(1)=0; for j=2:m; d(j)=dl(1) dl(j); end clear imag t=[0:5:5*(m 1)]; plot(t,d) figure plot (xtrans,ytrans,' *',xc,yc,'o') %Convert to actual length px=0.087;%set the pixel size xt=xtrans*px; yt=ytrans*px; xca=xc*px; yca=yc*px; dt=d*px; %move the activation center as the origin of coordinates xt0=xt xca; yt0=yt yca; xca0=0; yca0=0; %move the original position of the nucleus as the origin of coordinates xtt=xt xt(1); ytt=yt yt(1); xcat=xca xt(1); ycat=yca yt(1); rotangle=atand(ycat/xcat);

PAGE 145

145 APPENDIX C MATLAB CODE FOR CENT ROSOME POSITION ANAL YSIS Tracking centrosome positions in experiments described in Chapter 3 and Chapter 4. %Start of centrosome.m clc clear all close all n=1; %Starting image for analysis m=28; %Ending image for analysis threshold=2; % Set thresholding parameter, this depends on the image quality num=5; % This is the size around the correlation peak to which we fit a Paraboloid fitting, may be changed resp ective to the resolution of the image px=0.088;% Pixel size %Read the image files and collect information matfiles2 =dir((fullfile('F:','newRac1NuclearTranslation','newdata', 'centrosome','041311 3','centrosome','*.tif'))); size_matfiles2=size(matfiles2) ; for i =1:size_matfiles2(1) temp_image = imread(matfiles2(i).name); size_image=size(temp_image); imag{i}=temp_image; end %Crop a template from the first image [template, rectcrop] =imcrop(imag{1},[min(min(imag{n})),max(max(imag{n}))]); rectcrop(1)=round(rectcrop(1)); rectcrop(2)=round(rectcrop(2)); width=round(rectcrop(3)); height=round(rectcrop(4)); [template, rectcrop]=imcrop(imag{1},[rectcrop(1),rectcrop(2),width,height]);% We round this because the way MATLAB program works save _template{1}=template; save_rectcrop{1}=rectcrop; %Eliminating noise based on the threshhold given imag_centr{1}=template; imag_max=max(max(imag_centr{1})); size_imag=size(template); for i = 1:size_imag(1) for j = 1:size_imag(2) if imag_centr {1}(i,j)
PAGE 146

146 end %Use built in MATLAB functions to make a binary and calculate centroids L = bwlabel(imag_centr{1}); s = region props(L, 'Centroid'); %Record the centroid coordinates xcord(1)=s(1).Centroid(1)+rectcrop(1) 1;%coordinates of the centroids ycord(1)=s(1).Centroid(2)+rectcrop(2) 1; xCent(1)=xcord(1); yCent(1)=ycord(1); for g=2:m image=imag{g}; [c,corner_y(g),corner_x(g),xrefined,yrefined,cvalue]=process(template,image,n um); %This gives the corner position in the ith image of the old template (i.e. from the i 1 image). The main code passes three things to the custom he template from image i 1, image i and the parameter num. x_offset(g)=round(rectcrop(1)) corner_x(g); y_offset(g)=round(rectcrop(2)) corner_y(g); %this is the measured offset immediately between images cvalues(g)=cvalue; %Updated positions are calcula ted with the offsets and stored xcord(g)=xcord(g 1) x_offset(g); ycord(g)=ycord(g 1) y_offset(g); dc(1)=0; dc(g)= sqrt(x_offset(g)^2+y_offset(g)^2); %Updated positions are used to determine the template position in image i templatex=round(xcord(g) wid th/2); templatey=round(ycord(g) height/2); %Cropping new template for the next correlation [template,rectcrop]=imcrop(imag{g},[templatex,templatey,width,height]); %Refine the new template position by calculating the centroid imag_centr{g}=template; imag_max=max(max(imag_centr{g})); size_imag=size(template); for i = 1:size_imag(1) for j = 1:size_imag(2) if imag_centr{g}(i,j)
PAGE 147

147 L = bwlabel(imag_centr{g}); s = regionprops(L, 'Centroid'); %Store the centroid position as a secondary position for each time frame xCent(g)=s(1).Centroid(1)+rectcrop(1) 1; yCent(g)=s(1).Centroid(2 )+rectcrop(2) 1; %Redefine the new template positions using the calculated centroid templatex=round(xCent(g) width/2); templatey=round(yCent(g) height/2); [template,rectcrop]=imcrop(imag{g},[templatex,templatey,width,height]); %Save template and template coordinates save_template{g}=template; save_rectcrop{g}=rectcrop; end %Calculate the position in length x=xCent*px; y=yCent*px; xd=xcord*px; yd=ycord*px; %Plot positions, template, binary images as subimages together for l=n:m 1, image=imread(matfiles2(l).name); ima{l}=image; figure(l) subplot(1,4,1) imshow(ima{l}) subplot(1,4,2) imshow(imag{l},[min(min(imag{n})),max(max(imag{n}))]) hold on plot(xcord(l), ycord(l), 'bx') hold on plot(xCen t(l), yCent(l), 'rx') subplot(1,4,3) imshow(save_template{l},[min(min(imag{n})),max(max(imag{n}))]) subplot(1,4,4) imshow(imag_centr{l},[min(min(imag{n})),max(max(imag{n}))]) end

PAGE 148

148 APPENDIX D MODEL FOR MICROTUBULE MECHANIC S AND CENTROSOME CENTERING Microtubule Mechanics The relaxation time of a bent microtubule in a viscous fluid can be estimated from the equation of motion for small displacements, from the straight configuration (146, 172) (D 1) where is the mass per unit length of the filament, is the frictional resistance per unit length, and (52) is the flexural rigid ity. The solutions are a linear combination of trigonometric and hyperbolic functions with a wave vector k For a free ended segment in the lowest energy mode thus, for a typical segment length Neglecting inertia, in the overdamped limit the relaxation time is given by (D 2) The frictional resistance is given by where is the effective viscosity of the background fluid and the aspect ratio of the segment Estimates of the viscosity of cellular fluids range from 10 3 Pa s (water) to 10 1 Pa s (173, 174) and the corresponding relaxation times for segments are in the range 10 3 10 1 s. Observations of microtubule straightening in dynein inhibited cells indicate that the effective viscosity controlling the relaxation o f microtubules is much larger. We take a value Pa s, which gives a relaxation time of approximately 7 s for a 3 segment, similar to the experimental data in Figure 4 3A.

PAGE 149

149 Model for D ynein F orces In this model, the force exerted by a dynein molecule is assumed to depe nd on the displacement of the motor from the point of attachment to the microtubule (in a space fixed frame), and has contributions from the motion of the microtubule as well as the motor. Individual dynein molecules walk toward a microtu bule minus end at a speed v m that depends on the opposing force For simplicity, the force speed relation for dynein motors is taken as linear (109) ( D 3) where i s the speed of an unstressed motor, is the motor stall force, and t is a unit vector directed toward the plus end of the microtubule The force on the cytomatrix linkage increases as dynein translates and begins to exert tension, whi le motion of the microtubule itself can also contribute to a change in force. Assuming that the dynein linkages can be approximated as linear springs with stiffness D 4) where is the local velocity of the microtubule relative to the cytoskeleton. The time dependent force from a single motor follows from Equation D 3 and Equation D 4 with the initial condition ( D 5) where the timescale for the motor to stall Typical values for the parameters are: and (52, 109) leading to a timescale

PAGE 150

150 Dynein linkage s between the cytomatrix and the microtubule are assumed to dissocia te with a first order rate constant k off so that the probability of a linkage remaining at time is Thus the average force exerted by a single linkage over a time interval is The mean linkage lifetime is and thus there are linkages broken and reformed in the time interval T he average force exerted by a dynein linkage that is associating and dissociating from the microtubule is therefore ( D 6) where is the average force per linkage on a stationary actomyosin cortical network. The best estimate of the dynein off rate is so that the mean motor force is essentially the same as the maximum force, The force per unit length of the microtubule K can then be obt ained by multiplying Equation D 6 by the density of dynein linkages per unit length, ( D 7) where is the friction coefficient for lateral motion per dynein linkage Simulation M ethods. The nume rical simulations are based on a n algorithm for integrating the equations of motion of an elastic filament (175) In addition we include length changes in the microtubule to incorp orate polymerization and depolymerization of the microtubules. The length of the microtubule is taken to be a continuous variable, described by a stochastic differential equati on that inclu des switches to ca tastrophe (depolymerization) and recovery (polymerization) The parameters are taken from experimental

PAGE 151

151 measurements (116, 117) : and Segments are added or removed from each microtubule during the simulation, according to the calculated changes in length. The model for dynein motors Equation D 7, is included in the force balance that describes t he evolutio n of each microtubule, Equation 4 2, which leads to an expression for the microtubule velocity in terms of the coordinates The conformation of each microtubule is then updated with a time step of the order of 1 s. The dynamics of the centrosome were sim ulated by distributing microtubules around the centrosome with a uniform distribution of angles ; the microtubules are coupled to the centrosome by stiff springs In the experiments on animal cells many of the microtubules are observed to get pinned when th ey reach the cell boundary ; typically these microtubules buckle as shown in Figure 4 10 However it is sometimes observed that the microtubules slide along the cell boundary as shown in Figure 4 10A and occasionally microtubules apparently stop growing ( Fi gure 4 10B). We include the possibility of microtubule pinning by an angle dependent rate where is the angle between the tangent at the tip of the microtubule and the norm al to the cell surface. We take the rate constant and the unpi nning rate as k With this choice of kinetics about 2/3 of the microtubule s are pinned at any one time, similar to the experimental observations. The remaining microtubule s continue to polymerize as their tips slide along the cell surface. We also reduce the growth rate of pinned microtubules by 50% in comparison with the free microtubule s to account for stalling in the polymerization kinetics However, in simulations of in vitro microtubule growth, we

PAGE 152

152 allowed the microtubules to slide freely along the surface, mimicking the effect of the smooth glass walls. Simulations of M icrotubule B uckling The simulation illustrated in Fig ure 4 8 B models a minus ended microtubule after s evering, corresponding to the experiment shown in Fig ure 4 3 A. The microtubule is pinned at the plus end, but a similar result would follow from the frictional resistance of a n additional segment out of the field of view The simulation shows a small initi al buckle under the action of a continuous distribution of dynein motors. We observe a growth in the amplitude of the buckle with time, and a gradual pushing of the buckle towards the plus end. The simulation reproduces the key features of the experiment s hown in Fig ure 4 3 A and on a similar time scale. Variations in the assumed dynein off rate would be reflected by corresponding changes in the predicted timescale. Simulations of Centrosome C entering Simulations of centrosome centering are illust rated in Figure 4 9C and D We compare the dynamics of a radial array of polymerizing microtubules with and without the forces and friction from the dynein motors The parameters (motor density, motor friction, and friction from the background fluid) are t he same as in the single microtubule simulations. Simulations with motors show that microtubules are heavily buckled near the cell periphery whereas in the absence of motors the wavelen gths of the buckled microtubules are much longer. W ith dynein motors pu lling the microtubules, the centrosome centers with a time constant of about 10 min but simulations without motors suggest that polymerization forces are unable to center the centrosome under in vivo conditions. The centrosome remains essentially in place for the duration of the simulation (100 min).

PAGE 153

153 Figure 4 1 compares simulations of centrosome centering with experimental measurements of the fluctuations in centrosome position. The microtubule network develops by polymerization from a radial array of stubs (Figure 4 9C) during the first 5 min of the simulation. After the microtubule network has filled the cell, an initially off center centrosome moves towards the center with a relaxation time of approximately 12 min (time to decay to of its initial dis placement). The relaxation time in animal cells, measured from the autocorrelation function of fluctuations in centrosome position was about 8 min, with a standard deviation of 4 min (determined from 14 trajectories of 1 2 hours each). We have also simulat ed in vitro centering of an centrosome to compare with data in Ref. (77) In that work it was shown that polymerizing forces could center the centrosome by pushing from the cell boundary. The key changes in the simulation parameters were the cell size (12 m) and the viscosity of the background fluid (10 3 Pa s). In addition we eliminated microtubule pinning at the cell boundaries to model the slippage al ong the glass wall. The polymerization kinetics were tuned to produce microtubules with average lengths of approximately 20 and 40 m, again to correspond to the conditions of the experiments. The simulations reproduce the key featu res of the experiments (77) ; with the shorter (20 m) microtubules th e centrosome centers rapidly but with the longer (40 m) microtubules it drifts to a n off center location It can be seen that the longer microtubules buckle significantly and t herefore exert much less force than the shorter ones, which remain more or less straight. Under in vivo conditions, with a larger cell (40 m) and a larger viscosity of the background fluid (> 1 Pa s), polymerization forces are much too small to center the centrosome.

PAGE 154

154 It is interesting to compare the buckling of the microtubules with the in vivo sim ulation without motors Under in vivo conditions the microtubules buckle into higher order modes because the friction is so large that they do not relax to the m inimum energy state before the next segment polymerizes, whereas under in vitro conditions, with a much smaller viscosity, the microtubules buckle into the lowest order mode. Centrosome Relaxation Time The centrosome relaxation time can be estimated based on a model of linear rigid microtubule s. Assuming each microtubule is pinned at the cell periphery and the centrosome is moving with a velocity the lateral velocity of the microtubule segment at a distance l from the periphery is w here is the (time dependent) contour length of the microtubule; the tangential velocity is assumed to be constant since the microtubule is rigid. T he force exerted on the centrosome from a single microtubule is then found by integrating over the contour length ( D 8) Balancing the forces on the centrosome gives the velocity ( D 9) It follows from Equation D 9 that, when dynein motor forces dominate, the centrosome velocity is independent of dynein density. The time scale for centrosome centering can be estimated by considering a displacement of the centrosome from the cell center by a small distance along the x axis The force balance in the x direction is ( D 10)

PAGE 155

155 where represents the mean over a uniform angular distribution and is the length of the microtubule spanning the distance between the centrosome and the cell periphery. F or small displacements from the center of a circular cell, Evaluating the moments of in Equation D 10 to lowest order in ( D 11) Equation D 11 can be solved for to yield a linear relaxation equation for the displacement in terms of its time derivative where the relaxation time for centrosome cen tering is given by ( D 12) Significantly which is the characteristic time for the centrosome to center, is predicted to depend only on the cell size and molecular parameters not the number of microtubules. Different cell shapes only affect this result to a numerical pre factor. For the parameters we have chosen, min assuming a circular cell with the same area as the BCE cells.

PAGE 156

156 APPENDIX E FOR BLEACHING OF FREE PR OTEIN IN THE CYTOPLA SM The photobleaching rate constant will in general be different for the cytoplasm compared to a confined structure like the focal adhesion. This is because the confocal slice volume in which the majority of bleaching occurs does not encompass the whole cytoplasm, while it can encompass confined structures like the focal adhesion (~100 nm compared to a vertical cross section of the confocal plane of around 600 nm). A ssuming that the cytoplasm is well mixed, and that bleaching occurs in a smaller volume corresponding to the confocal slice volume, a balance on the free molecules during image capture yields (E 1) where is the photobleaching rate constant (s 1 ) for the free protein molecules and is the volume of the cytoplasm. Comparing this to equation (1) yields the effective bleaching constant for the cytoplasm: The effective bleaching rate constant is small for a large volume such as the cytoplasm due to the small value of For a confined structure like the focal adhesion or a promoter array in the nucleus, the rate constant will be larger as the volume of the structure is small (or comparable) compared to the confocal slice volume. Another consideration is the fact that fr ee molecules rapidly diffuse in and out of the confocal slice, resulting in less bleaching than would occur in a confined structure. Based on these considerations, the bleaching functions are assumed to be different for the free and bound species ( and ).

PAGE 157

157 APPENDIX F MATLAB CODE FOR FRAP MODELING AND SIMULAT ION These are Matlab program s for the modeling discussed in Chapter 5. All the notations are consistent through different programs unless otherwise specified. Simulation of Bleaching Dynamics in Image Capture Process T hese are the program s to generate Figure 5 1. Figure 5 2 used similar program. Comparison of Observed Dynamics and Actual Dynamics %Start of FRAPfig1A.m clear all close all clc m=12 ;%Number of frames a=0.9;%Value of alpha lamdaw2=0.2;%Value of lamda*omega interval=5;%Time intervals between frames omega=0.5;%Value of omega %observed dynmaics fw2=(exp(lamdaw2) 1)/lamdaw2;%function for intensity averaging due to image capture process g2=exp( lamd aw2);%function for bleaching due to image capture koff=0.2; tau=interval; % in seconds, this is the time between frames t=[0:1:m 1]; C(1) = a*fw2; for i = 2:m C(i)= a*g2^(i 1)*fw2; end figure(1) plot(t(2:m),C(2:m),'*k','markersize',12) %real dynamics s cale=tau/omega; tr=[0:0.5:(m 1)*tau]; Cr(1)=a; for i=1:m 1; for k=((i 1)*scale+2):((i 1)*scale+10); Cr(k)=Cr((i 1)*scale+1); end j=i*scale+1; Cr(j)=Cr(j 1)*g2; end hold on plot(tr/tau,Cr,':k','markersize',10) hold on for i=1:m 1; plot(tr((i 1)*scale+10)/tau,Cr((i 1)*scale+10),'ok','markersize',10)

PAGE 158

158 end Modeling of Normalized Bleaching Dynamics %Start of FRAPfig1B.m clear all close all clc m=12 ; a=1;%alpha starts at 1 because of normalization lamdaw2=0.2; interval=5; omega=0.5 ; %normalized observed intensity fw2=(exp(lamdaw2) 1)/lamdaw2; g2=exp( lamdaw2); koff=0.2; tau=interval; t=[0:1:m 1]; C(1)= 1; for i= 2:m C(i)= C(1)*g2^(i 1); end figure(1) plot(t(2:m),C(2:m),'*k','markersize',12) %normalized real dynamics scale=tau/omega; tr=[0:0.5:(m 1)*tau]; Cr(1)=1; for i=1:m 1; for k=((i 1)*scale+2):((i 1)*scale+10); Cr(k)=Cr((i 1)*scale+1); end j=i*scale+1; Cr(j)=Cr(j 1)*g2; end hold on plot(tr/tau,Cr,':k','markersize',10) hold on for i=1:m 1; plot(tr((i 1)*scale+10)/tau,Cr((i 1)*scale+10),'ok','markersize',10) end

PAGE 159

159 Simulation of FRAP This is the program used in Figure 5 3. The p rogram can be adjusted to simulate different and and immobile fraction Here we use as an example %Start of FRAPfig3A.m clear all close all clc m=11; a=0.4; b=0.3;%Value for immobile fraction lamdaw1=0.2;%Value of lamda*omega at the focal adhesion lamdaw2=0.000000000001;%Value of lamda*omega for the free protein C0=1;%this is normalized intensity before bleaching alpha=a; imfrac=b; beta=[1 imfrac imfrac];%mobile and immobile fraction g1=exp( lamdaw1);%bleaching function for focal adhesion g2=exp( lamdaw2);%bleaching function for free protein ktau=1;%Value of tau*koff frac=0;%ratio of koff for mobile and immobile, it's zero for completely immobile Ktau=[ktau ktau*frac]; t=0:1:m 1; Cm(1)=alpha*beta(1);%normalized intensity of the first frame after bleaching for mobile fraction Ci(1)=alpha*beta(2);%normalized intensity of the first frame after bleaching for immobile fraction for i = 2:m Cm(i)= C0*beta(1)*(g2^(i 2) (g2^(i 2) Cm(i 1)/beta(1))*exp( Ktau(1)))*g1;%normalized intensity of mobile fractio n Ci(i)= C0*beta(2)*(g2^(i 2) (g2^(i 2) Ci(i 1)/beta(2))*exp( Ktau(2)))*g1;%normalized intensity of immobile fraction end C=Cm+Ci;%normalized total fluorescence figure(1) plot(t,C,' *k','MarkerSize',10) Model Fitting of FRAP These are the programs used in model fitting in Figure 5 4. Define Fitting Function of FRAP Model This is for focal adhesion FRAP fitting

PAGE 160

160 %Start of FA3T3FRAPpara.m function sse=FA3T3FRAPpara(param,input1,input2,input3,input4,input5,data1) C0=input1;%define input parameter 1 to 5 al pha= input2; L=input3; tau=input4; imfrac=input5; lamdaw1=param(1); % define parameters 1 to 5 that need to be estimated by fitting lamdaw2=param(2); g1=exp( lamdaw1); g2=exp( lamdaw2); koff=param(3); beta=[1 imfrac imfrac]; Cm(1)=alpha*beta(1); Ci(1)=alpha*beta(2); for i = 2:L Cm(i)= C0*beta(1)*(g2^(i 2) (g2^(i 2) Cm(i 1)/beta(1))*exp( tau*koff))*g1; Ci(i)= C0*beta(2)*Ci(i 1)/beta(2)*g1; end C=Cm+Ci; Error1=C data1; sse=sum(Error1.^2);% When curvefitting, a typical quantity to minimize is the sum of squares error Fitting Focal Adhesion FRAP Experiment D ata to the Model %Start of FA3T3FRAPfitting clear all close all clc load 3T3FRAP020212cell16.mat%load data from experiment C0=1; %for FRAP without bleaching intra=intreca'/C0a;%normalize i ntensity by deviding experiment data by the intensity before bleaching t=time;%time series of the experiment alpha=intra(1); L=length(intra);%number of frames in the experiment tau=t(2) t(1);%time interval in the experiment imfrac=im;%iimmobile fraction calculated from independent experiment of the same focal adhesion Starting=[0.1, 0.1, exp( tau*0.2)];%initial guess for the fitting options=optimset('Display','iter'); Estimates=fminsearch(@FA3T3FRAPpara,Starting,options,C0,alpha,L,tau,imfrac,in tra);%call fitting function "FA3T3FRAPpara.m" %read values from fitting and plot the result Cm(1)=alpha*(1 imfrac); Ci(1)=alpha*imfrac; g1=exp( Estimates(1)); g2=exp( Estimates(2)); for i=2:L;

PAGE 161

161 Cm(i)=C0*(1 imfrac)*(g2^(i 2) (g2^(i 2) Cm(i 1)/(1 imfrac))*exp( tau*Estimates(3)))*g1; Ci(i)=C0*imfrac*Ci(i 1)/imfrac*g1; end lamdaw1a=Estimates(1); lamdaw2a=Estimates(2); g1a=g1; g2a=g2; koffa=Estimates(3); Ca=Cm+Ci; figure(1) plot(t+5,intra,'*k') hold on plot(t+5,Ca,'r') hold on plot(t(1),C0,'^k','markersize',10) Define Fitting Function for Free Molecule Intensity This is for bleaching of free protein fitting. %Start of bleachpara.m function sse=bleachpara(params,input1,input2,data1) intf=input1; L=input2; lamdaw2=params(1); g2=exp( lamdaw2); Cf(1)=intf(1); for i=2:L; Cf(i)=Cf(1)*g2^(i 1); end error=Cf data1; sse=sum(error.^2); Fitting Free Molecule Intensity in Focal Adhesion FRAP Experiments %Start of FRAPbleachingfitting.m clear all close a ll clc load 3T3FRAP020212cell16.mat intf=intb'/intb ; L=length(intf); t=time ; tau=t(2) t(1); Starting=rand(1,1); options=optimset('Display','iter'); Estimates=fminsearch(@bleachpara,Starting,options,intf,L,intf); lamdaw2=Estimates(1); g2=exp( Estimates(1)); Cf(1)=intf(1); for i=2:L;

PAGE 162

162 Cf(i)=Cf(1)*g2^(i 1); end plot(t,intf,'*') hold on plot(t,Cf,'r') Simulation of FRAP with Multiple Time Intervals This is the program used for simulation in Figure 5 6. %Start of MultiTausimu.m clear all close all clc m=10; a=0.2; lamdaw1=0.2; lamdaw2=0.000000000001; interval=2; omega=0.2; C0=1; alpha=a; g1=exp( lamdaw1); g2=exp( lamdaw2); koff=0.2; tau1=interval;%the first time interval t1=0:m; C(1)=alpha; alpha(1)=alpha; for i=2:m+1 C(i)= C0 *(g2^(i 2) (g2^(i 2) alpha(i 1))*exp( (tau1 omega)*koff))*g1; alpha(i)=(g2^(i 2) (g2^(i 2) alpha(i 1))*exp( (tau1 omega)*koff))*g1; end scale=tau1/omega; tr1=0:omega:m*tau1; Cr(1)=C(1); ralpha(1)=a; for i=1:m; for k=((i 1)*scale+2):(i*scale); Cr(k)=C0*(1 (1 ralpha(i))*exp( (tr1(k) (i 1)*tau1)*koff)); end j=i*scale+1; Cr(j)=Cr(j 1)*exp( lamdaw1); ralpha(i+1)=Cr(j)/C0; end tau2=2*interval;%the second time interval t2=m+1:2*m; alpha(m+1)=C(m+1)/C0;

PAGE 163

163 for i=m+2:2*m+1 C(i)= C0*(g2^(i 2) (g2^(i 2) alpha(i 1))*exp( (tau2 omega)*koff))*g1; alpha(i)=(g2^(i 2) (g2^(i 2) alpha(i 1))*exp( (tau2 omega)*koff))*g1; end scale2=tau2/omega; tr2=(m*tau1+omega):omega:(m*tau1+m*tau2); ralpha(m+1)=alpha(m+1); for i=m+1:2*m; for k=((i m 1)*scale2+2):((i m)*scale2); Cr(m*scale+k)=C0*(1 (1 ralpha(i))*exp( (tr2(k 1) m*tau1 (i m 1)*tau2)*koff)); end j=m*scale+(i m)*scale2+1; Cr(j)=Cr(j 1)*exp( lamdaw1); ralpha(i+1)=Cr(j)/C0; end tau3=3*interval;%the third time interval t3=2*m+1:3*m; alpha(2*m+1)=C(2*m+1)/C0; for i=2*m+2:3*m+1 C(i)= C0*(g2^(i 2) (g2^(i 2) alpha(i 1))*exp( (tau3 omega)*koff))*g1; alpha(i)=(g2^(i 2) (g2^(i 2) alpha(i 1))*exp( (tau3 omega)*koff))*g1; end scale3=tau3/omega; tr3=(m*tau1+m*tau2+omega):omega:(m*tau1+m*tau2+m*tau3); ralpha(2*m+1)=alpha(2*m); for i=2*m+1:3*m; for k=((i 2*m 1)*scale3+2):((i 2*m)*scale3); Cr(m*scale+m*scale2+k)=C0*(1 (1 ralpha(i))*exp( (tr3(k 1) m*tau1 m*tau2 (i 2*m 1)*tau3)*koff)); end j=m*scale+m*scale2+(i 2*m)*scale3+1; Cr(j)=Cr(j 1)*exp( lamdaw1); ralpha(i+1)=Cr(j)/C0; end ts=1:m; t=[tau1*t1 tau1*m+tau2*ts tau1*m+tau2*m+tau3*ts]; tr=[tr1 tr2 tr3]; plot(t+5,C,' *k','markersize',12) hold on plot(tr+5,Cr,':k') hold on plot(t(1),C0,'^k','markersize',10)

PAGE 164

164 LIST OF REFERENCES 1. Jaalouk, D. E., and J. Lammerding. 2009. Mechanotransduction gone awry. Nat Rev Mol Cell Biol. 10:63 73. 2. Starr, D. A. 2009. A nuclear envelope bridge positions nuclei a nd moves chromosomes. J Cell Sci. 122:577 586. 3. Zhang, X., R. Xu, B. Zhu, X. Yang, X. Ding, S. Duan, T. Xu, Y. Zhuang, and M. Han. 2007. Syne 1 and Syne 2 play crucial roles in myonuclear anchorage and motor neuron innervation. Development. 134:901 908. 4. Wang, N., J. D. Tytell, and D. E. Ingber. 2009. Mechanotransduction at a distance: mechanically coupling the extracellular matrix with the nucleus. Nat Rev Mol Cell Biol. 10:75 82. 5. Gomes, E. R., S. Jani, and G. G. Gundersen. 2005. Nuclear movement regulated by Cdc42, MRCK, myosin, and actin flow establishes MTOC polarization in migrating cells. Cell. 121:451 463. 6. Baye, L. M., and B. A. Link. 2008. Nuclear migration during retinal development. Brain Res. 1192:29 36. 7. Stearns, T. 1997. Motorin g to the finish: kinesin and dynein work together to orient the yeast mitotic spindle. J Cell Biol. 138:957 960. 8. Gros Louis, F., N. Dupre, P. Dion, M. A. Fox, S. Laurent, S. Verreault, J. R. Sanes, J. P. Bouchard, and G. A. Rouleau. 2007. Mutations in SYNE1 lead to a newly discovered form of autosomal recessive cerebellar ataxia. Nat Genet. 39:80 85. 9. Burke, B., and C. L. Stewart. 2002. Life at the edge: the nuclear envelope and human disease. Nat Rev Mol Cell Biol. 3:575 585. 10. Mattout, A., T. Dechat, S. A. Adam, R. D. Goldman, and Y. Gruenbaum. 2006. Nuclear lamins, diseases and aging. Curr Opin Ce ll Biol. 18:335 341. 11. Warren, D. T., Q. Zhang, P. L. Weissberg, and C. M. Shanahan. 2005. Nesprins: intracellular scaffolds that maintain cell architecture and coordinate cell function? Expert Rev Mol Med. 7:1 15. 12. Levy, J. R., and E. L. Holzbaur. 2008. Dynein drives nuclear rotation during forward progression of motile fibroblasts. J Cell Sci. 121:3187 3195. 13. Zhou, K., M. M. Rolls, D. H. Hall, C. J. Malone, and W. Hanna Rose. 2009. A ZYG 12 dynein interaction at the nuclear envelope defines cyt oskeletal architecture in the C. elegans gonad. J Cell Biol. 186:229 241.

PAGE 165

165 14. Fridolfsson, H. N., N. Ly, M. Meyerzon, and D. A. Starr. 2010. UNC 83 coordinates kinesin 1 and dynein activities at the nuclear envelope during nuclear migration. Dev Biol. 338: 237 250. 15. Crisp, M., Q. Liu, K. Roux, J. B. Rattner, C. Shanahan, B. Burke, P. D. Stahl, and D. Hodzic. 2006. Coupling of the nucleus and cytoplasm: role of the LINC complex. J Cell Biol. 172:41 53. 16. Starr, D. A. 2011. Watching nuclei move: Insight s into how kinesin 1 and dynein function together. Bioarchitecture. 1:9 13. 17. Worman, H. J., and G. G. Gundersen. 2006. Here come the SUNs: a nucleocytoskeletal missing link. Trends Cell Biol. 16:67 69. 18. Starr, D. A., and J. A. Fischer. 2005. KASH n Karry: the KASH domain family of cargo specific cytoskeletal adaptor proteins. Bioessays. 27:1136 1146. 19. Wilhelmsen, K., M. Ketema, H. Truong, and A. Sonnenberg. 2006. KASH domain proteins in nuclear migration, anchorage and other processes. J Cell S ci. 119:5021 5029. 20. Tzur, Y. B., K. L. Wilson, and Y. Gruenbaum. 2006. SUN domain proteins: 'Velcro' that links the nucleoskeleton to the cytoskeleton. Nat Rev Mol Cell Biol. 7:782 788. 21. Zhang, Q., J. N. Skepper, F. Yang, J. D. Davies, L. Hegyi, R. G. Roberts, P. L. Weissberg, J. A. Ellis, and C. M. Shanahan. 2001. Nesprins: a novel family of spectrin repeat containing proteins that localize to the nuclear membrane in multiple tissues. J Cell Sci. 114:4485 4498. 22. Padmakumar, V. C., T. Libotte, W. Lu, H. Zaim, S. Abraham, A. A. Noegel, J. Gotzmann, R. Foisner, and I. Karakesisoglou. 2005. The inner nuclear membrane protein Sun1 mediates the anchorage of Nesprin 2 to the nuclear envelope. J Cell Sci. 118:3419 3430. 23. Zhang, Q., C. D. Ragnauth, J. N. Skepper, N. F. Worth, D. T. Warren, R. G. Roberts, P. L. Weissberg, J. A. Ellis, and C. M. Shanahan. 2005. Nesprin 2 is a multi isomeric protein that binds lamin and emerin at the nuclear envelope and forms a subcellula r network in skeletal muscle. J Cell Sci. 118:673 687. 24. Hale, C. M., A. L. Shrestha, S. B. Khatau, P. J. Stewart Hutchinson, L. Hernandez, C. L. Stewart, D. Hodzic, and D. Wirtz. 2008. Dysfunctional connections between the nucleus and the actin and mic rotubule networks in laminopathic models. Biophys J. 95:5462 5475.

PAGE 166

166 25. Houben, F., C. H. Willems, I. L. Declercq, K. Hochstenbach, M. A. Kamps, L. H. Snoeckx, F. C. Ramaekers, and J. L. Broers. 2009. Disturbed nuclear orientation and cellular migration in A type lamin deficient cells. Biochim Biophys Acta. 1793:312 324. 26. Huxley, H. E. 1957. The double array of filaments in cross striated muscle. J Biophys Biochem Cytol. 3:631 648. 27. Burkhardt, J. K., C. J. Echeverri, T. Nilsson, and R. B. Vallee. 19 97. Overexpression of the dynamitin (p50) subunit of the dynactin complex disrupts dynein dependent maintenance of membrane organelle distribution. J Cell Biol. 139:469 484. 28. Corthesy Theulaz, I., A. Pauloin, and S. R. Pfeffer. 1992. Cytoplasmic dynein participates in the centrosomal localization of the Golgi complex. J Cell Biol. 118:1333 1345. 29. Ebneth, A., R. Godemann, K. Stamer, S. Illenberger, B. Trinczek, and E. Mandelkow. 1998. Overexpression of tau protein inhibits kinesin dependent trafficki ng of vesicles, mitochondria, and endoplasmic reticulum: implications for Alzheimer's disease. J Cell Biol. 143:777 794. 30. Pilling, A. D., D. Horiuchi, C. M. Lively, and W. M. Saxton. 2006. Kinesin 1 and Dynein are the primary motors for fast transport of mitochondria in Drosophila motor axons. Mol Biol Cell. 17:2057 2068. 31. Roux, K. J., M. L. Crisp, Q. Liu, D. Kim, S. Kozlov, C. L. Stewart, and B. Burke. 2009. Nesprin 4 is an outer nuclear membrane protein that can induce kinesin mediated cell polari zation. Proc Natl Acad Sci U S A. 106:2194 2199. 32. Tsai, J. W., K. H. Bremner, and R. B. Vallee. 2007. Dual subcellular roles for LIS1 and dynein in radial neuronal migration in live brain tissue. Nat Neurosci. 10:970 979. 33. Fridolfsson, H. N., and D A. Starr. 2010. Kinesin 1 and dynein at the nuclear envelope mediate the bidirectional migrations of nuclei. J Cell Biol. 191:115 128. 34. Lye, R. J., M. E. Porter, J. M. Scholey, and J. R. McIntosh. 1987. Identification of a microtubule based cytoplasm ic motor in the nematode C. elegans. Cell. 51:309 318. 35. Paschal, B. M., S. M. King, A. G. Moss, C. A. Collins, R. B. Vallee, and G. B. Witman. 1987. Isolated flagellar outer arm dynein translocates brain microtubules in vitro. Nature. 330:672 674.

PAGE 167

167 36. Paschal, B. M., H. S. Shpetner, and R. B. Vallee. 1987. MAP 1C is a microtubule activated ATPase which translocates microtubules in vitro and has dynein like properties. J Cell Biol. 105:1273 1282. 37. Vale, R. D. 2003. The molecular motor toolbox for in tracellular transport. Cell. 112:467 480. 38. Asai, D. J., and M. P. Koonce. 2001. The dynein heavy chain: structure, mechanics and evolution. Trends Cell Biol. 11:196 202. 39. Gennerich, A., and R. D. Vale. 2009. Walking the walk: how kinesin and dynein coordinate their steps. Curr Opin Cell Biol. 21:59 67. 40. Reck Peterson, S. L., A. Yildiz, A. P. Carter, A. Gennerich, N. Zhang, and R. D. Vale. 2006. Single molecule analysis of dynein processivity and stepping behavior. Cell. 126:335 348. 41. Mosley Bishop, K. L., Q. Li, L. Patterson, and J. A. Fischer. 1999. Molecular analysis of the klarsicht gene and its role in nuclear migration within differentiating cells of the Drosophila eye. Curr Biol. 9:1211 1220. 42. Kracklauer, M. P., S. M. Banks, X. Xie, Y. Wu, and J. A. Fischer. 2007. Drosophila klaroid encodes a SUN domain protein required for Klarsicht localization to the nuclear envelope and nuclear migration in the eye. Fly (Austin). 1:75 85. 43. Malone, C. J., L. Misner, N. Le Bot, M. C. Tsai, J. M Campbell, J. Ahringer, and J. G. White. 2003. The C. elegans hook protein, ZYG 12, mediates the essential attachment between the centrosome and nucleus. Cell. 115:825 836. 44. Albrecht Buehler, G. 1984. Movement of Nucleus and Centrosphere in 3T3 Cells. In Cancer Cells: The transformed phenotype. A. J. Levine, G. F. V. Woude, W. C. Toop, and J. D. Watson, editors. Cold Spring Harbor Laboratory, Newyork. 87~96. 45. Allen, V. W., and D. L. Kropf. 1992. Nuclear rotation and lineage specification in Pelveti a embryos. Development. 11. 46. Ji, J. Y., R. T. Lee, L. Vergnes, L. G. Fong, C. L. Stewart, K. Reue, S. G. Young, Q. Zhang, C. M. Shanahan, and J. Lammerding. 2007. Cell nuclei spin in the absence of lamin b1. J Biol Chem. 282:20015 20026. 47. Pomerat, C M. 1953. Rotating nuclei in tissue cultures of adult human nasal mucosa. Exp Cell Res. 5:191 196.

PAGE 168

168 48. Lee, J. S., M. I. Chang, Y. Tseng, and D. Wirtz. 2005. Cdc42 mediates nucleus movement and MTOC polarization in Swiss 3T3 fibroblasts under mechanical shear stress. Mol Biol Cell. 16:871 880. 49. Gerashchenko, M. V., I. S. Chernoivanenko, M. V. Moldaver, and A. A. Minin. 2009. Dynein is a motor for nuclear rotation while vimentin IFs is a "brake". Cell Biol Int. 33:1057 1064. 50. Friedl, P., K. Wolf, and J. Lammerding. 2011. Nuclear mechanics during cell migration. Curr Opin Cell Biol. 23:55 64. 51. Vicente Manzanares, M., K. Newell Litwa, A. I. Bachir, L. A. Whitmore, and A. R. Horwitz. 2011. Myosin IIA/IIB restrict adhesive and protrusive signaling to generate front back polarity in migrating cells. J Cell Biol. 193:381 396. 52. Howard, J. 2001. Mechanics of Motor Proteins and the Cytoskeleton. Sinauer, Sunderland, MA. 53. Huxley, A. F. 1957. Muscle structure and theories of co ntraction. Prog Biophys Biophys Chem. 7:255 318. 54. Huxley, A. F., and R. Niedergerke. 1954. Structural changes in muscle during contraction; interference microscopy of living muscle fibres. Nature. 173:971 973. 55. Huxley, H., and J. Hanson. 1954. Chan ges in the cross striations of muscle during contraction and stretch and their structural interpretation. Nature. 173:973 976. 56. Folker, E. S., C. Ostlund, G. W. Luxton, H. J. Worman, and G. G. Gundersen. 2011. Lamin A variants that cause striated muscl e disease are defective in anchoring transmembrane actin associated nuclear lines for nuclear movement. Proc Natl Acad Sci U S A. 108:131 136. 57. Luxton, G. W., E. R. Gomes, E. S. Folker, E. Vintinner, and G. G. Gundersen. 2010. Linear arrays of nuclear envelope proteins harness retrograde actin flow for nuclear movement. Science. 329:956 959. 58. Martini, F. J., and M. Valdeolmillos. 2010. Actomyosin contraction at the cell rear drives nuclear translocation in migrating cortical interneurons. J Neurosci. 30:8660 8670. 59. Chancellor, T. J., J. Lee, C. K. Thodeti, and T. Lele. 2010. Actomyosin tension exerted on the nucleus through nesprin 1 connections influences endothelial cell adhesion, migration, and cyclic strain induced reorientation. Biop hys J. 99:115 123.

PAGE 169

169 60. Sims, J. R., S. Karp, and D. E. Ingber. 1992. Altering the cellular mechanical force balance results in integrated changes in cell, cytoskeletal and nuclear shape. J Cell Sci. 103 ( Pt 4):1215 1222. 61. Maniotis, A. J., C. S. Chen, and D. E. Ingber. 1997. Demonstration of mechanical connections between integrins, cytoskeletal filaments, and nucleoplasm that stabilize nuclear structure. Proc Natl Acad Sci U S A. 94:849 854. 62. Buxboim, A., I. L. Ivanovska, and D. E. Discher. 2010. Matrix elasticity, cytoskeletal forces and physics of the nucleus: how deeply do cells 'feel' outside and in? J Cell Sci. 123:297 308. 63. Wu, Y. I., D. Frey, O. I. Lungu, A. Jaehrig, I. Schlichting, B. Kuhlman, and K. M. Hahn. 2009. A genetically encoded photoactivatable Rac controls the motility of living cells. Nature. 461:104 108. 64. Tassin, A. M., and M. Bornens. 1999. Centrosome struc ture and microtubule nucleation in animal cells. Biol Cell. 91:343 354. 65. Vernos, I., and E. Karsenti. 1996. Motors involved in spindle assembly and chromosome segregation. Curr Opin Cell Biol. 8:4 9. 66. Waters, J. C., R. W. Cole, and C. L. Rieder. 19 93. The force producing mechanism for centrosome separation during spindle formation in vertebrates is intrinsic to each aster. J Cell Biol. 122:361 372. 67. Cole, N. B., and J. Lippincott Schwartz. 1995. Organization of organelles and membrane traffic by microtubules. Curr Opin Cell Biol. 7:55 64. 68. Danowski, B. A., A. Khodjakov, and P. Wadsworth. 2001. Centrosome behavior in motile HGF treated PtK2 cells expressing GFP gamma tubulin. Cell Motil Cytoskeleton. 50:59 68. 69. Savoian, M. S., and C. L. Ri eder. 2002. Mitosis in primary cultures of Drosophila melanogaster larval neuroblasts. J Cell Sci. 115:3061 3072. 70. Verstraeten, V. L., L. A. Peckham, M. Olive, B. C. Capell, F. S. Collins, E. G. Nabel, S. G. Young, L. G. Fong, and J. Lammerding. 2011. Protein farnesylation inhibitors cause donut shaped cell nuclei attributable to a centrosome separation defect. Proc Natl Acad Sci U S A. 108:4997 5002. 71. Salmon, W. C., M. C. Adams, and C. M. Waterman Storer. 2002. Dual wavelength fluorescent speckle m icroscopy reveals coupling of microtubule and actin movements in migrating cells. J Cell Biol. 158:31 37.

PAGE 170

170 72. Waterman Storer, C. M., and E. D. Salmon. 1997. Actomyosin based retrograde flow of microtubules in the lamella of migrating epithelial cells inf luences microtubule dynamic instability and turnover and is associated with microtubule breakage and treadmilling. J Cell Biol. 139:417 434. 73. Brangwynne, C. P., F. C. MacKintosh, S. Kumar, N. A. Geisse, J. Talbot, L. Mahadevan, K. K. Parker, D. E. Ingb er, and D. A. Weitz. 2006. Microtubules can bear enhanced compressive loads in living cells because of lateral reinforcement. J Cell Biol. 173:733 741. 74. Inoue, S., and E. D. Salmon. 1995. Force generation by microtubule assembly/disassembly in mitosis and related movements. Mol Biol Cell. 6:1619 1640. 75. Tran, P. T., L. Marsh, V. Doye, S. Inoue, and F. Chang. 2001. A mechanism for nuclear positioning in fission yeast based on microtubule pushing. J Cell Biol. 153:397 411. 76. Howard, J. 2006. Elastic and damping forces generated by confined arrays of dynamic microtubules. Phys Biol. 3:54 66. 77. Holy, T. E., M. Dogterom, B. Yurke, and S. Leibler. 1997. Assembly and positioning of microtubule asters in microfabricated chambers. Proc Natl Acad Sci U S A. 94:6228 6231. 78. Burakov, A., E. Nadezhdina, B. Slepchenko, and V. Rodionov. 2003. Centrosome positioning in interphase cells. J Cell Biol. 162:963 969. 79. Dujardin, D. L., and R. B. Vallee. 2002. Dynein at the cortex. Curr O pin Cell Biol. 14:44 49. 80. Zhu, J., A. Burakov, V. Rodionov, and A. Mogilner. 2010. Finding the cell center by a balance of dynein and myosin pulling and microtubule pushing: a computational study. Mol Biol Cell. 21:4418 4427. 81. Vogel, S. K., N. Pavi n, N. Maghelli, F. Julicher, and I. M. Tolic Norrelykke. 2009. Self organization of dynein motors generates meiotic nuclear oscillations. PLoS Biol. 7:e1000087. 82. Bicek, A. D., E. Tuzel, A. Demtchouk, M. Uppalapati, W. O. Hancock, D. M. Kroll, and D. J. Odde. 2009. Anterograde microtubule transport drives microtubule bending in LLC PK1 epithelial cells. Mol Biol Cell. 20:2943 2953. 83. Ingber, D. E. 2003. Tensegrity I. Cell structure and hierarchical systems biology. J Cell Sci. 116:1157 1173.

PAGE 171

171 84. Pala zzo, A. F., H. L. Joseph, Y. J. Chen, D. L. Dujardin, A. S. Alberts, K. K. Pfister, R. B. Vallee, and G. G. Gundersen. 2001. Cdc42, dynein, and dynactin regulate MTOC reorientation independent of Rho regulated microtubule stabilization. Curr Biol. 11:1536 1541. 85. Kaufman, E. N., and R. K. Jain. 1990. Quantification of transport and binding parameters using fluorescence recovery after photobleaching. Potential for in vivo applications. Biophys J. 58:873 885. 86. Kaufman, E. N., and R. K. Jain. 1991. Meas urement of mass transport and reaction parameters in bulk solution using photobleaching. Reaction limited binding regime. Biophys J. 60:596 610. 87. Mueller, F., T. S. Karpova, D. Mazza, and J. G. McNally. 2012. Monitoring dynamic binding of chromatin pro teins in vivo by fluorescence recovery after photobleaching. Methods Mol Biol. 833:153 176. 88. Mueller, F., D. Mazza, T. J. Stasevich, and J. G. McNally. 2010. FRAP and kinetic modeling in the analysis of nuclear protein dynamics: what do we really know? Curr Opin Cell Biol. 22:403 411. 89. Phair, R. D., S. A. Gorski, and T. Misteli. 2004. Measurement of dynamic protein binding to chromatin in vivo, using photobleaching microscopy. Methods Enzymol. 375:393 414. 90. Phair, R. D., and T. Misteli. 2001. Kinetic modelling approaches to in vivo imaging. Nat Rev Mol Cell Biol. 2:898 907. 91. Stenoien, D. L., K. Patel, M. G. Mancini, M. Dutertre, C. L. Smith, B. W. O'Malley, and M. A. Mancini. 2001. FRAP reveals that mobil ity of oestrogen receptor alpha is ligand and proteasome dependent. Nat Cell Biol. 3:15 23. 92. Wagner, S., S. Chiosea, M. Ivshina, and J. A. Nickerson. 2004. In vitro FRAP reveals the ATP dependent nuclear mobilization of the exon junction complex prote in SRm160. J Cell Biol. 164:843 850. 93. Deschout, H., J. Hagman, S. Fransson, J. Jonasson, M. Rudemo, N. Loren, and K. Braeckmans. 2010. Straightforward FRAP for quantitative diffusion measurements with a laser scanning microscope. Opt Express. 18:22886 22905. 94. Smisdom, N., K. Braeckmans, H. Deschout, M. vandeVen, J. M. Rigo, S. C. De Smedt, and M. Ameloot. 2011. Fluorescence recovery after photobleaching on the confocal laser scanning microscope: generalized model without restriction on the size of t he photobleached disk. J Biomed Opt. 16:046021.

PAGE 172

172 95. Stasevich, T. J., F. Mueller, A. Michelman Ribeiro, T. Rosales, J. R. Knutson, and J. G. McNally. 2010. Cross validating FRAP and FCS to quantify the impact of photobleaching on in vivo binding estimates Biophys J. 99:3093 3101. 96. Sprague, B. L., R. L. Pego, D. A. Stavreva, and J. G. McNally. 2004. Analysis of binding reactions by fluorescence recovery after photobleaching. Biophys J. 86:3473 3495. 97. Sprague, B. L., and J. G. McNally. 2005. FRAP an alysis of binding: proper and fitting. Trends Cell Biol. 15:84 91. 98. Braeckmans, K., K. Remaut, R. E. Vandenbroucke, B. Lucas, S. C. De Smedt, and J. Demeester. 2007. Line FRAP with the confocal laser scanning microscope for diffusion measurements in sm all regions of 3 D samples. Biophys J. 92:2172 2183. 99. Kang, M., C. A. Day, E. DiBenedetto, and A. K. Kenworthy. 2010. A quantitative approach to analyze binding diffusion kinetics by confocal FRAP. Biophys J. 99:2737 2747. 100. Kang, M., C. A. Day, K. Drake, A. K. Kenworthy, and E. DiBenedetto. 2009. A generalization of theory for two dimensional fluorescence recovery after photobleaching applicable to confocal laser scanning microscopes. Biophys J. 97:1501 1511. 101. Kang, M., and A. K. Kenworthy. 20 08. A closed form analytic expression for FRAP formula for the binding diffusion model. Biophys J. 95:L13 15. 102. van Royen, M. E., P. Farla, K. A. Mattern, B. Geverts, J. Trapman, and A. B. Houtsmuller. 2009. Fluorescence recovery after photobleaching ( FRAP) to study nuclear protein dynamics in living cells. Methods Mol Biol. 464:363 385. 103. Russell, R. J., S. L. Xia, R. B. Dickinson, and T. P. Lele. 2009. Sarcomere mechanics in capillary endothelial cells. Biophys J. 97:1578 1585. 104. Chancellor, T. J., J. Lee, C. K. Thodeti, and T. Lele. Actomyosin tension exerted on the nucleus through nesprin 1 connections influences endothelial cell adhesion, migration, and cyclic strain induced reorientation. Biophys J. 99:115 123. 105. Fink, J., M. Thery, A. Azioune, R. Dupont, F. Chatelain, M. Bornens, and M. Piel. 2007. Comparative study and improvement of current cell micro patterning techniques. Lab Chip. 7:672 680. 106. Paddock, S. W., and G. Albrecht Buehler. 1988. Rigidity of the nucleus during nuclear rotation in 3T3 cells. Exp Cell Res. 175:409 413.

PAGE 173

173 107. Quintyne, N. J., S. R. Gill, D. M. Eckley, C. L. Crego, D. A. Compton, and T. A. Schroer. 1999. Dynactin is required for microtubule anchoring at centrosomes. J Cell Biol. 147:3 21 334. 108. King, S. J., C. L. Brown, K. C. Maier, N. J. Quintyne, and T. A. Schroer. 2003. Analysis of the dynein dynactin interaction in vitro and in vivo. Mol Biol Cell. 14:5089 5097. 109. Toba, S., T. M. Watanabe, L. Yamaguchi Okimoto, Y. Y. Toyoshi ma, and H. Higuchi. 2006. Overlapping hand over hand mechanism of single molecular motility of cytoplasmic dynein. Proc Natl Acad Sci U S A. 103:5741 5745. 110. Khatau, S. B., C. M. Hale, P. J. Stewart Hutchinson, M. S. Patel, C. L. Stewart, P. C. Searson D. Hodzic, and D. Wirtz. 2009. A perinuclear actin cap regulates nuclear shape. Proc Natl Acad Sci U S A. 106:19017 19022. 111. Yamada, M., S. Toba, T. Takitoh, Y. Yoshida, D. Mori, T. Nakamura, A. H. Iwane, T. Yanagida, H. Imai, L. Y. Yu Lee, T. Schroe r, A. Wynshaw Boris, and S. Hirotsune. mNUDC is required for plus end directed transport of cytoplasmic dynein and dynactins by kinesin 1. EMBO J. 29:517 531. 112. Thery, M., V. Racine, M. Piel, A. Pepin, A. Dimitrov, Y. Chen, J. B. Sibarita, and M. Borne ns. 2006. Anisotropy of cell adhesive microenvironment governs cell internal organization and orientation of polarity. Proc Natl Acad Sci U S A. 103:19771 19776. 113. Brock, A. L., and D. E. Ingber. 2005. Control of the direction of lamellipodia extension through changes in the balance between Rac and Rho activities. Mol Cell Biomech. 2:135 143. 114. Evans, E. 2001. Probing the relation between force -lifetime -and chemistry in single molecular bonds. Annu Rev Biophys Biomol Struct. 30:105 128. 115. Genn erich, A., A. P. Carter, S. L. Reck Peterson, and R. D. Vale. 2007. Force induced bidirectional stepping of cytoplasmic dynein. Cell. 131:952 965. 116. Gliksman, N. R., R. V. Skibbens, and E. D. Salmon. 1993. How the transition frequencies of microtubule dynamic instability (nucleation, catastrophe, and rescue) regulate microtubule dynamics in interphase and mitosis: analysis using a Monte Carlo computer simulation. Mol Biol Cell. 4:1035 1050. 117. Shelden, E., and P. Wadsworth. 1993. Observation and quantification of individual microtubule behavior in vivo: microtubule dynamics are cell type specific. J Cell Biol. 120:935 945.

PAGE 174

174 118. Harper, S. M., L. C. Neil, and K. H. Gardner. 2003. Structural basi s of a phototropin light switch. Science. 301:1541 1544. 119. Yao, X., M. K. Rosen, and K. H. Gardner. 2008. Estimation of the available free energy in a LOV2 J alpha photoswitch. Nat Chem Biol. 4:491 497. 120. Wang, X., L. He, Y. I. Wu, K. M. Hahn, and D. J. Montell. 2010. Light mediated activation reveals a key role for Rac in collective guidance of cell movement in vivo. Nat Cell Biol. 12:591 597. 121. Machacek, M., L. Hodgson, C. Welch, H. Elliott, O. Pertz, P. Nalbant, A. Abell, G. L. Johnson, K. M. Hahn, and G. Danuser. 2009. Coordination of Rho GTPase activities during cell protrusion. Nature. 461:99 103. 122. Lauffenburger, D. A., and A. F. Horwitz. 1996. Cell migration: a physically integrated molecular process. Cell. 84:359 369. 123. Chen, W. T. 1981. Mechanism of retraction of the trailing edge during fibroblast movement. J Cell Biol. 90:187 200. 124. Amy, R. L., and R. Storb. 1965. Selective mitochondrial damage by a ruby laser microbeam: an electron microscopic study. Science. 150:756 758. 125. Kumar, S., I. Z. Maxwell, A. Heisterkamp, T. R. Polte, T. P. Lele, M. Salanga, E. Mazur, and D. E. Ingber. 2006. Viscoelastic retraction of single living stress fibers and its impact on cell shape, cytoskeletal organization, and extracellular matrix mechanics. Biophys J. 90:3762 3773. 126. Stachowiak, M. R., and B. O'Shaughnessy. 2009. Recoil after severing reveals stress fiber contraction mechanisms. Biophys J. 97:462 471. 127. Tanner, K., A. Boudreau, M. J. Bissell, and S. Kumar. 2010. Dissecting regional variations in stress fiber mechanics in living cells with laser nanosurgery. Biophys J. 99:2775 2783. 128. Heisterkamp, A., I. Z. Maxwell, E. Mazur, J. M. Underwood, J. A. Nickerson, S. Kumar, and D. E. Ingber. 2005. Pulse energy dependence of su bcellular dissection by femtosecond laser pulses. Opt Express. 13:3690 3696. 129. Shen, N., D. Datta, C. B. Schaffer, P. LeDuc, D. E. Ingber, and E. Mazur. 2005. Ablation of cytoskeletal filaments and mitochondria in live cells using a femtosecond laser n anoscissor. Mech Chem Biosyst. 2:17 25.

PAGE 175

175 130. Berns, M. W., J. Aist, J. Edwards, K. Strahs, J. Girton, P. McNeill, J. B. Rattner, M. Kitzes, M. Hammer Wilson, L. H. Liaw, A. Siemens, M. Koonce, S. Peterson, S. Brenner, J. Burt, R. Walter, P. J. Bryant, D. van Dyk, J. Coulombe, T. Cahill, and G. S. Ber ns. 1981. Laser microsurgery in cell and developmental biology. Science. 213:505 513. 131. Gabel, C. V. 2008. Femtosecond lasers in biology: nanoscale surgery with ultrafast optics. Contemporary Physics. 49(6):391 411. 132. Colombelli, J., A. Besser, H. Kress, E. G. Reynaud, P. Girard, E. Caussinus, U. Haselmann, J. V. Small, U. S. Schwarz, and E. H. Stelzer. 2009. Mechanosensing in actin stress fibers revealed by a close correlation between force and protein localization. J Cell Sci. 122:1665 1679. 133. Lele, T. P., J. Pendse, S. Kumar, M. Salanga, J. Karavitis, and D. E. Ingber. 2006. Mechanical forces alter zyxin unbinding kinetics within focal adhesions of living cells. J Cell Physiol. 207:187 194. 134. Farhadifar, R., J. C. Roper, B. Aigouy, S. Eato n, and F. Julicher. 2007. The influence of cell mechanics, cell cell interactions, and proliferation on epithelial packing. Curr Biol. 17:2095 2104. 135. Tinevez, J. Y., U. Schulze, G. Salbreux, J. Roensch, J. F. Joanny, and E. Paluch. 2009. Role of corti cal tension in bleb growth. Proc Natl Acad Sci U S A. 106:18581 18586. 136. Mayer, M., M. Depken, J. S. Bois, F. Julicher, and S. W. Grill. 2010. Anisotropies in cortical tension reveal the physical basis of polarizing cortical flows. Nature. 467:617 621. 137. Moore, J. K., V. Magidson, A. Khodjakov, and J. A. Cooper. 2009. The spindle position checkpoint requires positional feedback from cytoplasmic microtubules. Curr Biol. 19:2026 2030. 138. Stiess, M., N. Maghelli, L. C. Kapitein, S. Gomis Ruth, M. Wi lsch Brauninger, C. C. Hoogenraad, I. M. Tolic Norrelykke, and F. Bradke. 2010. Axon extension occurs independently of centrosomal microtubule nucleation. Science. 327:704 707. 139. Colombelli, J., E. G. Reynaud, J. Rietdorf, R. Pepperkok, and E. H. Stelz er. 2005. In vivo selective cytoskeleton dynamics quantification in interphase cells induced by pulsed ultraviolet laser nanosurgery. Traffic. 6:1093 1102. 140. Wakida, N. M., C. S. Lee, E. T. Botvinick, L. Z. Shi, A. Dvornikov, and M. W. Berns. 2007. Las er nanosurgery of single microtubules reveals location dependent depolymerization rates. J Biomed Opt. 12:024022.

PAGE 176

176 141. Brangwynne, C. P., G. H. Koenderink, E. Barry, Z. Dogic, F. C. MacKintosh, and D. A. Weitz. 2007. Bending dynamics of fluctuating biopoly mers probed by automated high resolution filament tracking. Biophys J. 93:346 359. 142. Bicek, A. D., E. Tuzel, D. M. Kroll, and D. J. Odde. 2007. Analysis of microtubule curvature. Methods Cell Biol. 83:237 268. 143. Tao, W., R. J. Walter, and M. W. Ber ns. 1988. Laser transected microtubules exhibit individuality of regrowth, however most free new ends of the microtubules are stable. J Cell Biol. 107:1025 1035. 144. Brangwynne, C. P., F. C. MacKintosh, and D. A. Weitz. 2007. Force fluctuations and polymerization dynamics of intracellular microtubules. Proc Natl Acad Sci U S A. 104:16128 16133. 145. Yamada, M., S. Toba, T. Takitoh, Y. Yoshida, D. Mori, T. Nakamura, A. H. Iwane, T. Yanagida, H. Imai, L. Y. Yu Lee, T. Schroer, A. Wynshaw Boris, and S. Hirotsune. 2010. mNUDC is required for plus end directed transport of cytoplasmic dynein and dynactins by kinesin 1. EMBO J. 29:517 531. 146. Landau, L. D., and E. M. Lifs hitz. 1986. Theory of elasticity (Course of Theoretical Phisics). Butterworth Heinemann. 147. Manneville, J. B., M. Jehanno, and S. Etienne Manneville. 2010. Dlg1 binds GKAP to control dynein association with microtubules, centrosome positioning, and cell polarity. J Cell Biol. 191:585 598. 148. Mitchison, T., and M. Kirschner. 1984. Dynamic instability of microtubule growth. Nature. 312:237 242. 149. Dogterom, M., J. W. Kerssemakers, G. Romet Lemonne, and M. E. Janson. 2005. Force generation by dynamic microtubules. Curr Opin Cell Biol. 17:67 74. 150. Dogterom, M., and B. Yurke. 1998. Microtubule Dynamics and the Positioning of Microtubule Organizing Centers. Phys. Rev. Lett. 81:485 488. 151. Foethke, D., T. Makushok, D. Brunner, and F. Nedelec. 2009. Force and length dependent catastrophe activities explain interphase microtubule organization in fission yeast. Mol Syst Biol. 5:241. 152. Manneville, J. B., and S. Etienne Manneville. 2006. Pos itioning centrosomes and spindle poles: looking at the periphery to find the centre. Biol Cell. 98:557 565. 153. Pinot, M., F. Chesnel, J. Z. Kubiak, I. Arnal, F. J. Nedelec, and Z. Gueroui. 2009. Effects of confinement on the self organization of microtu bules and motors. Curr Biol. 19:954 960.

PAGE 177

177 154. Tolic Norrelykke, I. M. 2010. Force and length regulation in the microtubule cytoskeleton: lessons from fission yeast. Curr Opin Cell Biol. 22:21 28. 155. Lindemann, C. B., and A. J. Hunt. 2003. Does axonemal dynein push, pull, or oscillate? Cell Motil Cytoskeleton. 56:237 244. 156. Benson, D. M., J. Bryan, A. L. Plant, A. M. Gotto, Jr., and L. C. Smith. 1985. Digital imaging fluorescence microscopy: spatial heterogeneity of photobleaching rate constants in in dividual cells. J Cell Biol. 100:1309 1323. 157. Rodrigues, I., and J. Sanches. 2010. Denoising of LSFCM images with compensation for the photoblinking/photobleaching effects. Conf Proc IEEE Eng Med Biol Soc. 2010:4292 4295. 158. Dragestein, K. A., W. A. van Cappellen, J. van Haren, G. D. Tsibidis, A. Akhmanova, T. A. Knoch, F. Grosveld, and N. Galjart. 2008. Dynamic behavior of GFP CLIP 170 reveals fast protein turnover on microtubule plus ends. J Cell Biol. 180:729 737. 159. Lel e, T. P., and D. E. Ingber. 2006. A mathematical model to determine molecular kinetic rate constants under non steady state conditions using fluorescence recovery after photobleaching (FRAP). Biophys Chem. 120:32 35. 160. Lele, T. P., C. K. Thodeti, J. Pe ndse, and D. E. Ingber. 2008. Investigating complexity of protein protein interactions in focal adhesions. Biochem Biophys Res Commun. 369:929 934. 161. Sharp, Z. D., M. G. Mancini, C. A. Hinojos, F. Dai, V. Berno, A. T. Szafran, K. P. Smith, T. P. Lele, D. E. Ingber, and M. A. Mancini. 2006. Estrogen receptor alpha exchange and chromatin dynamics are ligand and domain dependent. J Cell Sci. 119:4101 4116. 162. Bulinski, J. C., D. J. Odde, B. J. Howell, T. D. Salmon, and C. M. Waterman Storer. 2001. Rapi d dynamics of the microtubule binding of ensconsin in vivo. J Cell Sci. 114:3885 3897. 163. Meyerzon, M., H. N. Fridolfsson, N. Ly, F. J. McNally, and D. A. Starr. 2009. UNC 83 is a nuclear specific cargo adaptor for kinesin 1 mediated nuclear migration. Development. 136:2725 2733. 164. Schneider, M., W. Lu, S. Neumann, A. Brachner, J. Gotzmann, A. A. Noegel, and I. Karakesisoglou. 2010. Molecular mechanisms of centrosome and cytoskeleton anchorage at the nuclear envelope. Cell Mol Life Sci. 68:1593 1610.

PAGE 178

178 165. Ralston, E., Z. Lu, N. Biscocho, E. Soumaka, M. Mavroidis, C. Prats, T. Lomo, Y. Capetanaki, and T. Ploug. 2006. Blood vessels and desmin control the positioning of nuclei in skeletal muscle fibers. J Cell Physiol. 209:874 882. 166. Toivola, D. M., G. Z. Tao, A. Habtezion, J. Liao, and M. B. Omary. 2005. Cellular integrity plus: organelle related and protein targeting functions of intermediate filaments. Trends Cell Biol. 15:608 617. 167. Wilhelmsen, K., S. H. Litjens, I. Kuikman, N. Tshimbalanga, H. Janssen, I. van den Bout, K. Raymond, and A. Sonnenberg. 2005. Nesprin 3, a novel outer nuclear membrane protein, associates with the cytoskeletal linker protein plectin. J Cell Biol. 171:799 810. 168. Fuchs, E., and I. Karakesisoglou. 2001. Bridging c ytoskeletal intersections. Genes Dev. 15:1 14. 169. Prince, J. L., and R. B. Dickinson. 2003. Kinetics and Forces of Adhesion for a Pair of Capsular/Unencapsulated Staphylococcus Mutant Strains. Langmuir. 19 (1):154 159. 170. Lee, J., Y. L. Wang, F. Ren, and T. P. Lele. 2010. Stamp wound assay for studying coupled cell migration and cell debris clearance. Langmuir. 26:16672 16676. 171. Gupton, S. L., W. C. Salmon, and C. M. Waterman Storer. 2002. Converging populations of f actin promote breakage of associated microtubules to spatially regulate microtubule turnover in migrating cells. Curr Biol. 12:1891 1899. 172. Love, A. E. H. 1944. A treatise on the mathematical theory of elasticity. Dover. 173. Bicknese, S., N. Periasamy, S. B. Shohet, and A. S. Verkman. 1993. Cytoplasmic viscosity near the cell plasma membrane: measurement by evanescent field frequency domain microfluorimetry. Biop hys J. 65:1272 1282. 174. Fushimi, K., and A. S. Verkman. 1991. Low viscosity in the aqueous domain of cell cytoplasm measured by picosecond polarization microfluorimetry. J Cell Biol. 112:719 725. 175. Ladd, A. J. C., and G. Misra. 2009. A symplectic in tegration method for elastic filaments. J Chem Phys. 130:124909.

PAGE 179

179 BIOGRAPHICAL SKETCH Jun Wu was born in Chishui City, Guizhou Province, China in 1983 to Daili Wu and Songhui Liu. He graduated from Zunyi Hangtian High School in Zunyi, Guizhou P rovince, China in 2001. In 2005, he received a Bachelor of Science degree in chemical engineering and t echnology from Tianjin University, Tianjin, China. He joined Dr. y and studied computational fluid dynamics of crystallization separation process. He received a Master of Science degree in chemical d ngineering from Tianjin University in 2007. He entered the PhD p rogram in the Chemical Engineering D epartment at the Unive 2007. His research was focused on microtubule and nuclear mechanics in fibroblasts and endothelial cells He earn ed his Doctor of Philosophy in chemical e ngineering from the University of Florida in 2012.